United States Office of Research and EPA/600/R-02/097
Environmental Protection Development December 2002
Agency Washington DC 20460
Fish Physiology,
Toxicology, and Water
Quality
Proceedings of the Sixth
International Symposium,
La Paz B.C.S. Mexico
January 22-26, 2001
-------
EPA/600/R-02/097
December 2002
FISH PHYSIOLOGY, TOXICOLOGY,
AND WATER QUALITY
Proceedings of the Sixth International
Symposium, La Paz, B.C.S. Mexico
January 22-26, 2001
Edited By
Robert V. Thurston
Fisheries Bioassay Laboratory
Montana State University
Bozeman, Montana 59717
Published By
Ecosystems Research Division
Athens, Georgia 30605
National Exposure Research Laboratory
Office of Research and Development
U.S. Environmental Protection Agency
Research Triangle Park, North Carolina 27711
-------
NOTICE
The views expressed in these Proceedings are those of the individual authors and do not
necessarily reflect the views and policies of the U.S. Environmental Protection Agency (EPA).
Scientists in EPA's Office of Research and Development have authored or coauthored papers
presented herein; these papers have been reviewed in accordance with EPA's peer and
administrative review policies and approved for presentation and publication. Mention of trade
names or commercial products does not constitute endorsement or recommendation for use by
EPA.
11
-------
FOREWORD
Joint ecological research involving scientists and environmental managers from every
country in the world is essential if global environmental problems are to be solved. Recognition
of this international aspect of environmental protection is reflected in the joint activities
undertaken under Annex 3, Item 4 of the United States of America-People's Republic of China
Protocol for Environmental Protection. This component of the protocol provides for cooperative
research on the effects of pollution on freshwater organisms, environmental processes, soils,
surface water and groundwater, and on the application of pollutant transport and transformation
models.
Specific areas of cooperation in environmental research include: inorganic chemical
characterization and measurement; inorganic chemical transport and transformation process
characterization; biological degradation process characterization; oxidation/reduction process
characterization; field evaluation of selected transport, exposure and risk models; and application
of models for environmental decision-making concerning organic pollution in semi-arid
conditions, heavy metal pollution, and permissible loading of conventional and toxic pollutants
in rivers. Activities include seminars, workshops, joint symposia, training programs, joint
research, and publications exchange.
This is the sixth international symposium to bring together researchers from the U.S.,
China, and other countries to report on and exchange information in the area offish physiology,
toxicology, and water quality. The sixth symposium was held in La Paz, B.C.S. Mexico,
January 22-26, 2001. Scientists from 15 countries presented 25 papers at the symposium
sponsored by the U.S. Environmental Protection Agency, The American Fisheries Society, The
Canadian Society of Zoologists, The Centre Regional de Investigacion Pesquera, La Paz,
Mexico, and Montana State University. The five previous symposia were held in Guangzhou,
China, September 14-16, 1988; in Sacramento, California, USA, September 18-19, 1990; in
Nanjing, China, November 3-5, 1992; in Bozeman, Montana USA, September 19-21, 1995; and
in Hong Kong, China, November 10-13, 1998.
Symposia are effective means of fostering cooperation among scientists from different
countries as environmental organizations seek to gain the information necessary to predict the
effects of pollutants on ecosystems and apply the results on a global scale. The symposia
provide a forum through which distinguished scientists from laboratories and institutions from
several countries can exchange scientific knowledge on environmental problems of concern to
the U.S. Environmental Protection Agency and the international environmental community.
Rosemarie C. Russo, Ph.D.
Director
Ecosystems Research Division
Athens, Georgia
in
-------
DEDICATION
The Proceedings of this International Symposium is dedicated to the late Robert Vance
Thurston, Research Professor in the Department of Chemistry and Biochemistry and Director of
the Fisheries Bioassay Laboratory at Montana State University. Dr. Thurston was the manager
and organizer for 13 years of this international symposium series, "Fish Physiology, Toxicology,
and Water Quality", held biannually and rotated among the United States, China, and Mexico
with Europe planned for 2003. This symposium series brings together researchers from all over
the world (North and South America, China and other Asian countries, Western and Eastern
Europe, Russia and Ukraine) to present papers and discuss research findings on the effects of
pollutants on organisms, the behavior of pollutants in freshwater and marine systems, and the
modeling and management of aquatic systems. Dr. Thurston was also the Editor of the
Symposium Proceedings, which published the papers presented in the Symposia. The Symposia
series was jointly sponsored over the years by the U.S. Environmental Protection Agency,
Montana State University, the American Fisheries Society, and several professional societies,
research institutes, and universities abroad. Dr. Thurston was involved over many years in
collaborative research with scientists from many countries, and it is fitting that this Proceedings
honors his memory. He will be missed by his many colleagues and friends.
IV
-------
ABSTRACT
Scientists from fifteen countries presented papers at the Sixth International Symposium
on Fish Physiology, Toxicology, and Water Quality held in La Paz, Baja, Mexico,
January 22-26, 2001. These Proceedings include 25 papers presented in sessions convened over
four days. Papers addressed the effects of hypoxia and anoxia on the physiology of fishes and
aquatic invertebrates as a global phenomenon, the role of adenosine as a universal promoter of
fish survival under hypoxia, the effects of hypoxia on fish species, and the specific effects of
hypoxia and anoxia in: temperate estuaries, the continental shelf, the deep sea environment,
shallow eutrophic lakes, and the subtropical environment. Water quality papers included:
general discussions on hypoxia, effects of anoxia on the marine sulfur cycle, effects of
hypoxia/anoxia on major ion and redox chemistry, physical effects of anoxia on sediment biota
morphology, hypoxia in the Gulf of California, effect of hypoxia on the ecological conditions of
coastal estuaries, nonpoint source pollution effects on coastal hypoxia, modeling effects of
climate change on hypoxia, and the use of euthrophication modeling to assess water quality and
ecological endpoints.
-------
CONTENTS
Page
FOREWORD iii
DEDICATION iv
ABSTRACT v
ACKNOWLEDGMENTS viii
PHYSIOLOGY AND TOXICOLOGY
Adenosine - A universal promoter of survival under hypoxic conditions.
G.E. Nilsson 1
Anemia and polycythaemia affect levels of ATP and GTP in fish red blood cells.
A.L. Val, B. Wicks, andD.J. Randall 11
Nitrogen and anaerobic metabolism and hemocyanin levels in the white shrimp
Litopenaeus vannamei exposed to short-term hypoxia. S. Gomez-Jimemez,
M.E. Lugo-Sanchez, AM. Guzmdn-Partida, G. Garcia-Sdnchez, and
R. Barraza-Guardado 21
The effect of aquatic hypoxia on fish.
W.L. Poon, C.Y. Hung, andD.J. Randall 31
Impact of hypoxia on Atlantic cod in the northern Gulf of St. Lawrence.
J-D. DutilandD. Chabot 51
Sensitivity of sturgeons to environmental hypoxia: A review of physiological and
ecological evidence. D.H. Secor andE.J. Niklitschek 61
Hypoxia and anoxia in small temperate estuaries: Patterns of oxygen deficiency, effects,
and recovery. J. Franco, E. Aspillaga, I. Muxika, V. Perez, O. Solaum, and
A. Borja 79
The effect of seasonally severe hypoxia on continental shelf fauna.
N.N. Rabalais 95
Adaptations of demersal fish species in a nutrient-rich embayment of the Ionia Sea (Greece).
V. Vassilopoulou, C. Papaconstantinou, andE. Caragitsou 107
Oxygen minimum zone influence on the community structure of deep-sea benthos.
LA. Levin 121
VI
-------
Response of benthic fauna and changing sediment redox profiles over a hypoxic gradient.
R. Rosenberg, H.C. Nilsson, andRJ. Diaz 133
Fish responses to organic matter loading and to hypoxia in shallow eutrophic lakes.
A. Tuvikene, A. Jarvalt, R. Laugaste, andE. Pihu 147
Hypoxia/anoxia in Lake Vortsjarv, Estonia.
L. Tuvikene, P. Noges, and T. Noges 163
Responses of benthic communities to hypoxia in a sub-tropical environment: Problems
and hypotheses. R. Wu 171
Hypoxia and anoxia as global phenomena.
R.J. Diaz 183
WATER QUALITY
Chemical processes in the anoxia zones of the Baltic Sea.
V-J. Sukyte 203
Metal and major-ion redox chemistry of the hypoxic and anoxic zones: An overview.
G. W. Bailey 219
The different faces of anoxia in the Baltic Sea.
H. Rumohr 273
Hypoxic waters in the Gulf of Mexico: Origin, distribution, and possible consequences.
J. V. Macias-Zamora andF.D-Hinojosa 285
The ecological condition of estuaries: A focus on the Atlantic Ocean and Gulf of Mexico
Coasts of the United States. W.H. Benson andJ.K. Summers 297
Human influences on coastal hypoxia: Examples from Chesapeake Bay Watershed.
D.L. Correll 311
Implications of global climate changes for coastal and estuarine hypoxia: Hypotheses,
observations, and models for the northern Gulf of Mexico. D. Justic,
N.N. Rabalais,andR.E. Turner 323
Eutrophication modeling capabilities for water quality and integration toward
ecological endpoints. R.F. Carousel andR. C. Russo 345
vn
-------
ACKNOWLEDGMENTS
Organizing and presenting a symposium and preparing the proceedings is frequently a
complex task, particularly when participants represent organizations in several countries. For
this reason, the cooperative work of those involved from the sponsoring agencies - the U.S.
Environmental Protection, the American Fisheries Society, the Canadian Society of Zoologists,
The Centre Regional de Investigacion Pesquera, and Montana State University - is greatly
appreciated. The hard work of members of the organizing committee, Nancy N. Rabalais,
Universities Marine Consortium, USA; David J. Randall, University of British Columbia,
Canada; and Vinicio Macias-Zamora, University of Baja California North, Mexico, is gratefully
acknowledged. The scientists, engineers, and environmental managers who prepared the papers
and participated in the symposium are, of course, deserving of primary recognition. In
particular, recognition is accorded to the session chairpersons, who assured efficient functioning
of the symposium: Drs. Rabalais, Randall, Robert J. Diaz, Virginia Institute of Marine Science,
USA, and Donald F. Boesch, University of Maryland, USA. Able assistance in preparing this
was provided by Sheila Walker, who also prepared the final draft. The invaluable contribution
made by Dr. Robert R. Swank, Jr. in reviewing the manuscripts is gratefully acknowledged.
Robert V. Thurston
Chairman of the Symposium
Vlll
-------
IX
-------
ADENOSINE - A UNIVERSAL PROMOTER OF SURVIVAL
UNDER HYPOXIC CONDITIONS
Goran E. Nilsson1
ABSTRACT
Falling cellular ATP levels are life threatening for most animals. In energy deficient
animal tissues, adenosine can be formed by dephosphorylation of the phosphorylated adenosines,
ATP, ADP and AMP. Thus, during hypoxia and anoxia, rising adenosine levels have been
detected in several tissues in various animals, from invertebrates to mammals. Adenosine has
nearly all the properties that can be expected of a promoter of hypoxia tolerance. When released
extracellularly, adenosine activates purinergic receptors (Ai and A2 receptors) that both
stimulates ATP production, by mediating increased blood flow and glycogenolysis, and reduces
ATP consumption, by depressing neuronal activity and transmitter release. Thus, rising
adenosine levels will induce changes that help restore ATP levels and energy charge. Adenosine
receptors have been identified in fish tissues, and there is now evidence for a role of adenosine in
hypoxic/anoxic survival in teleosts, elasmobranches, and cyclostomes. In crucian carp
(Carassius camssius), possibly the most anoxia tolerant fish species, adenosine mediates
increased brain flow during anoxia. Moreover, blocking adenosine receptors in this species
results in an increased metabolic rate during anoxia, indicating a role of adenosine in anoxic
metabolic depression. Also in salmonids and a hagfish, adenosine appears to function to reduce
metabolic rate during hypoxic conditions.
INTRODUCTION
During periods of reduced oxygen availability, the primary problem for the cell is the
resultant slow down or cessation of oxidative phosphorylation, leaving the cell with glycolysis as
the only option for producing ATP. Unfortunately, the ATP yield of glycolysis is less than a
tenth of that of the complete oxidation of glucose. The result of hypoxia is, therefore, often a
drop in the cellular ATP levels. The energetic crisis that hypoxia brings to cells, tissues, and
organisms often results in elevated levels of adenosine — an event closely linked to the falling
ATP concentration. As we shall see, the rise in adenosine is also likely to be a key life-saving
event during hypoxia.
ADENOSINE METABOLISM
Adenosine is produced during the enzymatic breakdown of the high energy purines ATP,
ADP, and AMP. It is formed from AMP, either directly (by 5'-nucleotidase) or indirectly via
IMP and inosine (Figure 1). The most important source of AMP is probably intracellular ATP.
Division of General Physiology, Department of Biology, University of Oslo, P.O.Box 1051, N-0316 Oslo
Norway.
1
-------
However, it may also be derived from extracellular ATP released as a neurotransmitter or a co-
transmitter (since ATP is sometimes present in the synaptic vesicles of other neurotransmitters).
Intracellularly, an important enzymatic step acting to decrease the adenosine level appears to be
the ATP dependent adenosine kinase reaction, which forms AMP from adenosine. During
energy deficiency, when ATP consumption exceeds ATP production, there is a net increase in
the AMP available for adenosine formation. At the same time, the ATP level falls, which will
slow down the adenosine kinase reaction. Consequently, in an energetically compromised cell,
adenosine levels increase substantially, both intracellularly and extracellularly (Kaufman 1985
Hagberg et al. 1987). In the mammalian brain, for example, anoxia, ischemia and hypoglycemia,
as well as excessive neuronal activity, leads to substantially increased levels of adenosine
(Fredholm 1995).
Inhibited
transmitter
release
Al receptor
I n o s i n e -*— 4 aenaswe
Aaenostne
A2 receptor
\
Vasodilation
Figure 1. Adenosine in brain: Adenosine is derived from intracellular ATP break-down caused
by anoxic/ischemic energy deficiency or neuronal over-excitation. It may also be
produced from ATP released as a neurotransmitter (or co-transmitter). Extracellularly,
adenosine activates AI receptors, causing a decrease in neurotransmitter release and
neuronal excitability. When binding to A2 receptors, adenosine causes vasodilation and,
thereby, a local increase in cerebral blood flow.
-------
Adenosine is most often referred to as a neuromodulator rather than a neurotransmitter.
Although extracellular adenosine has powerful effects on neurotransmission, it does not appear
to be stored in vesicles, like the classical neurotransmitters. Since adenosine lacks the charged
phosphate groups, it is more hydrophobic than ATP, ADP or AMP, so it may diffuse out of the
cells through the lipid cell membrane. However, most of the adenosine release appears to
involve nucleotide transporters, which are directional (driven by the Na+ gradient) as well as
equilibrative (driven by the adenosine gradient) (For reviews see Rudolphi et al. 1992, Fredholm
1995).
Although most of our knowledge of how adenosine acts as a neuromodulator, particularly
on the cellular and intracellular levels, comes from mammalian studies; adenosine receptors are
widespread among vertebrates. They have, for example, been shown to occur in both fish and
reptilian brains (Rosati et al. 1995, Lutz and Manuel 1999).
Adenosine receptors have long been divided into two major types, the AI and A2
receptors (Stiles 1991), but more recent evidence has clearly established the presence of an AS
receptor as well as two sub-types of the A2 receptor (A2a and A2t>). Although cloning studies
have revealed striking similarities in the amino acid sequences of the mammalian adenosine
receptors, they can be distinguished by their differential interaction with a variety of adenosine
analogues (Fredholm 1995).
The AI and A2 receptor types are blocked by methylxanthines such as caffeine,
theophylline or aminophylline (a dimer of theophylline that is rapidly cleaved in tissues).
Several derivatives of these xanthines have proven to be specific antagonists for the different
adenosine receptor sub-types, while derivatives of adenosine are used as receptor sub-type
specific antagonists.
The best known effects of adenosine are neuronal inhibition and vasodilation. The AI
adenosine receptors appear to be primarily responsible for the inhibitory effects that adenosine
has on neuronal excitability and neurotransmission, while the vasodilatory effects of adenosine
operates via the A2 receptors. In most tissues, activation of AI receptors reduces adenylate
cyclase activity, and, thus, cAMP production, whereas A2 receptors stimulate adenylate cyclase.
In both cases, the receptors are coupled to adenylate cyclase via a G protein. AI receptor
activation has also been found to open K+ channels, and probably also regulates Ca2+ channels.
Clearly, the hyperpolarising effect of opening K+ channels must underlie some of the inhibitory
effects that adenosine has on nerve cell activity.
As in mammals, adenosine has been shown to inhibit neurotransmitter release in fish
brain (Oshima 1989), acting through AI receptors (Rosati et al. 1995). The latter study also
indicated the presence of A2b receptors in the brains of goldfish (Carassius auratus).
In contrast to the well known vasodilatory actions of adenosine in mammals (acting
through A2 receptors), the most striking effect of injecting adenosine into rainbow trout
(Oncorhynchus mykiss) is a profound vasoconstriction of gill arteries, evidently mediated by an
activation of AI receptors (Sundin and Nilsson 1996). The reason why adenosine is such a
potent vasoconstrictor offish gill vasculature remains unclear. However, the same study showed
3
-------
that adenosine simultaneously cause a decreased systemic vascular resistance. In fact, if
adenosine is superfused onto the brain of crucian carp (Nilsson et al. 1994) or freshwater turtles
(Trachemys scriptd) (Hylland et al. 1994), a substantial increase in brain blood flow occurs,
indicating cerebral vasodilation.
ADENOSINE AS A MEDIATOR OF HYPOXIC SURVIVAL
In the medical literature, adenosine has been termed a "retaliatory metabolite" for several
tissues, including the heart and brain, since it is produced during conditions of energy
insufficiency and subsequently acts to reduce energy consumption and increase energy supply
(Newby et al. 1990). Taking the actions of adenosine into account, this molecule appears
perfectly suited as a promoter of cellular survival in situations likely to cause energy deficiency,
such as hypoxia or anoxia. For many aquatic vertebrates, such situations are probably much
more common than for mammals.
Adenosine may, in fact, perform a key role in anoxic survival in those few vertebrates
known to tolerate prolonged anoxia: the North American freshwater turtles of the genera
Trachemys and Chrysemys, and the crucian carp. While these animals are noted for being able to
maintain their brain ATP levels for days of anoxia, limited but significant temporary drops in
levels of ATP and ADP are seen in the oxygen deprived freshwater turtle brain early, during the
transition period of hypoxia to anoxia (Lutz et al. 1984, Kelly and Storey 1988). Nilsson and
Lutz (1992) suggested that this drop in ATP occurs in the turtle when the production of ATP
from both aerobic (hypoxic) and anaerobic glycolysis is insufficient to meet the normal demand
for ATP. The drop in ATP and other phosphorylated adenosines was postulated to result in a
concomitant rise in adenosine. In fact, a 10-fold increase in the extracellular level of adenosine
was shown to occur in the turtle brain during the initial 1-3 hours of anoxia (Figure 2) (Nilsson
and Lutz 1992). In turtles, there is now evidence that adenosine acts as a signal of energy
insufficiency, and is directly involved in initiating the changes that result in stimulated
glycolysis as well as in a drastic reduction of the metabolic rate (Perez-Pinzon et al. 1993, Pek
and Lutz 1997, Buck and Bickler 1998, Lutz and Manuel 1999).
Adenosine is widely regarded as having an important neuroprotective role during
ischemia and hypoxia in the mammalian brain (Rudolphi et al. 1992). In these situations,
adenosine is released into the extracellular space (Van Wylen et al. 1986) of the mammalian
brain where it acts to produce an increase in cerebral blood flow (Collis 1989), to stimulate
glycogenolysis (Magistretti et al. 1986), and to decrease neuronal excitability as well as to
suppress excitatory neurotransmitter release (Stone 1991, Prince and Stevens 1992). In the rat
hippocampus, it has been shown that adenosine delays the onset of harmful hypoxic
depolarization, acting via the AI receptors (Lee and Lowenkopf 1993).
-------
s
3
4,,
-
1
-100
I 110 200 300
in
410
Figure 2. Anoxia induced increase in extracellular adenosine measured by intracerebral
microdialysis in the freshwater turtle brain (striatum). From Nilsson and Lutz (1992).
Nevertheless, the ability of adenosine to protect the anoxic mammalian brain is clearly
limited, the survival time still being in the order of minutes. By contrast, in the anoxia-tolerant
animal the protective role of adenosine may be the first step in brain metabolic reduction
regulation. Indeed, adenosine may even be involved in regulating the whole body metabolic
depression that is characteristic of species that are anoxia-tolerant. In support of this is an
experiment showing that administration of the adenosine receptor blocker aminophylline causes
the crucian carp to increase its anoxic metabolic rate, measured as the rate of production of
ethanol — the principal end product of anaerobic glycolysis for this species (Figure 3) (Nilsson
1991). The same study showed that the normoxic metabolic rate, measured as Qj. consumption,
is unaffected by aminophylline treatment. Here, it should be mentioned that recent studies
suggest that adenosine also has an important energy saving role in fish liver. Thus, treating
goldfish and rainbow trout (Oncorhynchus mykiss) hepatocytes with adenosine leads to
reductions in both protein synthesis and Na+/K+ pump activity (Krumschnabel et al. 2000).
Adenosine may also promote anoxic survival of anoxia-tolerant vertebrates by
stimulating increased cerebral blood flow, and thereby glucose delivery to the brain. In fact,
both crucian carp and freshwater turtles show an elevated level of brain blood flow during
anoxia that can be inhibited by aminophylline treatment (Hylland et al. 1994, Nilsson et al.
1994). There are
5
-------
experimental indications that such a change in glucose delivery to the brain is of particular
importance. Superfusing the anoxic isolated turtle cerebellum with theophylline (an unspecific
adenosine receptor blocker) or the specific adenosine Ai-receptor blocker 8-
cyclopentyltheophylline (8-CPT) causes rapid depolarization (Perez-Pinzon et al. 1993). In
contrast, superfusing the anoxic in situ brain with the adenosine receptor antagonists
(theophylline, 8-CPT) does not produce depolarization (Pek and Lutz 1997). The reason for this
difference is probably that the in vitro preparation is energy-limited, depending on diffusion of
glucose through the slice to meet its energy needs, while in the in situ brain glycolytic delivery is
greatly enhanced during anoxia through an increased brain blood supply and hyperglycemia.
*
o
yj
100
in
Figure 3. Evidence for a role of adenosine in anoxic metabolic depression in crucian carp.
Treating crucian carp with the adenosine receptor blocker aminophylline causes an
elevation of the rate of ethanol release to the water. Ethanol is the main metabolic end
product of anaerobic ATP production in this species, and it is released to the water
through the gills. (Thus, this species does not have to suffer from increasing lactate levels
during anoxia.) From Nilsson, 1991).
ADENOSINE - AN ANCIENT SIGNAL FOR HYPOXIA
There may be a very ancient history behind the role of adenosine as a metabolic
depressor during periods of hypoxia in vertebrates. There is evidence of such an effect in the
Pacific hagfish (Eptatretus stouti), a representative of the most primitive group of vertebrates,
the cyclostomes. Bernier et al. (1996a) found that injecting the adenosine receptor blocker,
-------
theophylline, into hypoxic rainbow trout and hypoxic Pacific hagfish resulted in significantly
higher concentrations of plasma lactate and a more pronounced acidosis in both species,
indicating an enhanced rate of glycolysis. In the rainbow trout, there was also an increased rate
of glycogen and creatine phosphate depletion in the heart. At the same time, theophylline was
without effect on normoxic control fish of both species. A parallel study using theophylline
suggested a feed-back role for adenosine in depressing catecholamine release in response to
anoxia in both the rainbow trout and the Pacific hagfish (Bernier et al. 1996b).
The function of adenosine as an endogenous metabolic depressant may be even older
than the vertebrates. In the marine worm Sipunculus nudus, an anoxia-tolerant invertebrate,
adenosine levels have been found to increase in the nervous system following exposure to either
anoxia or hypercapnia - two conditions known to induce metabolic depression in this species
(Figure 4) (Reipschlager et al. 1997). The highest adenosine levels were found in worms
exposed simultaneously to anoxia and hypercapnia. Moreover, injections of adenosine caused a
marked depression of the oxygen consumption of control specimens. The experiments indicated
that adenosine was acting through the nervous system of the worm rather than directly on the
muscular body wall. Finally, theophylline was found to have a significant ability to block the
depressed oxygen consumption of hypercapnic worms.
6-
A-
2-
0-
Sipij/tcijti
T
Control
Hy
ills
Rtt
percapni
tfl/£
a
T
Anoxia
Hi
/percapn
+ Anoxia
ia
Figure 4. Adenosine levels increase in the nervous system of the marine worm Sipunculus nudus
when it is exposed to hypercapnia and anoxia (alone or combined). This species tolerates
anoxia well and shows metabolic depression in response to both anoxia and hypercapnia.
Data from Reipschlager et al. (1997).
-------
REFERENCES
Bernier, N.J., J. Harris, J. Lessard, and DJ. Randall. 1996a. Adenosine receptor blockage and
hypoxia-tolerance in rainbow trout and Pacific hagfish. I. Effects on anaerobic
metabolism. Journal of Experimental Biology 199:485-495.
Bernier, N.J., J. Fuentes, and DJ. Randall. 1996b. Adenosine receptor blockage and hypoxia-
tolerance in rainbow trout and Pacific hagfish. II. Effects on plasma cathecholamines
and erythrocytes. Journal of Experimental Biology 199:497-507.
Buck, L.T., and P.E. Bickler. 1998. Adenosine and anoxia reduce N-methyl-D-aspartate
receptor open probability in turtle cerebrocortex. Journal of Experimental Biology
201:289-297.
Collis, M.G. 1989. The vasodilator role of adenosine. Pharmacology Therapeutics 41:143-162.
Fredholm, B.B. 1995. Modulation of neurotransmitter release by heteroreceptors. In:
Neurotransmitter Release and its Modulation, S. J. Bunn and D. A. Powis (eds.),
Cambridge University Press, United Kingdom, pp. 104-121.
Hagberg, H., P. Andersson, J. Lacarewicz, I. Jacobson, S. Butcher, and M. Sandberg. 1987.
Extracellular adenosine, inosine, hypoxanthine, and xanthine in relation to tissue
nucleotides and purines in rat striatum during transient ischemia. Journal of
Neurochemistry 49:227-231.
Hylland, P., G.E. Nilsson, and P.L. Lutz. 1994. Time course of anoxia induced increase in
cerebral blood flow rate in turtles: evidence for a role of adenosine. Journal of Cerebral
Blood Flow and Metabolism 14:877-881.
Kaufman, S. 1985. Regulatory properties of phenylalanine, tyrosine and tryptophan
hydroxylases. Biochemical Society Transactions 13:433-436.
Kelly, D.A., and K.B. Storey. 1988. Organ-specific control of glycolysis in anoxic turtles.
American Journal of Physiology 255:R774-R779.
Krumschnabel, G., C. Biasi, and W. Wieser. 2000. Action of adenosine on energetics, protein
synthesis and K(+) homeostasis in teleost hepatocytes. Journal of Experimental Biology
203:2657-2665.
Lee, K.S., and T. Lowenkopf 1993. Endogenous adenosine delays the onset of hypoxic
depolarization in the rat hippocampus in vitro via an action at Aj receptors. Brain
Research 609:313-315.
Lutz, P.L., and L. Manuel. 1999. Maintenance of adenosine AI receptor function during long-
term anoxia in the turtle brain. American Journal of Physiology 276:R633-R636.
-------
Lutz, P.L., P. McMahon, M. Rosenthal, and TJ. Sick. 1984. Relationships between aerobic and
anaerobic energy production in turtle brain in situ. American Journal of Physiology
247:R740-R744.
Magistretti, P.J., P.R. Hof, and J.L. Martin. 1986. Adenosine stimulates glycogenolysis in
mouse cerebral cortex: a possible coupling mechanism between neuronal activity and
energy metabolism. Journal of Neuroscience 6:2553-2562.
Newby, A.C., Y. Worku, and P. Meghji. 1990. Adenosine: a retaliatory metabolite or not?
News in Physiological Sciences 5:67-70.
Nilsson, G.E. 1991. The adenosine receptor blocker aminophylline increases anoxic ethanol
excretion in the crucian carp. American Journal of Physiology 261:R1057-R1060.
Nilsson, G.E., P. Hylland, and C.O. Lofman. 1994. Anoxia and adenosine induce increased
cerebral blood flow in crucian carp. American Journal of Physiology 267: R590-R595.
Nilsson, G.E., and P.L. Lutz. 1992. Adenosine release in the anoxic turtle brain as a mechanism
for anoxic survival. Journal of Experimental Biology 162:345-351.
Oshima, N. 1989. Adenosine inhibits the release of neurotransmitters from melanosome-
aggregating nerves offish. Comparative Biochemistry and Physiology C. Comparative
Pharmacology 93:207-211.
Pek, M., and P.L. Lutz. 1997. Role for adenosine in channel arrest in the anoxic turtle brain.
Journal of Experimental Biology 200:1913-1917.
Perez-Pinzon, M., P.L. Lutz, T. Sick, and M. Rosenthal. 1993. Adenosine, a "retaliatory"
metabolite, promotes anoxia tolerance in turtle brain. Journal of Cerebral Blood Flow
and Metabolism 13:728-732.
Prince, D.A., and C.F. Stevens. 1992. Adenosine decreases neurotransmitter release at central
synapses. Proceedings of the National Academy of Science (USA) 89:8585-8590.
Reipschlager, A., G.E. Nilsson, and H.O. Former. 1997. A role for adenosine in metabolic
depression in the marine invertebrate Sipunculus nudus. American Journal of Physiology
272: R350-R356
Rosati, A.M., U. Traversa, R. Lucchi, and A. Poli. 1995. Biochemical and pharmacological
evidence for the presence of AI but not A2a adenosine receptors in the brain of the low
vertebrate teleost Carassius auratus (goldfish). Neurochemistry International 26:411-23.
Rudolphi, K.A., P. Schubert, F.E. Parkinson, and B.B. Fredholm. 1992. Adenosine and brain
ischemia. Cerebrovascular and Brain Metabolism Reviews 4:346-369.
Stiles, G.L. 1991. Adenosine receptors: physiological regulation and biochemical mechanisms.
News in Physiological Sciences 6:161-165.
-------
Stone, T.W. 1991. Receptors for adenosine and adeninine nucleotides. General Pharmacology
22:25-31.
Sundin, L., and G.E. Nilsson. 1996. Branchial and systemic roles of adenosine in rainbow trout,
an in vivo microscopy study. American Journal of Physiology 271: R661-R669.
Van Wylen, D.G.L., T.S. Park, R. Rubio, and R.M. Berne. 1986. Increases in cerebral
interstitial fluid adenosine concentration during hypoxia, potassium infusion, and
ischemia. Journal of Cerebral Blood Flow and Metabolism 6:522-528.
10
-------
ANEMIA AND POLYCYTHANEMIA AFFECT LEVELS OF ATP
AND GTP IN FISH RED BLOOD CELLS
Adalberto L. Val1, Bev Wicks2, and David J. Randall3
ABSTRACT
Exposure to either environmental hypoxia or decreased oxygen carrying capacity of the
blood results in oxygen shortage at the tissue level in fish. A wide range of physiological and
biochemical adjustments have evolved to maintain tissue oxygen supply during such constraints.
These adjustments include the regulation of the levels of adenosine triphosphate (ATP) and
guanosine triphosphate (GTP) in the red blood cells. ATP and GTP are negative modulators of
hemoglobin-oxygen affinity (Hb-O2), i.e., they bind to deoxygenated hemoglobin, decreasing its
oxygen affinity. Compared to ATP, GTP has a greater modulating effect on Hb-O2 affinity. A
decrease in the levels of ATP and GTP in the red blood cells offish experiencing environmental
hypoxia has been reported for the great majority offish species so far studied. Whenever GTP is
present in the red blood cells, its concentration is adjusted faster than that of ATP. This
adjustment has been explained as an adaptation to hypoxia as it secures oxygen uptake at the
gills. In fish experiencing anemia (decreased peripheral erythrocyte count), or
methahemoglobinaemia (increased level of oxidized hemoglobin), situations that result in a
decrease of oxygen carrying capacity, an improvement of oxygen unloading at the tissue level is
needed. These situations have been analysed in rainbow trout (Oncorhynchus mykiss) and in
pirarara (Phractocephalus hemiliopterus). Anemia is caused either by the reduction in
circulating red blood cells or by the conversion of hemoglobin to methahemoglobin. In contrast
to environmental hypoxia, anemia elicits an increase in ATP and GTP levels in the red blood
cells. The kinetics for the adjustments of ATP and GTP in anemic animals is also a rapid
process, but the mechanism controlling the concentration of these compounds in the red blood
cells remains unknown. Finally, the effect of polycythaemia on red cell ATP and GTP levels
will be discussed.
INTRODUCTION
Changes in environmental oxygen level may occur naturally as observed in tropical
swamps and floodplain areas and in temperate water bodies as a consequence of a thermocline,
halocline, or pynocline. The amount of dissolved oxygen in waters of the Amazon depends on
the interactions of biological, chemical, and physical factors affected by the regular and seasonal
river water level oscillation. In addition to these seasonal changes, extreme variations also tend
to occur in a very short period of time. In floodplain lakes, oxygen level often drops from
Manaus, Brazil, 2UBC, Vancouver, Canada, 3City University, Hong Kong.
11
-------
oversaturated levels at noon to zero at night. These mixed patterns of dissolved oxygen
associated with temperatures ranging from 20 up to 40°C impose several challenges for fish
living in these environments. In temperate water bodies, hypoxia also represents a significant
environmental pressure for fish, though due to different processes.
Anthropogenic activities have resulted in hypoxia over large areas of temperate and
tropical, marine- and fresh-water bodies. This type of hypoxia causes severe physiological
problems to fish, because it is regularly associated with the release of several harmful chemicals
to the environment that impair oxygen uptake. These chemicals include: (1) complex mixtures
such as crude oil that reside on the top of the water column reducing the access of air-breathers
(Val and Almeida-Val 1999); (2) nitrite that oxidizes hemoglobin to methahemoglobin, an
anionic form unable to reversibly bind to hemoglobin (Bartlett et al. 1987, Knudsen and Jensen
1996); and (3) PCB and heavy metals that cause anaemia. So, the adjustments fish have
developed during evolution to improve oxygen transfer may be maladaptive under these new
environmental challenges caused by men.
Fish depend on aerobic metabolism. Oxygen is needed to oxidize food material to
produce energy that is vital in all organismal activity. Reduced oxygen supply to tissues causes a
myriad of problems including reduced growth, abnormal protein synthesis, migration
impairment, low fertility, and mortality. Fish respond to changes in environmental and blood
oxygen content by adjusting biochemical, physiological, and behavioral processes to minimize
disturbances in oxygen transfer to their tissues. The adjustments are designed to facilitate
oxygen uptake at the gas exchange surfaces and oxygen unloading to the tissues under both long-
and short-term hypoxia exposure. Among these adjustments are variations in the functional
properties of hemoglobin that result partly from its molecular structure and partly from changes
caused by pH and heterotropic factors like erythrocytic phosphates that bind to
deoxyhemoglobin, decreasing its affinity to oxygen (Weber et al. 2000).
ATP and GTP are the two most important allosteric factors influencing HB-O2 affinity
found in fish erythrocytes ((Nikinmaa 1990, Val 1993, Weber 1996,Val 2000). In addition to
ATP and GTP, other organic phosphates have been detected in fish, such as 2,3
diphosphoglycerate (2,3DPG) and inositol pentaphosphate (TPP) (Isaacks etal. 1977, Bartlett
1978, Val 1993). In all fish species studied so far, these phosphates act as negative modulators
of HB-O2 affinity. The magnitude of their effect decreases in the following order: IPP > GTP >
ATP > 2,3DPG.
The decrease in the concentration of erythrocytic phosphates in fish experiencing hypoxia
has been explained as a physiological adjustment to maintain tissue oxygenation (Wood and
Johansen 1972, Greaney and Powers 1978, Weber 1979, Val etal. 1995, Val 2000). Whenever
GTP is present, its concentration decreases faster than that of ATP. However, the mechanisms
controlling the erythrocytic concentration of these organic phosphates are unknown. The present
paper analyses the effects of two different types of anemia and the effects of polycythanemia on
the levels of ATP and GTP in species of tropical (Phractocephalus hemiliopterus and Colossoma
macropomum) and temperate fish (Oncorhynchus mykiss).
12
-------
EFFECTS OF ANEMIA ON ERYTHROCYTIC PHOSPHATES
Anemia is a manifestation of some underlying disorder and can be divided into three
broad groups on the basis of the fundamental pathogenic mechanisms resulting in the lowered
erythroid values: (1) blood loss, (2) impaired erythropoiesis, and (3) red blood cell and/or
hemoglobin destruction. In fish, anemia is observed in animals exposed to stressful conditions
that result in acute hemorrhaging, bacterial and viral infection, external parasitism, red cell
destruction, iron deficiency, reduced activity of certain hematopoietic factors, and hemoglobin
oxidation. A reduction in the oxygen-carrying capacity of the blood caused by anemia produces
tissue hypoxia that elicits a number of corrective mechanisms to compensate for the interference
with oxygen transport (Val 1995, Brauner and Wang 1997). These corrective mechanisms are
not mediated by adrenaline and noradrenaline since no major changes of these hormones have
been observed in experimental anemic fish (Iwama et al. 1987).
In contrast to environmental hypoxia, anemic fish must improve oxygen unloading to the
tissues rather than securing increased oxygen uptake at the gills. Thus, it is likely that an
increase in the concentration of erythrocytic phosphate, namely ATP and GTP, occurs in anemic
fish. In fact, an experimental reduction of 15 and 25% of blood volume in rainbow trout resulted
in a significant increase in nucleotide triphosphate (NTP) in the red blood cells as reported by
(Lane 1984, Vorger and Ristori 1985). These authors, however, did not produce normovolemic
anemia, and may have imposed some extra energy-consuming processes to maintain blood
volume. In 1994, Val and co-workers (Val et al. 1994) analysed this question in rainbow trout
producing a progressive normovolemic anemia, and reported similar results, i.e., the lower the
oxygen carrying capacity as a consequence of decreased erythroid values, the higher the
concentration of ATP and GTP inside the red blood cells.
Recently, we have returned to this point to analyse the effect of anemia on the red blood
cell levels of erythrocytic phosphates of a tropical fish species. Tropical fish survive higher
environmental temperatures compared to the temperate rainbow trout and this imposes a
challenge in maintaining tissue oxygen supply that would be exacerbated by the effects of
anemia. Thus, we chose to impose experimental anemia on the giant, tropical fresh water fish
pirarara, Phractocephalus hemiliopterus. Pirarara is a carnivore that lives near the bottom of
rivers of the Amazon where there are periodic episodes of hypoxia. We analysed juvenile
pirarara weighing between 3 and 4 Kg (the animal grows up to 200 Kg). The experimental
animals had 25% of their blood volume, estimated as 5% of the body weight, replaced by saline.
Twenty-four hours later they were analysed for ATP and GTP levels in their red blood cells.
As expected, the levels of NTP (ATP plus GTP) increased by approximately 30%, though
changes of GTP levels were higher than those of ATP (Figure 1). This situation did not differ
from that previously reported for rainbow trout, except for the higher levels of GTP in pirarara.
Whenever GTP is present in the red blood cells, its concentration decreases faster than that of
ATP in animals exposed to hypoxia (Jensen and Weber 1985a, Val 1993). The present data from
pirarara and rainbow trout suggest that the regulation of GTP levels is also faster than the
regulation of ATP levels in anemic fish. Because GTP is more effective in reducing the Hb-C>2
affinity than ATP, this faster adjustment improves tissue oxygenation.
13
-------
Control
Anaemia
NTP
ATP
6TP
Figure 1. Effect of 25% blood removal on the red blood cell erythrocytic levels of ATP and GTP
of pirarara (Phractocephalus hemiliopterus). The blood volume removed was replaced
by saline to achieve normovolemic anaemia. ATP and GTP red blood cell levels were
estimated 24 hours later using HPLC as described by Schulte et al. 1992.
In contrast to anemia caused by reduction of erythroid values, methahemoglobinanemia,
i.e., increased levels of oxidized hemoglobin, results in decreased oxygen-carrying capacity
without changes in the number of circulating red blood cells and total hemoglobin. To analyse
the effect of methahemoglobin levels on erythrocytic phosphates, oxidation of hemoglobin was
induced by intraperiotoneal sodium nitrite injection in rainbow trout. A level of 30% of
methahemoglobin was achieved with a dose of 75 mg kg"1 of sodium nitrite. Compared to the
controls, methahemoglobinanemic animals had a higher proportion of NTP to functional Hb,
though only a small net increase in NTP levels was observed (Figure 2). Jensen and co-workers
(Jensen et al. 1987) attributed the small net increase in NTP levels they observed in carp exposed
to ImM of nitrite to cell shrinkage. These data suggest that the increase in the ratio of NTP to
non-oxidised hemoglobin results in a decrease of FIb-02 affinity that facilitates oxygen unloading
to the tissues in anemic fish. More data is needed, however, before we can have a clear picture
of the equilibrium of oxidised/non oxidised Fib and both ATP and GTP.
14
-------
.0
3.0
2.5
2.0
1.5 -
1.0
0.5
0.0
0
10 15 20 25 30
MetaHb (%)
Figure 2. Effect of methahemoglobin on the red blood cell erythrocytic levels of NTP
(ATP+GTP) in rainbow trout (Oncorhynchus mykiss). NTP levels were corrected to
functional hemoglobin. Increases in methahemoglobin were induced by intraperitoneal
injection of sodium nitrite. See text for details.
EFFECTS OF POLYCYTHANEMIA ON ERYTHROCYTIC PHOSPHATES
Polycythanemia or erythrocytosis is an increase in the peripheral erythroid values
(erythrocyte count, hemoglobin level, and hematocrit) above normal. Polycythanemia occurs
when the circulating red blood cells increase due to abnormally lowered fluid intake, marked loss
of body fluid, and defective oxygen saturation of arterial blood. In fish, increased red blood cell
count is often observed both in tropical and temperate fish experiencing stressful environmental
conditions, and has been related to adrenergic stimulation of the spleen (Wells and Weber 1990,
Moura 1994). This increase in the number of circulating red blood cells is a physiological
adjustment to maintain oxygen transfer to tissues. However, there is a limitation in increasing
red blood cell counts related to the extra work needed to pump a more viscous blood. In
Amazonian tropical fish, hematocrit varies from 14 up to 60% in facultative air-breathers
exposed to hypoxia. According to Wells and Weber (1991), the oxygen transport capacity
15
-------
decreases as hematocrit values diverge from the optimum, estimated as 30%, in rainbow trout.
Therefore, similar to anemia, polycythanemia may compromise oxygen transfer to tissue,
eliciting corrective mechanisms.
To analyse the effect of polycythanemia on red blood cell erythrocytic phosphate levels,
two groups of rainbow trout (donor and experimental) had an indwelling cannulae implanted in
the dorsal aorta and were allowed to recover for 48 hours. Blood doping was planned to result in
a 4% step increase in hematocrit under a normovolemic conditions. The animals were allowed to
recover for 12 hours in a darkened Perspex chamber after blood doping and then blood samples
were collected for analysis of ATP and GTP. The blood samples were immediately processed.
The hematocrit of experimental animals was increased up to 55%. Above this point, the
animals either showed significant bleeding through the skin, eyes and fins or did not survive the
experimental procedure, suggesting an impairment of oxygen transfer. From the control values
up to the highest achieved hematocrit, a continuous increase in NTP (ATP+GTP) levels was
observed (Figure 3), though the differences from the control were not significant for the first step
increases in hematocrit. These data suggest that the increase in circulating red blood cells would
be effective in improving oxygen transfer up to the point that it does not compromise oxygen
transfer. Above this limit, polycythanemia elicits a series of compensatory mechanisms that
includes the adjustment of red blood cell erythrocytic ATP and GTP levels to compensate for the
eventual decrease in oxygen transfer. Increased levels of ATP and GTP reduce Hb-O2 affinity,
facilitating oxygen unloading to the tissues even at relatively high partial pressure of oxygen.
o
a
CL
2.4
2.0 ]
b3 1-6
V) ~~c.
s 1 ? \
a 0 1.^1
§ ^ 0.8
o "i 0.4
0.0 ^
O OTC
ATP + GTP
10
9
8
7
6
5
o
_Q
0 10 20 30 40 50 60 70
Hematocrit (%)
Figure 3. Effect of polythanemia on red blood cell erythrocytic levels of NTP (ATP+GTP) in
rainbow trout (Oncorhynchus mykiss). Data for oxygen transport capacity are from Wells
and Weber 1991.
16
-------
The observed increase in NTP levels in polycythanemic fish were, in fact, much higher.
The ATP content was drastically reduced in the young compared to mature red blood cells of
rainbow trout (Lane 1984). This suggests that the initial lower levels of ATP in the young red
blood cells released into the circulation were compensated by a higher rate of ATP synthesis that
could not be observed in our experiments because we used mature red blood cells for doping.
TIME REQUIRED IN ADJUSTING ERYTHROCYTIC NTP LEVELS
In fish exposed to hypoxia, adjustments of the concentration of ATP and GTP are fast
enough to compensate for short-term changes in dissolved oxygen. They occur within minutes
in some tropical fish species (Val 2000), within one hour in rainbow trout exposed to deep
hypoxia (Tetens and Lykkeboe 1985), or by 24 hours in tench exposed to hypoxia-hypercapnia
(Jensen and Weber 1985b). A new level of NTP was observed within 12 hours for
polycythanemic and 24 hours for anemic fish, but we do not know the exact time course of the
response in fish facing these disturbances.
SUMMARY
Together, results of this study indicate that the regulation of red blood cell erythrocytic
levels of ATP and GTP in fish is influenced by multiple control mechanisms, i.e., there are at
least two signalling processes: one external, governed by the levels of water dissolved oxygen;
and one internal, governed by the rate of oxygen transfer to tissues. Despite the fact that both
environmental hypoxia and impairment of oxygen transfer generate tissue hypoxia, these two
conditions produce opposite effects on red blood cell erythrocytic levels of ATP and GTP.
Environmental hypoxia results in a decrease of ATP and GTP; anemia and polycythanemia result
in an increase in the concentration of these allosteric effectors of Hb-C>2 affinity.
ACKNOWLEDGEMENTS
The present paper is based upon work supported by INPA, CNPq, SUDAM and NSERC.
ALV is recipient of a Research Fellowship from The Brazilian National Research Council
(CNPq).
17
-------
REFERENCES
Bartlett, G.R. 1978. Phosphates in red cells of two South American osteoglossids: Arapaima
gigas and Osteoglossum bicirrhosum. Canadian Journal of Zoology 56: 878-881.
Bartlett, G.R., A.R. Schwantes, and A.L. Val. 1987. Studies on the influence of nitrite on
methahemoglobin formation in amazonian fishes. Comparative Biochemistry and
Physiology 86: 449-456.
Brauner, C., and T. Wang. 1997. The optimal oxygen equilibria curve: A comparison between
environmental hypoxia and anemia. American Zoologist 37: 101-108.
Greaney, G.S., and D.A. Powers. 1978. Allosteric modifiers offish hemoglobins: in vitro and in
vivo studies of the effect of ambient oxygen and pH on erythrocyte ATP concentrations.
Journal of Experimental Zoology 203: 339-350.
Isaacks, R.E., H.D. Kim, G.R. Bartlett, and D.R. Harkness. 1977. Inositol pentaphosphate in
erythrocytes of a freshwater fish, pirarucu (Arapaima gigas). Life Sciences 20: 987-990.
Iwama, G.K., R.G. Boutilier, T.A. Heming, D.J. Randall, and M. Mazeaud. 1987. The effects of
altering gill water flow on gas transfer in rainbow trout. Canadian Journal of Zoology 65:
2466-2470.
Jensen, F.B., N.A. Andersen, and N. Heisler. 1987. Effects nitrite exposure on blood respiratory
properties acid-base and electrolyte regulation in the carp (Cyprinus carpio). Journal of
Comparative Physiology B 157: 533-541.
Jensen, F.B., and R.E. Weber. 1985a. Kinetics of the acclimational responses of tench to
combined hypoxia and hypercapnia. Journal of Comparative Physiology B 156: 205-211.
Jensen, F.B., and R.E. Weber. 1985b. Kinetics of the acclimational responses of tench to
combined hypoxia and hypercapnia. I. Respiratory responses. Journal of Comparative
Physiology B 156: 197-203.
Knudsen, P.K., and F.B. Jensen. 1996. Recovery from nitrite-induced methahemoglobinanemia
and potassium balance disturbances in carp. Fish Physiology and Biochemistry 16: 1-10.
Lane, H.C. 1984. Nucleoside triphosphate changes during the peripheral life-span of erythrocytes
of adult rainbow trout (Salmo gairdneri). The Journal of Experimental Zoology 231: 57-
62.
Moura, M.A.F. 1994. Efeito da anemia, do exercicio fisico e da adrenalina sobre o ba9o e
eritrocitos de Colossoma macropomum (Pisces). MSc thesis. PPG INPA/FUA. 84pp
18
-------
Nikinmaa, M. 1990. Vertebrate Red Blood Cells. Adaptations of function to respiratory
requirements. Springer-Verlag, Heidelberg.
Schulte, P.M., C.D. Moyes, and P.W. Hochachka. 1992. Integrating metabolic pathways in post
exercise recovery of white muscle. Journal of Experimental Biology 166: 181-195.
Tetens, V., and G. Lykkeboe. 1985. Acute exposure of rainbow trout to mild and deep hypoxia:
O2 affinity and O2 capacitance of arterial blood. Respiration Physiology 61: 221-235.
Val, A.L. 1993. Adaptations of fishes to extreme conditions in fresh waters. In: The vertebrate
Gas Transport Cascade. Adaptations to Environment and Mode of Life. J.E.P.W. Bicudo,
(ed.) CRC Press, Boca Raton, pp. 43-53.
Val, A.L. 1995. Oxygen transfer in fish: morphological and molecular adjustments. Brazilian
Journal of Medical and Biological Research 28: 1119-1127.
Val, A.L. 2000. Organic phosphates in the red blood cells offish. Comparative Biochemistry and
Physiology 125: 417-435.
Val, A.L., and V.M.F. Almeida-Val. 1999. Effects of crude oil on respiratory aspects of some
fish species of the Amazon. In: Biology of Tropical Fish. A.L. Val and V.M.F. Almeida-
Val (eds.), INPA, Manaus. pp. 277-291.
Val, A.L., J. Lessard, and D. Randall. 1995. Effects of hypoxia on rainbow trout (Oncorhynchus
mykiss): intraerythrocytic phosphates. The Journal of Experimental Biology 198: 305-
310.
Val, A.L., C.F. Mazur, R.H. Salvo-Souza, and G. Iwama. 1994. Effects of experimental anemia
on intra-erythrocytic phosphate levels in rainbow trout, Oncorhynchus mykiss. Journal of
Fish Biology 45: 269-279.
Vorger, P., and M.T. Ristori. 1985. Effects of experimental anemia on the ATP content and the
oxygen affinity of the blood in the rainbow trout (Salmo gairdneri). Comparative
Biochemistry and Physiology 82: 221-224.
Weber, R., A. Fago, A.L. Val, A. Bang, M.L. Van Hauwaeert, S. De Wilde, F. Zal, and L.
Moens. 2000. Isohemoglobin differentiation in the biomodal-breathing Amazon catfish
Hoplosternum littorale. The Journal of Biological Chemistry 275: 17297-17305.
Weber, R.E. 1979. Erythrocytic nucleoside triphosphates inhibit oxygen linked binding of carbon
dioxide by carp hemoglobin. Acta Physiologica Scandinavica 105: 16A-17A.
Weber, R.E. 1996. Hemoglobin adaptations in Amazonian and temperate fish with special
reference to hypoxia, allosteric effectors and functional heterogeneity. In: Physiology and
Biochemistry of the Fishes of the Amazon. A.L. Val, V.M.F. Almeida-Val, and D.J
Randall (eds.), INPA, Manaus, pp. 75-90.
19
-------
Wells, R.M., and R.E. Weber. 1991. Is there an optimal hematocrit for rainbow trout,
Oncorhynchus mykiss (Walbaum)? An interpretation of recent data based on blood
viscosity measurements. Journal of Fish Biology 38: 53-65.
Wells, R.M.G., and R.E. Weber. 1990. The spleen in hypoxic and exercised rainbow trout. The
Journal of Experimental Biology 150: 461-466.
Wood, S.C., and K. Johansen. 1972. Adaptation to hypoxia by increased HbO2 affinity and
decreased red cell ATP concentration. Nature 237: 278-279.
20
-------
NITROGEN AND ANAEROBIC METABOLISM
AND HEMOCYANIN LEVELS IN THE WHITE SHRIMP
Litopenaeus vannamei EXPOSED TO SHORT-TERM HYPOXIA
S. Gomez-Jimenez1, M.ELugo-Sanchez1, A.M. Guzman-Partida1, G. Garcia-Sanchez1
and R. Barraza-Guardado2
ABSTRACT
In summer, high temperatures are combined with severe nocturnal hypoxia in shrimp
farming ponds in Northwest Mexico. Under these conditions the white shrimp Litopenaeus
vannamei, have shown some mortality when the duration of hypoxia times are longer than
3-4 h. Hemolymph ammonia, total protein, hemocyanin, lactate levels, and ammonia fluxes were
measured in the white shrimp, Litopeneaus vannamei challenged to short-term severe hypoxia
(10% oxygen saturation) for 30, 60, 90, 120 or 180 min at 28°C. The principal objective was to
evaluate changes in the physiology of this species when being exposed to hypoxia over similar
times to those that might encounter during aquaculture farming in Northwest Mexico.
The data obtained indicate that the mean ammonia efflux rate of this species under
control conditions (normoxia) at 28°C was 8.00 ± 0.88 jimol g"1 h"1; this value being reduced to
3.81 ± 0.19 jimol g"1 h"1 after 180 min exposure to hypoxia. Mean normoxic blood hemocyanin
levels were 1.04 ± 0.11 mmol/L with the lowest and highest value of 0.82 ± 0.06 mmol/L and
1.42 ± 0.08 mmol/L being measured after 180 m and 120 m of hypoxia, respectively.
Hemolymph lactate level during normoxic conditions was 3.44 mg/dL, increasing to 45.39
mg/dL after a 180 minute exposure to hypoxia. Large variability in data within individual time
periods was observed. Mortalities for the various time periods of hypoxia were: 0% (normoxia),
0% (30 minutes), 13% (60 minutes, 90 minutes, and 120 minutes) and 26% (180 minutes),
respectively. Results are discussed in relation to the physiological ecology of the unique
aquiculture farming conditions of L. vannamei in Northwest Mexico.
Centra de Investigation en Alimentation y Desarrollo, A.C. (CIAD, AC) Apdo. Postal 1735, Hermosillo, Sonora
2-
83000, Mexico. Departamento de Investigaciones Cientificas y Tecnologicas de la Universidad de Sonora
(DICTUS). Apdo. Postal 1819, Hermosillo, Sonora 83000, Mexico.
21
-------
INTRODUCTION
The white shrimp, Litopenaeus vanamei, is a native species of the Pacific west coast, and
is found from Sonora State, Mexico in the Gulf of California to Peru in South America. This
species has been farmed in the northwest Mexican region since 1973, and has become a key
cultured species since shrimp aquaculture has evolved into an important industry in this region.
However, low dissolved oxygen concentrations, particularly in early morning hours, have been a
problem associated with high primary productivity in the shrimp ponds. Furthermore, it is well
known that oxygen solubility depends on water temperature (Withers 1992). The extreme
climate that characterizes this semiarid region causes high water temperatures to be reached
especially during the summer, further lowering the dissolved oxygen in the ponds. The extent of
shrimp tolerance to periods of severe hypoxia had only been assessed by empirical observation
of the farmers who noticed that periods longer than 4 h produced mortality. Exposure to severe
hypoxia may affect several metabolic and physiological processes and, under such conditions,
compensatory mechanisms to meet energy demand may be required. Aquatic invertebrates,
especially burrowing species, have shown some capacity to maintain their aerobic respiration
under moderate hypoxia (Butler et al. 1978, Zou et al. 1993). However, when the animals are
further stressed and low oxygen levels are maintained and/or prolonged, the majority of
crustaceans shift to anaerobic metabolism and lactate is normally the only end-product of
anaerobic glycolysis (Teal and Carey 1967, Gade 1983).
Ammoniotelism is a characteristic of aquatic crustaceans. Ammonia dominates amongst
the nitrogenous compounds during various catabolic reactions and is easily excreted across the
gills by diffusional movement and/or ionic exchange mechanisms (Kormanik and Cameron
1981) without any further processing under normoxia. However, when the overall metabolic rate
is impaired nitrogen metabolism may also be expected to change, and when environmental
conditions worsen to severe hypoxia, a change in ammonia production and efflux rates may
occur (Hagerman and Szaniawska 1994). Thus, the objective of this study was to assess the
effect of short-term severe hypoxia on nitrogen anaerobic metabolism of L. vannamei under
experimental conditions that might mimic aquaculture farming conditions in Northwest Mexico.
MATERIAL AND METHODS
Adult specimens of white shrimp L. vannamei with an average weight of 13.88 ± 0.44 g
were kindly donated by the University of Sonora-DICTUS from their experimental farming
ponds in Bahia Kino-Mexico. The shrimps were transported to CIAD-Hermosillo using a large
refrigerated container with oxygenated seawater. Upon arrival the animals were transferred into
an aerated and closed fiberglass container. The animals were kept in the system for 10 days prior
to starting the experiments. They were fed ad libitum with commercial pelletized food. Food
was denied 48 h prior to initiating experimental procedures. Seawater was monitored daily for
temperature, salinity, dissolved oxygen, and ammonia levels. Water exchange ratios were
performed when needed to assure good seawater quality throughout the experiment.
22
-------
EXPERIMENTAL CONDITIONS AND SAMPLING
Two hours before the start of the experiment, eight animals from the stock container were
placed in six individual glass aquaria, 5 L of seawater added and each system aerated. Nitrogen
gas (99.99% purity) was bubbled through the system to achieve an oxygen tension of 10%
oxygen saturation within narrow limits (5% of the desired value). Lengths of exposure to
hypoxia were 0, 30, 60, 90, 120, and 180 minutes. Oxygen tensions were monitored
continuously in the aquaria containing the shrimps by a dissolved oxygen electrode (Yellow
Springs Instrument). All the aquaria used were covered with small styrofoam spheres; this
diminished contact with the atmosphere and reduced the need for N2 bubbling. All experiments
were made at S = 38 ppt, T = 28°C. Water samples (1.5 mL) were taken for the measurement of
ammonia efflux rates using 1.7 mL polypropylene microcentrifuge tubes at timed intervals
during the environmental hypoxia; water samples were immediately frozen at -30°C to await
analysis. Hemolymph samples (200 jiL) were taken after 30, 60, 90, 120 and 180 m in the
environmental hypoxia period. Hemolymph were collected at the same time interval via the
pereiopod sinus using disposable 1 mL syringes with a 29 gauge needle. Hemolymph samples
(20 jiL) were immediately diluted (1:40) with Tris buffer (0.05 M) in a 10 mm quartz cuvette
and the absorbance measured at 335 nm against a distilled water blank using a Gary 2 Varian UV
spectrophotometer. The hemocyanin concentrations were estimated using an extinction
coefficient, Elcm, of 17.26 as applied to Carcinus maenas hemocyanin (Hagerman and Weber
1981). One hemolymph sub-sample (20 jiL) was precipitated with an equal addition of cold 6%
perchloric acid for lactate measurements. The remaining hemolymph samples were immediately
frozen at -30°C to await analysis.
All aquariums used in the experiments were first thoroughly cleaned with 10% HC1
solution followed by rinsing with distilled water to minimize bacterial action on the nitrogen
excreted by the animals. A container without animals was also sampled as above to estimate the
effect of microbial activity on ammonia efflux rates. Control animals (8) were always in 100%
oxygen saturation and sampled under the same conditions as the experimental animals. All
experiments were performed inside temperature-controlled rooms.
Analytical Procedure
Total Ammonia: Total ammonia in water and hemolymph was quantified using a flow
injection/gas diffusion (FIGD) technique described by Clinch etal. 1988. Ammonia excretion
rates were calculated from changes in ammonia concentration in the water (|imol N/L) multiplied
by the volume (L) and factored by time (h) and mass (g).
Protein: Hemolymph total protein content was measured according to Bradford 1976 using
bovine serum albumin (BSA) as a standard.
Lactate Concentration: As an index of anaerobic metabolism, the concentration of lactate was
determined in hemolymph. Hemolymph samples were mixed with equal volumes of 6 % cold
perchloric acid (PCA), centrifuged (10,000 g, 5 min), and then neutralized with 2.5 M
23
-------
(20 % sample volume). The supernatant from each sample was then assayed in duplicate for
lactate using colourimetric kits from Sigma Chem. Corp. (cat. No. 735).
Statistics: Data of the various parameters were subjected to Analysis of Variance (1-way
ANOVA) and Tukey-HSD test for multiple comparisons. All statistical significance tests were
atthePO.05 level.
RESULTS AND DISCUSSION
The ammonia excretion rates of the control group (normoxia or 100% oxygen saturation)
and the group of shrimp exposed to different lengths of hypoxia (10% oxygen saturation) are
summarized in Table 1. The control group mean ammonia excretion rate of 8.00 jimol/g'h is
higher than the value of 2.5 |imol/g/h reported in the shrimp Palaemonetes varians and the value
of 1.6 |imol/g/h reported in Crangon crangon (Hunter and Uglow 1993). However, those values
were measured at temperatures of 17 and 18.5°C, respectively and the values reported in this
study were taken at 28°C. The excretion of ammonia by crustaceans is greatly dependent on
environmental and dietary factors, for example temperature. Ammonia excretion rate generally
increases as temperature increases (Regnault 1987). The significantly higher ammonia excretion
rate found here after 30 and 60 min of hypoxia might be a response to an increase in (escape)
swimming activity observed during the first hour of exposure to hypoxia. Many crustacean
species have been shown to have some form of locomotory response to a marked lowering of
oxygen tensions. For example, Hagerman and Ostrup 1980 reported an increase in locomotory
activity in Palaemon adspersus. The reduced ammonia excretion rate of Litopenaeus vannamei
after 120 and 180 minutes (Table 1) showed a significantly decreased in metabolic activity. This
pattern of reduced rate of ammonia efflux under severe hypoxia has been observed to occur in
some species and may be a general trend amongst crustaceans (Regnault and Aldrich 1988).
Table 1. Mean ammonia excretion rate of white shrimp exposed to different periods of hypoxia.
Hypoxia exposure
(minutes)
0 (Control)
30
60
90
120
180
Ammonia excretion rate1
(umol/g/hr)
8.00+0.883
20.4+1.8b
23.1+2.3b
8.94+0.823
5.44+0.593
3.81+0.193
Values are mean + SE, N=8, in all cases. The differences of comparison of pairs with different letters were p <0.05
24
-------
The normoxic mean hemolymph concentration found in this study for Litopenaeus
vannamei of 746.78 + 72.3 |j,mol/L (Figure 1), is greater than values reported for cold-water
shrimp species (Hunter and Uglow 1993). This difference might be attributed to a higher
metabolic activity of semitropical shrimp species such as L. vannamei. The combination of the
lowered hemolymph ammonia levels and the decreased ammonia excretion rates found after 180
minutes of hypoxia (Table 1 and Figure 1) indicate ammonia production inhibition. This
paralleled the decrease in metabolic activity observed in shrimp that remained quiescent.
900 n
800-
-T 700
_i
a. o 600
I. =5- 500
cu
CO
-| 400
E 300
co 200
100
0
0
50 100
Time (min)
150
200
Figure 1. 1 Litopenaeus vannamei. Concentration of ammonia obtained from the
hemolymph of shrimp under normoxia (0 min) and different hypoxia exposure
times.
The hemolymph protein and hemocyanin levels of the control and treated groups are
shown in Table 2. The normoxic value of hemocyanin of 1.04 + 0.11 mmol/L for L. vannamei
corresponds well with values reported for other crustacean species (Magnum 1983). Synthesis
of new hemocyanin may be a natural response when faced with prolonged or more or less
permanent hypoxia (Hagerman 1986). Our results show a general pattern (see Table 2) for
hemocyanin similar to that for ammonia excretion. Specifically, there is significant increase up
to 120 minutes of hypoxia, which might reflect hemocyanin synthesis during this time.
However, hemocyanin response requires further research.
25
-------
Table 2. Hemolymph protein and hemocyanin levels in white shrimp exposed to different
periods of hypoxia.
Hypoxia exposure
(minutes)
0 (Control)
30
60
90
120
180
Protein1
(mg m/L)
104+13.0b
152+14.53
156+5.93
139+11. 9a
149+8.63
135+8. la
Haemocyanin1
(mmol/L)
1.04+0.11b
1.14+0.05bc
0.92+0.063
1.31+0.08d
1.42+0.08d
0.82+0.063
Values are mean + SE, N=8, in all cases. The differences of comparison of pairs with different letters were p <0.05
80 n
0
50
100
Time (min)
150
200
Figure 2. Litopenaeus vannamei. Concentration of lactate obtained from the
hemolymph of shrimp under normoxia (0 min) and different hypoxia exposure
times.
26
-------
Hemolymph lactate levels are shown in Figure 2. Hemolymph lactate was always
detected in L. vannamei. The normoxic value of 3.40 mg/dL is similar to levels in other
crustacean species as reported by Bridges and Brand (1980). The increased levels of hemolymph
lactate found in L. vannamei after 2 h of severe hypoxia were significantly different than those
found for the animals at normoxia (p > 0.05). Thus, anaerobic metabolism dominates when
oxygen no longer can fuel the basal metabolism in the tissues. After exposure to severe hypoxia,
the lactate levels in the hemolymph of L. vannamei rose rapidly, indicating a rapid use of
anaerobic pathways to meet energy demands. The lactate production rate after 2 hours of severe
hypoxia was as high as 28.33 mg/dL/h. This pattern of rapid transition to anaeubic respiration in
crustaceans during exposure to hypoxia has also been reported in the crabs Eriocheir sinensis
(Zou et al. 1996) and Nephrops norvegicus (Hagerman et al. 1990). In L. vannamei, lactate
accumulates under potentially lethal conditions so that, unless more favourable oxygen
conditions occur within 1 to 3 h, the shrimp will die.
The semitropical shrimp Litopenaeus vannamei thus appears to be a species that can not
tolerate severe hypoxia for times longer than 60 minutes. Furthermore, during severe hypoxic
periods L. vannamei turns to anaerobic metabolism, but at summer temperatures in the farming
ponds this will be lethal within a few hours. Our data suggest that greater care must be taken
relative to farm pond conditions, especially during summer, and a means of providing additional
aereation to the farm ponds, especially during critical hours, should be provided.
ACKNOWLEDGMENTS
We are grateful to Rene Valenzuela-Miranda for his technical support and Gisela
Carvallo-Ruiz and Carlos Verdugo-Salazar for their help in sampling. We would also like to
thank Ana Maria Calderon for the facilities provided.
REFERENCES
Bradford, M.M. 1976. A rapid and sensitive method for the quantification of micrograms
quantities of protein utilizing principle of protein dry binding. Analytica Biochemistry
72:248-254.
Bridges, C.R., and A.R. Brand. 1980. The effect of hypoxia on oxygen consumption and blood
lactate levels of some marine Crustacea. Comparative Biochemistry and Physiology
65A:399-409.
Butler, P., E. Taylor, and B. McMahon. 1978. Respiratory and circulatory changes in the lobster
(Homarus vulgaris) under long term exposure to moderate hypoxia. Journal of
Experimental Biology 73:131-146.
27
-------
Clinch, J.R., PJ. Worsfold, and F.W. Sweeting. 1988. An automated spectrophotometric field
monitor for water quality parameters: Determination of ammonia. Analytica Chimica
Acta 214:401-407.
Gade, G. 1983. Energy metabolism of arthropds and mollusks during environmental and
functional anaerobiosis. Journal of Experimental Zoology 228:415-429.
Hagerman, L., and J. Ostrup. 1980. Seasonal and diel activity variations in the shrimp Palaemon
adspersus from a brackish, non-tidal area. Marine Ecology Progress Series 2: 329-335.
Hagerman, L., and R.E. Weber. 1981. Respiratory rate, hemolymph oxygen tension, and
hemocyanin level in the shrimp Palaemon adspersus (Rathke). Journal of Experimental
Marine Biology and Ecology 54:13-20.
Hagerman, L. 1986. Hemocyanin concentration in the shrimp Crangon crangon (L.) after
exposure to moderate hypoxia. Comparative Biochemistry and Physiology 85A: 721-724.
Hagerman, L., T. Sondergaard,. K. Weile, D. Hosie, and R.F. Uglow. 1990. Aspects of blood
physiology and ammonia excretion in Nephrops norvegicus under hypoxia. Comparative
Biochemistry and Physiology 97A:51-55.
Hagerman, L., and A. Szaniawska. 1994. Hemolymph nitrogen compounds and ammonia efflux
rates under anoxia in the brackish water isopod Saduria entomon. Marine Ecology
Progress Series 103:285-289.
Hunter, D.A., and R.F. Uglow. 1993. Handling-induced changes in hemolymph ammonia
concentration and ammonia excretion rate of Crangon crangon (L.). Ophelia. 38 (2):
137-147.
Kormanik, G.A., and J.N. Cameron. 1981. Ammonia excretion in the seawater blue crab
(Callinectes sapidus) occurs by diffusion and not by Na+/NH4+ exchange. Journal of
Comparative Physiology B 141:457-462.
Magnum, C. 1983. Oxygen transport in the blood. In: The Biology of Crustacea. L.H. Mantel
(ed). Academic Press, New York, Volume 5 pp. 373-429
Regnault, M., and J.C. Aldrich. 1988. Short-term effect of hypoxia on ammonia excretion and
respiration rates in the crab Carcinus maenas. Marine Behaviour and Physiology 6:257-
271.
Regnault, M. 1987. Nitrogen excretion in marine and fresh-water Crustacea. Biological Reviews
62:1-24.
Teal, J., and F. Carey. 1967. The metabolism of marsh crabs under conditions of reduced oxygen
pressure. Physiological Zoology 40:83-91.
28
-------
Withers, P.C. 1992. Comparative Animal Physiology. Saunders College Publishing. Jovanovich
Publisher, Orlando, Florida, U.S.A.
Zou, E.,, and N. Du. 1993. The effects of acute progressive hypoxia on the respiration rate of the
Chinese crab, Eriocheir sinensis. Zoological Research 14: 327-334.
Zou, E., N. Du., and W. Lai. 1996. The effects of severe hypoxia on lactate and glucose
concentrations in the blood of the Chinese freshwater crab Eriocheir sinensis (Crustacea:
Decapoda). Comparative Biochemistry and Physiology 114A: 105-109.
29
-------
30
-------
THE EFFECTS OF AQUATIC HYPOXIA ON FISH
W.L. Poon, C.Y. Hung and DJ. Randall1
ABSTRACT
Aquatic hypoxia is a common and frequent event. Fish show a variety of responses to
hypoxia that increase in magnitude as hypoxia becomes more severe. Fish reduce food intake,
leading to a reduction in growth. Reproduction is inhibited, and both fertilization success and
larval survival are compromised. Fish attempt to maintain oxygen intake during hypoxia by
increasing gill ventilation and oxygen delivery via increased circulation. Energy utilization is
decreased, associated with a shift from aerobic to anaerobic metabolism. There are several
strategies for decreasing energy expenditure. These include moving to water at lower
temperature, and reducing activity, reproduction, feeding, and protein synthesis. Transcription is
reduced, mediated by increased levels of hypoxia-inducing factor 1 (HIF-1), which also up-
regulates genes involved in erythropoiesis, capillary growth and glucose transport. HIF-1 may
also be involved in hypoxia-induced apoptosis. All these responses are directed at maintaining
cellular oxygen homeostasis and reducing energy expenditure, thereby augmenting survival of
the animal during hypoxia. In general, the actions of toxicants are exacerbated during hypoxia,
through a variety of mechanisms. Some species are much more tolerant of hypoxia than others,
leading to differential survival during extended periods of hypoxia.
INTRODUCTION
Oxygen levels in the atmosphere began to increase about 2 billion years ago with the
advent of photosynthesis, reaching about 15% of the atmosphere during the Cambrium, some
600 million years ago. The balance between oxygen production and consumption has varied
over time, resulting in oxygen levels as high as 35% of the atmosphere in the Carboniferous and
Permian periods, about 300 million years ago (Dudley 1998). Since then, oxygen levels have
stabilized at around 21%. Although carbon dioxide levels in the atmosphere are increasing due
to the burning of fossil fuels, there is no measurable change in present oxygen levels because of
the very large stores of oxygen in the Earth's surface.
The proportion of oxygen in the biosphere dissolved in water is very small because of the
low solubility of oxygen in water. Rates of diffusion in water are also slow, being 10"5 of that in
air. Distribution of oxygen in the water column thus depends on mixing from the surface layers
where photosynthesis occurs. At depth there is no light and, therefore, no photosynthesis. Thus,
hypoxia occurs at depth in unmixed waters. At a particular depth in the oceans where light is
very reduced, there are many organisms supported by food dropping from the upper lighted zone.
Department of Biology and Chemistry, City University of Hong Kong, Kowloon, Hong Kong, SAR, China.
31
-------
These organisms consume oxygen, but little is produced by photosynthesis, creating a minimum
oxygen (hypoxic) layer that extends throughout the oceans. This minimum oxygen layer, or
zone, extends for vast areas and can be as low as 0.5 mg O2/L.
During the night, in the absence of photosynthesis, there is often a measurable drop in
oxygen levels in water, especially in tropical lakes and lagoons, because stores are small, mixing
is limited, and transfer between the atmosphere and the water is slow. This nocturnal hypoxia is
often associated with oxygen levels that exceed atmospheric solubility equilibrium levels during
the day. Thus many aquatic systems are subject to large diurnal oscillations in oxygen levels,
being hyperoxic during the day and hypoxic at night. Ice reduces both light penetration and the
transfer of oxygen into water from the atmosphere. As a result, hypoxia is common in ice-
covered lakes.
Our aquatic ancestors evolved in water, when oxygen levels were about half of present
day levels, whilst mammals evolved in atmospheric levels approaching those of the present.
Aquatic animals, breathing water, are subjected to frequent and sometimes unpredictable
changes in the oxygen content of their medium. Thus, it can be expected that these animals have
highly evolved mechanisms for surviving aquatic hypoxia, perhaps more sophisticated than those
of the more studied terrestrial mammals.
Most humans live close to the coast or along rivers and deposit their waste into
neighboring waters. This usually results in the addition of nutrients to the water resulting in
eutrophication that, along with decomposition of waste and fertilizer run-off, increases oxygen
utilization and causes aquatic hypoxia. The result is that coastal regions, especially river
estuaries, are experiencing increased levels of hypoxia, particularly in summer. Thus, aquatic
hypoxia has been severely exacerbated by anthropogenic inputs over the past fifty years. The
consequences are a reduction in biodiversity and biomass, the flourishing of hypoxia resistant
species, and the removal of commercially important species. Fish vary in their ability to survive
hypoxia, and the increased bouts of aquatic hypoxia are having a marked effect on fish numbers
and distribution. In this review, we will discuss the responses offish to hypoxia at the whole
animal, organ and tissue level.
RESPONSES OF FISH TO HYPOXIA
Hypoxia and Food Intake
Acquisition of food and its digestion and assimilation are major energy expenditures (up
to 60%) of fishes (Dam and Pauly 1995). A number of factors affect food intake by fish,
including water temperature, salinity, photoperiod, density, fish size, social interactions, food
availability, hormonal interactions, pollutants, etc. (Jobling 1994). Decreased oxygen
availability is also considered a major factor in determining food intake. Low dissolved oxygen
is a type of stress frequently found in fish farms characterized by high fish densities and polluted
fresh or marine waters.
32
-------
Several studies have investigated the relationship between hypoxia and fish food intake.
Randolph and Clemens (1976) found that feeding patterns of channel catfish varied with
temperature and oxygen availability. During summer time, when oxygen content dropped below
5 mg/L, the fish adjusted or missed their daily feeding period (Figure 1). Rainbow trout
(Oncorhynchus mykiss) reduced appetite when oxygen saturation fell below approximately 60%
(Jobling 1994). Similar results have been obtained from European sea bass (Dicentrachus
labrax, L) (Thetmeyer etal. 1999), blue tilapia (Oreochromis aureus) (Papoutsoglou and Tziha
1996), channel catfish (Ictaluruspunctatus) (Buentello et al. 2000), juvenile turbot (Pichavant et
al. 2001) and common carp (Cyprinus carpio. L) (Zhou unpublished data). All these fish
experienced reduced growth. The RNA quantity reflects the anabolic activity of the fish whereas
the DNA content represents the standard reference (Jobling 1994). In common carp, the
RNA/DNA ratio was significantly reduced in the white muscle offish exposed to 1 mg O2/L or
lower (Zhou et al. 2001) (Figure 2).
Figure l.(a) Percentage of twenty marked (25 and 46 cm in length) channel catfish feeding at
various temperatures during March, showing that small and large fish preferred different
temperature ranges for feeding (n = 632). (b) Percentage of twenty marked (25 and 46
cm in length) channel catfish feeding at various oxygen concentrations in large ponds
during July and August, showing that small and large fish fed over different oxygen
ranges (n = 840), but all stopped feeding at low oxygen levels (from Randolph and
Clemens 1976).
33
-------
* C rsa r v ti ::?|ii'
*
1
Figure 2. Specific growth rates and the RNA/DNA ratio of common carp after exposure
to normoxia and hypoxia for 1 or 4 weeks. (**P < 0.01; P < 0.001; n = 4 -8; mean
+ SE, from Zhou etal. 2001).
It is not surprising that fish stop feeding during hypoxia. When oxygen availability in
the water drops to a level that cannot support aerobic metabolism, fish shift to anaerobic
pathways for energy production. Subsequently, metabolic depression occurs to minimize energy
expenditure (see below for details). Fish reduce or stop feeding completely during hypoxic
conditions, presumably because food digestion is energetically demanding. The reduction in
food intake could also be due to the fact that under hypoxic condition fish become less active and
the energy required for normal locomotion is used instead to sustain basal metabolic process.
Using cDNA microarray gene expression profiling, Gracey etal. (2001) showed that genes
involved in the glycolytic metabolic pathway, muscle contraction and locomotion are all down-
regulated in the muscle cells of the euryoxic fish, Gillichthys mirabilis, when exposed to low
oxygen levels. On the other hand, several genes involved in gluconeogenesis, as well as others
that play important roles in the suppression of cell proliferation and growth, were up-regulated in
the liver during hypoxia.
The mechanisms governing the reduction in food intake under hypoxia, at present, are
unclear. Some evidence suggests that corticotrophin-releasing hormone (CRH) is a strong
appetite-suppressor in mammals (for review, see Morley 1987). Comprehensive studies on the
effect of CRH on fish food uptake were conducted by means of intacerebroventricular
administration (De Pedro etal. 1993, 1995, 1997; Bernier and Peter 2001) and by intraperitoneal
implantation (Bernier and Peter 2001). Bernier and Peter (2001) showed that urotensin I (UI), a
34
-------
member of the CRH family of peptides, was a more potent inhibitor than CRH, and that both
CRH and UI appeared to be dose-dependent in suppressing food intake (Figure 3).
a.C-
*••
•*••"" rr
0.1 ' C« 'K1:d liifla
0! p€ptt06 1?!Q ,' g 8W)
Figure 3. Hill plots demonstrating the ED50 of corticotrophin-releasing hormone (CRH) (ED50 =
43.Ing/g BW) and urotensin I (UI) (EDso =38 ng/g BW) injections on food intake (FI) in
goldfish. Regression analysis and test for parallelism (analysis of covariance) indicated
that UI was significantly (p<0.05) more potent than CRH in suppressing food intake.
(From Bernier and Peter 2001).
In addition, other peptides such as Bombesin (BBS) and Cholecystokinin (CCK) have
also been found to have roles in gastrointestinal regulation, and probably exert anorectic effects
when injected into fish (for review, see De Pedro and Bjornsson 2001). Nonetheless, not much
work has been carried out to investigate the effects of these appetite-suppressive hormones in
relation to the reduction of food intake in fish exposed to hypoxia.
Hypoxia and Reproduction
Hypoxia has profound effects on the process of reproduction; including puberty, gonadal
development and fertility (Bentley 1998). Studies have shown that mammals, when subjected to
high altitude, have delayed puberty and a prolonged sexual maturation period. Decreased
testicular size, Leydig cell number, and reduced sperm viability have been found in men at high
altitude. For women at high altitude, decreased fertility, a reduction in the frequency of
ovulation and a lowered amount of sex hormones (estradiol, progesterone, prolactin) have been
reported (Ducsay 1999).
Very little has been published on the relationship between hypoxia and reproduction in
fish. However, in the past decade, hypoxia in coastal marine waters has been associated with a
major change in fish species composition with a reduction in the number of demersal fishes.
One possible explanation of such a phenomenon is the impairment of gonadal development and
eventually increased failure in spawning, fertilization, hatching, and survival. Zhou (2001)
studied the effects of hypoxia on reproduction of the common carp (Cyprinus carpid). Gonad
development was reduced when fish were exposed to hypoxia for 8 weeks (Figure 4). The
underdeveloped gonads had significant reduction in the number of spermatocytes and
-------
spermatids, lowered incidents of mitosis, decreased lobular diameter of testes, and reduced
sperm motility (Figure 5). In female carp, oocytes from hypoxic fish remained in the early
stages of the developmental process, whereas normoxic female carp had oocytes that were near
completion of the developmental process. Successful spawning females were 71.4% in the
normoxic group, significantly higher than the hypoxic group (8.3%). There was a rapid decrease
in the percentages of fertilization success (99.4% in normoxia and 55.5% in hypoxia); hatching
(98.8% in normoxia and 17.2% in hypoxia); and survival of larvae (93.7% in normoxia and
46.4% in hypoxia) (Figure 6). In addition, hormone assays showed that there were significant
decreases in serum testosterone, estradiol (Figure 7 and Figure 8) and triiodothyronine in carp
exposed to hypoxia. This data implies that spermatogenesis and oogenesis, which are controlled
by neural-endocrine signals, are highly affected by hypoxia. In fact, evidence has shown that
cytochrome P450 enzymes, which require oxygen and are involved in steriodgenesis, are down-
regulated during hypoxia (Galal and du-Souich 1999). Hence, one may infer that reduced
oxygen availability may decrease steriod production, which, in turn, delays development and
affects the normal growth of gonads in the common carp.
Figure 4. Lobule diameter of testes of C. carpio upon exposure to 7.0 and 1.0 mg O2fL for 8
weeks. A value significantly different from the control is indicated by an asterisk (n = 1-
ll,mean±SE). ***;p< 0.001. (From Zhou 2001).
Figure 5. Sperm motility of C. carpio upon exposure to 7.0 or 1.0 mg O2/L for 12 weeks. Values
significantly different from the control are indicated by asterisks (n = 6, mean +SE). *:p
< 0.05. VCL: mean curviline velocity; VAP: angular path velocity; VSL: mean straight
line velocity. (From Zhou 2001).
36
-------
.|B>
.
Mitettfig LtfrMlam m it
to
Figure 6. (A) Percentage survival of eggs to larvae; (B) Percentage of fertilization, hatching rate
and larval survivorship of C.carpio upon exposure to 7.0 or 1 mg C>2/L for 12 weeks.
Values significantly different from the control are indicated by asterisks, (n = 6, mean +
SE). (**:/K0.01; ***p<0.00l). (From Zhou 2001).
The mechanisms underlying the defects are unknown at the moment. Nevertheless, it is
known that the physiology associated with reproductive capability is closely related to stress in
general. Serum cortisol levels were significantly higher when fish were stressed by various
means (Sjoerd 1997, Schreck etal. 2001). Plasma cortisol levels were significantly elevated
from 78 ± 9 ng/ml to 735 ± 424 ng/ml and 270 ±37 ng/ml in surviving and non-surviving
rainbow trout, respectively, after 4.5 hours of hypoxic exposure (van Raaji et al. 1996).
However, whether cortisol has any negative effect on reproduction is still the subject of much
debate.
It could also be possible that the reduction in body weight delays the onset of puberty in
fish. Since fish reduce food intake during hypoxic conditions (Thetmeyer et al. 1999, Pichavant
et al. 2001) to minimize the energetic costs associated with feeding, gonadal development might
be affected since overall growth is retarded in fish subjected to hypoxic conditions. Moreover,
the limited energy resources available to be allocated to eggs and sperm during hypoxia might
account for the reduced progeny viability.
Hypoxia has become a profound environmental problem in the recent years. Studies to
investigate its effect on reproduction in fish would definitely help to minimize the loss of this
natural resource and protect fish species from extinction.
37
-------
fT3& 1 i1 ma:"'
, _ , jsi£Ji]Sf£iSs£!i---.
4
; ,„
•Swetfcs
;~" -Cmgl"1
SS3 1 angf*
§
Figure 7. Levels of (A) testosterone and (B) estradiol in male C. carpio upon exposure to 7.0 or
1.0 mg O2/L for 4 or 8 weeks. Values significantly different from the control are
indicated by asterisks (n = 6-7, mean +SE). *p<0.05; **p< 0.01; ***;p < 0.001. (From
Zhou 2001).
I i
Figure 8. Levels of testosterone and estradiol in female C. carpio upon exposure to 7.0 or 1.0
mg C>2/L for 8 weeks. Values significantly different from the control are indicated by
asterisks (n = 6-7, mean +SE). *p<0.05; ***;p < 0.001. (From Zhou 2001).
38
-------
Maintaining Oxygen Delivery During Hypoxia
Fish respond to hypoxia initially by maintaining oxygen delivery in the face of reduced
oxygen content of the medium. This response is rapid and is the first line of defense. If the
hypoxia is prolonged, then the fish reduces energy expenditure. If the animal cannot maintain
energy supply by aerobic means, then it up-regulates anaerobic pathways.
Fish invoke a number of behavioral responses when subjected to aquatic hypoxia. Some
species move to the air-water interface and skim the oxygen rich surface water (Val 1995).
Many other fish species breathe air. In this case they usually do not use gills, but have some
other modified region of the body that acts as a respiratory organ (Graham 1997). Many species
try to leave the hypoxic region that, in water, can be patchy and discontinuous. They move to
water of a lower temperature, reducing body temperature and, therefore, energy requirements.
Fish limit movement, often remaining stationary for days, to reduce energy consumption
when exposed to hypoxia. The swimming speeds of Atlantic cod (Schurmann and Steffensen
1994), yellowfin tuna (Korsmeyer et al. 1996), coho salmon and largemouth bass (Dalberg et al.
1968) all are reduced during hypoxia; swimming may even cease in order to minimize energy
consumption (Figure 9).
!
Figure 9. The effect of dissolved oxygen concentration on maximum swimming speed of coho
salmon (Oncorhynchus kisutch) at 20°C and largemouth bass (Microperus salmoides) at
25°C. (Velocity measurements were at 10 minute intervals). (From Heath 1995).
39
-------
Fish gills contain internal and external oxygen chemoreceptors. Hypoxia stimulates these
chemoreceptors, increasing ventilation of the gills to maintain the delivery of oxygen to the
respiratory surface. Heart rate is reduced (Peyraud-Waitzenegger and Soulier 1989) but, in some
cases, is associated with an increase in stroke volume. This changing pattern of blood flow
increases the gill diffusing capacity and augments the transfer of oxygen into the blood (Randall
1982). Decreased red blood cell phosphate levels result in an increase in hemoglobin oxygen
affinity and this facilitates oxygen uptake at the gills. The decrease in erythrocytic phosphate
levels is rapid enough to compensate for circadian oscillations in dissolved oxygen in the
environment. All-major modulators of vertebrate hemoglobin-oxygen affinity appeared during
the radiation offish. As in mammals, anemia results in an increase in erythrocytic phosphate
levels facilitating oxygen unloading to tissues (Val and de Almeida-Val 1995). Blood
erythrocyte levels are increased initially due to release from the spleen and then subsequently
due to erythropoiesis in response to the hormone, erythropoietin (EPO), produced by the kidney.
Hypoxia results in increased EPO levels in the kidney and spleen of rainbow trout (Kakuta and
Randall, unpublished observations). Exposure to 30% oxygen-saturated water resulted in kidney
EPO increases that peaked after 24 hours exposure (Figure 10). There was an initial increase in
blood hemoglobin levels, probably due to red blood cell release from the spleen, whereas an
observed increase in blood hemoglobin levels after 6 days was probably due to erythropoiesis.
In contrast, exposure to lower levels of hypoxia (55% oxygen saturation) was associated with a
kidney EPO increase after 6 days (Figure 11) with no change seen at 1 and 3 days.
c
s
15
•o.
i
o
Q
T"
>^
I
SO
8h
24h
72h
Figure 10. EPO in rainbow trout kidney, hypoxia = 30% oxygen saturation in water.
40
-------
5W
¥
It
s
fc
? 20
t
i
1
- 10
o
CL
u
| Control
^ Hypoxia
ffifefr!Sv5
Livmt
Tiasnie
Figure 11. Erythropoietin levels in various tissues of rainbow trout following 6 days hypoxia at
55% oxygen saturation in water.
Increased intracellular adenosine levels during hypoxia result from the metabolism of
ATP/ADP/AMP and s-adenosylhomocysteine to adenosine. In mammals, adenosine is
transported into the extracellular space and can reach concentrations of 0.1 to 0.3 |jM that last
only a few seconds because adenosine is either rapidly taken-up by cells or converted to inosine.
Thus, adenosine only acts locally and very briefly. Adenosine has many actions in mammals,
including reducing energy demand and at the same time increasing oxygen supply by
vasodilation of blood vessels (see review by Poulsen and Quinn 1998). It appears that adenosine
plays a similar role of matching energy supply and demand during hypoxia in fish (Bernier et a/.
1996 a&b). Adenosine levels in the brain of carp, however, remain elevated for prolonged
periods (hours) and are associated with an increased cerebral blood flow (Nilsson et a/. 1994).
DOWN REGULATION OF METABOLISM
During hypoxia, oxygen supply is limited and there is a shift from aerobic to anaerobic
metabolism, as well as a reduction in energy expenditure. The common sole, Solea solea, a
benthic flatfish, reduces its resting metabolism by 27% and 48% when exposed to acute hypoxia
of 12% and 6% saturation, respectively (Dalla Via et al. 1994). During chronic hypoxia, sole
attempt to escape from the scene (Dalla Via et al. 1998). The european eel lowers its oxygen
requirements during hypoxia by decreasing its overall metabolic rate by 70% (van Ginneken et
al. 2001). As discussed earlier, reduced food intake and digestion rate, reduced activity and egg
and sperm production, and moving to a colder temperature all contribute to the reduction in
energy expenditure. The protein turnover rate is reduced during prolonged anoxia in crucian
carp (Smith et al. 1999). At the cellular level, energy conservation is associated with a marked
reduction in protein synthesis. Smith et al. (1996) found that when crucian carp were exposed to
48h anoxia, there was more than a 95% reduction of protein synthesis rate in liver, 53% in heart,
52% in red muscle and 56% in white muscle, but no change in synthesis rate was found in the
41
-------
brain. Glycogen stores in the liver are reduced during hypoxia (Zhou et al. 2000) and are the
main energy source during prolonged hypoxia exposure in support of anaerobic metabolism.
Reduced Na+/K+ pump activity and Ca2+ cycling also can reduce energy consumption,
the so called "Channel arrest" (Hochachka 1986). A reduction in cell membrane permeability,
which in turn reduces the energy cost of maintaining transmembrane ion gradients, has been
reported in hepatocytes and neurons subjected to anoxia (Boutilier 2001).
ANAEROBIC METABOLISM
Aerobic and anaerobic metabolism are both involved in producing energy during
hypoxia (van den Thillart and van Waarde 1985). Under normal oxygen concentrations or mild
hypoxia, aerobic metabolism is dominant; however, anaerobic metabolism is dominant under
deep and chronic hypoxic conditions. Astronotus ocellatus, an amazon fish, decreases its
standard metabolic rate under hypoxia, but shifts to anaerobic metabolism only under deep
hypoxia or anoxia (Muusze et al. 1998). It appears that the degree of metabolism depression
and/or energy generation by aerobic or anaerobic means is species dependent, and even tissue
specific (van Ginneken et al. 1995). Goldfish and crucian carp generate energy by
decarboxylating pyruvate to acetaldehyde by dehydrogenation (van den Thillart and van Waarde
1985). Lactic acid is also converted to ethanol and carbon dioxide (Stangl and Wegener 1996).
This may be one of the reasons why goldfish and crucian carp can survive deep hypoxia or even
anoxia.
HIF-1
Hypoxia-inducible factor 1 (HIF-1), a transcription factor, was found to be up- regulated
during hypoxia. HIF-1 was first extracted from hypoxic cells by Semenza and Wang (1992) and
its DNA binding activity to the Hypoxia-Responsive Element (HRE) of hypoxia responsive
genes demonstrated (Wang and Semenza 1993 and review by Fandrey 1995). HIF-1 is a
heterodimer composed of two subunits, namely HIF-la and HIF-113 (also known as Aryl
Hydrocarbon Receptor Nuclear Translocator, ARNT). HIF-la and HIF-113 are constitutively
expressed. They both belong to the basic Helix-Loop-Helix-Per/ARNT/AhR/Sim (bHLH-PAS)
transcription factor family.
HIF-la is the HIF-1 subunit that is regulated by hypoxia. HIF-la protein was shown to
be rapidly degraded by 26S proteasome in the ubiquitin-proteasome system during normoxic
conditions (Salceda and Caro 1997; Sutler et al. 2000). However, it is stabilized by the
chaperone protein and heat shock protein 90 (Hsp90), translocates to the nucleus and dimerizes
with HIF-lp to form the HIF-1 complex under hypoxia (see review by Minet et al. 2001). In the
nucleus, HIF-1 binds to the consensus sequence 5'-RCGTG-3' in the HRE (see review by
Semenza 2000) of some hypoxia responsive genes such as vascular endothelial growth factor
(VEGF, involved in angiogenesis), erythropoietin (EPO, involved in erythropoiesis), glucose
transporter (GLUT, involved in glycolysis) and many other genes to induce or repress their
expression. In other words, HIF-1 is involved in a number of mechanisms and is an important
mediator of cellular and systemic oxygen homeostasis (Jonathan and Ratcliffe 1998; Seagroves
etal. 2001; Semenza 200la & b) (Figure 12).
42
-------
Cell
Figure 12. Hypoxia-inducible factor la (HIF-la) protein was degraded by proteasome 26S
under normoxia, but under hypoxia it is stabilized and hyterodimerized with hypoxia-
inducible factor 1|3 (HIF-113) in the nucleus to form HIF-1. HIF-1 protein binds to the
hypoxia responsive element (FIRE) of hypoxia responsive genes such as vascular
endothelial growth factor (VEGF), erythropoietin (EPO) and glucose transporter (Glut) to
regulate their expression. Modified from Brahimi-Horn et al. 2001.
Recent publications have shown that both phosphorylation and dephosphorylation can
activate HIF-1 (see review by Minet et al. 2001), indicating that the signaling mechanisms
triggered by hypoxia are complicated. Takagi et al (1998) demonstrated that the upregulation of
Glutl by HIF-1 was mediated by adenosine using bovine retinal endothelial cells, but the details
of the mechanism are not yet known.
HIF-1 has not been cloned from many species so far. In fish, only rainbow trout
(Oncorhynchus mykiss) HIF-1 has been described (Soitamo et al. 2001). Partial sequences have
been obtained from zebrafish (Danio rerio), grass carp (Ctenopharyngodon idelld) and
Janpanese medaka (rice fish) (Oryzias latipes).
HIF-1 has also been found to be involved in hypoxia-induced apoptosis. Apoptosis
(programmed cell death) is an ATP dependent process. In the past, it was believed that hypoxia
was associated with necrosis, an ATP independent process. However, more recent evidence
indicates that apoptosis is induced by hypoxia (Carmeliet et al. 1998; Minet et al. 2000; Riva et
al. 1998). Apoptosis represents controlled cell death and is presumably less damaging than
43
-------
necrosis, which can lead to inflammation and disease. There are a number of apoptotic
pathways, but the mechanism for hypoxia-induced apoptosis is still unclear. It is expected that
the nature of the hypoxia-induced apoptotic pathway may vary with the tissue and species, as
well as the level of hypoxia.
SUMMARY
Fish are often exposed to hypoxic conditions in the water and hypoxia frequency, lengths
and severity have all gotten worse in the last few decades due to anthropogenic inputs. Much of
vertebrate evolution occurred at much lower oxygen levels than exist at present. Undoubtedly,
the major responses evolved in aquatic ancestors, and these may be attenuated in terrestrial
vertebrates existing in relatively high oxygen concentrations in the atmosphere. There are many
similarities between fish and mammalian response to hypoxia, although much more is known
about the response of mammals. Responses observed in fish that are dissimilar to those observed
in mammals include: (1) erythrocytic phosphate levels decrease during aquatic hypoxia
increasing hemoglobin oxygen affinity; and (2) the heart slows but cardiac output remains
unchanged (there are exceptions); the resulting pattern of blood flow augments gill oxygen-
diffusing capacity.
There are also a few possible differences between fish and mammal response at the
cellular level, but little is known about this aspect of the responses offish to hypoxia. In all
cases, there is a marked reduction in fish energy expenditure during hypoxia. The relative role of
temperature change, starvation, adenosine production, and HIF-1 expression in metabolic
depression in fish is not clear. What is clear is that, in fish exposed to hypoxia, there is reduced
exercise and they move to lower temperatures. Food intake is reduced and this leads to reduced
growth and reproduction. The action of many toxicants is exacerbated during hypoxia. Fish are
the most diverse group of vertebrates, with an enormous number of species, and the responses to
hypoxia are known in only a few species. Even so, it is clear that some species offish are much
more tolerant to hypoxia than others.
ACKNOWLEDGEMENT
This work was supported in part by a Competitive Earmark Research Grant from the
Research Grants Council of Hong Kong awarded to DJ. Randall, R.S.S. Wu and Y.C. Kong (No.
9040658).
REFERENCES
Bentley, PJ. 1998. Comparative Vertebrate Endocrinology. Cambridge University Press,
Cambridge.
Bernier, N.J., J. Harris, J. Lessard, and DJ. Randall. 1996a. Adenosine receptor blockade and
hypoxia-tolerance in rainbow trout and hagfish. I. Effects on anaerobic metabolism.
Journal of Experimental Biology 199:485-495.
44
-------
Bernier, N.J., J. Fuentes, and DJ. Randall. 1996b. Adenosine receptor blockade and hypoxia-
tolerance in rainbow trout and hagfish. II. Effects on plasma catecholamines and
erythrocytes. Journal of Experimental Biology 199:497-507.
Bernier, N.J., and R.E. Peter. 2001. Appetite-suppression effects of urotensin I and
corticotropin-releasing hormone in Goldfish (Carassius arratus). Neuroendocrinology
73:248-260.
Boutilier, R.G. 2001. Mechanisms of cell survival in hypoxia and hypothermia. The Journal of
Experimental Biology 204: 3171-3181.
Brahimi-Horn, C., E. Berra, and J. Pouyssegur. 2001. Hypoxia: the tumor's gateway to
progression along the angiogenic pathway. Trends in Cell Biology 11: S32-S36.
Buentello, III, J.A., D.M. Gatlin, and W.H. Neill. 2000. Effects of water temperature and
dissolved oxygen on daily feed consumption, feed utilization and growth of channel
catfish (Ictaluruspunctatus). Aquaculture 182: 339-352.
Carmeliet, P., Y. Dor, J.M. Herbert, D. Fukumura, K. Brusselmans, M. Dewerchin, M.
Neeman, F. Bono, R. Abramovitch, P. Maxwell, C.J. Koch, P. Ratcliffe, L. Moons,
R.K. Jain, D. Collen, E. Keshert, and E. Keshet. 1998. Nature 394: 485-490.
Dalla Via, J., G. Van den Thillart, O. Cattani, and A. De Zwaan. 1994. Influence of long-term
hypoxia exposure on the energy metabolism of Solea solea. II. Intermediary metabolism
in blood, liver, and muscle. Marine Ecology Progress Series 111: 17-27'.
Dalla Via, D., G. Van den Thillart, O. Cattani, and P. Cortesi. 1998. Behavioural responses and
biochemical correlates in Solea solea to gradual hypoxic exposure. Canadian Journal of
Zoology 76: 2108-2113.
Dam, V., and D. Pauly. 1995. Simulation of the effects of oxygen on feed intake and growth of
Nile tilapia, Oreochromis niloticus L. Aquaculture Research 26: 427-440.
De Pedro, N., A.L. Alonso-Gomez, B. Gancedo, M.J. Delgado, and M. Alonso-Bedate. 1993.
Role of corticotropin-releasing factor (CRF) as a food intake regulator in goldfish.
Physiology and Behavior 53: 517-520.
De Pedro, N., B. Gancedo, A.L. Alonso-Gomez, M.J. Delgado, and M. Alonso-Bedate. 1995.
Alternations in food intake and thyroid tissue content by corticotropin-releasing factor
in Tunia tinea. Journal of Physiology and Biochemistry 51: 71-76.
De Pedro N., A.L. Alonso-Gomez, B. Gancedo, A.I. Valenciano, M.J. Delgado, and M.
Alonso-Bedate. 1997. The effect of a-helical CRF (9-41) on feeding in goldfish:
involvement of cortisol and catecholamines. Behavioral Neuroscience 3: 398-403.
45
-------
De Pedro, N., and B.T. Bjornsson. 2001. Regulation of food intake by neuroprptides and
hormones. In: Food Intake in Fish. D. Houlihan, T. Boujard, and M. Jobling (eds.).
Blackwell Science, Maiden , MA.
Ducsay, C.A. 1999. Hypoxia, Effect on reproduction. In: Encyclopedia of Reproduction
Volume 2. pp. 769-775 Academic Press, New York.
Dudley, R. 1998. Atmospheric oxygen, giant Paleozoic insects and the evolution of aerial
locomotor performance. Journal of Experimental Biology 201: 1043-1050.
Fandrey, J. 1995. Hypoxia-inducible gene expression. Respiration Physiology 101:1-10.
Galal, A., and P. du-Souich. 1999. 21-aminosteroids prevent the down-regulation of hepatic
cytochrome P450 induced by hypoxia and inflammation in conscious rabbits. British
Journal of Pharmacology 128: 374-379.
Gracey, A.Y., J.V. Troll, and G.N. Somero. 2001. Hypoxia-Induced Gene Expression Profiling
in the Euryoxic Fish, Gillichthys mirabilis. Proceedings of National Academy of
Sciences of the United States of America 98: 1993-1998.
Graham, J.B. 1997. Air-breathing fishes: evolution, diversity, and adaptation. Academic Press,
San Diego. 299pp.
Heath, A.G. 1995. Water pollution and fish physiology. Lewis Publishers, Boca Raton. 359pp.
Hochachka, P.W. 1986. Defence strategies against hypoxia and hypothermia. Science 231: 234-
241.
Jobling, M. 1994. Fish Bioenergetics. Volume 13: 1-44 Chapman and Hall. London.
Jonathan, M.G., and P.J. Ratcliffe. 1998. Hypoxia and the regulation of gene expression.
Molecular Medicine Today 4:122-129.
Korsmeyer, K.E., H. Dewar, N.C. Lai, and J.B. Graham. 1996. Tuna aerobic swimming
performance: Physiological and environmental limits based on oxygen supply and
demand. Comparative Biochemistry and Physiology 113B: 45-56.
Minet, E., G. Michel, J. Remacle, and C. Michiels. 2000. Role of HIF-1 as a transcription factor
involved in embryonic development, cancer progression and apoptosis (review).
International Journal of Molecular Medicine 5: 253-259.
Minet, E., G. Michel, D. Mottet, M. Raes, and C. Michiels. 2001. Transduction pathways
involved in hypoxia-inducible factor-1 phosphorylation and activation. Free Radical
Biology and Medicine 31: 847-855.
46
-------
Morley, I.E. 1987. Neuropeptide regulation of appetite and weight. Endocrine reviews 8: 256-
287.
Muusze, B., J. Marcon, G. van den Thillart, and V. Almeida-Val. 1998. Hypoxia tolerance of
Amazon fish respirometry and energy metabolism of the cichlid Astronotus ocellatus.
Comparative Biochemistry and Physiology 120A: 151-156.
Nilsson, G.E., P. Hylland, and C.O. Lofman. 1994. Anoxia and adenosine induce increased
cerebral blood flow rate in crucian carp. American Journal of Physiology 267: R590-
R595.
Papoutsoglou, S.E., and G. Tziha. 1996. Blue tilapia (Oreochromis aureus) growth rate in
relation to dissolved oxygen concentration under recirculated water conditions.
Aquaculture Engineering 15: 181-192.
Peyraud-Waitzenegger, M., and P. Soulier. 1989. Ventilatory and circulatory adjustments in the
European eel (Anguilla auguilla L.) exposed to short term hypoxia. Experimental
Biology (Berlin) 48: 107-122.
Pichavant, K., J. Person-Le-Ruyet, N.L. Bayon, A. Severe, A.L. Roux, and G. Boeuf. 2001.
Comparative effects of long-term hypoxia on growth, feeding and oxygen consumption
in juvenile turbot and European sea bass. Journal of Fish Biology 59: 875-883.
Poulsen, S-A., andR.J. Quinn. 1998. Adenosine Receptors: New Opportunities for Future
Drugs. Bioorganic and Medicinal Chemistry 6: 619-642.
Randall, D.J. 1982. The control of respiration and circulation in fish during exercise and
hypoxia. Journal of Experimental Biology 100: 275-288.
Randolph, K.N., and H.P. Clemens. 1976. Some factors influencing the feeding behaviour of
channel catfish in culture ponds. Transactions of American Fisheries Society 105: 718-
724.
Riva, C., C. Chauvin, C. Pison, and X. Leverve. 1998. Cellular physiology and molecular
events in hypoxia-induced apoptosis. Anticancer Research 18: 4729-4736.
Salceda, S., and J. Caro. 1997. Hypoxia-inducible factor lalpha (HIF-lalpha) protein is rapidly
degraded by the ubiquitin-proteasome system under normoxic conditions. Its
stabilization by hypoxia depends on redox-induced changes. The Journal of Biological
Chemistry 272: 22642-22647.
Schreck, C.B., W. Contreras-Sanchez, and M.S. Fitzpatrick. 2001. Effects of stress on fish
reproduction, gamete quality, and progeny. Aquaculture 197: 3-24.
47
-------
Schurmann, H., and J.F. Steffensen. 1994. Spontaneous swimming activity of Atlantic Cod
GadusMorhua exposed to graded hypoxia at three temperatures. Journal of
Experimental Biology 197: 129-142.
Seagroves, T.N., H.E. Ryan, H. Lu, E.G. Wouters, M. Knapp, P. Thibault, K. Laderoute, and
R.S. Johnson. 2001. Transcription factor HIF-1 is a necessary mediator of the Pasteur
effect in mammalian cells. Molecular and Cellular Biology 21: 3436-3444.
Semenza, G.L., and G.L. Wang. 1992. A nuclear factor induced by hypoxia via de novo protein
synthesis binds to the human erythropoietin gene enhancer at a site required for
transcription activation. Molecular and Cellular Biology 12:5447-5454.
Semenza, G.L. 2000. HIF-1: mediator of physiological and pathophysiological responses to
hypoxia. Journal of Applied Physiology 88: 1474-1480.
Semenza, G.L. 2001a. Hypoxia-inducible factor 1: oxygen homeostasis and disease
pathophysiology. Trends in Molecular Medicine 7: 345-350.
Semenza, G.L. 2001b. HIF-1, C>2, and the 3 PHDs: How animal cells signal hypoxia to the
nucleus. Cell 107:1-3.
Sjoerd, E.W.B. 1997. The stress response in fish. Physiological Reviews 77: 591-625.
Smith, R.W., D.F. Houlihan, G.E. Nilsson, and J.G. Brechin. 1996. Tissue-specific changes in
protein synthesis rates in vivo during anoxia in crucian carp. American Journal of
Physiology 40: R897-R904.
Smith, R.W., D.F. Houlihan, G.E. Nilsson, and J. Alexandre. 1999. Tissue-specific changes in
RNA synthesis in vivo during anoxia in crucian carp. American Journal of Physiology
277: R690-R697.
Soitamo, A.J., C.M. Rabergh, M. Gassmann, L. Sistonen, and M. Nikinmaa. 2001.
Characterization of a hypoxia-inducible factor (HIF-1 a) from rainbow trout. The
Journal of Biological Chemistry 276: 19699-19705.
Stangl, P., and G. Wegener. 1996. Calorimetric and biochemical studies on the effects of
environmental hypoxia and chemicals on freshwater fish. Thermochimica Acta
271:101-113.
Sutler, C.H., R.E. Laughne, and G.L. Semenza. 2000. Hypoxia-inducible factor la protein
expression is controlled by oxygen-regulated ubiquitination that is disrupted by
deletions and missense mutations. Proceedings of the National Academy of Sciences of
the United States of America 97: 4748-4753.
48
-------
Takagi, H., King, G.L., and L.P. Aiello. 1998. Hypoxia up-regulates glucose transport activity
through an adenosine-mediated increase of GLUT 1 expression in retinal capillary
endothelial cells. Diabetes 47: 1480-1488.
Thetmeyer, H., U. Waller, K.D. Black, S. Inselmann, and H. Rosenthal. 1999. Growth of
European sea bass (Dicentrarchus labrax L.) under hypoxia and oscillating oxygen
conditions. Aquaculture 174: 355-367.
Val, A.L. 1995. Oxygen transfer in fish: morphological and molecular adjustments. The
Brazilian Journal of Medical and Biological Research 28: 1119-1127.
Val, A.L., and V.M.F. de Almeida-Val. 1995. Fishes of the Amazon and their environment:
physiological and biochemical aspects. Zoophysiology. Volume 32. Springer-Verlag,
Heidelberg. 224p.
Van den Thillart, G., and A. van Waarde. 1985. Teleosts in hypoxia: Aspects of anaerobic
metabolism. Molecular Physiology 8: 393-409.
Van Ginneken, V., G. van den Thillart, A. Addink, and C. Erkelens. 1995. Fish muscle energy
metabolism measured during hypoxia and recovery: An in vivo of 31P-NMR study.
American Journal of Physiology 288(5 part 2): Rl 178-R1187.
Van Ginneken, V.J.T., M. Onderwater, O.L. Oliver, and G.E.E.J.M. van den Thillart. 2001.
Metabolic depression and investigation of glucose/ethanol conversion in the European
eel (Anguilla anguilla Linnaeus 1758) during anaerobiosis. Thermochimica Acta 373:
23-30.
Van Raaji M.T.M., D.S.S. Pit, P.H.M. Balm, A.B. Steffens, and G.E.E.J.M. van den Thillart.
1996. Behavioral strategy and the physiological stress response in rainbow trout
exposed to severe hypoxia. Hormones and Behavior 30: 85-92.
Wang, G.L., and G.L. Semenza. 1993. Characterization of hypoxia-inducible factor 1 and
regulation of DNA binding activity by hypoxia. The Journal of Biological Chemistry
268:21513-21518.
Zhou, B.S. 2001. Effects of hypoxia on metabolism, bioenergetics, growth and reproduction of
the common carp (Cyprinus carpio L.). Ph.D. Thesis. City University of Hong Kong.
Zhou, B.S., R.S.S. Wu, D.J. Randall, and P.K.S. Lam. 2001. Bioenergetics and RNA/DNA
ratios in the common carp (Cyprinus carpio) under hypoxia. Journal of Comparative
Physiology B 171: 49-57.
Zhou, B.S., R.S.S Wu, D.J. Randall, P.K.S. Lam, Y.K. Ip, and S.F. Chew. 2000. Metabolic
adjustments in the common carp during prolonged hypoxia. Journal of Fish Biology 57:
1160-1171.
49
-------
50
-------
IMPACT OF HYPOXIA ON ATLANTIC COD
IN THE NORTHERN GULF OF ST. LAWRENCE
Jean-Denis Dutil1 and Denis Chabot1
ABSTRACT
Oxygen levels range from 70 to 30% saturation or less in waters deeper than 175 m in the
Gulf of St. Lawrence. The potential impact of hypoxia on two stocks of Atlantic Cod (Gadus
morhua) that live in the Gulf of St. Lawrence was investigated. In a first experiment (84 days,
three meals per week), cod were raised under six oxygen treatments between 45 and 93%
saturation. Growth and food ingestion were limited by oxygen below 70% saturation. In a
second experiment (56 days), we varied meal frequency (one, three, and seven meals per week)
for cod held at two regimes of dissolved oxygen (40 and 90% saturation). In normoxia, growth
was significantly reduced at one meal per week compared to three and seven meals per week. In
hypoxia, however, there was no difference in growth between the three feeding frequencies. At
one meal per week, growth was equally poor at both oxygen levels. At three and seven meals
per week, growth was faster in normoxia than in hypoxia, although this was significant only at
the highest feeding frequency. In both experiments, food consumption explained practically all
of the variability in growth rate.
INTRODUCTION
To assess the areal extent of hypoxic zones requires a precise definition of the term
"hypoxia". While many authors focus on a concentration of dissolved oxygen below 2 mg/L as a
threshold value for aquatic environments, such an arbitrary limit may be unsuitable when
examining potential impacts on any one given species. Species differ in their basic oxygen
requirements, and oxygen requirements increase as energy-demanding metabolic processes are
mobilized. Fishes have developed several mechanisms to secure more oxygen from their
environment in critical situations such as low oxygen availability (Hoar and Randall 1984).
When the partial pressure of oxygen in the environment drops below some critical limit,
however, the pressure gradient between blood and water may not allow the fish to deliver as
much oxygen to its tissues as needed to meet metabolic requirements associated with ingestion,
digestion, growth and activity. Thus, critical thresholds may vary through time in demersal fish
species and are best described in terms of partial pressure of oxygen or of percent saturation.
Atlantic cod (Gadus morhua) is a demersal fish species that inhabits the North Atlantic
Ocean and adjacent marine areas including the Baltic Sea and the Gulf of St. Lawrence.
Reduced exchanges of water with the North Sea and human derived inputs promoting
eutrophication may explain the low oxygen levels observed in the Baltic Sea. In contrast,
^inistere des Peches et des Oceans, Institut Maurice-Lamontagne, 850 route de la Mer, Mont-Joli, Quebec, Canada
G5H 3Z4.
51
-------
hypoxia is generally considered to be a natural feature of bottom waters in the Gulf of St.
Lawrence. Labrador current waters mix with Gulf stream waters and penetrate into the Gulf of
St. Lawrence through Cabot Strait (Figure 1). Typically, deep waters in Cabot Strait are 60 to
70% saturated in oxygen. As these deep waters progress towards the head of the estuary, they
become progressively more depleted in oxygen (Figure 2). While surface waters are saturated
with oxygen throughout the Gulf of St. Lawrence, oxygen availability decreases with depth and
is minimal in the warmer bottom layer below 200 m, particularly in the Laurentian and Esquiman
channels (Gilbert et al. 1997).
0
100
£ 200
Temperature (°C) and O2 (mg/l)
-2 0 2 4 6 8 10 12
Q.
0)
O
300
400
500
Figure 1. Temperature and dissolved oxygen profiles near Cabot Strait, 16/05/91 (from
D' Amours 1993).
The amount of time spent by cod in the channels in the Gulf of St. Lawrence is unknown.
Recent evidence suggests that cod distribution has shifted to deeper waters, possibly in response
to a cooling event in the surface and mid-water layers, with cod being found in deeper waters in
the 1990s than before the cooling event occurred (Castonguay etal. 1999). Thus, some segments
of the cod population may have encountered low oxygen tensions in their routine feeding
activities or when migrating seasonally in and out of the Gulf of St. Lawrence. In order to
determine whether current levels of hypoxia in the Laurentian and Esquiman channels have
potential impacts on the survival, feeding, growth and swimming capacities of cod, we have
conducted a series of experiments designed to determine cod performance under hypoxic
52
-------
conditions. This short paper reviews some impacts of low oxygen availability on some aspects
of the cod physiology.
Figure 2. Dissolved oxygen levels (in percent saturation) 1 m off the bottom in the Gulf of St.
Lawrence, August 1992. The white line is the 200 m isobath.
MATERIALS AND METHODS
Survival Experiments
Survival of two size classes (45.2 ± 4.2 cm and 57.5 ± 3.8 cm) of cod was determined at
two temperatures (2 and 6°C) representative of the range of temperatures in the bottom waters of
the Gulf of St. Lawrence (Gilbert et al. 1997). Cod were transferred directly from normoxic to
hypoxic conditions and mortalities were assessed periodically (1, 3, 6, 12 hours and then every
12 hours) over a period of 96 hours. For each size and temperature combination, two sets of
experiments were conducted in which 60 cod were exposed to hypoxic waters ranging from 14 to
42% saturation (10 cod in each tank). Using PROBIT analyses, we determined two lethal
thresholds, i.e., saturation levels at which 5 and 50% of the fish died after 96 hours of exposure.
Feeding and Growth Experiments
Food ingestion, gross conversion efficiency and growth in length and weight of cod
averaging 44.2 ±3.1 cm in length and 715 ± 188 g in weight (somatic condition factor 0.81 ±
0.10) were determined for 120 cod exposed to hypoxic waters ranging from 45 to 93% saturation
(six levels of saturation, 20 fish per tank) over a period of 84 days. The individually-tagged cod
(Visible Implant Tags, Northwest Marine Technologies, Shaw Island, Washington) were fed
53
-------
three times a week with frozen capelin, a natural prey item in the cod diet. Each meal consisted
of feeding the fish ad libitum over a 1-hour period with surplus food removed. The experiment
was conducted at 10°C. To determine whether meal frequency had an effect on the relationship
between cod growth and oxygen availability, a shorter (46 days) experiment was conducted at
10°C. Individually tagged cod (nine or 10 cod per tank) were fed frozen capelin ad libitum for
one hour, once, three times, and seven times a week either in normoxia (>90% oxygen
saturation) or hypoxia (45% oxygen saturation) with replicates.
Stomach Contents
During trawl surveys conducted by the Department of Fisheries and Oceans to assess the
size of the northern Gulf of St. Lawrence cod stock, a length-stratified subsample of cod was
selected after each tow for stomach examination. For each stomach, an index of stomach
fullness was calculated as:
Fullness Index = 10000 • C • FL
-3
where C is stomach content mass in g, and FL is fork length in cm. For the period 1993-1998,
stomachs collected during the feeding season (July-October) were grouped by 50-m strata
according to the depths that the fish were caught. Mean stomach fullness and the proportion of
empty stomachs (in percent) were calculated for each depth category.
RESULTS AND DISCUSSION
Atlantic cod mortalities occurred at dissolved oxygen levels above the lowest levels
observed in the bottom waters of the Gulf of St. Lawrence. Percent oxygen saturation below
which 50% of the fish died within 96 hours was 21.0% (confidence interval 19.9-22.1) while
c
o
'^
1
•4-i
re
o>
"re
•4-i
0)
30
25
20
15
10
5
Error bars = 95% Cl
-A- -LC,
50
20 40 60
Exposure time (h)
80
100
Figure 3. Dissolved oxygen levels (in percent saturation) that killed 5% (squares) and 50%
(triangles) at each of the sampling periods. The final values (96 hours) were adopted to
define hypoxia tolerance in cod (Plante et al. 1998).
54
-------
percent oxygen saturation below which 5% of the fish died in the same period was 28.1% (25.8-
30.5) (Figure 3) (Plante et al. 1998). This latter value can be considered the incipient lethal
oxygen threshold for cod. No size (45 vs. 58 cm in lengeth) or temperature (2 and 6°C) effects
were observed. Lower tolerance limits previously obtained for cod (Sundnes 1957, Scholz and
Waller 1992, Schumann and Steffensen 1992) were determined using different methods, and so
could not be directly compared. Short-term exposure to severe hypoxia (6 hours, 18-20%
saturation) at 5°C resulted in a marked ventilatory response accompanied by metabolic acidosis
and a marked hematological response suggesting that catecholamines were being released
(Claireaux and Dutil 1992). Severe metabolic perturbations, particularly a sharp increase in
plasma lactate, were also observed and six out of 29 fish died during the period of observation
(Ibid.). Potentially lethal oxygen levels occur in the deep channels of the Gulf of St. Lawrence
(Figure 1), particularly in the estuary and west of Anticosti (Gilbert etal. 1997). D'Amours
(1993) examined dissolved oxygen and cumulative distribution of cod with depth in a stratified
random survey and observed fewer cod in hypoxic zones than would be expected from the extent
of such zones, suggesting cod avoided potentially lethal hypoxic conditions.
Non-lethal hypoxia depressed growth rate in cod and was not associated with metabolic
perturbations (Chabot and Dutil 1999). While cod increased in length and mass at all levels of
oxygen saturation during the growth experiment, growth was slowest at 45% saturation and
increased with oxygen availability (Figure 4). This relationship became non-significant beyond
65-75% oxygen saturation. Compared to size increases observed at higher levels of dissolved
oxygen, length increment was 35% less and weight increment was 52% less at 45% saturation.
As a result, the increase in condition factor was 59% less at 45% saturation. These results are
consistent with similar studies conducted on several other fish species (e.g. Secor and Gunderson
1998). During mild hypoxia (6 hours, 38-40% saturation), few metabolic perturbations occurred
in cod but a strong hyperventilatory response was observed (Claireaux and Dutil 1992). Cod
inhabiting waters deeper than approximately 150-200 m in the Gulf of St. Lawrence would thus
be expected to grow more slowly than cod living in shallower waters under similar food and
temperature conditions. A large proportion of the cod stock in the northern Gulf of St. Lawrence
appear to live in growth-limiting, low oxygen conditions during the late summer feeding period
(D'Amours 1993).
Growth rate depression was due to reduced food ingestion during hypoxia (Chabot and
Dutil 1999). Daily ingestion rate correlated with dissolved oxygen content, with 90% of the
variability among tanks explained by oxygen availability. Change in mass was, in turn, very
closely correlated with ingestion (Figure 5). Thus, oxygen availability limits growth through a
loss of appetite. High energy demands associated with post-prandial mechanisms (Soofiani and
Hawkins 1982) may trigger a negative feedback on physiological mechanisms controlling
appetite or behavioral processes associated with feeding activities. This finding is consistent
with field observations on stomach fullness for the July-October period (Figure 6). Stomachs
contained progressively less food, and an increasing proportion of empty stomachs was found
with increasing depth.
55
-------
ou
E 70
E.
^ 60
c 50
CD
E
§ 40
—
30
900
35800
' "'
^ 700
^
~ 600
| 500
CD
£ 400
c
300
200
0.3
^
.E
-£ 0.2
CD
E
CD
O
.E 0.1
n
: a _ .
: T
|
: 65%! T .,.-4-, T
X i ~ Hr" i """Sb
1 XT'
: f"
: c c c c "
BBS B
! A , A , , , ,
-b T
]
' 73%.- i-,,. j
~ T* — — T" 3 — """V.. — '
T .'' /' 1 "B
4/' \ ,/' "" \
ff* T " "~""""-~"~*---:IM •L
- lg*' ''" i
/' i
X
4^"
^ c c c c
BBS
-, .A. .,..,,,..,,,,,,,,,,..,,,..
c
I
T ,••'''
jsC.^_ T / -*• 1
T t-r^ """'--¥
xK i
T ff'r 1
1 X" "™
tf
i c c c c
BBS
A A
. . , , i , , , , i , , , , i , , , , i , , , , i , , , ,
^
O
08 CD
3
CD
0.6 jf.
^
3
0.4 &
10 -
q
CD
8 3
CD
«— 1-
6 ^
S
4 ****
0.004
D
Q]
0.003 =
o
03
0.002 ^
CD
5'
0.001 *
n
40 50 60 70 80 90 100
Dissolved oxygen (% saturation)
Figure 4. Increase in length (L), mass (M) and condition factor (K) in cod fed 3 times a
week over 12 weeks at 6 levels of dissolved oxygen and 10°C; letters indicate treatments that
were not statistically different (From Chabot and Dutil 1999).
56
-------
35
30
25
20
15
10
29
28
27
26
25
a
24
y = 23.204 + 0.067x FT= 0.92
40 50 60 70 80 90 100
Dissolved oxygen (% saturation)
Figure 5. Daily ingestion rate at different levels of hypoxia (a) and relationship between
daily change in mass and daily ingestion (b) in cod at 10°C (From Chabot and Dutil 1999).
I 1
70
60
50 5s
o>
40 B
o
30 3
20
10
0 50 100 150 200 250 300 350
Depth (m)
Figure 6. Index of stomach fullness (circles) and percent of empty stomachs (squares) of
cod as a function of depth. Samples were obtained during the summer fishing season and from
the annual research survey conducted by the Department of Fisheries and Oceans in the northern
Gulf of St. Lawrence (1993-1998).
57
-------
When given access to a larger amount of food through more frequent meals, cod exposed
to hypoxia still experienced a slower growth rate than cod reared in normoxia. Cod fed only
once a week had a very similar growth rate and final liver-somatic index under both hypoxia and
normoxia conditions (Figure 7). Cod fed three and seven times a week in normoxia grew faster
and had larger livers than cod fed once a week. In contrast, more frequent meals had no impact
on the specific growth rate and final liver-somatic index of cod in hypoxia. Cod reared in
normoxia versus hypoxia differed in specific growth rate (seven meals a week) and final liver-
somatic index (three and seven times a week). Again, differences among tanks in the daily
growth of cod were essentially explained by differences in daily ingestion (see Figure 8),
indicating that post-prandial mechanisms were not affected by low dissolved oxygen in the
environment. Digestion and subsequent anabolic processes require large amounts of ATP, and
hence create a great demand for dissolved oxygen. In cod, oxygen consumption increased
linearly with food intake with satiated, resting fish having a metabolic rate close to the maximum
rate observed in active fish (Soofiani and Hawkins 1982). Claireaux etal. (2000) suggested that
the observed decrease in food ingestion reported in Chabot and Dutil (1999) represented a
behavioural adaptation to the oxygen-mediated reduction in metabolic scope. Specifically,
perhaps fish having fed shortly before being exposed to hypoxic regurgitated ingested food when
oxygen the saturation was decreased and then swallowed any regurgitated food when oxygen
saturation was increased.
"S
p
.0
'o
2.
tn
X
90% saturation) and hypoxia (45% saturation)
at 10°C for a period of 46 days. Horizontal lines link
treatments that did not differ (normoxia: solid line; hypoxia:
dashed line). Stars indicate significant differences between
oxygen treatments, for given meal frequency.
U>
0
w o
0,
o,
• Hypoxia
A Normoxia
= -0.034 + 0.029x ?=0.98
10 15 20 25
Ingestion (g/day)
30
35
Figure 8. Relationship between daily change in mass and
daily ingestion for cod fed one, three, or seven times a week
in hypoxia and normoxia at 10°C over 46 days.
58
-------
When examining the areal extent of hypoxic zones and predicting the practical impact of
low oxygen levels on cod survival, thresholds may be considered, for example, one for growth
and one for mortality. Further studies on cod active metabolism under hypoxia may provide a
third threshold value relevant to our understanding of predator-prey relationships and migration
routes in and out of the Gulf of St. Lawrence. Cod from the northern Gulf may be less
productive than other stocks not only because they live in cold water (Brander 1995, Dutil et al.
1999), but also because deep waters in the northern Gulf are hypoxic (Gilbert et al. 1997) and
some segments of the cod stock are found in deep waters (D1 Amours 1993). Growth is a
significant determinant of cod surplus production in the northern Gulf (Dutil et al. 1999, Dutil et
al. 2000), and hence growth determinants such as oxygen availability must be considered and
threshold values precisely defined in order to improve our ability to forecast stock status.
REFERENCES
Brander, K. M. 1995. The effect of temperature on growth of Atlantic cod (Gadus morhud).
ICES Journal of Marine Science 52: 1-10.
Castonguay, M., C. Rollet, A. Frechet, P. Gagnon, D. Gilbert, and J.-C. Brethes. 1999.
Distribution changes of Atlantic cod (Gadus morhud) in the northern Gulf of St.
Lawrence in relation to an oceanic cooling. ICES Journal of Marine Science 56: 333-344.
Chabot, D. and J.-D. Dutil. 1999. Reduced growth of Atlantic cod in non-lethal hypoxic
conditions. Journal of Fish Biology 55: 472-491.
Claireaux, G. and J.-D. Dutil. 1992. Physiological response of the Atlantic cod (Gadus morhud)
to hypoxia at various environmental salinities. Journal of Experimental Biology 163: 97-
118.
Claireaux, G., D.M. Webber, J.-P. Lagardere, and S.R. Kerr. 2000. Influence of water
temperature and oxygenation on the aerobic metabolic scope of Atlantic cod (Gadus
morhud). Journal of Sea Research 44: 257-265.
D'Amours, D. 1993. The distribution of cod (Gadus morhud) in relation to temperature and
oxygen level in the Gulf of St. Lawrence. Fisheries Oceanography 2: 24-29.
Dutil, J.-D., M. Castonguay, D. Gilbert, and D. Gascon. 1999. Growth, condition, and
environmental relationships in Atlantic cod (Gadus morhud) in the northern Gulf of St.
Lawrence and implications for management strategies in the Northwest Atlantic.
Canadian Journal of Fisheries and Aquatic Science 56: 1818-1831.
59
-------
Dutil, J.-D., M. Castonguay, D. Gilbert, and D. Gascon. 2000. Production analyses for cold-
water and warm-water stocks and their use to predict surplus production. ICES CM
2000/C: 12. Report of the ICES/GLOBEC Workshop on the Dynamics of Growth in Cod:
Working Document 1.5, 5 p.
Gilbert, D., A.F. Vezina, B. Pettigrew, D.P. Swain, P.S. Galbraith, L. Devine, and N. Roy. 1997.
Etat du golfe du Saint-Laurent: conditions oceanographiques en 1995. Canadian
Technical Report of Hydrography and Ocean Sciences (ISSN: 0711-6772).
Hoar, W. S. and DJ. Randall. 1984. Gills: anatomy, gas transfer and acid-base regulation.
Academic Press. 456 p.
Plante, S., D. Chabot, and J.-D. Dutil. 1998. Hypoxia tolerance in Atlantic cod (Gadus
morhud). Journal of Fish Biology 53: 1342-1356.
Scholz, U. and U. Waller. 1992. The oxygen requirements of three fish species from the
German Bight: cod Gadus morhua., plaice Pleuronectesplatessa, and dab Limanda
limanda. Journal of Applied Ichthyology 8: 72-76.
Schurmann, H. and J. F. Steffensen. 1992. Lethal oxygen levels at different temperatures and
the preferred temperature during hypoxia of the Atlantic cod, Gadus morhua. Journal of
Fish Biology 41: 927-934.
Secor, D. H. and T.E. Gunderson. 1998. Effects of hypoxia and temperature on survival,
growth, and respiration of juvenile Atlantic sturgeon, Acipenser oxyrinchus. Fisheries
Bulletin 96: 603-613.
Soofiani, N. M. and A.D. Hawkins. 1982. Energetic costs at different levels of feeding in
juvenile cod, Gadus morhua. Journal of Fish Biology 21: 577-592.
Sundnes, G. 1957. On the transport of live cod and coalfish. Journal du Conseil international
pour 1' Exploration de la Mer 22: 191-196.
60
-------
SENSITIVITY OF STURGEONS TO ENVIRONMENTAL HYPOXIA: A REVIEW OF
PHYSIOLOGICAL AND ECOLOGICAL EVIDENCE
D.H. Secor1 and E.J. Niklitschek2
ABSTRACT
In this essay, three lines of evidence are developed that sturgeons in the Chesapeake Bay
and elsewhere are unusually sensitive to hypoxic conditions: 1. In comparison to other fishes,
sturgeons have a limited behavioral and physiological capacity to respond to hypoxia. Basal
metabolism, growth, feeding rate, and survival are sensitive to changes in oxygen level, which
may indicate a relatively poor ability of sturgeons to oxyregulate. 2. During summertime,
temperatures >20°C amplify the effect of hypoxia on sturgeons and other fishes due to a
temperature oxygen "squeeze" (Coutant 1987). In bottom waters, this interaction results in
substantial reduction of habitat; in dry years, sturgeon nursery habitats in the Chesapeake Bay
may be particularly reduced or even eliminated. 3. While evidence for population level effects
due to hypoxia is circumstantial, there are corresponding trends between the absence of Atlantic
sturgeon reproduction in estuaries like the Chesapeake Bay where summertime hypoxia
predominates on a system-wide scale. Also, the recent and dramatic recovery of shortnose
sturgeon in the Hudson River (4-fold increase in abundance from 1980 to 1995) may have been
stimulated by improvement of a large portion of the nursery habitat that was restored from
hypoxia to normoxia during the period 1973-1978.
INTRODUCTION
Sensitivity by sturgeons and other fishes to temperature, oxygen and their interaction is
evaluated experimentally through respirometry. As an example, the basal metabolism of the
stellate sturgeon Acipenser stellatus, measured over a range of temperatures and oxygen levels,
increases with temperature, but is only affected by oxygen at lower oxygen levels, above which
there is little response (Figure 1). The point of inflection in the curve of metabolic response to
dissolved oxygen (DO) is called the critical concentration. Oxygen levels below that point will
constrain metabolism, growth, swimming activity, and feeding rate (Fry 1971, Chiba 1988,
Kaufmann and Wieser 1992). As basal metabolism increases due to increased temperature, the
critical concentration becomes higher (Figures 1-3). In other words, as basic metabolic
requirements increase with temperature, so too will oxygen demand. This increase in oxygen
demand will outpace increased oxygen availability at higher temperatures due to decreased
oxygen solubility.
Chesapeake Biological Laboratory, University of Maryland Center for Environmental Science,
Solomons, MD 20688. 2Universidad Austral de Chile, Portales #73, Chile.
61
-------
Threshold BO
I Critical DO
I \
£ 70G
li e3C
I *nr
nar *J*J
g 34)(!
a
!>»'
100
o -
•D 1
si
st
D
5 S
1 9
Figure 1. Basal metabolic responses of young-of-the-year stellate sturgeon (A. stellatus). Data
from Winberg 1956, Figure 20. Critical DO is specified as the inflection point indicating
the point of metabolic responsiveness. Threshold DO indicates lethality where basal
metabolism can no longer be maintained.
Crtfca! OC -
Russian Kuf
:: Beluga, hybrid
#! Russiin s-Pjfqeo-n-2
OB
E
O
Q
Threshold DC -
i
Sly f
Russian tiJ
10 2R
TEMP
Figure 2. Critical and Threshold DO concentrations for Eurasian sturgeons (A. gueldemtaedtii, A.
stellatus, H. huso XA.ruthensus, A. baeris) as a function of temperature. Data from
Klyashtorin 1976, Figure 3 (DO levels were recalculated from partial pressures to
concentration). Critical DO is specified as the inflection point indicating the point of
metabolic responsiveness. Threshold DO indicates lethality where basal metabolism can
no longer be maintained. Russian sturgeon-1 and Russian sturgeon-2 designate two
separate experiments. All data are for young-of-the-year stage sturgeons.
62
-------
At very low oxygen concentrations, metabolism decreases rapidly to nil and the fish dies.
This is termed the threshold concentration (Figures 1-3). Both critical and threshold
concentrations were substantially higher for sturgeons in comparison to freshwater fishes (Figure
3). Critical concentrations for Eurasian sturgeons (A. guldenstddtii, A. baeri, A. stellatus and the
hybrid Huso huso x A. ruthensus) ranged between 25-60% saturation, increasing with
temperature. At 20°C, critical concentration was 3.6 mg/L (42% saturation) (Figures 2,3), which
was ca. 20% higher than that reported for rainbow trout. At 24°C, critical concentration was 4.5
mg/L (54% saturation) for the Eurasian sturgeons. The heightened sensitivity of metabolism to
oxygen levels may be characteristic of sturgeons, and has been ascribed to an inefficiently
functioning oxyregulatory system. Klyashtorin (1982) concluded that ancestral morphological
and physiological traits caused sturgeons to be less efficient in respiration than other fishes.
These traits included less efficient gill ventilation, low cardiac performance (Agnisola etal.
1999), and lower affinity of hemoglobin to oxygen.
5-
4- f--"'"1"' ' Critscal DO
,.-•" -
•-• "
J' >.• ^,,.- " ' ' ''•
DO
—"• -
0 10 20 30
TEMP
Figure 3. Critical and Threshold DO concentrations for various fish species. Data from
Klyashtorin 1976, Figure 6 (DO levels were recalculated from partial pressures to
concentration). Critical DO is specified as the inflection point indicating the point of
metabolic responsiveness. Threshold DO indicates lethality where basal metabolism can
no longer be maintained. Data for Eurasian sturgeons represents a pooled mean response.
All data are for young-of-the-year stage sturgeons, trout and carp.
In experiments on Atlantic and shortnose sturgeons A. oxyrinchus and A. brevirostrum,
Niklitschek (2001) observed substantial reductions in routine metabolism, consumption, feeding
metabolism, growth, and survival at 40% vs. 70% DO (Figures 4-6). Again the effect of oxygen
level was conditional on temperature. In comparison to normoxia at 20°C, 40% DO saturation
63
-------
(3.3 mg/L) yielded a 33% and 30% reduction in growth, a 29% and 27% reduction in
consumption, and a 23% and 17% reduction in routine metabolism, for Atlantic and shortnose
sturgeons, respectively. At 27°C and 40% saturation (2.9 mg/L), there was 77% and 69%
reduction in growth, a 38% and 45% reduction in consumption, and a 28% and 21% reduction in
routine metabolism for Atlantic and shortnose sturgeon, respectively. Because routine
metabolism was investigated rather than basal metabolism, estimates of critical DO
concentrations were not possible. In a separate laboratory study, Secor and Gunderson (1998)
reported 2- and 4-fold reductions in growth rate due to hypoxia at 26 and 19°C, respectively.
Atlantic
fcr •
i r -
Shortnose
" ;:
Oxpfisn S-rfl-iiiiflinii f<
-t a: .c
Figure 4. Effect of DO and temperature on consumption by Atlantic and shortnose
sturgeon young-of-the-year. 10-d laboratory experiments conducted by
Niklitschek(2001).
64
-------
Attefe
Figure 5. Effect of DO and temperature on growth of Atlantic and shortnose
sturgeon young-of-the-year. 10-d laboratory experiments conducted by
Niklitschek(2001).
All-antic
I
Dissolved oiyg*!* saturation |%|
Figure 6. Effect of DO and temperature on long-term survival (20-45 d trials) of Atlantic
and shortnose sturgeon young-of-the-year. Laboratory experiments conducted by
Niklitschek (2001). Bars represent standard errors.
65
-------
Few studies have addressed lethal effects of hypoxia on sturgeons (Table 1). Jenkins et al.
(1994) observed 86 - 100% mortality for 25-64 day old fish in an acute 6 hr exposure to 2.5 mg/L
DO at 22.5°C (30% saturation). Older juveniles (100-310 days old) experienced 12-20 %
mortality under the same conditions. Short-term exposure to 3.0 mg/L (35% saturation) resulted
in 18-38% mortality for juveniles 20-77 days in age. No mortality was observed for exposures to
> 3.5 mg/L (42% saturation). Long term exposure (10 days) of Atlantic sturgeon young-of-the-
year juveniles (150-200 days old) to hypoxia at 26°C (37%-44% DO saturation; 2.8 - 3.3 mg/L)
resulted in complete mortality over the ten day period in three of four replicates (Secor and
Gunderson 1998). The fourth replicate experienced 50% mortality. At 20°C and hypoxia at
27%-37% saturation (2.3 - 3.2 mg/L), the latter researchers observed 12-25% mortality. No
mortality was observed for normoxic treatments. In preliminary experiments, Niklitschek (2001)
observed lethality for young-of-the-year Atlantic and shortnose sturgeon reared at 30% oxygen
saturation (27°C; 2.35 mg/L) for 24 hours. Thus, he selected 40% saturation to observe sub-lethal
physiological effects and avoid loss and suffering of experimental animals. At this "sub-lethal"
level (2.8 mg/L at 28°C; 3.3 mg/L at 20°C), daily survival rates were significantly reduced in
comparison to survival under normoxia (Figure 6). Threshold oxygen levels (levels below which
basal metabolism cannot be sustained) reported for Eurasian sturgeons occurred at 1.7-2.0 mg/L
(25% saturation) for temperatures 20-24°C (Figure 2; Klyashtorin 1975). At 20°C, threshold
oxygen concentrations were about 20% higher for rainbow trout than for Eurasian sturgeons, but
at temperatures less than 18°C, threshold values were similar between the sturgeon and rainbow
trout (Figure 3).
In a direct comparison between juvenile white sturgeon and striped bass, the bass growth
rate was substantially more depressed under hypoxia than that for the white sturgeon A
transmontanus (Cech et al. 1984). Sturgeon growth rates under hypoxia (90 torr ~ 4.5 mg/L) and
temperatures 20-25°C were 20-25% lower than those under normoxia (130 torr ~ 7 mg/L).
Striped bass reared under identical conditions experienced an additional 7-9% growth rate
depression under hypoxia. During the 34-day long experiment, higher mortalities of both species
occurred under relatively mild hypoxia in comparison to normoxic treatments.
Beyond metabolic response, sturgeons undertake other physiological and behavioral
responses to hypoxia. Niklitschek (2001) observed that egestion levels for Atlantic and shortnose
sturgeon juveniles increased significantly under hypoxia, indicating that consumed food was
incompletely digested. This response could serve as a useful means to shunt energy towards
respiration and other life support functions. Behavioral studies indicate that Atlantic sturgeon and
shortnose sturgeon are quite sensitive to ambient conditions of oxygen and temperature. In a
series of choice experiments, juvenile sturgeons consistently selected normoxic over hypoxic
conditions (Figure 7; Niklitschek 2001). On the other hand, larval stage Siberian sturgeon A
baeri did not actively avoid or disperse from experimental hypoxic conditions (2-3 mg/L at 19°C;
Khakimullin 1988). Beyond escape or avoidance, sturgeons respond to hypoxia (< 40%
saturation) through increased ventilation, increased surfacing (to ventilate relatively oxygen-rich
surficial water), and decreased swimming and routine metabolism (Nonnette et al. 1993, Croker
and Cech 1997, Secor and Gunderson 1998, Niklitschek 2001). Historically, sturgeons were at
66
-------
Table 1. Results of lethality experiments on shortnose sturgeon and Atlantic sturgeon.
Species and test
temperature
Shortnose, 22.5°C
Atlantic, 26°C
Atlantic, 20°C
Atlantic, 20-26°C
(a) Jenkins et al. 1993
Age (days) D.O. (mg/L)
46
110
25
32
64
104
20
39
77
103
19
90
150-200
150-200
150-200
(b) Secorand
2.0
2.0
2.5
2.5
2.5
2.5
3.0
3.0
3.0
3.0
3.5
3.5
2.8-3.3
2.3-3.2
>5.0
Gunderson
Survival Reference and comments
rate (%)
8 Ref. (a) Fig. 2, mean of 6 replicates, 6-hour test
78
0
o
J
14
88
62
82
67
100
78
100
12.5 Ref. (b) Mean of 4 replicates, 10-day test,
4 tanks open to air
78 Ref. (b) Mean of 4 replicates, 10-day test,
2 tanks were sealed and 2 tanks were open to air
100 Ref. (b)
1998
"ti:rI«M»i»
Figure 7. Selection of hypoxia (40% saturation) over normoxia (70 or 100% saturation) by Atlantic and
shortnose sturgeon young-of-the-year based on experiments conducted by Niklitschek (2001).
Positive % of choices indicates selection of higher DO, negative % indicates selection of lower
DO level.
67
-------
the center of a debate on whether or not fishes could oxyconform to hypoxia: that is, could fish
tolerate declines in dissolved oxygen delivery to their tissues? (Burggren and Randall 1978,
Klyashtorin 1982, Nonnette et al. 1993). Oxyregulation by the typical vertebrate mechanism of
anaerobic metabolism has since been demonstrated for sturgeons (Nonnette etal. 1993,
McKenzie et al. 1995), providing evidence against physiological oxyconformity. Still, behavioral
means to oxyregulate in sturgeons may be relatively more important since critical concentrations
are higher for sturgeons than other teleosts (Klyashtorin 1982). As evidence of this, white
sturgeon juveniles were 3- to 9-fold less active under hypoxia (80 mm Hg ~ 5 mg/L or 50%
saturation at 16°C) than under normoxia (Croker and Cech 1997). Secor and Gunderson (1998)
observed that lethality to hypoxia increased when surfacing behavior was blocked.
In summary, sturgeons endemic to the Chesapeake Bay, and sturgeons in general, are
unusually sensitive to hypoxia in terms of their metabolic and behavioral responses. The critical
concentration at which sturgeons metabolically respond to dissolved oxygen is higher or similar
to that of rainbow trout. Bioenergetic and behavioral responses indicate that young-of-the-year
juveniles (-30 to 200 days old) will experience lost production in those habitats with less than
60% oxygen saturation. For summertime temperatures (22-27°C), this level corresponds to
dissolved oxygen concentrations of 4.3 - 4.7 mg/L DO. Acute and chronic lethal effects for
shortnose and Atlantic sturgeon were observed at levels of 3.3 mg/L at summertime temperatures.
Threshold concentrations for Eurasian sturgeons were somewhat lower, ranging between 1.7 and
2.0 mg/L for temperatures 20-24°C.
A Habitat Squeeze for Chesapeake Bay Sturgeons?
Coutant and Benson (1990) proposed that habitats of Chesapeake Bay striped bass were
severely curtailed during summer months due to thermal preference compounded by hypoxia.
During summer, striped bass adults select narrow lenses within the water column characterized by
< 25°C water and normoxia (Cheek et al. 1985). In many instances, it was expected that such
summertime refugia would curtail prey availability and growth. Coutant's concept, known
popularly as the "habitat squeeze," was a prevailing hypothesis to describe the decline of striped
bass during the 1980s. Using bioenergetic models (Hartman and Brandt 1995), Brandt and Kirsch
(1993) mapped habitat suitability for striped bass in the Chesapeake Bay, and predicted
substantial habitat restrictions in summer, leading to negative or static growth during the summer
months.
Sturgeons are particularly vulnerable to a habitat squeeze (i.e., synergism between
temperature and dissolved oxygen effects on habitat availability) due to their demersal lifestyle
and unique bioenergetic responses to hypoxia. While sturgeons do occasionally surface, they
depend almost exclusively on benthic substrates and bottom waters for spawning, feeding,
migration, and refuge from predation or stressful environments (e.g. flow and temperature
refugia). Their specialized underslung jaw, diets, electrosensory and olfactory systems, poor
vision, body form, and heterocercal tail are but a few features that attest to their demersal lifestyle
(Burggren 1978, Bemis andKynard 1997, Carlson and Simpson 1987, Haley 1999, Secor et al.
2000). Atlantic sturgeons are known to occur at depths between 1 m to greater than 25 m;
shortnose sturgeons have been observed between 1 and 12 m (Kieffer and Kynard 1997, Savoy
68
-------
and Shake 2000, Welsh et al. 2000). In the Chesapeake Bay during winter, Atlantic sturgeon
select deeper habitats occurring in the deep channel (Secor et al. 2000, Welsh et al. 2000). Thus,
sturgeons are not limited by bathymetry within the Bay and would be expected to utilize sub-
pychnocline waters contingent upon water quality.
In dissertation research, Niklitschek (2001) developed spatially explicit bioenergetic
models for Chesapeake Bay sturgeons to evaluate the influences of water quality on sturgeon
habitats. This is appropriate for sturgeons because opportunities to directly observe habitat use is
curtailed given their threatened or extirpated status. Habitat value was assigned based upon
expected growth and survival responses, predicted by a bioenergetics model filtered through
Chesapeake Bay Program Monitoring data (1990-1999). Potential production rates for young-of-
the-year Atlantic and shortnose sturgeons in the Chesapeake Bay reflected the strong seasonal
cycle in potential production driven by the interaction of temperature, salinity and dissolved
oxygen. Summer represented the most critical season in which hypoxia and high temperature
conditions caused severe habitat fragmentation for both species, restricting suitable habitat to a
small fraction of the Bay (Figure 8). Negative potential production areas closely mirrored
hypoxic regions occurring in the middle mainstem, as well as the lower Patuxent, Potomac and
Rappahannock rivers, major tributaries on the western shore of the Chesapeake Bay. Negative
production areas were also predicted near the Bay mouth where very high salinities masked the
otherwise improved conditions of dissolved oxygen and temperature. As a result, null or positive
production was restricted to very limited areas that coincide only in part between the two sturgeon
species.
For Atlantic sturgeon, summer refuges for an average year were restricted to the upper
Bay between the Magothy River and the Susquehanna Flats. Slightly negative areas were
expected around Fishing Bay-Nanticoke River, between the Severn and Choptank rivers and in
the upper Potomac River. For shortnose sturgeon, most of the suitable habitat in the mainstem
would be restricted to the Bay head above the Sassafras River. Other areas of positive production
included the upper sections of the Potomac and James rivers, as well as most of the Nanticoke
River. The total area supporting positive production (suitable habitat) under average July
conditions corresponded to 1,586 and 1,076 km2, for Atlantic and shortnose sturgeons,
respectively. These surface areas represent only 8.5% and 5.8% of the total surface area of the
Bay mainstem and tidal sections of its tributaries, respectively. The best summer condition for
both sturgeons was predicted for 1996, where suitable habitat reached circa 4,200 km2 for
Atlantic sturgeon and 2,050 km2 for shortnose sturgeon. In 1996, average temperature and
salinity were the second lowest in the study period and the average dissolved oxygen conditions
were above the study period average. This year also exhibited the highest July freshwater inflow
of the study period. The worst conditions for Atlantic sturgeon were observed in July 1999,
where suitable habitat was down to about 1 km2. Also in 1999, overall salinity was the highest in
the study period time series. This higher than usual salinity reduced the habitat value of the upper
Bay section, which is typically the most productive section of the Chesapeake Bay due to its
favorable temperature and oxygen conditions.
69
-------
O t/3
3- ^3
3 fa
O ~-
w fa
o f^
tr OQ
to ^
§1
^ OQ
2,
X ^
ft,
01
fa
s^S
X
ft
a
B
OQ O
vo
vo
O
VO
VO
VO
>
r-K
fa"
r-1-
o'
fa
3
e
•*t
so
70
-------
In summary, due to avoidance, sub-lethal, or lethal effects sturgeon summertime habitats
are expected to be restricted in comparison to historical times when hypoxia was less prevalent
(Officer et al. 1984, Cooper et al. 1991). The habitat squeeze phenomenon is particularly relevant
to sturgeons due to the synergism of temperature, dissolved oxygen and salinity effects during
their first year of life. In dry years, the interaction of high salinity, warm temperatures, and
hypoxic waters in summer severely reduces sturgeon habitats and in some years can virtually
eliminate all habitats for Atlantic sturgeon. The fragmented distribution and small volume of
productive habitats imposed by summertime hypoxia represents a substantial hurdle to overcome
in the restoration of Chesapeake Bay sturgeons.
Chesapeake Bay Sturgeons: Where Are They Now?
Shortnose sturgeon is in all probability extirpated from the Chesapeake Bay. Shortnose
sturgeon recently observed in the region of Susquehanna flats and in the Potomac River have been
confirmed to be immigrants from the Delaware Bay population, presumably having migrated
through the C&D Canal (Grunwald et al. in review). The status of Atlantic sturgeon in the
Chesapeake Bay is less certain (Grogan and Boreman 1998). There has been no evidence of
reproduction in the Maryland portion of the Chesapeake Bay for over 25 years (Secor 1995).
Recently, evidence of reproduction (capture of young-of-the-year Atlantic sturgeon) occurred for
the York and Rappahannock Rivers (NMFS 1998). Based upon historical catch data, Secor
(2000) estimated that 19th century adults may have numbered > 100,000 in the Chesapeake Bay.
Their numbers were decimated by over-exploitation at the turn of the 19th century. During the
past 100 years, there has been no evidence of recovery.
Secor and Gunderson (1998) hypothesized that due to their unusual sensitivity to hypoxia,
Atlantic sturgeon may have been extirpated from Maryland Bay waters due, in part, to the
increased prevalence of hypoxia in the 20th century. Atlantic sturgeon populations recovered in
South Carolina estuaries and in the Hudson River following relief from the intensive 19th century
fishing (Smith et al. 1985, NMFS 1998), but not in the Chesapeake Bay. Secor (2000) recorded
that cycles of overfishing and recovery for this species occur over a 50-year period. Prevalence of
hypoxia increased in critical habitats of sturgeons due to post-WWII agricultural practices
(synthetic fertilizers) and residential development (increased wastewater). These factors may
have curtailed any 20th century cycle of recovery within the Chesapeake Bay. There are also
other important factors that may have prevented recovery, including lost spawning grounds due to
siltation and reductions in abundance due to incidental catch (there has been no directed fishery
for sturgeons in over a century in the Chesapeake Bay). Circumstantial evidence for a
relationship between increased hypoxia and loss of sturgeons this past century includes the
absence of sturgeons in systems that are prevalently hypoxic in summer (Collins et al. 2000).
71
-------
1Jic_jyybam
9 ALR.iNV
JP ATHENS
5 0
.**- IN
. J&-' CE
• . '•-
K.m
Figure 9. Region of persistent hypoxia known as the Albany Pool in the Hudson River. This
region overlaps broadly with shortnose sturgeon nursery habitat that occurs throughout
the freshwater tidal reach of the estuary.
The recent recovery of shortnose sturgeon abundance in the Hudson River is also
consistent with the idea that hypoxia may be important in controlling sturgeon populations. Bain
et al. (2000) has compared mark-recapture population estimates for the period 1980 - 1995.
During this period, sub-adult and adult shortnose sturgeon increased from about 13,000 to 57,000
- a four-fold increase in abundance! This yields a population growth rate of 10% per year that is
remarkably high for a long-lived, late maturing species such as shortnose sturgeon. Prior to 1974,
a pervasive hypoxic/anoxic summertime region known as the "Albany Pool" (Figure 9)
overlapped approximately 40% of the expected nursery habitat for shortnose sturgeon (i.e. 40% of
the tidal freshwater area of the Bay). Levels of pervasive hypoxia there (<30% saturation) would
have been lethal to shortnose sturgeon juveniles (Figures 10, 11). Few fish, sturgeon or
otherwise, were documented in the in the 60 km river stretch of the "Albany Pool" during July -
October (Leslie et al. 1988). In 1974, >80% of the region's wastewater began to receive
secondary and tertiary treatment, and in less than two years the system recovered fully to
normoxia (Figures 10, 11). Subsequent monitoring data has revealed a dramatic faunal recovery
in the number offish species returning to the Albany Pool region (Leslie et al. 1988). In 1983, a
relatively strong year-class (high abundance of juveniles) was reported by two independent
studies (Carlson and Simpson 1987, Dovel et al. 1992). Such year-classes may have stimulated
recovery of Hudson River shortnose sturgeon during the past 20 years.
72
-------
In summary, absence or diminished populations of sturgeons correspond to systems where
summertime hypoxia is prevalent. Recovery of normoxia to the Hudson River estuary
corresponded with a remarkable recovery of shortnose sturgeon there. These case studies provide
circumstantial evidence that summertime hypoxia might substantially diminish population
recovery or perhaps even lead to extirpation. Sturgeons represent the only resource species in the
Chesapeake Bay that is threatened or extinct. Future recovery of sturgeons coincident with
improved water quality in the Chesapeake Bay would be a quite a success story.
i
*«-i
PJ
E
>««•*
O
O
Albany (river km 227)
ID
8
Yew
0
Jun Jyl Sep Oct
Month
Figure 10. DO levels for an "Albany Pool" station during summer and fall months.
73
-------
8
eh
O
Q
m
6
SO 120 150
km
180
210
240
Figure 11. DO levels throughout the Hudson River estuary before (1967) and after (1978) system
recovery from hypoxia. The "Albany Pool" region upriver is shown in gray, the NY City
effect area in gray at the mouth.
REFERENCES
Agnisola, C., D.J. McKenzie, D. Pellegrino, P. Bronzi, B. Tota, and E.W. Taylor. 1999.
Cardiovascular responses to hypoxia in the Adriatic sturgeon (Acipenser naccarii). Journal
of Applied Ichthyology 15: 67-72.
Bain, N.B., N. Haley, D.L. Peterson, K.K. Arend, K.E. Mills, and P.J. Sullivan. 2000. Shortnose
sturgeon of the Hudson River: an Endangered Species recovery story. Biology,
Management and Protection of Sturgeon Symposium Pre-Print. EPRI. Palo Alto,
California, USA.
Barrowman, T.M. and M.L. O'Shea. 1996. Long-term improvements in water quality due to
sewage abatement in the Lower Hudson River. Estuaries 19: 890-900.
Bemis, W.E. and B. Kynard. 1997. Sturgeon rivers: an introduction to acipenseriform
biogeography and life history. Environmental Biology of Fishes 46: 167-183.
74
-------
Boyle, R.H. 1969. The Hudson River; a natural and unnatural history. Norton, New York.
pp. 304.
Brandt, S.B. and J. Kirsch. 1993. Spatially explicit models of striped bass growth potential in
Chesapeake Bay. Transactions of American Fisheries Society 122: 845-869.
Burggren, W.W. and DJ. Randall. 1978. Oxygen uptake and transport during hypoxic exposure
in the sturgeon Acipenser transmontanus. Respiration Physiology 34: 171-183.
Carlson, D.M. and K.W. Simpson. 1987. Gut contents of juvenile shortnose sturgeon in the
upper Hudson Estuary. Copeia 1987: 796-802.
Cheek, T.E., MJ. Van Den Avyle, and C.C. Coutant. 1985. Influence of water quality on
distribution of striped bass in a Tennessee River impoundment. Transactions of American
Fisheries Society 114: 67-76.
Chiba, K. 1988. The effect of dissolved oxygen on the growth of young striped bass. Nippon
Suisan Gakkaishi 54: 599-606.
Claireaux, G. and J. Lagardere P. 1999. Influence of temperature, oxygen and salinity on the
metabolism of the European sea bass. Journal of Sea Research 42: 157-168.
Collins, M.R. S.G. Rogers, T.I.J. Smith, and M.L. Moser. 2000. Primary factors affecting
sturgeon populations in the southeastern United States. Fishing mortality and degradation
of essential fish habitats. Bulletin of Marine Science 66: 917-928.
Cooper, S.R. and G.S. Brush. 1993. A 2,500-year history of anoxia and eutrophication in
Chesapeake Bay. Estuaries 16: 617-626.
Coutant, C.C. 1987. Thermal preference: when does an asset become a liability. Environmental
Biology of Fishes 18: 161-172.
Coutant, C.C. and D.L. Benson. 1990. Summer decline in habitat suitability for striped bass in
Chesapeake Bay: reflections on a population decline. Transactions of American Fisheries
Society 119:757-778.
Crocker, C.E. and JJ. Chech. 1997. Effects of environmental hypoxia on oxygen consumption
rate and swimming activity in juvenile white sturgeon, Acipenser transmontanus, in
relation to temperature and life intervals. Environmental Biology of Fishes 50: 382-389.
Croker, C.E. and JJ. Cech. 1996. The effects of hypercapnia on the growth of juvenile white
sturgeon, Acipener transmontanus. Aquaculture 147: 293-299.
75
-------
Croker, C.E. and JJ. Cech. 1997. Effects environmental hypoxia on oxygen consumption rate and
swimming activity in juvenile white sturgeon, Acipemer transmontamis, in relation to
temperature and life intervals. Environmental Biology of Fishes 50: 383-389.
Dadswell, M. J. 1979. Biology and population characteristics of the shortnose sturgeon,
Acipemer brevirostrum Le Seuer 1818 (Osteichthyes: Acipseridae), in the Saint John
Estuary, New Brunswick, Canada. Canadian Journal of Zoology 57:2186-2210.
Dovel, W.L. and T.J. Berggren. 1983. Atlantic sturgeon of the Hudson Estuary, New York. New
York Fish and Game Journal 30: 140-172.
Dovel, W.L., A.W. Pekovitch, and T.J. Berggren. 1992. Biology of the shortnose sturgeon
(Acipenser brevirostrum LeSueur, 1818) in the Hudson River Estuary, New York,. In
Estuarine Research in the 1980s. C.L. Smith (ed.). State University New York, Stony
Brook, New York. pp. 187-227.
Fry, F. E. J. 1971. The effect of environmental factors on the physiology offish. In: Fish
Physiology.Volume VI. W. S. Hoar and D. J. Randall (eds.). Academic Press, New York.
pp. 1-98.
Grogan, C. S., and J. Boreman. 1998. Determining the probability that historical populations of
fish species are extirpated. North. American Journal of Fish Management 18:522-529.
Gross, MR, J. Repka, C. Robertson, D.H. Secor, and W. Van Winkle.2000. Conserving ancient
fish: life history and demography of sturgeon. American Fisheries Society Symposium
28:13-30.
Grunwald, C., J. Stabile, J. R. Waldman, R. Gross, and 1.1. Wirgin. In review. Population
genetics of shortnose sturgeon, Acipenser brevirostrum, based on sequencing of the
mitochondrial DNA control region. Molecular Ecology 11(10):1885-1898.
Haley, N.J. 1999. Habitat characteristics and resource use patterns of sympatric sturgeons in the
Hudson River Estuary. M.S. Thesis, University of Massachusetts, Amherst,
Massachusetts. 124 pp.
Hartman, K.J. and S.B. Brandt. 1995. Comparative energetics and the development of
bioenergetics models for sympatric estuarine piscivores. Canadian Journal of Fisheries and
Aquatic Science 52: 1647-1666.
Jenkins, W.E., T.I.J. Smith, L.D. Heyward and David M. Knott. 1995. Tolerance of shortnose
sturgeon, Acipenser brevirostrum, juveniles to different salinity and dissolved oxygen
concentrations. Proceedings Southeastern Association of Fish and Wildlife Agencies
47:476-484.
76
-------
Hildebrand, S. F., and W. C. Schroeder. 1927. Fishes of the Chesapeake Bay. U.S. Bureau of
Fisheries, Washington, D.C. 388 pp.
Kaufmann, R. and W. Wieser. 1992. Influence of temperature and ambient oxygen on the
swimming energetics of cyprinid larvae and juveniles. Environmental Biology of Fishes
33: 87-95.
Khakimullin, A.A. 1987. Oxygen preferendum (preferred oxygen regime) of hatchery young of
the Siberian sturgeon, Acipenser baeri. Voprosy Ikhtiologii 1987: 693-696.
Kieffer, M. C., and B. Kynard. 1993. Annual movements of shortnose and Atlantic sturgeon in
the Merrimack River, Massachusetts. Transactions of American Fisheries Society
122:1088-1103.
Klyashtorin, L.B. 1982. The ability of sturgeons (Acipenseridae) to regulate gas exchange.
Ichthyology : 141-144.
Klyashtorin, L.B. 1976. The sensitivity of young sturgeons to oxygen deficiency. Journal of
Ichthyology 16: 677-681.
Leslie, J.A., K.A. Abood, E.A. Maikish, and PJ. Keeser. 1988. Recent dissolved oxygen trends
in the Hudson River,] In: Fisheries Research in the Hudson River. C.L. Smith (ed). State
University of New York, Albany, p. 287-303
McKenzie, D.J., J.F. Steffensen, C.L. Bolis, P. Bronzi, and E.W. Taylor. 1995. Effects of diet on
spontaneous locomotor activity and oxygen consumption in Adriatic sturgeon Acipenser
naccarii. Fish Physiology and Biochemistry 14: 341-355.
National Marine Fisheries Service. 1998. Status review of Atlantic sturgeon (Acipenser
oxyrinchus oxyrinchus). NMFS, Gloucester, Massachusetts, USA. 124 pp.
Niklitschek, E. J. 2001. Bioenergetics Modeling and Assessment of Suitable Habitat for Juvenile
Atlantic and Shortnose Sturgeons in the Chesapeake Bay. Ph.D. thesis, University of
Maryland, College Park, Maryland, USA.
Nonnotte, G., V. Maxime, J.P. Truchot, P. Williot, and C. Peyraud. 1993. Respiratory responses
to progressive ambient hypoxia in the sturgeon, Acipenser baeri. Respiration Physiology
91:71-82.
Officer, C. B., R. B. Biggs, J. L. Taft, L. E. Cronin, M. A. Tyler, and W. R. Boynton. 1984.
Chesapeake Bay anoxia: Origin, development, and significance. Science (Washington,
D.C). 223:22-27.
Paerl, H.W., J.L. Pinckney, J.M. Fear, and B.L. Peierls. 1998. Ecosystem responses to internal
and watershed organic matter loading: consequences for hypoxia in the eutrophying
Neuse River Estuary, North Carolina, USA. Marine Ecology Progress Series 166: 17-25.
77
-------
Petersen, J.K. and L. Pihl. 1995. Responses to hypoxia of plaice, Pleuronectesplatessa and dab,
Limanda limanda, in the southeast Kattegat: distribution and growth. Environmental
Biology of Fishes 43: 311-321.
Ruer, P.M., JJ. Cech, and S.I. Doroshov. 1987. Routine metabolism of the white sturgeon,
Acipemer trammontanus: Effect of population density and hypoxia. Aquaculture 62:
45-52.
Sauriau, P.-G., J.-P. Robin, and J. Marchand. 1993. Les Mortalites estivales du muletZ/'za
ramada en loire: dysfunctionnement cyclique du role de voie de migration d'un estuarie.
Bulletin of Ecology 24: 87-91.
Savoy, T. and D. Shake. 2000. Atlantic sturgeon, Acipenser oxyrinchus., movements and
important habitats in Connecticut waters. Biology, Management and Protection of
Sturgeon Symposium Pre-Print. EPRI. Palo Alto, California, USA.
Secor, D.H. 2000. Atlantic sturgeon fisheries and stock abundance during the late 19th Century.
American Fisheries Society Symposium. 28:89-98.
Secor, D.H., E. Niklitschek, J. T. Stevenson, T. E. Gunderson, S.Minkkinen, B. Florence, M.
Mangold, J. Skjeveland and A.Henderson-Arzapalo. 2000. Dispersal and growth of
yearling Atlantic sturgeon Acipenser oxyrinchus released into the Chesapeake Bay.
Fisheries Bulletin 98(4): 800-810.
Secor, D. H. and J. R. Waldman. 1999. Historical abundance of Delaware Bay Atlantic sturgeon
and potential rate of recovery. In: Life in the slow lane: ecology and conservation of long-
lived marine animals. J.A. Musick (ed). American Fisheries Society Symposium 23: 203-
216.
Secor, D.H. and T.E. Gunderson. 1998. Effects of hypoxia and temperature on survival, growth,
and respiration of juvenile Atlantic sturgeon, Acipenser oxyrinchus. Fisheries Bulletin. 96:
603-613.
Secor, D. H. 1995. Chesapeake Bay Atlantic sturgeon: Current status and future recovery.
Chesapeake Biological Laboratory, Solomons, Maryland, USA. 11 pp.
Smith, T.I.J. 1985. The fishery, biology, and management of Atlantic sturgeon, Acipenser
oxyrhynchus, in North America. Environmental Biology of Fishes 14: 61-72.
Welsh, S.A., S.E. Eyler, M.F. Mangold, and AJ. Spells. 2000. Capture locations and growth
rates of Atlantic sturgeon in the Chesapeake Bay. American Fisheries Society Symposium
28:183-194.
Winberg, G.G. 1960. New Information on Metabolic Rate in Fishes. Fisheries Research Board of
Canada, Nanaimo, British Columbia, Canada. 234 pp.
78
-------
HYPOXIA AND ANOXIA IN SMALL TEMPERATE ESTUARIES:
PATTERNS OF OXYGEN DEFICIENCY, EFFECTS, AND RECOVERY
Javier Franco1, Elizabeth Aspillaga1, Ifiigo Muxika1,
Victor Perez1, Oihana Solaun1, and Angel Borja1
ABSTRACT
The estuaries of the Basque Country (Northern Spain), which are drowned river valleys,
are small in size and highly affected by river discharges and tides. They show large differences
in geomorphology, hydrographs, and anthropogenic impacts. In recent years, sewerage schemes
have been carried out in some basins and, for the rest, they are under construction or being
planned. In this work, we present the main patterns of dissolved oxygen content variability
(both, spatially and temporally) in these estuaries, with a special emphasis on the conditions that
lead to hypoxic and anoxic situations. The effects of low oxygen concentrations on various
ecological communities will be analysed using the data from monitoring programs in two
systems highly affected for many years by urban and industrial wastewaters. These discharges
produced hypoxic and anoxic conditions that were reflected in the benthic communities, with
very impoverished fauna in the systems and absence of organisms near the discharge outlets.
Sewerage schemes for these estuaries are being developed. Implementation of these schemes is
improving the environmental quality of some estuaries, both from the point of view of water
quality (e.g. higher oxygen concentrations), and the partial recovery of benthic communities
(with the presence of fauna in reaches previously azoic).
INTRODUCTION
Oxygen is a key element in the metabolic processes of fishes and invertebrates. Oxygen
deficiency is one of the most important stress factors of aquatic organisms (Llanso 1992) and it is
perhaps the most widespread anthropogenically induced deleterious effect in the marine
environment that causes localized mortality of benthic macrofauna (Diaz and Rosenberg 1995).
Dissolved oxygen in coastal marine environments has changed drastically in the last several
decades, and hypoxia — defined as occurring when oxygen concentration declines below 2 mg
O2/L (Dauer et al. 1992) or 2 ml O2/L (Diaz and Rosenberg 1995) — is now a common estuarine
phenomenon and is occurring more frequently in many marine ecosystems.
Hypoxia can cause adverse effects on benthic macrofauna at several levels. It causes
mortality, changes in behaviour, reductions in growth, decreases in biomass and diversity, and
changes in the abundance and species composition of benthic assemblages (Diaz and Rosenberg
1995). Tolerance to hypoxia varies among phyla, order, and species, and is also dependent on
Department of Oceanography and Marine Environment, AZTI Foundation (Technological Institute for Food and
Fisheries) Herrera Kaia - Portu Aldea z/g 20110 - Pasaia (Gipuzkoa) - Spain.
79
-------
the life history characteristics of infaunal species. In estuaries, hypoxia usually increases the
abundance of small, short-lived species and decreases the abundance of large, long-lived species
(Holland et al. 1987, Llanso 1992). As a consequence, in hypoxic areas large infauna usually are
present at lower diversity, abundance and biomass, while small infauna usually occur at higher
density and biomass (Dauer etal. 1992, Diaz and Rosenberg 1995).
The effects of hypoxia on benthic fauna depend to a great extent on the intensity and
duration of hypoxic episodes. Few benthic invertebrates can survive a complete lack of oxygen;
long hypoxic episodes have more severe consequences than short episodes and extended periods
of hypoxia will kill even the most tolerant species (Diaz and Rosenberg 1995).
Recovery of benthic communities following hypoxic events depends primarily on the
interaction between the complexity of the community affected and the severity and duration of
hypoxia which, in their turn, are dependent on hydrodynamic and mixing processes (Diaz and
Rosenberg 1995). Initial recovery of benthic communities after hypoxia usually follows the
Pearson and Rosenberg (1978) enrichment-disturbance model. Recovery of an affected
community to a "mature" community following the return to normoxic conditions can take
several years. There is no large system that has fully recovered after development of persistent
hypoxia or anoxia. The only exceptions may be relatively small systems where point source
discharges have ceased and recovery was initiated from surrounding non-affected areas
(Rosenberg 1976).
Dissolved oxygen concentrations in temperate estuaries are affected by several physical,
chemical, and biological processes that together can produce great dissolved oxygen variability
at different spatial and temporal scales (Litaker et al. 1987, Kenney et al. 1988, Litaker et al.
1993). The seasonal cycle of dissolved oxygen usually presents a sharp decline in summer from
the effects of high respiration and mineralization rates, higher stratification of the water column
(lower diffusion to bottom waters), low river flows (higher retention times), and lower oxygen
solubility at higher temperatures (Taft et al. 1980, Kemp et al. 1992). In fact, hypoxia is a
common phenomenon in many temperate systems.
The estuaries of the Basque Country (northern Spain) are small in size and highly
affected by river discharges and tides. They exhibit great differences in geomorphology,
hydrography and anthropogenic impacts. The upper reaches of some of these systems can be
considered stressed environments. Wastewaters are discharged into these areas with high
nutrient and suspended solid loads resulting in high chlorophyll concentrations in spring and
summer and low oxygen concentrations. Benthic communities are very impoverished in some of
these systems. In recent years, sewer-wastewater treatment systems have been installed in some
basins and, for the rest, they are under construction or being planned.
The Estuaries of the Basque Country
The estuaries of the Basque Country are drowned river valleys formed during the
Flandrian transgression (approx. 5,000 years ago). They are located in the Cantabrian coast, in
the Bay of Biscay (north of Spain; Figure 1). The rivers that form these systems are short
because the main mountains in the area are close to and parallel to the coast in an E-W
80
-------
Figure 1. Map of the Basque Coast and location of the estuaries.
direction. Annual rainfall in the area varies between 1,000 and more than 2,500 mm with an
increasing gradient from west to east.
Tides are semidiurnal, with two highs and two lows each day. Tidal amplitudes vary
between 1.5 m on neap tides and 4.5 m on spring tides.
Estuary lengths vary between 2 km in the Lea estuary to 22 km in the Nervion estuary
(Table 1). These are shallow systems and intertidal areas are very important in some of them.
Although all these systems are very influenced by river discharges and tides, the relative
importance of these two factors is very different among them. The relative importance of river
flow (based on the estuary volume/river flow ratio) is highest in the Deba estuary and lowest in
the Nervion estuary (Table 1).
There are important human settlements in the river courses (especially in the lower
reaches) and estuarine areas, explaining why anthropogenic impacts on these systems is high.
The most populated and industrialized areas are those of the Nervion (the most industrialized and
populated area of the Atlantic Spanish coast) and Oiartzun estuaries. In fact, these two systems
(and, especially, the first one) are strongly polluted in their upper reaches (Borja et al. 2001).
The Main Patterns of Dissolved Oxygen in Medium-Polluted Temperate Systems
In temperate estuaries, dissolved oxygen usually presents its annual minimum from mid-
spring to mid-fall, which is related to the higher rate of breakdown of organic matter at higher
temperature, the lower solubility of oxygen with increasing temperature, and the longer retention
times associated with the lower river flow discharges. This is also the pattern in the
81
-------
Table 1. Main geomorphological and hydrological features of the Basque estuaries. See Figure
1 for location of the estuaries. Data from Villate et al. (1989), Garcia de Bikufia and Docampo
(1990), and Diputacion Foral de Gipuzkoa (2000).
Estuary
Barbadian
Nervion
Butron
Oka
Lea
Artibai
Deba
Urola
Oria
Urumea
Oizrtzun
Bidasoa
Basin
area
(km2)
127
1755
174
178
84
101
534
364
888
279
87
700
River
flow
(m3/second)
2.9
35.6
4.7
3.6
1.8
2.5
14.0
8.0
25.7
16.5
4.8
28.7
Estuary
length
(km)
4.4
22.0
8.0
12.5
2.0
3.5
5.5
5.7
11.1
7.7
5.5
11.1
Estuary
depth
(m)
0-5
0-30
0-10
0-10
0-5
0-10
0-5
0-10
0-10
0-10
0-20
0-10
Estuary
volume
(m3xl06)
-
200
0.69
3.29
-
-
0.35
-
2.10
-
-
7.05
Estuary
volume /
River flow (days)
-
65.0
1.7
10.6
-
-
0.3
-
0.9
-
-
2.8
estuaries of the Basque Country. As an example, in the Urumea estuary minimum annual
oxygen concentrations are normally reached in summer or in early fall (Figure 2). Values lower
than 2 mg/L were recorded in 1988, 1989, 1990, and 1994, and no values lower than 4 mg/L
have been recorded since then. Maximum annual concentrations are normally reached in winter.
The annual ranges of concentration have been lower in the last 6 years compared to the
beginning of the period, which is more related to the increase in annual minima than to the
decrease in annual maxima. Two main factors seem to be responsible for these trends. On the
one hand, the first years of this period were very dry with rainfall in summer being the minimum
values of the period (Figure 2a). On the other hand, in the middle of the 1990s a great amount of
wastewater that had been directly discharged to the estuary was diverted to an outfall located on
the coastline. This reduced the organic and nutrient loads to the estuary, acting as a positive
factor for water oxygenation. Mean seasonal dissolved oxygen concentrations decrease seawards
(Figure 2b), which is explained by the fact that the areas that still receive wastewater loads are
the middle and the outer reaches of the estuaries.
82
-------
Figure 2. Main patterns of dissolved oxygen variability in the Urumea estuary, (a) Dissolved
oxygen concentrations from 1988 to 2000. Mean (all the stations), minimum and
maximum values are shown. Total rainfall in summer (July, August and September) is
also presented, (b) Mean (1988-2000) seasonal concentrations at each sampling station.
Location of sampling stations in the estuary is shown. Data provided by the Diputacion
Foral de Gipuzkoa from a monitoring programme since 1988.
In the Oka estuary, which is a 12.5-km long system, hypoxia develops in spring and
summer (Figure 3). Saturation percentages lower than 30% are present in the inner estuary in the
lower layers of the water column. This area receives the discharges of an out-of-date sewage
treatment plant (STP) and organic matter and nutrient concentrations are very high, transparency
is very low, and chlorophyll concentrations at the surface can reach 100 mg/L in spring and
summer (Franco 1994). So, large amounts of organic matter, both from waste loads and from in
situ primary production, are available for bacteria. Short-term oxygen variability can be very
high, with concentrations much higher in the evening than in the early morning due to the day-
night cycles of photosynthesis and respiration (Franco etal. 1996).
In these small estuaries, high river flows have great effects on the hydrological, physico-
chemical, and biological features. Under these conditions, water masses are largely advected
seawards and these systems can be almost completely occupied with fresh or low-salinity waters.
As these "new" waters are normally well oxygenated, the overall effect of elevated river flows
consists of a general reoxygenation of these systems (Figure 3).
83
-------
-4 r
Figure 3. Spatial (horizontal and vertical) distribution of dissolved oxygen in the Oka estuary
under normal and high river discharge conditions. Location of sampling stations in the
estuary is shown.
Dissolved Oxygen and Benthic Communities in Heavily Polluted Systems
The Nervion estuary (Figure 1) is a 22-km long system located in the conurbation area of
Bilbao, the most populated and industrialized area of the Atlantic Spanish coast. The estuary has
two main zones (Figure 4): an inner channel of about 15 km length with depths between 5 and 10
m, which is strongly stratified; and an outer area, with depths from 10 to 30 m, which is slightly
stratified and comprises more than 90% of the total volume (Urrutia 1986).
This estuary has received large amounts of wastes, both from urban and industrial origin,
during several decades. This has led to a very deteriorated system with very low oxygen
concentrations and the absence of fauna along the main part of the system. In 1979 the Sewerage
Scheme for the area was approved. The scheme comprises more than 170 km of pipes and two
STP, with a total investment of about 600 million USD and approximately 1,000,000 inhabitants
connected into it. Construction will be completed by 2004-2005. Physico-chemical primary
treatment has been operating since 1984 and biological treatment will come into operation in
2001. The main STP, the Galindo plant, discharges treated waters into the middle reach of the
estuary. The main objective of the Scheme is the recovery of the biota along the system, and a
water quality standard of 60% oxygen saturation was established. Since 1990, the Consorcio de
Aguas Bilbao Bizkaia carries out monthly or bimonthly surveys of the water quality in the
estuary. In addition, three surveys per year are carried-out on zooplankton communities, and one
survey per year on sediments, benthic communities, and demersal fauna.
84
-------
Dissolved oxygen in the Nervion estuary presents a clear seasonal pattern, with large
differences along the main axis of the system and between surface and bottom waters (Figure 4).
Minimum dissolved oxygen concentrations occur in summer. In summer, hypoxia is normally
observed in the inner estuary in bottom waters, and almost in the entire inner channel in the
surface waters. Bottom waters present an increasing gradient of dissolved oxygen seaward
during fall, winter, and spring, while minimum concentrations are measured in the middle
estuary at the water surface. This spatial pattern is explained by several factors: (a) the higher
retention times of bottom waters in the inner estuary, which move landward and seaward along
the main axis of the system without going out; (b) the presence of sediments very enriched in
organic matter (normally between 5 and 20%) and hence with a high dissolved oxygen demand;
(c) the slow diffusion of oxygen to bottom waters due to the large stratification; and (d)
especially, the discharge of the waters treated in the Galindo STP.
D
H
X
o
Distance from tfo? tidal lirnit (kna)
2.1
River Nervion
Figure 4. Seasonal distribution of dissolved oxygen in the Nervion estuary. Mean values (1989-
2000) in surface and bottom water are presented. Location of sampling stations in the
estuary is shown.
Considering all the data from these surveys to date, more than half of the oxygen values
are lower than 60% saturation, and about a quarter of them are lower than 20% saturation. In
recent years, a slight improvement of the oxygen content in the water has been observed, with
increasing mean annual concentrations and decreasing percentages of samples with hypoxia.
Nevertheless, the water quality standard (60% oxygen saturation) is still far from being reached.
In benthic communities there is a general increasing trend in the number of taxa seaward
with maximum abundance in the boundary between the inner channel and the outer area (Figure
5a). The two most upriver stations are very impoverished and fauna has only been found in 5 of
the last 12 years (mainly in the last 5 years), although the number of taxa has never been higher
85
-------
than four. Oxygen is the key environmental factor to explain this distribution (Gonzalez-Oreja
and Saiz-Salinas 1998). Bottom waters in the inner estuary are hypoxic most of the time,
especially in spring and summer. The presence of fauna becomes permanent from about 10 km
from the tidal limit seaward; this is also the zone in which bottom waters are rarely hypoxic
(Figure 5b). In this area, the number of taxa has varied between two and 17 during the last 12
90
80-
70-
80-
50
40
30
20
10
0-
1DDODD
10000
1000
100
10
1
D
100
80 -
60 -
40 -
20 -
0
Sea
RUtr NerutJi
NUMBER OFTAXA
a
5 101 102 106 110 111
ABUNDANCE
6 5 101 102 106 110 111
STATIONS
Oxygen: percentage of values < 20 It sat.
Stations 8
101
••»•••• ALL DATA
-*— SPRING- SUIWMER
0 5 10 15
DISTANCE FROM THE TIDAL LIMIT (km)
20
Figure 5. Benthic communities distribution and hypoxia in the Nervion estuary, (a) Mean (1989-2000),
minimum and maximum values of number of taxa and abundance (number of individuals per m2,
logarithmic scale) in each sampling station are presented. Location of sampling stations in the
estuary is shown, (b) Percentage of samples with oxygen saturation lower than 20% in bottom
water for all the data, and for only spring and summer data. Correspondence between distance
from the tidal limit and location of some sampling stations is also presented.
86
-------
years, with a mean value of about seven; the total abundance has varied between 30 and 26,600
individuals per m2. These benthic communities are dominated by annelids. The opportunistic
and moderately-hypoxia resistant polychete Capitella capitata is normally the dominant species,
and other common taxa are the hypoxia resistantMalacocerosfuliginosus., the oligochetes, and
the gastropod Hydrobia ulvae.
Along with the slight improvement in water column oxygenation, benthic communities
have shown some signs of recovery and colonization in the middle and upper reaches (Franco et
al. 200la).
The importance of oxygen as a limiting factor for the establishment of healthy benthic
communities in this estuary has been shown in laboratory bioassay experiments (Saiz-Salinas et
al. 1996, Saiz-Salinas and Gonzalez-Oreja 2000). When juveniles of the polychete Hediste
diversicolor (a common and characteristic species of muddy sediments in temperate estuaries)
were placed both in sediments obtained from the Nervion estuary (very polluted) and in control
sediments from the Butron estuary (Figure 1) and located in a well oxygenated environment, no
differences in survivorship among control and polluted sediments were found; indeed, growth
was higher in the polluted sediments, probably because of their higher organic content.
However, similar experiments carried out with the bivalve Scrobiculariaplana resulted in
significantly lower survivorship in the chemically contaminated sediments from the Nervion
estuary. These results reflect differences between these two species in their sensitivity to oxygen
deficiency and chemical contamination. It can be stated that oxygen is a main factor for the
recovery of the fauna, although sediment contamination could delay this process once dissolved
oxygen conditions improve.
Based on the observed relationship between abiotic variables and benthic community
structure, a future scenario has been proposed for this system once dissolved oxygen reaches the
water quality standard established in the Sewerage Scheme (Gonzalez-Oreja and Saiz-Salinas
1998, Saiz-Salinas and Gonzalez-Oreja 2000). According to this scenario, macrozoobenthic
biomass will increase in most of the estuary.
The Oiartzun estuary is located in the eastern zone of the Basque coast (Figure 1). Its
fluvial system drains an area of approximately 87 km2, and the mean flow to the estuary is
approximately 4.8 m3/second. It is a small (5 km long) estuarine system but its depth is
relatively great (> 10 m in most of the system), partially because of continuous dredging to
facilitate navigation in its harbour. Until 1996, the estuary received the direct discharges of
several municipalities, both through the river loads and directly to the estuary. Since, the
particular geomorphological and hydrological features of this system produce very limited
renovation rates, the area showed great environmental deterioration, with very polluted water and
sediments and a complete lack of fauna in the upper reaches (Zaballa 1985, Borja et al. 1996).
In the summer of 1996, and as a transitory solution in the context of the Sewerage
Scheme for the area, many of the discharges were diverted to an outfall located in the shore at
about 1.5 km to the west of the mouth of the Oiartzun estuary. Since then, more discharges have
87
-------
been diverted to the outfall. The Sewerage Scheme will eventually consist of a STP and
submerged outfall that will discharge the treated water into the Bay of Biscay approximately 1.5
km from the coast at about 50 m depth. The Diputacion Foral de Gipuzkoa has conducted
several field surveys during the last few years in order to look for changes in the environmental
quality of the estuary and adjacent coastal area (Franco et al. 200 Ib). This public institution has
conducted a regular monitoring program in all the estuaries of the province since 1989.
Before discharges were diverted to the outfall, oxygen concentrations in the water
column were very low, with hypoxia (< 2 ml/L) usually present in the uppermost stations and
nearly anoxic conditions in summer (Table 2, Figure 6). Related to this, benthic communities in
the upper reaches of the Oiartzun estuary were absent or very impoverished, both in terms of
number of taxa and biomass. In the rest of the estuary the number of taxa increased seaward.
The opportunistic polychetes, Capitella capitata and Pseudopolydorapaucibranchiata, were the
dominant species in the upper and lower estuary, respectively (Borja et al. 1996). The latter
species reached very high densities (more than 9,500 individuals per m2 in the middle of the
estuary).
Table 2. Dissolved oxygen and benthic communities in the Oiartzun estuary before and after the
waste diversion at each sampling station.
Before diversion1
Station Dissolved No. of Dominant species
oxygen taxa
1 2.35(0.13) 0
no fauna
4.18 (2.91) 5 Capitella capitata
5.19(4.57) 25 P. paucibranchiata
4 2.55(0.31) 0
no fauna
6.75 (6.36) 29 P. paucibranchiata
Dissolved No. of
oxygen taxa
After diversion2
Dominant species
M. fuliginosus (1997, 1998),
y ' A. prismatica (1999), C. capitata (2000)
6 05 18 5 C8 97^1 P' ligni (1997)' P' Paucibranchiata (1998)-
( ' A. prismatica (1999), C. capitata (2000)
P. paucibranchiata (1997, 1998, 1999),
t°y -m^ou) c capitata (2000)
49* \n ^ C' CaPitata (1997' 1999' 2000)'
y ' M. fuliginosus (1998)
P. paucibranchiata (1997),
7.62 31 (26-35) C. capitata (1998, 2000),
Nematodes (1999)
^ean dissolved oxygen concentrations (mg/L) at each sampling station considering data at several depths from 31
surveys between 1989 and 1996. Values in parenthesis mean summer bottom values; Number of taxa
detected at each sampling station in 1995; Dominant species in terms of abundance.
2Mean dissolved oxygen concentrations (mg/L) at each sampling station considering data at several depths from 12
surveys between 1996 and 2000; Mean and range of number of taxa detected at each sampling station
between 1997 and 2000 (summer surveys); Dominant species in terms of abundance.
-------
Sea
River
Oiartzun
Figure 6. Location of sampling stations in the Oiartzun estuary.
Similar to the Nervion estuary, dissolved oxygen is the main limiting environmental
factor in the health of the benthic communities in the Oiartzun estuary. This latter system does
not contain any intertidal areas, so benthic communities are always dependent on oxygen
conditions in the overlaying waters.
Partial Recovery Processes
The Oiartzun estuary can serve as an example of how benthic communities respond to
changes in water column oxygenation conditions. As has been noted, in the summer of 1996
most wastewater discharges were diverted out of the estuary. Since then dissolved oxygen has
dramatically increased in the estuary (Figure 7), both in surface and bottom waters, and
concentrations after the diversion are statistically significantly higher (p <0.01; Mann-Whitney U
test) than before the diversion (Table 2).
Coincident with the oxygen level increases, the benthic communities have improved:
fauna is present throughout the system, a rise in the number of taxa has been observed in the
estuary (Table 2, Figure 8), some groups absent before the diversion are now present, and the
community seems to be more equilibrated in terms of relative contribution to the total number of
taxa by the main groups, i.e., annelids, arthropods, and molluscs (Figure 8).
Nevertheless, since 1996 benthic communities in the entire Oiartzun estuary system have
been dominated by opportunistic and hypoxia resistant species like the polychetes, Capitella
capitata, Malacocerosfuliginosus, Pseudopolydorapaucibranchiata, and Polydora ligni, and the
bivalve Abraprismatica. In 2000, Capitella capitata was the dominant species in the entire
estuary in terms of density (Table 2). Therefore, although benthic communities show signs of
some degree of recovery, the system is still clearly impacted.
89
-------
140
120
100
80
eo
40
INNER ESTUARY
F-9J S-35 F-9S
LU 120
o
> 100 J
0 80 H
80
40 •
20 •
0
F-96 S-ST F-3? S-98 F-98 S-93 F-93 SO]
•4---SUHF.
-• IOTTOU
MIDDLE ESTUARY
F-9S S-96 F-96
F-9T
DATE
F-9S S-99 F-9§ S-IB
Figure 7. Dissolved oxygen (% sat.) history in the Oiartzun estuary. Mean low-high tide values
in the surface and bottom waters from fall 1994 to summer 2000 in the upper and in the
middle estuary are presented. The approximate date of the waste diversion is indicated.
Figure 8. Benthic communities history in the Oiartzun estuary. Relative contribution of the
main groups to the total number of taxa. The total number of taxa detected each year is
also shown. The approximate date of the waste diversion is indicated.
90
-------
CONCLUSIONS
Oxygen is a key environmental factor for maintenance of benthic communities in
temperate estuaries. In small estuarine systems, dissolved oxygen in the water column varies at
several spatial and temporal scales. Interannual and seasonal variability are mainly explained by
river flow variations and the seasonal cycle of insolation and temperature. Short-time variability
is mainly dependent on sporadic high river discharges and biological processes (photosynthesis
and respiration). Nevertheless, these general patterns are greatly affected by anthropogenic loads
of nutrients and organic matter (eutrophication) and in some systems oxygen variability is more
associated with these anthropogenic disturbances than with natural factors. Benthic communities
clearly reflect the degree of human impact on these systems. In very polluted estuaries (in terms
of organic matter enrichment, both in situ or allochtonous) benthic communities are absent or
very impoverished, and dominance by opportunistic and/or resistant species is observed in
stressed environments. Oxygen deficiency plays an important role in these processes.
Implementing sewage treatment works in some of these estuary systems clearly demonstrated
that water column oxygenation improvement is followed by the recolonisation or enhancement
of benthic communities in previously azoic or very impoverished areas. Nevertheless, these
general recovery scenarios could be delayed by the presence of contaminants in the sediments
once water column oxygenation improves.
ACKNOWLEDGMENTS
The Departamento de Obras Hidraulicas y Urbanismo de la Diputacion Foral de
Gipuzkoa provided data from the Urumea and Oiartzun estuaries. The Consorcio de Aguas
Bilbao Bizkaia provided data from the Nervion estuary. The Departamento de Ordenacion del
Territorio, Vivienda y Medioambiente del Gobierno Vasco provided data from the Oiartzun
estuary. The Institute Nacional de Meteorologia provided meteorological information. Benthic
communities were mainly analysed by S. C. de Investigation Submarina (INSUB).
REFERENCES
Borja, A., R. Castro, and A. Uriarte. 1996. Estudio de la calidad biologica de los sedimentos de
los estuarios interiores de los rios Bidasoa, Oiartzun y Urumea. Azti for Departamento
de Obras Hidraulicas y Urbanismo de la Diputacion Foral de Gipuzkoa. pp. 105.
Borja, A., J. Franco, MJ. Belzunce, and V. Valencia. 2001. Red de vigilancia y control de la
calidad de las agues litorales del Pais Vasco: otofio 1999 - verano 2000. Azti-Labein for
Departamento de Ordenacion del Territorio, Vivienda y Medio Ambiente, Gobierno
Vasco. pp. 430.
Dauer, D.M., AJ. Rodi, Jr., and J.A. Ranasinghe. 1992. Effects of low dissolved oxygen events
on the macrobenthos of the lower Chesapeake Bay. Estuaries 15: 384-391.
91
-------
Diaz, R.J., and R. Rosenberg. 1995. Marine benthic hypoxia: a review of its ecological effects
and the behavioural responses of benthic macrofauna. Oceanography and Marine
Biology: an Annual Review 33: 245-303.
Diputacion Foral de Gipuzkoa, 2000. Los rios de Gipuzkoa. Estaciones de aforo y calidad.
pp. 72.
Franco, J. 1994. Variabilidad espacio-temporal de la biomasa y production del fitoplancton el
estuario de Urdaibai. Tesis doctoral. Universidad del Pais Vasco. pp.201.
Franco, J., A. Ruiz, and E. Orive. 1996. Factores y escalas de variabilidad del oxigeno disuelto
en el estuario de Urdaibai. Ozeanografika 1: 43-64.
Franco, J., A. Borja, R. Castro, O. Solaun, MJ. Belzunce, V. Perez, and F. Villate. 2001a.
Seguimiento ambiental de los estuarios del Nervion, Barbadun y Butron durante 2000.
Azti for Consorcio de Aguas Bilbao Bizkaia. pp. 273.
Franco, J., A. Borja, O. Solaun, MJ. Belzunce, J. Bald, and V. Valencia. 2001b. Campafia de
medicion de variables biologicas y fisico-guimicas en el estuario del rio Oiartzun y area
costera proxima a cala Murgita. Azti for Departamento de Obras Hidraulicas y
Urbanismo de la Diputacion Foral de Guipuzkoa. pp. 197.
Garcia de Bikufia, B., and L. Docampo. 1990. Limnologia de los rios de Vizcaya. Teorias,
aplicaciones e implicaciones biologicas. Dpto. Urbanismo, Vivienda y Medio Ambiente,
Gobierno Vasco. pp. 200.
Gonzalez-Oreja, J.A., and J.I. Saiz-Salinas. 1998. Exploring the relationships between abiotic
variables and benthic community structure in a polluted estuarine system. Water
Research 32: 3799-3807.
Holland, A.F., A.T. Shaughnessy, and M.H. Hiegel. 1987. Long-term variation in mesohaline
Chesapeake Bay macrobenthos: spatial and temporal patterns. Estuaries 10: 370-378.
Kemp, W.M., P.A. Sampou, J. Garber, J. Turtle, and W.R. Boynton. 1992. Seasonal depletion
of oxygen from bottom waters of Chesapeake Bay: roles of benthic and planktonic
respiration and physical exchange processes. Marine Ecology Progress Series 85: 137-
152.
Kenney, B.E., W. Litaker, C.S. Duke, and J. Ramus. 1988. Community oxygen metabolism in a
shallow tidal estuary. Estuarine, Coastal and Shelf Science 27: 33-43.
Litaker, W., C.S. Duke, B.E. Kenney, and J. Ramus. 1987. Short-term environmental variability
and phytoplankton abundance in a shallow tidal estuary. I. Winter and summer. Marine
Biology 96: 115-121.
92
-------
Litaker, W., C.S. Duke, B.E. Kenney, and J. Ramus. 1993. Short-term environmental variability
and phytoplankton abundance in a shallow tidal estuary. II. Spring and fall. Marine
Ecology Progress Series 94: 141-154
Llanso, RJ. 1992. Effects of hypoxia on estuarine benthos: The lower Rappahannock River
(Chesapeake Bay), a case study. Estuarine, Coastal and Shelf Science 35: 491-515.
Pearson, T.H., and R. Rosenberg. 1978. Macrobenthic succession in relation to organic
enrichment and pollution of the marine environment. Oceanography and Marine
Biology: an Annual Review 16:229-311.
Rosenberg, R. 1976. Benthic faunal dynamics during succesion following pollution abatement
in a Swedish estuary. Oikos 27:414-427
Saiz-Salinas, J.I., G. Frances, and X. Imaz. 1996. Uso de bioindicadores en la evaluation de la
contamination de la Ria de Bilbao. Servicio editorial, Universidad del Pais Vasco.
pp. 128.
Saiz-Salinas, J.I., and J.A. Gonzalez-Oreja. 2000. Stress in estuarine communities: lessons from
the highly-impacted Bilbao estuary (Spain). Journal of Aquatic Ecosystem Stress and
Recovery 7: 43-55.
Taft, J.L., W.R. Taylor, E.O. Hartwig, and R. Loftus. 1980. Seasonal oxygen depletion in
Chesapeake Bay. Estuaries 3: 242-247.
Urrutia, J. 1986. Estudio de la estructura y funcionamiento del estuario del Nervion en relation
a la dinamica del fitoplancton. Tesis Doctoral. Universidad del Pais Vasco. pp. 279.
Villate, F., J. Franco, A. Ruiz, and E. Orive. 1989. Caracterizacion geomorfologica e
hidrologica de cinco sistemas estuaricos del Pais Vasco. Kobie XVIII: 157-170.
Zaballa, K. 1985. Estudio de las Taxocenosis Anelidianas en las Rias de Guipuzcoa: Bidasoa
(Fuenterrabia), Pasajes, Urumea (San Sebastian), Oria y Urola (Zumaya). Tesis de
Licenciatura. Universidad del Pais Vasco.
93
-------
94
-------
THE EFFECTS OF SEASONALLY SEVERE HYPOXIA
ON CONTINENTAL SHELF FAUNA
Nancy N. Rabalais1
ABSTRACT
Severe hypoxia occurs over broad areas of the Louisiana shelf most summers, coincident
with the habitat of commercially important species, such as penaeid shrimp (Farfantepenaeus
aztecus and Litopenacus setiferus). Significant decreases in species richness, abundance and
biomass of benthic organisms occur under severe hypoxia/anoxia. Effects at episodically-
hypoxic sites are less severe or negligible, but masked by high variability consistent with pulses
of biological productivity on a river-influenced shelf. Short-lived hypoxic episodes may not
lessen habitat value for fisheries species in other systems and in fact may facilitate predation
upon the benthos. As oxygen concentration falls on the Louisiana shelf, fishes move away from
the area before the benthos are stressed, in contrast to other areas of the coastal ocean. If the
duration or severity of hypoxia worsens, community composition and trophic interactions are
likely to shift and affect energy transfer. Areas on the periphery of severe hypoxia and spring
recruitment in hypoxic areas are characterized by opportunistic species. While biomass may be
periodically high or turnover of opportunists may provide suitable prey, the overall productivity
of the benthic system, transfer to other trophic levels, and secondary production, including
fisheries, in general are not known.
INTRODUCTION
Bottom-water hypoxia (< 2 mg O2/L) is often a secondary response of an estuarine or
coastal system to eutrophication. The dissolved oxygen conditions of many major coastal
ecosystems around the world have been affected adversely by the process of eutrophication
(Diaz and Rosenberg 1995). The zone of bottom-water hypoxia on the northern Gulf of Mexico
continental shelf west of the Mississippi River delta is one of the largest zones in the world's
coastal ocean, exceeded only by the coastal areas of the Baltic (84,000 km2; Rosenberg 1985)
and the northwestern shelf of the Black Sea (20,000 km2; Tolmazin, 1985). From 1993 to 1997,
the size of the Gulf of Mexico hypoxic zone was consistently greater than 16,000 km2 in mid-
summer, and reached 20,000 km2 in mid-summer of 1999 (Rabalais etal. 1999, Rabalais and
Turner 2001). On the southeastern Louisiana shelf (Figure 1), critically depressed dissolved
oxygen concentrations (< 2 mg/L) occur below the pycnocline from as early as late February
through early October and nearly continuously from mid-May through mid-September.
We previously documented the variability of two benthic communities over two annual
cycles of oxygen stress and recovery (Rabalais etal. 1993, 1995). In this summary, we discuss
Louisiana Universities Marine Consortium, Chauvin, Louisiana.
95
-------
the data from a single year for the two stations. Continuous recordings of bottom-water
dissolved oxygen concentrations characterized the two study areas as one with seasonally severe
and persistent hypoxia and the second with aperiodic or moderate hypoxia (Rabalais et al. 1994).
This paper documents the differences in the seasonal decline of benthos—specifically
abundance, species richness, and assemblage composition—within two differing hypoxia
regimes on the southeastern Louisiana shelf. We also consider the implications of hypoxia-
stressed benthic communities to trophic interactions that might ultimately affect fisheries
resources.
STUDY AREAS AND OXYGEN ENVIRONMENT
Sites adjacent to two oil production platforms (WD32E and ST53A) in two oil fields
(Figure 1) were sampled in April and June-October 1990 for benthic communities and
supplementary hydrographic and sedimentary data. The West Delta site (Station WD32E) was
closer to the Mississippi River delta in a sedimentary regime that was predominantly silty
sediments (85 to 90%) with some clay and sand. Sediments at Station ST53 A, 60 km farther to
the west, were predominantly sandy silts with little clay fraction. Sediment total organic carbon
(TOC) values were consistently low across the study area, typically less than 1.0% (Rabalais et
al. 1993, 1995). Sediment chlorophyll a and phaeopigment concentrations were consistent with
season and distance from the Mississippi River (Rabalais et al. 1992). Petroleum related
hydrocarbons and trace metals were mostly at background levels for benthic stations reported
here (Rabalais et al. 1993).
30.0
29.5H
29.0
28.5H
>75%
>50%
>25%
<25%
50km
93.5
92.5
91.5
90.5
89.5
Figure 1. Frequency of occurrence of bottom-water hypoxia (dissolved oxygen < 2
mg/L) in mid-summer of 1985-1999. Identified are study areas ST53A (=ST),
WD32 (=WD).
Bottom waters at Station ST53 A were severely depleted in dissolved oxygen and often
anoxic for most of the continuous record from mid-June through mid-August and for much of the
month of September in 1990 (Figure 2). Hydrogen sulfide was detected in bottom-water samples
on several occasions in June, July and August 1990. In contrast, hypoxia occurred at Station
WD32E for only 50% of the total record, hypoxic events were shorter in duration than at Station
ST53 A, and there was a strong diurnal pattern in the oxygen time series (Rabalais et al. 1994).
The record of dissolved oxygen at Station WD32E was most coherent with the diurnal bottom
96
-------
water pressure signal, which suggested the importance of tidal advection in the variability of that
oxygen record. Wind-induced mixing was insufficient to aerate the water column prior to the
outbreak of cold air fronts in late September and early October at which time a relaxation in the
stratification also occurred due to thermal cooling. Lack of strong winds and changes in bottom-
water temperature suggested that reoxygenation (at Station ST53 A in late August and at Station
WD32E for most of the record) resulted from lateral advection.
8
§ 5
O)
O 3
0> 2 -
.<2 0
Q
ST53A
O)
CD
5
X
O
1
O
<" n
(/) 0
Jun
I
WD32E
no data
Aug
I
Sep
Jun
Jul
Aug
Sep
Figure 2. Time series plots of near-bottom dissolved oxygen concentration at Stations
ST53A and WD32E in 1990 (modified from Rabalais et al. 1994). Arrows
indicate dates of benthos collections; April not shown.
BENTHIC COMMUNITIES
There were statistically significant differences between study sites with regard to both
number of species and number of individuals, across all months (April 1990-October 1990) and
any single month (Rabalais et al. 200la). In July, there was a general seasonal decline in
populations at both sites, but the decline was much more precipitous at Station ST53 A, the
severely hypoxic station, than at Station WD32E with aperiodic hypoxia. The decline in
populations at Station WD32E continued into September and October, but the benthic
community at Station ST53 A showed a slight recovery during that period.
97
-------
Station WD32E
Species richness was similar in April and June 1990, decreased in July and August, then
decreased further in September and October (Figure 3). Peak abundance occurred in June 1990
followed by a mid-summer and fall decline. Polychetes were a large component of the benthic
community at Station WD32E, but other major taxonomic groups exceeded the polychetes in
April 1990 and August 1990 and were half the total in June 1990. Polychetes dominated in July,
September and October of 1990. The benthic community at Station WD32E was diverse, with a
complement of pericaridean crustaceans, bivalves, gastropods and other taxa. Dominant species
for most months were Paraprionospiopinnata and Mediomastus ambiseta. The abundance of
Armandia maculata increased in August 1990. Changes in several dominant species through
1990 were evident including Prionospio cristata, Nephtys incisa, Magelona sp. I, Magelona sp.
H, Ampharete sp. A and Oweniafusiformis.
1990 Taxonomic Composition of West Delta 32E
.2-
°3
o>
a
in
<*H
o
•—
0)
.0
I
• Other
CH Sipuncula
• Ophiuroidea
• Crustacea
• Gastropoda
E3 Bivalvia
Nemertea
Polychaeta
10
Mediomastus ambiseta
Paraprionospio pinnata
Ampharete sp. A
Magelona sp. H
Magelona sp. I
Tellina versicolor
Armandia maculata
6 7
Month
10
Figure 3. 1990 Taxonomic composition of West Delta 32E. Number of species within
taxonomic groups (total for ten 0.02-m2 cores) and mean number of
individuals/m2 (n = 10) at WD32E for months indicated in 1990 (modified from
Rabalais et al. 1995).
98
-------
Station ST53A
Species richness was lowest in August and September 1990, and was also low in July and
October 1990 (Figure 4). Species richness in June and April 1990 was similar and
approximately six times greater than in July through October 1990. Abundance of individuals
was high in April 1990, higher in June 1990; then the number of individuals was low from July
through September. There was a slight recovery of individuals in October. While polychetes
comprised most of the species at Station ST53A, composition by other major taxonomic groups
was fairly high in April (13 taxa) and June (11 taxa) of 1990, then reduced to four to six major
taxa in July through October 1990. The polychetes Ampharete sp. A, Paraprionospiopinnata
and Mediomastus ambiseta were common in spring and early summer of 1990. As hypoxia
worsened, the community was reduced to the polychetes Ampharete sp. A andMagelona sp. H
and the sipunculan Aspidosiphon sp. Only Magelona sp. H and Aspidosiphon sp. maintained any
significant population levels in August 1990. During September and October 1990, the overall
increase in number of individuals was due primarily to the recruitment of Paraprionospio
pinnata andArmandia maculata and sustained levels of Magelona sp. H and Aspidosiphon sp.
1990 Taxonomic Composition of South Timbalier 53B
150 -\
Other
Sipuncula
Ophiuroidea
Crustacea
Gastropoda
Bivalvia
Nemertea
Polychaeta
Mediomastus ambiseta
Paraprionospio pinnata
Ampharete sp. A
Magelona sp. H
Owenia fusiformis
Aspidosiphon sp.
Clymenella torquata
6 7
Month
Figure 4. 1990 Taxonomic composition of South Timbalier 54A. Number of species within
taxonomic groups (total for ten 0.02-m2 cores) and mean number of individuals/m2
(n = 10) at ST53A for months indicated in 1990 (modified from Rabalais et a/. 1995).
99
-------
SEVERE VS. APERIODIC HYPOXIA
The composition of the benthic communities on the southeastern Louisiana continental
shelf reflected differences in sedimentary regime, the seasonal input of organic material and
seasonally severe hypoxia/anoxia. There was a precipitous reduction in species, abundance and
biomass of macroinfauna at the station exposed to severe and continuous hypoxia during mid-
summer (summarized in Table 1). At the intermittently hypoxic site there was a seasonal decline
in both species richness and abundance that was not obviously related to oxygen, but could be
attributed to a general decrease in organic material supply and/or increased predation. Except
during periods of severe hypoxia at Station ST53 A, Station WD32E had fewer species and lower
abundances, which likely reflected the latter's predominantly silt sediments.
Table 1. Characteristics of Louisiana continental shelf benthos subject to seasonally severe
hypoxia.
Reduced species richness
Severely reduced abundances (but never azoic)
Low biomass
Limited fauna (none with direct development)
Characteristic resistant species in fauna (e.g., a few polychetes and sipunculans)
Limited recovery following abatement of oxygen stress
The number of major taxonomic groups at Station WD32E was fairly consistent over
time indicating the lack of influence of severe hypoxia (either during the summer or in
successive years) on the benthic community. In contrast, at Station ST53A (severe summer
hypoxia every year), there was limited diversity of major taxa through most of the year and
especially during the period of severe hypoxia, despite higher sand content of the sediments.
The fauna at Station WD32E (other than polychetes) was composed of pericaridean crustaceans,
bivalves, gastropods and ophiuroids that were mostly absent at Station ST53 A. The taxonomic
diversity at Station ST53 A in spring and fall was the result of species with planktonic larvae, not
individuals with direct development; e.g., ampeliscid amphipods were essentially nonexistent in
the South Timbalier fauna in the two periods studied, 1985-86 and 1990-91 (Rabalais etal. 1989,
N. N. Rabalais, unpublished data; this study).
The summer "hypoxia" fauna at the South Timbalier sites was composed mostly of the
polychete Magelona sp. H and the sipunculan Aspidosiphon sp. Similar population levels of
these were maintained throughout the year. Paraprionospio pinnata peaked during spring and
fall recruitment periods at Station ST53 A and dominated the macroinfauna at Station WD32E
similar to other Louisiana-Texas inner shelf areas exposed to intermittent hypoxia (cf. Harper et
a/., 1981, 1991). P. pinnata is a highly fecund, multiple-spawning, ubiquitous member of the
benthic macroinfauna of the northwestern Gulf of Mexico shelf (Mayfield, 1988). The
opportunist capitellid polychete Mediomastus ambiseta and the surface deposit-feeding
polychete Ampharete sp. A, which are capable of readily exploiting the freshly deposited organic
material, were also dominant spring recruits at Station ST53 A. Opportunistic bivalves, such as
100
-------
Abra aequalis at the inshore Texas shelf site subject to aperiodic, but severe, hypoxia (Harper et
al. 1981, 1991) andMulinia lateralis at a southwestern Louisiana shelf site also subject to
aperiodic, but severe, hypoxia (Gaston 1985, Gaston and Edds 1994) were never common
members of the benthic community at Stations WD32E or ST53A.
IMPLICATIONS FOR FISHERIES RESOURCES
It is apparent that demersal fish and invertebrates, including the commercially important
penaeid shrimps, are not usually found where the oxygen concentration falls below 2 mg/L
(Pavela etal. 1983, Leming and Stuntz 1984, Renaud 1986, Craig et al. 2001, and Zimmerman
and Nance 2001), although some shrimps and invertebrates such as stomatopods have been seen
in submersible video tapes at oxygen concentrations as low as 1.5 mg/L (Rabalais etal. 2001b).
A large area of essential habitat for demersal-feeding organisms (up to 20,000 km2) is eliminated
in summer along the Louisiana shelf. Although these calculations of hypoxic zone size are
usually limited to single, 5-day survey estimates, some surveys repeated within 2 to 3 weeks
indicate a persistence to the distribution and size of the zone, at least in mid-summer (Rabalais et
al. 1999, Rabalais and Turner 2001). Data for the whole shelf are lacking for other times of the
summer, but hypoxia can often be widespread and severe on the southeastern shelf for much of
May-September.
In other hypoxia-affected estuarine and shelf environments, predators may benefit from a
hypoxia-stressed benthos, either during or immediately following hypoxia. Infauna that have
moved closer to the sediment-water interface may be more easily preyed upon (Diaz et al. 1992,
Pihl et al. 1992, Pihl 1994, and Nestlerode and Diaz 1998). This is not likely the case for the
severely affected areas of the southeastern Louisiana shelf (e.g., South Timbalier sites) for three
reasons: (1) the remaining surviving fauna is not predominantly at the sediment surface, (2) fish
predators are excluded from the zone of hypoxia and not seen by either direct observations or
video (Rabalais etal. 2001b) and (3) the presence of intact moribund and stressed benthic
organisms at the sediment surface is evidence for the absence of larger predators (Ibid).
Following the abatement of hypoxia in the fall, there was either a slight increase in biomass
predominantly by small, opportunistic polychetes (Station ST53 A), or no increase (Station
ST53B, not reported here) (Rabalais etal. 1995). Thus, a substantial area of feeding habitat is
removed from the foraging base of demersal organisms for months at a time. The proportion of
this unsuitable habitat to the whole of the Louisiana shelf is not known. Nematodes, while
reduced in abundance at more severely-affected stations (15-m depth) than inshore stations (8-m
depth) in the South Timbalier study area averaged about 1,200 individuals per 10 cm2 through
the year, but harpacticoid copepods were virtually eliminated by summer hypoxia (Murrell and
Fleeger 1989). The insensitivity of nematode densities to oxygen deficiency, or sometimes
increase, under severe hypoxia (Josefson and Widbom 1988, Levin et al. 1991, and Cook et al.
2000) may make these surviving meiofaunal organisms potential food for foraging fishes. The
relative suitability of this potential nematode food to demersal feeders on the shelf compared to
harpacticoid copepods and macroinfauna is not known. Fishes would not be potential predators
during mid-summer severe hypoxia, but nematodes may be suitable prey for some foragers
during the fall after hypoxia dissipates.
101
-------
Periods prior to severe hypoxia during spring recruitment have significantly higher
biomass in the form of small, opportunistic, surface-dwelling polychetes that should serve as a
readily available food source, except that the biomass levels vary from spring to spring. Areas
on the inshore periphery of severe hypoxia (intermittently or moderately affected) maintain
populations of opportunistic species, but do experience summer decreases in biomass that may
be due either to oxygen stress, reduced food supply or increased predation. Diaz and Solow
(1999) pointed out that these types of benthic communities did not store large amounts of energy
as biomass to buffer the ecosystem against the pulsing of energy and usually supported boom
and bust cycles. On the offshore periphery of the Louisiana shelf hypoxic, benthic populations
appeared to be relatively unaffected. However, abundances decreased with depth (Gaston et al.
1998), and probably biomass decreased as well if accepted continental shelf oxygen depth
gradients are applicable to the Louisiana continental shelf Through an annual cycle, therefore,
there are areas potentially without suitable food resources for extended periods, and other areas
with highly variable populations of opportunistic species that would be suitable prey for
demersal feeders.
While biomass in hypoxia-affected habitats on the Louisiana shelf may be periodically
high with opportunistic species, the overall productivity of the benthic system, transfer to other
trophic levels, and secondary production in general are not known. A high recruitment of larval
Mediomastus., Paraprionospio, Ampharete and other polychetes that have high growth rates
utilize the readily available organic matter fluxed to the seabed, and eventually provide suitable
food for demersal feeders and thereby contribute to a high, but temporary transfer of carbon to
higher trophic levels. These organisms do not persist through the severe summer hypoxia, and
increase in their biomass in fall is low. Their demise is predicted to be due to low oxygen and
not predation (i.e., no transfer of carbon), since the predators vacate the area before the decline in
benthos begins. Meroplankton, dominated by larval Paraprionospiopinnata, are distributed
throughout the water column in the summer and are more abundant when bottom water oxygen
is hypoxic than normoxic, but these larvae do not recruit to the benthos (Powers et al. 2001), or,
if they do, die immediately. Larger larvae in the overlying waters may have either delayed
metamorphosis, or they emigrated from the sediments under extreme oxygen stress (sensu
Wetzel et al. 2001). A higher secondary production based on high turnover of individuals does
not appear to be the case during the period of severe hypoxia.
Despite reduced suitable habitat and apparent reduced food resources at times of the year,
demersal fishery production remains high and must be supported by the available benthic
production (Chesney and Baltz 2001). The overall secondary production, however, may have
been affected or shifted within the context of decadal changes in primary production and
worsening hypoxia stress. Zimmerman and Nance (2001) found a correlation between the
reduction in total brown shrimp (Farfantepenaeus aztecus) catch in recent years as the mid-
summer size of the hypoxic zone increased and a recent decline in the catch per unit effort in the
brown shrimp fishery that corresponds with the expansion of hypoxia. This decline, however,
may be as likely related to other environmental factors or a combination of hypoxia and other
factors. Diaz and Solow (1999) provided evidence that annual productivity for some systems
with severely stressed habitats as a result of hypoxia was lower, but that this trend was not
consistent across habitat types. As more estuarine and coastal areas worldwide are exposed to
worsening oxygen stress, benthic communities will become more severely stressed. Benthic
102
-------
communities of the South Timbalier area (e.g., Station ST53A) are extremely stressed with
limited recovery and may be symptomatic of worsening oxygen conditions on the Louisiana
shelf. The relative area of such oxygen-stressed habitats on the Louisiana shelf has the potential
to affect carbon transfer to higher trophic levels, but at present the relative proportion of such
habitats is not known.
REFERENCES
Chesney, E.J., and D.M. Baltz. 2001. The effects of hypoxia on the northern Gulf of Mexico
coastal ecosystem: A fisheries perspective. In: Coastal Hypoxia: Consequences for
Living Resources and Ecosystems, N. N. Rabalais and R. E. Turner (eds.), Coastal and
Estuarine Studies 58, American Geophysical Union, Washington, D.C., pp. 321-354.
Cook, A.A., P.J.D. Lambshead, L.E. Hawkins, N. Mitchell, and L.A. Levin. 2000. Nematode
abundance at the oxygen minimum zone in the Arabian Sea. Deep-Sea Research, Part II
47: 75-85.
Craig, J.K., L.B. Crowder, C.D. Gray, CJ. McDaniel, T.A. Kenwood, and J.G. Hanifen. 2001.
Ecological effects of hypoxia on fish, sea turtles, and marine mammals in the
northwestern Gulf of Mexico. In: Coastal Hypoxia: Consequences for Living Resources
and Ecosystems, N. N. Rabalais and R. E. Turner (eds.), Coastal and Estuarine Studies
58, American Geophysical Union, Washington, D.C., pp. 269-291.
Diaz, R.J., and A. Solow. 1999. Ecological and Economic Consequences of Hypoxia: Topic 2
Report for the Integrated Assessment of Hypoxia in the Gulf of Mexico. NOAA Coastal
Ocean Program Decision Analysis Series No. 17, NOAA Coastal Ocean Program, Silver
Spring, Maryland.
Diaz, R.J., and R. Rosenberg. 1995. Marine benthic hypoxia: a review of its ecological effects
and the behavioural responses of benthic macrofauna. Oceanography and Marine Biology
Annual Review 33: 245-303.
Diaz, R.J., RJ. Neubauer, L.C. Schaffner, L. Pihl, and S.P. Baden. 1992. Continuous
monitoring of dissolved oxygen in an estuary experiencing periodic hypoxia and the
effect of hypoxia on macrobenthos and fish. In: Marine Coastal Eutrophication, R.A.
Vollenweider, R. Marchetti, and R. Viviani (eds.), Science of the Total Environment,
Supplement Number 0048-9697, pp. 1055-1068.
Gaston, G.R. 1985. Effects of hypoxia on macrobenthos of the inner shelf off Cameron,
Louisiana. Estuarine, Coastal and Shelf Science 20: 603-613.
Gaston, G.R., and K.A. Edds. 1994. Long-term study of benthic communities on the continental
shelf off Cameron, Louisiana: A review of brine effects and hypoxia. Gulf Research
Reports 9: 57-64.
103
-------
Gaston, G.R., B.A. Vittor, B. Barrett, and P.S. Wolfe. 1998. Benthic Communities of Louisiana
Coastal Waters, Draft Technical Bulletin No. 45. Louisiana Department of Wildlife and
Fisheries, Marine Fisheries Division, Baton Rouge, Louisiana.
Harper, D.E., Jr., L.D. McKinney, R.R. Salzer, and RJ. Case. 1981. The occurrence of hypoxic
bottom water off the upper Texas coast and its effects on the benthic biota. Contributions
in Marine Science 24: 53-79.
Harper, D.E., Jr., L.D. McKinney, J.M. Nance, and R.R. Salzer. 1991. Recovery responses of
two benthic assemblages following an acute hypoxic event on the Texas continental
shelf, northwestern Gulf of Mexico. In: Modern and Ancient Continental Shelf Anoxia,
R. V. Tyson and T. H. Pearson (eds.), Geological Society Special Publication 58, pp. 49-
64.
Josefson, A.B., and B. Widbom. 1988. Differential response of benthic macrofauna and
meiofauna to hypoxia in the Gullmar Fjord basin. Marine Biology 100: 31-40.
Leming, T.D., and W.E. Stuntz. 1994. Zones of coastal hypoxia revealed by satellite scanning
have implications for strategic fishing. Nature 310: 136-138.
Levin, L.A., C.L. Huggett, and K.F. Wishner. 1991. Control of deep-sea benthic community
structure by oxygen and organic-matter gradients in the eastern Pacific Ocean. Journal of
Marine Research 49: 763-800.
Mayfield, S.M. 1988. Aspects of the Life History and Reproductive Biology of the Worm
Paraprionospio pinnata. M. S. thesis, Department of Biology, Texas A&M University,
College Station, Texas.
Murrell, M.C., and J.W. Fleeger. 1989. Meiofauna abundance on the Gulf of Mexico
continental shelf affected by hypoxia. Continental Shelf Research 9: 1049-1062.
Nestlerode, J., and R.J. Diaz. 1998. Effects of periodic environmental hypoxia on predation of a
tethered polychaete, Glycera americana: implications for trophic dynamics. Marine
Ecology Progress Series 172: 185-195.
Pavela, J.S., J.L. Ross, and M.E. Chittenden. 1983. Sharp reductions in abundance of fishes and
benthic macroinvertebrates in the Gulf of Mexico off Texas associated with hypoxia.
Northeast Gulf Science 6: 167-173.
Pihl, L. 1994. Changes in the diet of demersal fish due to eutrophication-induced hypoxia in the
Kattegat, Sweden. Canadian Journal of Fisheries and Aquatic Science 51: 321-336.
Pihl, L., S.P. Baden, R.J. Diaz, and L.C. Schaffner. 1992. Hypoxia induced structural changes
in the diets of bottom-feeding fish and Crustacea. Marine Biology 112: 349-361.
104
-------
Powers, S.P., D.E. Harper, Jr., andN.N. Rabalais. 2001. Effects of Hypoxia/Anoxia on the
Supply and Settlement of Benthic Invertebrate Larvae. In: Coastal Hypoxia:
Consequences for Living Resources and Ecosystems, N. N. Rabalais and R. E. Turner
(eds.), Coastal and Estuarine Studies 58, American Geophysical Union, Washington,
D.C., pp. 185-210.
Rabalais, N.N., and R.E. Turner. 2001. Hypoxia in the Northern Gulf of Mexico: Description,
Causes and Change. In: Coastal Hypoxia: Consequences for Living Resources and
Ecosystems, N. N. Rabalais and R. E. Turner (Eds.), Coastal and Estuarine Studies 58,
American Geophysical Union, Washington, D.C., pp. 1-36.
Rabalais, N.N., L.E. Smith, and D.F. Boesch. 1989. The effects of hypoxic water on the benthic
fauna of the continental shelf off southeastern Louisiana. In: Proceedings: Ninth Annual
Gulf of Mexico Information Transfer Meeting, U.S. Minerals Management Service, OCS
Study MMS 89-0060, New Orleans, Louisiana, pp. 147-151.
Rabalais, N.N., R.E. Turner, and Q. Dortch. 1992. Louisiana continental shelf sediments:
indicators of riverine influence. In: Nutrient Enhanced Coastal Ocean Productivity
Workshop Proceedings. Publ. No. TAMU-SG-92-109, Texas A&M University Sea
Grant College Program, College Station, Texas, pp. 77-81.
Rabalais, N.N., L.E. Smith, E.B. Overton, and A.L. Zoeller. 1993. Influence of Hypoxia on the
Interpretation of Effects of Petroleum Production Activities, OCS Study/MMS 93-0022.
U.S. Dept. of the Interior, Minerals Management Service, Gulf of Mexico OCS Region,
New Orleans, Louisiana.
Rabalais, N.N., WJ. Wiseman, Jr., and R.E. Turner. 1994. Comparison of continuous records
of near-bottom dissolved oxygen from the hypoxia zone of Louisiana. Estuaries 17: 850-
861.
Rabalais, N.N., L.E. Smith, D.E. Harper, Jr., and D. Justic. 1995. The effects of bottom water
hypoxia on benthic communities of the southeastern Louisiana continental shelf.
MMS/OCS Study 94-0054, U.S. Minerals Management Service, Gulf of Mexico OCS
Region, New Orleans, Louisiana.
Rabalais, N.N., R.E. Turner, D. Justic, Q. Dortch, and W.J. Wiseman, Jr. 1999.
Characterization of hypoxia: Topic 1 Report for the Integrated Assessment of Hypoxia in
the Gulf of Mexico. NOAA Coastal Ocean Program Decision Analysis Series No. 16,
NOAA Coastal Ocean Program, Silver Spring, Maryland.
Rabalais, N.N., L.E. Smith, D.E. Harper, Jr., and D. Justic. 2001a. Effects of seasonal hypoxia
on continental shelf benthos. In: Coastal Hypoxia: Consequences for Living Resources
and Ecosystems, N.N. Rabalais and R.E. Turner (eds.), Coastal and Estuarine Studies 58,
American Geophysical Union, Washington, D.C., pp. 211-240.
105
-------
Rabalais, N.N., D.E. Harper, Jr., and R.E. Turner. 2001b. Responses of nekton and demersal
and benthic fauna to decreasing oxygen concentrations. In: Coastal Hypoxia:
Consequences for Living Resources and Ecosystems, N.N. Rabalais and R.E. Turner
(eds.), Coastal and Estuarine Studies 58, American Geophysical Union, Washington,
D.C., pp. 115-128.
Renaud, M.L. 1986. Hypoxia in Louisiana coastal waters during 1983: implications for
fisheries. Fishery Bulletin 84: 19-26.
Rosenberg, R. 1985. Eutrophication—The future marine coastal nuisance? Marine Pollution
Bulletin 16:227-231.
Tolmazin, R. 1985. Changing coastal oceanography of the Black Sea. I. Northwestern shelf.
Progress Oceanogr. 15: 2127-276.
Wetzel, M.A., J.W. Fleeger, and S.P. Powers. 2001. Effects of hypoxia and anoxia on
meiofauna: a review with new data from the Gulf of Mexico. In: Coastal Hypoxia:
Consequences for Living Resources and Ecosystems, N.N. Rabalais and R.E. Turner
(eds.). Coastal and Estuarine Studies 58, American Geophysical Union, Washington,
D.C., pp. 165-184.
Zimmerman, R.J., and J.M. Nance. 2001. Effects of hypoxia on the shrimp fishery of Louisiana
and Texas. In: Coastal Hypoxia: Consequences for Living Resources and Ecosystems,
N.N. Rabalais and R.E. Turner (eds.). Coastal and Estuarine Studies 58, American
Geophysical Union, Washington, D.C., pp. 293-310.
106
-------
ADAPTATIONS OF A DEMERSAL FISH SPECIES IN A NUTRIENT-RICH
EMBAYMENT OF THE IONIAN SEA (GREECE)
V. Vassilopoulou1, C. Papaconstantinou1, andE. Caragitsou1
ABSTRACT
The Amvrakikos Gulf, a semi-enclosed embayment in northeastern Greece, is subjected
to considerable organic enrichment. The length and age frequency distribution, the degree of
sexual maturity, and the diet of red mullet in the Gulf have been investigated in order to gain a
better understanding of the biological and ecological responses of this fish to eutrophication
features. Our data showed that the vast majority of red mullet collected in the Gulf were
relatively young (1 and 2 years old) specimens, which had not reached sexual maturity. Mean
length-at-age appeared to be smaller than in other areas of Greece, and growth in weight was
allometric, reflecting possibly the poor somatic condition of these fish. The feeding intensity of
red mullet appeared to be reduced, particularly in the inner part of the Gulf, which had the lowest
dissolved oxygen concentrations and smaller diversity and abundance values of species
belonging to the benthic community. A bathymetric barrier to the distribution of red mullet
seemed to exist in the Amvrakikos Gulf because of the unfavorable conditions prevailing in
deeper areas, causing larger specimens offish to conduct an ontogenetic migration to deeper
waters in the Ionian Sea.
INTRODUCTION
Anthropogenous nutrient and organic carbon loading are the causes of eutrophication in
many river, estuarine, and coastal systems. Decomposition of organic matter sinking to the
bottom combined with stratification of the water column result in the formation of vertical
gradients of dissolved oxygen concentrations that lead to hypoxic/anoxic conditions, particularly
in the bottom layer. Behavioral responses of various fishes to these low dissolved oxygen
concentrations have been studied worldwide (Pihl et al. 1991, Duque et al. 1998, Paerl et al.
1998, Plante et al. 1998). Most commonly, fish appear to avoid hypoxic waters (Plante et al.
1998, Schurmann et al. 1998), while there are cases when oxygen availability in the environment
determines migration patterns (Ochumba et al. 1993, Snelling et al. 1993, Statkus 1998).
Moreover, there are several reports of reduced growth of freshwater fishes exposed to hypoxia
(Stewart et al. 1967, Andrews et al. 1973, Weber and Kramer 1983, Pedersen 1987, Thetmeyer
et al. 1999), but information pertinent to growth reduction of marine fishes under hypoxic
conditions is limited. This is possibly due to the fact that effects of hypoxia in marine
environments is less easily isolated/detected. Chabot and Dutil (1999) studying Atlantic cod
(Gadus morhua L.) and Pichavant et al. (2000) studying turbot (Scophthalmus maximus L.)
reared at different levels of dissolved oxygen stated that oxygen availability had a strong
negative influence on fish growth rate.
National Centre for Marine Research, Hellinikon, 16604, Athens, Greece.
107
-------
One of the most important European wetlands subjected to high organic enrichment in
recent years is the Amvrakikos Gulf in the northeastern Mediterranean Sea. Eutrophication,
along with the distribution of water properties in the Gulf, resulting in strong stratification of the
upper layer (Friligos et al. 1997) create hypoxic conditions in the deeper waters (Friligos and
Koussouris 1977), that seriously affect the benthic community (Bogdanos etal. 1989), as well as
the demersal fish assemblages of the area (Papaconstantinou and Caragitsou 1990). The
objective of the present study is to provide information on the length and age frequency
distribution, sexual maturity, and the feeding habits of red mullet (Mullus barbatus L.) in order
to gain a better understanding of the biological/ecological responses of the red mullet stock
related to eutrophication features. In terms of economic value, the red mullet is the most
important demersal fish in the Amvrakikos Gulf.
MATERIAL AND METHODS
Area of Study
The Amvrakikos Gulf is a semi-enclosed embayment in northwestern Greece, with a
maximum depth of approximately 60 m. It connects to the Ionian Sea by a shallow (< 10m), 600
m wide channel. The rivers Louros and Arachthos discharge into its northern reach after having
drained the surrounding, extensively cultivated plains. In the delta area a coastal lagoon system
has formed, which is one of the largest in the Mediterranean region (Diapoulis et al. 1991). In
the delta area there are also a number of aquaculture farms. Domestic sewage discharges from
the city of Preveza, in the northwestern part of the gulf, as well as from other coastal towns and
villages contribute to nutrient enrichment of the waters. Nutrient concentrations, particularly
high values of ammonia, silicate, and phosphate, highlight the eutrophied character of the area
(Friligos et al. 1997). In addition, salinity values remain very low throughout the year (16.8-34.8
ppt), resulting in strong stratification of the water column (Ibid.). Low dissolved oxygen
concentrations (< 2 mg/L) occur in the deeper areas of the gulf (> 29 m), particularly in the
eastern (inner) area (Friligos et al. 1987), creating hypoxic conditions that cause significant
changes in the structure and function of benthic communities (Bogdanos et al. 1989).
Sampling Procedure
During five seasonal sampling cruises, (November 1996; February, May, July, and
September 1997), 1,518 red mullets were collected at 13 sites distributed throughout the
Amvrakikos Gulf (Figure 1) using trammel nets with a mesh size of 17mm. The depths of the
sampling stations were restricted to between 3 and 18m, since at depths greater than 18m there
weren't any red mullets in the catches. For each specimen, length to the nearest millimeter,
weight to the nearest gram, and sex were recorded. Gonadal maturity was determined according
to Nikolsky's scale: Stages I, II, III immature gonads; IV, V mature; VI spent (Nikolsky 1976).
Otolith pairs were extracted, cleaned and stored in plastic vials containing 100% glycerin. The
otoliths (sagittae) of the fish were extracted and stored in glycerin, while the stomachs were
stored in 6% buffered formaldehyde solution.
108
-------
39.05-
39.00-
sass-
LON GIT UDE
E)
Figure 1. Location of sampling stations in the Amvrakikos Gulf (NW Greece). Stations in the
eastern (inner) part are marked differently from those in the western (outer) part.
Sample and Data Analysis
Fish age determination was based on otolith readings. For an age reading, otoliths were
placed on a concave watch glass having a black background and containing glycerin; they were
then examined under the reflected light of a stereoscope at a magnification of 25x. Two distinct
types of rings were observed, the transition from one to the other being quite pronounced;
opaque rings appeared white and translucent rings appeared dark. Translucent rings, continuing
around the entire circumference of the otolith, were considered annual growth zones (annuli).
Fork length (FL) - whole body weight (W) regressions were calculated according to the formula
W=aFLb. Differences in the slopes of the length-weight regressions between sexes were
investigated using analysis of covariance (ANCOVA).
For analysis of feeding habits, samples were segregated by month and area of capture
(western-outer and eastern-inner part of the gulf). The degree of fullness (L) was estimated for
individual stomachs according to Lebedev's scale (Lebedev 1946), ranging from 0 (empty) to 5
(stomach fully distended with food). Measures of food intake were determined by calculating an
index of fullness (B: wet weight of stomach contents as a percentage of total body weight of each
predator; Hyslop 1980). Index values were log-transformed and subjected to multiple analysis of
variance (MANOVA) to evaluate the order of importance of season and area of capture in
determining the variability in stomach contents. Prey items were identified to the most precise
taxonomic level possible, and counted and weighed to the nearest 0.01 g. Percent number (Cn)
and weight (Cw) of each prey item was calculated.
109
-------
An ordination technique (multi-dimensional scaling) using the Bray-Curtis similarity
coefficient (Bray and Curtis 1957) was performed on the Cw of each prey item in each season in
the two areas of the Amvrakikos Gulf, using the PRIMER algorithms (Clarke and Warwick
1989). An analysis of similarities (ANOSIM) randomisation routine was then applied to test
whether dietary samples were significantly different, using again the PRIMER package. To
establish which species contribute most to separating one group from another, the similarity
percentage breakdown procedure (SIMPER), comparing the mean percent weight contribution
(Cw) of each prey taxa within one group with that of another group, was used. The contribution
of each species to the Bray-Curtis similarity coefficient was calculated after root-root
transformation, and the species were then ranked in order of their contribution to separating two
groups, percent and cumulative percent (Warwick and Clarke 1991) present in sampled
stomachs.
Diet similarity was determined using Schoener's index (Schoener 1970). The index
values were compared with Langton's convention (Langton 1982), in which values of 0.00-0.29,
0.30-0.60 and greater than 0.60 define low, medium, and high similarity, respectively.
Prey diversity in the diet, which corresponds to food niche breadth (Scrimgeour and Winterbourn
1987), was calculated using the Shannon-Wiener index (Pielou 1966).
RESULTS
Age and Growth
Analysis of variance applied to red mullet fork length data did not reveal any significant
differences (P>0.05) relative to season, area of capture (inner or outer part of the Gulf) or sex of
the specimens, so data were combined. From Figure 2a, showing the length frequency
distribution of the 1518 red mullet collected in the Amvrakikos Gulf, it is obvious that modal
lengths at capture were 120 to 140mm FL, while there were very few specimens with lengths
greater than 160 mm. Moreover, young-of-the-year specimens (FL<100 mm) were never
collected, possibly due to the selectivity of the net used in the framework of the present study.
Data from a previous study, however, using a net with a mesh size of 10 mm suggested that the
smallest group of red mullet (60-80 mm FL) were collected close to the channel connecting the
Gulf to the open sea area (Caragitsou, unpublished data).
40
30
g
2U
10
0
in I I
10 11
—
12
FL
— i
13
(rr
14 15 16
m)
50
40
gso
20
10
1 2 3
Age groups
Figure 2. (a) Length and (b) age frequency distribution of the red mullet collected in the
Amvrakikos Gulf.
Age readings were performed twice, approximately 2 months apart. Annual marks were
quite distinct and appeared clearly, while non-annual marks when occurring were unclear and did
110
-------
not form continuous rings. The study of the marginal growth of the red mullet sagittae revealed
that in a year's time one opaque and one translucent ring is formed. The vast majority (98.1%)
of red mullet collected in the Amvrakikos Gulf belonged to the 1+ and 2+ age classes, while
there was also a small proportion (1.9%) of older (3+) specimens (Figure 2b).
The evaluation of the age-length key (Table 1) revealed age variation of 2 to 3 years
within a single length group. Of the specimens sampled, 77.2% had a fork length belonging to
the 120-140 mm length interval
Table 1. Age-length key constructed from red mullet collected in the
Amvrakikos Gulf.
Length interval
101-110
111-120
121-130
131-140
141-150
151-160
Mean FL (mm)
I
(%)
2.0
11.7
24.4
9.8
1.0
123
Age Group
II
(%)
16.6
25.9
6.8
134
III
(%)
0.5
1.0
0.5
145
ANCOVA applied to the FL-W regressions did not reveal any significant differences
between sexes (F=3.31, P>0.05). In both sexes, however, growth in weight was allometric, and
in particular the red mullet specimens from the Amvrakikos Gulf seemed to be thinner for their
length.
Fork length - total weight regressions were calculated for males and females:
Females: W=2.47-10'4FL2'44,
Males: W=6.98-10'5FL2'71,
r=0.998
r2=0.997
111
-------
Gonadal Maturity
In May the peak spawning season for fish species in Greek waters (Vassilopoulou 1987),
only 12% of specimens collected in the Amvrakikos Gulf were found to be sexually mature.
This could be possibly due to the fact that red mullet of the present study were relatively young
specimens, while it is known that the species does not reach sexual maturity until after the
second year of life (Vassilopoulou and Papaconstantinou 1991).
Feeding Intensity
Although there were very few specimens with empty stomachs, the degree of stomach
fullness of most specimens ranged between 1 (traces of food in the stomach) and 2 (small portion
of food in the stomach) of the Lebedev scale (Lebedev 1946). The degree (L) and the index (B)
of stomach fullness in each month and area of capture are reported in Table 2. Multiple analysis
of variance applied to the data revealed that both month and area of capture had a significant
effect on L and B (MANOVA: FL=7.356, FB=7.031, P< 0.001). Both L and B were significantly
lower in late spring and summer.
Geographically, red mullet from the western (outer) part of the Gulf exhibited higher L
and especially B values, as compared to those caught in the eastern (inner) part.
Table 2. Total numbers of individual stomachs analysed, number of empty stomachs, mean
degree (L) and mean index (B) of stomach fullness with 95% confidence intervals (a) in
the five sampling months, (b) in the two parts of the Amvrakikos Gulf.
(a) Analysis according to month
No. of stomachs
No. of empty stomachs
Degree of fullness (L)
Index of fullness (B)
No. of stomachs
No. of empty stomachs
Degree of fullness (L)
Inex of fullness (B)
November February
92 84
0 0
May
71
5
2.39+0.23 2.15+0.19 1.38+0.18
0.91+0.08 0.85+0.09 0
(b) Analysis
Western
270
1
2.17+0.13
0.90+0.07
42+0.15
according to
July
63
0
1.87+0.22
0.59 + 0.11
location
Eastern
128
4
2.02+0.17
0.65+0.07
September
93
0
2.48+0.19
1.15+0.08
112
-------
Temporal and Spatial Variations in Stomach Contents
Overall, in terms of percentage by weight, the stomach contents consisted mostly of
polychetes, molluscs (bivalves and gastropods), crustaceans (mainly amphipods, decapods and
mysids), and sipunculids, while the rest of the prey taxa, although eaten regularly, appeared to be
of minor dietary importance (Figure 3).
Othe
Mollusc
olychaeta
Figure 3. Major prey taxa found in the stomachs of red mullet from the Amvrakikos Gulf.
Ordination applied to the seasonal dietary replicates from the two parts of the Gulf
(western/outer and eastern/inner) revealed that at the arbitrary 70% similarity level replicates
from the western part formed one group (group I), while three replicates from the eastern part
formed another group (group II) (Figure 4). The spring replicate from the eastern part was
separated from the rest. ANOSIM revealed that the groupings identified by MDS ordination
differed significantly (P < 0.05). Results derived from the SIMPER routine suggested that
polychaetes of the family Sipunculidae and various isopod species that were encountered more
frequently in the western part, were the major discriminating taxa for the two groups. The
dissimilarity of the spring replicate in the eastern part of the Amvrakikos Gulf seemed to arise
from the small number of prey taxa, found in the stomachs of the species during that sampling
period. A striking difference, however, occurring in the diet of the species during the latter
season in the whole Amvrakikos Gulf area was the absence of bivalves from the stomach
contents, which played an important dietary role during all other seasons.
113
-------
Group I
Me
Figure 4. MDS plot of the recent weight contribution (Cw) of each prey item in each
season/month (F: February, M: May, J: July, S: September, N: November) in the
two parts (w: western, e: eastern part) of the Amvrakikos Gulf.
Dietary Overlap and Niche Breadth
Dietary overlap, calculated on the basis of percentage by weight (Cw), of prey found in
fish stomachs indicated changes in the prey spectrum in relation to season. There was
considerable dietary similarity (S > 0.60) in November, July, and September (Table 3a), while
the decrease of the species' niche breadth in February and May, particularly in the eastern part of
the Gulf (Table 3b) appeared to have a negative effect on dietary overlap values. In general, fish
collected in the western part of the gulf appeared to have a wider dietary breadth compared to
those caught in the eastern part (Table 3b), reflecting possibly the existence of more prey taxa in
the outer part of the gulf.
Table 3. (a) Schoener's index of dietary overlap (S) between red mullet collected in different
months and (b) Shannon-Wiener index (H) reflecting the dietary breadth of specimens
collected during the five samplings in the western (outer) and eastern (inner) part of the
Amyrakikos Gulf.
February
May
July
September
November
February
May
July
September
November
0.37
0.63
0.84
0.77
(a) Schoener's index
February May
-
0.43
0.34 0.49
0.23 0.57
(b) Shannon- Wiener index
Western part Eastern part
2.77 2.35
2.34 1.95
2.44 1.91
2.87 2.24
2.68 1.98
July
-
0.77
114
-------
DISCUSSION
Our data showed that the vast majority of red mullet collected in the Amvrakikos Gulf
had a size of 120-140 mm, being relatively young specimens (1 to 2 years old), while the
occurrence of larger fish seemed to be rather opportunistic. Mean length-at-age appeared to be
smaller than in other Greek waters (Vassilopoulou and Papaconstantinou 1991), suggesting
possibly a reduced growth rate for the species in the Gulf. Growth in weight was allometric
indicating the existence of thinner specimens for their length, which possibly reflects relatively
poor somatic condition for red mullet in the Amvrakikos Gulf.
The study of the gonadal maturity of red mullet revealed that in May, when red mullet
spawn in Greek waters (Vassilopoulou 1987), few specimens with mature gonads were collected
in the Gulf. This could be due to the fact that the red mullet stock in the Amvrakikos Gulf was
comprised of relatively small and, hence, sexually immature specimens.
A shift of red mullet to deeper waters as they grow has been mentioned before
(Vassilopoulou and Papaconstantinou 1988). In striped mullet (M. surmuletus), the existence of
an ontogenetic movement to deeper waters appeared to be triggered by first maturity (Machias et
al. 1998). If this is also the case for red mullet, it appears reasonable to hypothesize that
unfavourable environmental conditions in deeper hypoxic areas of the gulf possibly cause the
migration of larger specimens to the Ionian Sea where the species appears to reproduce,
according, at least, to the results of an ichthyoplankton survey study (Anonymous 1989).
Migration of demersal fish and crustaceans, affected by hypoxia, to normoxic waters was also
mentioned by Pihl et al. (1991). The ontogenetic migration of Dover sole (Microstomus
pacificus) to deeper waters is a general pattern; abundance values, however, were lower at depths
where lowest oxygen concentrations occurred (Jacobson and Hunter 1993).
Red mullet in the Amvrakikos Gulf are benthophagous, feeding on infaunal and epifaunal
invertebrates, primarily polychetes, bivalves and crustaceans. Seasonal changes appeared in the
feeding habits of Gulf specimens, and were also found in red mullet from other parts of the
Mediterranean Sea (Haidar 1970, Caragitsou and Tsimenidis 1982). A number of authors have
shown that, as the density of a particular prey type declines, a predator may switch to feeding on
another prey that is more abundant (Murdoch 1969, Cornell 1976, Hume and Northcote 1985,
Davidson 1986). The lack of data on prey availability in the study area did not allow similar
observations.
The feeding intensity of red mullet in the Amvrakikos Gulf appeared to be reduced,
particularly in the inner part of the Gulf and during the winter-spring period. The latter coincides
with the period of increased river flow, which seemed to be critical for the benthos of the area
(Bogdanos et al. 1989). Quantitative, as well as qualitative, differences existed in the diet of red
mullet between the two parts of the Amvrakikos Gulf; larger quantities of food were found in the
stomachs of specimens collected in the western Amvrakikos Gulf. Qualitatively, in the latter
area there was a greater diversity among stomach contents, as compared to the eastern part. A
difference in the diversity and abundance of benthic species was also established between the
two parts of the Gulf; higher diversity and abundance values coincided with the western part
115
-------
(Bogdanos etal. 1989). Reduced feeding intensity and appetite under hypoxic conditions has
been reported for other fish (Chabot and Dutil 1999, Thetmeyer et al. 1999), and appeared to
produce reduced growth. In the Amvrakikos Gulf, although there were indications for reduced
growth and poor condition of red mullet, as well as reduced feeding intensity, it was not possible
to isolate hypoxia effects, from those of interacting factors.
Summing up, the red mullet stock in the Amvrakikos Gulf appeared to consist of
relatively young, sexually immature fish, and although trawl fishing is prohibited in the Gulf,
this Gulf stock is fully vulnerable to coastal fisheries. The fact that the species dwells in the Gulf
the first 2 years of its life when energy requirements and allocation for growth should be at
maximum, inasmuch as the growth increment only reaches 70% of the species' greatest
attainable size (Vassilopoulou and Papaconstantinou 1991) possibly suggests that the
Amvrakikos Gulf does not constitute a favourable habitat for young mullet. Then, the mullet
shift to open seawaters where they spawn and do not return. The latter observation could be also
attributed to conditions prevailing in deeper areas of the Gulf, making a significant portion of the
benthic habitats in the Gulf uninhabitable for red mullet and possibly for other demersal fish
species, thereby limiting the fishery potential of the area. The previous discussion highlights the
degradation of the Amvrakikos Gulf ecosystem that should be thoroughly investigated in order to
provide the appropriate data that would assist in improved management of this ecologically and
economically important area of Greek waters.
REFERENCES
Andrews, J.W., T. Murai, and G. Gibbons. 1973. The influence of dissolved oxygen on the
growth of channel catfish. Transactions of the American Fisheries Society 102: 835-838.
Anonymouns 1989. The Fisheries of the Amvrakikos Gulf. In: Final Report on the
Oceanographic Study of the Amvrakikos Gulf. Vol. 5. National Centre for Marine
Research Editions, Athens, pp. 1-72 (in Greek).
Bogdanos, C., P. Panagiotidis, N. Simboura, and V. Mastori. 1989. Study of the Benthos of the
Amyrakikos Gulf. In: Final Report on the Oceanographic Study of the Amvrakikos Gulf
4. Biological Oceanography. National Centre for Marine Research Editions, Athens, pp.
3 8-47 (in Greek).
Bray, J.R., and J.T. Curtis. 1957. An ordination of the upland forest communities of southern
Wisconsin. Ecological Monographs 27: 325-349.
Caragitsou, E., and N. Tsimenides. 1982. Seasonal changes and comparative analysis of the
food of the red mullet (Mullus barbatus) in Thracian Sea. Thalassographica 5: 41-61.
Chabot, D., and J. Dutil. 1999. Reduced growth of Atlantic cod in non-lethal hypoxic
conditions. Journal of Fish Biology 55: 472-491.
116
-------
Clarke, K.R., and R.M. Warwick. 1989. Lecture notes prepared for the training workshop on
the statistical treatment and interpretation of marine community data. FAO/IOC/UNEP,
Athens, September 1989, Part 11, 84 pp.
Cornell, H. 1976. Search strategies and the adaptive significance of switching in some general
predators. American Naturalist 110: 317-320.
Davidson, RJ. 1986. Mussel selection by the paddle crab Ovalipes catharus: evidence of a
flexible foraging behaviour. Journal of Experimental Marine Biology and Ecology 102:
281-299.
Diapoulis, A., Th. Koussouris, I. Bertahas, and G. Photis. 1991. Ecological stresses on a delta
area in western Greece. Toxicological and Environmental Chemistry 31-32: 285-290.
Duque, A.B., D.C. Taphorn, and K.O. Winemiller. 1998. Ecology of the corporo, Prochilodus
mariae (Characiformes, Prochilodontidae), and status of annual migrations in western
Venezuela. Environmental Biology of Fishes 53: 33-46.
Friligos, N., and T. Koussouris. 1977. Preliminary chemical, physical and biological
observations in the gulf of Amvrakikos. Thalassia Jugoslavia 13: 53-58.
Friligos, N., P. Dagre, R. Psyllidou-Giouranovits, and G. Pappas. 1987. Oceaonographic Study
of the Amvrakikos Gulf. National Center for Marine Research Technical Report Volume
III: Chemical Oceanography, 13-24. (In Greek)
Friligos, N., E.T. Balopoulos, and R. Psyllidou-Giouranovits. 1997. Eutrophication and
hydrography in the Amvrakikos Gulf. Fresenius Environmental Bulletin 6: 21-26.
Haidar, Z. 1970. L'oecologie du rouget (Mullus barbatus) en Adriatique orientate. Acta
Adriatica 14: 1-94.
Hume, J.M.B., and T.G. Northcote. 1985. Initial changes in use of space and food by
experimentally segregated populations of Dolly Varden (Salvelinus malmd) and cutthroat
trout (Salmo clarki). Canadian Journal of Fisheries and Aquatic Science 42: 101-109.
Hyslop, E.J. 1980. Stomach content analysis - a review of methods and their application.
Journal of Fish Biology 17: 411-429.
Jacobson, L.D., and J.R. Hunter. 1993. Bathymetric demography and management of Dover
sole. North American Journal of Fisheries Management 13:405-420.
Langton, R.W. 1982. Diet overlap between Atlantic cod, Gadus morhua, silver hake,
Merluccius bilinearis and fifteen other Northwest Atlantic finfish. Fishery Bulletin 80:
745-759.
117
-------
Lebedev, N.V. 1946. Elementary populations of fish. Zoologicheskii Zhurnal 25: 121-135 (in
Russian).
Machias, A., S. Somarakis, andN. Tsimenides. 1998. Bathymetric distribution and movements
of red mullet Mullus surmuletus. Marine Ecology Progress Series 166: 247-257.
Murdoch, W.W. 1969. Switching in general predators: experiments on predator specificity and
stability of prey predations. Ecological Monographs 39: 335-354.
Nikolsky, G.V. 1976. The Ecology of Fish. 6th Ed. Academic Press, New York, 352 pp.
Ochumba, P.B.O., M. Gophen, and L. Kaufman. 1993. Changes in oxygen availability in the
Kenyan portion of Lake Victoria: Effects on fisheries and biodiversity. The Lake
Victoria Ecosystem Workshop: People, Fisheries, Biodiversity, and the Future of Lake
Victoria., Edgerton Research Laboratory, New England Aquarium, Boston,
Massachussets (USA), 33p. (non vidi).
Paerl, H.W., J.L. Pinckney, J.M. Fear, and B.L. Peierls. 1998. Ecosystem responses to internal
and watershed organic matter loading: consequences for hypoxia in the eutrophying
Neuse River Estuary, North Carolina, USA. Marine Ecology Progress Series 166: 17-25.
Papaconstantinou, C., and E. Caragitsou. 1990. Abiotic parameters effects on the abundance of
fish populations in the Amvrakikos Gulf. Proc. 2nd Hellenic Cypriotic Congress of
Chemistry and Environment, Greek and Cypriotic Union of Chemists, Athens, pp. 416-
420 (in Greek).
Pedersen, C.L. 1987. Energy budgets for juvenile rainbow trout at various oxygen
concentrations. Aquaculture 62: 289-298
Pichavant, K., J. Person-Le-Ruyet, N. Le Bayon, A. Severe, A. Le Roux, L. Ouemener, V.
Maxime, G. Nonnote, and G. Boeuf. 2000. Effects of hypoxia on growth and
metabolism of juvenile turbot. Aquaculture 188: 103-114
Pielou, E.G. 1966. Shannon's formula as a mean of specific diversity: its use and misuse.
American Naturalist 100: 463-465.
Pihl, L., S.P. Baden, and RJ. Diaz. 1991. Effects of periodic hypoxia on distribution of
demersal fish and crustaceans. Marine Biology 108: 349-360.
Plante, S., D. Chabot, and J. Dutil. 1998. Hypoxia tolerance in Atlantic cod. Journal of Fish
Biology 53: 1342-1356.
Schoener, T.W. 1970. Nonsynchronous spatial overlap of lizards in patchy habitats. Ecology
51:408-418.
118
-------
Schurmann, H., G. Claireaux, and H. Chartois. 1998. Changes in vertical distribution of sea
bass (Dicentrarchus labrax L.) during a hypoxic episode. Hydrobiologia Volume
371/372: 207-213.
Scrimgeour, G.J., and MJ. Winterbourn. 1987. Diet, food resource partitioning and feeding
periodicity of two riffle-dwelling fish species in a New Zealand river. Journal of Fish
Biology 31: 309-324.
Snelling, J.C., C.B. Schreck, C.S. Bradford, L.E. Davis, and C.H. Slater. 1993. Migratory
characteristics of spring chinook salmon in the Willamette River. Annual report. Oregon
State University, Corvallis (USA). Department of Fisheries and Wildlife, 47 pp.
Statkus, R. 1998. Biology and fishery industry of smelt (Osmerus eperlanus) in Lithuanian
economic zone of the Baltic Sea. Fishery and Aquaculture in Lithuania. Lithuanian
Society of Hydrobiologists, 3: 115-124 (non vidf).
Stewart, N.E., D.L. Shumway, and P. Dourdorff. 1967. Influence of oxygen concentration on
growth of juvenile largemouth bass. Journal of the Fisheries Research Board of Canada
24: 475-494.
Thetmeyer, H., U. Waller, K.D. Black, S. Inselmann, and H. Rosenthal. 1999. Growth of
European sea bass (Dicentrarchus labrax L.) under hypoxic and oscillating oxygen
conditions. Aquaculture 174: 355-367.
Vassilopoulou, V. 1987. Sexual maturity of red mullet (Mullus barbatus) in the gulfs of Korin-
thiakos and Patraikos and in the Ionian Sea. Ptoe. 2nd Hellenic Symposium on
Oceanography and Fisheries, National Centre for Marine Research, Athens, pp. 565-570
(in Greek).
Vassilopoulou, V., and C. Papaconstantinou. 1988. Distribution with depth and catches per unit
effort of the hake and the red mullet in the western coastal waters of Greece. United
Nations Food and Agricultural Organization Fisheries Report 394:174-180.
Vassilopoulou, V., and C. Papaconstantinou. 1991. Aspects of the biology and dynamics of red
mullet (Mullus barbatus} in the Aegean Sea. FAO Fisheries Report 477: 115-126.
Warwick, R.M., and K.R. Clarke. 1991. A comparison of some methods for analysing changes
in benthic community structure. Journal of the Marine Biology Association of the United
Kingdom 71: 225-244.
Weber, J.M., and D.L. Kramer. 1983. Effects of hypoxia and surface access on growth,
mortality, and behavior of juvenile guppies, Poecilia reticulata. Canadian Journal of
Fisheries and Aquatic Science 40: 1583-1588.
119
-------
120
-------
OXYGEN MINIMUM ZONE INFLUENCE ON THE
COMMUNITY STRUCTURE OF DEEP-SEA BENTHOS
Lisa A. Levin1
ABSTRACT
Mid-water oxygen minimum zones impinge on continental margins in vast regions of the
eastern Pacific Ocean, off west Africa, and in the Arabian Sea. Where this occurs, benthic
communities can exhibit distinct features, including reduced body size, altered densities, shifts in
taxonomic composition, reduced species richness, and elevated dominance. Increased incidence
of chemosynthetic nutritional pathways and reduced bioturbation activity are also characteristic
of oxygen minimum zone assemblages. Adaptation of benthic faunas to low oxygen within these
regions includes long, thin body forms, elaborate respiratory surfaces, and the presence of
hemoglobin. Many questions remain concerning controls on faunal abundance, trophic
pathways, physiological adaptations and the consequences of reduced diversity within oxygen
minima. Answers to these questions will help us understand past and future incidences of
hypoxia.
INTRODUCTION
Oxygen minimum zones (OMZs) are regions of the ocean with hypoxic waters, where
oxygen concentrations typically are < 0.5 mg/L (or < about 20|iM). They usually occur in mid-
water at upper bathyal depths (200-1200 m) (Wyrtke 1966, 1973). Where they intercept the
continental margin, they produce great effects on benthic ecosystems. The goal of this
presentation is to describe general features of the benthic environment within OMZs and to
review what is known about the effects of OMZs on benthic communities and organisms.
OMZs generally form where strong upwelling leads to high surface productivity that
subsequently dies, sinks and degrades, thereby depleting oxygen within the water column. OMZ
formation is most intense in regions of sluggish circulation and where there are source waters
already low in oxygen. OMZs occur in much of the eastern Pacific Ocean, in the Arabian Sea,
and off West Africa (Figure 1) (Kamykowski and Zentara 1990). Deep-water hypoxia also is
found in deep basins in the southern California borderland, and in some fjords (Diaz and
Rosenberg 1995).
Typically, a vertical profile of oxygen concentration through an OMZ exhibits a steep
drop in oxygen at the upper boundary, within the top 100-200 m. Below this, there is a zone of
continuous low oxygen with concentrations often « 0.1 ml/L. The thickness of the OMZ varies
regionally, and is strongly influenced by circulation. Off Mexico and in the Arabian Sea,
^ntegrative Oceanography Division, Scripps Institution of Oceanography, La Jolla, California 92093-0218 USA.
121
-------
the OMZ is about 1000 m thick (Wyrtki 1973), but off Peru, Chile and California, the OMZ is
only a few hundred m thick (Wyrtki 1966). The lower OMZ boundary exhibits a gradual
increase in oxygen with water depth.
ISO" 120
Figure 1. Distribution of oxygen minimum zones in the world oceans (modified from Diaz and
Rosenberg, 1995).
OMZs differ from many shallow-water hypoxic regions in exhibiting stable, persistent
low oxygen over ecological time scales, such that sessile species will live out many generations
in continuous low oxygen. At the upper OMZ boundary, oxygen concentrations can vary with
internal tides (Levin et al. 1991) or larger-scale oceanographic forces such as El Nino (Tarazona
et al. 1988), but such variation usually has little effect on core dissolved oxygen values. OMZ
intensity and distribution vary on geological time scales. Shifts in productivity or circulation
over thousands of years are thought to drive expansion and contractions of OMZs, both
vertically and horizontally (Tyson and Pearson 1991, Rogers 2000).
122
-------
Where oxygen minimum zones impinge on the seafloor, they create strong gradients in
bottom-water oxygen concentration. Associated with these low oxygen values are high organic
carbon contents of the sediments. Percent organic carbon (POC) values of 4-6% are typical of
the sediments of many OMZs (Levin and Gage 1998), but off Peru values can reach 16% (Levin
et al, submitted). There is generally an inverse relationship between bottom-water oxygen
concentration and sediment POC in bathyal sediments.
Another feature of many OMZs is the presence of large, filamentous, sulfur-oxidizing,
nitrate-reducing bacteria (Jorgensen and Gallardo 1999). These are typically Thioploca, but
sometimes include Beggiatoa, and may form mats, tufts or a thin, grass-like cover (Jorgensen
and Gallardo 1999, Levin personal observation).
BENTHIC RESPONSE TO OMZ CONDITIONS
Community Structure
Many aspects of benthic ecosystems vary within oxygen minimum zones. This issue was
first explored by Sanders (1969) for the benthos off Walvis Bay, West Africa. Since then,
additional research has been carried out in the eastern Pacific Ocean (e.g., Levin et al. 1991,
Bernhard et al. 2000, Neira et al. 2001, Levin et al. submitted) and the Arabian Sea (Gage et al.
2000). Aspects of animal community structure, including size, abundance, taxonomic
composition, diversity and lifestyles are distinct within sediments intercepted by an OMZ.
Ecosystem functions, such as bioturbation and trophic pathways, also vary within, compared to
beneath, OMZs.
Perhaps the most inclusive benthic system response to OMZ conditions is altered size
structure. At very low oxygen levels (<0.1 ml/L), the fauna often consists of meiofaunal size
organisms (protozoans and metazoans); macrofauna and megafauna are typically rare or absent
(Levin et al. 1991, Levin personal observation). Such conditions exist in the Santa Barbara
Basin (Bernhard et al. 2000) and on Volcano 7, a seamount off Mexico, where the summit
protrudes into the eastern Pacific OMZ (Levin et al. 1991). On Volcano 7, bottom-water
dissolved oxygen concentration increases downslope in a linear fashion. Bacteria are abundant
at the summit, where oxygen is lowest, but little else lives there except nematodes. The
macrofauna are rare; megafauna are absent or sparse (Levin et al. 1991, Wishner et al. 1995).
Presumably, small-bodied animals have an advantage in severe hypoxia by having a larger
surface area:volume ratio. Multiple regressions were employed to examine relationships among
environmental factors and densities of bacteria, meiofauna and macrofauna (Levin et al. 1991).
Results of these analyses suggest that the densities of bacteria and metazoan meiofauna are
related largely to measures of organic matter availability (e.g., Chlorophyll a, POC, PON).
Bottom-water oxygen concentration was correlated only with macrofaunal densities.
Megafaunal densities were not tested.
Distinct abundance trends are seen within OMZs for some taxa but not others. Total
densities of meiofauna are never reduced within OMZs, and often reach maximum bathyal
123
-------
values within OMZs, presumably due to abundant particulate food and/or reduced predation
intensity (Neira etal. 2001). In contrast, abundances of macrofauna are sometimes depressed.
They often exhibit a maximum or peak where oxygen levels climb even slightly, to
concentrations of 0.1 - 0.2 ml/L. This pattern has been observed off central California, West
Africa, and Mexico (Volcano 7), and on the Oman margin in the Arabian Sea (Figure 2). It has
sometimes been called a boundary effect, but often these local dissolved oxygen maxima occur
well above the OMZ lower boundary (technically defined as 0.5 ml/L). Some sort of
physiological threshold appears to create this pattern. Once the oxygen level rises sufficiently, a
small number of tolerant species are able to attain exceptionally high densities due to the great
food supply. We have not observed this effect off Peru (Levin et al. submitted). Similar
thresholds occur in the megafauna as illustrated by counts on Volcano 7 (Wishner et al. 1990,
1995). Extraordinarily high densities of megafauna can be found near OMZ boundaries.
At the community level, we see distinct taxonomic trends associated with OMZs.
Among macrofauna, annelids dominate. Echinoderms and other heavily calcified taxa are often
reduced in density within OMZs (Levin et al. 1991, 1997, 2000). Crustaceans and molluscs are
even less tolerant groups (Diaz and Rosenberg 1995), although there are certain taxa (Ampelisca,
Astyrispermodesta, lepidomeniomorph aplacophorans) that are exceptions (Levin, unpublished
data).
We might expect to see the same macrofaunal taxa distributed globally within OMZs.
However, there is a surprising amount of variation. The top four species at stations with the
lowest oxygen levels in four regions of the world each are distinct (Table 1). Polychetes
dominate off Oman, where oxygen concentration is just above 0.1 ml/L. At 400 m on the Oman
margin, most individuals belong to two species, a spionid Prionospio sp. and a cirratulid,
Aphelochaeta sp. (Levin et al. 1997). Off Peru, a single species of gutless, tubificid oligochete,
Olavius crassitunicatus (Phallodrilinae) comprises most of the macrofauna (Levin et al.
submitted). Surprisingly, a gastropod (Astyrispermodestd) is typically found near the sediment-
water interface of the Santa Barbara Basin (L. Levin and J. Bernhard, unpublished). On Volcano
7, where the summit is covered by coarser foraminiferal sands, aplacophorans and polychetes
(including pogonophorans) dominate (Levin et al. 1991). At the edge of the Santa Barbara Basin
(555 m), oligochetes also dominate, but there is a surprising diversity of taxa present, including
crustaceans, echinoderms and aplacophorans (Beaudreau, 1999). At the Basin center, there are
no macrofauna although one meiofaunal polychete appears in low numbers on a 0.3 mm screen.
Among the meiofauna, nematodes, and calcareous foraminifera are most tolerant of low
dissolved oxygen levels (Levin et al. 1991, Cook et al. 2000, Gooday et al. 2000, Neira et al.
2001). This is illustrated by varying ratios of nematodes to harpacticoid copepods within the
metazoan meiofauna. On the summit of Volcano 7 and on the Peru margins, nematode to
harpacticoid ratios are very high (500:1 and 65:1, respectively) within the OMZ (Levin etal.
1991, Neira et al. 2001). Beneath the OMZs, the ratios are much lower. Based on studies of
meiofauna off Oman, Peru and Mexico, harpacticoid copepods and agglutinated foraminifera
appear to be especially intolerant to low oxygen (Gooday et al. 2000, Neira et al. 2001, Levin et
al. submitted).
124
-------
DEPTH (m)
CJQ
A U H-
0000
a a o a
a a o a
i
„ D
D— ux
0 °
3 \
3 \
3 \
" \ /
80 a /
= V
_,0
w
e
- e
e
H*
e
e
**
W
- e
e
e
w
o
e
55
o
•*•
5
w
e
§-8
r 3 5.-
< I' ^
5. 3. o
g
N
(11
O
D
||
^X
o
3
CO
CD
=5
Q.
CD
CO
_.
l£3
o^
(SS N* —
o en o en
o o o o
o o o o
^ ^ ^ l 1 — i r~- a
Lh r^
n [] pi^^'^x /
^
* \
-
» \
03 \
^ fl
HHf —
s
- O
e
s
M
9
e
e
«
e
s
e
e
e
S-
'
en
a
r^ C 3
lit
ft
<
5'
8 ~
a
o
o
en
o
o
tn
o
1
e
s. O B.
^ o
CO
en
o
o
u
o
a
o
N
en
D
D
O
O
O
U1
o
o
a
a
a
tn
C fa
"
25
o
a
t/J
n
u
125
-------
Table 1. Listing of the four top-ranked macrofaunal species in different bathyal OMZ settings.
Data are from Levin et al. (1991, 1997, and unpublished) and Beaudreau (1999).
Rank
1.
Oman Margin
400m
(0.13 ml/L)
Taxon (% of total)
Prionospio sp. (63)
(Polychaeta)
Peru Margin
300m
(0.02 ml/L)
Olavius crassitunicatus (83)
(Oligochaeta)
Volcano 7
750m
(0.08 ml/L)
Lepidomeniidae (47)
(Aplacophora)
Santa Barbara Basin
Center -590 m
(0.055 ml/L)
Nerillidae
(Polychaete)
Edge -550 m
(0.055 ml/L)
Tubificidae (44)
(Oligochaete)
Aphelochaeta sp. (27) Turbellaria (13)
(Polychaeta)
Tharyxsp. (3)
(Polychaeta)
Astrys permodesta (2)
(Gastropoda)
Cirrophorus lyra (18) Astrys permodesta Ampelisca sp. (25)
(Cirratulidae) (Gastropoda) (Amphipoda)
Pogonophoran (9)
(Polychaeta)
Holothuroidea (25)
Cossura sp. (1)
(Polychaeta)
Sigambra sp. (1)
(Polychaeta)
Protodorvillea sp. (8)
(Polychaeta)
Aplacophora (6)
The taxonomic shifts described above translate into changes in species diversity both
with respect to dominance and species richness. Dominance of macrobenthos is extraordinarily
high within OMZs. In a survey of 5 OMZ regions, the top ranked species comprise 47-87% of
the total macrofauna (Table 1). Accompanying this high dominance is reduced species richness.
Graphical representation of Rank 1 dominance and rarefaction measure of species richness
(Esioo) as a function of dissolved oxygen level for bathyal sites around the world indicate that the
effects of oxygen on diversity are evident only at oxygen levels below about 0.3 or 0.4 ml/L
(Figure 3). Possible causes of reduced species richness within OMZs include loss of species
within taxa that are generally much less tolerant to low oxygen, for example the echinoderms,
crustaceans and molluscs. However, reductions in richness also occur within tolerant taxa such
as the annelids (Levin and Gage 1998). Organic enrichment may also contribute to reduced
diversity, independent of oxygen level. Separating the effects of hypoxia from those of organic
enrichment within OMZs is difficult. Large-scale, multiple regression studies by Levin and
Gage (1998) suggest that within the polychetes, oxygen exerts greatest control on species
richness, while organic matter availability has more influence on measures of dominance and
evenness. In the Arabian Sea, one measure of food availability, sediment pigment concentration,
explained 70-90% of variation in indices of macrofaunal species richness, information index,
dominance, and evenness (Levin et al. 2000).
Functional Processes
Trophic Pathways. One might expect that lying beneath the most productive waters in
the world, the OMZ benthos would rely on heterotrophic consumption of this production.
126
-------
However, recent findings suggest that chemosynthesis plays an important role in OMZ systems
in several ways (Levin et al.unpublished data). Numerous OMZ species possess endosymbiotic,
sulfide-oxidizing bacteria that fix and translocate carbon to the host, or episymbiotic bacteria
that may also provide food to the host. Still other species consume the free-living bacteria, or
prey on species that do, or on species with symbionts. One of the most interesting examples of
an OMZ species with chemoautotrophic endosymbionts is the gutless oligochete Olavius
crassitunicatus, the dominant taxon at 300 m depth off Peru where oxygen concentrations are <
0.02 ml/L. This latter species possesses three types of subcuticular bacteria, at least one of
which oxidizes sulfur (Giere and Krieger 2001). This oligochete comprises 83% of the
macrofauna present at this site and attains densities of over 13,500 individuals/m3 (Levin et al.
submitted). Other examples of symbiont-bearing taxa within OMZs include pogonophorans on
Volcano 7, nerilid polychetes and nematodes with episymbionts in the Santa Barbara Basin, and
lucinid clams on the Oman margin and in the Santa Barbara Basin. A recent paper by Bernhard
et al. (2000) has shown symbioses to be the norm for protists and meiofaunal metazoans in the
Santa Barbara Basin.
Bioturbation. Animal activities such as bioturbation and bioirrigation enhance
oxygenation and solute transport, and speed the remineralization of organic matter. In general,
bioturbation is reduced within OMZs (Savrda and Bottjer 1991). Under extreme hypoxia or
anoxia, all bioturbating organisms are absent and laminae or varves often form. Bioturbation of
1
75-
a*
a
a 50-
S
o
Q
25-
0-
>
A
• *
•
•
•
• _
• * ».
:•
: .-v •
••;
ou -
50-
40-
f
H 30-
(»
20-
10-
I
n-
..
•
0
•
•
•
A
• *
• •*
* »*
> ^
0123
Bottom -Water Oxygen (ml/1)
0123
Bottom -Water Oxygen (ml/1)
Figure 3. Plot of Rank 1 Dominance and Expected Species Richness (Esioo) as a function
of bottom-water oxygen concentration for bathyal macrofauna from the eastern
Pacific and northern Indian Oceans.
recent sediments has been quantified in two OMZ regions, on the Peru and Oman margins. In
both cases the mixed layer depth, determined by Pb-210 and Th-234 profiles, is much thinner
within than beneath the OMZ (Smith et al. 2000, Levin et al. submitted). Specifically, on the
Peru margin, particle mixing rates, measured by Th-234, were much lower within the OMZ (14
127
-------
cm2/year) relative to that at a station beneath the OMZ (80-100 cm2/year) (Levin et al.
submitted). However, no reduction in particle mixing rate was observed within, as opposed to
beneath, the Oman margin OMZ, but a longer-lived tracer (Pb-210) was used in this study
(Smith etal. 2000).
ADAPTATIONS TO PERMANENT HYPOXIA
Adaptations of biota to OMZ conditions were reviewed by Childress and Seibel (1998),
largely for planktonic organisms. These authors emphasized that animals living in OMZs must
adapt to limited oxygen availability, but not to complete depletion. For even at very low oxygen
concentrations, there is enough oxygen available in the water if organisms can access it.
Childress and Seibel (1998) proposed three general approaches that OMZ taxa can use to cope
with low oxygen: (1) increase the effectiveness of oxygen uptake; (2) lower their metabolic
demands; and (3) switch to anaerobic metabolism. The authors argue that the first method is the
most widely encountered approach. OMZ fauna do show lower metabolic oxygen requirements
than shallow water relatives, but other deep-water species not living in OMZs do as well. The
third approach is used mainly by vertically migrating plankton that can pay back oxygen debts
incurred during daily migrations (up or down) to better-oxygenated water. In general, all of
these possible adaptations are poorly studied in benthic species.
Childress and Seibel (1998) also proposed five general methods by which organisms can
increase the effectiveness of their oxygen uptake. Four of these have been observed in OMZ
benthic macrofauna. Increased gill surface area is evident in amepliscid amphipods, a group that
occurs in OMZs off Oman, Chile, Peru, and California (Levin, unpublished data). Elongated,
proliferated and numerous branchiae appear to be adaptations to permanent hypoxia in some
spionid and dorvilleid polychetes (Lament and Gage 2000, Levin personal observation).
Cossurid polychetes have exceptionally long median antennae within the Oman margin OMZ
(Lament and Gage 2000). Increased gill surface area has also been documented in mid-water
mysids, fishes, and cephalopods (Childress and Seibel 1998). Another possible adaptation,
reduced diffusion distances, may explain the success of small, thin, elongated taxa such as
oligochetes and nematodes within OMZs. Development of respiratory pigments (e.g.,
hemoglobins) with high affinity for oxygen has been observed in benthic fish (Sebastolobus
alascanus) as well as pelagic fishes that live in the OMZ. Such adaptations have also been used
by molluscs and other organisms where hemoglobins have been observed (e.g.,Amygdalum
politum) (Levin, unpublished observations). A variety of behavioral adaptations, including
vertical migration of plankters (Childress and Seibel 1998) and ontogenetic migrations (Wishner
et al. 2000), have been documented. Aplacophoran molluscs within the OMZ on the summit of
Volcano 7 seem to live with their mantle permanently open, a possible adaptation to increase
respiration effectiveness (A. Scheltema, Woods Hole Oceanographic Institution, personal
communication). Increased ventilatory ability and circulation capacity have been documented in
midwater crustaceans as possible adaptations to OMZs, but have not been but studied in benthic
species.
128
-------
REMAINING QUESTIONS
In general, OMZ benthos are poorly studied, and there exist more questions than answers
about these systems. Key ecological questions that remain to be answered include the following:
1. What really controls standing stock in OMZs? Do oxygen and organic matter
availabilities interact in determining abundances? What is the role of sulfides? Food cannot be
the sole determinant of community structure because the most organic-rich system in the world
off Peru has only small bodied organisms and low biomass density (Levin et al. submitted).
2. What are the physiological adaptations of benthic OMZ animals? Are there enzymatic
adaptations? Do chemoautotrophic symbionts play a role?
3. What is the relative importance of chemosynthesis-based versus photosynthesis-based
nutrition in OMZs?
4. What are the functional consequences of low diversity in OMZs? The effects of low
diversity on ecosystem-level processes of production and remineralization are of considerable
interest. OMZs with their low diversity may be a particularly good place to study these.
CONCLUSIONS
Understanding the structure and function of modern OMZ faunas can help us to
understand the past and possibly to predict the future. Modern OMZ faunas are considered
analogs for construction of paleoecological low-oxygen models (Savrda and Bottjer 1991).
These include biofacies (body fossil), ichnofacies (trace fossil) and bioturbation (particle mixing)
models. These models link the body or trace fossils of animals and the amount of sediment
mixing to the oxygen level of overlying waters, or in some cases to productivity or organic
matter availability. Scientists use this information to reconstruct the conditions in ancient seas.
Studies of places such as the Peru margin, which are affected by interannual oxygen shifts, can
help us understand system responses. For example, we have learned from studies of the Peru
margin that bioturbation by non-feeding, symbiont-bearing forms can occur under almost anoxic
conditions (Levin et al. submitted). It is also likely that modern OMZs can provide clues about
how a shallow water system might change should it move from episodic to permanent hypoxia.
Finally, the study of modern OMZs can reveal the types of adaptations that animals might
undergo or be selected for over ecological and evolutionary time.
Studies of OMZ benthos to date have been hampered largely by limited access to deep-
water systems in remote parts of the world. As scientists begin to understand the importance of
these systems for nutrient cycling and for evaluating adaptations to extreme environments, our
knowledge of these systems should increase tremendously.
129
-------
ACKNOWLEDGMENTS
Much of the oxygen minimum zone research discussed here has been supported by grants
to the author from the Office of Naval Research, the National Science Foundation, the North
Atlantic Treaty Organization, and the University of California Ship Funds. Collaborations with
Joan Bernhard, Adam Cook, John Gage, Victor Gallardo, Dimitri Gutierrez, Andrew Gooday,
Peter Lament, Lauren Mullineaux, Carlos Neira, Anthony Rathburn, Javier Sellanes, Craig
Smith, and Karen Wishner have been particularly helpful. I thank Ann Beaudreau, Cindy
Huggett, Chris Martin, David James, Joshua Hillman, Guillermo Mendoza, Leslie Harris, Larry
Lovell and many others for assistance in the laboratory and with identifications.
REFERENCES
Beaudreau, A. 1999. Macrofaunal community responses to hypoxia in the Santa Barbara Basin.
UCSD SURF Program undergraduate report. 44 pp.
Bernhard, J.M., K.R. Buck, K.R. Farmer, and S.S. Browser. 2000. The Santa Barbara Basin is a
symbiosis oasis. Nature 403: 77-80.
Childress, J.J., and B.A. Seibel. 1998. Life at stable low oxygen levels: adaptations of animals
to oceanic oxygen minimum layers. Journal of Experimental Biology 201: 1223-1232.
Cook, A.A, P.J. Lambshead, L.E. Hawkins, L.E., N. Mitchell, and L.A. Levin. 2000. Nematode
abundance at the Oxygen Minimum Zone in the Arabian Sea. Deep-Sea Research 47: 75-
85.
Diaz, R.J., and R. Rosenberg. 1995. Marine benthic hypoxia: A review of its ecological effects
and the behavioral responses of benthic macrofauna. Oceanography and Marine Biology
Annual Reviews 33: 245-303
Gage, J.D., L.A. Levin, and G.A. Wolff. 2000. Benthic processes in the deep Arabian Sea.
Deep-Sea Research 47: 1-2.
Giere, O., and J. Krieger. 2001. A triple bacterial endosymbiosis in a gutless oligochaete
(Annelida). Ultrastructural and immunocytochemical evidence. Invertebrate Biology
120: 41-49.
Gooday, A.J., J.M. Bernhard, L.A. Levin, and S. Suhr. 2000. Foraminifera in the Arabian Sea
OMZ and other oxygen deficient settings: taxonomic composition, diversity and relation
to metazoan faunas. Deep-Sea Research 47: 25-54.
Jorgensen, B.B., and V.A. Gallardo. 1999. Thioploca spp: filamentous sulfur bacteria with
nitrate vacuoles. FEMS Microbiology Ecology 28: 301-313.
130
-------
Kamykowsky, D., and S.J. Zentara. 1990. Hypoxia in the world ocean as recorded in the
historical data set. Deep-Sea Research 37: 1861-1874.
Lament, P. A., and J.D. Gage. 2000. Morphological responses of macrobenthic polychaetes to
low oxygen on the Oman continental slope, NW Arabian Sea. Deep-Sea Research 47: 9-
24.
Levin, L.A., and J.D. Gage. 1998. Relationships between oxygen, organic matter and the
diversity of bathyal macrofauna. Deep-Sea Research 45:129-163.
Levin, L.A., J.D. Gage, P. Lament, L. Cammidge, C. Martin, A. Patience, and J. Crooks. 1997.
Infaunal community structure in a low-oxygen, organic-rich habitat on the Oman
continental slope, NW Arabian Sea. In: The Responses Of Marine Organisms to Their
Environment, L.E. Hawkins and S. Hutchinson (eds). University of Southampton,
pp. 223-230.
Levin, L.A., J.D. Gage, C. Martin, and P. A. Lament. 2000. Macrobenthic community structure
within and beneath the oxygen minimum zone, NW Arabian Sea. Deep-Sea Research
Special Volume: Benthic Processes in the Arabian Sea. Deep-Sea Research 47: 189-
226.
Levin, L. A., G. Gutierrez, A. Rathburn, C. Neira, J. Sellanes, P. Munoz, V. Gallardo, and M.
Salamanca. Submitted. Benthic processes on the Peru Margin: a transect across the
oxygen minimum zone during the 1997-98 El Nino. Progress in Oceanography.
Levin, L.A., C.L. Huggett, and K.F. Wishner. 1991. Control of deep-sea benthic community
structure by oxygen and organic-matter gradients in the eastern Pacific Ocean. Journal of
Marine Research 49: 763-800.
Neira, C., L. Sellanes, L.A. Levin, and W.E. Arntz. 2001. Meiofaunal distributions on the Peru
margin: relationship to oxygen and organic matter availability. Deep-Sea Research 48:
2453-2472.
Rogers, A.D. 2000. The role of oceanic oxygen minimum zones in generating biodiversity in
the deep sea. Deep-Sea Research 47: 119-148.
Sanders, H. 1969. Benthic marine diversity and the Stability-Time Hypothesis. Brookhaven
Symposium in Biology 22: 71-81.
Savrda, C.E., and D.J. Bottjer. 1991. Oxygen-related biofacies in marine strata: an overview
and update. In: Modern and Ancient Continental Shelf Anoxia, R.V. Tyson, and T.H.
Pearson (eds.), Geological Society, Special Publication 58, pp. 201-219.
131
-------
Smith, C.R., L.A. Levin, DJ. Hoover, M. Cremer, G. McMurtry, and J.D. Gage. 2000.
Bioturbation across the oxygen minimum zone on the Arabian-Sea Slope. Deep Sea
Research Part II 47: 227-257.
Tarazona, J., H. Salzwedel, and W. Arntz. 1988. Positive effects of "El Nino" on macrobenthos
inhabiting hypoxic areas of the Peruvian upwelling system. Oecologia 76: 184-190.
Tyson, R.V., and T.H. Pearson. 1991. Modern and ancient continental shelf anoxia: an
overview. In: Modern and Ancient Continental Shelf Anoxia, R.V. Tyson, and T.H.
Pearson (eds.), Geological Society, Special Publication 58, pp. 1-24.
Wishner, K., CJ. Ashjian, C. Gelfman, M. Gowing, L. Kann, L. Levin, L. Mullineaux, and J.
Saltzman. 1995. Pelagic and benthic ecology of the lower interface of the eastern
tropical Pacific oxygen minimum zone. Deep-Sea Research 42: 93-115.
Wishner, K.F., M.M. Gowing, and C. Gelfman. 2000. Living in suboxia: ecology of an Arabian
Sea oxygen minimum zone copepod. Journal of Limnology and Oceanography 45: 1576-
1593.
Wishner, K., L. Levin, M. Gowing, and L. Mullineaux. 1990. Involvement of the oxygen
minimum in benthic zonation on a deep sea mount. Nature 346: 57-59.
Wyrtki, K. 1966. Oceanography of the eastern Pacific Ocean. Oceanography and Marine
Biology Annual Reviews 4: 33-68.
Wyrtki, K. 1973. Physical oceanography of the Indian Ocean. In: The Biology of the Indian
Ocean, B. Zeitzschel (ed.) Springer-Verlag, Berlin, pp. 18-36.
132
-------
RESPONSE OF BENTHIC FAUNA AND CHANGING SEDIMENT
REDOX PROFILES OVER A HYPOXIC GRADIENT
Rutger Rosenberg1, Hans C. Nilsson1, and Robert J. Diaz2
ABSTRACT
The Koljofjord is an enclosed, stratified fjord on the Swedish west coast with
hypoxic/anoxic bottom water during most of the year. In the winter of 1999-2000, the water in
the entire fjord was re-oxygenated after a period of stagnation, but the following summer oxygen
concentrations declined to below 1 ml/L between 20 and 40 m depths. The objectives of this
study were to investigate the structure of benthic communities along a depth gradient of
declining oxygen concentration and the impact of fauna on sediment redox conditions. The
vertical distribution of the fauna in the sediment was restricted to the upper few centimetres.
Dominant species at most stations were the burrower Capitella capitata and the tube-builder
Pseudopolydora antennata. The species found in the fjord are probably not particularly tolerant
of hypoxia, but they have life-history traits that facilitate a rapid colonisation following
improved oxygen conditions. The depth of the redox potential discontinuity (RPD) layer, a
recognisable division zone between oxidised (sub-oxic) and reduced chemical conditions, is
dependent on infaunal activity, e.g. burrows, tubes, and feeding voids. Measurement of apparent
RPD (aRPD) from sediment profile images (SPIs) compared well to electrode measurement of
RPD. We conclude that a digital analysis of aRPD from images has many advantages compared
to RPD measurements by electrodes.
INTRODUCTION
Disturbance gradients elicit various responses from the benthos depending upon the
source of stressor in combination with temporal and spatial factors. As benthic communities are
changed, biologically mediated geochemical cycles are also altered. Systems stressed by organic
matter have received the most attention, with many having a well documented fit to the response
model developed by Pearson and Rosenberg 1978, (e.g. Heip 1995, Nilsson and Rosenberg
2000).
Dissolved oxygen is a key factor in regulating both benthic community complexity and
many biogeochemical cycles, such as sulphur and nitrogen (Aller 1979, Yingst and Rhoads
1980, Jenkins and Kemp 1984). In many marine and estuarine systems, the flux of dissolved and
particulate substances across the sediment-water interface is regulated by benthic organisms.
Department of Marine Ecology, Goteborg University, Ristineberg Marine Research Station, 450 34 Fiskebackskil,
Sweden. 2Virginia Institute of Marine Science, College of William and Mary, Gloucester Point, Virginia 23062,
USA.
133
-------
These organisms mix and irrigate surface sediments, and by this activity create a thin layer of
oxidised sediments between the water column and deeper anaerobic sediments. During periods
of hypoxia (dissolved oxygen < 2 ml/L) changes in community structure and behaviour lead to
altered geochemical profiles in the sediments as bioturbation, to support macrofaunal activity,
declines from lack of oxygen.
Initially, it is the behaviour of organisms that is changed. This leads to first order
reductions in bioturbation rates that are reversible if the hypoxia is not severe or long-term.
When hypoxia leads to mortality, then second order changes in bioturbation rates occur that are
not reversed until community structure is restored.
To investigate the degree to which benthic communities and geochemistry are correlated,
we examined their relationship along a hypoxic gradient in the Koljofjord, a fjord located on the
west coast of Sweden (Figure 1). The Koljofjord is part of the fjordic system around the island
of Orust. To the west it is connected to the Skagerrak by an 8 m sill and to the east by a 12 m sill
with the adjoining Havstensfjord. A halocline that fluctuates around 15m stratifies the
Koljofjord and prevents mixing of surface and bottom waters, which makes Koljofjord prone to
development of hypoxia. Salinity above the halocline is 22%o to 25%o and below it is about
28%o. Hypoxic conditions are alleviated by the exchange of the bottom water that takes place
during the winter at irregular intervals of 1 to 8 years (Gustafsson and Nordberg 1999).
Figure 1. Map of the Koljofjord with the six stations labelled by depth. Shaded areas are
shallower than 10m.
134
-------
In several fjords along the Swedish Skagerrak coast, annual minimum oxygen
concentrations in the bottom water declined significantly in the period 1951 to 1984 (Rosenberg
1990). As a consequence, the benthic fauna has been reduced, particularly in the deeper parts of
the fjords where hypoxia/anoxia is most frequent (Josefson and Rosenberg 1988, Nilsson and
Rosenberg 1997, Gustafsson andNordberg 1999). However, historical data on the commercially
important shrimp, Pandalus borealis, indicated that it was abundant in the Koljofjord in 1909,
but in 1910 only few shrimp were caught (Bjorck 1913). Bjorck (1913) suggested that the
decline in shrimp was due to low dissolved oxygen, but no measurements were made.
The oxygen concentrations in the Koljofjord have a variable pattern (Figure 2). In early
1999, the oxygen concentrations were high at depths of 10 and 15m, but zero at 40 m depth. In
September oxygen dropped to below 1 ml/L even at 10 m depth. In the winter of 1999-2000, the
water in the fjord was re-oxygenated with gradually greater concentrations in the shallowest
depths. Shortly after this, a steep decline occurred at all depths. As an example, at 22 m depth
the oxygen concentration peaked in March at 6.2 ml/L and declined to 0.7 ml/L at the time of
sampling in early August. During the steepest decline from March to May at this depth, the
declining rate was 0.08 ml/L/day. During sampling in August, the oxygen concentration was 1.0
ml/L at 15 m depth and less deeper down.
E G
O
C r
p •!'
e
* ,
o "
(I \/fh \V
i i
Date
Figure 2. Dissolved oxygen concentrations (ml/L) from January 1999 to August 2000 at the
water depths 10, 15, 20, 30 and 40 m. The recordings were made close to station 40
(courtesy of Bohuslans Vattenvardsforbund).
135
-------
The objectives of this study were to investigate the structure of benthic communities
along a depth gradient of declining oxygen concentration and the impact of the fauna on redox
conditions in the sediment. The six stations investigated in the Koljofjord were located along a
depth gradient from 10 to 40 m, with the 40 m station being closest to the location of long-term
dissolved oxygen monitoring. The depth of the redox potential discontinuity (RPD) layer is a
recognisable division zone between oxidised (sub-oxic) and reduced chemical conditions in the
sediment (Fenchel 1970, Lyle 1983, Santschi et al. 1990). The oxidised part appears as rust-
brown in color; the reduced layer below this is generally grey-green or black. In the present
study, we compared two methods to assess the depth of the RPD in the sediment. Measurements
were made both with electrodes and by digitally analysing the depth of RPD in sediment profile
images (SPIs). The SPI technique has proven useful in this context in a number of other studies
(e.g. Rhoads and Germano 1986, Nilsson and Rosenberg 2000).
MATERIAL AND METHODS
Samples for infauna and sediment were taken in the Koljofjord at the following depths:
10, 15, 18, 22, 30 and 40 m on 31 July and on 2 and 4 August 2000. The depths are used as
station numbers (Figure 1). At each station, five deployments were made with a sediment profile
camera, and three samples for infauna were taken with a 0.05 m2 Ponar grab. A digital CCD
camera (Canon Power Shot Pro 70) took vertical in situ pictures through a prism (30 x 15 cm) as
described in Nilsson and Rosenberg (1997). After each deployment, the sediment profile image
(SPI) was transferred to a computer and stored. The contrasts in the SPIs were enhanced in
Adobe Photoshop 5.0. The depth of the apparent redox potential discontinuity (aRPD) was
measured, using a software programme (NUT image 1.6), as the distance from the sediment
surface to the borderline between rust-brown and green-grey or sometimes even black sediment.
This color borderline indicated the shift between ferric (Fe+3) and ferrous (Fe+2) ions (e.g.
Mortimer 1941, 1942). In each image, the maximum and minimum aRPD was measured, the
mean aRPD calculated as the area of aRPD coverage divided by the width of the image, and the
benthic habitat quality (BHQ) index was calculated. This index characterises sediment
structures, sub-surface structures, and the aRPD. The BHQ index varies between 0 and 15, and
is related to the faunal successional stages in the Pearson-Rosenberg model, where low values
indicate a disturbed benthic fauna and a thin aRPD, and high values indicate a diverse fauna and
deeper lying aRPD (Nilsson and Rosenberg 1997, Nilsson and Rosenberg 2000). The grab
samples were washed on a 1-mm sieve and preserved in 70% ethanol. Biomass was reported as
ethanol wet weight.
Three sediment cores were taken at each station with a gravity corer (4.5 cm diameter).
RPD was measured with 10 different electrodes placed simultaneously and vertically through
pre-drilled holes in the cores at distances of 1 cm between measurements. The electrodes were
15 cm long with a 1 mm diameter. The electrodes were platinum with a colomel electrode used
as reference; the recorder was a Radiometer (Copenhagen, pH meter 22). A value of+240 mV
was added to the potential measured, which then corresponds to the shift between insoluble
ferric and soluble ferrous ions (Mortimer 1941, 1942, Fenchel 1969, Lyle 1983). At station 18,
only one core was successfully analysed. The 0-1 cm layer of the sediment was analysed for
total carbon and nitrogen with a Carlo Elba Elemental Analyzer. The water on top of the cores
136
-------
was siphoned off, immediately preserved, and analysed for oxygen content by the Winkler
method. Historical, monthly records of oxygen concentrations shown in Figure 2 are from the
monitoring programme of Bohuslans Vattenvardsforbund county.
RESULTS
The benthic fauna in the Koljofjord was depauperate with a dominance of polychetes
(Table 1). The total number of species collected was 12. At 18 m and deeper, only the two
polychetes Pseudopolydora antennata and Capitella capitata were found. The total abundance
at stations 10 and 15 was slightly above 200 individuals/m2 and less at greater depths. The
biomass exceeded 1 g/m2 only at station 10. Variations were generally large and indicative of
patchy distributions.
In the SPIs from stations down to 30 m water depth, the top sediment appeared oxidised
and benthic fauna was present (Figure 3). Tubes of P. antennata were seen on the sediment
surface in images from 15 to 30 m depth, and one long, thin tube of Euchonepapillosa appeared
at 10m. The rust-brown colour in the images is indicative of the oxidised (sub-oxic) zone in the
sediment, and the dark colour shows the reduced zone and presence of iron sulphide. At 10 m
depth, animal bioturbation has occurred, as a mean, in the upper 1.9 cm of the sediment, which is
equal to the aRPD (Table 2). On station 15 and 18, polychete tubes and burrows made the aRPD
to appear jagged. The greatest mean aRPD was recorded at 2.9 cm on station 15. The aRPD
varied between images at the same station (Table 2), e.g. at stations 18 and 22 the maximum
aRPD in some images were 2.7 to 2.8 cm deep in the sediment, whereas the minimum aRPD was
only 0.3 cm or less. This suggests a patchy distribution in redox conditions at similar depths. At
station 40, the sediment was black with four laminated layers, which indicates that periods of no
or minor bioturbation had occurred earlier. The white patch on that sediment surface appeared to
be a mat of sulphur bacteria, Beggiatoa spp.
The functional relationship between the mean RPD measured from cores and SPIs
(aRPD) was analysed using linear regression (Figure 4). The mean RPD of 0.6 cm compared to
the aRPD of 0.0 cm at station 40 contributes to an intercept of 0.21. This value should have been
close to the origin if the RPD was zero.
The mean BHQ index (Table 2) was greatest at station 15 (7.8) and lowest at station 40
(1.2). Low indices are indicative of environmental disturbance. Values < 4 are suggested to
indicate a disturbed fauna with the presence of the pioneering benthic successional stage I.
These index values were common at stations 22, 30, and 40. The BHQ indices at the other
stations were between 5.2 and 7.8, and were assigned to successional stage II, a transitory stage
between a pioneering community and an "equilibrium" stage community (III).
The content of total C and N in the surficial sediment was highest at stations 15 to 30 and
lowest at stations 10 and 40 (Table 2).
137
-------
Table 1. Macrobenthic species, abundance, and biomass (standard deviation n=3) in the
Koljofjord in August 2000.
Depth
(m)
10
15
18
22
30
40
Species
Gattyana cirrosa
Pectinaria koreni
Scalibregma inflatum
Euchone papillosa
Abra alba
Microdeutopus gryllotalpa
Malacoceros fuliginosus
Hydrobia spp.
Totals
Pseudopolydora antennata
Scalibregma inflatum
Euchone papillosa
Trochochaeta multisetosa
Eteone longa
Capitella capitata
Abra alba
Totals
Pseudopolydora antennata
Capitella capitata
Totals
Pseudopolydora antennata
Totals
Capitella capitata
Pseudopolydora antennata
Totals
Capitella capitata
Pseudopolydora antennata
Totals
Individuals
perm2
73
47
27
20
20
7
7
7
207
100
33
33
27
20
13
7
233
60
20
80
7
7
47
47
93
13
7
20
(S.D)
61
31
23
0
20
12
12
12
169
122
42
58
31
20
12
12
295
69
20
89
12
12
31
42
72
12
12
23
Biomass
perm2
1.46
0.75
1.25
0.02
1.55
0.04
0.02
0.04
5.13
0.05
0.60
0.07
0.01
0.14
0.02
0.07
0.95
0.05
0.07
0.13
0.01
0.01
0.21
0.11
0.32
0.07
0.01
0.07
(S.D)
2.0
0.8
1.3
0.0
2.5
0.1
0.0
0.1
6.8
0.1
0.6
0.1
0.0
0.2
0.0
0.1
1.1
0.1
0.1
0.2
0.0
0.0
0.2
0.1
0.3
0.1
0.0
0.1
138
-------
Figure 3. Sediment profile images from the six sampling depths. Contrasts are digitally
enhanced. SWI = sediment water interface, RPD = redox potential discontinuity, OX =
oxidised sediment, RED = reduced sediment, BU = burrow, ET = tube ofEuchone
papillosa, PT = tubes of Pseudopolydora antennata, BE = sulphur bacteria Beggiatoa
spp.
139
-------
Table 2. Sediment characteristics at various depths in the Koljofjord.
Depth (m)
10
15
18
22
30
40
Mean RPD
1.8
2.6
1.0
0.7
0.5
0.6
Mean aRPD
1.9
2.9
1.5
0.9
0.6
0.0
Max. aRPD
3.6
7.2
2.7
2.8
1.3
0.0
Min. aRPD
0.6
0.9
0.4
0.0
0.3
0.0
BHQ
5.2
7.8
6.0
3.7
4.0
1.2
C (%)
2.9
5.4
6.0
6.6
6.1
3.1
N (%)
0.3
0.6
0.6
0.7
0.7
0.3
RPD measured from cores and apparent RPD (aRPD) measured from the SPIs are presented as means (cm);
maximum and minimum aRPD are from individual images. Mean Benthic Habitat Quality (BHQ) indices are from
all images, and mean total carbon (C) and nitrogen (N) are from the top 0-1 cm sediment layer.
C
a.
ec
e
a
I
y * Q.rGx + 0.21
R- = O.B8
10m ,-'"
0
1 2
RPD
Figure 4. Linear regression of mean redox potential discontinuity (RPD) measured by electrodes
from cores, and mean apparent redox potential discontinuity (RPD) measured from
sediment profile images.
140
-------
DISCUSSION
Redox Conditions in the Sediment
Animals are dependent on dissolved oxygen for their respiration. Most of the
macrofauna on soft bottoms are buried in the sediment and pump oxygenated water down into
their burrows and tubes. Since dissolved oxygen penetrates only a few millimetres into the
sediment by molecular diffusion (Revsbech et al. 1980), animal irrigation is the main process
whereby dissolved oxygen is transported deep into the sediment. Bioturbation, irrigation, and
building of structures by the infauna are important activities in maintaining oxidised conditions
deep in the sediment. A diverse benthic fauna with many functional groups (Pearson and
Rosenberg 1987) has been shown to correlate with a deep aRPD (Nilsson and Rosenberg 2000).
In the sub-oxic zone, NO"3, Mn+4, and Fe+3 act as oxidising agents (Aller 1988), and may react
continuously with each other in bioturbated layers with high organic content (Santschi et al.
1990). In aquatic sediments, the concentration of Fe+3 frequently exceeds that of other electron
acceptors, and sulphur reduction only takes place when the other electron acceptors are
exhausted (Ibid).
In most studies, redox potentials in sediments are measured by inserting electrodes into
the sediment at different depths (e.g. Fenchel 1969). Thus, the recordings will be influenced by
the activity and construction of the animals. If the electrode is gradually pushed down vertically
from the sediment surface, the sediment structure and pore water content may be affected and
have an impact on the redox measurements. In the present study, we used pre-drilled holes in
the cores and inserted the electrodes horizontally. This allowed independent measurements with
several different electrodes, but a drawback was that it was not possible to make measurements
at the same distance from the sediment surface in different cores unless the sediment in the core
was pushed upwards. In this study, the vertical resolution was only 1 cm. To move the sediment
may, however, distort the redox conditions. Moreover, when sediment cores are collected, the
sediment is compacted, especially if the core diameter is small.
In the present study, linear regression analysis showed good agreement between
measurements of RPD and aRPD (Figure 4). Thus, the mean aRPD measured as the border
between rust-brown sediment and green-grey or black sediment in the images correlated to the
mean RPD in the cores (corrected with +240 mV). Others (Mortimer 1941, 1942, Fenchel 1969)
have also recorded a colour shift at this voltage. The greatest RPD values were recorded at
station 15, where the animal activity penetrated deepest into the sediment. The surficial
sediment at station 40 smelled of H^S, indicative of ferrous ions being exhausted (Mortimer
1941, 1942). The patch ofBeggiatoa spp. in one image at station 40 is indicative of the zone
between reduced sediment with H2S and overlying water containing at least some oxygen
(J0rgensen and Revsbech 1983, Rosenberg and Diaz 1993).
Organic enrichment of the sediment surface leads to increased oxygen consumption and
reduces the depth distribution of the RPD. High carbon content of the surficial sediment is
indicative of a high oxygen demand. A high sedimentation rate may also lead to increased
141
-------
oxygen demand in the near-bottom water, and this demand may be even higher than in the
surficial sediments (Rydberg et al. 1990). Temporal changes in oxygen concentrations have
been shown to correlate with changes in the depth distribution of the RPD (Rosenberg 1977).
Use of the SPI technique for assessing thickness of the RPD has many advantages. SPI
allows an in situ observation of the sediment. The width examined in the present study was 15
cm and the thickness of the sediment is dependent on the penetration depth of the prism
(maximum 30 cm). As seen in the images (Figure 3), the aRPD is not level, but undulates due to
animal structures, such as burrows and tubes. The minimum and maximum aRPD can vary over
several centimetres in one image at these oxygen stressed stations (Table 2). A digital
measurement of the area of the RPD is probably the best method to get an accurate assessment of
the mean depth distribution of the aRPD in a particular image. The possibility of digitally
enhancing the contrast in the images is an advantage. Smearing on the front plate of the prism
can, for some sticky sediments, cause problems in the interpretation of part of an image. Where
this is the case, that part of the image should be left-out of the analysis.
Faunal Response to Hypoxia
The critical oxygen concentration for the survival of coastal benthic fauna is around 0.7
ml/L (Nilsson and Rosenberg 2000). Based on the oxygen recordings in the Koljofjord it seems
that little to no fauna could have been present in September 1999 at < 15 m water depth. At 15
m, the oxygen concentration was 0.78 ml/L, and so only few if any species could have survived.
After this date, colonisation might have occurred, peaking in spring 2000 when higher oxygen
concentrations prevailed. By July, oxygen levels were again low at 30 and 40 m depths, e.g. <
0.6 ml/L. Few species can persist through such great changes in dissolved oxygen concentration.
The vertical distribution of fauna was restricted to the upper few centimetres of the
sediment. Such a narrow distribution has been found in other oxygen stressed fjords (Rosenberg
1977), with the fauna pushed upwards as the RPD layer depth becomes more shallow (Ankar and
Jansson 1973). The only sub-surface feeders in the Koljofjord were the polychetes Capitella
capitata and Scalibregma inflatum. The biomasses were very low and only exceeded lg/m2 at
the shallowest station. Sandnes et al. (2000) demonstrated that biomass correlated strongly with
sediment mixing rates. Thus, the impact on physical and chemical processes in the sediment
may be significantly reduced when the benthic fauna is continuously stressed by hypoxic/anoxic
events. Organisms recruited during periods of normoxia (> 2 ml/L) in areas that experience
hypoxia/anoxia tend to be smaller, opportunistic species that have life histories that can be
completed during the periods of normoxia. Larger bodied and long-lived species, which also
tend to be dominant bioturbators such as sea urchins, are likely to be eliminated by hypoxia after
settlement. Polychetes with opportunistic features were the dominants in the Koljofjord similar
to that recorded at other oxygen-stressed areas (Pearson and Rosenberg 1978). For example, in
Chesapeake Bay on the east coast of the United States, oxygen-stressed benthic communities
were dominated by polychetes, particularly opportunistic spionides (Llanso 1992).
142
-------
Pearson and Rosenberg (1978) state that several Polydora species and Capitella capitata
are rapid colonisers of defaunated areas. In the Koljofjord, Pseudopolydora antennata was the
most common species occurring at all stations < 15 m. Capitella sp. and specimens of the genera
Pseudopolydora have been found to be fast colonisers of defaunated sediments following
hypoxic events in Japan (Tsutsumi 1987). Tubes of P. antennata are clearly seen in the images,
and at low dissolved oxygen concentrations it is likely to be an advantage to have tubes
stretching higher up into the microgradients of oxygen. Similar behaviour has been shown for
several tube-building polychetes in Swedish fjords, e.g. Pectinaria koreni (Nilsson and
Rosenberg 1994), Euchone papillosa (Nilsson and Rosenberg 1997) andMelinna cristata
(Nilsson and Rosenberg 2000). The bivalve Abra alba and polychete Pectinaria koreni observed
in the shallow areas of the Koljofjord also were rapid colonisers following hypoxic events (Arntz
1981, Rosenberg and Loo 1988). The tolerance limit of A. alba to low dissolved oxygen
concentrations is around 0.8 ml/L (Rosenberg et al. 1991).
The benthic species found in the Koljofjord are probably not particularly tolerant of
hypoxia, and only Malacoceros fuliginosa is listed as a tolerant species in the review by Diaz
and Rosenberg (1995). Rather, the species observed have life history traits that facilitate rapid
colonisation of oxygen-stressed areas when conditions improve (Gray 1979). In several areas
along the Swedish and Norwegian Skagerrak coast, a temporal decline in oxygen concentrations
at the bottom has been demonstrated during periods of the 20th century (Rosenberg 1990,
Johannessen and Dahl 1996, Aure et al. 1996). The main reason for this decline was suggested
by the authors to be eutrophication. In enclosed and stratified areas, as exemplified by the
Koljofjord, such decline leads to an impoverished benthic fauna and unpredictable food supply
for demersal fish.
143
-------
REFERENCES
Aller, R.C. 1979. Relationships of tube-dwelling benthos with sediment and overlying water
chemistry. In: Marine Benthic Dynamics. K.R. Tenore and B.C. Coull (eds.),
University of South Carolina Press, Columbia, pp. 285-308.
Aller, R.C. 1988. Benthic fauna and biogeochemical processes in marine sediments: the role of
burrow structures. In: Nitrogen Cycling in Coastal Marine Environments. T.H.
Blackburn and J. S0rensen (eds.), J. Wiley and Sons, New Delhi, pp. 301-338.
Ankar, S., and B.O. Jansson. 1973. Effects of an unusual, natural temperature increase on a
Baltic soft bottom community. Marine Biology 18: 9-18.
Arntz, W. 1981. Zonation and dynamics of macrobenthos biomass in an area stressed by
oxygen deficiency. In: Stress Effects on Natural Ecosystems. G.W. Barrett and R.
Rosenberg (eds.), John Wiley and Sons, Chichester, pp. 215-225.
Aure, J., D. Danielssen, and R. Saetre. 1996. Assessment of eutrophication in Skagerrak coastal
waters using oxygen consumption in fjordic basins. ICES Journal of Marine Science 53:
589-595.
Bjorck, W. 1913. Bidrag till kannedom om nordhafsrakans (Pandalus borealis Kr.) utbredning
och biologi i Kattegatt och Skagerack. Svenska Hydrogrask-biologiska kommissionens
skrifter, 4: 1-11.
Diaz, R.J., and R. Rosenberg. 1995. Marine benthic hypoxia - review of ecological effects and
behavioural responses of marine macrofauna. Oceanography and Marine Biology
Annual Review 33: 245-303.
Fenchel, T. 1969. The ecology of marine microbenthos IV. Structure and function of the
benthic ecosystem, its chemical and physical factors and the microfauna communities
with special reference to the ciliated protozoa. Ophelia 6: 1-182.
Fenchel, T.M., and R.J. Riedl. 1970. The sulfide system: a new biotic community underneath
the oxidized layer of marine sand bottoms. Marine Biology 7: 255-268.
Gray, J.S. 1979. Pollution-induced changes in populations. Philosophical Transactions of the
Royal Society of London 286: 545-561.
Gustafsson, M., and K. Nordberg. 1999. Benthic foraminifera and their response to
hydrography, periodic hypoxic conditions and primary production in the Koljo fjord on
the Swedish west coast. Journal of Sea Research 41: 163-178.
144
-------
Help, C. 1995. Eutrophication and zoobenthos dynamics. Ophelia 41: 113-136.
Jenkins, M.C., and W.M. Kemp. 1984. The coupling of nitrification and denitrification in two
estuarine sediments. Limnology and Oceanography 29: 609-619.
Johannessen, T., and E. Dahl. 1996. Declines in oxygen concentrations along the Norwegian
Skagerrak coast, 1927-1993: a signal of ecosystem changes due to eutrophication?
Limnology and Oceanography 41: 766-778.
Josefson, A.B., and R. Rosenberg. 1988. Long-term soft-bottom faunal changes in three
shallow fjords, west Sweden. Netherlands Journal of Sea Research 22: 149-159.
J0rgensen, B.B., and N.P. Revsbech. 1983. Colorless sulfur bacteria, Beggiatoa spp. and
Thiovolum spp., in ©2 and JH^S microgradients. Applied and Environmental
Microbiology 45: 1261-1270.
Llanso, R.J. 1992. Effects of hypoxia on estuarine benthos: the lower Rappahannock River
(Chesapeake Bay), a case study. Estuarine and Coastal Shelf Science 35: 491-515.
Lyle, M. 1983. The brown-green color in marine sediments: a marker of the Fe(III)-Fe(II) redox
boundary. Limnology and Oceanography 28: 1026-1033.
Mortimer, C.H. 1941-42. The exchange of dissolved substances between mud and water in
lakes. Journal of Ecology 29, 30: 280-329, 147-201.
Nilsson, H., and R. Rosenberg. 1994. Hypoxic response of two marine benthic communities.
Marine Ecology Progress Series 115: 209-217.
Nilsson, H.C., and R. Rosenberg. 1997. Benthic habitat quality assessment of an oxygen
stressed fjord by surface and sediment profile images. Journal of Marine Systems 11:
249-264.
Nilsson, H.C., and R. Rosenberg. 2000. Succession in marine benthic habitats and fauna in
response to oxygen deficiency: analysed by sediment profile imaging and by grab
samples. Marine Ecology Progress Series 197: 139-149.
Pearson, T.H., and R. Rosenberg. 1978. Macrobenthic succession in relation to organic
enrichment and pollution of the marine environment. Oceanography and Marine Biology
Annual Review 16: 229-311.
Pearson, T.H., and R. Rosenberg. 1987. Feast and famine: structuring factors in marine benthic
communities. The 27th Symposium of The British Ecological Society Aberystwyth
1986. J.H.R. Gee and P.S. Ciller (eds.). Blackwell Scientific Publications, Oxford, pp.
373-395.
145
-------
Revsbech, N.P., J. S0rensen, and T.H. Blackburn. 1980. Distribution of oxygen in marine
sediments measured with microelectrodes. Limnology and Oceanography 25: 403-411.
Rhoads, D.C., and J.D. Germano. 1986. Interpreting long-term changes in benthic community
structure: a new protocol. Hydrobiologia 142: 291-308.
Rosenberg, R. 1977. Benthic macrofaunal dynamics, production, and dispersion in an oxygen-
deficient estuary of west Sweden. Journal of Experimental Marine Biology and Ecology
26: 107-133.
Rosenberg, R. 1990. Negative oxygen trends in Swedish coastal bottom waters. Marine
Pollution Bulletin 21: 335-339.
Rosenberg, R., and RJ. Diaz. 1993. Sulfur bacteria (Beggiatoa spp.) mats indicate hypoxic
conditions in the inner Stockholm Archipelago. Ambio 22: 32-36.
Rosenberg, R., B. Hellman, and B. Johansson. 1991. Hypoxic tolerance of marine benthic
fauna. Marine Ecology Progress Series 79: 127-131.
Rosenberg, R., and L.O. Loo. 1988. Marine eutrophication-induced oxygen deficiency: effects
on soft bottom fauna, western Sweden. Ophelia 29: 213-225.
Rydberg, L., L. Edler, S. Floderus, and W. Graneli. 1990. Interaction between supply of
nutrients, primary production, sedimentation and oxygen consumption in SE Kattegat.
Ambio 19: 134-141.
Sandnes, J., T. Forbes, R. Hansen, B. Sanders, and B. Rygg. 2000. Bioturbation and irrigation
in natural sediments, described by animal-community parameters. Marine Ecology
Progress Series 197: 169-179.
Santschi, P., P. Hohener, G. Benoit, and M. Buchholtz-ten. 1990. Chemical processes at the
sediment-water interface. Marine Chemistry 30: 269-315.
Tsutsumi, H. 1987. Population dynamics of Capitella capitata (Polychaeta; Capitellidae) in an
organically polluted cove. Marine Ecology Progress Series 36: 139-149.
Yingst, J.Y., and D.C. Rhoads. 1980. The role of bioturbation in the enhancement of microbial
turnover rates in marine sediments. In: Marine benthic dynamics, K.R. Tenore and B.C.
Coull (eds.). University of South Carolina Press, Columbia, pp. 407-421.
146
-------
FISH RESPONSES TO ORGANIC MATTER LOADING
AND TO HYPOXIA IN SHALLOW EUTROPHIC LAKES
Arvo Tuvikene1, Ain Jarvalt1, Reet Laugaste1, and Ervin Pihu1
ABSTRACT
Discharge of nutrients into the lakes Vortsjarv and Peipsi has increased since the 1950s
up to the beginning of the 1990s, and this nutrient loading has influenced fish communities in
these lakes. There has been an increase in the abundance of fish species favored by
eutrophi cation and a decrease offish species sensitive to eutrophi cation. Two types of seasonal
oxygen depletion have been recorded in these lakes: in Lake Peipsi summertime nocturnal
hypoxia, and in Lake Vortsjarv hypoxia in early spring under ice. Some cases of serious fish kill
due to hypoxia are registered in these lakes. Due to a decrease in the nutrient content, the
eutrophi cation process within the lakes has slowed down, but the probability offish kill events
are still a major concern in Lakes Vortsjarv and Peipsi.
INTRODUCTION
Dissolved oxygen less than 2 mg/L is harmful to most aquatic organisms. In inland water
bodies, when substances with high oxygen demand reduce oxygen concentration, oxygen can
become a factor limiting the abundance offish. In a few local cases, hypoxia can sometimes be
due to a natural disaster, but the increased areas of water affected by hypoxia in recent times are
mostly due to human activities. Over-enrichment of lake water with nutrients, especially
phosphorus, stimulates outbreaks of algae blooms that consume oxygen from the water when
decomposing. During the winter months the remarkable reduction of light by ice, and especially
snow cover, leads plants to consume more oxygen than they produce. In small lakes, winter fish
kills occur rather frequently, but in large lakes kills are rare.
Fish respond to hypoxia in a variety of ways; one way is by maximizing oxygen uptake
and another is by economizing oxygen use (Randall 1982, Yoshikawa et al. 1995). Survival time
of organisms during exposure to environmental stresses that limit energy availability is directly
related to the degree of metabolic depression achieved (Hand 1996); therefore, feeding is
strongly affected by reduced oxygen levels (Kramer 1987). There are two main strategies used
by fish to transport oxygen in blood (Perry and McDonald 1993): (1) more active fish species
like salmonids utilize low-affinity hemoglobin in conjunction with high arterial hemoglobin
saturation; and (2) more sluggish fish like cyprinids appear to utilize a relatively high-affinity
hemoglobin in conjunction with low arterial hemoglobin saturation. The first group offish has a
lower ability to extract oxygen from water compared to the other. However, salmonids usually
inhabit well-aerated waters. On the other hand, cyprinids and other relatively inactive fishes
have a high efficiency of oxygen extraction from water and, therefore, they have a
^imnological Station, Institute of Zoology and Botany, Estonian Agricultural University, 61101 Rannu, Tartu
County, Estonia.
147
-------
better chance to survive in an oxygen-depleted environment. Of course, freely mobile fish can
usually avoid hypoxia by moving into more oxygenated water (Bejda et al. 1987, Kramer 1987).
Eutrophication is considered one of the most serious problems in many fresh and marine
waters around the world. Nutrient loads into Estonia's largest lakes, Lake Peipsi and Lake
Vortsjarv have become one of the main local concerns. Discharge of nutrients into these lakes
has increased steadily since the 1950s into the early 1990s. These nutrient loads have affected
the abundance and species composition offish communities of both lakes.
Investigations in the drainage area of these lakes (i.e., the drainage area of the Narva
River) have recorded some areas in these large lakes with seasonally depleted oxygen levels (< 2
mg/L). In these areas, there are mainly two types of oxygen depletion. In Lake Peipsi, under
certain hydrometeorological conditions (low water level, prolonged calm and hot weather, strong
water bloom) summer night hypoxia can occur (Pihu and Kangur 2000). In Lake Vortsjarv,
hypoxia takes place mainly under ice in early spring, especially during years of low water level
(Jarvalt and Pihu 2002).
There is much public concern about the reduction in fish abundance, especially
commercially important species, in Lakes Peipsi and Vortsjarv. The decline offish stocks may
be partly due to over-fishing, or partly due to habitat degradation resulting from eutrophication.
During recent years, fish stocks can no longer support previous fishing levels in either lake (Pihu
1998, Pihu and Kangur 2000).
The objective of this paper is to summarize the knowledge of the responses offish
communities affected by eutrophication and irregular oxygen depletion in the large lakes of
Estonia. Lakes Peipsi and Vortsjarv have been studied for several decades, but early data during
hypoxic and anoxic events are lacking. Only since the late 1990s has more frequent and
complete data appeared to prove the fish kills are due to hypoxia.
Water Quality in Lakes, Eutrophication, and Oxygen Depletion
The Narva River drainage basin (area 56,225 km2) contains two large, shallow lakes,
Lake Peipsi and Lake Vortsjarv (Figure 1). Lake Peipsi (3,555 km2) is located on the Estonian-
Russian border and is the fifth largest lake in Europe. Lake Peipsi is relatively shallow (mean
and maximum depth 7.1 m and 15.3 m, respectively) and its three parts (L. Peipsi s.s. sensu
stricto, i.e. in a restricted sense, L. Lammijarv, L. Pihkva) each have a different trophic status.
Lake Peipsi is intensively aerated and very seldom suffers from oxygen depletion. Due to its
large area and shallow depth, stratification is generally limited to a few months during
summertime, and is rather unstable. In its larger northern part, L. Peipsi s.s. is mesotrophic to
moderately eutrophic (mean content of Ntotai about 700mg/m3, Ptotai 35 mg/m3) with a mean water
transparency of 2 m. The southern parts, Lakes Lammijarv and Pihkva, are highly eutrophic
(mean values of Ntotai above 1000 mg/m3, Ptotai63 mg/m3, transparency 1 m) (Starast et al. 1999).
The cyanobacteria Gloeotrichia echinulata, Anabaenaflos-aquae and Aphanizomenon flos-
aquae (all strong N-fixers) are the main dominants in the lakes causing water blooms in summer
and early autumn. The diatoms Aulacoseira islandica and Stephanodiscus binderanus dominate
148
-------
in the cool period. Phytoplankton biomass does not track the dynamics of the water nutrient
concentrations since the dominating cyanobacteria can fix nitrogen, and some of them, such as
Aphanizomenon flos-aquae and Planktothrix agardhii, are able to store phosphorus in their cells
in large quantities. Therefore, they are not nutrient limited (Konopka 1989). Phytoplankton
blooms are usually caused by hydrophysical and weather conditions.
GULF OF FINLAND
Figure 1. Lake Vortsjarv and Lake Peipsi as objects of concern in Estonia.
Sample sites, ordered by stream direction:
1 - L. Vortsjarv; 2 - Emajogi R. at Tartu; 3 - Emajogi R., mouth;
4 - L. Peipsi, Praaga; 5 - L. Peipsi, middle; 6 - Narva R., head;
7 - Narva reservoir; 8 - Narva R., mouth; 9 - Narva Bay;
10 - Gulf of Finland, littoral.
The oxygen conditions in Lake Vortsjarv are generally favorable for fishes due to its
shallowness and to wind action (area 270 km2, mean and maximum depth 2.8 m and 6.0 m,
respectively). The content of nutrients in the lake fluctuates from eutrophic to hypertrophic;
mean values being eutrophic for total phosphorus (45-55 mg/m3) and hypertrophic for nitrogen
(1200 mg/m3). Phytoplankton biomass is mainly light-limited due to its high abundance and lake
detritus content, which cause a low transparency (about 0.3-1 m in vegetation period). The main
dominants (Limnothrix redekei, L. planktonica, Planktolyngbya limneticd) are filiform
cyanobacteria that cannot fix N2. N-fixers (cyanobacteria Aphanizomenon spp.) dominate in the
low water level years when the N:P ratio is < 10 (Huttula and Noges 1998) due to intensive
internal P-loading, and during the clear water period in June (Anabaena lemmermannif).
149
-------
The major non-point sources of nutrients into these lakes are fertilizers and animal
manure. Major point sources for Lake Peipsi are wastewater from the cities of Tartu and Pskov.
According Loigu etal. (1999), the modeled nitrogen and phosphorus loads entering Lake Peipsi
are 19,000 metric tons and 580 metric tons per year, respectively. Seven percent of the external
nitrogen load originates from point pollution sources, half of the load comes from agriculture,
and 22% is of natural origin. Of the phosphorus load, 36% comes from point pollution sources
and 38% from agriculture. Generally, the nutrient content decreases from upstream of the Narva
River drainage area to downstream (Figure 2). This is because the lakes in this drainage area
behave as nutrient traps. Agriculture in the Baltic countries changed drastically at the beginning
of the 1990s. Due to poor financial conditions, the use of mineral fertilizers dropped by a factor
of two from 1992-1999 (Status of Estonian Environment 2000). Although the use of fertilizers
has decreased considerably since 1990, no clear decrease in the nutrient concentrations has so far
been detected in the drainage area of the Narva River. However, since the early 1990s, the
nutrient load from the catchment area into Lake Peipsi has decreased (Suits and Jaani 2000); a
clear decline in Ntotai has occurred, and to a lesser extent a decrease in Ptotai has also been seen. A
significant decline in the N:P ratio was observed, and the low N:P ratio is favorable for
cyanobacteria.
o"
456
Sample sites
10
Figure 2. Nutrients and Chi a in the Narva River basin, August 2000 (See Figure 1 for sample
site locations).
Eutrophication resulting in low oxygen levels, coupled with certain hydrometeorological
conditions, is an important water quality concern in this area. In Lake Peipsi, hypoxia sometimes
happens on summer nights during hot and windless weather. Low oxygen content is usually
observed in Lake Vortsjarv from February through April, and it is primarily caused by low water
level and extended periods of ice cover.
150
-------
General Effects of Eutrophication on the Fish Communities of the Lakes
A number of properties make fish one of the most relevant groups of biota as a biological
indicator of eutrophication. Fishes occupy the functional link between trophic levels in aquatic
ecosystems. The composition offish communities directly indicates certain properties of a water
body, e.g. through the productivity gradient.
Dissolved oxygen availability may affect the ecology of fishes more through the
availability of energy for locomotion, growth and reproduction than through its direct effect on
survival (Kramer 1987). Habitat shifts and the use of alternative breathing modes, such as air
breathing or aquatic surface respiration, not only alter oxygen availability, but also influence the
risk of predation and food availability.
The Narva River drainage area is a relatively productive ecosystem that provides a
variety of resources, including fishes, recreation, and water supplies. The total annual fish catch
in Lake Peipsi has been fluctuating considerably, between 6,300 and 15,100 metric tons, over the
last 60 years, depending mostly on the abundance of dwarf smelt (Pihu and Kangur 2000). In
Lake Vortsjarv, the annual catch has been fluctuating between 111 and 677 metric tons during
this same period. According to the fishery classification, Lake Peipsi is a "smelt-bream-
pikeperch" lake, while Lake Vortsjarv is a "pikeperch-bream" lake.
There are 34 permanent fish species living in Lake Peipsi and 32 in Lake Vortsjarv (Pihu
1998, Pihu and Kangur 2000), but in this paper we will only discuss briefly the changes in the
most commercially important species.
According to Rask and Peltonen (1999), it is possible to use indicator fish species to
measure the eutrophi cation level. The abundances of indicator species in the large lakes of
Estonia have changed historically. In general, there has been an increase in the abundance of
fish species favored by eutrophi cation, like bream (Abramis bramd) and pikeperch (Sander
lucioperca), and a decrease for fish species sensitive to eutrophication, like whitefish (Coregonus
lavaretus maraenoides), vendace (Coregonus albuld) and burbot (Lota lota) (Haberman and
Jarvalt 1984, Krause and Palm 2000, Pihu and Kangur 2000) (Figures 3A, 3B, and 3C). In Lake
Vortsjarv natural eutrophication was already apparent at the beginning of the 20th century;
however, whitefish and vendace were still commercially important fish species at the end of the
1930s and 1950s, respectively. Today these latter two species are very seldom seen in the lake
(Figures 3 A, 3B). On the contrary, the abundance of pikeperch started to increase in the early
1960s, partly due to eutrophication and partly due to changes in fishing practices (Figure 3C)
(Pihu 1998).
151
-------
60
50
40
CO
o
15
D
c
c
10
Whitefish
Years
9nn -,
1RD
c/) 1 fin
.2 140
-C 1 on
o '^u
IS inn
o luu
Qfl
(0 °U
3 Rn
<^ AC\
9n
„
1 i JH
n oo o
T> CO U
3) O) C
Vendace
n
In
flnl II
TCDOCMinoot-Tri^ococDrocMinoo
TinmcDCDCDi^i^i^oooooooororoo)
DO5O5O5O5O5O5O5O5O5O5O5O5O5O5O5
Years
ftnn -,
ynn
w ROD
5
- ^nn
^
rn 400
o ^uu
c
c onn
<^ ^uu
n
L
c
c
i
D
O
7>
Pikeperch
llnlnll nil
CD in CD co
co in in CD
O) O) O) O)
1
i-
c
c
I
1
J
Btt
T- in CD
D i^ r^ i^
J5 O5 O5 O)
Years
i 1
1
1
co r^ T-
00 00 O5
O) O) O)
I
in
O)
O)
c
c
c
D
D
D
Figure 3. Annual fish catch from Lake Vortsjarv, 1935-1999.
152
-------
In Lake Peipsi, the commercially important fish species that have experienced reduced
abundance due to eutrophication are also whitefish, vendace, and burbot. There has been a
corresponding remarkable rise in the number of pikeperch as a result of the emergence of several
strong pikeperch generations. The decrease in the number of vendace was probably connected
with warm winters and bad ice conditions in the late 1980s and early 1990s, which affects the
development of the embryos. The main spawning places of whitefish are located in the southern
part of Lake Peipsi s.s. where the lake bottom is naturally covered by sand or gravel. However,
due to eutrophication the suitable spawning areas are diminishing as the gravelly and sandy
bottom is covered by mud.
The stock of whitefish in Lake Peipsi has never been large. It has been historically
suppressed mainly by overfishing and high water temperature in summer, and now by shortage
of spawning places, eutrophication and parasites (Krause and Palm 2000). According to
Shirkova (1974) the hearts of whitefish in Lake Peipsi are broadly infected with larvae of the
parasite Tetracotyle intermedia.
Pikeperch has been reported to be favored in eutrophic lakes (Erm 1981). Due to changes
in fishing practices and accelerated eutrophication, the stock of pikeperch has been growing over
the last 15-18 years in Lake Peipsi. According to Lappalainen and Lehtonen (1995), the year-
class strength of pikeperch correlated positively with the mean summer water temperature and
negatively with wind speed from certain directions. Increased turbidities have been suggested to
provide better survival conditions for pikeperch (Erm 1981). Good indicators of the trophic
status of lakes are also littoral fish, like sculpin (Coitus gobio) and gudgeon (Gobio gobio). In
Lake Vortsjarv, the abundances of these latter two species have dropped over the last 30 years
due to eutrophication (Haberman and Jarvalt 1984).
Hansson and Rudstam (1990), Hendrikson (1991) and Lappalainen etal. (2001) have
shown that there have also been the changes in fish populations and fish catches in the
eutrophicated marine areas of the Gulf of Finland. Their studies indicated that sea trout (Salmo
trutta), whitefish, ide (Leuciscus idus), burbot and northern pike (Esox lucius) all have become
more rare, whereas roach (Rutilus mtilus), white bream (Blicca bjoerkna), and ruffe
(Gymnocephalus cernuus) have become more common.
Although eutrophication reduces nocturnal dissolved oxygen levels, the effect is not
normally lethal, however, fish growth rate may be reduced. Fishes may choose not to remain in
oxygen-depleted water and rise to the surface to breathe the oxygenated surface layer.
Laboratory studies have shown that the growth rate for juvenile winter flounder
(Pseudopleuronectes americanus) exposed to low and diurnally fluctuating levels of dissolved
oxygen were reduced as compared to that for juvenile winter flounder exposed to a continually
high level of dissolved oxygen (Bejda et al. 1992).
According to Kerr and Ryder (1988), water quality in lakes is an important predictor of
fish yield. However, based on studies of Ranta and Lindstrom (1993) in Finnish lakes, fishing
effort was more important in affecting fish yield than was water quality. Their conclusion is in
accordance with the results of our large lake studies where the changes in fishing practices have
had substantial impacts on fish yield (Pihu 1998).
153
-------
Behavior of Fish During Hypoxia
Avoidance reactions of fishes to hypoxia have been studied by many authors. According
to Kramer (1987) in the case of hypoxia, fish can use the following behavioral responses: change
in activity, increase use of air breathing, increase use of aquatic surface respiration, and vertical
or horizontal habitat changes. A rather common behavioral response reported many times is an
increase in general activity and movement toward the water surface where higher oxygen
concentrations are expected (Bejda et al. 1987, Poulin et al. 1987, Freon and Misund 1999).
This is common during summertime, but also seen in winter when lakes are covered with ice.
Weltzier with co-authors (1999) looked at the avoidance reaction of inland silverside (Menidia
beryllind) in laboratory experiments. In these low dissolved oxygen experiments, the fish larvae
began swimming at high speed, four times faster than the maximum speed observed in well-
oxygenated water, and the larva always moved in an upward direction. According to Magnuson
et al. (1985) fishes also try to move upward in hypoxic ice-covered lakes.
On the other hand, hypoxia can depress swimming activity in some fishes. According to
Crocker and Cech (1997), juvenile white sturgeon (Acipenser transmontanus) decreased overall
energy expenditure during hypoxia via reduction in spontaneous swimming activity in order to
increase survival during widespread or prolonged environmental hypoxia. It is also reported that
the activity of guppies (Poecilia reticulatd) increased during hypoxia if they had surface water
access, but decreased if surface access was not possible (Weber and Kramer 1983). In a recent
unpublished study in our laboratory, bream (Abramis brama\ during exposure to an oxygen
level around 1 mg/L at 10°C, increased the use of aquatic surface respiration if they had surface
access. Many fish species show seasonal changes in low oxygen tolerance and they are usually
most sensitive during summer and less in spring (Hlohowskyj and Wissing 1987).
To detect how catch per unit of effort (CPUE) depends on the oxygen content in water,
fish were regularly caught from Lake Vortsjarv with monitoring gill nets in winter/spring 1994-
1995 (high water level) and 1995-1996 (low water level). The acquired data showed that CPUE
is negatively affected by water level and is also influenced by oxygen content (Figure 4). The
mean CPUE was 3.6-fold greater in the high water level year (1994-1995) compared with the
low water level year (1995-1996). During the second half of winter, fish became less active and
this was more pronounced during the low water level year (Figure 4). There were small trends
toward lower CPUE at decreased oxygen content. Based upon trawling results just after ice
break-up, we concluded that fish stayed in the same area and did not leave. They regulated their
activity downward to economize the use of oxygen. Similar results have been obtained by other
authors; for example, Casselman (1978) showed that pike (Esox lucius) catch per unit effort
under the ice decreased in case of extreme hypoxia.
154
-------
A
10
~
LU
CL
O
-CPUE
Dissolved oxygen
• Poly. (CPUE)
16
14
+ 12
10
+ 8
~-6
4
+ 2
0
co co
O)
B
CPUE
Dissolved oxygen
Poly. (CPUE)
0
35027 35039 35057 35067 35079 35094 35118 35145 35156 35170
Figure 4. Catch per unit effort (CPUE) and water oxygen content during winter/spring 1994-
1995 (A, high water level) and 1995-1996 (B, low water level). Poly. (CPUE) - 4-order
polynomial trendline.
155
-------
Lake Peipsi: Outbreaks of Water Bloom and Summer Kills
Calm weather, high temperature (>25°C), high pH (9), and nutrient enrichment are the
essential combination of conditions for phytoplankton blooms in the lake. The most commonly
known algal bloom events in Lake Peipsi involve blooms of cyanobacteria, which occur almost
every summer.
Cases of serious fish kills have been recorded in the lake. In July 1959, the kill of dwarf
smelt (Osmerus eperlanus eperlanus morpha spirinchus) in Lake Pihkva was obviously due to
nightly oxygen depletion. There was a heavy water bloom oiAphanizomenon flos-aquae,
Anabaena flos-aquae and Microcystis aeruginosa during this warm summer (Semenova 1966);
unfortunately, the environmental parameters were not determined.
In July 1972, another kill of smelt occurred in Lake Pihkva and the southern part of Lake
Peipsi s.s. The long-lasting, calm, warm weather caused high water temperatures (over 28°C),
producing a heavy cyanobacterial bloom (Gloeotrichia echinulata, Anabaena flos-aquae and
Aphanizomenon flos- aquae) and severe nightly oxygen depletion (Kuderskij and Fedorova
1977). In July 1988, in the southern part of Lake Peipsi s.s., a dwarf-smelt kill was probably due
to the cyanobacterial bloom of Planktothrix agardhii and Limnothrix redekei, both indicators of
hypertrophy. The daytime pH was about 9 and oxygen saturation at 26%.
In May 1989, a kill of dwarf smelt occurred in the northern part of the lake under the ice;
pH measured at ice breakup was over 10. The bloom of Aulacoseira islandica was the probable
cause of the rise in pH and increased NH3 concentration.
Why were only the dwarf smelt dying during the above cited hypoxic events? Most
probably there is more than one reason, and these kills seem to be complex phenomena. One
factor that may favor the high mortality of dwarf smelt is that this species was infected with the
parasite Tetracotyle in large numbers in Lake Pihkva (up to 100%). Dwarf smelt in Lake Peipsi
were also infected, but not to the same extent as in Lake Pihkva. This parasite can damage the
heart of the fish (Shirkova 1966) and, thus, negatively affect blood circulation. Dwarf smelt is a
schooling fish, and on the one hand the high adaptive significance of schooling is important in
the antipredatory, feeding, and migratory behavior of this species, as well as diminishing the
energetic expenditure for swimming (Pavlov and Kasumyan 2000). On the other hand, the level
of dissolved oxygen is more crucial for schooling fish, especially for large schools with high
packing density (Freon and Misund 1999). According to MacFarland and Moss (1967), there
can be depletion of oxygen inside densely packed schools, especially in the rear part. It may also
be that schools of dwarf smelt inhabit places close to the lake bottom where most probably the
greatest nocturnal oxygen depletion occurs.
Unfortunately, the environmental parameters, especially oxygen content, were not
determined immediately after the fish kill, but were measured some days later. For this reason,
the conclusions drawn are based only on indirect signs. Fish kills under the ice are usually
visible only after ice breakup, sometimes even later, and it is impossible to reconstruct exactly
the winter conditions producing the kill.
156
-------
Lake Vortsjarv: Winter Kills During Low Water Levels
According to fishermen and data from the Vortsjarv Limnological Station, fish kills in
Lake Vortsjarv occurred in 1939, 1940, 1948, 1964, 1967, 1969, 1978, and 1996 (Kirsipuu and
Tiidor 1987, Noges and Noges 1998). Most kills happened in late winter (in March), and were
most probably caused by lack of oxygen under the ice when snow cover was thick and
phytoplankton photosynthesis was lacking. In summer 1964 a massive fish kill occurred, but the
probable cause was a heavy rainfall leading to the inflow of pesticides from surrounding fields
along with the wastewater from a starch factory. In April-May and June 1987, a serious fish kill,
mainly pikeperch and eel (Anguilla anguilld), occurred on the eastern side of the lake. In June,
the cause was probably the runoff of pesticides from surrounding fields, in April-May winter
oxygen depletion (Kirsipuu and Tiidor 1987).
In April 1996, an eel kill occurred in the narrow and shallow southern part of Lake
Vortsjarv, caused by serious hypoxia (< 0.3 mg/L O2) that was measured over a two week period
in the lake (Figure 4). However, we did not notice other fish species dying. According to
different estimates, 10-201 of eel died in that kill (Noges and Noges 1998). The most probable
cause was that the eel were trapped in the ice due to the extremely low water level and thick ice
cover.
Use of the Lake Vortsjarv outlet dam to regulate the water level to minimize oxygen
depletion is being considered.
CONCLUSIONS
Due to a decrease of nutrient content in the Narva River drainage, the eutrophication
process has been slowed, but the question remains: can we expect the occurrence of large fish
kills in the future? Probably due to high water temperature in combination with certain
meteorological and climatic factors, massive mortality of smelt and other fishes can still occur in
Lake Peipsi. Predictions for Lake Vortsjarv are much more problematic because low water level
in combination with thick snow cover can also produce oxygen depletion. Eutrophi cation
remains a major concern in Lakes Vortsjarv and Peipsi, despite that the eutrophi cation process
has been slightly reduced.
ACKNOWLEDGEMENTS
This study was supported by the Estonian Science Foundation (grant No. 3694) and US
EPA Grant No. CR 827711-01. We thank Robert V. Thurston for correcting the language of the
manuscript.
157
-------
REFERENCES
Bejda, A.J., B.A. Phelen, and A.L. Studholme. 1992. The effect of dissolved oxygen on the
growth of young- of-the-year winter flounder, Pseudopleuronectes americanus.
Environmental Biology of Fishes 34: 321-327.
Bejda, A.J., A.L. Studholme, and B.L. Olla. 1987. Behavioral responses of red hake, Urophycis
chuss, to decreasing concentration of dissolved oxygen. Environmental Biology of Fishes
19: 261-268.
Casselman, J.M. 1978. Effects of environmental factors on growth, survival, activity, and
exploitation of northern pike. American Fisheries Society Special Publication 11: 114-
128.
Crocker, C.E., and JJ. Cech, Jr. 1997. Effects of environmental hypoxia on oxygen consumption
rate and swimming activity in juvenile white sturgeon, Acipenser transmontanus, in
relation to temperature and life intervals. Environmental Biology of Fishes 50: 383-389.
Erm, V. 1981. Pikeperch.Valgus, Tallinn, 128 pp. (in Estonian).
Freon, P., and O.A. Misund. 1999. Dynamics of pelagic fish distribution and behaviour: effects
on fisheries and stock assessment. Fishing News Books. Great Britain. The University
Press, Cambridge, pp. 348.
Haberman H., and A. Jarvalt. 1984: Vortsjarve kalastiku kvalitatiivsetest ja kvantitatiivsetest
muutustest. In: J. Haberman (ed.). Vortsjarve okosiisteemi seisund. Eesti NSV Teaduste
Akadeemia, Tartu, pp. 106-112 (in Estonian).
Hand, S.C. 1996. Downregulation of cellular metabolism during environmental stress:
Mechanisms and Implications. Annual Review of Physiology 58: 539-563.
Hansson, S., and L.G. Rudstam. 1990. Eutrophication and Baltic fish communities. Ambio 19
(3): 123-125.
Hendrikson, S.-H. 1991. Effects offish farming on natural Baltic fish communities. Marin
Aquaculture and Environment 22: 85-104.
Hlohowskyj, I, and T.E. Wissing. 1987. Seasonal changes in low oxygen tolerance of fantail,
Etheostomaflabellare, rainbow, E. caeruleum, and greenside, E. blennioides, darters.
Environmental Biology of Fishes 18: 277-283.
Huttula, T., and T. Noges (eds.). 1998. Present state and future fate of Lake Vortsjarv. Tampere,
150pp.
158
-------
Jarvalt, A., and E. Pihu. 2002. Influence of water level on the stocks and catches offish in Lake
Vortsjarv. Proc. Estonian Acad. Sci. Biol. Ecol. (in press).
Kerr, S.R., and R.A. Ryder. 1988. The applicability offish yield indices in freshwater and
marine ecosystems. Limnology and Oceanography 33: 973-981.
Kirsipuu, A., and R. Tiidor. 1987. Koige eest tuleb maksta. Eesti Loodus 12: 807-811 (in
Estonian).
Konopka, A. 1989. Metalimnetic cyanobacteria in hardwater lakes: buoyancy regulation and
physiological state. Limnology and Oceanography 34: 1174-1184.
Kramer, D.L. 1987. Dissolved oxygen and fish behavior. Environmental Biology of Fishes 18:
81-92.
Krause, T., and A. Palm. 2000. Peipsi whitefish, Coregonus lavaretus maraenoides Poljakow, in
Lake Peipsi. Proceedings of the Estonian Academy of Sciences. Biology. Ecology 49: 91-
97.
Kuderskij, L.A., and G.V. Fedorova. 1977. Snizhenie zapasov snetka v bol'shikh vodoemakh
Severo-ZapadaEvropejskoj chasti SSSR v 1973-1975 gg. In: Rybokhozyajstvennoe
izuchenie vnutrennikh vodoemov, 20: 3-8 (in Russian).
Lappalainen, J., and H. Lehtonen. 1995. Year-class strength of pikeperch (Stizostedion
lucioperca L.) in relation to environmental factors in shallow Baltic Bay. Annales
Zoologici Fennici 32: 411-419.
Lappalainen, J., M. Rask, H. Koponen, and S. Vesala. 2001. Relative abundance, diet and growth
of perch (Percafluviatilis) and roach (Rutilus rutilus) at Tvarminne, northen Baltic Sea,
in 1975 and 1997: responses to eutrophication. Boreal Environment Research 6 (2): 107-
118.
Loigu, E., A. lital, S. Karlova, U. Leisk, K. Pahel, U. Suits, M. Trapido, A. Vassiljev, and I.
Veldre. 1999. Pollution load and water quality of rivers in the basin of Lake Peipsi. In:
Lake Peipsi. E. Pihu and A. Raukas (eds.). Environmental Information Centre. Ministry
of the Environment of Estonia, Tallinn, 233-234 (in Estonian).
Mac Farland, W.N., and S.A. Moss. 1967. Internal behaviour in fish schools. Science 156: 260-
262.
Magnuson, J.J., A.L. Beckel, K. Mills, and S.B. Brant. 1985. Surviving winter hypoxia:
behavioral adaptations of fishes in northern Wisconsin winterkill lakes. Environmental
Biology of Fishes 14: 241-250.
159
-------
Noges, P., and T. Noges. 1998. The effect of fluctuating water level on the ecosystem of Lake
Vortsjarv, Central Estonia. Proceedings of Estonian Academy of Sciences. Biology.
Ecology 47, 2: 98-113.
Pavlov, D.S., and A.O. Kasumyan. 2000. Patterns and mechanisms on schooling behavior in
fish: A review. Journal of Ichthyology 40 (2): 163-231.
Perry, S. F., and G. McDonald. 1993. Gas exchange. In: The Physiology of Fishes, D. H. Evans
(ed.). CRC Press. Boca Raton, Florida, USA: 251-278.
Pihu, E. 1998. Fishes and fisheries management in Lake Vortsjarv. Limnologica 28 (1): 91-94.
Pihu, E., and A. Kangur. 2000. Main changes in the ichthyocoenosis of Lake Peipsi Since the
1950s. Proceedings of the Estonian Academy of Sciences 49: 81-90.
Poulin, R., N.G. Wolf, and D.L. Kramer. 1987. The effect of hypoxia on the vulnerability of
guppies (Poecilia reticulata, Poeciliidae) to an aquatic predator (Astronotus ocellatus,
Cichlidae). Environmental Biology of Fishes 20: 285-292.
Randall, DJ. 1982. The control of respiration and circulation in fish during exercise and
hypoxia. Journal of Experimental Biology 100: 275-288.
Ranta, E., and K. Lindstrom. 1993. Theory offish yield versus water quality in lakes. Annales
Zoologici Fennici 30: 71-75.
Rask, M., and H. Peltonen. 1999. Fish monitoring in lakes in the Nordic countries. In:
Proceedings of a workshop in monitoring of large lakes. A. Peltonen and M. Viljanen
(eds.). University of Joensuu. Publications of Karelian Institute, pp.193-200.
Semenova, N.I. 1966. Vliyanie osennego razvitiya diatomovykh vodoroslej na gidrokhimicheskij
rezhim yugo-vostochnoj chasti Chudskogo ozera. In: Hydrobiological Researches, 4.
Tallinn, 71-74 (in Russian).
Shirkova, A.P. 1966. Epidemic of Tetracotyle among dwarf smelt in Lake Pihkva. In: E. Pihu
and Mjating (eds.). Hydrobiology and fisheries of Lake Peipsi-Pihkva. Valgus, Tallinn,
323-325 (in Russian).
Shirkova, A.P. 1974. The Chudskoy whitefish. Izv. GosNIORKh 83: 131-135 (in Russian).
Starast, H., T. Mols, A. Lindpere, and A. Milius. 1999. Hiidrokeemia. In Peipsi. E .Pihu. and A
Raukas. ( eds.). Keskkonnaministeeriumi Info-ja Tehnokeskus, Tallinn, pp. 56-65.
Status of Estonian Environment. 2000. Keskkonnaministeeriumi Info-j a Tehnokeskus. Tallinn
2000: pp. 96 (in Estonian).
160
-------
Suits, U., and A. Jaani. 2000. Hydrological peculiarities and ecological situation of Lake Peipsi.,
In: Nordic Hydrological Conference 2000, vol. 1. Nordic Association for Hydrology,
Uppsala, Sweden. T. Nilsson (ed.), Nordic Programme Report 46:185-190.
Weber, J.-M., and D.L. Kramer. 1983. Effects of hypoxia and surface access on growth,
mortality, and behavior of juvenile guppies, Poecilia reticulata. Canadian Journal of
Fisheries and Aquatic Sciences 40: 1583-1588.
Weltzier, F.-A., K.B. D0ving, and W.S. Carr. 1999. Avoidance reaction of yolk-sac larvae of the
inland silverside Menidia beryllina (Atherinidae) to hypoxia. The Journal of
Experimental Biology 202: 2869-2876.
Yoshikawa, H., Y. Ishida, K. Kawata, F. Kawai, and M. Kanamori. 1995.
Electroencephalograms and cerebral blood flow in carp, Cyprinus carpio, subjected to
acute hypoxia. Journal of Fish Biology 46, 114-122.
161
-------
162
-------
HYPOXIA/ANOXIA IN LAKE VORTSJARV, ESTONIA
Lea Tuvikene1'2, PeeterNoges1'2, Tiina Noges1'2
ABSTRACT
The consequences of hypoxia and anoxia in certain regions or in the whole water column
of Lake Vortsjarv, Estonia, have been analyzed. Lake Vortsjarv is a large, shallow lake located
in southern Estonia. The most dramatic consequences of hypoxia/anoxia in Lake Vortsjarv have
been fish kills, although it has not always been clear whether low oxygen conditions or some
other factor caused these. Phosphorus supply in the water column is another significant factor
connected to the oxygen conditions in the Lake. The upper 10-cm layer of sediments in the Lake
is usually well oxygenated, and phosphorus is mainly bound to iron (III). This finding is
supported by an Fe/P mass ratio of 26-30. Due to the large surface area, the Lake is strongly
affected by water level and wind. Big storms during low water periods have caused the release
of noteworthy amounts of soluble, reactive phosphorus into the water column from resuspended,
deeper, anoxic sediments. Summer stratification, which occurs on some hot and windless days,
supports oxygen depletion by bacteria at the sediment surface and increases phosphorus release
from the sediments.
INTRODUCTION
Oxygen conditions in shallow lakes are usually good and rarely become unfavorable for
lake biota. Still, under certain circumstances, hypoxia or even anoxia can appear in some regions
or in the whole water column of a shallow lake. In temperate lakes, thick, long-lasting ice and
snow cover can cause serious oxygen deficiency. This frequently leads to fish kills in small,
shallow lakes (Scheffer 1998). In large, shallow lakes, the problem arises only when freezing
takes place down to very near the bottom. In addition to lowering water levels, thereby reducing
water volume, other factors or combination of factors can cause fish kills, e.g. lethal
concentrations of toxicants or oxygen deprivation caused by eutrophication.
Oxygen status is an important factor in determining whether phosphorus is either
entrapped by or released from sediments. An aerobic upper layer of sediments acts as a
phosphorus trap for a lake, since phosphorus in that layer is usually bound to insoluble iron (III)
calcium and phosphate complexes (Wetzel 1983, Bostrom etal. 1985, Lofgren 1987, Scheffer
1998). These complexes are reduced to soluble iron (II) phosphate and calcium phosphate under
anaerobic conditions, causing phosphorus release to the water column. Phosphorus liberation
Estonian Agricultural University, Institute of Zoology and Botany, Vortsjarv Limnological Station, Rannu, Tartu
County, Estonia, 61101. 2University of Tartu, Institute of Zoology and Hydrobiology, 46 Vanemuise St., Tartu,
Estonia, 51014.
163
-------
from aerobic sediments is also possible (Lofgren 1987). Specifically, in the case of a low
sediment Fe/P molar ratio, adsorption sites on the iron (III) surfaces become saturated with
phosphorus and, therefore, the bonding of phosphorus to sediments is restricted. Another way
for phosphorus to be released from aerobic sediments to the water column is by formation of
anaerobic microzones at the sediment-water interface due to microbial activity enhanced by
increased temperature (Ibid). The phosphorus retention capacity of sediment is increased by
calcareous water (Lofgren 1987, Scheffer 1998).
Shallow lakes with a large surface area are greatly influenced by wind. Big storms
during low water periods can stir-up the bottom sediments, and substantial enrichment of the
water column with phosphorus from the resuspended, deeper, anoxic sediments takes place.
Summer stratification, though quite rare in shallow lakes, may appear during some hot and
windless days. During intensive microbial respiration, oxygen is used-up and the sediment-water
interface becomes anoxic.
STUDY SITE AND DATA BASE
Lake Vortsjarv is the second largest lake in the Baltic states (mean depth 2.8 m,
maximum depth 6 m, surface area 270 km2, drainage area 3374 km2) (Figure 1). The lake is
greatly affected by large fluctuations in the water level (mean annual amplitude 1.4 m), and is
covered with ice on the average of 135 days a year, from the end of November until the end of
April. The Secchi depth reading is 0.5 - 1.0 m during the ice-free period, principally limited by
stirred-up bottom sediments. The lake is eutrophic with total nitrogen concentrations at 1-2
mg/L and total phosphorus about 50 |ig/L. The total number of bacteria ranges from <1 to 6
million cells/mL, and the chlorophyll a concentration ranges from almost 0 to 84 mg/L3.
Normally, the whole water column of Lake Vortsjarv is saturated with dissolved oxygen (D.O.).
The lake is sensitive to changes in the watershed and climate. Low-water periods are especially
dangerous to the ecosystem, causing an increase in sediment resuspension, acceleration of
nutrient cycling, and an improvement of light conditions in the water column. Winter oxygen
depletion can occur due to a significantly lower oxygen storage capacity and a great amount of
easily degradable organic matter that had been produced during the preceeding vegetation period
(Huttula and Noges 1998). The lowest water level in recorded history (since the 1880s) was
registered in the lake in 1996, and in the winter 1995-1996 an anoxic water column was observed
in Lake Vortsjarv for the first time.
For our analysis of oxygen conditions in Lake Vortsjarv, we used the state monitoring
database for the lake, available since 1968. We also used data from the scientific database of the
Vortsjarv Limnological Station of the Institute of Zoology and Botany of Estonian Agricultural
University.
164
-------
GULF OF FINLAND
Figure 1. The location of Lake Vortsjarv.
RESULTS AND DISCUSSION
Occurrence of Hypoxia/Anoxia in Lake Vortsjarv
Small Water Volume, Extended Ice Cover, and Thick Snow Layer. According to long-
term data, winter anoxia is very rare in the surface layer of water in Lake Vortsjarv (Figure 2).
Still, low winter D.O. values <5 mg/L in the bottom water layer occur almost every year. Winter
oxygen values <1 mg/L were common in the 1970s after the occurrance of maximum summer
phytoplankton biomasses, which lead to high oxygen consumption in the sediments. Occasional
low winter oxygen levels in the bottom water layer were also registered in the 1980s.
^
t£*
U
10 -
A
I
i """"•"•-- ---L._ T j T T
— ^---"--p--^
i
i i i i i
Nov
Dec
Jan
Feb
March
Apr
Figure 2. Average (± standard deviation) winter dissolved oxygen concentration in the surface water
layer of Lake Vortsjarv, 1968-1997.
165
-------
On 18 March 1996, for the first time during 30 years of measurements, total anoxia was
registered in the whole water column throughout large areas of the lake (Figure 3). Oxygen
conditions improved by the end of March due to inflowing snow melt waters. The massive
anoxia event was the result of several concurrent circumstances — a small amount of water, a
cold winter, and thick ice- and snowcover. Oxygen reduction reached its maximun rate, 670
mg/m2/day, whereas the mean oxygen decrease rate under the ice cover was 100 mg/m2/day
(Figure 4).
15 --
10 --
5 --
0
Reserve
Deficiency rate
800
-- 600
"- 400
-- 200
.- 0
-200
0/j
0
Q
1995
Date
Figure 3. Oxygen reserve and deficiency rate in Lake Vortsjarv in winter 1995-1996.
400 -
-a o
t.
200
-400
-600
-i"
Nov-Dec Dec-Jan Jan-Feb Feb-Mar Mar-Apr
Figure 4. Yearly average (± standard deviation) oxygen deficiency rate in Lake
Vortsjarv, 1968-1997.
166
-------
High Temperature and Lack of Wind. Summer stratification is not common in Lake
Vortsjarv, but it may appear and last if there is a succession of very hot and windless days
(Figure 5). Oxygen is used-up during intensive microbial respiration, and the sediment-water
interface becomes anoxic, causing rapid phosphorus release to the overlying water. For
formation of anoxic zones at the sediment-water interface, one day of stratification has been
shown to be enough (Lofgren 1987).
10 -
_i 8 -
O)
£ 6-
9 4-
2 -
0 -
r; , absol. value of vertical temperature difference
1 \ L vertical oxygen difference
,; •.
/ I A AJI
X^M / VU - ^
\ : ..!-..: ^-jf, ^T\.-^ V 1'^^'- ^i. f^^**-"^ -«-»-'J""s. , "I . _~L- --»•
cd cd o *-" *-" *-" >^>^S^^ £>f) p( +^ +^ ^
C3 C3 Q^ ,^w ,^w MH C^ C^ 3 H-j H-j 3 0^ JZ JZ O
*""* *""* rr | ^> ^> -^^ ^J ^J ^~> 1 i <^ ^/^ ^^ ^^ ^7"
/
/
1 1
O
Q
oo
(N
(N
(N
(N ^
(N
O
(N
Figure 5. Difference between bottom and surface water dissolved oxygen
concentration as a function of temperature in Lake Vortsjarv, January-
December 1995.
Summer Microbial Activity. Anoxic (micro)zones at the sediment surface can appear
even when the oxygen concentration in the bottom water layer is high. This is caused by high
microbial activity in summer, and may result in release of a considerable amount of sediment
phosphorus to intersticial water (Lofgren 1987).
Consequences of Hypoxia/Anoxia in Lake Vortsjarv
Fish Kills. Thick, long-lasting ice cover frequently leads to hypoxic conditions in small
shallow waters, killing virtually all fish life (Scheffer 1998). Lake Vortsjarv is quite a large lake,
and winter hypoxic conditions do not always result in such drastic consequences. Fish kills in
Lake Vortsjarv have been documented in the years 1939, 1948, 1967, 1969, 1978, 1987
(Kirsipuu and Tiidor 1987), and 1996 (unpublished data), but never did all the fish die. In 1969
and 1978, the possible reason for fish kills, besides the oxygen deficiency, may have been high
CO2 concentration in the water (50 mg/L at the bottom) resulting from the decomposition of
167
-------
organic matter (Kirsipuu and Tiidor 1987). In 1987, no connection between the fish kill and
oxygen conditions was observed. The main cause was concluded to be pollutant runoff from
agriculture (Kirsipuu and Tiidor 1987).
Minimum tolerable D.O. concentration was considered 2-2.5 mg/L for bream (Abramis
brama), and 3.5 mg/L for pike-perch (Sander luciopercd) (Kirsipuu and Tiidor 1987). Recent
laboratory experiments have shown that fish can survive at much lower D.O. concentrations, e.g.
the estimated l-hourLC50at 5, 15, and 25°C for pike perch, 1+year old, was 0.5, 0.7, and 1.1
mg O2/L, respectively (A. Tuvikene, unpublished data). The winter anoxia extending to the
whole water column in most parts of the Lake Vortsjarv in 1996 did not cause large fish kills.
Probably, most of the pelagic fishes could find some refuge near the river mouths where oxygen
conditions were better, but the oxygen depletion did cause a large kill of eel (Anguilla anguilld)
(10-20 metric tons according to different opinions), which dig into mud for hibernation.
Phosphorus Release from the Sediment-Water Interface. Processes at the sediment-water
interface are extremely complex. Here, oxygen conditions are very important in phosphorus
immobilization or release. Phosphorus is usually bound as insoluble iron (III) phosphate only
under aerobic conditions, although other factors such as pH, alkalinity and redox potential
impact this binding.
In Lake Vortsjarv, the Fe/P mass ratio in sediments of 26-30 (Huttula and Noges 1998) is
sufficient to bind phosphorus. The upper 10 cm sediment layer is usually aerobic, and
phosphorus there is bound mainly to iron (III) (Noges and Kisand 1999). Still, phosphorus
retention capacity of aerobic sediments depends greatly on temperature, pH, and water hardness.
In summer, in addition to high temperature, pH is usually low due to intensive primary
production, and this reduces the capability of aerobic sediments to bind phosphorus. On the
other hand, the water of Lake Vortsjarv is four to eight times oversaturated with calcite during
the vegetation period, and settled calcite raises the phosphorus buffering capacity of sediments
(Starast 1982).
Heavy storms, especially during low water conditions, will resuspend a thick sediment
layer and a considerable amount of phosphorus can then be released to the water column from
the deeper anoxic sediments. For example, on 19 September 1996, at the lowest recorded water
level, right after a north wind of 15 m/second, the concentration of soluble, reactive phosphorus
was increased 23 times compared to that before the storm (from 6 |ig/L to 140 |ig/L). The rise of
total phosphorus was 4.9 times.
SUMMARY
Oxygen conditions in Lake Vortsjarv are usually favorable for lake biota. Reduced
oxygen conditions are not always the only cause offish kills in the lake. Lake Vortsjarv acts
normally as a phosphorus sink due to the well-oxygenated upper layer of sediments and a
favorable Fe/P mass ratio (26-30). Large amounts of soluble, reactive phosphorus can
168
-------
occasionally be released from sediments to the water column. This mostly happens when deeper
anoxic sediment layers are mixed into the water column during storms, or when the sediment
surface becomes anoxic during summer stratification.
ACKNOWLEDGEMENTS
This research was supported by the Estonian Science Foundation Grant No. 4080, by the
5th Framework Program of the European Commission (Contract EVK l-CT-1999-00039), and by
the United States Environmental Protection Agency Grant No. CR 827711-01-0.
REFERENCES
Bostrom, B., I. Ahlgren, and R.T. Bell. 1985. Internal nutrient loading in a eutrophic lake,
reflected in seasonal variations of some sediment parameters. Internationale Vereinigung
fur theoretische und angewandte Limnologie. Verhandlungen, 22, 3335-3339.
Huttula, T., and T. Noges (eds.). 1998. Present state and future fate of Lake Vortsjarv. Results
from Finnish-Estonian joint project in 1993-1997. Pirkanmaa Regional Environment
Centre, Tampere, 150 pp.
Kirsipuu, A., and R. Tiidor. 1987. We must pay for everything. Estonian Nature, pp. 806-811
(in Estonian).
Lofgren, S. 1987. Phosphorus retention in sediments - implications for aerobic phosphorus
release in shallow lakes. Acta Universitatis Upsaliensis. Comprehensive Summaries of
Uppsala Dissertations from the Faculty of Science, Uppsala, 21 pp.
Noges, P., and A. Kisand. 1999. Forms and mobility of sediment phosphorus in shallow,
eutrophic Lake Vortsjarv (Estonia). Internationale Revue der gesamten Hydrobiologie,
84(3), 255-270.
Scheffer, M. 1998. Ecology of Shallow Lakes. Chapman & Hall, London, 357 pp.
Starast, H. 1982. Regularities of phosphate and nitrate content dynamics of the Vortsjarv Lake
basin, In: Eesti NSV jarvede niiiidisseisund (The nowadays state of lakes of Estonian
SSR). H. Haberman, K. Paaver and H. Simm (eds.). Eesti NSV Teaduste Akadeemia.
(in Estonian).
Wetzel, R.G. 1983. Limnology. 2nd edition. Saunders College Publ., Philadelphia.
169
-------
170
-------
RESPONSES OF BENTHIC COMMUNITIES TO HYPOXIA IN A SUB-TROPICAL
ENVIRONMENT: PROBLEMS AND HYPOTHESES
Rudolf Wu1
ABSTRACT
Hypoxia affects several thousand hectares of aquatic habitat around the world, and has
caused mass mortality of marine animals, benthic defaunation, and decline in fisheries
production in many places. In Hong Kong, eutrophication has resulted in regular occurrence of
hypoxia and defaunation in Tolo Harbor. The benthic community, however, has been restored to
its original state by rapid winter recolonization. The abundance and dominance of predatory
gastropods show an increase along a gradient of hypoxia in the Harbor, reflecting changes in the
trophic structure in relation to oxygen. No significant changes in percentage of deposit-feeders,
however, were found along the same gradient. We hypothesize that predators are more oxygen
demanding, and therefore would be more sensitive to hypoxia, while deposit feeders are more
tolerant. Field manipulation experiments and multivariate statistics were used to examine
recolonization and succession of benthos in defaunated sediments, and also the time for
recovery. Again, rapid recolonization was found after defaunation. No significant difference in
abundance or species richness was observed between defaunated sediment and the natural
benthic community after 15 months, suggesting that a stable community had been achieved
within a short time. Often, the occurrence of hypoxia in the natural environment is associated
with elevated levels of ammonia, hydrogen sulphide, and paniculate organic materials. The
interactions between hypoxia and these water quality parameters, however, are poorly known.
The inability to isolate effects from individual factors and their interactions makes it difficult to
attribute the observed ecological effects to hypoxia.
INTRODUCTION
Hypoxia is defined as dissolved oxygen less than 2.8 mg O2fL (equivalent to 2 ml O2fL or
91.4 mM) (Diaz and Rosenberg 1995). Hypoxia can be a natural phenomenon caused by vertical
stratification such as the formation of haloclines and thermoclines (Rosenberg etal. 1991, Pihl et
al. 1992, Hoback and Barnhart 1996). More often, however, hypoxia is caused by eutrophi cation
and oxygen-demanding organic pollution (Pihl et al. 1992, Dalla Via et al. 1994, Peckol and
Rivers 1995, Gamenick et al. 1996, Sandberg 1997, Wu and Lam 1997, Aarnio et al. 1998,
Mason 1998). Nowadays, hypoxia or anoxia affecting thousands of km2 has commonly been
reported in North and South America, Africa, Europe, India, South-east Asia, Australia, Japan,
and China (Nixon 1990, Diaz and Rosenberg 1995, Wu 1999). Mass mortality of fishes and
marine animals, defaunation of benthic populations, and decline in fisheries production, caused
Biology & Chemistry Department, City University of Hong Kong, Tat Chee Ave., Kowloon, Hong Kong.
171
-------
by hypoxia, have been reported in many parts of the world (Baden et al. 1990, Diaz and
Rosenberg 1995, Lu and Wu 2000, Wannamaker and Rice 2000). In some marine systems with
extremely limited water exchange and receiving excessive anthropogenic inputs of nutrients (e.g.
Gulf of Finland and the central areas of the Baltic, Black, and Caspian Seas), bottom water has
become permanently hypoxic or anoxic.
An increase in nutrient level has clearly occurred in coastal waters all over the world
during the last few decades (Table 1). This increase is primarily attributable to intensive farming,
application of fertilizers, deforestation, and discharge of domestic wastewaters. For example,
analysis of organic matter in sediment cores from the Chesapeake Bay have shown a marked
increase in organic carbon (35-50%) since 1934. Analyses of lipid biomarker distribution
revealed a substantial change in the sources of the deposited organic matter. The observed
changes coincide well with fertilizer application and human population growth in the watershed,
and also with the onset of eutrophic and hypoxic conditions in Chesapeake Bay (Zimmerman and
Canuel 2000). Results of this study further suggest that anthropogenic activities may exert a
substantial influence on carbon cycling processes in coastal systems. Gabric and Bell (1993)
estimated that the magnitude of global anthropogenic flux of N and P is comparable to that of
natural flux. There is little doubt, therefore, that such a huge nutrient input has contributed much
to eutrophication and hypoxia in marine coastal systems worldwide.
Table 1. Increase in nutrients in coastal waters reported from various places around the world in
recent decades.1
Period
1970-90
1970-90
1960-80
1930-80
1950-90
1930-90
1970-90
Location
North Sea, and coasts
of China and Japan
Coast of Germany
Black Sea
Dutch Sea
Wadden Sea
Queensland
Baltic Sea
Element and number
Of times increased
Both P and N
N1.7
P 20, N 5
P2,N4
P4,N3
P5,N3
P 1.5, N 4.5
JData compiled from GESAMP (1990), Moss et al. (1992), Vollenweider et al. (1992), Gabric and Bell (1993),
Bell and Elmetri (1995), De-Jonge et al. (1996).
172
-------
There is good scientific evidence to support that frequency of occurrence, severity, and
area affected by hypoxia have all increased in the last two decades. For example, the area of
hypoxia in the northern Gulf of Mexico has increased from 9,000 km2 (1985-1992) to 16,000-
20,000 km2 (1993-1999). Likewise, areas of the East China Sea affected by hypoxia were less
than 1,000 km2 in 1980, and had increased to 13,700 km2 in 1999 (Rabalais 2001). Indeed,
hypoxia caused by eutrophication and organic pollution is now considered to be among the most
pressing water pollution problems in the world (GESAMP 1990, Goldberg 1995, Wu 1999).
Unfortunately, there are very good reasons to believe that the problem of marine hypoxia is
likely to increase. First, the world population is expected to double by 2020, and most of that
population growth will occur near coastal cities (Garble and Bell 1993, UNEP 1991). It is
unlikely that sewage treatment facilities can catch up with this rapid population growth, and
hence further increase in nutrients loadings can be expected into coastal waters. Second, further
increase in the use of fertilizers, deforestation and release of nitrogen oxides into the atmosphere
is expected (Nixon 1990). Third, global warming caused by the green house effect will warm-up
surface water more quickly, thus promoting the formation of thermoclines. Freshwater runoff
will also likely increase due to increased urbanization and intensive agriculture. This will not
only add to nutrient loading, but also enhance the formation of haloclines (Justic etal. 2001).
Changes in Community Structure
Benthic communites are relatively long-lived and stable, and their condition is often
indicative of environmental conditions and pollution. Mass mortality of benthos and fishes over
large areas due to hypoxia has been reported in marine coastal areas all over the world, and
sensitive species have been permanently or periodically removed in many places (Wu 1982,
Dauer 1993, Diaz and Rosenberg 1995). Tolerances of a variety of benthic species to hypoxia
and anoxia have been well documented (Rosenberg et al. 1991, Diaz and Rosenberg 1995). In
general, the critical dissolved oxygen concentration for survival of most benthic organisms is
around 2.8 mg C>2/L, while certain species can tolerate 0.5 to 1 mg C>2/L for several days to
weeks (Rosenberg 1980). Polychetes, for example, are among the marine organisms most
tolerant to stress associated with organic loading and low oxygen, and are, therefore, often used
as environmental indicators for hypoxia and organic pollution (Levin 2000). Thus, hypoxia may
eliminate sensitive species (or animal groups) but encourage the proliferation of a few tolerant
species in a community (Dauer 1993).
In coastal marine waters, hypoxia causes a major change in species composition. Pihl
(1994) demonstrated that hypoxia in the Kattegat, Sweden caused significant, long-term changes
in the diet of demersal fishes, and he further related this to changes in species composition of
benthic macrofauna. In contrast, a study by Sagasti et al. (2000) showed that many epifaunal
species have high hypoxia tolerance, and epifaunal communities in areas of Chesapeake Bay
exposed to brief hypoxic episodes and moderate hypoxia (0.5 to 2 mg O2/L) can persist with
little change in species composition and abundance.
Hypoxia does not only change species composition, but also alters community structure
and decreases both species diversity and species richness in benthic communities (Wu 1982,
Dauer et al. 1992, Dauer 1993). There is also a general tendency for suspended feeders to be
replaced by deposit feeders, and macrobenthos to be replaced by meiobenthos and metazoans
173
-------
(Diaz and Rosenberg 1995, Levin 2000). In Chesapeake Bay, reduced dominance by
equilibrium species (e.g. long-lived bivalves and maldanid polychaetes) was observed in
hypoxia-affected areas (Dauer etal. 1992). Dauer (1993) reported alteration of biomass
distribution amongst species groups, with less biomass of deep-dwelling and equilibrium species
and more biomass of opportunistic species found in hypoxic areas.
Changes in Trophodynamics
Most studies have reported changes in community structure in response to hypoxia.
Relatively few studies have been carried out to study effects of hypoxia on trophodynamics.
Tolo Harbor is an embayment to the northeast of Hong Kong and connects with the outer Bay
(Mirs Bay) by a long, narrow channel (Figure 1). Because of the "bottle-necked" topography,
the flushing rate is poor and the water residence time in the inner Harbor is ca. 28 to 42 days
(Wu 1982). In the 1980's, the inner Harbor received untreated sewage from 0.2 million people
and slurry from 21,000 pigs, with a total BOD loading of 2,870 tonnes per year. Results of a
long-term monitoring program showed a clear decrease in dissolved oxygen going from the
Channel to the Harbor (Wu and Lam 1997). A monthly trawl survey was carried-out for 2 years
at five stations to study the temporal and spatial changes in the epibenthic community along the
hypoxic gradient (Wu 1982) (see Figure 1.). The results showed that the number of animals and
species, biomass and species diversity (Shannon's function and evenness) all were higher at
stations in the Channel, where oxygen levels were higher, than at stations in the inner Harbor,
where the water was hypoxic. Along an gradient of increasing hypoxia, a decrease in dominance
and abundance of predatory gastropods (Murex trapa, Nassarius crematus, N. siguinjorensis,
Turricula nelliae) was found along the hypoxic gradient (Figure 2). However, no significant
change in the percentage of deposit feeders was found along the hypoxia gradient. In
Chesapeake Bay, Breitburg et al. (1994) also showed that hypoxia altered the absolute and
relative importance of predators offish larvae, thereby causing significant changes in trophic
pathways.
•
t-.f
J"
WLs-
HUSH HAY
c?
^
~
« i*
./ l/f
~
,,
"-9,
V
/
Figure 1 . Location and topography of Tolo Harbor and the hypoxic gradient. Lines indicate five
trawl stations.
174
-------
Normoxiu
Hypo\ia
Predator
Tttrrifuki
m
Deposit Feed el's
Figure 2. Changes in dominance of predatory gastropod species along the hypoxic
gradient in Tolo Harbor. No change in deposit feeding species was observed
along this same gradient.
Mcio benthos
[Demersal fish
[, Macro be n I lies
[Su\ pen d c'd feed er%
i Diversity
[Species richness
[Larger body «iiie
[ Predators?
Lfr-sclettccl species?
\l ic rofla-gcl tales,
Smaller body size
r- selected species?
-Shorter food chain?
chai
in
Figure 3. A generalized diagram showing changes in community structure and function expected to occur
along a hypoxic gradient. Possible changes, which warrant further study, are also shown and
marked with a question mark.
Hypothetically, replacement of K-selected species by r-selected species (MacArthur and
Wilson 1967) and a complex food chain by a simple food chain should occur in communities
under hypoxic stress, although these hypotheses have not been tested in previous studies.
Known general and possible changes in community structure and function along a hypoxic
gradient are summarized above in Figure 3.
175
-------
Recovery of Benthic Community
Only a few studies have been carried out to test how long it takes for benthic
communities to recover from hypoxia, even though this question is highly relevant to
environmental management. Boesch and Rosenberg (1981) opined that recovery of benthic
communities following hypoxia depends primarily on the severity and duration of hypoxia, as
well as the complexity and composition of the community. In the Gullmarsfjord, for example,
hypoxia (0.28 mg O2/L) eliminated all macrobenthic fauna below 115m, and the community was
still not restored 18 months after the collapse (Josefson and Widbom 1988). Likewise, there was
no evidence that the nematode assemblage structure was returning to its pre-hypoxia state 1 year
after this hypoxic event (Austen and Widbom 1991).
On the inner continental shelf of New Jersey, USA, a mass mortality of the benthic
community induced by hypoxia covered an area of 8,000 km2 for 2 months, but was rapidly
recolonized by pelagic larvae of opportunistic species (Falkowski et al. 1980). However, many
previously dominant species failed to return within 1 year after hypoxia, and it was further
suggested that several years are required for recovery of benthic communities (Boesch and
Rosenberg 1981). The study in the Kattegat, Sweden showed little to no recovery in the benthic
community 2 years after defaunation caused by hypoxia (Rosenberg etal. 1992). Likewise, only
36% of initial benthic biomass was recorded 3 years after a hypoxia-induced benthic defaunation
in the northern Adriatic Sea (Stachowitsch 1991). A study on benthic recovery after a major
hypoxic event in Pomeranian Bay, southern Baltic Sea, showed that recolonization was still not
complete 2 years later at sites suffering from severe hypoxia. At sites moderately affected,
species composition and abundance returned to normal conditions within 2 years, but total
biomass was still lower (Powilleit and Kube 1999). In a review, Diaz and Rosenberg (1995)
concluded that recovery of a disrupted community is a lengthy process and usually takes several
years. They further postulated that no large system has fully recovered after development of
persistent hypoxia or anoxia. The only exception may be recovery on a smaller scale, for
example, where point source discharges cease, and recovery is initiated from surrounding, non-
affected areas (Rosenberg 1976).
A 2-year field study carried-out in a subtropical region, Tolo Harbor, Hong Kong,
showed that hypoxia caused regular mass mortality of epibenthos in the summer. The benthic
community was, however, soon restored to its original state by rapid winter recolonization,
although it is not clear whether the re-established community reached an equilibrium stage (Wu
1982).
An experimental study was carried out by Lu and Wu (2000) to study the patterns of
recolonization and succession of macrobenthic infauna in defaunated sediment, with a view to
determine the time required for the macrobenthic infauna to reach a stable, established
community after defaunation caused by hypoxia (Figure 4). Natural sediment was defaunated,
placed in trays, and these were placed randomly at subtidal of a pristine site. Five trays were
sampled monthly for the subsequent 15 months after deployment. In parallel, five grab samples
were taken monthly from natural sediment at the same site. Species composition, abundance,
176
-------
dominant species and diversity (H' and J) were determined at each sampling for the defaunated
and natural sediments, and then compared against each other using univariate statistics and
multidimensional scaling (MDS). Analysis of similarities (ANOSIM) was also carried-out to
test the differences in community composition between sample groups (Clarke and Warwick
1994).
Natural Airlrie-tlort month ^ Drfautiaietl
n t
5
Per MBHilt l
15 nwnllu
uiict1, Npects cum post nn
Diversity, Evenness, Species richness
Bray Cfui'tii Similarity, way ANOSIM i
Figure 4. Experimental design for studying benthic recovery after defaunation (Lu and
Wu 2000).
Initial recolonization occurred rapidly in the sediment trays within the first month after
their deployment. The results showed that benthic composition in the defaunted sediment
became 60% similar to that of the natural community within 5 months, and 94% of species
occurring in the natural community were also found in the defaunated sediment within 15
months. No significant difference could be found in abundance, species number, diversity and
species composition between the defaunated sediment and natural sediment after 4 to 10 months.
The results indicated that, by and large, a stable community was reached less than 15 months
after defaunation, although minor variations in species composition were still discernible at 15
months the between defaunated and natural sediments. Boesch and Rosenberg (1981) postulated
that the resilience of a benthic community, after hypoxia, depends on the constituent species, life
cycles, and reproductive patterns. The rapid recovery of the marine macrobenthic infauna in this
experimental study (Lu and Wu 2000) also agreed well with the results of our recovery field
study in Tolo Harbor, Hong Kong (Wu 1982). It is hypothesized that benthic recovery from
hypoxia in subtropical and tropical environments may be more rapid than their temperate
counterparts. Such rapid recovery may be attributed to the differences in community
composition and life cycles of species between temperate and sub-tropical benthos (Alongi 1989,
1990). Lesser inter-species competition in defaunated sediment may also allow more species to
colonize (or coexist). Further research on recolonization of marine biota following hypoxic
events is required, since such information is not only of great ecological interest, but also of
considerable practical use in environmental impact assessment and coastal management.
177
-------
Problems
Under natural conditions, many responses of benthic communities to hypoxia observed in
the field are confounded by other factors. For example, it is well known that hypoxia caused by
organic pollution is often associated with elevated levels of ammonia, hydrogen sulphide and
particulate organic matter (Figure 5). Apart from the stress effects contributed by each
individual factor, responses of benthic population and communities to hypoxia may be
significantly confounded by the interactions of these factors. For example, the isopod Saduria
entomon is able to synthesize haemocyanin to enhance oxygen uptake during hypoxia (1.8 - 2.9
mg O2/L), but haemocyanin synthesis is inhibited when the isopod is exposed to hypoxia in the
presence of hydrogen sulphide (150 (jM/L) (Hagerman and Vismann 1993). Magaud et al.
(1997) found that mortality of rainbow trout (Oncorhynchus mykiss) exposed simultaneously to
0.5 mg/L un-ionized ammonia and hypoxia (1.7 mg/L) was significantly higher than when
exposed to ammonia or hypoxia at these same levels alone, suggesting a synergistic effect.
Multiple regression analysis showed that hypoxia primarily affects patterns of species richness,
while organic enrichment mostly affects dominance and evenness of polychetes (Levin 2000). In
general, interactions between hypoxia and confounding factors commonly occurring in the
environment are poorly known. Factorial design laboratory experiments should be carried-out to
test the interactions of hypoxia and the likely confounding factors. Multivariate statistics may
also be used to determine the importance of these interactions under field situations.
Orsumte
\asi
Pollution
^articulate
Organic
'
Ammonia
Hvpoxia ;•-..
/
. •- t /
_,..-- Hydrogen
sulphide
i Biological/ecological cllcc-ls \
Svncr$>i stic? Additive?
•* •e.'-'
Figure 5. Some possible confounding factors of hypoxia resulting from organic pollution.
178
-------
ACKNOWLEDGEMENTS
Part of this study was supported by a research grant from the Research Grant Council of
the Hong Kong Government (No. 9040187).
REFERENCES
Aarnio, K., E. Bonsdorff, and A. Norkko. 1998. Role of Halicryptus spinulosus (Priapulida) in
structuring meiofauna and settling macrofauna. Marine Ecology Progress Series 163:
145-153.
Alongi, D.M. 1989. Ecology of tropical soft-bottom benthos: A review with emphasis on
emerging concepts. Review of Biology in the Tropics 37: 85-100.
Alongi, D.M. 1990. The ecology of tropical soft-bottom benthic ecosystems. Oceanogrraphy and
Marine Biology Annual Review 28: 381-496.
Austen, M.C., and B. Wibdom. 1991. Changes in and slow recovery of a meiobenthic nematode
assemblage following a hypoxic period in the Gullmar Fjord basin, Sweden. Marine
Biology 111: 139-145.
Baden, S.P., L.O. Loo, L. Pihl, and R. Rosenberg. 1990. Effects of eutrophication on benthic
communities including fish: Swedish west coast. Ambio 19: 113-122.
Bell, P.R.F., and I. Elmetri. 1995. Ecological indicators of large scale eutrophi cation in the Great
Barrier Reef lagoon. Ambio 24: 208-215.
Boesch, D.F., and R. Rosenberg. 1981. Response to stress in marine benthic communities. In:
G.W. Barrett and R. Rosenberg (eds.), Stress Effects on Natural Ecosystems, Chichester:
John Wiley, pp. 179-200
Breitburg, D.L., N. Steinberg, S. DuBeau, C. Cooksey, and E.D. Houde. 1994. Effects of low
dissolved oxygen on predation on estuarine fish larvae. Marine Ecology Progress Series
104: 235-246.
Clarke, K.B., and R.M. Warwick. 1994. Changes in Marine Communities: an Approach to
Statistical Analysis and Interpretation. Natural Environment Research Council, United
Kingdom.
Dalla Via, J., G. Van den Thillart, O. Cattani, and A. De Zwaan. 1994. Influence of long-term
hypoxia exposure on the energy metabolism of Solea solea. II. Intermediary metabolism
in blood, liver and muscle. Marine Ecology Progress Series 111: 17-27'.
179
-------
De-Jong, V.N., J.F. Bakker, and M. Van-Stralen. 1996. Recent changes in the contribution of
River Rhine and North Sea to the eutrophication of the western Dutch Wadden Sea.
Netherland Journal of Aquatic Ecology 30: 27-39.
Diaz, J.R., and R. Rosenberg. 1995. Marine benthic hypoxia: a review of its ecological effects
and the behavioural responses of benthic macrofauna. Oceanography and Marine Biology
Annual Review 33: 245-303.
Dauer, D.M., A.J. Rodi, Jr., and J.A. Ranasinghe. 1992. Effects of low dissolved oxygen events
on the macrobenthos of the lower Chesapeake Bay. Estuaries 15: 384-391
Dauer, D.M. 1993. Biological criteria, environmental health and estuarine macrobenthic
community structure. Marine Pollution Bulletin 26: 249-257.
Falkowski, P.G., T.S. Hopkins, and J.J. Walsh. 1980. An analysis of factors affecting oxygen
depletion in the New York Bight. Journal of Marine Research 38: 479-506.
Gabric, A.J., and P.R.F. Bell. 1993. Review of the effects of non-point nutrient loading on
coastal ecosystems. Australian Journal of Marine and Freshwater Research 44:73-85.
Gamenick I, A. John, K. Vopel, and O. Giere. 1996. Hypoxia and sulphide as structuring factors
in macrozoobenthic community on the Baltic Sea Shore: Colonization studies and
tolerance experiments. Marine Ecolology Progress Series 144: 73-85.
GESAMP (Group of Experts on the Scientific Aspects of Marine Pollution). 1990. The State of
the Marine Environment. United Nations Environmental Programme Regional Sea
Reports and Studies. No. 115. UNEP Blackwell Scientific Publications, Oxford, United
Kingdom.
Goldberg, E.D. 1995. Emerging problems in the coastal zone for the twenty-first century. Marine
Pollution Bulletin 31: 152-158.
Hagerman, L., and B. Vismann. 1993. Anaerobic metabolism, hypoxia and hydrogen sulphide in
the brackish water isopod Saduria entomon (L.). Ophelia 38: 1-11.
Hoback, W.W., and M.C. Barnhart. 1996. Lethal limits and sublethal effects of hypoxia on the
amphipod Gammaruspseudolimnaeus. Journal of the North American Benthologic
Society 15: 117-126.
Josefson, A.B., and B. Widbom. 1988. Differential response of benthic macrofauna and
meiofauna to hypoxia in the Gullmar Fjord basin. Marine Biology 100: 31-40.
Justic, D., N.N. Rabalais, and R.E. Turner 2001. Implication of global climate change for coastal
and estuarine hypoxia; hypothesis, observations, and models for the northern Gulf of
Mexico. Presented at the Sixth International Symposium on Fish Physiology,
Toxicology, and Water Quality. La Paz, Mexico 22-26 January 2001.
180
-------
Levin, L.A. 2000, Polychetes as environmental indicators: response to low oxygen and organic
enrichment. Bulletin of Marine Science 67: 668
Lu, L., and R.S.S. Wu. 2000. An experimental study on recolonization and succession of marine
macrobenthos in defaunated sediment. Marine Biology 136: 291-302.
MacArthur, R.H., and E.O. Wilson. 1967. The Theory of Island Biogeography. Princeton
University Press, Princeton, New Jersey.
Magaud, H., B. Migeon, P. Morfm, J. Garric, and E. Vindimian. 1997. Modeling fish mortality
due to urban storm run-off: Interacting effects of hypoxia and un-ionized ammonia.
Water Research 31:211-218.
Mason, Jr., W.J. 1998. Macrobenthic monitoring in the Lower St. Johns River, Florida.
Environmental Monitoring Assessment 50: 101-130.
Moss, A.J., G.E. Rayment, N. Reilley, and E.K. Best. 1992. A preliminary assessment of
sediment and nutrient transport from Queensland coastal catchment. Report of
Queensland Department of Environment and Heritage and Department of Primary
Industry, Brisbane, 33pp.
Nixon, S. 1990. Marine eutrophication: a growing international problem. Ambio 19: 101.
Pihl, L., S.P. Baden, R.J. Diaz, and L.C. Schaffner. 1992. Hypoxia-induced structural changes in
the diet of bottom-feeding fish and Crustacea. Marine Biology 112: 349-361.
Pihl, L. 1994. Changes in the diet of demersal fish due to eutrophication-induced hypoxia in the
Kattegat, Sweden. Canadian Journal of Fisheries and Aquatic Science 51: 321-336.
Powilleit, M., and J. Kube. 1999. Effects of severe oxygen depletion on macrobenthos in the
Pomeranian Bay (southern Baltic Sea): a case study in a shallow, sublittoral habitat
characterised by low species richness. Journal of Sea Research 42: 221-234
Peckol P., and J.S. Rivers. 1995. Physiological responses of the opportunistic macroalgae
Cladophora vagabunda (L.) and Gracilaria tikvahiae (McLachlan) to environmental
disturbances associated with eutrophication. Journal of Experimental Marine Biology and
Ecology 190: 1-16.
Rabalais, N.N. 2001. The effects of seasonally severe hypoxia on continental shelf benthos.
Presented at the 6th International Symposium on Fish Physiology Toxicology and Water
Quality. La Paz, Mexico 22-26 January 2001.
Rosenberg, R. 1980. Effects of oxygen deficiency on benthic macrofauna in fjords. In: Fjord
Oceanography. H. J Freeland, D. M. Farmer, and C. D Levings. (eds.). New York:
Plenum Press, pp. 499-514.
181
-------
Rosenberg, R. 1976. Benthic faunal dynamics during succession following pollution abatement
in a Swedish estuary. Oikos 27: 414-427.
Rosenberg R., B. Hellman, and B. Johansson 1991. Hypoxic tolerance of marine benthic fauna.
Marine Ecology Progress Series 79: 127-131.
Rosenberg, R., L.O. Loo, and P. Moller. 1992. Hypoxia, salinity and temperature as structuring
factors for marine benthic communities in an eutrophic area. Netherland Journal of Sea
Research 30: 121-129.
Sandberg, E. 1997. Does oxygen deficiency modify the functional responses of Saduria entomon
(Isopoda) to Bathyporeiapilosa (Amphipoda)? Marine Biology 729: 499-504.
Sagasti, A., L.C. Schaffner, and J.E. Duffy. 2000. Epifaunal communities thrive in an estuary
with hypoxia episodes. Estuaries 23: 474-487.
Stachowitsch, M. 1991. Anoxia in the Northern Adriatic Sea: rapid death, slow recovery. In: RV
Tyson, and TH Pearson (eds.), Modern and Ancient Continental Shelf Anoxia, London:
Geological Society (Special Publication 58), pp 119-129.
UNEP (United Nations Environmental Programme). 1991. The State of the World Environment.
UNEP. Nairobi, 48 pp.
Wannamaker, C.M., and J.A. Rice. 2000. Effects of hypoxia on movements and behaviour of
selected estuarine organisms from the southeastern United States. Journal of
Experimental Marine Biology and Ecology 249: 145-163.
Wu, R.S.S. 1982. Periodic defaunation and recovery in a sub-tropical epibenthic community in
relation to organic pollution. Journal of Experimental Marine Biology and Ecology 64:
253-269.
Wu, R.S.S., and P.K.S. Lam 1997. Glucose-6-phosphate dehydrogenase and lactate
dehydrogenase in the green-lipped muscle (Perna virids): Possible biomarkers for
hypoxia in the marine environment. Water Research 11: 2797-2801.
Wu, R.S.S. 1999. Eutrophication, trace organics and water-borne pathogens: Pressing problems
and challenge. Marine Pollution Bulletin 39: 11-22.
Vollenweider, R.A., R. Marchetti, and R. Viviani. 1992. Marine Coastal Eutrophication. Elsevier
Science Publishers B.V. p.3.
Zimmermann, A.R., and E.A. Canuel. 2000. A geochemical record of eutrophication and anoxia
in Chespeake Bay sediments: anthropogenic influence on organic matter composition.
Marine Chemistry 69: 117-137
182
-------
HYPOXIA AND ANOXIA AS GLOBAL PHENOMENA
Robert J. Diaz1
ABSTRACT
No other environmental variable of such ecological importance to estuarine and coastal
marine ecosystems around the world has changed so drastically, in such a short period of time, as
dissolved oxygen. While hypoxic and anoxic environments have existed through geological
time, their occurrence in shallow coastal and estuarine areas appears to be increasing, most likely
accelerated by human activities. Several large systems, with historical data, that never reported
hypoxia at the turn of the century (e.g. Kattegat, the sea between Sweden and Denmark) now
experience severe seasonal hypoxia. Synthesis of literature pertaining to benthic hypoxia and
anoxia revealed that the oxygen budgets of many major coastal ecosystems have been adversely
affected, mainly through the process of eutrophication (the production of excess organic matter).
It appears that many ecosystems that are now severely stressed by eutrophication-induced
hypoxia are threatened with the loss of fisheries, loss of biodiversity, and alteration of food webs.
INTRODUCTION
A review of literature pertaining to ecological effects of hypoxia and anoxia revealed that
the oxygen budgets for major coastal ecosystems around the world have been adversely affected
mainly through the process of eutrophication. Eutrophication, the production of excess organic
matter (Nixon 1995), fuels the development of hypoxia and anoxia when combined with water
column stratification. Many ecosystems have reported some type of monotonic decline in
dissolved oxygen levels through time, with a strong correlation between human activities and
declining dissolved oxygen (for example: Gulf of Trieste, Italy; Kattegat, Sweden-Denmark). In
some ecosystems prone to water column stratification, the linkage of human activity to hypoxia
is less obvious (for example: Chesapeake Bay, Maryland-Virginia; Port Hacking, Australia).
Hypoxia related to anthropogenic activities appears to develop within a system as a result
of the cumulative effects of eutrophication. Many times hypoxia is not noticed until higher level
ecosystem effects are manifested. For example, in the Kattegat, hypoxia did not become a
prominent environmental issue until the collapse of a Norway lobster fishery several years after
hypoxic bottom waters were first reported. The northern Gulf of Mexico is representative of
severely stressed coastal ecosystems that currently experience seasonal hypoxia, but have not
experienced hypoxia-related loss of fisheries. Over the last several decades, hypoxia in the
northern Gulf of Mexico has affected benthic invertebrate communities, but there is no clear
signal of hypoxia in fisheries landings statistics (Rabalais et al. 2001, Chesney and Baltz 2001).
The shallow, northwest continental shelf of the Black Sea (which is not part of the deep central
basin anoxia) is typical of ecosystems that have experienced drastic reductions in bottom
1 College of William and Mary, School of Marine Science, Virginia Institute of Marine Science, Gloucester Pi, VA
23062, USA.
183
-------
fisheries due to hypoxia. Since the 1960's, increasing hypoxia and anoxia have been blamed for
the replacement of the highly valued demersal fish species with less desirable planktonic
omnivores. Of the 26 commercial species fished in the 1960's, only six still support a fishery
(Mee 1992).
This article presents a brief overview and update of hypoxic conditions in estuarine and
marine systems around the world.
HYPOXIA DEFINITION
Oxygen is necessary to sustain the life of all fishes and invertebrates, and when the
supply of dissolved oxygen is cut-off or the consumption rate exceeds resupply, oxygen
concentrations decline beyond the point that sustains most animal life. This condition of low
dissolved oxygen is known as hypoxia. The point at which various animals suffocate varies, but
generally effects start to appear when oxygen drops below 2 ml O2/L or 2.8 mg O2/L; for sea
water this is about 18% of saturation (see summary table in Diaz and Rosenberg 1995). The two
principal factors that lead to the development of hypoxia, and at times anoxia, are water column
stratification that isolates the bottom water from exchange with oxygen-rich surface water, and
the decomposition of organic matter in the isolated bottom water that reduces oxygen levels.
Both factors must be at work for hypoxia to develop and persist.
OXYGEN AROUND THE WORLD
On a geological time scale, low dissolved oxygen environments (hypoxia and anoxia)
were major factors in shaping evolution of life (see Caplan and Bustin 1999 as a recent
example). However, much of the current hypoxia and anoxia in shallow coastal and estuarine
areas is recent in origin, within the last 50 to 100 years, and closely associated with human
activities. The worldwide distribution of hypoxic zones is centered on major population centers
and closely associated with developed watersheds that deliver large quantities of nutrients, the
most important of which is nitrogen, to coastal seas (Howarth et al. 1996).
Within the last 40-50 years, dissolved oxygen conditions of many major coastal
ecosystems around the world have been adversely affected through the process of eutrophication.
Most of these coastal systems have recorded a steady decline in dissolved oxygen through time,
in most cases starting from initial oxygen measurements, usually in the 1950's (Rosenberg 1990).
The declining trend in dissolved oxygen lagged about 10 to 20 years behind the post World War
II trend of increased chemical fertilizer use. For systems that have historical data from the turn-
of-the-century, the declines in oxygen levels appear to have started in the 1950's and 1960's.
However, for the Baltic Sea, declining dissolved oxygen levels were noted as early as the 1930's
(Fonselius 1969). A summary of systems experiencing low dissolved oxygen problems indicated
a positive association between the likelihood of hypoxia and large population centers (Table 1).
184
-------
From historical perspective, it is clear that many of the systems that are currently hypoxic
(Table 1) were not when they were first studied. The best examples of systems with long-term
data come from Europe, where benthic hypoxia was not reported prior to the 1950's in the Baltic
Sea proper (Fonselius 1969), 1960's in the northern Adriatic (Justic 1987), 1970's in the Kattegat
(Baden et al. 1990a), and the 1980's on the northwest continental shelf of the Black Sea (Mee
1992). Except in areas of natural upwelling, such as off Peru and Central America (Tarazona et
al. 1988) and west Africa's Namibian shelf (Hamukuaya et al. 1998), or near oceanic oxygen
minimum zones (OMZ), such as in the Arabian Sea (Gooday et al. 2000), coastal hypoxia does
not appear to be a natural condition.
By the 1970's, estuarine and coastal ecosystems around the world were becoming over-
enriched with organic matter (Nixon 1995), and many of them manifested hypoxia for the first
time (Diaz and Rosenberg 1995). Once it occurred, hypoxia quickly became an annual event and
a prominent feature affecting energy flow processes in the ecosystem (Elmgren 1989, Pearson
and Rosenberg 1992). From the 1980's to the present, the distribution of hypoxia around the
world has not changed appreciably in a positive way (Diaz and Rosenberg 1995, 2001). Only in
systems that have experienced intensive regulation of nutrient or carbon inputs have oxygen
conditions improved, for example the Hudson River, New York, and Delaware River,
Pennsylvania-New Jersey. There are many examples of small-scale reversals in hypoxia
associated with improvements in treatment of sewage and pulp mill effluents as early as the
1970s (Rosenberg 1972, 1976). Temporary improvements have also been seen in systems with
changes in hydrology or nutrient inputs, such as the Black Sea, Baltic Sea, and northern Gulf of
Mexico.
The occurrence of hypoxia was closely linked to eutrophication as early as the 1980s.
For example, in the German Bight, van Pagee et al. (1983) found that from 1930 to 1983 there
was an increase in nutrients that corresponded with an increase in the duration and severity of
hypoxia. In all recent cases, as listed in Table 1, hypoxia appears to be a result of general
ecosystem eutrophication. So it is difficult or impossible to separate the effects of hypoxia
verses eutrophication on ecosystem functioning. Eutrophication is also closely linked to a
system's secondary productivity, and to a point enhances biomass and fisheries yield (Caddy
1993). The critical point is the appearance of severe hypoxia or anoxia in the system, which has
the potential to produce mass mortality of both benthic and pelagic species. The general effect
of eutrophication is to favor species with opportunistic life histories and the increased organic
matter added to the system tends to increase biomass. Systems that experience hypoxia or
anoxia also tend to be dominated by opportunistic species, but benthic biomass is reduced since
the low dissolved oxygen tends to cause mortality. However, eutrophication has a
preconditioning effect on benthic fauna, which tends to lessen the acute response of the system to
hypoxia when it finally does occur. This is the reason some systems that experience mild
hypoxia show no acute effect, such as the York River, Virginia. The most common form of
hypoxia is annual, one event per year, with the most common response being mortality of
benthos followed by recolonization upon the return of normal oxygen conditions (Table 1).
185
-------
(D
d,
'1 1
.2 >
o ^-,
3 also experience anoxia. Hyj
rter than a year; Annual, year'
£H O
C J
tfl M
£|
1 -^
o J3
"S M
g 2
^ M
CO
1 "
(D ^
"&5 .-H
tS "g
1 'S
0 ^
I a"
lasis on ben
;er than a ye
I 2
§ M
§ "I
•S ^
's s
o -^
S on
I 1
i s
3 S
§0
0
a °o
•-J O
O.S
S"1 QJ
^ s
• •^ rt
§ "S
Mg
ummary of anthrop
as: Aperiodic, ever
C/3
-H'
2
H
.2
O
J2
^
5 1
I -i
a °
.2 &
S o
1 1
— •§
1 u
§ ^3
3 OJ
lH 0
0 C
related to summer <
event; Mortality, r
&
1
g
Q
0
ti
on
2
^
|
0
%
&
-2
S
O
-fj1
'S
1
o
0
n occurs bu
_o
_N
_O ^
O >-j
g rt
tl
1 1
rt 1
51 >,
H
P o
II
"u 'S
3 ^
i- °
a I
§" 2
Change, dynamics
structure; Annual, i
Reference
||
CM .H
II
C 0
QJ qj
pq ^
M QJ
-H C
S o
J3 OH
E (Ij
"rt
;nthic Faun
Response
PQ
,0
QJ
^
QJ
S
I"
O
O
QJ
t/3
01
•2
0 QJ
& ^
a E^
Fallesen and Jorgensen 199 '.
Moser 1998
1-1
o> "^
i|
>> >,
ass Mortali
ass Mortali
S S
>1
QJ
^ t/3
s S
B
a s
0> 1>
Q ^
Arhus Bay
Barnegatt Inlet
13 13
3 3
II
OO
OS
& o\
•-H ^
Gonzales-Oreja and Saiz-Sa
Tulkki 1965, Leppakoski 19
1
3
>>
Mortality
ass Mortali
s
rt
(D
C/3
.S •-
t&m
Bilbao Estuary
Bornholm Basin
13 13
3 3
II
Mallinefa/. 1999
Holland et al. 1987
Mee 1992, Zaitsev 1993
•^ "^ "^
III
^ "^ "^
1 — ' (D W
-a 1 §
M jd (D
•&1
.S .S t:
a) S 0
HI
s
,C
i"S «
0 J w
Q rt ^
z S m
"M
S |
-f^ ^
t/3 -H
'S ^
Cape Fear River
Chesapeake Bay I
Continental Shelf
13 13 13
333
III
•0
«j
"rt -^
^ O
a) o
CC f~^
Boesch and Rababais 1991,
Turner 2001, Rabalais et al.
13
3
"^
$
QJ
&
Mortality
cs
G
.2
'-2
'3
o
1-1
if
_o
J3
Continental Shelf
13
3
*
S
1 1
^fe
Q 3
Ritter and Montagna 1999
Brongersma- Sanders 1957, '.
al. 1996, Bentley and Nittro
"2 13
o 3
•3 1
^^
Reduced
Mortality
>.
1 1
QJ QJ
H O
>:
Corpus Christi Ba
Eckernforde Bay
13 13
3 3
II
Thielefa/. 1995
*OJ
M
1)
l-i
1
0>
o
3
w
(D
.0
3
13
3
*
Friligos and Zenetos 1988,
Theodorou 1996
*c3
1
>-.
ass Mortali
s
Q
i
ass Mortali
s
1
Gulf of Trieste
13
3
•*
Simunovic et al. 1999
Josefson and Widbom 1988
13
3
QJ
(/]
QJ
&
01
>1
ass Mortali
s
a
T3
QJ
01
Gullmarsfjord
13
3
*
O
OO
o
M
o
*c3
1
&
o
s
rt
3
m
BO
3
0
3
13
3
^
OO
o\
'J2
s
>->
o
M
s
1
c
3
3
3
13
3
^
186
-------
Mortality Stressed . Nakataefa/. 1997
Suzuki and Matsukawa 1987
nmark Mass Mortality Collapse Norway Multi-year Baden et al. 1990a, Josefson and Jensen
Lobster 1992, Rosenberg et al. 1992
Mass Mortality Stressed Annual Arntz 1981, Rumohr 1986, Weigelt 1990
Lopez-Jamar et al. 1995
13 13 13
3 =5 3
§
H
rfi OH
O 01
Kiel Bay
La Coruna Bay
13 13
3 3
Mortality Stressed Annual Baden et al. 1990b,
Rosenberg and Loo 1988
Mass Mortality . . Yamamuro et al. 1998
Sweden
>^
&
B
1
2
13
3
1
Lake Shinji
13
3
Mass Mortality None Annual Jorgensen 1980, Hylleberg 1993
Denmark
"H
0
tc>
13
"^
3
^t-
C\
OS
"3
"^
^c
GO
1
o
Alabama
Mobile Bay
13
3
Glenn ef a/. 1996
aj
2
i
rt
m
c
rf
'S
o
PH
"^
3
Mortality . Annual Rainer and Fitzhardinge 1981
Australia
Port Hacking
13
3
*0
t^
Os
T3
!a
^
o
(2
T3
C
OJ
e
o>
O
o>
1
OJ
^
0
><
(D
Raritan Bay
13
3
00
Os
I
o
'S
H
W^
^sO
o>
•^
J
o
s
1
"s
British Co:
Saanich Inlet
13
3
00
c\
Os
1
o
1>
o
&
OJ
^
Sea of Azov
*rt
=s
Mortality . Annual Imabayashi 1986
I
rt
OJ
C/3
T3
M
"c
O
(D
01
13
3
Mortality . Annual Mason 1998
Mass Mortality Reduced shellfish . Souchu et al. 1998
production
Mass Mortality . Annual Wu 1982
^3 dj
3 °
E &
St. Johns River
Thau
13 13
3 3
BO
O
O
Tolo Harbor
—
1
Mortality . Annual Tsutsumi 1987
Mortality Stressed Reduced Josefsen and Rosenberg 1988
al
3 (D
M
O
(D -(->
oa
0> j_,
S ^
s2|
13 13
187
-------
France Mass Mortality Collapse of cockle industry . Rybarczyk et al. 1996
c Baie de Somme
0
•c
0
'C
1
T3
o>
cc
1>
55
1
o
M
rf
1
£
q_r
J2
T3
0
OH
^
Wadden Sea Mortality . . deJongeefa/. 1994
Angola . . . LeLoeuff 1999
c Wadden Sea
c West Africa She]
-73 T3
0 0
OH OH
<, <,
ON
ON
0>
ON
OJ
CM
0
1
&
T3
(D
X
&
x
0
rt
(D
X
_0
rf
m
c Wismar Bay
o
OH
^
Alabama Mortality Loss of oyster . Rikard et al. 2000
Bon Secour Bay
;2
.0
PH
Florida Mortality . . Lapointe and Matzie 1996
France . . . Abril et al. 1999
Florida Keys
Gironde Estuary
-H -H
0 0
'C 'C
PH PH
i Louisiana Mortality Loss of large clams . Abadie and Poirrier, 2000
Lake Pontchartra
;2
.0
PH
Italy Mortality . Annual Tagliapietra ef al. 1998
Palude della Ros
;2
.0
PH
France Mass Mortality Reduced aquaculture Annual Guyoneaud et al. 1998
Prevost Lagoon
;2
.0
PH
/er Virginia Mortality Stressed Annual Llanso 1992
Florida . Fauna migrate . Leonard and McClintock 1999
Rappahannock R
St. Joseph Bay
O Q
0 0
'C 'C
PH PH
•O
S3
"f
rt
T3
§
c
•3
^
OJ
1
O
rt
3
E
St. Lucie River
Q
,0
PH
Virginia None Stressed No Change PMet al. 1991, Diaz et al. 1992
York River
o
0
•c
PH
00
t~
c\
~5
^
g
(D
i
Q
T3
(D
T3
OJ
cc
(D
GO
tn
0
rt
-------
Sg
' — 1 , 1
^ £P
•*-, OJ
II
a -o
>> |
2 "o
^ cS
>i >^
s s
•e e
0 0
"T3 rt
GulfofFinla
Sweden-Nor
Gulf of Finland, Deep
Idefjord
c c
QJ QJ
a^ t>3
2 2
PH PH
00
Os
<^
M
1
Califorinia
Santa Monica Basin
S
03
e
CM
oo
ON
"M
"QJ
1
Tl
>
o>
ffl
Scheldt Estuary
c
CM CM
0>
TO
a
M
C
(D
Taiwan
Tan Shui Estuary
s
VI
2
i
il
ffl Q
"^
"5
o
TO GO
j3 o>
O 2:
Fosu Lagoon
Horseshoe Lagoon
'$ '$
2 2
13 D
00
Os
"3
^
Lassm
G
Indian Oceal
.1
s
£
2
1
s
^_l
1 i
"3 "c
ll
bfl oj
Mondego River
Pettaquamscutt River
9- '$
2 2
II
s ^
OS
<^
1
E
Denmark
Roskilde Fjord
•$
2
1
1
1
OH ^
J3 S
CO ^"
189
-------
SUMMARY
Up to the 1950s, reports of mass mortality of marine animals caused by lack of oxygen
were limited to small systems that had histories of oxygen stress. In the 1960s, the number of
systems with reports of hypoxia-related problems started to increase with the 1970s and 1980s as
the period with the most initial reports of hypoxia. By the 1990s, most estuarine and marine
systems in close proximity to population centers had reports of hypoxia or anoxia. It appears
that the number of systems being affected by hypoxia/anoxia through time is increasing.
Coastal and estuarine hypoxia does not appear to be a natural condition, except in areas
influenced by OMZs, upwelling, or enclosed fjordic systems. The main factor in development of
hypoxia in coastal and estuarine systems has been the input of excess nutrients that leads to
eutrophication. The determination of population or ecosystem level effects from hypoxia is
complicated by many factors that include inadequate data on historic trends of species
populations and dissolved oxygen concentrations, and the interaction of multiple stressors such
as fishing pressure, and habitat loss. Hypoxia and anoxia are among the most widespread
deleterious anthropogenic effects in estuarine and marine environments. The effects of hypoxia
may be reversible with the reduction of nutrients or organic inputs to a system, which would lead
to a reduction or elimination of hypoxia.
REFERENCES
Abadie, S.W., and M.A. Poirrier. 2000. Increased density of large Rangia clams in Lake
Pontchartrain after the cessation of shell dredging. Journal of Shellfish Research 19:481-
485 .
Abril, G., H. Etcheber, P.L. Hir, P. Bassoullet, B. Boutier, and M. Frankignoulle. 1999.
Oxic/anoxic oscillations and organic carbon mineralization in an estuarine maximum
turbidity zone (The Gironde, France). Limnology and Oceanography 44:1304-1315.
Andersin, A.B., J. Lassig, L. Parkkonen, and H. Sandier. 1978. The decline of macrofauna in the
deeper parts of the Baltic proper and the Gulf of Finland. Kieler Meeresforschungen,
Sonderheft 4:23-52.
Arntz, W. 1981. Zonation and dynamics of macrobenthos biomass in an area stressed by oxygen
deficiency. In Stress Effects on Natural Ecosystems, Barrett, G.W: and Rosenberg, R.
(eds.), John Wiley and Sons, Chi Chester, pp. 215-225.
Baden, S.P., L.O. Loo, L. Pihl, and R. Rosenberg. 1990a. Effects of eutrophi cation on benthic
communities including fish - Swedish west coast. Ambio 19:113-122.
Baden, S.P., L. Pihl, and R. Rosenberg. 1990b. Effects of oxygen depletion on the ecology,
blood physiology and fishery of the Norway lobster Nephrops norvegicus. Marine
Ecology Progress Series 67:141-155.
190
-------
Bentley, S.J., and C.A. Nittrouer. 1999. Physical and Biological Influences on the Formation of
Sedimentary Fabric in an Oxygen-restricted Depositional Environment: Eckernfoerde
Bay, Southwestern Baltic Sea. Palaios 14.
Blay, J., Jr., and F. Dongdem. 1996. Preliminary observations on the benthic macrofauna of a
polluted coastal lagoon in Ghana (West Africa). Tropical Ecology 37:127-133.
Boesch, D.F., andN.N. Rabalais. 1991. Effects of hypoxia on continental shelf benthos:
comparisons between the New York Bight and the Northern Gulf of Mexico. In: Modern
and Ancient Continental Shelf Anoxia. R.V. Tyson and T.H. Pearson (eds.). Geological
Society Special Publication No. 58, London, pp. 27-34.
Brockmann, U., G. Billen, and W. Gieskes. 1988. North Sea Nutrients and Eutrophication . In:
Pollution of the North Sea: An Assessment. W. Salomons, B. Bayne, E. Duursma, U.
Forstner (eds.), Springer-Verlag, New York. pp. 348-389.
Brongersma-Sanders, M. 1957. Mass mortality in the sea. In: Treatise on Marine Ecology and
Paleoecology, Volume 1, J.W. Hedgpeth (ed.), Waverly Press, Baltimore, pp. 941-1010.
Brosnan, T.M., and M.L. O'Shea. 1996. Long-term improvements in water quality due to sewage
abatement in the lower Hudson River. Estuaries 19:890-900.
Caddy, J. 1993. Toward a comparative evaluation of human impacts on fishery ecosystems of
enclosed and semi-enclosed seas. Review of Fishery Science 1:57-96.
Caplan, M.L., and R.M. Bustin. 1999. Devonian-Carboniferous Hangenberg mass extinction
event, widespread organic-rich mudrock and anoxia: causes and consequences
Palaeogeography, Palaeoclimatology, Palaeoecology 148:187-208.
Chamberlain, R., and D. Hayward. 1996. Evaluation of water quality and monitoring in the St.
Lucie Estuary, Florida. Water Resources Bulletin 32:681-696.
Chechun, T.Y. 1998. Nutrition of the Acipenseridae of the Sea of Azov under present conditions.
Journal of Ichthyology 38:147-154.
Chesney, E.J. and D.M. Baltz. 2001. The effects of hypoxia on the northern Gulf of Mexico In:
Coastal Hypoxia Consequences for Living Resources and Ecosystems. N.N. Rabalais and
R.E. Turner (eds.). American Geophysical Union, Washington, D.C. pp. 321-354.
Christensen, J., and T. Packard. 1976. Oxygen utilization and plankton metabolism in a
Washington fjord. Estuarine and Coastal Marine Science 4:339-347
D'Andrea, A.F., N.I. Craig, and G.R. Lopez. 1996. Benthic macrofauna and depth of bioturbation
in Eckernfoerde Bay, southwestern Baltic Sea. Geo-Marine 16:155-159.
191
-------
DeJonge V., W. Boynton, C. D'Elica, R. Elmgren, and R. Welsh. 1994. Responses to
developments in eutrophication in four different North Atlantic estuarine systems. In:
Changes in Fluxes in Estuaries: Implications from Science to Management. K. Dyer, and
R. Orth (eds.), Olsen and Olsen, Fredensborg.
Desprez, M., H. Rybarczyk, J.G. Wilson, J.P Ducrotoy, F. Sueur, R. Olivesi, and B. Elkaim.
1992. Biological impact of eutrophi cation in the Bay of Somme and the induction and
impact of anoxia. Netherlands Journal of Sea Research 30:149-159.
Dethlefsen, V., and H. v. Westernhagen. 1983. Oxygen deficiency and effects on bottom fauna in
the eastern German Bight 1982. Meeresforschung 60:42-53.
Diaz, RJ. and R. Rosenberg. 2001. Overview of anthropogenically-induced hypoxia effects on
marine benthic fauna. In: Coastal Hypoxia Consequences for Living Resources and
Ecosystems. N.N. Rabalais and R.E. Turner (eds.). American Geophysical Union,
Washington, D.C. pp. 129-146.
Diaz, R.J., RJ. Neubauer, L.C. Schaffner, L. Pihl, and S.P. Baden. 1992. Continous monitoring
of dissolved oxygen in an estuary experiencing periodic hypoxia and the effect of
hypoxia on macrobenthos and fish. Science of the Total Environment, Supplement 1992,
1055-1068.
Donnelly, T.H., P.W. Ford, D. McGregor, and D. Allen. 1999. Anthropogenic changes to a
billabong in New South Wales. 1. Lagoon evolution and phosphorus dynamics. Marine
and Freshwater Research 50:689-698.
Dumont, HJ. 1998. The Caspian Lake: History, biota, structure, and function. Limnology and
Oceanography 43:44-52.
Dyer, M.F., J.G. Pope, P.D. Fry, RJ. Law, and J.E. Portmann. 1983. Changes in fish and benthos
catches off the Danish coast in September 1981. Journal of the Marine Biological
Association of the United Kingdom 63:767-775.
Elmgren, R. 1989. Man's impact on the ecosystem of the Baltic Sea: Energy flows today and at
the turn of the Century. Ambio 18:326-332.
Engle, V.D., J.K. Summers, and J.M. Macauley. 1999. Dissolved oxygen conditions in northern
Gulf of Mexico estuaries. Environmental Monitoring and Assessment 57:1-20.
Fallesen, G., andH.M. J0rgensen. 1991. Distribution of Nephtys hombergii and TV. ciliata
(Polychaeta: Nephtyidae) in Arhus Bay, Dennmark, with emphasis on the effect of
sewage oxygen deficiency. Ophelia Supplement 5:443-450.
192
-------
Flemer, D.A., W.L. Kruczynski, B.F. Ruth, and C.M. Bundrick. 1999. The relative influence of
hypoxia, anoxia, and associated environmental factors as determinants of macrobenthic
community structure in the northern Gulf of Mexico. Journal of Aquatic Ecosystem
Stress and Recovery 6:311-328.
Flindt, M.R., L. Kamp-Nielsen, J.C. Marques, M.A. Pardal, M. Bocci, G. Bendoricchio, J.
Salomonsen, S.N. Nielsen, and S.E. Jorgensen. 1997. Description of the three shallow
estuaries: Mondego River (Portugal), Roskilde Fjord (Denmark) and the Lagoon of
Venice (Italy). Ecological Modelling 102:17-31.
Fonselius, S.H. 1969. Hydrography of the Baltic deep basins III. Fishery Board of Sweden,
Hydrography Series Report No.23. pp. 1-97.
Friligos, N. and A. Zenetos. 1988. Elefsis Bay anoxia: nutrient conditions and benthic
community structure. Marine Ecology 9:273-290.
Fritz, C., E. LaBrecque, J. Tober, PJ. Behr, and I. Valiela. 1996. Shrimp in WaquoitBay:
Effects of nitrogen loading on size and abundance. General Scientific Meetings of the
Marine Biological Laboratory, Woods Hole, MA. 191:326-327.
Garlo, E.V., C.B. Milstein, and A.E. Jahn. 1979. Impact of hypoxic conditions in the vicinity of
Little Egg Inlet, New Jersey in summer 1976. Estuarine and Coastal Marine Science
8:421-432.
Gillibrand, P.A., W.R. Turrell, D.C. Moore, and R.D. Adams. 1996. Bottom water stagnation
and oxygen depletion in a Scottish sea loch. Estuarine, Coastal and Shelf Science 43:217-
235.
Glenn, S., M. Crowley, D. Haidvogel, and Y. Song. 1996. Underwater observatory captures
coastal upwelling event off New Jersey. EOS 77:223-236.
Gooday, A.J., J.M. Bernhard, L.A. Levin, and S.B. Suhr. 2000. Foraminifera in the Arabian Sea
oxygen minimum zone and other oxygen-deficient settings: taxonomic composition,
diversity, and relation to metazoan faunas. Deep-Sea Research Part II 47:25-54.
Gong, C. 1998. An Organic and Isotopic Geochemical Study of Santa Monica Basin Sediments,
Offshore California. Dissertation Abstracts International Part B: Science and
Engineering, University Microfilms International 59:2076.
Gonzales-Oreja, J., and J. Saiz-Salinas. 1998. Loss of herterotrophic biomass structure in an
extreme estuarine environment. Estuarine, Coastal and Shelf Science 48:391-399.
Guyoneaud, R., R. De Wit, R. Matheron, and P. Caumette. 1998. Impact of macroalgal dredging
on dystrophic crises and phototrophic bacterial blooms (red waters) in a brackish coastal
lagoon. Oceanologica acta 21:551-561.
193
-------
Hamukuaya, H., MJ. O'Toole, and P.MJ. Woodhead. 1998. Observations of severe hypoxia and
offshore displacement of Cape Hake over the Namibian shelf in 1994. South African
Journal of Marine Science 19:57-59.
Harper, D.E., L.D. McKinney, J.M. Nance, and R.B. Salzer. 1991. Recovery responses of two
benthic assemblages following an acute hypoxic event on the Texas continental shelf,
northwestern Gulf of Mexico. In: Modern and Ancient Continental Shelf Anoxia. R.V.
Tyson and T.H. Pearson (eds.). Geological Society Special Publication No. 58, London.
pp. 49-64.
Harper, D.E., L.D. McKinney, R.B. Salzer, and RJ. Case. 1981. The occurrence of hypoxic
bottom water off the upper Texas coast and its effects on the benthic biota. Contributions
in Marine Science 24:53-79.
Holland, A.F., A.T. Shaughnessy, and M.H. Hiegel. 1987. Long-term variation in Mesohaline
Chesapeake Bay macrobenthos: spatial and temporal patterns. Estuaries 10:370-378.
Howell, P., and D. Simpson. 1994. Abundance of marine resources in relation to dissolved
oxygen in Long Island Sound. Estuaries 17:394-402.
Howarth, R.W., G. Billen, D. Swaney, A. Townsend, N. Jaworski, K. Lajtha, J.A. Downing, R.
Elmgren, N Caraco, T. Jordan, F. Berendse, J. Freney, V. Kudeyarov, P. Murdoch, and Z.
Zhao-Liang. 1996. Regional nitrogen budgets and riverine N & P fluxes in the drainage
to the North Atlantic Ocean: natural and human influences. Biogeochemistry 35:75-139.
Hylleberg, J. 1993. Extinction and immigration of benthic fauna, the value of historical data from
Limfjorden, Denmark. Symposium Mediterranean Seas 2000, Santa Margherita Ligure,
23-27. September 1991, 1st Scientific Ambientali Marine, Santa Margherita Ligure, Italy.
pp. 43-70.
lizuka, S., and S. Min. 1989. Formation of anoxic bottom waters in Omura Bay. Engan Kaiyo
Kenkyu. 2:75-86.
Imabayashi, H. 1986. Effect of oxygen-deficient water on the settled abundance and size
composition of the bivalve Theora lubrica. Bulletin of the Japanese Society of Scientific
Fisheries. 5:391-397.
Jeng, W-L., and B-C. Han. 1996. Coprostanol in a sediment core from the anoxic Tan-Shui
Estuary, Taiwan. Estuarine, Coastal and Shelf Science 42:727-735.
J0rgensen, B.B. 1980. Seasonal oxygen depletion in the bottom waters of a Danish fjord and its
effect on the benthic community. Oikos 34:68-76.
Josefson, A.B., and B. Widbom. 1988. Differential response of benthic macrofauna and
meiofauna to hypoxia in the Gullmar Fjord basin. Marine Biology 100:31-40.
194
-------
Josefson, A.B., and J.N. Jensen. 1992. Effects of hypoxia on soft-sediment macrobenthos in the
southern Kattegat. In: Marine Eutrophication and Population Dynamics. G. Colombo, I.
Ferrari, V. U. Ceccherelli and R. Rossi (eds.), Olsen and Olsen, Fredensborg. pp. 21-28.
Josefson, A.B., and R. Rosenberg. 1988. Long-term soft-bottom faunal changes in three shallow
fjords, west Sweden. Netherlands Journal of Sea Research 22, 149-159.
Justic, D. 1987. Long-term eutrophication of the northern Adriatic Sea. Marine Pollution Bulletin
18:281-284.
Kaiser, J., and S. Lutter. 1998. Do we have the right strategies to combat eutrophi cation in the
Wadden Sea? A critical review of current policies. Ecosystem Research in the Wadden
Sea Area. 9. International Scientific Wadden Sea Symposium, Norderney, Germany, 5-8
Nov 1996. Senckenbergiana maritima 29:17-24.
Kitching, J.A., F.J. Ebling, J.C. Gable, R. Hoare, A.A.Q.R. McLeod, and T.A. Norton. 1976. The
ecology of Lough Ine. XIX. Seasonal changes in the western trough. Journal of Animal
Ecology 45:731-758.
Laine, A.O., H. Sandier, A-B. Andersin, and J. Stigzelius. 1997. Long-term changes of
macrozoobenthos in the eastern Gotland Basin and the Gulf of Finland (Baltic Sea) in
relation to the hydrographical regime. Netherlands Journal of Sea Research 38:135-159.
Lapointe, B.E., and W.R. Matzie. 1996. Effects of stormwater nutrient discharges on
eutrophi cation processes in nearshore waters of the Florida Keys. Estuaries 19:422-435.
Le Loeuff, P. 1999. The benthic macrofauna of the variable salinity waters ecosystems along the
Atlantic coast of tropical Africa; biodiversity variations with the current climatic
conditions (rainfall) and the regional climatic history. Zoosystema, 21:557-571.
Lenihan, H.S., and C.H. Peterson. 1998. How habitat degradation through fishery disturbance
enhances impacts of hypoxia on oyster reefs. Ecological Applications 8:128-140.
Leonard, C.L., and J.B. McClintock. 1999. The population dynamics of the brittlestar
Ophioderma brevispinum in near- and farshore seagrass habitats of Port Saint Joseph
Bay, Florida. Gulf of Mexico Science 17:87-94.
Leppakoski, E. 1969. Transitory return of the benthic fauna of the Bornholm Basin, after
extermination by oxygen insufficiency. Cahiers de Biologie Marine 10:163-172.
Llanso, R.J. 1992. Effects of hypoxia on estuarine benthos: the lower Rappahannock River
(Chesapeake Bay), a case study. Estuarine, Coastal and Shelf Science 35:491-515.
Lopez-Jamar, E., O. Francesch, A.V. Dorrio, and S. Parra. 1995. Long-term variation of the
infaunal benthos of La Coruna Bay (NW Spain): results from a 12-year study (1982-
1993) Scientia marina 59:49-61.
195
-------
Mallin, M.A., M.H. Posey, G.C. Shank, M.R. Mclver, S.H. Ensign, and T.D. Alphin. 1999.
Hurricane effects on water quality and benthos in the Cape Fear watershed: Natural and
anthropogenic impacts. Ecological Applications 9:350-362.
Mason, W.T., Jr. 1998. Macrobenthic monitoring in the Lower St. Johns River, Florida.
Environmental Monitoring and Assessment 50:101-130.
May, E. 1973. Extensive oxygen depletion in Mobile Bay, Alabama. Limnology and
Oceanography 18:353-366.
Mee, L.D. 1992. The Black Sea in crisis: A need for concerted international action. Ambio
21:278-286.
Mirza, F.B., and J.S. Gray. 1981. The fauna of benthic sediments from the organically enriched
Oslofjord, Norway. Journal of Experimental Marine Biology and Ecology 54:181-207.
Morrisey, D.J. 2000. Predicting impacts and recovery of marine-farm sites in Stewart Island,
New Zealand, from the Findlay-Watling model. Aquaculture 185:257-271.
Moser, F.C. 1998. Sources and Sinks of Nitrogen and Trace Metals, and Benthic Macrofauna
Assemblages in Barnegat Bay, New Jersey. Dissertation Abstracts International Part B:
Science and Engineering, University Microfilms International 58:5849.
Nakata, K., S-I. Sugioka, and T. Hosaka. 1997. Hindcast of a Japan Sea Oil Spill. Spill Science
and Technology Bulletin 4:219-229.
Niermann, U., E. Bauerfeind, W. Hickel, and H.V. Westernhagen. 1990. The recovery of benthos
following the impact of low oxygen content in the German Bight. Netherlands Journal of
Sea Research 25:215-226.
Nixon, S.W. 1995. Coastal marine eutrophication: a definition, social causes, and future
concerns. Ophelia 41:199-219.
Pearson, T.H., and A. Eleftheriou. 1981. The benthic ecology of Sollom Voe. Proceedings of the
Royal Society of Edinburgh 808:241-269.
Pearson, T.H., and R. Rosenberg. 1992. Energy flow through the SE Kattegat: a comparative
examination of the eutrophi cation of a coastal marine ecosystem. Netherlands Journal of
Sea Research 28:317-334.
Paerl, H.W., J.L. Pinkney, and S.A. Kucera. 1995. Clarification of the structural and functional
roles of heterocysts and anoxic microzones in the control of pelagic nitrogen fixation.
Limnology and Oceanography 40:634-638.
196
-------
Paerl, H.W., J.L. Pinckney, J.M. Fear, and B.L. Peierls. 1998. Ecosystem responses to internal
and watershed organic matter loading: Consequences for hypoxia in the eutrophying
Neuse River Estuary, North Carolina, USA. Marine Ecology Progress Series 166:17-25.
Petersen, C.G.J. 1915. On the animal communities of the sea bottom in the Skagerak, the
Christiania Fjord and the Danish waters. Report from the Danish Biological Station 23:1-
28.
Pihl, L., S.P. Baden, and RJ. Diaz. 1991. Effects of periodic hypoxia on distribution of demersal
fish and crustaceans. Marine Biology 108:349-360.
Pohl, C., U. Hennings, I. Petersohn, and H. Siegel. 1998. Trace metal budget, transport,
modification and sink in the transition area between the Oder and Peene rivers and the
southern Pomeranian Bight. Marine Pollution Bulletin 36:598-616.
Mallin, M.A., M.H. Posey, G.C. Shank, M.R. Mclver, S.H. Ensign, and T.D. Alphin. 1999.
Hurricane effects on water quality and benthos in the Cape Fear watershed: Natural and
anthropogenic impacts. Ecological Applications 9:350-362.
Powilleit, M., and J. Kube. 1999. Effects of severe oxygen depletion on macrobenthos in the
Pomeranian Bay (southern Baltic Sea): a case study in a shallow, sublittoral habitat
characterised by low species richness. Netherlands Journal of Sea Research 42:221-234.
Prena, J. 1995a. Effects of eutrophication on macrobenthos zonation in Wismar Bay (western
Baltic Sea). Archiv fur Hydrobiologie. Stuttgart 133:245-257
Prena, J. 1995b. Temporal irregularities in the macrobenthic community and deep-water
advection in Wismar Bay (western Baltic Sea). Estuarine, Coastal and Shelf Science
41:705-717.
Rabalais, N.N., and Turner, R.E. 2001. Hypoxia in the northern Gulf of Mexico: description,
causes and change. In: Coastal Hypoxia Consequences for Living Resources and
Ecosystems. N.N. Rabalais and R.E. Turner (eds.). American Geophysical Union,
Washington, D.C. pp. 1-36.
Rabalais, N.N., Turner, R.E., Wiseman, Jr. W.J. andBoesch, D.F. 1991. A brief summary of
hypoxia on the northern Gulf of Mexiaco continental shelf: 1985-1988. In Modern and
Ancient Continental Shelf Anoxia. R.V. Tyson and T.H. Pearson (eds.). London:
Geological Society Special Publication No. 58, pp. 35-47.
Rabalias, N.R., L.E. Smith, D.E. Harper, Jr. 2001. Effects of seasonal hypoxia on continental
shelf benthos. In: Coastal Hypoxia Consequences for Living Resources and Ecosystems.
N.N. Rabalais and R.E. Turner (eds.). American Geophysical Union, Washington, D.C.
pp.185-210
197
-------
Rainer, S.F., and R.C. Fitzhardinge. 1981. Benthic communities in an estuary with periodic
deoxygenation. Australian Journal of Marine and Freshwater Research 32:227-243.
Reish, DJ. 1955. The relation of polychaetous annelids to harbor pollution. Public Health Report
70:1168-1174.
Reizopoulou, S., Thessalou-Legaki, M., and A. Nicolaidou. 1996. Assessment of disturbance in
Mediterranean lagoons: An evaluation of methods. Marine Biology. Berlin, Heidelberg
125:189-197.
Richards, F.A. 1965. Dissolved gases other than carbon dioxide. In: Chemical Oceanography,
Volume 1, J.P. Riley and G. Skirrow, (eds.). Academic Press, New York. pp. 197-226.
Rikard, F.S., R.K. Wallace, D. Rouse, and I. Saoud. 2000. The effect of low oxygen on oyster
survival during reef restoration efforts in Bon Secour Bay, Alabama. Journal of Shellfish
Research 19:640.
Ritter, C., and P.A. Montagna. 1999. Seasonal hypoxia and models of benthic response in a
Texas bay. Estuaries 22:7-20.
Rosenberg, R. 1980. Effects of oxygen deficiency on benthic macrofauna in fjords. In: Fjord
Oceanography, H.J. Freeland, D.M. Farmer and C.D. Levings (eds.), Plenum Publishing
Corp., New York. pp. 499-514.
Rosenberg, R. 1972. Benthic faunal recovery in a Swedish fjord following the closure of a
sulphite pulp mill. Oikos 23:92-108.
Rosenberg, R. 1976. Benthic faunal dynamics during succession following pollution
abatement in a Swedish estuary. Oikos 27:414-427.
Rosenberg, R., and R.J. Diaz. 1993. Sulphur bacteria (Beggiatoa spp.) mats indicate hypoxic
conditions in the Inner Stockholm Archipelago. Ambio 22:32-36.
Rosenberg, R., and L.O. Loo. 1988. Marine eutrophication induced oxygen deficiency: effects on
soft bottom fauna, western Sweden. Ophelia 29:213-225.
Rosenberg, R. 1990. Negative oxygen trends in Swedish coastal bottom waters. Marine Pollution
Bulletin 21:335-339.
Rosenberg, R., J.S. Gray, A.B. Josefson, and T.H. Pearson. 1987. Petersen's benthic stations
revisited. II. Is the Oslofjord and eastern Skagerrak enriched? Journal of Experimental
Marine Biology and Ecology 105:219-251.
Rosenberg, R., L.O. Loo, and P. Moller. 1992. Hypoxia, salinity and temperature as structuring
factors for marine benthic communities in a eutrophic area. Netherlands Journal of Sea
Research 30:121-129.
198
-------
Rumohr, H. 1986. Historical evidence for eutrophication processes in the Kieler Bucht (western
Baltic). Meeresforschung 31:115-123.
Rybarczyk, H., B. Elkaim, J.G. Wilson, and N. Loquet. 1996. Eutrophication in the Baie de
Somme: Consequences and impacts on the benthic population. Oceanologica acta
19:131-140.
Santos, S.L. and J.L. Simon. 1980. Marine soft-bottom community establishment following
annual defaunation: larval or adult recruitment. Marine Ecology Progress Series 2:235-
241.
Sanukida, S., H. Okamato, and M. Hitomi. 1984. Alteration of pollution indicator species of
macrobenthos during stagnant period in eastern Hiuchi Sound. Bulletin Japanese Society
of Scientific Fisheries 50:727.
Schimmel, S.C., SJ. Benyi, and CJ. Strobel. 1999. An assessment of the ecological condition of
Long Island Sound, 1990-1993. Environmental Monitoring and Assessment 56:27-49.
Sindermann, C., andR. Swanson. 1980. Chapter 1. Historical and Regional Perspective. In:
Oxygen Depletion and Associated Benthic Mortalities in New York Bight, 1976. R.L.
Swanson and CJ. Sindermann (eds.) Rockville. U.S. Department of Commerce National
Oceanic and Atmospheric Administration, pp. 1-16.
Simunovic, A., C. Piccinetti, and M. Zore-Armanda. 1999. Kill of benthic organisms as a
response to an anoxic state in the northern Adriatic (a critical review). Acta Adriatica
40:37-64.
Souchu, P., A. Gasc, Y. Collos, A. Vaquer, H. Tournier, B. Bibent, and J-M eslous-Paoli. 1998.
Biogeochemical aspects of bottom anoxia in a Mediterranean lagoon (Thau, France).
Marine Ecology Progress Series 164:135-146.
Stachowitsch, M. 1991. Anoxia in the Northern Adriatic Sea: rapid death, slow recovery. In:
Modern and Ancient Continental Shelf Anoxia. R.V. Tyson and T.H. Pearson (eds.).
London: Geological Society Special Publication No. 58. pp. 119-129.
Stanley, D.W., and S.W. Nixon. 1992. Stratification and bottom-water hypoxia in the Pamlico
River Estuary. Estuaries 15:270-281.
Suzuki, T., and Y. Matsukawa. 1987. Hydrography and budget of dissolved total nitrogen and
dissolved oxygen in the stratified Mikawa Bay, Japan. Journal of the Oceanographic
Society of Japan 43:37-48.
Tagliapietra, D., M. Pavan, and C. Wagner. 1998. Macrobenthic community changes related to
eutrophication in Palude della Rosa (Venetian Lagoon, Italy). Estuarine, Coastal and
Shelf Science 47:217-226.
199
-------
Tarazona, J., H. Salzwedel and W. Arntz. 1988. Positive effect of "El Nino" on macrozoobenthos
inhabiting hypoxic areas of the Peruvian upwelling system. Oecologia 76:184-190.
Tenore, K.R. 1972. Macrobenthos of the Pamlico River estuary, North Carolina. Ecological
Monographs 42:51-69.
Thiel, R., A. Sepulveda, R. Kafemann, and W. Nell en. 1995. Environmental factors structuring
the fish community of the Elbe estuary. Journal of Fisheries Biology 46:47-69.
Theodorou, AJ. 1996. Long-term environmental effects of raw sewage sea disposal in Elefsis
Bay (Saronikos Gulf, Greece). Partnership in Coastal Zone Management, Samara
Publishing Ltd., Cardigan, UK. pp. 547-556.
Thomassin, B.A., P. Gourbesville, B. Gout, and A. Arnoux. 1998. Impact of an industrial and
urban sewage off a coral fringing reef at Mauritius (Indian ocean): modeling of plumes,
distribution of trace metals in sediments and effects of the eutrophication on coral reef
communities. Oceans 98 — Engineering for sustainable use of the oceans: conference
proceedings 28 September - 1 October 1998, Nice (France)., IEEE/OES, Piscataway, NJ
(USA), vol. 1, pp. 301-305
Thouvenin, B., P. Le Hir, and L.A. Romana. 1994. Dissolved oxygen model in the Loire Estuary.
In: Changes in Fluxes in Estuaries: Implications from Science to Management. Olsen and
Olsen, Fredensborg, Denmark, pp. 169-178.
Tsutsumi, H. 1987. Population dynamics of Capitella capitata (Polychaeta, Capitellidae) in an
organically polluted cove. Marine Ecology Progress Series 36:139-149.
Tulkki, P. 1965. Disappearance of the benthic fauna from the basin of the Bornholm (southern
Baltic) due to oxygen defiency. Cahiers de Biologie Marine 6:455-463.
Tunnicliffe, V. 1981. High species diversity and abundance of the epibenthic community in an
oxygen-deficient basin. Nature 294:354-356.
Valentin, J., D. Tenenbaum, A. Bonecker, S. Bonecker, C. Nogueira, R. Paranhos, and M.C.
Villac. 1999. Hydrobiological characteristics of the GuanabaraBay (Brazil). Journal de
Recherche Oceanographique 24:33-41.
Verlaan, P.A.J., M. Donze, and P. Kuik. 1998. Marine vs fluvial suspended matter in the Scheldt
Estuary. Estuarine, Coastal and Shelf Science 46:873-883.
Weigelt, M. 1990. Oxygen conditions in the deep water of Keil Bay and the impact of inflowing
salt-rich water from the Kattegatt. Meeresforschung 33:1-22.
Weisberg, S.B., H.T. Wilson, P. Himchak, T. Baum, and R. Allen. 1996. Temporal trends in
abundance offish in the tidal Delaware River. Estuaries 19:723-729.
200
-------
Welsh, B., R. Welsh, and M. DiGiacomo-Cohen. 1994. Quantifying hypoxia and anoxia in Long
Island Sound. In: Changes in Fluxes in Estuaries: Implications from Science to
Management. K. Dyer, and R. Orth (eds.): Olsen and Olsen. Fredensborg: Olsen and
Olsen. pp. 131-137.
Westernhagen, H.V., and V. Dethlefsen. 1983. North sea oxygen deficiency in 1982 and its
effects on the bottom fauna. Ambio 12:264-266.
Wilkin, R., and H.L. Barnes. 1997. Pyrite formation in an anoxic estuarine basin. American
Journal of Science 297:620-650.
Wu, R.S.S. 1982. Periodic defaunation and recovery in a subtropical epibenthic community, in
relation to organic pollution. Journal of Experimental Marine Biology and Ecology
64:253-269.
Yamamuro, M., N. Oka, and J. Hiratsuka. 1998. Predation by diving ducks on the biofouling
mussel Musculista senhousia in a eutrophic estuarine lagoon. Marine Ecology Progress
Series 174:101-106.
Zaitsev, Y.P. 1993. Impacts of eutrophication on the Black Sea fauna. In: Fisheries and
Environment Studies in the Black Sea System, Food and Agricultural Organization of the
United Nations, Rome, Italy, pp. 64-86.
201
-------
202
-------
CHEMICAL PROCESSES IN THE ANOXIC ZONES
OF THE BALTIC SEA
Vida-Judita Sukyte1
ABSTRACT
The mechanism of oxidation of H2S in Baltic Sea water has been modelled in the
laboratory based on our hydrochemical investigation data from the Baltic Sea. Our preliminary
potentiometric and analytical investigations have enabled us to evaluate existing mechanisms of
sulfide oxidation in seawater, and to propose a new one that acts through a complicated chain
reaction caused by active forms of oxygen. This new mechanism involves formation of
intermediate, unstable sulfur compounds including HSO~ and HSO2~. These latter compounds,
reacting with other products of a chain process, form the final products that have been
analytically observed: hyrdrogen sulfide (HSn~), elemental sulfur (S°), thiosulfate (S2O32~), and
sulfate (SO42~). The rates of oxidation of H2S in air-saturated distilled water and seawater have
been measured as a function of temperature (5° and 15°C) and salinity (18 and 22%o) at pH 7.2.
At 15°C and pH 7.2, our results expressed as the half-life of oxidation (ti/2) are: ti/2 = 67 hours in
distilled water, ti/2 = 34 hours in 18%o seawater, and ti/2 = 23 hours in 22%o seawater. The
shorter half-lives were obtained for the most saline seawater. We also found that the rate of
oxidation was independent of H2S concentration and quantified the pseudo-first-order rate
constants for the oxidation of H2S in water and seawater.
INTRODUCTION
At very irregular intervals, some parts of the bottom of the Baltic Sea may be entirely
devoid of fauna for months or years as a result of oxygen deficiency and the subsequent presence
of hydrogen sulfide. At its worst, these desert-like areas may comprise up to 100,000 km2,
equivalent to approximately 25% of the total surface area of the Baltic Sea. The low oxygen and
hydrogen sulphide stress factors fluctuate intermittently and irregularly (Andersin etal. 1977,
Voipio 1981). Hydrogen sulfide is oxidized in the sea water redox zone by microbiological as
well as by chemical processes (Leonov and Aizatullin 1987). Using antiseptics to suppress
microbiological activity, it has been determined that H2S is oxidized by a chemical pathway
(Sorokin 1970). The products of H2S oxidation by oxygen in sea water have been determined by
analytical and indirect methods. The basic products are as follows: elemental sulfur (S°),
polysulfides (Sn2"), sulfites (SO32"), thiosulfates (S2O32"), and sulfates (SO42"). In one report,
dithionite (S2O42"), dithionate (S2Oe2"), and tetrathionate (S^e2") were found among other
products when the pH was less than seven (Leonov and Aizatullin 1987).
Given the variety and number of oxidation products, one can assume that the mechanism
of oxidation is very complicated and still not fully understood. It is hypothesized that H2S
Associate Professor, Kaunas University of Technology, Kaunas, Lithuania.
203
-------
oxidation proceeds in accordance with a chain reaction mechanism (autocatalytic reaction) with
Sn2" ions serving as catalysts (Bowers et al. 1966, Chen and Morris 1971). However, other
oxidation reaction products, such as SC>42" and S2O32" ions, act as inhibitors and slow the reaction.
Oxidation of H2S is both slowed-down and speeded-up by various catalysts, such as calcium or
magnesium ions, heavy metals, phosphates, pH, contact surfaces, and organic substances. Such
a great sensitivity of reaction kinetics to various admixtures and media parameters makes it
difficult to assess the theoretical H^S oxidation rate in seawater based on studies under natural
conditions. The most reliable path is actual analytical data on Baltic Sea seawater combined
with laboratory experiments modeling natural conditions. For this purpose, H2S oxidation has
been modeled in water of the Baltic Sea under laboratory conditions.
MATERIALS AND METHODS
Anaerobic processes in the Baltic Sea bottom have been investigated using seasonal
expeditions carried-out 1981-1991 by the Klaipeda Hydrometeorological Observatory, and by
the Lithuania Environmental Protection Department and Marine Research Center (Ministry of
Environment) 1991-2000 under the National Baltic Sea Monitoring Programme (BMP) (Figure
1).
Figure 1. Location of sampling stations: 31- The Fazo Deep (205 m); 37 - The Gotland Deep
(249 m); 55 - Gdansk Deep (105 m); 62n - The Bornholm Basin (105 M); 69n - The
Karlso Deep (114 m); 71n - Landsort Deep (459 m).
204
-------
Determinations of H2S by a colorimetric method with p-phenylenediamine-
dihydrochloride (Fonselius 1976) were carried-out almost 20 years ago (Anonymous 1976-
1991). Since 1981, H2S has been analyzed routinely at a depth of 80 m and deeper. Hydrogen
sulfide oxidation modeling using water from the Baltic Sea was carried out under laboratory
conditions (Kaunas Technological University) in 1991-1998 and 2000.
Modeling of Hydrogen Sulfide Oxidation Reactions
Modeling of H2S oxidation was carried-out in samples of water from the Baltic Sea under
laboratory conditions. Recrystallized Na2S was used to prepare a 0.2 M standard solution in
freshly boiled, distilled water, and kept isolated from the air. In the absence of dissolved oxygen
(O2), the mean H2S concentration in the Baltic Sea is about 0.1 mM. The solubility of O2 at a
temperature of 5°C is 55 cmVdm3 or 1.5 mM. Thus, in order to have a sufficient amount of
dissolved O2 for oxidation, 0.1-1 mM concentrations of Na2S solution were used for laboratory
testing.
Test solutions of Na2S were made-up in distilled water and seawater and stored in flasks
with corks slightly loose so that O2 might penetrate. After dissolving the Na2S in distilled water,
the pH of the resultant solution was 8.5; pH was reduced to the seawater level of 7.5 by adding a
drop of 0.1 M HC1. Upon dissolving Na2S in seawater, the pH level was unchanged, apparently
due to the buffering capacity of the seawater. The tests were carried-out at a temperature of
20°C. Periodically, samples were taken and titrated by an iodine (I2) solution in an alkaline
medium in order to determine the amount of sulfide had not been oxidized. A platinum
electrode, together with a saturated calomel electrode as a reference electrode, was used to
determine the end-point.
RESULTS
Formation and Change of Hydrogen Sulfide Zones
The redox conditions in the Baltic Sea bottom water and sediment surface have fluctuated
widely during the last 400 years (Hallberg 1974). O2 deficiency, as well as the formation of H2S,
in water layers below the salinity halocline at a depth of 70-90 m and in isolated basins was
observed 70 years ago (Granqvist 1932). The irregular interchange of aerobic and anaerobic
periods is presumably determined by the intensity of the inflow of the North Sea water
interacting with anthropogenic impacts (Fonselius 1962, 1969, FIELCOM 1996).
The O2 regime in the deep bottoms can be dated back to 1980. According to Nehring and
Matthaus (1991), there has been no dissolved O2 detected in the Gotland Deep at a depth of
248 m since 1980, and at a depth of 200 m since 1983. According to our data for 1984-1991,
dissolved O2 has never been observed in the Gotland Deep below 200 m. During the same
period in the Gdansk Deep at a depth of 100 m and below, H2S was observed to be occasionally
formed, and this H2S zone persisted for 4-5 months. In the Gotland Deep, however, a H2S zone
was constantly being formed, with the upper boundary at a depth of 130-140 m.
205
-------
During the period 1981-1991, slight decreases in H2S concentration were observed in the
bottom layer of the Gotland Deep, and for the most recent 3 years there has been an apparent
positive trend (Figure 2).
H2S, cm3 dm"
""3
3.0 -
2.0 -
10 -
o o
82 83 83 S4-
&6
68 89 90 -51
3 "3
H2S, cm3 dm
3.0 -
2.0 -
10 -
0.0
i — i — i — i
90
92
Figure 2. Long-term trends of annual means of H2S concentrate in the bottom water (below 200
m) of the Gotland Deep (source Anonymous 1976-1991).
The meteorological, hydrological and hydrographic conditions during the last two
decades of the second millenium showed some remarkable features that had a strong impact on
the Baltic Sea environment. The hydrographic conditions are characterized by continuation of
the stagnation period in the deep water of the eastern Gotland Basin during most of the
206
-------
assessment period. Extreme variations in hydrographic parameters have been observed in the
deep water of the central part of the Baltic Sea (Matthaus 1990, Nehring and Matthaus 1991). In
the course of the last 16 years, the salinity and temperature have decreased. The temperatures at
the beginning of the stagnation period were the highest ever measured near the bottom (Figure
3c; cf. Fonselius 1977), and the salinity observed at the end was the lowest recorded since the
beginning of regular measurements in the Baltic Sea (Figure 3a). The H2S concentrations in the
deep water of the Eastern Gotland Basin reached the highest value ever measured in the Baltic
Sea, and, in contrast, the O2 concentrations in the near-bottom layers of the western Gotland
Basin, Landsort and Karlso Deeps were observed to increase (Figure 3b).
1890 1910 1930 1950 1970 1990
1870 1890 1910 1930 1950 1970 1990
Oxygen Content/cm3W
3,
; 2'
1
0' -
-1
-2'
lS7B""(886""V9To""r9131o""1r915H""f976""T991o"
!.• • ' .
;:•• ;• •
. •<• •/•<,
Bernhelm Deep
80m.
** ... ;C
. »' % *»•
Gollend Deep
200m.
Faro Deep
150m.
'.•'•" v
Landsort Deep
400 m.
*. »
Karlso Deep
100 m.
:;
:
»
3
1™
Temperature / C
Bernhelm Deep
Gollend Deep
200 m.
Faro Deep
150m.
Landsort Deep
400 m.
Karlso Deep
100 m.
1910 1930 1950 1970 19
1870 1890 1910 1930 1950 1970 1990
Figure 3. Long-term variations in (a) salinity, (b) oxygen, and (c) temperature in the
deep water of the Baltic Sea during the present century. (Black bars: assessment
period) (RELCOM 1996)
The year 1990 marked a change in the general water exchange pattern between the North
Sea and the Baltic Sea. Small inflows to the Baltic Sea occurred in early 1990 and during the
turn of both 1990 to 1991 and 1991 to 1992. In January 1993, after 16 years of stagnation, a
major inflow occurred in the eastern Gotland Basin (Hakanson et al. 1993). During a 3-week
period of strong westerly winds, a total of about 310 km3 of water, 150 km3 of which was highly
saline (> 17%o) and oxygenated, entered the Baltic Sea and resulted in a large increase in the
volume of the Baltic Sea (70 cm above mean water level).
The stagnation period was finally terminated by inflows of lower magnitude in December
1993 and in March 1994 (Matthaus et al. 1994). During spring 1994, O2 concentrations of about
3-3.8 cm3/dm3 were measured between 170 m and the bottom of the Gotland Deep. These were
the highest concentrations observed at this station since the 1930s (Figure 3b). For the first time
since 1977, the whole central part of the Baltic Sea was free of H2S.
207
-------
An intermediate redox zone, where both dissolved O2 and H2S are observed, exists in the
deep layers of the Baltic Sea above the H2S zone (Table 1). Hydrogen sulfide oxidation in this
redox zone is one of the main chemical processes upon which changes of other hydrochemical
variables depend, as well as does the general condition in the bottom layers of the Baltic Sea.
Table 1. Background ranges of pH, O2, and H2S in the redox layer of the Baltic Sea in
1981-1991.
Stations
24
25
27
30
31
36
37
38
42
43
50
55
62 n
78
69 n
71
71n
Depth
(m)
100-105
80-83
100-150
80-118
150
80-108
80-100
150
80-100
150-160
80-100
150-158
80-100
150
83
95-104
80-88
78
100-108
100
150-195
150
200-249
pH
7.32-7.36
7.56-7.88
7.18-7.35
7.46-7.47
6.88-7.56
7.16-7.51
7.00-7.68
7.19-7.51
7.17-7.786
7.31-7.62
7.30-7.47
7.18-7.67
7.27-7.50
7.32-7.75
7.28
7.28-7.44
7.36-7.72
7.21
7.20-7.49
7.33
6.96-7.36
7.00-7.85
7.21-7.43
02
(cm3/dm3)
1.04-1.71
6.60-3.73
0.25-0.99
1.94-2.30
0.24-0.68
0.29-2.32
1.39-3.63
0.09-0.91
0.54-3.49
0.39-0.91
2.62-4.17
0.23-0.51
0.81-5.55
0.49-2.72
0.76
0.14-2.11
0.13-2.58
1.08
0.23-6.02
0.27
0.12-1.08
0.13-1.11
0.15-073
H2S
(cm3/dm3)
0.06-0.18
0.06-0.08
0.11-0.17
0.08-0.15
0.08-0.10
0.06-0.17
0.06-0.10
0.07-0.70
0.08-0.14
0.07-0.16
0.07-0.08
0.16-0.17
0.06-0.19
0.15-0.18
0.15
0.06-0.14
0.09-0.16
0.06
0.07-0.19
0.08
0.06-0.18
0.06-0.17
005-0.15
In the central part of the Baltic Sea, the redox zone is between 80-150 m below the
surface. In the Landsort Deep (station 71n), the redox zone is between 140 and 250 m below the
surface. Dissolved O2 concentration in this zone was 0.0-1.0 cm3/dm3, although occasionally as
high as 5.5-6.0 cm3/dm3, and the pH was 6.9-7.8. Relative alkalinity affects the oxidation
process of H2S and the resulting sulfur forms in solution. The predominate form of H2S at pH 7
208
-------
is HS . The rates of redox reactions are partially dependent upon concentrations, and the
assumption is made that there are minimal biochemical processes occurring.
Many scientists have suggested models and empirical dependences for calculation of
some redox zone indices (Stanev 1986, Leonov and Aizatullin 1987); however, standard
expeditional investigations are not fully efficient to develop the necessary data for their testing.
Detailed investigations of vertical profiles of H2S and O2 concentration distributions, as well as
modeling in the laboratory, are also required.
Modeling and Mechanism of HiS Oxidation
A review of the curves of potentiometric titration (Figure 4) show that the redox potential
of a fresh H2S solution increased from -350 mV to +300 mV during the oxidation by I2.
300
200 -
100 -
E
LLf
-100
-200
-300 -
-400
10 12
-400
0, 01I2,cmJ
Figure 4. Potentiometric titration curves of 0.85 mM/dm3 H2S by I2 in alkaline solution: (1)
initial solution; (2) after 20 hours oxidation; (3) after 70 hours oxidation at 20°C.
• = seawater; A = distilled water.
Two abrupt potential in changes can be observed in the titration curves. The first change
(at 6 eqv. I2/mol H2S) corresponds to HS~ oxidation to the SOs2" ion:
HS" + 3I2 + 3H2O -» HSO3
6HI.
209
-------
The second change (at 8 eqv. I2/mol H2S) corresponds to full H2S oxidation, according to the
equation:
HS" + 4I2 + 4H2O -» HSO4" + 8HI.
During the period of H2S oxidation, the initial negative potential decreases, and after all the HS~
has been transferred to Sn2", it remains at -100 + 60 mV (Figure 4).
Kinetic H2S oxidation curves (Figure 5) show that the oxidation in sea water is slightly
slower than in distilled water; the reaction apparently impeded by the ionic composition of the
sea water. According to our calculations, intermediate oxidation products account for 10% of the
initial amount of H2S. The remainder is in the form of elemental crystalline sulfur (Sg), which is
precipitated, and by SC>42" ions, the origin of which is difficult to determine because of their
naturally high concentration in seawater.
10 r
50 100
Time, h
150
50
150
200
Time, h
Figure 5. Kinetic curves of H2S oxidation (pH 7.5) at 20°C: (1) 0.987 mM/dm3H2S; (2) 0.625
mM/dm3H2S. • = seawater, + = distilled water.
The Rate of Oxidation
Avrahami and Golding (1968) studied oxidation of S2" in alkaline solutions at pH 11-14
0 0 0 ••
and found that S " is initially oxidized to S2C>3 " and subsequently to 864 " ions. Ostlund and
Alexander (1963) found that the half-life of S2" oxidation in seawater saturated with O2 was 20
minutes. Cline and Richards (1969) concluded that oxidation of S2" in seawater is a complex
mechanism and its half-life is approximately 15 hours at 8-9°C. Chen and Morris (1972) stated
that the pH dependence of H2S oxidation by O2 dissolved in water is very complex and the half-
life of the reaction is about 50 hours at pH 7.94. Millero et al. (1987) investigated H2S oxidation
in sea water and in NaCl solutions and arrived at the conclusion that the half-life of S2" ion
oxidation is 50 hours and that the overall mechanism of the reaction is close to first order.
Leonov and Aizatulin (1987) investigated the kinetics of H2S oxidation in water of the Black Sea
210
-------
and suggested a mechanism for the reaction where one of the intermediate reactions is a tri-
molecular reaction; this being hard to believe. Our review of the previously cited articles and
other well-known data revealed great differences in the experiments and the results obtained.
Furthermore, there have been only a few investigations carried out under conditions similar to
that of the deep bottoms of the Baltic Sea (Almgren and Hagstrom 1974). Therefore, further
studies on the kinetics and mechanism of H2S oxidation in the waters of the Baltic Sea are
necessary.
We studied the oxidation of S2" ions by 62 dissolved in water as a function of salinity and
temperature. For our investigation, we used recrystallized Na2S-9H2O. The initial solution of
8.34 mM/dm3 Na2S was prepared by dissolving rinsed Na2S-9H2O crystals in twice-distilled and
degassed H2O. Sea water with salinity 18%o and 22%o was prepared from artificial sea water by
diluting it with distilled water. In 1992-1993, the concentration of H2S in water of the Gotland
Deep in the Baltic Sea was less than or equivalent to 2 cmVdm3 (Table 2). Therefore, for our
measurements we used Na2S solutions with concentrations of 0.5 mM/dm3 and 0.1 mM/dm3. To
make-up the test solutions, water saturated with air was used. The solubility of 62 in water at
5°C is 43.5 cm3/dm3 and at 15°C it is 35 cm3/dm3, i.e., 1.94 mM/dm3 and 1.56 mM/dm3
respectively. Accordingly, in these saturated solutions, the quantity of C>2 was sufficient to
oxidize the S2" ions present in our samples to SO42" ions.
Table 2. Seasonal means of H2S (cnrVdm3) in the Gotland Deep 1992-1993.
Depth
(M)
125
150
175
200
225
240
Winter
0.00
0.00
0.96
1.22
1.24
1.25
Spring
0.00
0.81
0.81
0.85
0.85
0.87
1992
Summer
0.00
0.78
1.35
1.32
1.35
1.39
Autumn
0.00
0.93
1.31
1.32
1.36
1.44
Winter
0.00
1.02
1.25
1.38
1.28
1.16
Spring
0.00
2.13
2.05
0.73
0.00
0.00
1993
Summer
0.00
0.10
0.63
0.00
0.00
0.00
Autumn
0.00
0.18
0.21
0.00
0.00
0.44
The oxidation studies were carried out at temperatures of 5 and 15°C and for dilution we
used sea water at the selected salinities saturated with air and air-saturated distilled water at these
temperatures. The pH of water in the areas of the Baltic Sea where H2S is formed is near 7.2.
For this reason, the pH of all samples was adjusted to 7.2 through the addition of dilute HC1 by
using a pH meter with glass and calomel electrodes. Samples during the oxidations were
thermostated at temperatures of 5 and 15°C. In order to determine the rate of S2" ion oxidation
by O2 dissolved in the water, we used a potentiometric titration method to measure the residual
concentrations of non-oxidized S2". This involved measuring 12 concentration in alkaline
211
-------
solution with a platinum electrode and a saturated calomel electrode. All the measurements were
made in solutions periodically reaerated, thus ensuring an excess of O2 during the whole process.
DISCUSSION
Mechanism of Oxidation of H2S
As previously mentioned, the mechanism of oxidation of H2S by soluble O2 has not been
completely determined. Reactions sometimes presented in the literature, e.g. by Leonov (1987),
are as follows:
2HS" + O2 -» 2S° + 2OFT (1)
(n-l)S° + HS~^Sn2~ + H+ n = 2to8 (2)
These reactions are not correct because the product is S°, which later reacts with
dissolved H2S to form Sn2" ions. Besides, this reaction is tri-molecular, and the occurrence of
such reactions is rather doubtful.
Through investigation of S2O32" acid decomposition, H2S and SO2 interaction in
Wackenroder's liquid, and other polythionate reactions, it has been determined that S° appears in
the form of Sg rings. This molecule is formed through a number of intermediate products. For
example, S2O32" acid decomposition expressed by the simple equation
S2O32" + 2H+ -» S + H2SO3 (3)
proceeds through the intermediate products sulfan-monosulfonates (HSnSO3H) (Yanitskij et al.
1971), where n = 1 to 8. Only after the product HSgSO3H with the greatest amount of sulfur has
been formed does a molecule of Sg appear, and the rest of the sulfur is in the form of sulfurous
acid (H2SO3). Intermediate products of this reaction, HSnSO3H, were isolated as complex nitron
(Nt) salts.
From our investigations, we conclude that the main product of H2S oxidation is molecular
sulfur (Sg). We hypothesize that Sg results from a number of already identified intermediate
products, namely Sn2". Isolation of Sg by reaction (2) is doubtful. There are data in the literature
indicating that the initial H2S oxidation stage is a chain reaction, although the radicals or ion
radicals participating in the reaction are unknown. The most believable initiator of the chain
reaction is active oxygen (Ox), which can exist in several forms.
Thus, the first stage of the reaction would be a reaction with Ox:
HS~ + Ox -» HSO~ (4)
212
-------
This hypothetical, unstable compound could promote the further chain reaction:
HSCT + O2 -» HSO2~ + Ox (5)
The sulfoxylic acid anion, HSO2~, has been isolated as a salt and is a very active
compound. Further interaction of HSO2~ with HS~ can form polysulfides:
HSO2" + HS" -» HS2O" + OFT (6)
HS2O" + HS" -» HS3" + OFT (7)
When the reaction stage for HSgS" formation is reached, Sg is formed. However, HSO2~,
because of its reactivity, can react with other substances participating in the reaction, resulting in
hydrosulfite, sulfate, and thiosulfate:
HSO2~ + Ox -» HSO3" (8)
HSO2~ + O2 -» HSO4" (9)
HSO2" + HSO2" -» HS2O3" + OH" (10)
These reactions are much more probable since they are homogeneous and bimolecular;
similar mechanisms have been proven for other reaction systems. These reactions also
correspond to the contemporary view of sulfur chemistry, and are confirmed by the variety of
products observed during HS~ oxidation.
The Rate of Oxidation
The shape of the potentiometric titration curves (Figure 4), and the fact that all the
solutions remained transparent, even after a month, that is, they did not show any traces of
elemental sulfur, allow us to assume that sulfide oxidation is proceeding through an intermediate
stage of Sn2" and HSO2" formation, but not through that of S°.
We studied the rates of S2" ion oxidation by O2 dissolved in water using 0.5 mM/dm3
Na2S solutions in distilled water and sea water with salinities of 18%o and 22%o at temperatures
of 5 and 15°C. After completing oxidation experiments at 0.1, 0.5, and 1.0 mM/dm3 Na2S
concentrations in oxygen-saturated distilled water, we determined that the half-life of the S2" ion
oxidation reaction in distilled water is approximately 67 hours and is independent of S2" ion
concentration. Consequently, we concluded that in the presence of excess O2 at zero salinity, the
reaction proceeds by a zero-order mechanism.
We also observed that the salinity of sea water had an effect on the half-life of the S2"
oxidation process. In Figure 6, kinetics curves are drawn for the oxidation process. It may be
seen from Figure 6 that a 0.25 mM/dm3 concentration of S2" ions reacted most rapidly in the
more saline sea water (22%o), and more slowly in distilled water. The observed half-life for the
oxidation of S2" ions at a temperature of 15°C in water with a salinity of 22%o was about 3 hours,
213
-------
while that at a salinity of 18%o at the same temperature was about 13 hours. In distilled water the
half-life was approximately 67 hours (Table 3). By comparing the rates of S2" oxidation at two
different temperatures (Table 3), we determined that the temperature coefficient of reaction in
distilled water is 1.34, and in water with salinity of 18%o is 1.21. When the salinity was 22%o
and the temperature was 5°C, the concentration of unreacted S2" ions was smaller than for the
same oxidation period at 15°C. The greater solubility of Q^ in the colder water may be one of the
causes for this effect. The pseudo-first order mechanism of reaction is possible only when a
great excess of O2 is present such as in the case with the 0.5 mM/dm3
solution at 15°C. To confirm these assumptions, further investigations are required.
T \
- S " ion concentration test
0,5 I
0,45
0,4
0,35
j. *~
. /-— /—
' — * ""*-! f ^ + /_
K " j ,
0
12
60
72
24 .36 48
Time, h
Figure 6. Kinetic curves of oxidation of 0.5 mM/dm3 Na2S as function of concentration
of S2": (1) in distilled water; (2) in seawater with salinity 18%o; (3) in
seawater with salinity 22%o.
Table 3. Effect of salinity of water for sulfide ion oxidation at 5°C and pH 7,2.
Reaction medium and salinity, %o
Distilled water
Seawater, 18 %0
Seawater, 22 %0
Time of oxidation, (hours)
1.5
48
72
288
1.7
2.5
48
144
2
24
30
72
Concentration of sulfide, mM/dm3
0.285
0.27
0.255
0.23
0.29
0.275
0.249
0.22
0.28
0.248
0.245
0.2
214
-------
A comparison of our results, expressed as half-lives, with other workers indicated that
our results for distilled water are in good agreement with those obtained by O'Brien and Birkner
(1977) and Chen and Morris (1972). The seawater results are in good agreement with the work
ofMitteroetal. (1987).
We believe that our estimates of the oxidation rates of S2" should be useful in
understanding the formation of toxic chemicals in the anoxic zones, and as parameters for fate
and hazard assessment models.
In our further investigations, we have endeavored to isolate reaction products, both to
confirm our proposed reaction mechanism and to provide information required in assessing the
lexicological significance of these reactions and isolated compounds. The knowledge of the
significance of the effects of these sulfur compounds is far from complete.
ACKNOWLEDGEMENT
The author thanks Professor Robert V. Thurston, Fisheries Bioassay Laboratory,
Montana State University, Chairman of the Organizing Committee for an invitation to the Sixth
International Symposium "Hypoxia in the Aquatic Environment" and editing the Proceedings of
the La Paz Symposium.
REFERENCES
Almgren, T., and I. Hagstrom. 1974. The oxidation rate of sulfide in sea water. Water Research
8: 395-400.
Andersin, A.B., J. Lassing, and H. Sandier. 1977. Community structures of softbottom
macrofauna in different parts of the Baltic. In: N. F. Keegan, P. O'Ceidish, and P. Y. S.
Doaden (eds.), Biology of Benthic Organisms. Pergamon Press, Oxford, New York,
pp.7-20.
Anonymous. 1976-1991. Scientific-Technical Reports of Cruise of R/V Oceanograph, R/V Lev
Titov, and R/V Rudolph Samoilovich. Hydrometeorological Observatory, Klaipeda. (In
Russian).
Avrahami, M., and R.M. Golding. 1968. The oxidation of the sulfide ion at very low
concentrations in aqueous solutions. Journal of Chemical Society (A): 647-651.
Bowers, J.W., M.J.A. Fuller, and J.F. Packer. 1966. Autooxidation of aqueous sulfide solution.
Chemistry and Industry 2:65-66.
215
-------
Chen, K.Y., and J.C. Morris. 1971. Oxidation of aqueous sulfide by 62: 1. General
characteristics and catalytic influences. Advances in Water Pollution Research 2. p. III-
32/I-III32/17.
Chen, K.Y., and J.C. Morris. 1972. Kinetics of oxidation of aqueous sulfide by 62.
Environmental Science and Technology 6:529-537.
Cline J.D., and F.A. Richards. 1969. Oxygenation of hydrogen sulfide in sea water at constant
salinity, temperature and pH. Environmental Science and Technology 3:838-843.
Fonselius, S.H. 1962. Hydrography of the Baltic Deep Basins. Fishery Board of Sweden,
Series Hydrography, Report No. 13, Lund, p. 9.
Fonselius, S.H. 1969. Hydrography of the Baltic Deep Basins III. Fishery Board of Sweden,
Series Hydrography, Report No. 23, Lund, p. 36.
Fonselius, S.H. 1976. Determination of hydrogen sulfide. In: K. Grasshoff (ed.) Methods of
Seawater Analysis. Weinhein, pp. 71-78.
Fonselius, S.H. 1977. An inflow of unusually warm water into the Baltic deep basins.
Meddelande fran Havsfiskelaboratoriet, Lysekil, Hydrografiska avdelningen, Goteborg
229:1-15. (In Swedish).
Granqvist, G. 1932. Annual Report: Croisiere thalassologique et observations en lateaux
routiers en 1931. Havsforskningsinstitutets skriftNr. 81, p. 24. (In French).
Hallberg, R.O. 1974. Paleoredox conditions in the Eastern Gotland basin during the recent
centuries. Merentutkimuslaitoksen Julk./Havsforkningsinstitutts Skrift 238:3-16.
Hakansson, B., B. Broman, and H. Dahlin. 1993. The flow of water and salt in the Sound
during the Baltic major inflow event in January 1993. International Council for the
Exploration of the Sea. Statutory Meeting Dublin, Paper C. M. 1993/C:57.
HELCOM (Baltic Marine Environment Protection Commission - Helsinki Commission), 1996.
Third Periodic Assessment of the State of the Marine Environment of the Baltic Sea
1989-1993. Baltic Sea Environment Proceedings, No. 64B.
Leoniv, A.V., and T.A. Aizatullin. 1987. Kinetics and mechanism of oxidation of hydrogen
sulfide in seawater. Water Resources 1:89-103. (In Russian).
Matthaus, W. 1990. Long-term trends and changes of oceanological parameters during the
recent stagnation period in the deep water of the central Baltic Sea. Fischerei -
Forschung Wissenschaftlish Schriftreihe Rostock 28:25-34 (in German).
216
-------
Matthaus, W., D. Nehring, and G. Nausch. 1994. Effects of the inflows of salt-rich water during
1993 and early 1994 in the central Baltic Sea. International Council for the Exploration of
the Sea. C. M./Q.:3.
Millero, F.J., S. Hubinger, M. Fernandez, and S.Garnett. 1987. Oxidation of H2S in seawater as
a function of temperature, pH, and ionic strength. Environmental Science and
Technology 21:439-443.
Nehring, D., and W. Matthaus. 1991. Current trends in hydrographic and chemical parameters
and eutrophication in the Baltic Sea. International Revue Gesanten Hydrobiolgie 76:297-
316.
O'Brien, D.J., and F.G. Birkner. 1977. Kinetics of oxygenation of reduced sulfur species in
aqueous solution. Environmental Science and Technology 11:1114-1120.
Ostlund, G.H., and J. Alexander. 1963. Oxidation rate of sulfide in sea water. Journal of
Geophysical Research 68:3995-3997.
Skopintsev, B.A. 1975. Formation of the Present State of Chemical Composition of Water of
the Black Sea. Leningrad, Hydrometeoizdat, 336 pp. (In Russian).
Sorokin, J.I. 1970. Experimental research of the rate and mechanism of the oxidation of
hydrogen sulfide in the Black Sea by 35S. Oceanology 10:51-62. (In Russian).
Stanev, E. 1986. Determination of the depth of the redox zone in the Black Sea. Oceanology
26:439-445 (In Russian).
Voipio, A. 1981. The Baltic Sea. Elsevier Oceanographic Series No. 30. Elsevier, Amsterdam,
418pp.
Yanitskij, J.V., V.J. Zelionkaite, and V.J. Janitskis. 1971. Isolation of intermediate products of
acidic decomposition of thiosulfate. Journal of Inorganic Chemistry 16/3:617-621. (In
Russian).
217
-------
218
-------
METAL AND MAJOR-ION REDOX CHEMISTRY OF THE
HYPOXIC AND ANOXIC ZONES: AN OVERVIEW
George W. Bailey1
ABSTRACT
This Symposium seeks to understand the direct effect of hypoxia on aquatic biota at the
individual, population, community, and the ecosystem level. Another concern, however, is the
indirect effect of varying oxygen levels on the thermodynamics and kinetics of biogeochemical
processes, and resulting impacts on the transformation, speciation, bioavailability, and toxicity of
particulate-bound and dissolved forms of inorganic and organic contaminants as they pass
through or reside in the hypoxia/anoxic water column and underlying sediment. The basic tenets
of redox chemistry, microbial respiration, electron transfer, and the effect of oxygen levels on the
nature and utilization of terminal electron acceptors are presented and reviewed in detail. Redox
chemistry of O, Fe, Mn, Cr, Hg, C, P, I, S, As, Se, Te, Pu, and U is reviewed. The speciation in
seawater of important metals - Cu, Cd, Co, Cr, Fe, Mn, Ni, Pb, and Zn - metalloids — As, and
Se, and common constituents - Ca, Cl, COs, H, OH, Na, Mg, and 864 - are evaluated using
calculations from MINTEQA4, a geochemical equilibrium model, and literature findings.
Possible adverse effects of redox-sensitive, chemically active forms on aquatic biota present in
the hypoxic zone are discussed.
INTRODUCTION
Hypoxia (depleted dissolved oxygen levels) may be the most widespread
anthropogenically induced deleterious effect in estuarine and marine environments. Hypoxia
causes mortality of bottom-dwelling fauna, including important fishery species. Over the last
two to three decades, the number of coastal areas with seasonal hypoxia in bottom waters has
increased rapidly. Areas impacted include: the southern North and Baltic Seas (Brugmann et al.
1998, Modig 1998, Paerl 1998, Bianchi et al. 2000); the northern shelf of the Black Sea; the
northern Gulf of Mexico (Turner and Rabalais 1991, Justic and Rabalais 1993, 1995, 1996,
Trefry et al. 1994, Justic et al. 1997, Burkart and James 1999, CAST 1999, Engle et al. 1999,
Howarth 2001, Rabalais etal. 2001, Synder 2001, Winstanley 2001); the Sea of Cortez, Mexico;
the Chesapeake Bay, Maryland-Virginia (Diaz 2001); the Kattegat, Sweden-Denmark (Pearson
1992, Nordberg 1998); the Skagerrak, Dutch Wadden Sea (GESAMP 1991); the Long Island
Sound; the Sea of Trieste, Italy; the northern Adriatic Sea (Bertuzzi 1996, Legovic 1997); the
Inland Sea of Japan; the Great Barrier Reef of Australia (Moss et al. 1992); the English fresh
water lakes (Achterberg et al. 1997); and many Scandinavian fjords and Mediterranean bays
(CAST, 1999). This problem can only grow in severity. Currently, 70% of the world's
population (more than 4.2 billion people) live within 60 km of coastlines (UNEP 1991; Wu
2000), and this number will continue to increase. Since the cost of treating waste to remove
Ecosystems Research Division, National Exposure Research Laboratory, U.S. Environmental Protection Agency,
Athens, Georgia, 30605-2700, USA.
219
-------
N and P is expensive (tertiary treatment is more than four times as expensive as secondary
treatment), enormous quantities of wastewater and nutrients will continue to be discharged into
coastal waters in the future.
Eutrophication and Hypoxia
The main cause of hypoxia/anoxia is eutrophication, caused by the transport of excess
nutrients — principally nitrogen (but also dissolved phosphorous and silica) — into marine
systems. The onset of hypoxia occurs after organic matter, produced as a result of excessive
nutrient loads, settles toward the bottom to decompose in the denser, saltier, lower layers. The
decomposition process consumes more oxygen than can be re-supplied at the air-water interface.
Dissolved oxygen levels decrease below concentrations needed to sustain sensitive marine life.
Hypoxia is strongly correlated with inflow fresh water nutrient flux, primary marine production,
and carbon flux. Another factor causing hypoxia is fresh water/salt water column stratification
that isolates bottom water from the surface-air interface supply oxygen. Major effects of
euthrophication include increased sedimentation of organic carbon, proliferation of filamentous
algae, and changes in benthic communities. Anoxia, in contrast to hypoxia, is the virtual absence
rather than just depletion of dissolved oxygen in the water column or underlying sediments. The
subject of marine benthic hypoxia has been extensively reviewed (Diaz and Rosenberg 1995).
Of great importance is the observation that hypoxia occurs not only at the bottom of the
water column near the sediments, but well up into the water column. Hypoxia normally occupies
20 to 50% of the total water column, but under some conditions may encompass more than 80%
(CAST 1999). In the northern Gulf of Mexico, hypoxic waters are found between five and 30 m,
extending to 55 km from shore on the steeply sloped southern-eastern Louisiana coasts to
130 km offshore on the gradual slopes of the central and southwestern Louisiana shelf (Trefry et
al. 1994, Justic et al. 1995). Phytoplankton growth in the Gulf of Mexico was found to be very
rapid with growth rates controlled primarily by the supply of nitrogen via the Mississippi River
(Fahnenstiel and McCormick 1995, Lohrenz et at. 1997). In parts of the Black Sea, the hypoxic
water column may be 100 m or more in depth. Nearly 87% of the Black Sea water volume is
anoxic. A hydrogen sulfide layer lies 100 to 200 m below the surface, whose depth varies
seasonally. Hydrogen sulfide is even found in the shelf zones; its concentration varies between
1.5 and 2.25 ppm in the lower water column on the northwest shelf at depths of 10-30 m, and is a
direct result of euthrophi cation.
Organic and inorganic contaminants, as well as nutrients, are readily transported into
hypoxic or potentially hypoxic areas. Metals and metalloids are present in different forms as
shown in Figure 1. They are present as: aqua-cations or anions, bound exchangeable ions on the
surface of minerals, constituents of the mineral structure, complexes with a soluble or insoluble
ligand, and precipitated/co-precipitated crystalline or amorphous solids. Organic contaminants -
polar and nonpolar solutes - can also be present: dissolved in water; complexed to a metal ion;
sorbed to mineral, organic or organo-mineral surface; or present as an incursion complex within
the matrix of a macromolecule. A drastic decrease in the dissolved oxygen concentration can
220
-------
exert drastic changes in various geochemical processes - speciation, bioaccumulation,
bioavailability, mobility, transformation kinetics/pathways, (organic contaminants) and toxicity.
Similarly, for biological processes - [the transformation of organic matter in the water column
(pathways, kinetics, and form - particulate and dissolved)] - the oxygen content determines the
microbial ecology and, therefore, whether microbial respiration occurs by aerobic or anaerobic
pathways. Dissolved oxygen concentration along with pH determines the redox potential in the
water column, at particulate surfaces, and in the under-lying benthic sediment. This, in turn,
determines the oxidation-reduction state, the solubility and availability of certain metals and
metalloids, and their toxicity.
Metal-Pore Water and
Water Column Equilibria
z-(n-y)
[M(OH) (tyt) LJ]
T 4 . T , z-m .
-\-i: II Tij; \JI A
J r J M ^olid
If z-y 1\
[M(OH)yCH20)n.y] 4\
+i\ + \*
-H 1L+H +A __^ Ma
f ^^ff- MasoUd
[M(H,O) ]Z+1^^^A
Ji ^^,^+me"
-Lnl +L. ^^:=^^^rM('B,oi) fm
Jlr me n
Adsorption
Processes
K .
'
Solid Phases
•Minerals
-Silicates
-Oxides
-Carbonates
-Sulfides
•Hurnic Acid
•Microorganisms
-Bacteria
-Fungi
-Algae
Transport Pathways
Metal-Pore Water
B
Particle-Surface Water — ^Advection
\ /
(§\ /
\\ If
%\ /.v
Q\ /
Metal-Bottom Sediment
Mj Uptake
^ and Transport
Within the Fish
Reaction in
Fish Cell
Figure 1. Metal Behavior in the Aquatic Environment (After Bailey and Zhang 2000).
Sediments are important in determining redox and anoxic conditions. Under oxic
conditions, the water-sediment interface and a few mm below also exhibit oxic behavior and C>2
is the terminal electron acceptor (TEA). Beneath this upper sediment layer, the C>2 content
decreases rapidly resulting in a transition from a hypoxic to an anoxic state. In sediments after
221
-------
the 62 is depleted (due to biological oxidation of organic matter), first hypoxic then anoxic
conditions prevail. A series of rather stable vertical gradients are then formed (as in the water
column), and various TEAs are consumed in order of their decreasing redox potentials (Table 1).
The vertical sediment gradients (comparable to vertical depth in the water column) are a function
of the same variables as in the water column: organic matter input, microbial metabolic
capabilities, mineral types and content, and the geochemistry of the environment - marine vs.
estuarine vs. freshwater. The major difference between stratified water columns and stratified
sediments is the greater abundance of minerals in sediments. Minerals can act both as reactants
with and/or products of microbial metabolism and, therefore, can impact the microbial ecology
and metabolism of both environments, both structurally and functionally. Prokaryotes, e.g.,
bacteria, are the primary inhabitants of these stratified environments.
Table 1. Redox Chemistry: Sources, Microbiology and Energetics.
I O;. •- JfuMv
A primary difference between freshwater and marine sediments relates to the amount of
SO42" in the latter and the over-riding dominance of the sulfur cycle. However, in freshwater
sediments, CH4 formation is the terminal redox reaction, dominating carbon metabolism in the
anoxic freshwater zone. Therefore, an understanding of oxidation and reduction reactions is
imperative in order to predict inorganic and organic chemical speciation, transport,
transformation, and bioavailability within the water column and within the benthic sediment, and
222
-------
the transfer between various environmental compartments. Redox chemistry has been studied in
many areas including agriculture, environmental science and pollution control, geochemistry,
limnology, oceanography, and soil science. Much of the early work was done in soil science
(mineral weathering, soil genesis and morphology) and agriculture in connection with
mineralization (oxidation) of organic matter, ammonification, nitrification, dentrification,
nitrogen fixation, SC>42" reduction to S 2", and nutrient availability. Excellent reviews summarize
this work (Ponnamperuma 1972, Bartlett and James 1993). The importance of redox chemistry
in the subsurface/aquifer environment has been presented recently by Lovley (1997).
The purpose of this overview paper is to: (a) review biogeochemical processes affecting
dissolved oxygen level, in fresh, estuarine and marine waters, and to define the predominant
TEAs and resultant redox potentials; (b) understand the role of surface acidity and surface redox-
derived values on the redox potential at the solid-liquid interface; (c) examine the speciation of
the common constituents of sea water - Ca, Cl, COs, H, OH, Na, Mg, and 804; (d) assess the
effect of chemical composition of sea water on metal speciation (Cr, Fe, Mn, Cu, Ni, Zn, Cd),
metalloids (As, Se and Te), and actinides (Pu and U) — using a combination of literature values
and the application of the geochemical equilibrium model, MINTEQA4; and (e) examine the
redox chemistry of Fe, Mn, Cr, S, I, Hg, As, Se, Te, Pu, and U. Possible adverse effects of those
redox-sensitive, chemically active forms on aquatic biota present in the hypoxic zone are
presented and discussed. To facilitate this review, the basic tenets of redox chemistry are
presented and reviewed. ACCEDES software (LC % hour values) (ECOTOX 2002) will be used
where possible to help establish the concentrations of the metals and metalloids used in
speciation assessments.
BACKGROUND
Oxidation and Reduction (Redox)
Oxidation and reduction (redox) is the terminology used to describe the chemical process
that changes the electrical charge of a chemical element or compound by the gain or loss of an
electron. The change in charge occurs when electrons are transferred from one species to
another and is nothing more than a chemical reaction, i.e., there is an electron donor and there is
an electron acceptor. Redox processes, however, traditionally are represented by dividing the
reaction into two parts: one part comprises the species that give up electrons, denoted the
reductants; and the other part consists of the species that accept the electrons, or the oxidants.
The redox potential of a chemical solution has been characterized as an important
variable in chemistry since the 1920s (Conant 1926, Zobell 1946, Hewitt 1950, Bates 1959,
Clark 1960, Guenther 1975). Many studies (Ponnamperuma et al. 1967, Ponnamperuma 1972,
Liu 1985, Bartlett 1986, Lindsay 1988) have reported the critical role of redox potential in
natural environments such as soils or water/sediment systems. Conventional redox
measurements have been made in the presence of solid phases, but the role of electron
donors/acceptors on surfaces has not been delineated. Platinum electrodes have been used to
223
-------
measure the redox potential of porous media, but have certain well-known limitations
(Ponnamperuma 1972, Liu 1985). Therefore, a reproducible in situ method is needed to
characterize (Stumm and Morgan 1985) and estimate the redox potential at the interfacial region
of minerals, of microorganisms, of organic matter, and of composite organo- mineral surfaces.
The surface acidity (Bailey and Karickhoff 1973) and redox potential of iron-bearing minerals
are critical parameters in understanding redox chemistry of soils, sediment and aquifer systems
(Ponnamperuma 1972, Liu 1985, Stumm and Morgan 1985). Patrick and co workers (Turner
and Patrick 1968, Connell and Patrick 1968, Masscheleyn etal. 1990, 199lab, 1992, Patrick and
Jugsujinda 1992, and Patrick and Verio 1998) have extensively studied the factors influencing
redox potential of flooded soils and wetlands. Since details of the generalized surface acidity
function and surface redox potential are of major importance in this overview, equations for
these parameters are reviewed and described in the following sections.
Generalized Oxidation-Reduction Reactions
The fundamental oxidation-reduction reaction at equilibrium can be expressed as:
Oxidant + ne" = Reductant (1)
The quantitative relationship between Ehaq and an electron donor and acceptor in solution is
given by the classical Nernst equation:
that at 25°C becomes
f'k :: f
m_ (3)
where Ehaq = solution phase redox potential, V
E° = standard potential, V
T= temperature, °K
R = molar gas constant
F = Faraday's constant
n = stoichiometric coefficient
E° is the standard potential representing the capability of a redox couple to donate or
accept electrons under standard conditions (pH = 7.0 at 25°C and unit activity of reactants). For
a given couple, the ratio of an electron acceptor to an electron donor is the ratio of its reduced
form concentration divided by its oxidized form concentration. Therefore, if the
reduced/oxidized activity/concentration ratio is known, (equilibrium constant) and if the standard
potential and pH are known, it is possible to calculate Ehaq a priori.
224
-------
Generalized Surface Redox Potential
The surface redox potential, Ehss, of a porous medium reflects the steady state condition.
Generally, porous media approach equilibrium very slowly, if at all, i.e., disequilibrium exists
among various redox systems, e.g., O, S, N, Fe, Mn, and C species. Speciation of a metal ion in
the pore water governs its mobility and bioavailability. Metal ion speciation, in turn, is governed
by pH (hydrogen ion activity, /.e.,-log{H+}) and electron activity (quantity of electrons in the
system where speciation in this sense reflects the ion oxidation number). Recall that aqueous
solutions do not contain free protons or electrons. The pH measures the relative tendency of a
solution to accept or transfer protons; similarly pe measures the tendency to accept or transfer
electrons. Just as the activity of protons is very low at high pH, the activity of electrons is very
low at high pe. Thus a high pe indicates a relatively high tendency for oxidation while a low pe
indicates a high tendency for reduction. The general reaction can be written as:
(Reductant = electron donor) • "(Oxidant = electron acceptor) + ne" + mH (4)
where m is the stoichiometric coefficient of the hydrogen ion. The redox potential at the solid
surface, EhSS; is related to pH by the following relationship at 25°C (Ponnamperuma 1972; Liu
1985):
i'X'^ft /.«j;i4
-------
Where Ind and Indc are the conjugate indicator species concentrations differing by one proton, Ka
is the acidity constant describing the equilibrium between the two indicator species under ideal
aqueous solution behavior, and ss represents the surface phase.
Proton activity at or near a surface is not represented by bulk pH measurements.
Similarly, it may be expected that the electron activity at or near a surface, such as an iron-
bearing mineral surface, may not be represented by bulk phase redox potential measurements.
Because redox reactions involve the transfer of electrons, they also produce an electromotive
force (emf) that can be measured with a suitable electrode and volt meter. The emf is related to
the total Gibbs free energy (AG) of the reactants and products by the expression
AG= -nFE (8)
Where n is the number of moles of electrons involved in the reaction, F is the Faraday constant
(equal to 96,500 coulombs per electron), and E is the electrode potential. The electrode potential
of the standard state, E°, by analogy is given by
AG° = -nFE° (9)
Given the electrode potential, it is possible to calculate the free energy change for a given redox
reaction in terms of kJ mol"1 of product. The utilization of chemical energy by microorganisms
involves redox reactions. In biochemical systems, oxidation and reduction reactions frequently
involve the transfer of not just electrons, but hydrogen atoms as well. Redox reactions can
involve electrons only or hydrogen atoms only. When dealing with biological systems, redox
potentials are given relative to neutrality because the cytoplasm of a cell is neutral, or nearly so.
Redox Reactions in the Water Column and Porous Media
The redox potentials of both the water column and porous media are related to organic
matter content, mineral type, particle size and content, water regime, microbial activity, the
diffusion rate of oxygen, the presence of multiple redox couples, and pH. Recall that the
magnitude of change of Eh with change in pH is about 59 mv per one pH unit. We need to know
both the intensity and the capacity factor in the redox reaction if we are to predict the redox
potential of porous media. Eh expresses the degree of oxidation and reduction and, therefore, is
the intensity factor. The capacity factor is the relative proportion of oxidizable or reducible
material. The greater the amount of oxidizable organic matter, the lower the redox potential, i.e.,
a smaller positive number or a larger negative number. A platinum electrode is used to measure
the redox potential whether in the aqueous phase, in suspension, or in porous media, e.g..,
sediment. The electric field near charged surfaces may interfere with the readings of the
platinum electrode. The redox potential in the aqueous phase may be different from that where
charged surfaces are present. The same phenomenon applies for pH measurements. This is
called the suspension effect.
226
-------
Two classifications are used to define the redox state in aqueous and in porous media.
One defines the redox state according to Eh values measured by a platinum electrode (with its
above stated limitations). The following is an arbitrary classification of the relative redox states
according to measured Eh: strongly oxidizing (800 to 400 mv); moderately oxidizing (400 to 200
mv); moderately reducing (200 to -50 mv); reducing (-50- to-200 mv); and strongly reducing (-
200 to -400 mv). The Eh of water and porous media is bounded on the upper side by the
oxidation of water (O2 formation) and on the lower side by the reduction of water (H2
formation).
The second classification system of redox state is based upon the redox activity of the
microflora in the system, and is the one generally given and used in microbiology (Madigan et
al. 1997) and biochemistry (Stryer 1995, Garrett and Grisham 1999) text books. The redox state
is defined by the TEAs used by the microorganisms in the system. The poising of a porous
media refers to the ability of the porous media to resist change in Eh upon the addition of a small
amount of oxidant or reductant. To complete the picture, organisms can be classified in terms of
their energy sources. Chemotrophs use chemical compounds as the primary energy source.
They can be further subdivided into chemolithotrophs (use inorganic chemicals) and
chemoorganotrophs (use organic chemicals). Phototrophs use light as a primary energy source.
Phototrophs can be further divided into photolithrotrophs and photoorganotrophs.
Effect of Microbial Ecology and Dissolved Oxygen Concentrations on Transformation,
Speciation and Electron Transfer Processes
Microorganisms utilize a variety of processes for the biotransformation of metals
including bioreduction, biomethylation, biomineralization, and biosorption. In certain
environments, microorganisms are found in biofilm communities (bound to/residing upon a
surface) in which a suite of enzymatically driven reactions versus abiotic electron-transfer
reactions versus surface-complexation reactions can occur (Costerton and Phillip 2001).
In addition, mineral surfaces that act as substrates for the biofilm communities may affect
the microbial response to metal stress by potentially altering the bioavailability of the toxic
metals and/or the success of the detoxification strategy employed by the bacteria. Many
microorganisms favor being attached/bound to a surface through adhesion and ultimately
forming a biofilm given favorable spatial considerations.
Microbially mediated oxidation-reduction processes provide energy and reducing
equivalents necessary for cell growth and replication (Siciliano and Lean 2002). Various
elements act either as an electron donor or an electron acceptor undergoing a redox reaction that
alters the elements speciation, valency, and hence physicochemical characteristics. The redox
reaction is closely related to microbial metabolism and provides the organism with either an
essential element, energy or a detoxification process. In turn, microbial transformation of N, P,
S, actinides, metals, metalloids, and halogens regulates their bioavailability, toxicity, and
environmental impact in both the aquatic and terrestrial ecosystem.
227
-------
Electron Transfer Process and Mechanisms
Electron transfer is one of the primary reactions of transition metal complexes (Atwood
1991, Langmuir 1997). The reaction rate is a function of the concentration of the oxidant and the
reductant. Electron transfer mechanisms are of two basic types: (1) an inner-sphere mechanism
in which a ligand bridges the two metal centers, e.g., oxygen; and (2) an outer sphere mechanism
between metal centers with intact coordination spheres. Regardless of mechanism, a primary
factor in determining the rate is the changes in the bond lengths and angles necessary before
electron transfer can occur. Outer-sphere reactions can be described by theoretical methods.
Conjugation in a bridging ligand is electron transfer, although electron transfer can occur over
very long distances even in the absence of conjugation. The inner-sphere mechanism is
characterized by the formation of a binuclear transition state/intermediate during the electron
transfer. A very important facet of the inner-sphere process is the bridging ligand that forms part
of the coordination spheres of both the oxidant and the reductant. The bridging ligand must
function as a Lewis base toward both metal centers, and it must have two pairs of electrons that
can be donated to different centers. The electron transfer can be considered as several individual
reactions. The first reaction is the diffusion-controlled formation of a collision complex. The
second reaction is the formation of a complex, termed the precursor complex, in which the ligand
bridges the two metal centers, but in which the electron has not yet been transferred. The third
reaction is the electron transfer itself to the successor complex. The last reaction is the
dissociation from the successor complex. Any of these reactions can be rate determining.
In a microbial cell, the transfer of electrons in a redox reaction from the electron donor to
the electron acceptor involves one or more intermediates termed carriers (Madigan et al. 1997).
When such carriers are used, the initial donor is called the primary electron donor (FED) and the
final acceptor is called the terminal electron acceptor (TEA). The free energy change of the
complete reaction sequence is determined by the difference in reduction potentials between the
FED and the TEA. The transfer of electrons through the intermediates involves a series of redox
reactions, and the resulting free energy changes from these individual steps must sum to the
overall value obtained by considering only the starting and ending compounds. The intermediate
electron carriers can be divided into two classes: (1) those firmly attached to enzymes in the
cytoplasmic membranes and, (2) those freely diffusible. The fixed carriers function in
membrane-associated electron transport reactions. Freely diffusible carriers include the
coenzymes - nicotinamide-adenine-dinucleotide (NAD+) and NAD-phosphate (NADP+). NAD+
and NADP+ are hydrogen carriers and always transfer two hydrogen atoms to the next carrier in
the chain. Such hydrogen atom transfer is termed dehydrogenation.
Electron transport systems are composed of membrane-associated electron carriers.
These systems have two primary functions: (1) to accept electrons from an electron donor and
transfer them to an electron acceptor; and (2) conserve some of the energy released during the
electron transfer for synthesis of ATP. Several types of redox enzymes are involved in electron
transport.
228
-------
The electron transport process involves different molecular species (Garrett and Grisham
1999) including: (1) flavoproteins contain tightly bound flavin mononucleotide (FMN) or flavin
adenine dinucleotide (FAD) as prosthetic groups and may participate in one-or two- electron
transfer events; (2) ubiquinone (UQ) - coenzyme Q - that may participate in one-or two-electron
transfer events; (3) cytochromes are proteins containing heme prosthetic groups whose ligands
have varying coordination around the heme group that function by carrying or transferring
electrons and include cytochromes b, c,cj, a, and a3 that are one-electron transfer agents where
the heme iron is converted from Fe2+ to Fe3+ and back; (4) iron-sulfur proteins that participate in
one-electron transfer involving the Fe2+and Fe3+ states; and (5) protein-bound copper, a one-
electron transfer site that converts Cu+ and Cu2+. Nicotinamide adenine dinucleotide (reduced
form), NADH, serves as a link between the source of electrons - glycolysis, the tricarboxylic
acid cycle (TCA), and fatty acid oxidation - and the electron transport chain (UQ and the
cytochrome).
A general overview of the electron transport pathway is given below:
Glycolysis ^> NADH ^> NADH-UQ reductase ^> UQ ^> cytochrome reductase ^>
cytochrome c ^> cytochrome oxidase =^> C>2 + 4e" + 4H+ ^> IFkO (or an alternate TEA) (10)
The change in the redox status of the iron is as follows:
Cytochrome-Fe(II) => Cytochrome-Fe(III) + e (11)
The e is "captured" by the appropriate TEA present in the system
In order to understand this, we need to first understand the central mechanism of
microbial metabolism that allows microbial energy conservation to be viewed as a unified
feature. The chemiosmotic theory (Gottschalk and Blaut 1990) is shown in Figure 2. It proposes
that chemical energy of a variety of forms is transformed into an electrochemical potential across
a membrane termed a proton motive force (pmf). Essentially, this is how living organisms
harvest chemical (redox) energy from the environment and conserve it as biologically useful
energy in the form of adenosine triphosphate (ATP). The pmf is used to drive synthesis of ATP
via membrane-bound enzymes called ATPases, which utilize the energy in the electrochemical
gradient to drive the synthesis of high energy phosphate bonds that cells use to fuel cell
functions. Figure 2 also shows the location of the binding sites in the system and the reactions:
(1) NADH is on the matrix side of the membrane and pairs of electrons are transferred from
NADH to UQ; (2) cytochrome c is located in the intermembrane space side; and (3) cytochrome
oxidase is in the intermembrane space side.
229
-------
II
III
r ^ r ^ r
H* 2H+ 2H+ 2H+ 2-4 H
Outside (H-carrier) /^ (e'-carrier) (H-carrier),
nimn
Membrane
UJUJUJ
Inside
OH" / \
NADH+H NAD
2H
2H
red
ADP
+
P
ATP
Figure 2. Chemiosmotic Mechanism of Energy Conservation and Electron Transport i.e., how
organisms harvest chemical energy (redox) from the environment and conserve it for
useful energy. The basic features as shown in the figure numerically are: (I) A
semipermeable membrane that is impermeable to charged molecules and can be used to
separate the charges. Once charge separation is achieved, energy can be harvested. (II)
A vectorial electron transport chain, in which H-carriers and e-carriers alternate in the
flow of reducing power from substrate to oxidant. As electrons flow towards the oxidant,
protons are pumped to the outside of the membrane creating the electrochemical gradient
(a protonmotive force consisting of a combination of pH and charge gradient). (Ill) An
enzyme to convert the proton motive force into useful cellular energy. The enzyme
shown is the membrane-bound ATPase that allows the protons to flow back into the cell
through pores in the membrane it creates. (After Nealson and Stahl 1994).
Aerobic and Anaerobic Respiration Processes
The seven processes determining the redox status of water and porous media are (1)
aerobic respiration, (2) denitrification, (3) manganese reduction, (4) iron reduction, (5) sulfate
reduction, (6) methanogenesis, and (7) fermentation. The reactions were shown in Table 1. We
will examine this classification method in this section.
Redox reactions occurring in porous media are driven by energy derived from microbial
oxidation of organic matter. Microbial oxidation of various organic carbon compounds i.e.,
mineralization of C, N, and S in porous media, utilizes the oxidized forms of several inorganic
substances as TEAs. The most important TEA is oxygen because of its natural abundance in the
atmosphere, its ease of diffusion into porous media (diffusion constant of C>2 is 10,000 times
greater in the vapor phase then in a liquid phase), and its ease of reduction that limits the
reactivity of other TEAs as long as oxygen is present. In the absence of oxygen, certain
230
-------
microorganisms can switch over to alternate TEAs for the oxidation of organic matter and
introduced anthropogenic organic compounds. The initial step many times is the fermentation of
organic matter to yield acetate and hydrogen (see Figure 3) that are, in turn, oxidized by other
TEAs e.g., manganese-or iron-reducing microorganisms.
Figure 3. Model of Oxidation of Organic Matter to Carbon Dioxide with the Reduction of
Fe(III) to Fe(II) in Porous Media. (After Lovley 1997).
Aerobic respiration is a redox process in which oxygen serves as an electron acceptor and
as the result, large amounts of energy can be released (refer back to Table 1). Oxidation of
organic matter by microorganisms requires the use of TEAs. Aerobic respiration occurs in both
the water column and in the first few millimeters of the underlying sediment in the presence of
organic matter. This same sequence of TEAs can occur in sediment involving aerobic processes
at the sediment-water interface (if there is any oxygen remaining in the overlying water column),
progressing downward from the interface into the sediment column, i.e., aerobic respiration
followed sequentially by denitrification, manganese-(mineral phase)-reduction/dissolution, iron-
reduction/dissolution, sulfate reduction, methanogenesis and fermentation. Organic matter
present is respired to CO2 and H2O until consumption exceeds the amount of O2 that can be
delivered to the site by diffusion. This same sequence of TEAs can occur in aquifers. Reduction
of O2 is carried-out by true aerobes (refer back to Table 1), denitrification, i.e., reduction of NOs"
to N2(g) by facultative anaerobes, and reduction of Mn(IV) to Mn(II) and Fe(III) to Fe(II) by
facultative anaerobes. Reduction of SO42" to S2" and CO2 to CH4 are carried-out by true
anaerobes that can not function in the presence of O2. Fermentation can also result in CO2 as an
end product.
231
-------
MODELING METAL SPECIATION
To realize the goal of predicting metal ecotoxicity, we must first predict metal speciation
in solution and secondly at the site of action. Aquatic and biochemical systems comprise a large
number of components and interactions. Therefore, equilibrium metal speciation computations
for these systems can involve large arrays of equations that incorporate the principles of mass
action, charge balance, and mass balance to model such reactions as hydrolysis, complexation,
dissolution/precipitation, oxidation/reduction, ion exchange, and partition/adsorption. A number
of computational codes have been formulated to solve these arrays of equations using a
thermodynamic data base and an "equilibrium constant approach". These codes use the mass
action principle that relates the activities of free metal, free ligand and the metal-ligand complex.
The mass action relationships are linked to mass balance equations, resulting in sets of linear
equations. Generally, the assumptions are made that (1) solutions are in equilibrium with the
thermodynamic-predicted solid phases; and (2) only the most stable phase or phases can occur.
The equilibrium speciation of each component metal is obtained by simultaneously
solving the opposite set of linear equations. In the solution of chemical equilibrium models, it is
assumed that equilibrium exists for all reactions in the systems of interest. This assumption is
appropriate, because a majority of the interactions/reactions are known to occur rapidly.
Because of the general rapidity of these reactions, equilibrium predictions for aqueous and
adsorbed phases are generally reliable, except in the case where reactions are controlled by
diffusion; these latter reactions really are transport-controlled rather than chemically-controlled.
Caution also must be exercised when predictions involving solid-phase precipitation/dissolution
and redox transformations are involved because these depend on a generally moderate database
of chemical kinetics.
Several chemical equilibrium models have been developed including WATEQ (Truesdell
and Jones 1974), MINEQL (Westall etal. 1976), GEOCHEM (Mattigod and Sposito 1979),
GEOCHEM-PC (Parker, et. al., 1995), MINTEQ (Felmy etal. 1984), MINTEQA2 (Allison, et
al. 1991), GMIN (Felmy 1995), SOILCHEM (Sposito and Coves 1995), CHESS (Santore and
Driscoll 1995), ALCHEM (Schecher and Driscoll 1995) and C-SALT (Smith etal. 1995). For
more in-depth discussion of the subject of chemical equilibrium models, the reader is referred to
a publication by Loeppert et al. 1995.
What metal cations and/or anions are important in fish behavioral toxicology? Henry and
Atchison (1991) summarized the literature up through 1990 and found that the following
metals/metalloids are important: As, Cd, Cr, Cu, Se, and Zn. MINTEQA2 was used to assess the
effect of sea water composition on speciation of these four metals and two metalloids - As and
Se. Mn, Fe, Co, Ni, Pb, and U were also evaluated.
232
-------
Table 2. Speciation of Sea Water at pH = 8.2 and CO2 Concentration= 3.5 x 10"4and 2.0 x 10"3*
Atms.
Model/Component
Na
Ca
Mg
H
Cl
C03
S04
H2O
Specie
Na+
NaSO4 "
NaHC03(aq)
Ca2+
CaSO4(aq)
CaHCO3+
CaCO3(aq)
Mg2+
MgS04(aq)
MgHCO3+
MgCO3
HC03- (aq)
NaHCO3
MgHCO3+
H2CO 3 (aq)
CaHCO3+
Cl-
HCO3-
NaHCO 3(aq)
MgHCO3+
H2C03
CaHC03+
MgC03(aq)
NaCO3-
SO42-
Na SO4-
Mg S04 (aq)
Ca SO4 (aq)
MgOH+
OH-
CaOH+
Percent
98 (95)*
2 (2)
- (3)
81 (68)
15 (14)
3 (12)
1 (6)
82 (67)
13 (11)
5 (19)
- (3)
75 (76)
12 (11)
11 (9)
1 (1)
1 (1)
100 (100)
75 (72)
12(11)
11 (9)
1 (0)
1 (1)
- (2)
- (5)
40 (43)
30 (32)
24 (21)
6 (5)
64 (59)
33 (39)
3 (2)
233
-------
Table 3. Metal/Metalloid Speciation in Seawater at pH = 8.2 and CC>2 Concentration = 3.5 x 10"
1 ^ /"\ t f\~J A j_
and 2.0 x 10"J Atms.
Model/Component
As
Cd
Co
Cr
Cu
Fe(III)
Fe (II)
Mn
Ni
Pb
Se
Zn
Specie
H3AsO3
H2AsO3-
Cd2-+
CdCf
Cd(S04)- (aq)
Cd(C03) (aq)
Cd(S04)2 2" (aq)
Cd(CO3)2 2~
Co2+
cocr
Co(S04)(aq)
Co(C03)(aq)
CoH(CO3)
Cr042-
CuC03 (aq)
Cu(C03)22-
Fe(OH)2+
Fe(OH)3(aq)
Fe(OH) 4-
Fe2+
FeS04(aq)
FeHCO3 +
Mn2+
MnCf
MnCl2 (aq)
MnSO4(aq)
MnHCO3+
Ni2+
NiCf
NiSO4(aq)
NiC03(aq)
NiHCO3+
Pb2+
Pb(OH)+
Pb(CO3)2 2~
PbC03(aq)
PbHCO3+
Se042"
Zn2+
ZnSO4(aq)
Zn(SO4) 2~
ZnCO3
ZnHCO3+
Percent
94
6
59 (20)*
5 (2)
10 (4)
21 (39)
2 (0)
3 (35)
44
26
8
11
11
100
63
37
73
9
18
82
15
3
66
15
5
9
5
39 (14)*
17 (6)
7 (3)
22 (45)
16 (12)
2
2
35
57
4
100
45
8
2
40
5
234
-------
Model Input Parameters
The input parameters for an independent simulation run includes metal/metalloid type
and total concentration: As (0.11 umol/L), Cd (0.47 umol/L), Co (0.47 umol/L), Cr (0.01
umol/L), Cu,(0.63 umol/L), Fe(II, III) (0.47 umol/L), Mn (0.47 umol/L), Ni (0.66 umol/L), Pb
(0.31 umol/L), Se,(0.47 umol/L), and Zn (0.47 umol/L); plus pH 8.2; partial pressures of CO2 of
3.5 x 10"4 and 2 xlO"3 atm; ionic strength, molkg"1, and temperature, 10°C. Precipitation of solids
was allowed and activities were calculated by the Davies equation. The chemical composition of
seawater used in mM was: HCCV (2.38), SO4 2" (28.2), Cl' (545.0), Ca2+ (10.2), Mg2+(53.2), Na+
(468.0), and K+ (10.2) (Stumm and Morgan 1996). The effect of variable partial pressure of CO2
(3.5 xlO"4and 2 xlO"3 atm) on metal carbonate mineral formation was also assessed.
The MINTEQA2 speciation simulation results were shown in Tables 2 and 3. The
speciation of each element is summarized below. Redox values were also calculated for each
simulation.
Ion Speciation in Seawater
The speciation of ions in seawater at various CO2 levels was presented in Table 2. At the
normal CO2 level, the major speciation is the aqua species of Mg2+ (82%), Ca2+ (81%), Na+
(98%) and Cl" (100%); HCCV is the predominant species for the COs2" anion, although calcite,
aragonite, magnesite and/or dolomite (both ordered and disordered form), and huntite could all
precipitate. Sulfate is present as an aqua-anion and also could precipitate as gypsum. The
elevated level of CO2 favors an increase in carbonate-forming minerals and a lessening of
sulfate-forming minerals.
Metal and Metalloid Speciation in Seawater
Two of the anionic constituents - Se and Cr - are present at their highest valence state
(VI) while As is present at a valence state of (III) (refer back to Table 3). Cd occurs in six
different species; the major ones being the divalent aqua-Cd ion, i.e., [Cd(H2O)e]2+, and the
cadmium carbonate mineral, otavite. Elevated CO2 level again favors prevalence of the mineral
otavite and the aqueous Cd carbonate with a higher COs2" ratio. The presence of the aqua-
divalent cation specie is reduced by two-thirds with an elevated CO2 level. Copper occurs as a
carbonate species or precipitates in the absence of any organic complexing agent. Under highly
reducing conditions, the aqua - Cu1+ species may be present. Fe(III) occurs as a hydroxylated
species tending toward the formation of an Fe(III) oxide/oxyhydroxide. One would not expect
aqua-Fe(II), i.e., [Fe(H2O)e]2+, to exist in solution above pH 4.0. Except at a low redox
potential, Fe(II) would be oxidized to various Fe(III) species, i.e., like oxides and hydroxides.
Mn(II) occurs principally as an aqua-divalent cation species while Pb forms lead carbonate
(cerusite) and lead oxyhydroxy carbonate (white lead). Nickel occurs both as a aqua-divalent
cation and as nickel carbonate (hellyerite). Zinc is about equally distributed between the aqua-
divalent cation and zinc carbonate (Smithsonite). The elevated CO2 level again favors the
increasing prevalence of the metal carbonate precipitant. At low concentrations of heavy metals,
an alternative possibility is that a solid solution series of metals occur with calcite, rather than the
production of pure phases.
235
-------
Since there are many organic complexing agents in the water column their interactions
with the metals must also be considered in determining speciation. Some speciation ranking
results are shown in Table 4. It can be seen that Fe(III), Cu(II) and Pb(II) all have a high
proclivity to form organic complexes while Zn (II) has a lesser tendency to react and form an
organic complex.
Table 4. Metal Speciation Rankings.
Metal/Metalloid
Cd2+
Co2+
Cr3+, Cr6+
Cu2+
Fe3+
Mn2+
Nf+
Ptf+
Zn2+
As^+
Se5+
Speciation
Cd 2+ > CdCO3 (aq) > CdSO4 (aq)> CdCl+
Co 2+>CoCl+ >CoC03(aq) > CoHC03+>
Cr(OH)2+ » Cr(OH)3 (aq) » Cr(OH)2+>
CoSO4(aq)
CrO^ CrO4 2-
CuCO 3 (aq) » Cu(CO3)22- >: organic
organic » Fe(OH)2 1> Fe(OH4)- > Fe(OH)3(aq)
Mn2+ » MnCl+ > MnSO4 aq > MnCl 2 (aq)
Ni2+ > NiCO3(aq) > NiHCO3+> NiCl+ > NiSO4(aq)
organic » PbCO3(aq) > Pb(CO3)22- » PbHCO3+ > Pb2+
Zn 2+ > ZnCO3 (aq)> ZnSO4(aq) > ZnHCO
3+, organic
FfeAsOs » FfcAsOs-
SeO42-
(After Stumm and Morgan 1996).
METAL AND METALLOID REDOX CHEMISTRY
Redox and pH (and the presence, concentration and chemical composition of organic
ligands) determine the distribution of metals between the various chemical fractions - water
soluble, exchangeable, (e.g., with other cations on clay minerals with a net negative charge),
bound to iron and manganese oxides, bound to carbonates, complexed with soluble or to
insoluble organic ligands, bound to sulfides or co-precipitated with other metals to form a
crystalline or amorphous salt. We have seen previously that redox reactions control the
speciation of NO3", N, Mn, Fe, and S, and affects the speciation of As, Se, and Cr.
236
-------
In this section we will evaluate the effects of lowering the redox potential, i.e., becoming a more
negative value, on the valence state, solubility and prevalence of various elements (Mortimer and
Rae 2000).
Work of Patrick and coworkers (Connell and Patrick 1968, Turner and Patrick 1968,
Masscheleyn et al. 1990, 1991ab, 1992, Patrick and Jugsujinda 1992, Jugsujinda and Patrick
1995, Patrick and Verio 1998), dealing mainly with wetland environments, has shown the
following: Under reducing conditions, about two-thirds of the soluble Fe was complexed to
soluble organic matter. Soluble Mn, to the contrary was nearly all in the ionic form. About 90%
of the soluble Zn was complexed while only 9% of the Zn was adsorbed. The complexed soluble
Fe and Zn were bound to organic matter having a molecular weight greater than 25,000 D.
There were marked differences in the size distribution of the various organo-metal complexes
under different redox and pH conditions. The soluble Fe was associated with both the largest
and the smallest molecular size ranges of soluble organic matter. The effect of pH was more
evident in the smallest size range, with much more complexed Fe present at low pH. On the
other hand, Mn was associated with the smallest size range under all pH and redox conditions
reflecting its ionic nature. Hg and Pb were only associated with the largest size soluble
complexes and was little affected by pH and redox.
Fe(III)-stabilizing ligands such as bi-and multi-dentate carboxylates and phenolates
generally accelerate Fe redox reactions, while Fe(II)-stabilizing ligands such as phenanthroline
essentially stop Fe redox reactions because it is the Fe(II) species that, for example, reduces
Cr(VI) to Cr(III) and in turn becomes oxidized to Fe(III). Low redox values result in the
dissolution of Fe and Mn oxides and their enhanced concentration in the aqueous phase, the
oo n
reduction of 864 " to S Vpolysulfides/ S and the formation of metal sulfides of Cu, Zn, Cd, Hg,
Ni, and Pb even if the metals are in very low nanomolar concentrations. These metal sulfides are
very insoluble; pyrite (FeS2) being the most soluble and cinnabar (HgS) being the least soluble.
Precipitation of metal sulfides lowers dissolved metal ion concentration in solution and thus their
0 0
toxicity to the biota. Microbiological reduction of 864 " to S " promotes sulfide precipitation that
lowers the concentration of metals like Cd, Pb, Cu, Fe, Hg, and Zn in solution and therefore the
concentration that is bioavailable.
In an anoxic environment, a metal sulfide mineral like pyrite in the presence of an
another metal ion in solution with a higher stability constant with sulfide is dissolved
(competitive displacement) and the second metal sulfide is formed and precipitated (Lin et al.
1990, Simpson et al. 2000). The redox state determines the water solubility of metals like Fe,
Mn, and Co. A lower redox value changes the microbial ecology and species like Geobacter
prevail. These microbes use metal oxides of Fe, Mn, and Co as TEAs in exchange for energy for
growth and reproduction.
A gram-negative bacterium has been shown to reduce Cr(VI) to Cr(III) (either under
aerobic or anaerobic conditions) with the aqua-Cr(III) ion being adsorbed on the cell wall
surface, being present as an amorphous precipitate on the cell wall, or being present as a fine
grain precipitate, most probably Cr(OH)3 [mostly likely it is CrO(OH) - bracewellite or
237
-------
grimaldite] (McLean et al. 2000). Metal binding (of Cu, Fe, Au, and La) is greater in a biofilm
(except Cu) than that bound by planktonically grown cells of the same strain (Langley and
Beveridge 1999).
Supplying electrons toTEAs increases the pH of acidic media whereas removing electrons
lowers pH (Ponnamperuma, 1972). The primary driving force for increasing pH particularly in
bottom sediments is the dissolution of Fe(III) oxides (i.e., dissimilatory microbial reduction to
Fe(II) and its subsequent dissolution into the aqueous phase). The resulting increase in pH
lowers the activity of various transition metals through such processes as: (1) mineral
precipitation; (Lindsay 1979); (2) adsorption by deprotonated organic matter (Stevenson 1994);
and (3) adsorption onto an oxide surface (Karthikeyan et al. 1997).
In contaminated sediment, as the pH rises carbonates of Zn (Ma and Lindsay 1993), Cd
(Street etal. 1978) and Pb often control the activities of Zn(II), Cd(II) and Pb(II) (Brennan and
Lindsay 1996). In the case of Pb, this may be mitigated to some extent by the presence of humic
substances and the high affinity of Pb(II) for certain chemical functionalities within the humic
poly electrolyte structure (Lin et al. 1989, Jin et al. 1992). At low redox potential (see Table 1),
SO42" is reduced to S2" and where there are sufficient activities of both the metal cation and the
S2" anion, metal sulfides precipitate. When this occurs, carbonates and oxides dissolve to supply
the metal ion activities needed for precipitation. Until the controlling mineral supplying a metal
ion is depleted, there will be an equilibrium between the dissolving and precipitating mineral
containing that metal. We will examine first the effect of redox on the two principal metals in
aquatic systems - Fe and Mn; secondly on other metals - Cd, Cr, Cu, Hg, Pb, Zn; thirdly on
metalloids - As, Se, Te; and then on a halogen -1, the actinides - Pu and U - and finally C and P
in the next section.
Iron Redox Chemistry
Iron is the fourth most abundant element in the Earth's crust. Depending on
environmental conditions, Fe can form stable compounds in both the divalent and the trivalent
state. (Schwertmann and Fitzpatrick 1992). The form [(Fe(II) versus Fe(III)] in which iron is
found in nature depends on the pH and the C>2 concentration. The nature and properties of Fe
oxides have been summarized by Schwertmann and Taylor 1989; while those of Fe-bearing
smectites have been reviewed by Stucki 1988. Possible oxidation-reduction mechanisms in Fe-
bearing-phyllosilicates were evaluated (Stucki et al. 1996). One mechanism proposed to reduce
Fe(III) to Fe(II) in the lattice of smectites is the reaction of free radicals; in the case of dithionite
the proposed free radical is sulphoxylate (SCV0) (Gan etal. 1992). Fe(II) forms minerals such as
siderite, vivianite, or iron sulfide only in anoxic environments under weakly acidic to neutral
conditions. In the presence of C>2, Fe(II) is stable only under acidic conditions. At a neutral pH,
Fe(II) is quickly oxidized to Fe(III).
238
-------
Chemical Processes
Fe(III) forms minerals (16 different ferric iron oxides/oxyhydroxides are known) found
both in aerobic and anaerobic environments. In freshwater sediments, Fe(III) is the most
important TEA while SC>42" is the most important one in marine systems; both due to high
concentrations in their respective environments. Iron(III) plays a major role in controlling redox
in sediment, and may also be important in the overlying water column (particularly under
reducing and highly reducing conditions) through the ability to be a TEA, and form Fe sulfides
[pyrite ( FeS2), mackinawite (FeS),and greigite (Fes 84)]. The dissolution and precipitation of Fe
minerals are affected both by redox and pH. Fe is in solution at pH of four over a wide range of
redox values; precipitation starts with increasing pH in the alkaline range and with the depletion
of O2 (Patrick and Henderson 1980). Oxic conditions favor the presence of iron oxides, e.g..,
goethite.
Figure 3 indicated that there are both abiotic and biotic pathways for the Fe(III)/Fe(II)
couple. It also showed that complex formation of Fe(II) and Fe(III) on solids and with soluble
phases have a dramatic effect on the redox potential; therefore, electron transfer by the Fe(II)/
Fe(III) system can occur at pH =7 over the entire range of the stability of water (Eh -50 to
+1,100 mv). Table 1 and Table 4 revealed that the redox potential at pH = 7, E°H (pH =7),
decreases in the presence of most complex formers, particularly chelates with oxygen donor
atoms (strong Lewis base), because these ligands form stronger complexes with Fe(III) than with
Fe(II). Fe(II) complexes are stronger reductants than the aqua Fe2+ ion. This stabilization of the
Fe(III) oxidation state is also found with hydroxo-complexes and by binding with the O2" ligand
present in solid phases. Fe(II) minerals, therefore, are thermodynamically speaking strong
reductants. The couple Fe2SiC>4 (fayalite)/Fe3O4 (magnetite) has an EH similar to that for the
reduction of H2O to H2 (Bauer 1978). A surface complex of Fe2+ adsorbed inner-spherically
onto a oyxhydroxide surface is more reducing than the aqua-Fe2+ion (Stumm and Morgan
1996). The redox potentials for heme derivatives (see Table 5) clearly show the wide range of
possibilities involved in the bioinorganic system. Benzene anaerobic oxidation had been found
to occur in the Fe(III) reduction zone (Anderson et al. 1998) and in the SO42"-reduction zone
(Anderson and Lovley 2000).
Surfaces play an important role in abiotic transformation of the aqua- Fe(III) ion [and
Cu(II), V(V), and Cr(VI) aqua-ions]. Structural Fe(II) in magnetite and ilmenite
heterogeneously reduces these aqua ions over a pH range of 1 to 7 at 25°C (White and Peterson
1996). The half-cell potential for solid state oxidation [Fell] => ]Fe(III)] is -340 to -650 mv
making structural Fe(II) a stronger reducing agent than [Fe(H2O)e]2+ (-771 mv). Iron is released
into solution during redox reactions with magnetite and ilmenite and from the dissolution of the
oxide surface. Reduced Cr(III) and V(IV) species have been found via x-ray photoelectron
spectroscopy to reside on the oxide surface. Oxide redox potentials are determined by the
Fe(II)/Fe(III) composition of the oxide surface, and respond to aqueous ion potentials that
accelerate the oxidation process. Of note is that the ability of Fe(II) oxides to reduce transition
metals depends very strongly on the redox environment; these reactions are favored under anoxic
rather than aerobic conditions.
239
-------
Table 5. Standard Reduction Potentials for Biological and Inorganic Reduction
Half-Reactions.
. S.i - I! - * » I! »
h: ' - . > I .
vi • :n • :, ' MI, '
f vi/cis'/Ti,," - if;; s • :: f
I. '. :.x:<<:'vr>; - >?•_ ! * - ---
:ii
- VVrttl • II
- >. F
240
-------
Microbial Processes
The ability of microorganisms to reduce Fe(III) to Fe(II) has been known for nearly a
century. It was generally considered that most of the Fe(III) reduction in the environment was a
result of nonenzymatic processes (Allison and Scarseth 1942). The reason for this erroneous
conclusion was the work of Starkey and Halvorson (1927) who stipulated that microorganisms
brought about Fe(III) reduction by creating chemical conditions that favored a spontaneous
abiotic conversion of Fe(III) to Fe(II). Lovley (1991) pointed out that the 1927 study provided
no definitive data to support the hypothesis of abiotic Fe(III) reduction in microbial culture and
that the preponderance of evidence suggested that microorganisms reduce Fe(III) via enzymatic
processes.
There is growing evidence of biologically promoted dissolution and reduction of iron
oxides (Walker etal. 1989, Kostka andNealson 1995, Tratnyek 1995, Zachara etal. 1998) and
manganese oxides (Larsen et al. 1998). The metal reductase activity of a microorganism is
governed by its polyheme c-type cytochrome content (Drossman et al. 1988). This is clearly the
case for Desulfuromonias acetoxidam and Desulfovibrio; whether this will prove true for other
species is yet unknown. Figure 4 displays several possible mechanisms for the reduction of
Fe(III)s to Fe(II) s and/or the reduction and dissolution of Fe(III)s to Fe(II) aq. Note that humic
acid also may be a TEA where the functional group — quinone - accepts an electron from the
cytochrome reductase forming a semiquinone radical or hydroxyquinone. By completing the
"circuit", it may be possible that in the presence of humic acid Fe(III)s is reduced to Fe(II)s with
hydroxyqunione being oxidized to quinone. Hyperthermophilic microorganisms have been
found to reduce both Fe(III) and humic acid (Lovley et al. 2000). A role has been defined for
excreted quinone from cell walls in extracellular electron transfer (Newman and Kolter 2000).
Fe(ll)(aq)
Fe(ll)(aq;
FeflDaq
241
-------
Figure 4. Possible Mechanisms for the Fe(III) Solubilization and/or Reduction to Fe(II).
There is recent evidence that certain aerobic and facultative microorganisms have the
capability to excrete siderophores [highly Fe(III)-specific bidentate ligands] that migrate to
surfaces of iron-bearing minerals and complex Fe(III) into solution (Hersman et al. 1995)
thereby dissolving hematite. In another very significant finding, a c-type cytochrome was found
excreted as an extracellular electron carrier by Geobacter sulfurreducem into a growth medium
(Seeliger et al. 1998). Geobacter sulfurreducens cytochrome type-c reduces ferrihydrite, Fe(III)-
nitroacetic acid, Fe(III) citrate and MnO2 at high rates. Humic acids, S°, and anthraquinone
sulfate are also reduced, but more slowly. Geobacter sulfurreducens reduced the cytochrome-c
with acetate as a TEA and oxidized it with fumarate. Wolinella succinogenes reduced then
externally provided cytochrome-c of Geobacter sulfurreducens with molecular H2 or formate as
a TEA and oxidized it with fumarate or NO3". Experimentally, it was possible to establish a
coculture in which Geobacter sulfurreducens reduced the cytochrome-c with acetate and the
reduced cytochrome-c acted as an Fe(III) reductase for electron transfer to insoluble Fe(III)
oxyhydroxides, S°, MnO2 or other oxidized compounds. The reduced cytochrome-c also
transferred electrons to a partner bacteria, Enterbacter cloacae, which, in turn, affected a two-
electron reduction of nitroaromatic compounds (Nivinskas et al. 2000).
Under aerobic conditions, Fe is mainly present as Fe(III) oxyhydroxides including
Fe(OH)3(amorphous), FeOOH (goethite), and FeOOH (hematite) (Schwertmann 1991).
However, under anaerobic conditions two processes can occur: (1) Fe(III) oxides can be
transformed into mixed valency Fe oxides such as F 63(1)4 (magnetite) and FesO/^amorphous)
(Brennan and Lindsay, 1998); and (2) dissimilatory microbial reduction (Shewanella and
Geobacter species, see Figure 3, Lovley 1997 and Lower et al. 2001) of Fe(III)s to Fe(II)s and
dissolution of Fe(II)s to Fe(II)aq. Iron(III) can bind to both soluble dimeric and insoluble
polymeric forms of lignin (Guillon et al. 2001); the dimeric form produces two stable Fe(III)
complexes — FeL2+ and FeL(OH)+. The polymer has a great affinity for Fe(III) with minimum
sorption at pH 5. A pulsed ESR study showed surface oxidation by the Fe(III) cation of quinone
leading to Fe(II) and the presence of semiquinone radicals on the polymer surface with a radical
concentration of about 5 X 1017 spin/g. This semiquinone radical would be available to oxidize
S2" to a higher oxidation state (S°, SOs2" or possibly all the way to SC>42"). Interestingly,
Shewanellaputrefacien can reduce Fe(III) in Fe(III)-oxides to Fe(II) and also reduce SOs2" to S 2"
but not SC>42" to S 2". Five Geobacteraceae strains have been found (Coates et al. 1999) that can
reduce Fe (III) in Fe-oxide through dissimilatory reduction to Fe(II).
Anaerobic oxidation of Fe(II) minerals by lithotrophic, acidophilic and neutrophilic
bacteria has been recognized for many decades (Ghirose 1984, Harrison 1984). Anaerobic Fe(II)
oxidation was discovered only recently, with the isolation of phototrophic purple, non-sulfur
bacteria that were able to utilize Fe (II) as an electron acceptor in light (Wadden et al. 1993).
Also of note is that a NCV-reducing bacteria has been found that gains energy for growth by
oxidizing Fe(II) anaerobically (Straub et al. 1996). The energetics of iron reduction and
oxidation at neutral to alkaline pH differ substantially from those in the acid range. When the pH
is below 2.5, the standard redox potential of the Fe3+/Fe2+ couple is 771 mv (refer back to Table
5). Iron transformations at neutral to alkaline pH are carried out by different microorganisms
because they deal basically with different chemical species. Generally, Fe(III)-reducing bacteria
242
-------
live in surface-associated communities such as biofilms. The biochemistry of microbial Fe(III)
reduction is still only partly understood. Fe(III) reductase is detectable only in anaerobically
grown cells and is predominately found in the outer membrane (Lloyd et al. 1999). The c-type
cytochrome mentioned earlier play a major role in electron transfer. In the transfer of electrons
to Fe(III) oxides, the c-type cytochromes involved are localized mainly in the outer membrane
and in the periplasmic space (Myers and Myers 1997, Myers and Myers 2000) (see Figure 2).
Further evidence is seen in the work of Gaspard et al. (1998) where they found that Geobacter
sulfurreducnes Fe(III) reductase is a peripheral protein in the outer face of the outer membrane
and its activity is associated with c-type cytochrome that is oxidized by ferrihydrite.
The interface between the facultative anaerobe, Shewanellaputrefaciens, and the iron
oxide, goethite, differs (as determined by AFM force measurements) depending on whether the
environment is aerobic or anaerobic (Lower et al. 2001). Possibly, this is due to different
proteins (or different conformations of the same protein), e.g. a cytochrome reductase present in
the cell wall under anaerobic conditions compared to aerobic conditions. Causative factors can
include: (1) the chemical and structural features on the surface of the microorganism
(concentration and location of cytochromes and reductases and the physical structure of the outer
membrane); and (2) the nature and properties of the mineral surface — density and concentration
of Fe-O moieties, surface morphology, and crystallographic orientation. Sulfate-reducing
bacteria can directly reduce Fe (III) enzymatically, producing siderite (Fe COs) rather than iron
sulfide (Coleman et al. 1993). Dissimilatory reduction can also be beneficial in degrading
organic contaminants. Lovley et al. (1989) found that previously observed refractory aromatic
hydrocarbons, such as benzene, can be oxidized during the dissimilatory reduction of Fe(III) to
Fe(II). The presence of Fe(II)aq reflects the production of Fe(II) by microbial reduction of
Fe(III)-bearing mineral phases, although the sorption of Fe(II) to mineral surfaces, in particular
oxide surfaces, or precipitation of Fe-bearing minerals - pyrite, siderite and magnetite - as noted
earlier may be sinks for the Fe(II) produced.
Microbial enzymatic processes, rather than abiotic processes, are mainly responsible for
Fe(III) reduction to Fe(II) (Lovley et al. 1991); this only occurring in an anaerobic environment.
The oxygenation reaction of Fe(II) and the formation of Fe(III) oxyhydroxides are accelerated in
the presence of phyllosilicate clays and various aluminum oxides (Yiwei and Stumm 1994).
Polymerization of Fe(III) oxyhydroxides on clay surfaces occurs mainly in two, rather than
three, dimensions. These Fe(III)oxyhydroxides, in turn, can be reduced by fulvic acids and
Fe(III) ions released into the aqueous phase. Therefore, the Fe(III)-OH/Fe(II) couple is able to
function as an electron transfer mediator for the oxidation of organic matter by molecular 62,
either in the absence or presence of microorganisms, or as a supplement to microbial activity.
Bacteria influence the nature, rate and precipitation of iron-bearing minerals. Warren and
Ferris (1998) found that solid-phase partitioning of Fe(III) as hydrous ferric oxide (HFO) was
enhanced in the presence of a variety of bacteria over that found in abiotic controls. The start of
UFO formation occurred at lower pH values and in greater quantities at any given pH in the
presence of bacteria. Fe(III) reactions at bacterial surfaces follow a clear continuum between
243
-------
sorption and precipitation that can be quantitatively defined using geochemical principles, and
modeled using surface precipitation theory. This process is called biomineralization. Not only
can biotic-induced reduction (Shewanellaputrefaciem CN32) of an Fe(III) mineral, e.g.,
ferrihydrite to [Fe(H2O)6 ]2+ occur, but the bacteria can take Fe(II) into the cytoplasm and form a
ferrihydrite phase mineral (Glasauer et al. 2001). Warren and Ferris (1998) also found that
ferrihydrite immediately bound to the bacterial surface upon the introduction of the bacteria into
a ferrihydrite suspension. Strong chelating agents, such as bacterial siderophores, humic and
fulvic acids, citrate, and EDTA, were found to inhibit the growth and dissimilatory Fe(III)
reduction by Shewnalla sp in the presence of Fe(III) oxides (Hass et al. 2001).
Microorganisms apparently also reduce structural Fe(III) in iron-bearing alumino-
silicates (clay minerals) (Wu etal. 1988, Lovley et al. 1990, Kostka et al. 1996). The
mechanisms for this microbial reduction have yet to be identified, and many questions arise as to
the precise role of the microorganisms in these Fe-redox reactions. For instance, does the
reduction occur because of a membrane-bound process requiring intimate contact between the
clay mineral and organism, or is it due to extra-cellular or exudate compounds from the
organism? Is it an aerobic or anaerobic process? What are the metabolic sequences responsible
for reduction? As noted previously, the half-cell potential for solid state oxidation [Fell] =>
]Fe(III)] is -340 to -650 mv making structural Fe(II) a stronger reducing agent than [Fe(H2O)6]2+
(-771 mv). Contact between the microorganisms and the Fe(III)-bearing clay appears to be
required for some microorganisms (Lovley 1997), suggesting that this reduction is a membrane-
associated phenomenon. Fe(II) can also be oxidized to Fe(III) by anoxygenic, phototrophic
bacteria (Wadden et al. 1993).
Under oxic conditions nanomolar concentrations of organic ligands - natural (Millero
1998) and synthetic (Witter etal. 2000) can increase the solubility of Fe(III) from 32 to 65%,
apparently due to the formation of Fe(III)-organic ligand complexes. Decreasing pH increases
the solubility of Fe(III) minerals, e.g. Fe(III) solubility at pH 8 is 0.2 nmolar and 0.6 nmolar at
pH7.65.
The redox potential of metal sulfides varies among the different metal sulfides and as a
function of the metal cation reacting at the sulfide surface (see Table 6). However, in all cases
the reducing environment at the surface of pyrite is greater than the other metal sulfides studied,
and was independent of the reacting metal. The order of increasing reducing condition at the
metal sulfide surface is galena (PbS)< sphalerite (ZnS) < chalcocite (Cu2S) « pyrite (FeS2).
The water solubility increases in the same order, while the solubility product decreases in the
same order. Metal sulfide surfaces are very reactive. Lin et al. (1989) found that different types
of reactions occurred as a function of metal sulfide type and the nature of the reacting metal
aqua-ion. These reactions included: (1) reduction [Ag(I) was reduced to Ag(0)]; (2)
precipitation of the reacting/guest metal ion onto the respective metal sulfide (this occurs when
the solubility product of the guest metal ion is greater than that of the host metal ion with S 2";
also the host metal sulfide dissolves); and (3) surface complexation. Metal sulfides also were
found under anaerobic conditions to reduce nitrobenzene to aniline through an eight-electron
reaction (Yu and Bailey 1992); they also found that the nitrobenzene reduction was a solution,
not a surface phenomena, and that the sulfide mineral dissolution rate and its solubility
influenced the reaction rate.
244
-------
The acidity and redox character of surfaces, particularly iron-bearing mineral surfaces,
are important in biogeochemical processes. The data in Table 7 show that the Hammett surface
acidities for the iron bearing minerals, except pyrite, were more acidic than bulk pH values (Yu
and Bailey 1996). Therefore, the surface Eh values reflect a more oxidized environment than the
bulk Eh values of all iron-bearing minerals except pyrite.
Table 6. Redox Potential of Metal Sulfide Systems.
System
Cu/Sp
Ag/Sp
Cu/Py
Cr/Py
As/Py
EH (mv)
-46
-29
-219
-176
-201
System
Cu/Ch
Ag/Ch
Ag/Py
Ba/Py
~
EH (mv)
-58
-96
-229
-173
—
System
Cu/Gl
Ag/Gl
Cd/Py
Fe/Py
EH (mv)
-73
-21
-193
-188
Sp = sphalerite; Ch = chalcocite; Gl = galena; and Py = pyrite
(After Lin et al. 1989).
Table 7. Redox, Acidity, and Characterization of Six Fe-Bearing Mineral Suspensions.
Humic substances can act as a mediator - a TEA — for microbial catalyzed metal
reduction (Lovley et al. 1998). Geobacter metallireducem can transfer electrons to humics
during the dissimilatory reduction of structural iron in phyllosilicates and the crystalline
245
-------
Fe (III) forms, goethite and hematite. Electron shuttling between Fe(III)-reducing
microorganisms and Fe(III) via humics did not accelerate the microbial reduction of poorly
crystalline Fe(III) oxide, but did facilitate the reduction of Fe(III) forms that are not typically
reduced by microorganisms in the absence of humics. Similar experiments with a variety of
quinones (known components of humics) stimulated Fe(III) oxide reduction (Lovley et al. 2000).
Geothrix fermentem anaerobically degrades the anthropogenic compound, benzene, (via
anaerobic oxidation) to CC>2, resulting in the reduction of Fe(III)oxide. The persistence of
anthropogenic aromatic and aliphatic compounds can be affected by redox conditions. Heijman
et al. (1995) found that 10 monosubstituted nitroamines were stoichiometrically reduced to the
corresponding amine in laboratory aquifer columns; they proposed that the reduction occurred
primarily by a reaction with surface-bound iron species that served as mediators for the transfer
of electrons originating from microbial oxidation of organic compounds by iron-reducing
bacteria. Dissimilatory iron and/or manganese reduction is known to occur in the presence of
several organisms including anaerobic sulfur-reducing organisms, such as Geobacter
metallireducens and Desulfuromonas acetoxidan, and facultative aerobes, such as Shewanella
putrefaciens. These bacteria coupled carbon oxidation and growth with reduction of the cited
metals and inhibitors. Manganese (IV) and Fe(III) are efficient electron acceptors, similar to
NOs" in ability, and are capable of out-competing TEAs of lower potential such as SC>42" or CC>2
(Nealson and Saffarini 1994). Microorganisms also can catalyze the oxidation of H^ coupled
with the reduction of NO3", Mn(IV), Fe(III), SO42" and CO2 (Lovley and Goodwin 1988). Each
TEA reaction had a unique range of steady-state H2 concentration associated with it.
Manganese Redox Chemistry
In the natural environment, manganese is found as reduced, soluble, or adsorbed Mn(II),
and as insoluble Mn(III, IV) oxides. Manganese oxidation is accomplished by a great variety of
microorganisms, while Mn reduction (i.e., Mn (III, IV) oxide) to the aqua-Mn 2+ cations in
solution is accomplished only by strictly anaerobic bacteria (Gounot 1994). The rate of
reduction is a function of the degree of crystallinity (amorphous particles are reduced faster) and
particle size (rate increases exponentially with decreasing particle size). For example, Mn(IV)-in
MnO2 (e.g., birnessite) can be reduced to aqua-Mn 2+ cations through microbial dissimilatory
reduction in an anaerobic environment (Lovley and Phillips 1988, Lovley and Phillips 1992,
Lovley 1993).
Bacterial reduction of manganite (MnOOH) [Mn(III)] occurred in the presence of
Shewanella putrefaciens with the formation of aqua-Mn 2+ ions (Larsen et al. 1998). The rate of
reduction was optimal at pH 7.0 and 26°C, consistent with an enzymatic reaction. The rate of
reduction was proportional to the amount of manganite added but essentially independent of cell
concentration, indicating the dominating role of mineral surface properties on the kinetics of the
reduction reaction. Major differences were noted in the reduction rates when surface area was
varied. The importance of oxygen content and bacterial cell-mineral surface contact was also
noted. No Mn(III) reduction occurred when the samples were oxygenated or when the cells were
physically separated from the manganite crystals by a dialysis membrane. This suggests that
246
-------
physical contact between the cell and mineral surface is mandatory for the electron transfer
process to occur and metal reduction and solubilization to result. In the early stages of reduction,
scanning electron microscopy imaging showed close contact between the cells and the needle-
shaped mineral crystals. During the latter stages of the reduction process, the closely associated
cells were coated with a layer of extracellular polymeric material that had not been noted
previously, and the mineral surface was coated with a bio-film like layer of associated cells.
Mn(III, IV) oxides can also be reduced to the aqua-Mn2+ ions in the presence of Cr(III)
ions in solution (Risser and Bailey 1992, Risser and Bailey 1997). Manganese concentrations
can be quite high in (^-depleted water overlying sediments in which manganese reduction has
occurred (Thamdrup et al. 1996). Manganese is almost exclusively a marine oxidant, while iron
plays an important role in freshwater sediments. Oxidation by manganese can account for more
than 90% of total carbon oxidation in some marine sediment (Canfield et al. 1993). The
presence/accumulation of Mn(IV)oxide in anaerobic sediments, where Fe(III) reduction was the
TEA, removed all the dissolved Fe(II) from the aqueous phase, completely inhibited net Fe(III)
reduction and stimulated Mn(IV) reduction (Coates et al. 1998). This was apparently due to the
oxidation of Fe(II) to Fe(III) by MnOx, thereby inhibiting the microbial reduction of Fe(III).
Chromium Redox Chemistry
Cr(VI) is very toxic to humans and injurious to the environment. However, Cr(III), is the
major environmental species present under anoxic conditions, is present as a very insoluble oxide
and is nontoxic. Iron speciation influences the form of the adsorbed Cr (Abdel-Samad and
Watson 1997). Sorption increases with decreasing pH, reaching a maximum near pH 6.5. At
low redox levels (+100 mv), Fe(III) is reduced to soluble Fe(II), which in the presence of Cr(VI)
affects its reduction to Cr(III) (Masschelyn et al. 1992). Under moderately reduced conditions
(+500 to +100 mv), Cr behavior is dominated by Cr(VI) sorption and the reduction of soluble
Cr(VI) to insoluble Cr(III). The chemical nature of any complexing ligands also affects the
transformation kinetics of Cr(VI) to Cr(IV)( Buerge and Hug 1999).
Mercury Redox Chemistry
Although Hg is normally present in the environment at extremely low concentrations
(5.0 x 10 "9 M), because of its ability to be concentrated in living tissue and its high toxicity Hg is
of considerable environmental importance. The major form of Hg in the atmosphere is Hg(0),
which is volatile and oxidized to Hg(II) photochemically. The majority of the mercury in the
aquatic environment, is therefore, Hg(II). Once in the aquatic environment, Hg(II) is readily
bound to particulate matter and can be metabolized from there by microorganisms. The
principal microbial reaction is the methylation of Hg(II) yielding CH3Hg+. This latter compound
is water soluble and can be concentrated through the aquatic food chain primarily into fish and/or
247
-------
further methylated to (CH3)2Hg. A variety of microbial genera can methylate Hg(II) including
Bacillus, Escherichia, Aerobacter, andEnterbacter. Methylation increases the lipophilicity of
Hg and thereby its toxicity (Craig 1986). Other bacteria can further methylate (CH3 )2Hg+ by
utilizing S 2" in a dismutation reaction to produce (CH3)2Hg and HgS (Siciliano and Lean 2002).
Abiotic methylation of Hg has also been observed when the methane donor molecule (enzymes,
e.g., methionine synthetase, acetate synthetase and methane synthetase) has been added to water
containing Hg. Mercury transformation takes place in the atmosphere, in solution, and bound on
sediment. Starting from Hg(II), bacteria can methylate or dimethylate Hg. Thus, (CH3 )2Hg+ can
be reduced back to Hg(II) or CH3Hg+. CH3Hg+ can be reduced to produce CH4 and Hg(0), both
of which are volatile.
Redox Chemistry of Other Metals
A variety of metals are redox and/or pH sensitive, including Co, Cd , Ni, Zn, and Pb, and
their speciation will be different depending on the environmental state variables. Under oxic
conditions in sediments, Zn chemistry is controlled by the redox chemistry of Fe(III) and Mn(IV)
oxides through sorption/ligand exchange to the oxide surface (Gao et al. 1997). Cadmium
transformations are controlled by both Fe(III) and Mn(IV) oxides and carbonates. Under a
reducing condition, the behaviors of Zn and Cr are controlled primarily by sulfides and insoluble,
high molecular weight humic substances; the behavior of Cd is controlled by carbonates. When
the sediment redox potential increases, the affinity between Fe(III) and Mn(IV) oxides and Cd,
and Zn increases. When the sediment redox potential decreases, the affinity between carbonates
and Cd and Zn increases; the affinity between insoluble sulfides, large molecular humic
substances, and Cd and Zn increases; and the soluble levels of Cd, and Zn decreases. Results
suggest that reducing sediment conditions would reduce Cd and Zn toxicity, presumably through
the formation of Cd and Zn sulfides. Under aerobic conditions, the presence of reactive free
oxygen species may have an important influence on metal speciation (Green et al. 1990, Zhou
and Mopper 1990, Zepp et al. 1992, Blough and Zepp 1995).
REDOX CHEMISTRY OF SULFUR, CARBON, PHOSPHOROUS,
METALLOIDS, ACTINIDES, AND HALOGENS
Next we will we examine the redox chemistry of sulfur, carbon and phosphorous (the
major nutrients), metalloids, actinides, and the halogens.
Sulfate Redox Chemistry
The sulfate anion (SC>42") is an important constituent of seawater. There are, however,
several other species of sulfur that play an important role in the biogeochemistry of sulfur
including sulfide (S2"), elemental sulfur [(S(0)], thiosulfate (S2O32"), tetrathionate (S^e2"), sulfite
(SO3 2"), sulfur dioxide (SO2), sulfur trioxide (S03) and sulfhydyrl (R-SH). The significant
valence states/forms in the marine environment are -2, (S2"); 0, (S°) and +6, (SC>42") (Madigan et
al. 1997).
248
-------
Sulfate reduction to S2" is an eight-electron and an eight-proton reaction (see Table 1)
mediated by sulfate-reducing bacteria. The form in which S2" is present depends on the pH. At
high pH, the dominant form is S2". At neutral pH, HS" predominates. Below pH 6, H2S, a
gaseous product, is the major species. HS" and S2" are water-soluble, but H2S is not and readily
volatilizes. Even at neutral pH, some volatilization of H2S can occur because of the equilibrium
between HS" and H2S; as volatilization occurs, the equilibrium shifts toward H2S.
Microbially driven SO42" reduction can occur via two pathways - assimilative and
dissimilative. In assimilative SC>42" reduction, a variety of microorganisms can carry-out SC>42"
reduction converting HS" produced to organic sulfur. HS" is ultimately reformed by
decomposition of the organic sulfur via putrefaction and desulfurylation (Madigan etal. 1997).
Dissimilatory sulfate reduction is where SO42" acts as an electron acceptor and the reaction is
mediated by a variety of obligate anaerobes. A variety of electron donors can be used for
microbially driven SC>42" reduction including H2, lactate, pyruvate, fumarate, alcohols, acetate,
malate, propionate, butyrate, long chain fatty acids, benzoate, indole, and hexadecane. These
bacteria carry-out a cytochrome-based electron transport process that transfers electrons from the
energy source (oxidation of organic matter) to the SC>42" ion in adenosine phosphosulfate (APS)
and to SOs2", the first product of SC>42" reduction. Once SOs2" is formed, the subsequent
intermediary reactions proceed with HS" being the final reduction product.
Most marine SC>42"- reducing bacteria oxidize acetate as their sole energy source. These
organisms oxidize acetate to CO2 and reduce SC>42" to HS". Certain SC>42"- reducing bacteria are
capable of a unique form of energy metabolism called disproportionation, using sulfur
compounds of intermediate oxidation state - S2O3 2", SO3 2", and S°. For example, Desulfovibrio
sulfodismutant disproportionate S2C>32" to SC>42" and HS" with a free energy change of-21.9
kJ/reaction. Note that one sulfur atom of S2C>32" becomes more oxidized (forming SC>42") and the
other more reduced (forming HS"). Another disproportionation reaction involves SOs2" and
results in the formation of SO42" and HS" with a free energy change of -235.6 kJ/reaction. S° can
also be disproportionated into SO42" and HS", but this reaction is thermodynamically not favored
— + 40.8 kJ/reaction. However, if the HS" formed is oxidized back to S° by chemical reaction
with birnessite (MnOx) [forming also the aqua-Mn2+ ion in solution (free energy change of-140
kJ/reaction), the summation of these two reactions results in a favorable overall free energy
change of -100.6 kJ/reaction. Therefore, net sufficient energy is available to support the growth
of S°-disproportionating bacteria. However, unlike S2Os2 -and SOs2 "disproportionating
bacteria, S°-disproportionating bacteria require an electron acceptor such as Mn(IV) to drive the
energetics of the reaction.
Due to the necessity of organic electron donors (or molecular H2) derived from the
fermentation of organic compounds, SO42" reduction occurs most readily where organic matter is
prevalent. In many marine sediments or anoxic seawater columns, the rate of SO42" reduction is
carbon-limited and the rate can be greatly increased by the addition of organic matter. Marine
disposal of sewage sludge and garbage can lead to a marked increase in organic matter in
sediments and in the anoxic water column. This would result in an enhanced rate of SO42"
reduction and the formation of HS". Because HS" is toxic to many organisms (including man,
when in the gaseous form - H2S), formation of HS" by SO42" reduction is potentially
249
-------
hazardous. S2" is also toxic because it combines with the iron of cytochrome and other metal-
containing compounds in cells impeding their function, and disrupts disulfide bonds of proteins.
Formation of insoluble metal sulfides (Fe, Cu, Pb, Zn, Hg), however, may lower S2"
concentration in solution below levels toxic to marine/fresh water organisms.
Eighteen genera of dissimilatory-bacteria are currently recognized (Madigan etal. 1997)
and can be divided into two broad physiological sub-groups: sub-group I — Desulfovibrio,
Desulfuromonas, Desulfotomaculum, and Desulfobulbus; and subgroup II - Desulfobacter,
Desulfococcus, Desulfosarcina. Members of sub-group I utilize lactate, pyruvate, ethanol, and
certain fatty acids as carbon and energy sources, reducing SC>42" to H2S. Members of sub-group
II oxidize fatty acids to CO2; while reducing SC>42" to H2S.
In summary, the biogeochemistry of SC>42" concentration in marine and fresh water
systems reflects a balance between SC>4 " production via gypsum dissolution and pyrite oxidation
versus bacterial SC>42" reduction, pyrite or barite precipitation, and possibly sulfur or S2O32"
disproportionation reactions (Fredrickson and Onsott 2001).
Carbon Redox Chemistry
The process of hydrogen trophic methanogenesis is a result of the microbially driven
transformation of CO2(g) [C(IV)] into CH4(g) [(C-IV)]. The energetics of this process can be seen
in Table 1. The effect of TEAs and reductant type are discussed in the work of Peters and
Conrad (1996). The inhibitory mechanism of nitrate transformation and its denitrification
products (NO2", NO, N2O) on the production of CH4, and the rates and concentrations of the
reductants (H2, acetate, propionate, etc.) and oxidants [NO, N2O, Fe(III)] involved, were
evaluated in slurries of an anoxic Italian rice soil. Addition of N-compounds caused a complete,
but largely reversible, inhibition of methanogenesis. Nitrate, NO 2", and N2O significantly
decreased the H2(g) partial pressure below the threshold of the methanogens, thus shutting down
the exergonic production of methane (G>0, where AG= free energy of reaction). Furthermore,
significant production of the electron acceptors --Fe(III) and/or SO42" - was observed after
addition of NOs" and N2O, probably due to the oxidation of reduced iron and sulfur species using
the NOs" and/or N2O as electron acceptors. Methanogenic activity did not resume until all added
electron acceptors were reduced and the resulting increase in H2 level had reached the
methanogenic threshold again. Thus, competition for H2 with the denitrifying bacteria and the
Fe(III)- and SO42" -reducing bacteria seemed to be one important factor in the inhibition of
methanogenesis. Addition of rice straw to reduce competition for electron donors did not prevent
inhibition of methanogenesis after addition of NOs", but decrease the inhibition period. With the
addition of NO2" and NO, the toxic effects of these compounds may have been more important
than H2 competition. Although addition of NO2" or NO caused a decrease in the H2
concentration, exergonic methanogenesis from H2/CO2 occurred (G<0). Nevertheless, CH4
production was inhibited. Furthermore, although acetate concentrations were generally sufficient
for exergonic methanogenesis, CFLi production was completely inhibited. A model has been
developed of SO42"-reduction and methanogenesis in freshwater sediments (Lovley and Klug
1986).
250
-------
The H2 concentration behavior reflects production by water disproportionation reactions
and microbial fermentation of organic matter balanced against the microbial consumption of H2
linked to various terminal electron acceptor processes (Fredrickson and Onstott 2001). Acetate
concentration behavior represents the balance between production by microbial fermentation and
consumption by microbial respiration and metastable equilibrium with dissolved inorganic
carbon.
Phosphorus Redox Chemistry
Phosphorus occurs both in organic forms (phosphate esters, phospholipids, and nucleic
acids) and inorganic forms [phosphate salts of such metals as Fe(III)[ (FePO4*2H2O -strengite or
phosphosiderite)], Al(III) [(A1PO4 -2H2O - variscite)], Fe(II)[ (Fe3 (PO4)2 '2H2O - vivianite)]
and Ca(II) [(3Ca3P2(V 'Ca(OH)2 -hydroxyapatite)](Winchell and Winchell 1951). In addition,
the PC>43" anion via ligand exchange can replace the OH- group in kaolinite and become a
structural part of the octahedral sheet. Phosphate minerals can occur as a discrete solid
phase/particle or as a coating on the surface of other minerals. The structure may be crystalline
or amorphous, the former being more stable and less soluble then the latter. Phosphorus
constitutes an essential portion of adenosine triphosphate (ATP), which upon hydrolysis to
adenosine diphosphate (ADP) forms the basis for most energy transfer within the microbial cell.
Phosphate concentration in fresh or marine waters is frequently correlated with euthrophication.
Elevated PO43"concentrations contribute to accelerated growth of algae. pH influences the
H2P(V • "HPC^2" equilibria, with the former occurring under more acidic conditions.
Dissimilatory reduction and dissolution of Fe(III)-bearing phosphate minerals releases
[Fe(H2O)6 ]2+ and PO43" ions into the pore water or the overlying water column. Thus, diffusion
of PC>43" upward in the water column may contribute to euthrophi cation.
Arsenic, Selenium, and Tellurium Redox Chemistry
Arsenic, Se, and Te are metalloids, not metals. Generally, their concentration in the
water column is low due to a combination of sorption and precipitation reactions similar to that
found for phosphorus.
Arsenic
Arsenic occurs in three principle inorganic redox forms - As (V) [arsenate (AsC>43")],
As(III) [arsenite (AsOs3")] and As(-III) [ arsine gas (AsH3)], and one organic form -
dimethylarsenic acid [(CH3)2 HAsO4]. Under oxidizing conditions, As behavior is governed by
the redox chemistry of Fe(III) and Mn(IV) oxides. When sediment redox potential increases, the
affinity between Fe(III) and Mn(IV) oxides and As increases. When sediment redox potential
decreases, the affinity between insoluble sulfides, high molecular weight humic matter and As,
increases; the soluble As remains constant. Arsenate and AsOs3" can be reduced to AsH3 by
bacteria under anaerobic conditions and Desulfomicrobium strain Ben-RB and
251
-------
Desulfovibrio strain Ben-RA can produce (CHa^ HAsO4 and AsOs3" from AsC>43" (Macy et al.
2000). Arsenate reduction to AsOs3" occurs at a redox potential of + 150 mv (see Figure 5). In
high organic matter situations, As chemistry is governed by high molecular weight humic
substances, sulfides, and Fe- and Mn-oxides. The solubility of As remains low and constant
under both aerobic and anaerobic conditions when solutions of elevated dissolved organic matter
are present. Following dissolution of Fe(III) and Mn(III, IV)-oxides, As(V) is released into the
aqueous phase, reduced to As(III) and most likely precipitated as a metal sulfide (Carbonell-
Barrachina et al. 1999). The critical redox potentials in natural systems for As can be seen in
Figure 5. The reduction of As (V) to As(III) occurs in the +125 to +175 mv range.
Selenium
It is becoming more evident that biological cycling of Se is similar to that of sulfur,
which is located directly above Se in the Periodic Table. Selenium undergoes various redox
reactions that are biological in nature that may directly affect its redox state and, therefore, its
chemical properties and behavior in the environment. Selenium, like sulfur, undergoes both
assimilatory and dissimilatory reactions.
Six principal forms of selenium exist. These include: selenide (Se-II), elemental
selenium [Se(O)], selenite (SeO32") [(Se (IV)], selenate (SeO42") [Se(VI)], and two organic
selenic compounds - methyl-and dimethyl-selenic acid. Selenate is reduced to SeOs2" under
anaerobic conditions, and ultimately to Se(-II). Most bacteria that are capable of SeO42" or
AsC>43" reduction can also use several TEAs such as Fe(III), Mn(IV) and organic compounds. In
most instances, these bacteria exhibit a facultatively anaerobic form of metabolism (Madigan et
al. 1997). In the case of both metalloids, water solubility and toxicity increases with increasing
oxidation state. Chemical thermodynamics predicts that the reductive sequence should be NOs^
SeC>42" ~> MnO2 at pH 5 (Sposito et al. 1991). The latter authors found that microbial reduction
of SeC>42" to SeOs2" in soils occurred under anaerobic conditions, and that the rate of reduction
decreased in the presence of NOs". Soluble Mn levels increased concomitantly. Critical redox
potentials in natural systems for Se are also presented in Figure 5. The reduction of Se(VI) to Se
(IV) occurs in the +225 to +275 mv range, while that for the reduction of Se (IV) to Se(0,-II)
occurs in the -25 to - 50 mv potential range. Metalloid speciation depends on the prevailing
processes of dissolution, reduction, solution, ligand complexation and precipitation (refer back to
Figure 1). The pH and Eh control this equilibrium, even though the kinetics may be very slow.
In many instances, this equilibrium is irreversible or requires extremely long times to reach an
equilibrium.
252
-------
ANAEROBIC
HIGHLY
REDUCED
REDUCTION
REDUCED
MODERATELY
REDUCED
I AEROBIC
i REDUCTION
OXIDIZED
BIOLOGICAL REDUCTION REACTIONS
C02—•> CH4
(HEME DERIVATIES)
AsO43"—«-AsO/'
Glucose-* CO2+e"|
+ H!"
SeO,"—- Se° + Se2'
SO/--S2-
,-^ ti+ r^ 3-
Cr —>• Cr
NO;—» N,
SeO,,2"—* SeO,2"
Fej -"Fe-
Pu"+—>Pu3+ \ Mn4+—»Mn2*
O2 —«• H,O
1
-400 -200 -100 0 +100 +200
Redox Potential (mV) at pH 7
+300
+400
Figure 5. Critical Redox Potentials for Transformation of Various Redox Couples in Aquatic
Ecosystems. (After Lovley and Phillips 1992; (Rusin et al. 1994; DeLaune et al. 1995;
Councell et al. 1997).
Abiotic reduction of Se(VI) to Se(0) occurs in a suboxic environment in the presence of
"green rust" [Fe(II, III)] oxide (Myneni et al. 2001). These latter authors found that green rust
converts to goethite, lepidocrocite, maghemite, or magnetite depending on the rate of oxidation
and dehydration of the green rust. The Gibbs free energy change for the reaction is -671
kJ/reaction, which indicates the reaction is highly favorable to occur. The conversion of the
carbonate analog of green rust is favored in marine systems. When Se(VI) was present during
green rust precipitation, the aqueous form of Se(VI) decreased 48% in 36 sec and reduction
decreased slowly after that. In reactions with previously precipitated green rust surfaces, Se(VI)
reduced directly to Se(0). Interlayer-trapped Se(VI) formed bidentate, binuclear and edge-
253
-------
sharing complexes with Fe(II), and was reduced immediately to Se(IV), and then slowly
converted to Se(0) and Se(-II). Therefore, Se(VI) reduction by coprecipitation and adsorption
pathways can occur in an anoxic environment. Green rust may be the mediator for the abiotic
reduction of such elements as As(V), Te(VI), Cr(VI), and U(VI).
Reductive dissolution of Fe(III) oxides precipitates green rust and Se(0). Selenium can,
therefore, be reduced by both abiotic and biotic pathways. One possibility is that biotic
respiratory processes generate electrons (cytochrome reductase transport system) that use Fe(III)
oxide as a terminal electron acceptor (dissimilatory reduction), reducing Fe(III) to Fe(II) and
releasing Fe(II) into solution from which green rust is precipitated. As noted, this latter surface,
in turn, abiotically reduces Se(VI) ultimately to Se(-II).
Tellurium
Research over the last three decades has conclusively demonstrated that As, Se, and Te
can be methylated by microbes. For example, Basnayake et al. (2001) showed that
Pseudomonasfluorescens K27, a facultative anaerobe, can methylate Te, producing
dimethyltelluride from either sodium tellurate or sodium telluride; Te(0)is also produced by this
organism.
Actinide Redox Chemistry
The speciation and microbially driven-transformations of the actinides - Pu and U - are
examined.
Pseudomonas aeruginosa and Shewanellaputrefaciens (both are Fe(III) and SC>42" -
reducing bacteria) can use U(VI) as a TEA, reducing it to U(IV) (Abdelouas et al. 1998).
Sulfate was concomitantly reduced to S2" and the U(IV) precipitated as (U, Ca)C>2 on the surface
of the bacteria. Uranium in solution is complexed by carbonate, and may exist in this form in the
marine environment.
Anoxic conditions also may effect the reduction of U(VI) to U(IV) via Desulfovibrio
desulfur-icons, resulting in the precipitation of U(IV) (Lovley and Phillips 1992). Further work
(Lovley et al. 1991; Lloyd, 2001) showed that dissimilatory Fe(III)-reducing microorganisms
can derive energy for growth by reducing U(VI) to U(VI). Following O2 and N(V depletion,
U(VI) was reduced to U(IV) by a SO42"-reducing bacteria (Abdelouas et al. 2000). The uranium
IV was precipitated as a hydrated uranite (UO2 -xH2O). Fe(III) and Mn(IV) were reduced as
well.
Rusin et al. (1994) reported that under anaerobic conditions Fe(III)-reducing bacteria
solubilize plutonium hydrous oxide, PuO, in a similar fashion as was found for Fe(III) oxides.
The presence of nitriloacetic acid (NTA) facilitated PuO solubilization. In the presence of NT A,
90% of the PuO solubilized, while only 40% solubilized in the absence of NTA. One
explanation for the accelerated rate of reduction and solubilization in the presence of NTA is that
254
-------
the aqua-Pu(III) ion product of Pu(IV) reduction is complexed by NTA. The process of
complexation lowers the concentration of Pu(III) in solution, thus fostering the dissolution
reaction according to the law of mass action. In addition, lowering the concentration of Pu(III)
in solution may also decrease the adverse effect of Pu(III) ions on the growth and reproduction
of iron-reducing bacteria.
Iodine Redox Chemistry
The different oxidation states of iodine, like other elements, markedly affect its sorption,
bioavailability and transport in the environment. Councell et al. (1997) demonstrated the direct
microbial reduction of KV to I" at pH 7.0 by the SO42-reducing, anaerobic bacteria, Desufovibrio
desulfuricans. Under anaerobic conditions, soluble Fe(II) and S2", as well as FeS, can also
reduce IO3" to I".
METAL AND METALLOID TOXICITY
A major concern of hypoxic or anoxic conditions in the environment is the potential for
adverse effects on the biota. From Figure 1, we can see that a change in redox and pH affects the
speciation and distribution of metals/metalloids, and may directly impact the microbial
populations that determine metal speciation, transport and bioaccumulation. The oxyanions of
As, Cr, Mo, Se, and V are stable forms of elements of high oxidation state that cross cell
1 0
membranes using common PC>4 " and/or/SC>4 " carrier systems (Jennette 1981). Inside the cell,
these oxyanions may affect R-SH electron transfer reactions. Often these oxyanions act as
alternate substrates to form ester products that are hydrolytically unstable (compared with sulfate
and phosphate esters) and breakdown in aqueous solution. Arsenite and SeOs2" are both capable
of reacting with R-SH groups in proteins. There are some cells able to metabolize redox-active
oxyanions to more stable oxidation states. Reaction-specific enzymes may be involved in these
metabolic processes. The metabolites of these elements may form complexes with small
molecules, proteins, and nucleic acids that impair their ability to function properly. The divalent
ions of Cd, Co, Hg, Mn, Ni, and Pb may mimic essential divalent ions such as Mg, Ca, Fe, Cu,
and Zn. These former six ions may complex small molecules, enzymes and nucleic acids in such
a manner that the normal biochemical activities of these organic species are altered. Further
more, free radicals may be produced by the presence of these metal ions that can damage critical
cellular molecules.
In the case of Hg, both CH?Hg+ and (CHa^Hg bind to proteins and accumulate in muscle
tissue. CH3Hg+ is more thanlOO times more toxic than Hg(0) and Hg(II), and can be
concentrated in fish where it is a potent neurotoxin, eventually causing death. CH3Hg+is a major
environmental toxin and its accumulation is particularly by troublesome in fresh water lakes
where enhanced levels of CH?Hg+ have been detected in fish caught for human consumption.
255
-------
SUMMARY
The initiation of the euthrophication process, i.e., reducing dissolved O2 levels and
increasing the partial pressure of dissolved CO2, puts into effect a series of biophysicochemical
pathways that can affect the speciation of metals, transformation pathways, mechanisms of
chemical contaminant degradation, and the microbial ecology of the water column and
underlying sediments. Ultimately, euthrophication may result in the formation of anoxic
conditions in the water column and sediments. The ecosystem shifts from an oxic environment,
where Q^ serves as the terminal electron acceptor for electrons resulting from the bio-oxidation
of organic matter, to an anoxic condition, where C>2 is essentially absent.
When the dissolved 62 has been depleted, other solutes are utilized as TEAs. With
decreasing Eh, the microbial population changes from an aerobic bacterial population toward/to
an anaerobic suite of microorganisms. A sequential series of alternate TEAs come into play:
(1) NO 3" reduction (denitrification) processes result in a series of intermediate products -
NO2"(aq), NO2(g), NO(g) and ultimately to N2(g>; (2) dissimilatory reduction and dissolution of
MnOx to aqua-Mn2+ cations [capable of forming the more soluble MnCOs]; (3) dissimilatory
reduction and dissolution of Fe(III) oxides to the aqua-Fe(II) cation, a strong reductant when
bound on the surface of oxides and capable of reducing a variety of organic chemical
contaminants [e.g., nitrobenzene (aq) to aniline(aq)]; (4) dissimilative reduction of SC>42" to a
variety of products including S2C>32", SOs2", S2", polysulfides, elemental S (both forms), and
dissolved H^S. The presence of transition/heavy metals Zn(II), Cd(II), Hg(II), Cu(II), Co (II),
Mn(II), Pb(II) and Ni(II), results in the precipitation of metal sulfides, therefore, greatly lowering
the concentration of these metal ions in solution. Microorganisms that reduce SC>42" also can
ft o O
reduce a variety of metal oxyanions, e.g., CrC>4 ", AsC>4 " SeC>4 ", to a lower valance state at which
the metal cation can either precipitate or be methylated microbially. (Microbial reduction of
metal oxyanions to a lower valance state can also result in precipitation of the oxyanion onto the
microbial cell surface. The process is called biomineralization.); (5) methanogenesis via
microbial reduction of CC>2 to CFLi; and (6) fermentation.
Therefore, the concept of metal-based respiration is applicable to a wide variety of redox-
active metals, metalloids, actinides, halogens, and carbon that serve as TEAs in microbial
respiration. These microorganisms may also play an important role in the remediation of both
organic and metal contaminants in both aquatic and terrestrial environments. Although much has
been learned about the diversity of microorganisms responsible for this respiration process from
pure culture studies, more information is needed on the activity of these organisms in their native
environments.
The major redox pathway is electron transfer, and the major "vehicle" is cytochrome c
reductase. Apparently, the ligands coordinating to the Fe(II)-heme differs and, therefore, the
conformation (and thus the conformation energy/stability) differs resulting in widely varying
redox potentials. The electrons from Fe(II) cytochrome reductase are transferred to the
appropriate TEA, reducing it, and Fe(III) cytochrome oxidase is formed. Other electrons from
the oxidation of higher energy carbon sources react with the cytochrome c to form cytochrome
reductase.
256
-------
Stepwise reduction in going from an aerobic to an anaerobic system lowers the energy of the
system, but increases its stability (Nealson and Stahl 1997). A lowering of the pH and redox of
the system until SO42" reduction potential is reached generally increases the concentration of Fe,
Mn, and many of the transition metals in solution. Generation of HS" results in the formation and
precipitation (many times on the cell wall of bacteria) of metal sulfides, greatly lowering their
bioavailability.
Work to date indicates that the bioavailability and toxicity of metals and metalloid ions to
aquatic organisms depend strongly on the chemical speciation of the metal/metalloid ions, and in
turn, on solution conditions, especially pH and the chemical character and concentration of
various ligands. Speciation and bioavailability are related quantitatively in terms of the
thermodynamic stability of metal/metalloid complexes. Total metal/metalloid concentration in
solution is not a good indicator of toxicity.
ACKNOWLEDGEMENTS
The author gratefully acknowledges Drs. Jack Jones, Robert Swank, Jr. and John
Washington, ERD-Athens, and Dr. Duane Wolf, University of Arkansas, for their incisive
review of the manuscript and their invaluable comments and suggestions. The superb graphical
work done by the Computer Sciences Corporation, for the manuscript is also gratefully
acknowledged.
REFERENCES
Abdelouas, A., Y. Liu, W. Lutze, and H.E. Nuttall. 1998. Reduction of U (VI) to U(IV) by
indigenous bacteria in contaminated ground water. Journal of Contaminated Hydrology
35:217-233.
Abdelouas, A., W. Lutze, W. Gong, E. Nuttall, B.A. Sttietelmeier, and B.J. Travis. 2000.
Biological reduction of uranium in groundwater and subsurface soil. The Science of the
Total Environment 250:21-35.
Abdel-Samad, H., and R. Watson. 1997. An XPS study of the adsorption of chromate on
goethite (a-FeOOH). Applied Surface Science 108:371-377.
Achterberg, E., C.M.G. van den Berg, M. Boussemart, and D. Davison. 1997. Speciation and
cycling of trace metals in Esthwaite: A productive English lake with seasonal deep-water
anoxia. Geochimica et Cosmochimica Acta 61:5233-5253.
Allison, L.E., and G.D. Scarseth. 1942. A biological reduction method for removing free iron
oxides from soils and colloidal clays. Journal American Society of Agronomy 34:616-
623.
257
-------
Allison, J.D., D.S. Brown, and K.J. Novo-Gradac. 1991. MINTEQA2/PRODEFA2, A
geochemical assessment model for environmental systems: Version 3.0 User's Manual.
U.S. Environmental Protection Agency, Athens, Georgia. USA. EPA /600/3-91/021.
Anderson, R.T., J. N. Rooney-Vargas, and C.V. Gaw. 1998. Anaerobic benzene oxidation in the
Fe(III) reduction zone of a petroleum-contaminated aquifer. Environmental Science &
Technology 32:1222-1229.
Anderson, R.T., and D.R. Lovley. 2000. Anaerobic bioremediation of benzene under sulfate-
reducing conditions in a petroleum-contaminated aquifer. Environmental Science &
Technology 34:2261-2266.
Atwood, J.D. 1991. Inorganic and Organometallic Reaction Mechanisms, Second Edition.
Wiley, VCH, New York.
Bailey, G.W., and S.W. Karickhoff. 1973. An ultraviolet spectroscopic method for monitoring
surface acidity of clay minerals under varying water content. Clays and Clay Minerals
21:471-477.
Bailey, G.W., and Z.Z. Zhang. 2000. Towards predicting metal ecotoxicology: Applying
coordination theory, surface chemistry and simulation models. In: Proceedings of the
Fifth International Symposium on Fish Physiology, Toxicology, and Water Quality, R.V.
Thurston (ed). U.S. Environmental Protection Agency, Ecosystems Research Division,
Athens, Georgia, USA. EPA600/R-00/015. pp. 215-232.
Bartlett, RJ. 1986. Soil redox behavior In: Soil Physical Chemistry. D. L. Sparks, (ed.). CRC
Press: Boca Raton, FL. pp. 179-207.
Bartlett, R.J., and B.R. James. 1993. Redox Chemistry of Soils, Advances in Agronomy.
Academic Press: New York, volume 50:51-208.
Basnayake, R.S.T., J.H. Bius, O.M. Akpolat, and T.G. Chasteen. 2001. Production of dimethyl
telluride and elemental tellurium by bacteria amended with telluride or tellurate. Applied
Organometallic Chemistry 15:499-510.
Bates, R.G. 1959. Electrode Potentials: Treatise on Analytical Chemistry, Part I. Interscience,
New York.
Bauer, M.E. 1978. Thermodynamics of heterogeneous iron-carbon systems: Implications for the
terrestrial promotive reducing atmosphere. Chemical Geology 22:189-206.
Bertuzzi, A. 1996. Annual variation of benthic nutrient fluxes in shallow coastal waters (Gulf of
Trieste, Northern Adriatic Sea). Marine Ecology-pubblicazioni Delia Stazione Zoologica
DiNapoli 17:261-278.
258
-------
Bianchi, T.S. 2000. Breakdown of phytoplankton pigments in Baltic sediments: Effect of anoxia
and loss of deposit-feeding macro fauna. Journal of Experimental Marine Biology and
Ecology 25:161-183.
Blough, N.V., and R.G. Zepp. 1995. Reactive oxygen species in natural waters: Active oxygen
in chemistry. Blackie Academic & Professional, London, pp. 280-332.
Brennan, E.W., and W.L. Lindsay. 1996. The role of pyrite in controlling metal ion activities in
highly reduced soils Geochimica et Cosmochimica Acta 60: 3609-3618.
Brennan, E.W., and W.L. Lindsay. 1998. Reduction and oxidation effect on the solubility and
transformation of iron oxides. Soil Science Society America Journal 62: 930-937.
Brugmann, L., R. Hallberg, and C.V. Larsson. 1998. Trace metal speciation in sea and pore
water of the Gutland Deep, Baltic Sea. Applied Geochemistry 13:359-368.
Buerge, I.J., and S.J. Hug. 1999. Influence of mineral surfaces on chromium (VI) reduction by
iron(II). Environmental Science & Technology 33:4285-4291.
Burkart, M.R,. and D.E. James. 1999. Agricultural-nitrogen contributions to hypoxia in the Gulf
of Mexico. Journal of Environmental Quality 28: 850-859.
Canfield, D.E., B. Thamdrup, and J.W. Hansen. 1993. The anaerobic degradation of organic
matter in Danish coastal sediments: iron reduction, manganese reduction, and sulfate
reduction. Geochmica et Cosmochimica Acta 57:3867-3883.
Carbonell-Barrachina, A.A., Jugsujinda, R.D. DeLaune, W.H. Patrick, Jr., F. Burlo, S.
Sirisukhodom, and A. Anurakpongsatorn. 1999. The influence of redox chemistry and
pH on chemically active forms of arsenic in sewage sludge-amended soil, Environment
International 25:613-618.
CAST. 1999. Hypoxia in the Gulf of Mexico: Land and Sea Interactions. Council for
Agricultural Science and Technology. 53pp.
Clark, W.M. 1960. Oxidation-Reduction Potentials of Organic Systems. Williams and Wilkins,
Baltimore, MD.
Coates, J.D., E.E. Ridden, D.R. Lovley, D.J. Ellis, E.L. Blunt-Harris, and C.V. Gaw. 1998.
Recovery of humic-reducing bacteria from a diversity of environments. Applied and
Environmental Microbiology 64:1504-1509.
Coleman, M.L., D.B. Hedrick, and D.R Lovley. 1993. Reduction of Fe(III) in sediments by
sulfate-reducing bacteria. Nature 361:436-438.
Conant, J.B. 1926. The electrochemical formulation of the irreversible reduction and oxidation
of organic compounds. Chemical Review 3:1-40.
259
-------
Connell, W.E., and W.H. Patrick, Jr. 1968. Sulfate reduction in soil: Effect of redox potential
andpH. Science 159:86-87.
Costerton, J.W., and P.S. Phillip. 2001. Battling biofilms. Scientific American 285:3-9.
Councell, T.B., E.R. Landa, andD.R. Lovley. 1997. Microbial reduction of iodate. Water, Air,
and Soil Pollution 100:99-106.
Craig, PJ. 1986. Organomercury compounds in the environment, In Organometallic
Compounds in the Environment: Principles and Reactions. P. J. Craig (ed.). John Wiley
& Sons, New York, pp. 65-101.
Delaune, R.D., W.H.Patrick, Jr., and T. Guo. 1998. The redox -pH chemistry of chromium in
water and sediment. In: Metals in Surface Waters. H. E. Allen, A.W.Garrison, and
G.W. Luther III (eds.). Ann Arbor Press, Chelsea, Michigan, pp. 241-255.
Diaz, R.J., and R. Rosenberg. 1995. Marine benthic hypoxia - A review of its ecological effects
and the behavioral responses of benthic macro fauna. Oceanography and Marine Biology
33:245-303.
Diaz, R.J. 2001. Overview of hypoxia around the world. Journal of Environmental Quality
30:275-281.
Drossman, H., H. Johnson, and T. Mill. 1988. Structure activity relationships for environmental
processes. 1. Hydrolysis of esters and carbamates. Chemosphere 17:1509-1530.
ECOTOX Data Base Systems. 2002. National Health and Environmental Effects Laboratory,
Mid-Continent Ecology Division, U.S. Environmental Protection Agency.
http ://www. epa.gov/ecotox/.
Engle, V.D., J.K. Summers, and J.M. Macauley. 1999. Dissolved oxygen conditions in northern
Gulf of Mexico estuaries. Environmental Monitoring and Assessment 57:1-20.
Fahnenstiel, G.L., M.J. Mccormick, G.ALang., D.G. Redalje, S.S Lohrenz, M. Markowitz, B.
Wagoner, and H.J. Carrick. 1995. Taxon-specific growth and loss rates for dominant
phytoplankton populations from the Northern Gulf-of-Mexico. Marine Ecology-Progress
Series 117:229-239.
Felmy, A.R., D.C. Girvin, and E.A. Jenne. 1984. MINTEQ - A computer program for
calculating aqueous geochemical equilibria. U.S. Environmental Protection Agency,
Athens, GA, USA, EPA/6003-84-032.
Felmy, A.R. 1995. GMIN, A computerized chemical equilibrium program using a strained
minimization of the Gibbs free energy: Summary report. In: Chemical Equilibrium and
Reaction Models. R. H. Loeppert, A.P. Schwab, and S. Goldberg, (eds.). Soil Science
Society of America, Madison, Wisconsin. Special Publication Number 42, pp. 377-407.
260
-------
Fredrickson, J.K., and T.C. Onsott. 2001. Biogeochemical and geological significance of
subsurface microbiology. In: Subsurface Microbiology and Biogeochemistry. J. K.
Fredrickson, and M. Fletcher (eds.). John Wiley & Son, New York. pp. 3-38.
Gan, H., J.W. Stucki, and G.W. Bailey. 1992. Reduction of structural iron in ferruginous
smectite by free radicals. Clays and Clay Minerals 40:659-665.
Gao, T.Z., R.D. Delaune, and W.H. Patrick, Jr. 1997. The influence of sediment redox chemistry
on chemically active forms of arsenic, cadmium, chromium, and zinc in estuarine
sediment. Environment International 23:305-316.
Garrett, R.H., and C.M. Grisham. 1999. Biochemistry, Second Edition, Saunders College
Publishing. New York.
Gaspard, S., F.Vazquez, and C. Holliger. 1998. Localization of the iron(III) reductase of
Geobacter sulfurreducens. Applied and Environmental Microbiology 60:3188-3194.
GESAMP (Joint Group on the Scientific Aspect of Marine Pollution). 1990. The State of the
Marine Environment, UNEP Regional Seas Reports and Studies No. 115, UNEP.
Ghirose, W.C. 1984. Biology of iron-and manganese-depositing bacteria. Annual Reviews of
Microbiology 38:515-550.
Glasauer, S., S. Langley, and T.J. Beveridge. 2001. Precipitation of an intracellular Fe mineral
during reduction of ferrihydrite by Shewanellaputrefaciens CN 32. Eleventh Annual
V.M. Goldschmidt Conference, May 20-24, 2001, Hot Springs, VA.
Gottschalk G., and M. Blaut. 1990. Generation of proton and sodium motive forces in
methanogenic bacteria. Biochimica Biophysica Acta 1018:263-266.
Green, S.A., D.J. Simpson, G. Zhou, S. Ho, andN.V. Blough. 1990, Intramolecular quenching
of excited singlet-states by stable nitroxyl radicals. Journal of the American Chemical
Society 112:7337-7346.
Gounot, A.M. 1994. Microbial oxidation and reduction of manganese - Consequences in
groundwater and applications. FEMS Microbiology Reviews 14:339-349.
Guenther, W.B. 1975. Chemical Equilibrium. Plenum Press, New York.
Guillon, E., P. Merdy, M. Aplincourt, J. Dumonceau, and H. Vezin. 2001. Structural
characterization and iron(III) binding ability of dimeric and polymeric lignin models.
Journal of Colloid and Interface Science 239:39-48.
261
-------
Hass, J.R., N. Morgan, and TJ. Dichristina. 2001. Influence of organic chelation on
dissimilatory Fe(III) reduction of Shewanellaputrefaciens, Geological Society Annual
Meetings November 5-8.
Harrison, Jr., A. 1984. The acidophilic, thiobacillic, and acidophilic bacteria that share their
habitat. Annual Reviews of Microbiology 38:2865-2912.
Heijman, C.G., E. Grieder, C. Holliger, and R. Schwartzenbach. 1995. Reduction of
nitroaromatic compounds coupled to microbial iron reduction in laboratory aquifer
columns. Environmental Science & Technology29:775-783.
Henry, M.G., and GJ. Atchison. 1991. Metal effects on fish behavior - Advances in
determining the ecological significance of responses. In: Meal Ecotoxicolgy Concepts
and Applications. M.C. Newman and A.W. Mclntosh (eds.). Lewis Publishers, Chelsea,
Michigan, pp 131-144.
Hersman, L., T. Lloyd, and G. Sposito. 1995. Siderophores promoted dissolution of hematite.
Geochimica et Cosmochimica Acta 59:3327-3330.
Hewitt, L.F. 1950. Oxidation-reduction Potentials in Bacteriology and Biochemistry. Edinburgh,
United Kingdom.
Howarth, R.W. 2001. Hypoxia, fertilizer and the Gulf of Mexico. Science 292:1485-1485.
Jennette, K.W. 1981. The role of metals in carcinogenesis: Biochemistry and metabolism.
Environmental Health Perspectives 40:233-252.
Jin, X., G.W. Bailey, Y.S. Yu, and A.T.Lynch. 1996. Kinetics of single and multiple metal ion
sorption processes on humic substances. Soil Science 161:509-520.
Jugsujinda, A., and W.H. Patrick, Jr. 1995. Methane and water soluble iron precipitation under
controlled soil pH, and redox conditions. Communications in Soil Science and Plant
Analysis 27:2221-2227.
Justic, D., and N.N. Rabalais. 1993. Seasonal coupling between riverborne nutrients, net
productivity, and hypoxia. Marine Pollution Bulletin 26:184-189.
Justic, D., and N.N. Rabalais. 1995. Changes in nutrient structure of river-dominated coastal
waters - stoichiometric nutrient balance and its consequences. Estuarine Coastal and
Shelf Science 40:339-356.
Justic, D., and N.N. Rabalais. 1996. Effects of climate-change on hypoxia in coastal waters — A
doubled CC>2 scenario for the Northern Gulf of Mexico. Limnology and Oceanography
41:992-1003.
262
-------
Justic, D., N.N. Rabalais, and R.E. Turner. 1997. Impacts of climate-change on net productivity
of coastal waters - implications for carbon budgets and hypoxia. Climate Research
8:225-237.
Karthikeyan, K.G., H.A. Elliott, and F.S.Cannon. 1997. Adsorption and coprecipitation of
copper with the hydrous oxides of iron and aluminum Environmental Science &
Technology 31:2721-2725.
Kostka, I.E., and K.H. Nealson. 1995. Dissolution and reduction of magnetite by bacteria.
Environmental. Science & Technology 29:2535-2540.
Kostka, I.E., J.W. Stucki, and J. Wu. 1996. Reduction of structural Fe(III) in smectite by a pure
culture of Shewanellaputrefaciens strain MR-1: Clays and Clay Minerals 44:522-529.
Langley, S., and TJ. Beveridge. 1999. Metal binding by Psuedomonas aeruginosa PAO1 as
influenced by growth of the cells as a film. Canadian Journal of Microbiology 45:616-
622.
Langmuir, D. 1997. Aqueous Environmental Geochemistry. Prentice Hall, Upper Saddle River,
New Jersey.
Larsen, I, B. Little, K.H. Nealson, R. Ray, A. Stone, and J. Tian. 1998. Manganite reduction by
Shewanellaputrefaciens MR-4. American Mineralogist 83:1564-1572.
Legovic, T. 1997. When do phytoplankton blooms cause the most intense hypoxia in the
northern Adriatic Sea? Oceanologica Actra 20:91-97.
Lin, Y, G.W. Bailey, and A.T. Lynch. 1989. Metal affinities in single and multiple ion
adsorption reactions at sulfide mineral surfaces. Environmental Science (China) 1:15-
30.
Lin, Y, G.W. Bailey, and A.T. Lynch. 1990. Lanthanum adsorption and complexation at sulfide
mineral surfaces. In: Environmental Life Elements and Health. T. Jianan, P.J. Peterson,
L. Ribang, and W. Wuyi (eds.). Science Press. Beijing, pp. 105-115.
Lindsay, W.L. 1979. Chemical Equilibria in Soils. John Wiley & Sons, New York.
Lindsay, W.L. 1988. Solubility and redox equilibria of iron compounds in soils: In: Iron in Soil
and Clay Minerals, J.W.Stucki, B.A. Goodman, and U. Schwertmann (eds.). D. Reidel
Publishing Company, Dordrecht, pp. 37-62.
Liu, Z. 1985. Oxidation-reduction potential. In: Physical Chemistry of Paddy Soils, T. Yu (ed).
Science Press. Beijing, pp. 1-26.
263
-------
Lloyd, J.R., E.A. Blunt-Harris, and D.R. Lovley. 1999. The perimasmic 9.6 kilodalton c-type of
Geobacter sulfureducens is not an electron shuttle to Fe(III). Journal of Bacteriology
181:7647-7649.
Lloyd, J.R., J. Chesnes, S. Glasauer, DJ. Bunker, F.R. Livens, and D.R. Lovley. 2002.
Reduction of actinides and fission products by Fe(III)-reducing bacteria.
Geomicrobiology Journal 19:103-120.
Loeppert, R.H., A.P. Schwab, and S. Goldberg (eds.). 1995. Chemical Equilibrium and Reaction
Models, Madison, Wisconsin. Soil Science Society of America Special Publication
Number 42. 422pp.
Lohrenz, S.E., G.L Fahnenstiel, D.G. Redalje, G.A. Lang, X.G. Chen, and M. J. Dagg. 1997.
Variations in primary production of Northern Gulf-of-Mexico continental-shelf waters
linked to nutrient inputs from the Mississippi River. Marine Ecology-Progress Series
155:45-54.
Lovley, D.R. 1991. Dissimilatory Fe(III) and Mn(IV) reduction. Microbiology Reviews 55:259-
287.
Lovley, D.R. 1993. Dissimilatory metal reduction. Annual Reviews of Microbiology 47:263-
290.
Lovley, D.R. 1997. Microbial Fe(III) reduction in subsurface environments: Federation of
European Microbiology Societies Reviews 20:305-313.
Lovley, D.R., and M.J. Klug. 1986. Model for the distribution of sulfate reduction and
methanogenesis in freshwater sediments. Geochimica et Cosmochimica Acta 50:11-18.
Lovley, D.R., and S. Goodwin. 1988. Hydrogen concentration as an indicator of the
predominant terminal electron-accepting reactions in aquatic sediments. Geochimica et
Cosmochimica Acta 52:2993-3003.
Lovley, D.R., and E.A. Phillips. 1988. Manganese inhibition of microbial iron reduction in
anaerobic sediments. Geomicrobiology Journal 6:145-155.
Lovley, D.R., F. H Chapelle, and E.A. Phillips. 1990. Fe(III)-reducing bacteria in deeply buried
sediments of the Atlantic Coastal Plain. Geology 18:957-957.
Lovley, D.R., E.J. Phillips, Y.A Gorby, and E.R. Landa. 1991. Microbial reduction of uranium.
Nature 350:413-416.
Lovley, D.R., and E.J. Phillips. 1992. Reduction of uranium by Desulfovibrio desulfuricam.
Applied and Environmental Microbiology 58:850-858.
264
-------
Lovley, D.R., J.D. Coates, D.A. Saffarini, andDJ. Lonergan. 1997. Dissimilatory iron reduction.
In: Iron and Related Transition Metals in Metals in Microbial Metabolism, G. Winkleman
and CJ. Carano (eds.). Horwood, Switzerland, pp!87-215.
Lovley, D.R., EJ. Phillips, J.D. Coates, J.L. Frage, E.L Blunt-Harris, and L.A. Hayes. 2000.
Humic substances as mediator for microbially catalyzed metal reduction. Acta
Hydrochimica et Hydrobiologica 26:152-157.
Lovley, D.R., J.M. Tor, E.L Blunt-Harris, K. Kaskell, and M.Vargas. 2000. Reduction of humic
substances and Fe(III) by hyperthermophilic microorganisms. Chemical Geology
169:289-298.
Lower, S.K., M.F. Hochella, and TJ. Beveridge. 2001 Bacterial recognition of mineral
surfaces: Nanoscale interactions between Shewanella and alpha-FeOOH. Science
292:1360-1363.
Ma, Q.Y., and W.L.Lindsay. 1993. Measurements of free zinc2+ activity in uncontaminated and
contaminated soils using chelation. Soil Science Society of America Journal 57:963-967.
Macy, J,M., J.M. Santini, B.V. Pauling, A.H. ONeill, and L.I. Sly. 2000. Two new
arsenate/sulfate-reducing bacteria: mechanisms of arsenate reduction. Archives Of
Microbiology 173:49-57.
Madigan, M.T,, J.M. Martinko, and J. Parker. 1997. Brock Biology of Microorganisms. Eighth
Edition. Prentice Hall, Upper Saddle River, New Jersey.
Masschelyn, H., R.D. Deluane, and W.H. Patrick Jr. 1990. Transformation of selenium as
affected by sediment oxidation-reduction potential and pH. Environmental Science &
Technology 4:91-96.
Masschelyn, H., R.D. Deluane, and W.H. Patrick, Jr. 1991a. Effect of redox potential and pH on
arsenic speciation and solubility in a contaminated soil. Environmental Science &
Technology 25:1414-1419.
Masschelyn, H., R.D. Deluane, and W.H. Patrick, Jr. 1991b. Arsenic and selenium chemistry as
affected by sediment redox potential and pH.. Journal of Environmental Chemistry
20:522-527.
Masschelyn, H., J.J. Pardue, R.D. Deluane, and W.H. Patrick, Jr. 1992. Chromium redox
chemistry in a lower Mississippi Valley bottom and hardwood wetland. Environmental
Science & Technology 26:1217-1226.
Mattigod, S.V., and G. Sposito. 1979. Chemical modeling of trace metal equilibria in
contaminated soil solution using the computer, GEOCHEM. In: Chemical Modeling in
Aqueous Systems, E. A Jenne (ed.). American Chemical Society Symposium Series 92
American Chemical Society, Washington, D.C. pp. 837-855.
265
-------
McLean, J.S., TJ. Beveridge, and D. Phipps. 2000. Isolation and characterization of a
chromium-reducing bacterium from a chromated copper arsenate-contaminated site.
Environmental Microbiology 2:611-619.
Millero FJ. 1998 Solubility of Fe(III) in seawater. Earth Planet Science Letters 154:323-329.
Modig, H. 1998. Responses of Baltic benthic invertebrates to hypoxia events. Journal of
Experimental and Marine Biology 229:133-148.
Mortimer, R.J.G., and I.E. Rae. 2000. Metal speciation (Cu, Zn, Pb. Cd.) and organic matter in
oxic to suboxic salt marsh sediment, Severn Estuary, Southwest Britain. Marine
Pollution Bulletin 40:377-385.
Moss, A.J., G.E. Rayment, N. Reilly, and E.K. Best. 1992. A preliminary assessment of
sediment and nutrient export from Queensland coastal catchment, Report of Queensland,
Department of Environment and Heritage and Department of Primary Industry, Brisbane,
Australia. 33pp..
Myers, C.R., and J.M. Myers. 1997. Cloning and sequencing of cyma, a gene encoding
tetraheme cytochrome c required for reduction of Fe(III), fumarate, and citrate by
Shewanellaputrefaciens MR-1. Journal of Bacteriology 179:1143-1152.
Myers, C.R,.and J.M. Myers. 2000. Role of the tetraheme membrane cytochrome Cym A in
anaerobic transport in cells of Shewanella putrefaciens MR-1 with normal levels of
manaquinone. Journal of Bacteriology 182:67-75.
Myneni, S.C.B., T.K. Tokunaga, and G.E. Brown, Jr. 2001. Abiotic selenium redox
transformation in the presence of Fe(II,III) oxides. Science 278:1106-1109.
Nealson, K.H., and D. Saffarini. 1994. Iron and manganese in anaerobic respiration -
environmental significance, physiology, and regulation. Annual Review of Microbiology
48:311-343.
Nealson, K.H., and D.A. Stahl. 1997. Microorganisms and biogeochemical cycles: What can we
learn from layered microbial communities? In: Geomicrobiology Interactions Between
Microbes and Minerals, J. F Banfield, and K.H. Nealson (eds.). Mineralogical Society of
America, Washington D.C. pp. 5-34.
Newman, D.K., and R. Kolter. 2000. A role for excreted quinones in extracellular electron
transfer. Nature 405:94-97.
Nivinskas, H., A.F. Koder, R.L. Anusevicius, Z. Sarlauskas, J. Miller, and N. Cenas. 2000.
Two-electron reduction of nitroaromatic compounds by Enterobacter cloacae NAD(P)H
nitroreductase: Description of quantitative structure-activity relationships Acta
Biochimica Polonica 47:941-949.
266
-------
Nordberg, K. 1998. Oceanographic conditions in the deepest part of the Kattegat, Scandinavia
revealed through recent benthic foraminifera and hydrography. Estuarine Coastal and
Shelf Science 49:557-576.
Paerl, H.W. 1998. Ecosystem response to internal and watershed organic matter loading:
Consequences for hypoxia in the eutrophying Neuse river estuary, North Carolina.
Marine Ecology-Progress Series 166:17-25.
Parker, D.R., W.A. Norvell, and R.L.Chaney. 1995. GEOCHEM - PC - A chemical speciation
program for IBM and compatible personal computers. In: Chemical Equilibrium and
Reaction Models,: R.H. Loeppert, A.P Schwab, and S. Goldberg (eds.). Soil Science
Society of America. Madison, Wisconsin, Special Publication Number 42 pp. 253-270.
Patrick, W.H, Jr., and R.E. Henderson. 1981. A method for controlling redox potential in packed
soil cores. Soil Science Society of America Journal 45:35-37.
Patrick, W.H., Jr., and J.A. Jugsujinda. 1992. Sequential reduction and oxidation of inorganic
nitrogen, manganese, and iron in flooded soils. Soil Science Society of America Journal
56:1071-1073.
Patrick, W.H., Jr., and M. Verloo. 1998. Distribution of soluble heavy metals between ionic and
complexed forms in a saturated system as affected by pH and redox conditions. Water
Science and Technology 37:165-171.
Pearson, T.H. 1992. Energy flow through the Sattegat - A comparative examination of the
euthrophication of coastal marine ecosystems. Netherlands Journal of Sea Research
28:317-334.
Peters,V., and R. Conrad. 1996. Sequential reduction processes and initiation of CH4 production
upon flooding of oxic upland soils. Soil Biology and Biochemistry 28:371-382.
Ponnamperuma, N.N. 1972. The chemistry of submerged soils: Advances in Agronomy.
Academic Press, New York. Volume 24: 29-96.
Ponnamperuma, N.N., E.M. Tench,.and T. Lot. 1967. Redox equilibria in flooded soils. I. The
iron hydroxide system: Soil Science 103:374-382.
Rabalais, N.N. 1998. Consequences of the 1993 Mississippi River flood in the Gulf of Mexico.
Regulated Rivers-Research & Management 14:161-177'.
Rabalais N.N., W.J.Wiseman, R.E.Turner, B.K. Sen Gupta, and Q. Dortch. 1996. Nutrient
changes in the Mississippi River and system responses on the adjacent continental shelf.
Estuaries 19:386-407.
Rabalais, N.N., .R.E.Turner, and W.J. Wiseman. 2001. Hypoxia in the Gulf of Mexico. Journal
of Environmental Quality 30:320-329.
267
-------
Risser, J.A., and G.W. Bailey. 1992. Spectroscopic study of surface redox reactions with
manganese oxides. Soil Science 56:82-88.
Risser, J.A., and G.W. Bailey. 1997. Spectroscopic study of soil particle surface redox reactions.
Bulletin of Environmental Contamination and Toxicology 58:356-363.
Rusin, P.A, B.J.R. Strietelmeier, B.A. Tait, E.S.A. Palmer, P.S. Newton, andD.L. Clark. 1994.
Solubilization of plutonium hydrous oxide by iron-reducing bacteria. Environmental
Science & Technology 28:1686-1690.
Santore, R.C., and C.T. Driscoll. 1995. The CHESS Model for calculating chemical equilibria in
soils and solutions. In: Chemical Equilibrium and Reaction Models, R.H. Loeppert, A.P
Schwab, and S. Goldberg (eds). Soil Science Society of America. Madison, Wisconsin,
Special Publication Number 42, pp. 357-376.
Schecher, W.D, and C.T. Driscoll. 1995. ALCHEMI: A chemical equilibrium model to assess
the acid-base chemistry and speciation of aluminum in dilute solutions. /w:Chemical
Equilibrium and Reaction Models, R.H. Loeppert, A.P Schwab, and S. Goldberg (eds.).
Soil Science Society of America. Madison, Wisconsin, Special Publication Number, 42
pp. 325-356.
Schwertmann, U. 1991. Solubility and dissolution of iron-oxides. Plants and Soilsl30:l-25.
Schwertmann, U., andR.M. Taylor. 1989. Iron oxide: J.B. Dixon, and S.B. Weed (eds.)
Minerals in Soil Environments. Soil Science Society of America. Madison, Wisconsin.
pp. 379-438.
Schwertmann, U., and R.W. Fitzpatrick. 1992. Iron minerals in surface environments. In:
Biomineralization Processes of Iron and Manganese, H.C.W. Skinner, and R.W.
Fitzpatrick, (eds.), Catan, Cremlingen pp. 7-31.
Seeliger, S., R. Cord-Ruwisch, and B. Schink. 1998. A periplasmic and extracellular c-type
cytochrome Geobacter sulfurreducens acts a ferric ion reductase and as an electron
carrier to other acceptors or to partner bacteria. Journal of Bacteriology 180:3686-3691.
Siciliano, S.D., and D.R.S. Lean. 2002. Methyltransferase: An enzyme assay for microbial
methylmercury formation in acidic soils and sediments Environmental Toxicology and
Chemistry 21:1184-1190
Simpson, S.L., J. Rosner, and J. Ellis. 2000. Competitive displacement reactions of cadmium,
copper, zinc, and zinc added to a polluted, sulfidic, estuarine sediment. Environmental
Toxicology and Chemistry 19:1992-1999.
268
-------
Smith, G.R., K.K. Tanji, R.G. Burau, and JJ. Jurinak. 1995. C-SALT - A chemical equilibrium
model for multicomponent solutions. In: Chemical Equilibrium and Reaction Models,
R.H. Loeppert, A.P. Schwab, and S. Goldberg (eds.). Soil Science Society of America.
Madison, Wisconsin. Special Publication Number 42, pp. 289-324.
Sposito, G., and J. Cove. 1995. SOILCHEM on the Macintosh. In: Chemical Equilibrium and
Reaction Models, R.H. Loeppert, A.P. Schwab, and S. Goldberg (eds.). Soil Science
Society of America. Madison, Wisconsin, Special Publication Number 42, pp. 271-288.
Sposito, G., A. Yang, R.H. Neal, and A. Mackzum. 1991. Selenate reduction in alluvial soil.
Soil Science Society of America Journal 55:1597-1602.
Starkey, R.L., and H.O. Halverson. 1927. Studies on the transformation of iron in nature. II.
Concerning the importance of microorganisms in the solution and precipitation of iron.
Soil Science 24:381-402.
Stevenson, FJ. 1994. Humus Chemistry - Genesis, Composition, Reaction. Second Edition,
John Wiley & Sons, New York.
Straub, K.L., M. Benz, B. Schink, and G. Waddell. 1996. Anaerobic nitrate-dependent
microbial oxidation of ferrous iron. In: Iron in Soils and Clay Minerals. J.W. Stucki,
B.A. Goodman, and U. Schwertmann (eds.). D.Reidel Publishing Company, Dordrecht,
pp. 625-675.
Street, J.J, B.R.Sabey, and W.L.Lindsay. 1978. Influence of pH, phosphorus, cadmium, sewage
sludge, and incubation-time on solubility and plant uptake of cadmium. Journal of
Environmental Quality 7:286-290.
Stryer, L. 1995. Biochemistry. 4th Edition. W.H. Freeman, New York.
Stucki, J.W. 1988. Structural iron in smectites. In: Iron in Soils and Clay Minerals, J.W. Stucki,
B. A. Goodman and U. Schwertmann (eds.). D. Reidel Publishing Company, Dordrecht,
pp. 625-675.
Stucki, J.W., G.W.Bailey, and H. Gan. 1996. Oxidation-reduction mechanisms in iron-bearing
phyllosilicates. Applied Clay Science 10:417-430.
Stucki, J.W., J.B. Komadel, and H.T.Wilkinson. 1987. Microbial reduction of structural iron
(III) in smectite: Soil Science Society of America Journal 51:1663-1665.
269
-------
Stumm, W., and JJ. Morgan. 1985. The conceptual significance of pe. A comment on J. D.
Hostettler's paper, Electrode electrons, aqueous electrons and redox potentials in natural
waters. American Journal of Science 285:856-859.
Stumm, W., and JJ. Morgan. 1996. Aquatic Chemistry: Chemical Equilibrium and Rates in
Natural Waters. John Wiley & Sons, New York, pp. 425-515.
Synder, C.S. 2001. Hypoxia, fertilizer, and the Gulf of Mexico. Science 292:1485.
Thamdrup, B., D.E. Canfield, T.G. Ferdelman, R.N. Gluf, and S.J.K. Gunderdrn. 1996. A
biogeochemical survey of the anoxic basin, Golfo Dulce, Costa Rica. Reviews Biology of
the Tropics 44 (supplement 3): 19-33.
Tratnyek, P.G. 1995. Contaminant remediation with zero-valent metals - the utility of corrosion
in a complex geochemical environment: Abstracts of Papers of the American Chemical
Society 210:119.
Trefry, J.H., S. Metz, T.A.Nelsen, R.Trocine, andB.J. Eadie. 1994. Transport of particulate
organic-carbon by the Mississippi River and its fate in the Gulf-of-Mexico. Estuaries
17:839-849.
Truesdell, A.H., and B.E. Jones. 1974. WATEQ, a computer program for calculating chemical
equilibria in natural water. U.S. Geological Survey Journal of Research 2:321-326.
Turner, F.T., and W.H. Patrick, Jr. 1968. Chemical changes in waterlogged soils as a result of
oxygen depletion. Soil Science 4:53-65.
Turner, R.E., and N.N. Rabalais. 1991. Changes in Mississippi River water-quality this century.
Bioscience 41:40-147.
UNEP (United Nations Environmental Program). 1991. The state of the world environment.
Nairobi, 49pp.
Wadden, F., S. Schnell, S. Heising, A. Ehrenreich, B. Assmus, and B. Schink. 1993. Ferrous iron
oxidation by anoxygenic phototrophic bacteria. Nature 362:834-836.
Walker, S.G., C.A. Flemming, F.G.Ferris, T.J. Beveridge, and G.W. Bailey. 1989.
Physicochemical interaction of Escherichia coli cell envelopes and Bacillus subtilis cell
walls with two clays and ability of the composite to immobilize heavy metals from
solution. Applied and Environmental Microbiology 55:2976-2984.
Warren, L.A., and F.G. Ferris. 1998. Continuum between sorption and precipitation of Fe(III) on
microbial surfaces. Environmental Science & Technology 32:2331-2337.
270
-------
Westall, J.C., J.M. Zachara, and F.M.M. Morel. 1976. MINEQL; A computer program for
calculation of chemical equilibrium composition of aqueous systems.Technical Note 18.
Department of Civil Engineering, Massachusetts Institute of Technology, Cambridge,
Massachusetts.
White, A.F., and M.L. Peterson. 1996. Reduction of aqueous transition metal species on the
surfaces of Fe(II)-containing oxides. Geochimica et Cosmochimica Acta 60:3799-3814.
Winchell, A.N., and H. Winchell. 1951. Elements of Optical Mineralogy. An Introduction to
Microscopic Petrography. Part II: Description of Minerals. John Wiley, Fourth Edition,
New York, pp. 183-237.
Winstanley, D. 2001. Hypoxia, fertilizer and the Gulf of Mexico. Science 292:1486.
Witter, A.E., D.A. Hutchins, A. Butler, and G.W. Luther (III). 2000. Determination of
conditional stability constants and kinetic constants for strong model Fe-binding ligands
in seawater. Marine Chemistry 69:1-17.
Wu, R.S.S. 2000. Hypoxia in coastal waters: Pressing problems worldwide and their scientific
challenges. In: Proceedings of the Fifth International Symposium, on Fish Physiology,
Toxicology and Water Quality, Hong Kong, PRC. November 10-13 , 1998, R. V.
Thurston (ed.). U.S. Environmental Protection Agency Report/600/R-00/015, April 2000,
Ecosystems Research Division Athens, Georgia, USA. pp. 203-214.
Wu, J., C.B. Roth, and P.F. Low. 1988, Biological reduction of structural Fe in sodium-
nontronite. Soil Science Society of America Journal. 52:295-296.
Yiwei, D., and W. Stumm. 1994. Reactivity of aquatic iron (III) oxyhydroxides - Implications
for redox cycling of iron in natural waters. Applied Geochemistry 9:23-36.
Yu, Y.S., and G.W. Bailey. 1992. Reduction of nitrobenzene by four sulfide minerals: kinetics,
products and solubility. Journal of Environmental Quality. 21:1-8.
Yu, Y.S., and G.W. Bailey. 1996. Redox characterization of the surfaces of seven iron-bearing
minerals: Use of molecular probes and UV-visible spectroscopy: Clays and Clay
Minerals 44:573-586.
Zachara, J.M., J.F. Fredrickson, S. Li, D.W. Kennedy, S.C. Smith, andL. Gasman. 1998.
Bacterial reduction of crystalline Fe3+ oxides in single phase suspensions and subsurface
materials. American Mineralogist 83:1426-1443.
Zepp, R.G., B.C. Faust, and J. Hoigne. 1992. Hydroxyl radical formation in aqueous reactions
(pH 3-8) of iron(II) with hydrogen-peroxide - the photo-Fenton reaction. Environmental
Science & Technology. 26:313-319.
271
-------
Zhou, X.L., and K. Mopper. 1990. Determination of photochemically produced hydroxyl
radicals in seawater and fresh-water. Marine Chemistry 30:71-88.
Zobell, C.E. 1946. Studies on redox potential of marine sediments: American Association of
Petroleum Geologists Bulletin. 30: 477-513.
272
-------
THE DIFFERENT FACES OF ANOXIA IN THE BALTIC SEA
Heye Rumohr1
ABSTRACT
Long-term image monitoring (14 years) of the Baltic Sea with video, stills, and sediment
profile imaging (SPI) has revealed various facets of anoxia at the bottom of the Sea, many of
which are accompanied by wide-spread mats of Beggiatoa. Some features of this anoxic
environment cannot be retrieved with traditional (destructive) sampling because of their fragile
and delicate nature. Stages of organic enrichment, with fluffy flakes of growing size rolling over
the sediment, have been shown to result in a solid mat fixed by Beggiatoa. It is hypothesized
that circular structures around decaying organic sources (bivalves) are the subtidal equivalents to
the often observed black spots in the Wadden Sea. Milky coloring of the water column
coincided with a layer of hydrogen sulphide, 140 meters thick, in the open water of the Baltic
Sea proper, indicating sulfuric particles, possibly originating from sulfur oxidising bacteria in the
free water above the Sea bottom. By means of video and still photography, vagile predatory
polychaetes (Harmothoe spp.) living on the Beggiatoa mats after an inflow of oxygen-rich water
into the Baltic Sea have been found dead after the end of the normoxic period. SPI images from
all stations revealed the vertical structure of redox-mediated colour changes in the sediment
leading to stratification and lamination. Biological samples from the same stations showed
distinct changes in community patterns accompanying changes in the oxygen regime.
INTRODUCTION
Eutrophication of coastal waters in northern Europe led to increased interest in historical
oxygen data sets and time series that might shed some light on the dynamics and possible trends
in the parameters under investigation. International monitoring programmes were launched to
document the status of the environment and its changes. In the Baltic Sea, it was the HELCOM
monitoring program that continuously provided the environmental data and their assessment and
evaluation every 5 years. Oxygen was seen as the critical factor in most of the deep areas of the
Baltic Sea since it governs the existence of life and the maturity of the system.
The Baltic Sea is, in geological terms, a very young system. About 12,000 years after the
last glaciation, it is still in an evolutionary succession. The Baltic Sea has a special topography
being a series of basins divided by sills. It is an enclosed sea with a peculiar hydrography, i.e.
there are no tides; it is strongly stratified and needs major flushing with marine water from the
North Sea to sustain its density and oxygen content. The latest major inflow events occurred in
1977 and 1993, the longest period ever recorded between flushings. Most physiographic features
of the Baltic Sea can be found in the web version of the Ph.D. thesis of Unverzagt (2001).
Institut fur Meereskunde, Kiel, Germany.
273
-------
In recent decades, oxygen levels have dropped drastically and even free H2S has occurred
in the water. Also, salinity has dropped and the pycnocline reached the bottom of the central
Baltic Sea (90m) in the 1990s. As a result, the benthic fauna of the Baltic Sea has been heavily
affected over recent decades, and we now face a general decline of all benthic faunal parameters
such as species number, biomass, and numbers of individuals (Rumohr etal. 1996, Laine etal.
1997, Olenin 1997, Powilleit and Kube 1999).
There may have been temporal flashbacks over the years, but in general we are facing a
stable negative trend. The reasons for this may be sought both in the depth profile of the Baltic
Sea combined with climatic/hydrographical settings, as well as human impacts via nutrient
loading (hypertrophication) and other contaminations. An example of the latter is the Gdansk
Deep where the Vistula River carries the sewage of almost all of Poland into the Baltic Sea.
Results from long-term image monitoring (12 years) in the Baltic Sea with video, still
photography and sediment profile imaging have revealed various facets of anoxia at the bottom
of the Baltic Sea, many of which are accompanied by wide-spread mats ofBeggiatoa. Most of
these features cannot be retrieved with traditional (destructive) sampling because of their fragile
and delicate nature.
MATERIAL AND METHODS
The image data have been collected since 1986 with various underwater video cameras
manufactured by OSPREY (now Kongsberg-Simrad). Still photographs were taken with a
special OSPREY TVP camera comprised of a video and a stills camera in one housing, using the
same len with a prism, thereby preventing any parallax errors. Sediment profile images (SPI)
have been obtained with a slightly modified BENTHOS REMOTS camera (Rhoads and
Germano 1987, 1990) that has been used since 1986 at routine monitoring stations in the Baltic
Sea (Rumohr 1995). Hydrographical measurements were made with probes and water collection
bottles.
When trying to relate "bottom water values" to sea-floor level phenomena, one can run
into trouble because they may not relate well. Normally, there is only a defined distance to the
water surface in hydrographic sampling series and not to the seabed. So one can never tell at
what real distance from the sea-bed the "bottom water values" have been measured in the
supposed strong oxygen gradient just above the sediment. We tried to overcome this
shortcoming by always measuring 0.5 m above the sediment surface.
Grab sampling (0.1 m2 Van Veen) and dredge tows of 3 minutes at 2 knots with a 1-m
wide botanical dredge (Kieler Kinderwagen) complemented this monitoring programme
(Rumohr 1999). We surveyed the Baltic Sea from the Great Belt to the Aland Sea and the
Finnish Gulf on an annual basis, with a higher temporal resolution in the Kiel Bay/western Baltic
area. This paper will concentrate on the high variability of the visual aspects of anoxic bottoms
in the Baltic, based on 14 years of sampling, which could only be observed with non-destructive
techniques, i.e. imaging methods.
274
-------
RESULTS
Images of various hypoxic/anoxic situations in the Baltic Sea revealed the high variability
of the aspects of anoxic bottoms. The prime indicator was the presence ofBeggiatoa spp. that
grows at the boundary between sulfide and oxygen. All the steps leading to Beggiatoa mat
formation could be observed from single isolated patches (Figure 1) often found around decaying
organic sources like dead bivalves (Figure 2), or from advected algae drifters (Figure 3) located
in isolated areas and as wide-ranging, white covers (Figure 4a, b). The circular structures around
decaying organic sources (bivalves) are hypothized as the subtidal equivalents of the often
observed black spots in the Wadden Sea (Figure 5). In areas where we found free hydrogen
sulfide in the water, the bottom was covered with black flakes (Figure 6a, b) with no signs of
Beggiatoa. In the same water mass, we found a milky-white substance that we assumed to be
sulfur particles, possibly originating from sulfur oxidising bacteria in the free water above the
bottom. The depth distribution of the milky layer was in absolute concordance with the observed
H^S. Trawling removed the oxic sediment surface layers and left the sulfidic sediment
depressions filled with Beggiatoa "enamel" covers (Figure 7). Beggiatoa mats can also be a
summer feature of shallow waters, when drifting algae decay in shallow depressions and are
rapidly covered by Beggiatoa.
Figure 1. First isolated patches ofBeggiatoa spp. on suboxic sediments in the Arkona
Basin (48m).
275
-------
*
Figure 2. Beggiatoa spp. patches around large dead bivalves (Arctica islandicd) in the Arkona
Basin (46m).
Figure 3. Advected red algae drifters in the Arkona Basin (46m).
276
-------
*z
~'" , ,» 4, % m J •< «. '-j* • •»--Sv^ ":> jS*^
*" **- - - :- -" "*• --'''*
.« i
. *n '* ' * *
A*-.' J"
w- *. -^ •
^^^mmmsf&&^&v^
i-ain iiinF.. .*-.-=* JWT Jk, . •
-"*iji^
r-.
atH* "*• %
' vi/-'
' ' ^"V V-- -**
fe& v?:
SJjSfr '^- ^; ^S^$§fc'£v'v f^tk%
S"i!-?H **i'»y »*!. "- ''k ">\ j ,•! 'f~- •', *••«., ••i *^»A » «r* **JL*%!-*,J|>II:
A.'-'**™*'!* V* "fc^ \\."'° >,-*"*-
tJ"iKTP«»4L**.» 7 »* 3w,* ' * ,- •!.""* V *" * » • «•• •! ™ JE~il»^ w-i
^^*sir^ j»,?J'.%».^'»» ,=* -• ^*
1*]
Figure 4a
Figure 4b
Figure 4a, b. Solid cover with Beggiatoa spp.: (a) Gdansk Deep, 110m; (b) slope of Landsort
Deep 200m.
277
-------
Figure 5. Rings ofBeggiatoa around decaying organic matter (bivalves) in the Arkona Basin
(46m).
Figure 6a
278
-------
Figure 6b
Figure 6. Sediment profile images of black, sulfidic flakes on anoxic sediments beneath
containing water from Landsort Deep (a) at 380m, and (b) at 460m.
279
-------
Figure 7. Bottom trawl tracks "enameled" with Beggiatoa spp., where oxidized surface layer
was removed and sulfidic layer exposed.
DISCUSSION
Anoxic, "white", "infected" grounds were first reported by fishermen in the early
twentieth century from their fishing grounds in the Kiel Bay area (Rumohr 1986). They were, in
those days, mostly caused by cases thick layers of drifting algae that sedimented in shallow
depressions of the seafloor and started rotting there, causing local anoxia and a white cover of
Beggiatoa (Ibid). The same can be observed every summer, even in very shallow (1-3 m) areas
when algae lose parts of their thalli. Recent papers from the Aland Islands (Finland) discovered
this phenomenon not only as a threat to the oxygen conditions (Norkko and Bonsdorff 1996) but
also as a temporary habitat and means for dispersion of benthic invertebrates (Norkko et al.
2000).
Since the Baltic Sea is free of tidal currents, these fragile indicators stay in place on the
sediment surface as long as the chemical situation favors them. In tidal waters the Beggiatoa
mats are often swept away and form a considerable transport mechanism of sulfur in the benthic-
pelagic system (Grant and Bathmann 1987).
280
-------
The main cause for anoxia found in the deep basins is related to the long-term
eutrophi cation of the Baltic Sea in combination with its isolation from the world's oceans that
has turned the originally oligotrophic Baltic Sea into a sea with a trend toward that of the Black
Sea. Some basins, like the Central Eastern Gotland Basin, have been anoxic and void of
macrofauna for more than 70 years. The observed recolonization with vagile epifauna
(Harmothoe sp.) in 1995 was just a short episode that ended in the autumn of the same year, as
documented by hydrography and by the images of dead worms lying on the black sulfidic
sediment. (Rumohr, unpublished data.)
The analysis of data for years with oxygen deficiency has revealed a characteristic pattern
common for all "bad" years. We see that oxygen depletion in the autumn that is below the mean
is normally coupled with one, and sometimes two, depressions in the spring. The latter occur
most frequently in May and occasionally also in February. These depressions may be shifted for
some weeks according to the actual hydrographic conditions. In extremely bad years, like 1967,
1969, 1975, 1979-1981, and 1984, we consistently found an oxygen concentration in the spring
that was well below the long term mean. The effects of these depletions are readily seen on the
sediment surface.
When considering our findings in conjunction with earlier records (see Rumohr 1986), it
may be stated, with reservations, that oxygen depletions have repeatedly occurred in certain parts
of Kiel Bay and the Baltic Sea during the last 100 years. These events, nevertheless, were
restricted mainly to the autumn of those years, but it seems now the characteristics of this
phenomenon have changed. A new phenomenon, involving more widespread and intense
oxygen depletion, has been reported for Kiel Bay in 1969-71, in 1979-81, and in 1983-85
(Weigelt and Rumohr 1986). That autumnal hypoxic period that once lasted 1-4 weeks in any
given year occurs now in a cluster of 2-3 years at a time. The regenerative capability of the
ecosystem appears to be exhausted. One possible explanation for this could be that once the
benthos is damaged, it lacks the capability to "digest" the spring bloom, which as a consequence
goes directly to the bottom (Graf et a/., 1983). The outcome of this will be increased oxygen
demand by the decaying sedimented material that, in turn, will have a future detrimental
influence on the benthic fauna. Finally, the combination of these adverse factors with the
general cyclical/seasonal negative tendency in the oxygen content of the bottom waters of the
Baltic Sea will lead to acute oxygen depletion (as shown in the images) that is only occasionally
resolved by water renewal from the North Sea.
In conclusion, some sort of a "memory" in the sea-bed may be postulated that carries
"information" from the preceding year through the following and governs the subsequent events
to a certain extent. This "information" may stem from biological events, but also from
climatological or hydrographical features. The form this memory takes should be investigated in
future research.
281
-------
REFERENCES
Graf, G., R. Schulz, R. Peinert, and L.-A. Meyer-Reil. 1983. Benthic response to sedimentation
events during autumn to spring at a shallow-water station in the western Kiel Bight.
Marine Biology 77: 235-246.
Grant, J., and U. Bathmann. 1987. Swept away: resuspension of bacterial mats regulates
benthic-pelagic exchange of sulphur. Science 236: 1472-1474.
Laine, A., H. Sandier, A.B. Andersin, and J. Stigzelius. 1997. Long-term changes in
macrozoobenthos in the eastern Gotland Basin and the Gulf of Finland (Baltic Sea) in
relation to the hydrographical regime. Journal of Sea Research 38: 135-159.
Norkko, A., and E. Bonsdorff. 1996. Population responses of coastal zoobenthos to stress
induced by drifting algal mats. Marine Ecology Progress Series 140: 141-151.
Norkko, J., E. Bonsdorff, and A. Norkko. 2000. Drifting algal mats as alternative habitat for
benthic invertebrates. Journal of Experimental Marine Biology and Ecology 248: 79-
104.
Olenin, S. 1997. Benthic zonation in the eastern Gotland Basin, Baltic Sea. Netherlands
Journal of Aquatic Ecology 30: 265-282.
Powilleit, M., and J. Kube. 1999. Effects of severe oxygen depletion on macrobenthos in the
Pommeranian Bay (Southern Baltic Sea): a case study in a shallow, sublittoral habitat
characterised by low species richness. Journal of Sea Research 42: 221-234.
Rhoads, D.C., and J.D. Germano. 1987. Interpreting long-term changes in benthic community
structure: a new protocol. Hydrobiologia 142: 291-308.
Rhoads, D.C., and J.D. Germano. 1990. The use of REMOTS imaging technology for disposal
site selection and monitoring. Geotechnical Engineering of Ocean Waste Disposal ASTM
1097, K.R. Demars and R.C. Chaney (eds.) American Society for Testing and Materials,
Philadelphia, p.55-64.
Rumohr, H. 1986. Historische Indizien fur Eutrophierungerscheinungen (1875-1939) in der
Kieler Bucht (westliche Ostsee). Meeresforschung/Reports Marine Research 31: 115-
123.
Rumohr, H. 1995. Monitoring the marine environment with imaging methods. Scientia Marina
(Suppl. 1) 59: 129-138.
Rumohr, H., E. Bonsdorff, and T.H. Pearson. 1996. Zoobenthic succession in Baltic
sedimentary habitats. Archive for Fisheries and Marine Research 44: 179- 214.
282
-------
Rumohr, H. 1999. Soft bottom macrofauna: collection, treatment, and quality assurance of
samples. ICES Techniques in Marine Environmental Sciences No. 27, p. 19.
Unverzagt, S. 2001. Raumliche und zeitliche Veranderungen der Gebiete mit Sauerstoffmangel
und Schwefelwasserstoff im Tiefenwasser der Ostsee. Greifswalder Geographische
Arbeiten, Band 19, p. 121 www.uni-greifswald.de/~geograph/gga.
Weigelt, M., and H. Rumohr. 1986. Effects of wide-range oxygen depletion on benthic fauna
and demersal fish in Kiel Bay 1981-1983. Meeresforschung/Reports Marine Research
31: 124-136.
283
-------
284
-------
HYPOXIC WATERS IN THE GULF OF CALIFORNIA: ORIGIN, DISTRIBUTION,
AND POSSIBLE CONSEQUENCES
J. Vinicio Macias-Zamora1 and Francisco Delgadillo-Hinojosa1
ABSTRACT
Hypoxic and/or anoxic conditions have been reported in several parts of the oceans
around the world. Many of these areas have shown a large impact on its oxygen content by
anthropogenic inputs. The decomposition of large amounts of organic matter discharged at sea
has severely affected some areas. Two frequently mentioned areas are the Gulf of Mexico,
apparently due to large loadings of nutrients carried into the Gulf by the Mississippi River, and
the Baltic Sea, owing its decline in oxygen content to municipal discharges along its shore.
However, the Gulf of California has also shown a well-defined oxygen minimum zone (OMZ).
Our measurements in these waters have shown nearly total oxygen depletion values. The origin
of this large volume of water with oxygen concentration below hypoxia is thought to be due to
natural rather than anthropogenic causes. The contributing water masses have traveled long
distances, and their characteristic is that they are located at a depth below which natural oxygen
consumption exceeds production.
INTRODUCTION
Hypoxia has an adverse effect on the general quality of any given body of water and on
the organisms living in the sediments involved. It can also affect the type and number of
organisms in an aquatic system, and, ultimately, it may be detrimental to humans.
The term "hypoxia" has been variously defined at different concentrations of dissolved
oxygen. For example, Leming and Stuntz (1984) have defined hypoxia as occurring at oxygen
concentrations below 2.5 mg/L. Others have defined its limit by oxygen concentrations at 2
mg/L (Rabalais et al. 1996, Zimmerman and Canuel 2000). In addition, the value of 2.5 mg/L
has also been reported as a limit below which there is no evidence of the presence of organisms
such as shrimps and finfish (Leming and Stuntz 1984). Similarly, the definition for anoxia,
although literally indicating an absence of oxygen, has been defined by some at concentrations
<0.2 mg/L (Zimmerman and Canuel 2000).
Institute de Investigaciones Oceanologicas (UABC), Km 107 carr. Tij.-Ensenada, Apdo 453 Ensenada, Baja
California.
285
-------
The causes of hypoxia and anoxia are frequently linked to anthropogenic activities
resulting in greater inputs of organic matter and/or excess of nutrients into a body of water.
Because of nutrient enrichment, eutrophication, followed by phytoplankton decay, can result in
hypoxic or even anoxic conditions. Similarly, an excess load of organic carbon via municipal
and/or industrial discharges can result in direct depletion of oxygen through microbial oxidation
of the organic matter. These conditions are often worsened by restricted circulation of the water
mass in question. Human activities are, however, not the only causes of hypoxia and anoxia. In
the general oceanic circulation pattern, water masses in the Pacific Ocean have long been
reported as presenting a well-defined oxygen minimum zone (OMZ). Blake and Lissner (1997),
and references therein, have explained that the OMZ is the result of an excess of oxygen
consumption over oxygen supply. This excess consumption, in turn, is the result of frequent up-
welling events carrying nutrient-rich water to the surface with a consequent increase in
biological productivity. The sinking particles from this increased productivity reduces the
amount of dissolved oxygen at intermediate depths. Other authors (Childress and Seibel 1998)
have indicated that a difference exists between periodically hypoxic habitats and those of OMZs
in those in the OMZs are more permanent and usually occur over larger areas. In fact, because
hypoxia and anoxia are simply defined in terms of oxygen concentration, OMZs frequently
evidence hypoxia, and may even present anoxic conditions.
To our knowledge, no planned effort has been carried-out to map the OMZs of the Gulf
and, consequently, there is inadequate information on OMZ medium and long-term behavior.
Therefore, this work studied the spatial distribution of OMZs in the Gulf of California to enhance
our understanding of their origin and the possible consequences of their presence in this marginal
Sea.
STRUCTURE AND CIRCULATION OF WATER MASSES IN THE GULF
The structure of water masses in the Gulf of California (Figure 1) is complex. The Gulf
itself has been characterized as the only marginal sea of the Northeast Pacific Ocean. It lacks
significant riverine inputs, it is surrounded by arid and semi-arid land, and, consequently,
evaporation exceeds precipitation (Roden 1958, Bray and Robles 1991, Soto-Mardones et al.
1999). Estimations made by several authors place the value for evaporation at around 0.6 to 1.0
m/year (Roden and Groves 1959, Roden 1964, Beron-vera and Ripa 2000). Under summer
conditions, there is, however, a net gain of heat (Lavin and Organista 1988) from seawater of
sufficient magnitude to offset the loss of buoyancy. This effect results in the formation of a new
water mass in the northern part of the Gulf. This is the so-called Gulf of California water
(GCW) mass. The GCW water mass is characterized by both higher temperatures and salinities
(Figure 2). Because of its buoyancy, its circulation pattern occurs mainly in the upper few tenths
of meters.
286
-------
»*,_
x.
:
\
Snalea
Vt
Figure 1. Gulf of California and sampling sites for dissolved oxygen.
fr. ir'' ' ff.-'
C . -:^"
3 1k" ,*'•-"""
I _;,:::
6 - •'"''
f P'W
4 -\ . • .
3&:& 34.S 30.0 35.2 3G.-1 35.G
Figure 2. Water masses as defined by their T-S characteristics according to those described by
others. ESW = Equatorial surface Water. GCW = Gulf of California Water. SSW = Subsurface
Subtropical Water. PIW = Pacific Intermediate Water. PDW = Pacific Deep Water.
287
-------
To compensate for the loss of water by evaporation and the output of very saline GCW,
there is an input of deep water, composed mainly of Subsurface Subtropical Water (SSTW) and
Pacific Intermediate Water (PIW) masses into the Gulf (Figure 3). At depths of more than 1000
m, one can find the Pacific Deep Water (PDW). This water mass has been described as
originating in adjacent oceans. It enters the Pacific Ocean from the southwest between
Antarctica and New Zealand (Pickard and Emery 1990). The PIW is located around 500 to 1000
m depth. This water mass originates in the north and flows south in a clockwise manner. Above
these two water masses, one can find the Subsurface Subtropical Water mass (SSTW). This
water mass has been described as originating in the eastern Tropical Pacific (Alvarez-Borrego
and Schwartzlose 1979, Fernandez-Barajas et al. 1994).
GCW Q "' •• :!:;::;::; ESW
'" ;
PDW
Figure 3. Imaginary cut along the main axis of the Gulf of California. The main water masses
and their respective relative positions are shown. ESW = Equatorial Surface Water. GCW =
Gulf of California Water. SSTW = Subsurface Subtropical Water. PIW = Pacific Intermediate Water.
PDW = Pacific Deep Water.
Frequent up-welling along the main coast of Mexico, especially during winter, brings
nutrient-rich water to the surface, thereby supporting phytolankton blooms. Additionally,
Delgadillo-Hinojosa et al. (2001) have shown that the SSTW and probably the PIW, may be
contributors of nutrient-rich water to the surface layer of the Gulf in the Island region in the
northern part of the Gulf via the up-turning of a mix of intermediate water masses.
A typical Temperature-Salinity (T-S) diagram for the Gulf of California (Figure 2)
clearly defines the different characteristics of the water masses found in the Gulf. The profiles
288
-------
were obtained during the MEGAMARCO's (MEtals and GAses in the MAr of COrtez) I and II
oceanographic campaigns in 1996/97. Each profile corresponds to a station number shown in
Figure 1.
The diagrammatic structure of water masses present in the Gulf is shown in Figure 3.
The presence of these water masses can be detected in most oceanographic surveys as long as
they include salinity-temperature measurements extending from the surface to depths below
1000 m. The diagram shows the separation from or restricted circulation between the northern
part of the Gulf and the rest of the Gulf generated by the presence of sills in the Island region.
Due to topographic and other features related to productivity, the Gulf of California has
been divided into several regions (Gilbert and Allen 1943, Round 1967, Roden and Emilsson
1979). The northern end (Figure 1, stations 1-3) has been generally characterized as containing
the more saline waters (salinities above 35ppt). The wind regime and the mostly shallow depth
in this region help maintain a well-mixed water column generating a large load of sedimentary
materials in the water. The central region (Stations 8-10) is divided from the north by a series of
islands and sills that restrict the circulation of water with the rest of the Gulf. This region is
recognized as being the most productive (Alvarez-Borrego and Lara-Lara 1991), and where
upwellings and the up-turn of deeper water results in great enrichment of the surface water. To
gain some idea of the great biological productivity of the Gulf of California, we can compare
typical values reported for the plume region of the Mississippi River in the Gulf of Mexico of
320 g C/m2/yr (Rabalais et al. 1996) to values of 1,242 g C/m2/yr reported for the Gulf of
California (Gaxiola-Castro et al. 1995). This latter value is even larger than the one reported for
the upwelling region in Peru of 365 g C/m2/yr (Eganhouse and Venkatesan 1993).
A third region (Stations 12-17) has usually been assigned to a transitional zone consisting
of the area just south of the Islands and about half the distance to the mouth of the Gulf region.
This area has been reported as stratified for most of the summer, with little change in its water
structure. The separation between the central region and the south of the Gulf is ill defined. The
fourth region (Stations 19-21) is the one located near the mouth of the Gulf. It is generally
characterized as being the most oceanic part of the Gulf, possessing a complex circulation
patterns with the confluence of several water masses at different times during the year (Castro et
al. 2000).
The surface circulation in the Gulf of California is complex and seasonally. A series of
eddies with a counterclockwise circulation has been frequently reported from the mouth of the
Gulf to the central region (Beier and Ripa 1999). The circulation on the surface is associated
with wind stress. A predominant, northwesterly wind during the winter months (December -
April) results in up-welling events occurring mostly along the coasts of Sonora and Sinaloa. In
the northernmost part of the Gulf (to the north of Angel de la Guarda Island), Lavin et al. (1997)
provided evidence of the existence of a seasonally reversing gyro, cyclonic in spring/summer and
anticyclonic in fall/winter. This reversing gyro has been modeled by Beier (1997) and Beier and
Ripa (1999) using a two-layer, linear numerical model of the Gulf. They used only the annual
frequency forcing of wind, surface heat flux, and a baroclinic Kelvin wave at the mouth of the
Gulf. The gyros change their size and strength throughout the year.
289
-------
DISSOLVED OXYGEN CONCENTRATION DISTRIBUTION IN THE GULF
A typical dissolved oxygen profile for the Gulf is shown in Figure 4. In this figure, we
have indicated the approximate depth of each water mass of importance in the Gulf. Clearly, the
GCW contains oxygen concentrations well above hypoxia. In contrast, the SSTW and the PIW
water masses are the main contributors to hypoxic conditions. These latter two water masses, as
noted previously, appear to have different origins. Because they are located underneath the
photic zone, they tend to be the recipients of large amounts of paniculate organic material
raining down from the surface. Eventually, oxygen consumption exceeds the supply for these
two water masses. This has been demonstrated by others since these water masses exhibit very
well developed OMZ's, even before they enter the Gulf. Although we have discussed only the
SSTW and the PIW as the main contributors to hypoxic conditions, we also recognize that the
PDW can include a layer of around 400 to 500 m thick with oxygen concentrations of-2.5 mg/L
(Figure 4). However, due to the depth at which this PDW low oxygen layer is found, this water
mass is probably not a major contributor of nutrient-rich water to the surface layer.
012345678
0 "^l./W** • '*•* ~4 GCW
!:
i ! PIW
:QJ
a
*
Figure 4. Typical oxygen profile distribution (mg/L) showing the depth at which each water
mass is a contributor. GCW = Gulf of California Water. SSW = Subsurface Subtropical Water. PID
= Pacific Intermediate Water. PDW = Pacific Deep Water.
In Figure 5, we show the vertical oxygen distributions found for summer 1996 and winter
1997 in a transect running from the mouth to the north of the Gulf. The samples were collected
at the sites shown in Figure 1. Larger stratification can be observed for summer than for winter.
290
-------
Oxygen-rich water from the north appears to penetrate deeper during winter than during summer.
For example, the chosen oxygen concentration limit of 2.5 mg/L for hypoxia is located at around
100 m depth for all the Gulf in summer, but it is found below 200 m and even at 300 m depth at
the island region during winter. This deepening of the hypoxic condition can be attributed to the
mixing of water masses inside the Gulf promoted mainly by the interaction between tidal mixing
and the topographic features in the island area. Mixing allows the oxygenated water to penetrate
deeper and mix with the water layer having minimum oxygen concentration. This suggests that
although part of the GCW mass moves towards the mouth of the Gulf on the surface, another
part appears to get denser and sink underneath less dense water masses, and subsequent mixing.
r- ft"
-500
J—
100 200 300 400 SOQ 600
100 200 300 408 500 600 "Wil 80=0-
'ccs ' Km \
Figure 5. Latitudinal vertical distribution of dissolved oxygen (mg/L). The upper diagram
shows the summer concentrations and the lower one shows the oxygen distribution for
winter.
291
-------
To our knowledge, however, no effort has been made to specifically map the distribution
of anoxia and hypoxia in the Gulf of California. Our data shows oxygen profiles at specific
stations, although we can not comment on long term temporal trends, if any, for the OMZ
behavior within the Gulf.
ORIGIN OF THE HYPOXIA INSIDE THE GULF
The causes OMZs inside the Gulf are natural rather than anthropogenic. The OMZs
originate mainly because of the presence of "old" water inside the Gulf that comes from different
regions in the Northeast Pacific Ocean. For instance, under winter conditions, the SSTW signal
disappears from the Gulf, resulting in an intensified penetration of subsurface water coming from
the northeastern Pacific (Figure 5). Moreover, given the productivity exhibited by the Gulf,
hypoxia-anoxia is enhanced inside the Gulf due to the presence of frequent upwelling reported
both near the mainland of Mexico and on the Baja California coast. Additional information
bearing on the relative lack of anthopogenic impact is the fact that most rivers entering the Gulf,
especially the Colorado River, have been extensively dammed and the nutrient-rich discharges
severely restricted. Finally, most recently, an anthropogenic activity that could eventually have
impact on the OMZ is aquaculture and, in particular, shrimp farming. However, probably this is
more a local problem of potential nutrient enrichment. As such, it has negligible current effect
on the general distribution of oxygen poor waters in the Gulf.
ESTIMATION OF THE VOLUME OF HYPOXIC WATER IN THE GULF
It is clear that the volume of water having [©2] ~ 2.5 mg/L is highly irregular in the Gulf
of California. Because no study has been specifically designed to measure the distribution of the
OMZs, all estimates are essentially educated guesses. We have approximated this volume as a
regular shape (~ as a truncated pyramid) with an area at the mouth of-340 km2 and a truncated
top having an area of about 100 km2. This results in an estimate of-200,000 km3 of hypoxic
water. This volume represents on occasions as much as 40-60% of the total volume of water in
the Gulf. We do not know at present of any long term temporal trends, that is, if this volume is
growing or shrinking.
It is also important to emphasize that the origin of this hypoxic-anoxic water has not been
clarified. We believe that at least partially, the origin of this hypoxic water is natural. Mainly,
because as we have described above, the main component of water masses present inside the
Gulf of California are the PIW and PDW masses. These two water masses alone represent about
65% of the total volume inside the Gulf. Consequently, the origin of the OMZ for the PIW, the
PDW and the Gulfs OMZ must be the same. The general circulation belt strongly suggests that
this minimum originate mostly outside the Gulf due to the constant input of organic matter along
the path that the PIW and PDW masses travel.
292
-------
CONSEQUENCES OF HYPOXIC AND ANOXIC WATER IN THE GULF
The presence of such a large water volume with hypoxic and near anoxic conditions must
have some effect in shaping the distribution of living organisms in the Gulf. Large, free
swimmers have the capacity to get away from the presence of water with low-oxygen content.
However, there are organisms with restricted movement and sessile organisms located mainly in
sediments where this oxygen-poor water is in contact with the sediment. This is probably where
most of the low oxygen impact would be felt. If we estimate that the average water thickness
corresponding to the hypoxic water volume is about 1 km and the length of the Gulf is
somewhere around 1000 km, then the benthic area affected by the OMZ presence along the coast
must be near 1000 km2. This area must be doubled because the OMZ contacts both mainland
Mexico as well as the Baja California's eastern coast. If we take only half this number,
assuming that there are rocky bottoms, the 1000 km2 still represent a large area where living
organisms must adapt to survive in such extreme conditions.
One of the observable consequences of this low-oxygen water is the well-documented
presence of laminar sediments (Baumgartner et al. 1991). Under this condition, it is expected
that organic material exported to the bottom will be preserved for longer periods of time. This
organic matter consists mostly of refractory material with a relatively complex structure that is
resistant to attack by bacteria. In addition to the physical trapping of organic material, we would
expect that other materials, such as any excess input of trace metals produced by anthropogenic
activities, would reside in the sediment. At the same time, excessive input of materials during
rainy periods such as "El Nino" events are also expected to become preserved by the presence of
a water mass with very low or near zero oxygen content.
Another anticipated consequence of low-oxygen concentration that has been previously
reported is that when the oxygen content is less than 5|iM (0.16 |ig/L), denitrification or suboxic
metabolism occurs at the sediment-water interface. This can have an effect on the distribution of
chemical species both in the sediment and in the adjacent water column (Eganhouse and
Venkatesan 1993).
Finally, another probable outcome of the hypoxic/anoxic condition is that it may result in
a more restricted distribution of benthic organisms by generating anoxic sedimentary conditions.
To certain organisms, however, it may also offer advantages if they have the ability to adapt to
low-oxygen partial pressures, or to those that can swim away. It is a disadvantage for those that
cannot avoid its presence as has been explained by others (Childress et al. 1998).
ACKNOWLEDGEMENTS
We would like to thank CONACyT for providing the funds used to collect the oxygen
data (Projects 2045T and 2511PT). We would also like to thank EPA for the organization of
such wonderful meetings where knowledge and expertise are shared with so many researchers
from around the world.
293
-------
REFERENCES
Alvarez-Borrego, S., and J.R. Lara-Lara. 1991. The physical environment and primary
productivity of the Gulf of California. In: The Gulf and Peninsular Provinces of the
Californias. J.P. Dauphin and B.R.T. Simoneit (eds.). American Association of Petroleum
Geologists, Memoir 47. pp. 555-567.
Alvarez-Borrego, S., andR.A. Schwartzlose. 1979. Water Masses of the Gulf of California.
Ciencias Marinas 6: 43-63.
Baumgartner, T., V. Ferreira-Bartrina, and P. Moreno-Hentz. 1991. Varve formation in the Gulf
of California: A reconsideration of the origin of the dark laminae from the 20th century
varve record. In: The Gulf and Peninsular Provinces of the Californias. J.P. Dauphin and
B.R.T. Simoneit (eds.). American Association of Petroleum Geologists, Memoir 47.
pp.617-635 .
Beier, E., and P. Ripa. 1999. Seasonal gyros in the northern Gulf of California. Journal of
Physical Oceanography 29: 305-311.
Beron-Vera, J., and P. Ripa. 2000. Three dimensional aspects of the seasonal heat balance in the
Gulf of California. Journal of Geophysical Research 105: 11441-11457.
Blake, J.A., and L. Lissner. 1997. Physical description of the Santa Maria Basin and Western
Santa Barbara Channel. In: Santa Barbara Museum of Natural History. Taxonomic Atlas
of the Santa Maria Basin and the Western Santa Barbara Channel. U.S. Department of the
Interior. Volume 1. pp. 19-30.
Bray, N., and J.M. Robles. 1991. In: The Gulf and Peninsular Provinces of the Californias. J.P.
Dauphin and B.R.T. Simoneit (eds.). American Association of Petroleum Geologists,
Memoir 47. pp. 511 -553.
Castro, R., A.S. Mascarenhas, R. Durazo, and C. Collins. 2000. Seasonal variation of the
temperature and salinity at the entrance of the Gulf of California, Mexico. Ciencias
Marinas 26(4): 561-583.
Childress, J.J., and B.D. Seibel. 1998. Life at stable low oxygen levels: adaptation of animals to
oceanic oxygen minimum layer. Journal of Experimental Biology 201: 1223-1232.
Delgadillo-Hinojosa, F., J.V. Macias-Zamora, J.A. Segovia-Zavala, and S. Torres-Valdez. 2001.
Cadmium enrichment in the Gulf of California. Marine Chemistry 75:109-122.
294
-------
Eganhouse, R.P., and M.I. Venkatesan. 1993. Chemical Oceanography and Geochemistry. In:
Ecology of the Southern California Bight. A synthesis and interpretation. M.D. Dailey,
DJ. Reish, and J.W. Anderson (eds). pp. 71-188.
Fernandez-Barajas, M.E., M.A. Monreal-Gomez, and A. Molina-Cruz. 1994. Thermohaline
structure and geostrophic flow in the Gulf of California, during 1992. Ciencias Marinas
20(2): 267-286.
Gilbert, J.Y., and W.E. Allen. 1943. The phytoplankton of the Gulf of California obtained by the
"E.W. Scripps" in 1939 and 1940. Journal of Marine Research 5: 89-10.
Lavin, M.F., E. Beier, and A. Badan. 1997. Estructura hidrogafica y circulation del Golfo de
California: escala global, estacional e interanual. En: M.F. Lavin (ed). Contribuciones a la
Oceanografia Fisica en Mexico. Serie Monografias Geofisicas. Union Geofisica
Mexicana.
Lavin, M.F., and S. Organista. 1988. Surface heat flux in the northern Gulf of California. Journal
of Geophysical Research 93(C11): 14033-14038.
Leming, T.D., and W.E. Stuntz. 1984. Zones of coastal hypoxia revealed by satellite scanning
have implications for strategic fishing. Nature 310: 136-138.
Pickard, G.L., and W.J. Emery. 1990. Descriptive Physical Oceanography. An Introduction.
Pergamon Press, Oxford, England. 320 pp.
Rabalais, N.N., R.E. Turner, D. Justic, Q. Dortch, W.J. Wiseman, Jr., and B.K. Sen Guptka.
1996. Nutrient changes in the Mississippi River and system responses on the adjacent
continental shelf. Estuaries 19: 386-407.
Roden, G.I. 1958. Oceanographic and meteorological aspects of the Gulf of California. Pacific
Science 12: 21-45.
Roden, G.I. 1964. Oceanographic aspects of the Gulf of California. In: Marine Geology of the
Gulf of California: A symposium. Tj. H. van Andel and G.G. Shor, Jr. (eds.). American
Association of Petroleum Geologists, Memoir 3. pp. 30-58.
Roden, G.I., and G.W. Groves. 1959. Recent Oceanographic investigations in the Gulf of
California. Marine Research Journal 18: 10-35.
Roden, G.I., and I. Emilson. 1979. Physical Oceanography of the Gulf of California.
295
-------
Round, F.E. 1967. The phytoplankton of the Gulf of California. Part I. Its composition,
distribution and contribution to the sediments. Journal of Experimental Marine Biology
and Ecology 1: 76-97.
Soto-Mardones, L., S.G. Marinone, and A. Pares-Sierra. 1999. Time and spatial variability of sea
surface temperature in the Gulf of California. Ciencias Marinas 25: 1-30.
Zimmerman, A.R. and E.A. Canuel. 2000. A geochemical record of eutrophication and anoxia in
Chesapeake Bay sediments: anthropogenic influence on organic matter composition.
Marine Chemistry 69: 117-137.
296
-------
THE ECOLOGICAL CONDITION OF ESTUARIES: A FOCUS ON THE
ATLANTIC OCEAN AND GULF OF MEXICO COASTS OF THE UNITED STATES
William H. Benson1 and J. Kevin Summers1
ABSTRACT
Monitoring the estuaries of the Atlantic Ocean and Gulf of Mexico coastlines was
performed annually from 1990 to 1997 to assess ecological conditions on a regional basis for
four biogeographic provinces. These province estimates - Virginian, Carolinian, West Indian,
and Louisianian Provinces - are combined to provide an assessment of 87% of the estuarine area
of the United States and 96% of the estuarine area of the Atlantic and Gulf coasts. Combining
information over the 6 years of monitoring showed 34 ± 4% of the Atlantic and Gulf estuarine
sediments displayed poorer than expected biological conditions, based on benthic and finfish
community conditions, and 21 ± 4% of the area was characterized by low water clarity, the
presence of marine debris/noxious odors, or elevated fish tissue contaminants. More recently,
the U.S. Environmental Protection Agency has initiated a 5-year effort, known as Coastal 2000,
to build the scientific basis and the state capacity to monitor for the status of, and trends in, the
condition of estuaries of the United States.
INTRODUCTION
Estuaries are bodies of water that are balanced by freshwater and sediment influx from
rivers and the tidal actions of the oceans, thus providing transition zones between the freshwater
of a river and the saline environment of the sea. The result of this interaction is an environment
where estuaries, along with their adjacent marshes and seagrasses, can provide a highly
productive ecosystem that supports wildlife and fisheries, and contribute substantially to the
economy of coastal areas.
Coastal areas are the most developed in the United States. The coastal zone (comprising
17% of the nation's land mass) is home to more than 53% of the nation's population (NRC
2000). This pattern in coastal populations is increasing by 3,600 people per day, resulting in a
projected population increase of 27 million in the next decade
[http://state_of_the_coast.noaa.gov/bulletins/html/pop_01.html]. In addition to being a center of
population, coasts of the United States are a source of valuable commodities; for example, 31%
of the gross national product and 85% of commercially harvested fish depend on estuarine
habitats; 180 million people use coastal resources annually for swimming, diving, and boating,
while the estuaries receive discharges from numerous municipalities and industries (Cunningham
and Walker 1996, NRC 1997). Approximately $15 billion in public funds are spent annually on
^.S. Environmental Protection Agency, Office of Research and Development, National Health and Environmental
Effects Research Laboratory, Gulf Ecology Division, Gulf Breeze, Florida 32561 USA.
297
-------
outdoor marine and estuarine recreation in the 18 coastal states bordering the Atlantic Ocean and
the Gulf of Mexico (NOAA 1988).
Despite the importance of the coastal region to the nation's economy and well being,
little is actually known about the status and trends of critical environmental variables in coastal
regions. Other than coastal weather, water levels, and commercial fisheries, there are few
consistent measurements of the ecological condition of estuaries. There is, at present in the
United States, no nationally consistent, comprehensive monitoring program to provide the
information necessary for effective management and decision-making for coastal ecosystems.
However, for the past decade, the U.S. Environmental Protection Agency (U.S. EPA) Office of
Research and Development has been developing monitoring approaches and indicators that
could be used in a comprehensive monitoring program. The Environmental Monitoring and
Assessment Program (EMAP) has surveyed and assessed estuarine conditions in approximately
87% of the estuarine acreage in the continental United States between 1990 and 1997. The
information is reported here and is being used to initiate a synoptic national monitoring survey of
estuarine resources. The 1990 to 1997 surveys conducted along the Atlantic and Gulf of Mexico
coasts represent the data available to develop a baseline of ecological condition.
Methods
Regional surveys were conducted in the Virginian (1990-1993), Carolinian (1995-1997),
West Indian (1995-1996), and Louisianian (1991-1994) Provinces by sampling 100 to 150 sites
annually in each province. Sites were selected from three different strata using a probability-
based design (Summers etal. 1995). A simple classification system based on physical
dimensions was used to delineate the three sampling strata - large estuaries (> 250 km2,
length/mean width or aspect < 18), large tidal rivers (> 250 km2, aspect > 18), and small
estuarine systems (2-250 km2). Along the Atlantic and Gulf of Mexico coastlines, 1,516
estuaries totaling 74,744 km2 were identified that met the above criteria. These comprised 42
large estuaries (43,536 km2), 1,464 small estuaries or small tidal rivers (27,259 km2), and the
tidal portions of 10 large tidal rivers (3,949 km2). All sites were sampled during a 6- to 8-week
index period in late summer (July 15 - September 15). This time period was selected for
sampling in that it represents the time period most likely to show ecological effects due to
decreased dissolved oxygen conditions, increased contaminant availability, and increased human
usage.
Sites were characterized using selected indicators (Table 1) in order to determine the
status of components of ecological condition. The strategy for the selection of indicators for use
in EMAP-Coastal is described in detail in Knapp et al. (1990), Griffith et al. (1994), and Jackson
et al. [http://www.epa.gov/emap/ html/pubs/resdocs/]. Monitoring focused on indicators of
ecological response to stress and used measures of exposure to stress as a means of interpreting
that response. Indicators specific to a given region were added in some regions where the
addition of a specific indicator addressed a particular regional issue.
298
-------
Table 1. The ecological indicators used during the 1990-1996 EMAP-Coastal Monitoring
Program.
Indicator Type Indicator
Response Benthic community composition
Benthic abundance
Fish community composition
Pathologies in fish
Presence of submerged aquatic vegetation
Exposure Sediment contaminant concentrations
Sediment toxicity
Dissolved oxygen concentrations
Contaminant concentrations in fish tissue
Habitat Percent light transmittance
Salinity
Temperature
pH
Percent silt-clay
Grain size
At each site, a set of samples and data were taken using standardized methods. Triplicate
benthic samples were taken using a Young-modified Van Veen grab (440 cm2) in order to
provide data on the structure and composition of the benthic macroinvertebrate community. A
small core (60 cc) was extracted from each grab sample for sediment characterization (percent
silt-clay, grain size, and total organic carbon). Sediment samples for acid volatile sulfide
determinations were collected from mid-Atlantic and Gulf of Mexico sites. The remainder of
each grab was sieved through a 0.5 mm screen. Samples were preserved in 10% buffered, rose-
bengal formalin solution and stored at least 30 days prior to processing to assure adequate
fixation (Gaston et al. 1996). After 30 days, the stored samples were transferred from formalin
to ethanol, sorted, identified to species, and counted.
Additional sediment grabs were collected at each sampling site for sediment contaminant
analyses and toxicity bioassays. Samples were collected from a homogenate created at the site
from several (6-10) grabs from which the top 2 cm of sediment has been removed, placed in a
container, and thoroughly mixed until approximately 4 L of sediment had been obtained. This
mixture was apportioned for sediment chemistry analyses and toxicity testing.
Analyses for the determination of sediment characteristics included grain size, percent
silt-clay content, and total organic carbon at all sites. Grain size and silt-clay analyses were
initially determined by sieving through a 63 |im mesh sieve. Both the filtrate and the fraction
retained on the sieve were dried at 60°C and weighed to calculate the proportion of silts and
clays. Grain size determination was determined by further fractionation of the sieve-retained
299
-------
portion through additional sieves prior to drying. Total organic carbon was determined by
drying at least 5 g wet weight (wwt) of sediment for 48 hours; grinding to a fine consistency,
acidifying to remove inorganic carbon (e.g., shell fragments), igniting at 950°C, and measuring
the carbon dioxide evolved using an infrared gas analyzer.
Sediment samples for contaminant analyses were collected from the field homogenate at
each site. The sediments were analyzed by standard methods (U.S. EPA 1995) for the group of
contaminants listed in Table 2. Most of the sediment from the homogenate was used for
sediment bioassays. Toxicity tests were performed using the standard 10-day test method
(Swartz et al. 1985, ASTM 1990) with Ampelisca abdita, the tube-dwelling amphipod. Five
replicate tests were completed under static conditions for the listed time length at 20°C and 30
ppt.
Fishes, shrimps, and blue crabs were collected by trawling (depending upon location)
using a 5 m, high-rise otter trawl with a 2.5 cm mesh cod end in the Carolinian and Louisianian
Province; fish traps were used in the West Indian Province due to trawling restrictions in
Everglades National Park; and a 15 m, high-rise otter trawl with a 2.5 cm mesh cod end was used
in the deeper waters of the Virginian Province. The net was towed for 10 minutes against the
current at a speed of between 0.7-1.0 meter/second. All fishes and shellfishes caught in the
trawls and traps were identified to species, counted, and up to 20-30 individuals of each species
selected at random were measured to the nearest millimeter. All fishes were examined for
external pathologies. This inspection included body spinal alignment, lumps, bumps, bruises,
growths, opercular deformity, fin erosion, eye deformities, buccal cavity growths and
hemorrhages, parasitism, and overall body form.
Up to 10 individual target fishes/shellfishes were retained for tissue residue analysis, with
species depending upon geographic location (Table 3). The specimens were labeled, packed
with dry ice, and shipped to the appropriate laboratory where they were stored frozen until
analysis. Where available, four to 10 individuals of each species from each sampling site were
analyzed by compositing fillets into a homogeneous slurry. The edible portions of these fishes
and shellfishes included fillets with skin for Atlantic croakers, white perch, and seatrout; fillets
without skin for all catfish; tail meat for shrimp; and picked lump and claw meat for blue crabs.
This slurry was appropriately digested, extracted, and analyzed according to the methods of U.S.
EPA (1990).
Water quality information was collected at each site for instantaneous representations of
temperature, salinity, pH, and dissolved oxygen (DO). A Hydrolab Surveyor 2 (Hydrolab,
Austin, Texas, USA) equipped with a DO electrode was used to make the instantaneous
measurements between the hours of 9 a.m. and 4 p.m.. Vertical profiles of the water column at
meter intervals from surface to bottom were taken at all sites. Proportion of surface light
penetration was determined using a LICOR LI-1000 (LICOR, Lincoln, Nebraska, USA)
containing a submersible light sensor. Underwater readings at 1-m intervals were measured
simultaneously with ambient surface light. The ratio of these two measures provides a measure
of proportional light penetration based on incident light. The proportion reaching 1 m in depth
was used as an indicator of water clarity.
300
-------
Table 2. List of organic and inorganic compounds determined in both sediment and tissue
samples.
Polynuclear Aromatic Hydrocarbons (PAHs)
Acenaphthene
Anthracene
Benz(a)anthracene
Benzo(a)pyrene
Biphenyl
Chrysene
Dibenz(a,h)anthracene
Dibenzothiophene
2,6-dimethylnaphthalene
Fluoranthene
Fluorene
2-methylnaphthalene
1 -methy Inaphthalene
1 -methy Iphenanthrene
2,6-dimethylnaphtalene
Naphthalene
Pyrene
Benzo(b)fluoranthene
Acenaphthylene
Benzo(k)fluoranthene
Benzo(g,h,i)perylene
Ideno(l,2,3-c,d)pyrene
2,3,5-trimethy Inaphthalene
PCB Congener Number and Name
8 2,4'-dichlorobiphenyl
18 2,2',5-trichlorobiphenyl
28 2,4,4'-trichlorobiphenyl
44 2,2',3,5'-tetrachlorobiphenyl
52 2,2',5,5'-tetrachlorobiphenyl
66 2,3',4,4'-tetrachlorobiphenyl
101 2,2',4,5,5'-pentachlorobiphenyl
105 2,3,3',4,4'-pentachlorobiphenyl
110/77 2,3,3',4',6-pentachlorobiphenyl
3,3 ',4,4'-tetrachlorobiphenyl
118 2,3,4,4',5-pentachlorobiphenyl
126 3,3,4,4',5-pentachlorobiphenyl
128 2,2',3,3',4,4'-hexachlorobiphenyl
138 2,2',3,4,4',5'-hexachlorobiphenyl
153 2,2',4,4',5,5'-hexachlorobiphenyl
170 2,2',3,3',4,4',5-heptachlorobiphenyl
180 2,2',3,4,4',5,5'-heptachlorobiphenyl
187 2,2',3,4',5,5',6-heptachlorobiphenyl
195 2,2',3,3',4,4',5,6-octachlorobiphenyl
206 2,2',3,3',4,4',5,5',6-nonachlorobiphenyl
209 2,2'3,3',4,4',5,5',6,6'-decachlorobiphenyl
DDT and its Metabolites
2,4'-DDD
4,4'-DDD
2,4'-DDE
4,4'-DDE
2,4'-DDT
4,4'-DDT
Chlorinated Pesticides
other than DDT
Aldrin
Alpha-Chlordane
Dieldrin
Endosulfan I
Endosulfan II
Endosulfan sulfate
Endrin
Heptachlor
Heptachlor epoxide
Hexachlorobenzene
Lindane (gamma-BHC)
Mirex
Toxaphene
Trans-Nonachlor
Trace Elements
Aluminum
Antimony (sediment, only)
Arsenic
Cadmium
Chromium
Copper
Iron
Lead
Manganese (sediment, only)
Mercury
Nickel
Selenium
Silver
Tin
Zinc
301
-------
Table 3. Target species examined for residue analysis of edible tissue by province.
Species
Province
Virginian
Carolinian
West Indian Louisianian
Catfish
(Ictalums punctatus)
(Ameiums catus)
(Bagre marinus)
(Ariusfelis)
X
X
X
X
Atlantic croaker
(Micropogonias undulatus) X
Spot
(Leiostomus xanthums)
Shrimp
(Penaeus aztecus)
(Penaeus setiferus)
White Perch
(Morone americana) X
Weakfish
(Cynoscion regalis) X
Bluefish
(Pomatomus saltatrix) X
Winter Flounder
(Pleuronectes americanus) X
X
X
X
X
X
X
X
X
X
X
Anthropogenically-generated marine debris was determined from the contents of the
benthic grabs, fish trawls, and from surface floatables. The incidence and composition of this
debris was determined for each site location.
Data Analysis
All ecological indicators collected from the Atlantic and Gulf coasts were characterized
using cumulative distribution functions (Sokal and Rohlf 1981). These functions describe the
full distribution of these indicators in relation to their areal extent within the sampled province,
and were used primarily to determine the proportion of each province that is degraded with
respect to that indicator. All observations were weighted by the inclusion probability assigned to
each site location based upon the surface area associated with each site, and that represents the
probability of the sample's inclusion in the sampling design. For large estuaries in all provinces
and all estuaries in the West Indian Province, the inclusion probability was equal to the
hexagonal sampling space created by the design (280 km2 in large estuaries of the Virginian,
Louisianian, and Carolinian Provinces and 88 km in all estuaries of the West Indian Province)
302
-------
divided by the total area of the large estuaries included in the province sampling or, in the case
of the West Indian Province, all estuarine area sampled. For large tidal rivers and small tidal
rivers/estuaries in the Virginian, Louisianian, and Carolinian Provinces an alternate design and
analytical approach was used. For large tidal rivers, the inclusion probability associated with
each sampling segment was the surface area of the sampled segment divided by the area of the
estuarine portion of that large river. This included resources like the Potomac River, Indian
River Lagoon, Neuse River, and Mississippi River. For small estuaries and small tidal rivers in
these provinces, the inclusion probability for any small tidal river/estuary was equal to the total
surface area of that resource divided by the sum of the surface areas of the small resources
included in each year's survey. The approximate 95% confidence intervals for the province
level cumulative distribution functions were calculated based on Heimbuch etal. (1998).
Benthic indices in each of the biogeographic regions were created by combining multiple
metrics into a single, multi-metric index of benthic condition for each province (Engle and
Summers 1999, Engle etal. 1994, Hyland etal. 1998, 1999, Weisburg etal. 1997). These
indices integrate parameters of macrobenthic community structure, and are capable of
distinguishing polluted and unpolluted areas. While the indices are different in each province,
their components are largely the same (e.g., community biodiversity, abundance of pollution-
tolerant and pollution-sensitive species, proportional community composition) and each index
represents a relative measure of the condition of benthic resources in that province.
Guidelines used to assess potential for sediment degradation were the Long et al. (1995)
and Long and Morgan (1990) median values (ER-M) associated with biological effects. In
addition, the Long et al. (1995) 10% values (ER-L) were used to assess locations where some
contamination occurred at levels that had a low probability of resulting in biological effects.
Threshold values for province-wide ecological condition were determined through combinations
of the individual measures/indices based on an integration of literature values.
The proportion of estuarine area meeting acceptable human uses was determined by
combining data representing tissue residues in target species, proportional light penetration, the
presence/absence of marine debris, and the presence/absence of noxious odors from either the
water column or sediments. Poor conditions (exceeding threshold values, or presence of odors
and debris) of any of these measures were determined to constitute a poor human use condition.
RESULTS
Overall Condition
The overall health of the Southeast Atlantic Ocean and Gulf of Mexico coast estuaries is
good, based on data collected throughout the Southeast Atlantic and Gulf coasts from nearly
1000 stations sampled from 1990-1997. More specifically, about 56% of the estuaries are in
good condition for supporting plants, animals, and human uses. About 34% of the area of these
estuarine resources has poor benthic and fish community conditions while, 33% of the area has
303
-------
unacceptable levels of human-related uses.
Most of the biological communities in poor condition are benthic communities (bottom-
dwelling organisms). These poor conditions occur in areas of hypoxia, eutrophication, sediment
contamination, and habitat degradation. Depending upon location along the Atlantic and Gulf
coastlines, poor benthic conditions ranged from 27 to 35% of estuarine sediments. Aquatic life
is categorized as poor based on measures of biodiversity, increased abundances of pollution-
tolerant species, and decreased abundances of pollution-sensitive species. Less than 1% of
fishes examined (approximately 100,000 estuarine fishes) throughout the United States indicated
evidence of fin erosion, skin lesions, eye disorders, or gill problems.
People use and enjoy estuarine resources in many ways, including swimming, boating,
walking along the shore, animal husbandry, and aquaculture. Approximately 5 to 30% of the
Atlantic and Gulf Coast's estuarine waters are categorized as degraded for some human use.
There are three primary contributors to human use degradation: (1) water clarity, which affects
recreational activities; (2) accumulation of marine debris and presence of noxious odors, which
affects aesthetics and wildlife health; and (3) bioaccumulation of contaminants in edible portions
of fishes and shellfishes, which affect consumption.
Water Quality
Eutrophic conditions are based primarily on light penetration and dissolved oxygen
conditions. Clear waters are valued by society and contribute to the maintenance of healthy and
productive ecosystems. Losses of submerged aquatic vegetation can occur when light is
decreased due to turbid water associated with overgrowth of algae. Water visibility of < 10% at
1 m depth (10% of surface light reaches 1 m) is used to indicate poor conditions. This is
equivalent to being unable to see your hand in front of your face at a depth of 1 m. Poor light
penetration is a problem in 4% of estuarine waters, primarily in the western Gulf of Mexico and
western tributaries of Chesapeake Bay.
Low concentrations of dissolved oxygen often occur as a result of large algal blooms that
sink to the bottom and consume oxygen during the process of decay. Dissolved oxygen is a
fundamental requirement for all estuarine life. A threshold concentration of 4 to 5 ppm is used
by many states to set water quality standards. A concentration of 2 ppm is thought to be
extremely stressful to most estuarine organisms. Low levels of oxygen (hypoxia) or a lack of
oxygen (anoxia) often result from the onset of increased bacterial degradation of organic
materials, sometimes resulting in algal scums, fish kills, and noxious odors, as well as habitat
loss and degraded aesthetic values. These impacts result in a loss of tourism and recreational
water use. EMAP estimates that 4% of estuarine bottom waters are hypoxic (< 2 ppm) while
about 80% of waters maintain higher levels of dissolved oxygen (>5 ppm).
304
-------
Sediment Quality
Measurements of over 100 contaminants have been taken at each site including over 25
polynuclear aromatic hydrocarbons (PAHs), 22 poly cyclic biphenyl congeners (PCBs), total
PCBs, over 25 pesticides, and 15 metals. One to two percent of estuarine sediments in the
United States show concentrations of contaminants (PAHs, PCBs, pesticides, and metals) that
are above ER-M guidelines (the concentration of a contaminant associated with adverse effects
on estuarine organisms in the field and laboratory) while 10 to 29% of sediments have
contaminant concentrations that exceed the ER-L guidelines (the concentration having a low
probability of affecting organisms adversely). Most of the locations exceeding the ER-M
guidelines are located in the mid-Atlantic while the Gulf of Mexico coast contains many
locations that exceed the ER-L guidelines for five or more contaminants.
One of the challenges of assessing the magnitude of sediment contamination is
differentiating contaminants (organics, metals, and pesticides) that may occur naturally in the
earth's crust from those that are added from human activities. PCBs are relatively easy to
evaluate as they can only come from human activities. Similarly, with the exception of
arsenical, cyanide, microbial and botanical pesticides, most pesticides also come from human
activities. However, polynuclear aromatic hydrocarbons (PAHs), the above listed pesticides, and
metals can and do naturally occur in estuarine sediments. These measurements show that 40, 45,
and 75% of estuarine sediments are enriched with metals, PCBs, and pesticides, repectively,
from human sources.
Chemical analyses of sediments can provide information on the concentrations and
mixtures of potentially toxic substances in sediment samples. However, information gained
from these analyses alone provides no direct measure of the toxicological significance of the
chemicals. The Atlantic and Gulf Coastal Status and Trends Program and EMAP have
conducted surveys of sediment toxicity throughout the United States since 1981. Over 1000
locations have been tested using Ampelisca abdita, an amphipod that naturally occurs in
estuarine sediments. EMAP test results show that 10% of sediments in the estuaries of the
United States are toxic (resulting in significant mortalities) to amphipods exposed to sediments
for 10 days.
Mortality is not the only effect that contaminated sediments can have on benthic
organisms. Sub-lethal effects, including reductions in growth, changes in community structure
(biodiversity), and changes in abundance (reproduction), can occur as a result of exposure to
contaminated sediments. The EMAP benthic indices reflect changes in benthic community
diversity and the abundances and ratios of pollution-tolerant, and pollution-sensitive species.
Twenty-two percent of estuarine sediments are characterized by benthic communities that are
less diverse than expected, populated by greater than the expected number of pollution-tolerant
species, and contain fewer than expected pollution-sensitive species. To a large extent, these
differences appear to result from contaminated sediments, hypoxic conditions, habitat
degradation and eutrophication.
305
-------
Biotic Condition
Estuarine biota are negatively affected in about 34% of the estuarine area of the United
States. These effects include increased abundances of plankton, community changes in the
benthos, decreased abundances of fishes, increased incidences offish diseases, bioaccumulation
of contaminants in fish tissue, fish kills, and marine mammal mortalities. Earlier, the results of a
eutrophication assessment by the U.S. National Oceanic and Atmospheric Administration
(NOAA) showed that 22% of the Atlantic Ocean and Gulf Coast's estuarine area had high
concentrations of chlorophyll a (http://state_of_the_coast.noaa.gov/bulletins/html/eutro.html),
and EMAP has shown that 22% of benthic communities in the Atlantic and Gulf Coast's
estuaries are in poorer condition than expected (Summers 2001).
The frequency and type of gross pathologies on fishes taken in trawls of estuarine waters
are indicators of overall condition offish populations. Nearly 100,000 fishes were examined
from United States estuaries; only 454 of the fishes (0.5%) had external abnormalities (Table 4).
Of the fishes examined, bottom-feeding fishes (e.g., catfish) had the highest frequency of
disease. The number of fishes with multiple gross pathologies increased in areas where the
sediments contained high levels of multiple contaminants (Summers 2001).
Table 4. Proportion of fishes examined with external pathologies by province.
Province Number of Fishes Percent with pathologies
Virginian 13,421 0.4%
Carolinian 13,304 0.3%
Louisianian &
West Indian 64,100 0.7%
TOTAL 90,825
DISCUSSION
Six years of monitoring the estuaries along the Atlantic Ocean and Gulf of Mexico
coastlines have shown that monitoring ecological indicators of condition at a regional scale can
produce information that is useful to resource managers, particularly in identifying the extent of
observed problems. The probability-based sampling design and standardized methodologies
allowed for the collection of data that can be used in performing assessments throughout the
United States with a quantifiable level of uncertainty. These surveys represent the first of their
type in estuarine waters at large regional scales and with the capability of estimating condition
306
-------
with a known level of confidence. Other large-scale monitoring programs, such as the National
Status and Trends Program of NOAA, have stations that are located throughout the region or
nation. The NOAA stations are fixed and cannot readily be used to integrate data regionally to
assess overall condition. Only by assuming that biological populations at these fixed sites are
representative of the overall population, can they be used to assess overall condition. Rarely can
this assumption of representativeness be supported. Monitoring programs performed by
individual states (with some exceptions like Texas Park and Wildlife's fish survey) are also
based on fixed locations selected a priori based on known condition (e.g., a discharge is located
at the site, a bridge traverses the estuary at that point, a buoy exists at the location). Prior to
EMAP, estuarine regional assessments would have to bring together data collected from
different programs, at different scales, using non-standardized methodologies and attempt to
integrate the information. Using the EMAP-type probability design, changes in status and trends
in populations of estuarine resources can be determined within and between geographical
regions. While this approach was applied at the biogeographic region spatial scale, it is equally
useful at national, state, or local scales. To address all these scales, the design simply must be
adapted to the chosen scale or adapted, in a nested fashion, to represent multiple scales.
This form of re-adaptation of the EMAP approach to state and national scales is the basis
for U.S. EPA's National Coastal Assessment sampling program that will assess the condition of
estuarine resources within each of 25 coastal states and Puerto Rico over the period 2000 to
2001. These data from the 25 states will be integrated in the first national assessment of
estuarine condition with known confidence. The National Coastal Assessment and its
predecessor, EMAP-Estuaries, represent the first attempts by a large scale monitoring program to
incorporate common sampling methods over large geographic areas to estimate ecological
condition on an areal basis. By continuing these measurements, annually or on a fixed schedule
(e.g., 3-5 years) through the coming decades, changes in ecological status and trends can be
measured, assessed, and tracked objectively. This information can be used to determine whether
the environmental programs and policies of the United States and of individual states are
effectively protecting and/or restoring the estuarine environment. Conversely, the information
can pinpoint those programs and policies that are not having the desired effect on the
environment, and can then be used to modify, change, or restructure restorative efforts.
More research is necessary on the components of indices and their relationships to
ecological conditions and environmental stressors to assess their stability, accuracy, and validity.
These measurements, in combination with the probabilistic survey design approach, lie at the
core of regional and national assessments of ecological condition. Through the Clean Water Act
Plan (http://www.cleanwater.gov), multiple federal agencies have designed a multi-agency,
integrated research and monitoring program that will provide: (1) the necessary research to
improve these indicators on a continuous basis; (2) the assessment techniques to utilize them
better; and (3) a multi-spatial and multi-temporal scale monitoring program to collect the data for
all coastal resources [http://www.epa.gov/cwap]. The U.S. EPA National Coastal Assessment
Program represents the first tier of this proposed monitoring plan for estuaries.
307
-------
ACKNOWLEDGEMENT
The authors greatly appreciate the interactions with the "EMAP coastal crew" including,
but not limited to, Virginia Engle, John Macauley, Jeff Hyland, Fred Holland, Steve Weisberg,
John Paul, Brian Melzian, Charles Strobel, Robert Van Dolah, Amy Ringwood, Barbara Brown,
Oilman Veith, Rick Linthurst, Jay Messer, Ed Martinko, Tom Demoss, and Rick Kutz.
REFERENCES
ASTM (American Society for Testing and Materials). 1990. Standard Guide for Conducting 10-
Day Static Sediment Toxicity Tests with Marine and Estuarine Amphipods. E 1052.
Annual Book of ASTM Standards. Vol. 11, Philadelphia, Pennsylvania, USA.
Cunningham, C., and K. Walker. 1996. Enhancing public access to the coast through the
CZMA. Journal of Marine Education 14: 8-12.
Engle, V.D., J.K. Summers, and G.R. Gaston. 1994. Abenthic index of environmental
condition of Gulf of Mexico estuaries. Estuaries 17: 372-384.
Engle, V.D., and J.K. Summers. 1999. Refinement, validation and application of a benthic
condition index for northern Gulf of Mexico estuaries. Estuaries 22: 624-635.
Gaston, G.F., J.H.W. Bartlett, and R.W. Heard. 1996. Biomass variations of estuarine
macrobenthos preserved in ethanol and formalin. Estuaries 19:674-677.
Griffith, J.A., C.T. Hunsaker, and M.C. Barber. 1994. Ecosystem Monitoring and Ecological
Indicators: An Annotated Bibliography. EPA/620/R-94/021. U. S. Environmental
Protection Agency. Athens, Georgia, USA.
Heimbuch, D.G., H.T. Wilson, and J.K. Summers. 1998. Design-based estimators and power
analyses of trend tests for the proportion offish that exhibit gross pathological disorders.
Environmental Ecological Statistics 5: 65-80.
Hyland, J.L., L. Balthis, C.T. Hackney, G. McRae, A.H. Ringwood, T.R. Snoots, R.F. Van
Dolah, and T.L. Wade. 1998. Environmental quality of estuaries of the Carolinian
Province - 1995. Technical Memorandum. NOS ORCA 123. National Oceanic and
Atmospheric Administration, Silver Springs, Maryland, USA.
308
-------
Hyland, J.L., R.F. Van Dolah, and T.R. Snoots. 1999. Predicting stress in benthic communities
of southeastern U.S. estuaries in relation to chemical contaminants of sediments.
Environmental Toxicology and Chemistry 18: 2557-2568.
Knapp, C.M., D.R. Marmorek, J.P. Baker, K.W. Thornton, J.M. Klopatek, and D.F. Charles.
1990. The Indicator Strategy for the Environmental Monitoring and Assessment
Program. EPA/600/3-91/023. U.S. Environmental Protection Agency, Washington, DC,
USA.
Long, E.R., and L.G. Morgan. 1990. The potential for biological effects of sediment-sorbed
contaminants tested in the National Status and Trends Program. Tech. Memo. NOS OMA
52. National Oceanic and Atmospheric Administration, Rockville, Maryland, USA.
Long, E.R., K. MacDonald, S.L. Smith, and F.D. Calder. 1995. Incidence of adverse biological
effects within ranges of chemical concentrations in marine and estuarine sediments.
Environmental Management 19: 81-94.
NOAA (National Oceanic and Atmospheric Administration). 1988. Federal Plan for Ocean
Pollution Research Development and Monitoring: Fiscal Years 1988-1992. National
Oceanic and Atmospheric Administration, Rockville, Maryland, USA.
NRC (National Research Council). 1997. Striking a Balance: Improving Stewardship of
Marine Areas. National Academy Press, Washington, D.C., USA.
NRC (National Research Council). 2000. Clean Coastal Waters: Understanding and Reducing
the Effects of Nutrient Pollution. National Academy Press, Washington, DC, USA.
Sokal, R.R., and FJ. Rohlf. 1981. Biometry. New York, New York, USA.
Summers, J.K., J.F. Paul, and A. Robertson. 1995. Monitoring the ecological condition of
estuaries in the United States. Toxicology and Environmental Chemistry 49: 93-108.
Summers, J.K. 2001. Ecological condition of the estuaries of the Atlantic and Gulf coasts of the
United States. Environmental Toxicology and Chemistry 20: 99-106.
Swartz, R.C., W.A. DeBen, J.K. Jones, J.O. Lamberson, and F.A. Cole. 1985. Phoxocephalid
amphipod bioassay for marine sediment toxicity. In Cardwell, R.D., R. Purdy, and R.C.
Bahner (Eds.). Aquatic Toxicology and Hazard Assessment: Seventh Symposium, pp.
284-307. American Society for Testing and Materials, Philadelphia, Pennsylvania. USA.
U.S. EPA (U.S. Environmental Protection Agency). 1990. Environmental Monitoring and
Assessment Program-Near Coastal Demonstration Project: Laboratory Methods Manual.
EPA/600/R-91/001. Cincinnati, Ohio.
309
-------
U.S. EPA (U.S. Environmental Protection Agency). 1995. Environmental Monitoring and
Assessment Program (EMAP): Laboratory Methods Manual - Estuaries, Volume 1.
Biological and Physical Analyses. EPA/620/R-95/008. Narragansett, Rhode Island.
Weisberg, S.B., J.A. Ranasinghe, D.M. Dauer, L.C. Scaffner, RJ. Diaz, and J.B. Frithsen. 1997.
An estuarine benthic index of biotic integrity (B-IBI) for Chesapeake Bay. Estuaries 20:
149-158.
310
-------
HUMAN INFLUENCES ON COASTAL HYPOXIA:
EXAMPLES FROM THE CHESAPEAKE BAY WATERSHED
David L. Correll1
ABSTRACT
When humans convert the landscape from native vegetation to row crops and intensive
livestock production, nutrient losses to receiving waters convert these waters from nitrogen- or
phosphorus-limited to silicate- or light-limited systems. The resulting excessive primary
production brings about hypoxia. Interannual variations in seasonal and annual precipitation
change both the volume and the nutrient content of watershed discharge, resulting in changes in
the seasonality and spatial patterns of primary production and hypoxia in coastal waters.
Although winter and spring discharges of nitrogen and phosphorus from natural forests in the
Chesapeake Bay watershed can be significant, they are mostly in the form of organic forms. In
contrast, discharges from agricultural lands have much higher contents of nitrogen and
phosphorus and a much larger proportion is in the form of readily biologically available
inorganic nitrogen and phosphorus. Since discharges of dissolved silicate are not influenced as
much by weather or land use, discharges from agricultural watersheds, especially in unusually
wet seasons, are more likely to produce silicate limitation in receiving waters. In very wet
springs, the ratio of nitrogen to phosphorus in discharges from forested watersheds declines,
resulting in a higher probability of nitrogen limitation than in drier springs. This effect of
precipitation is more pronounced in discharges from cropland. The ratio of total organic carbon
to inorganic nitrogen also increases in wetter seasons, especially from cropland. Thus the
interacting effects of variations in precipitation and land use bring about changing probabilities
for the occurrence and extent of various nutrient effects in coastal receiving waters, but in all
weather conditions conversion of the watershed to agriculture increases nitrogen and phosphorus
discharges.
INTRODUCTION
Receiving waters normally respond to excessive nutrient loading by developing hypoxia.
Excessive nutrient loading, particularly with nitrogen and phosphorus, leads to phytoplankton
blooms. The deposition of the phytoplankton to bottom waters results in dissolved oxygen
depletion at rates that exceed oxygen inputs, thereby resulting in hypoxia and numerous
detrimental effects on benthic organisms and fishes (e.g. Boynton 2000). While I have chosen to
focus on Chesapeake Bay, in many ways other receiving water systems behave similarly, for
example, the Gulf of Mexico in the area affected by the Mississippi River plume (Rabalais et al.
2000) and freshwater Lake Okeechobee (Havens et al. 1996).
Smithsonian Environmental Research Center, Edgewater, Maryland, USA.
311
-------
Increased nutrient loading and estuarine responses leading to hypoxia are well
documented in Chesapeake Bay. The moving 2-year mean of river inflows is highly positively
correlated with chlorophyll a concentrations in surface waters and with annual average primary
production (Boynton and Kemp 2000). These river inflows are also highly correlated with
spring chlorophyll a deposition to the bottom and with rates of bottom water dissolved oxygen
depletion (Ibid). Annual nitrogen and phosphorus inputs to Chesapeake Bay are dominated by
diffuse sources from the watershed (Boynton et al. 1995). About 93% of the annual total
organic carbon inputs to Chesapeake Bay are generated internally rather than from diffuse
sources (Kemp et al. 1997).
Therefore, it is appropriate to ask the questions: how have changes in land use over
history affected nutrient inputs to Chesapeake Bay, and how do variations in weather affect
nutrient inputs from year to year? The purpose of this paper is to address these questions with
results from a long-term study of the nutrient dynamics of the Rhode River drainage basin.
Nutrient dynamics of the Rhode River watershed are probably better known than for any other
watershed. Although the Rhode River subwatershed is not necessarily representative of all parts
of the Chesapeake Bay watershed, or of other watersheds, there may be patterns that are
applicable elsewhere.
SITE DESCRIPTION AND METHODS
The watersheds studied herein are all subwatersheds of the Rhode River, a small tidal
tributary on the mid-western shore of Chesapeake Bay in Maryland, U.S.A. (38° 51' N, 76° 32'
W) in the inner Atlantic Coastal Plain physiographic province. The watershed has sedimentary
fine, sandy loam soils. Bedrock is about 1,000 m below the surface, but the Marlboro Clay layer
forms an effective aquiclude slightly above sea level throughout the watershed, causing each
subwatershed to have a perched aquifer. Overland storm flows, interflow, and groundwater
discharges move to the channel draining the subwatershed, where flows are measured and
volume-integrated samples taken continuously at V-notch weirs. Subwatershed slopes vary from
five to 11% and their sizes range from six to 1157 ha. Subwatersheds also differ in land use
from heavily row cropped to completely forested. For more detailed descriptions of the
subwatersheds see Correll (1977, 1981).
Precipitation data were taken at a weather station in a central location on the watershed.
Hydrologic and chemical data used in these analyses were taken from 1974 through 1998, except
for silicate, data for which was taken from 1984 through 1998. For details of sampling and
sample analysis see Correll et al. (1999a, b, c, d, 2000). For this article, the focus is on results
from two subwatersheds: watershed 110, hereafter referred to as the "forested watershed" or
"forest land", and watershed 109, hereafter referred to as the "cropland" or "crop watershed".
Watershed 110 is entirely forested, mostly with old-growth deciduous hardwoods that had never
been clear-cut. Watershed 109 was 64% row-cropped, while the remainder was a deciduous
hardwood riparian forest (see Peterjohn and Correll 1984). Prior to European settlement, most
of the Chesapeake Bay watershed was forested, so watershed 110 provides our best
approximation of pre-settlement nutrient dynamics.
312
-------
RESULTS AND DISCUSSION
Effects of Variations in Weather
Natural variations in seasonal precipitation have two effects on nutrient fluxes to the
Rhode River. When rains increase, more water moves from the watershed to the receiving
waters and, in the case of some nutrients, concentrations increase. Water yield increases with
precipitation as one might expect, sometimes in a non-linear manner (Correll et al. 1999a). In
the Rhode River subwatersheds agricultural land use has little effect on water yield, far less than
subwatershed size (Ibid.).
As winter and spring precipitation increases, the concentrations of inorganic phosphate,
organic phosphorus, nitrate, and organic nitrogen increase (Correll etal. 1999c, d), but silicate
concentrations remain about the same or may even decrease somewhat (Correll et al. 2000). As
a result, the fluxes of various fractions of nitrogen and phosphorus increase in a highly nonlinear
manner with increasing precipitation. Field data for these fluxes as a function of precipitation
are usually approximated best by either a power or an exponential equation (Correll et al. 1999c,
d). In the winter and spring, when high nutrient discharges are of most concern, in a very wet
season nutrient fluxes are typically from five to 100 times higher than for a very dry season
(Correll et al. 1999c, d). It is often assumed that these large increases in nutrient fluxes are the
result of overland storm flows and erosion. However, during major storms only the
concentrations of organic nitrogen and phosphorus exhibit major increases (Correll et al. 1999b).
Thus, the increased flux of these organic nutrients is to a large extent due to overland storm
flows, but the increases in nitrate flux are primarily due to increased leaching into groundwater,
which subsequently percolates to the streams between storms.
In wetter seasons the concentrations and fluxes of total organic carbon from the
watersheds also increase significantly (Correll et al. 2001). This increase is due primarily to
increased discharges of parti culate organic carbon in overland storm flows (Correll etal. 1999b,
2001).
Effects of Land Use
Like precipitation, land use also has a major effect on the concentrations and fluxes of
nutrients from the subwatersheds. In the winter and spring, nitrate fluxes from cropland were
100 to 200 times higher in dry seasons and 12 to 40 times higher in wet seasons than the fluxes
from native forest vegetation. Total organic nitrogen fluxes from cropland were three to eight
times higher in dry seasons and two to three times higher in wet seasons than from forest land
(Correll etal. 1999d).
For phosphorus in the winter and spring, inorganic phosphate fluxes from cropland were
four to nine times higher in dry seasons and five to eight times higher in wet seasons than the
fluxes from forest land. Organic phosphorus fluxes from cropland were four to eight times
higher in dry seasons and two to four times higher in wet seasons than from forest land (Correll
etal. 1999c).
313
-------
The combined effects of precipitation depth and land use on spring nitrate fluxes from the
watershed are illustrated in Figure 1. Observed nitrate fluxes from the forested watershed ranged
from essentially zero to about 250 g nitrate-N/ha-spring, while fluxes from the cropland
watershed ranged from about 200 to 4700 g nitrate-N/ha-spring. Although only 46 and 77% of
the variation in nitrate flux with precipitation was explained by the power function regressions,
both regressions were highly significant.
5000
4000
a
•a
OH
3000
X
200°
1000
crop NCI Flux = 6.62X"0, R2 = 0.46, P = 0.0006
forest N03 Flux = 0.000280X"6, R2 = 0.77, P = 0.00003
10
15 20 25 30 35 40
45 50 55
Precipitation (cm/spring)
Figure 1. Spring nitrate fluxes from a Rhode River, Maryland cropland watershed (round
dots and dashed line) and a Rhode River, Maryland forested watershed (square
symbols and solid line) as a function of depth of precipitation per spring. Each
symbol represents the integrated flux of nitrate for one complete season.
Atomic Ratios of Nitrogen to Phosphorus in Watershed Fluxes
Now that we have seen that nutrient fluxes vary with precipitation and land use, an
obvious question is: how do the ratios of nutrient fluxes vary with precipitation and land use?
The atomic ratios of total nitrogen (TN) to total phosphorus (TP), total inorganic nitrogen (TIN)
to total inorganic phosphorus (TIP), and total organic nitrogen (TON) to total organic
phosphorus (TOP) in spring watershed discharges all decline with increasing precipitation. For
the cropland watershed, these data are shown in Figure 2. Since the ratios involve two
independent variables, there is considerable scatter in the data, but all three regressions are
significant. In very dry springs, the nitrogen to phosphorus ratios are above the Redfield ratio of
15 to 16, but in very wet springs the ratios are below the Redfield ratio. Thus, a receiving water
such as the Rhode River shifts from potential phosphorus limitation in dry springs to potential
nitrogen limitation in wet springs.
314
-------
Somewhat surprisingly, the ratio of TIN to TIP in spring cropland discharges increases
less rapidly than in forest land discharges. Thus, the ratio of the cropland ratio to the forest land
ratio decreases about an order of magnitude with precipitation volume (Figure 3).
50
P-
H 40
o
•-P
03
O
1
PL,
H
20
10
0
80
60
_o
30
P-
o
H 25
O 20
.° 15
P^ 10
o
* TN/TP = 60.0el \ R" = 0.51, P = 0.006
* •
« TIN/HP = 114e('ao641x), R2 = 0.48, P=0.02
I
• TON/TOP = 19. leta°197X), R2 = 0.20, P=0.05
10
20 30 40 50
Precipitation Volume (cm/spring)
Figure 2. Atomic ratios of nitrogen to phosphorus in the integrated spring discharges from a
Rhode River, Maryland cropland watershed as a function of depth of precipitation. TN =
total N, TP = total P, TIN = total inorganic N, TIP = total inorganic P, TON = total
organic N, TOP = total organic P.
315
-------
- Ratio of Ratios = -0.223X + 11.37, R2 = 0.44, P = 0.002
12 h- '
PL,
l««4
I -
P
O
O 6 ^-
c+_ 4 i
O :
-
10 20 30 40 50
Precipitation Volume (cm/spring)
Figure 3. The ratio of the atomic ratio of total inorganic nitrogen to total inorganic
phosphorus in the integrated spring discharges from a Rhode River, Maryland
cropland watershed to the atomic ratio of total inorganic nitrogen to total
inorganic phosphorus in the spring discharges from a Rhode River, Maryland
forested watershed, as a function of depth of spring precipitation.
Atomic Ratios of Inorganic Nitrogen and Phosphorus to Dissolved Silicate
As we move from very dry springs to very wet springs, the ratio of TIN to dissolved
silicate (DSi) increases for both forest land and cropland (Figure 4). The ratios are much higher
and the increase much more rapid for cropland than for forest land. Dissolved silicate inputs to
receiving waters are an important factor in controlling phytoplankton taxonomic composition.
Diatoms require dissolved silicate, while many other groups of planktonic algae do not. For
diatoms, the Redfield atomic ratio of nitrogen to silicon is 1.0 to 1.3 (Correll 1987). All of the
measured discharges in Figure 4 had ratios well below the Redfield ratio, but cropland
discharges in wet springs began to approach this ratio.
The atomic ratio of TIP to DSi in watershed spring discharges also increased in wet
springs (Figure 5). This ratio was also higher for cropland discharges than those of the forest
discharges, but was usually lower than the Redfield ratio of about 0.075. Thus, for the Rhode
River, nitrogen and phosphorus are usually more likely to be limiting than dissolved silicate.
316
-------
0.7
_ crop TIN/DSi = 0.0114X + 0.050, R2= 0.44, P = 0.01
forest TIN/DSi = 0.00020X1'53, R2 = 0.56, P = 0.05
00
Q
Q
. ,-H
e
0.5
0.4
0.3
0.2
10
20
25
30
35
40
45
Precipitation (cm/spring)
Figure 4. Atomic ratios of total inorganic nitrogen to dissolved silicate in the integrated
spring discharges from a Rhode River, Maryland cropland watershed (round dots
and dashed line) and a Rhode River, Maryland forested watershed (square
symbols and solid line) as a function of depth of precipitation per spring.
Implications for Temporal and Spatial Patterns of Hypoxia in Receiving Waters
When watershed discharges mix into a receiving water body, the result is a spatial pattern
of nutrient concentrations that changes over time as the watershed input fluxes change.
Typically, in temperate latitudes, most of the watershed discharge is in the winter and spring. As
these flows mix into the receiving waters, the concentrations of nutrients are attenuated by such
processes as dilution, assimilation, dissimilation, and precipitation. Nutrient concentrations are
also augmented by such processes as nutrient recycling in the water column and regeneration
from bottom sediments. Within the main discharge plume, it may be that no nutrient is limiting,
so light penetration becomes the limiting factor for phytoplankton growth. However, as the
plume mixes into adjacent waters and attenuation of nutrient concentrations occurs, zones of
nutrient limitation will develop. If nitrogen or phosphorus limitation develops, then the
productivity of the overall plant population will be limited, barring the effects of nitrogen fixers.
If silicate limitation develops, then diatom populations will be limited, but not other plant forms.
It is the zone of either no nutrient limitation or only minor nutrient limitation where the potential
317
-------
for hypoxia is strongest. The longer favorable nutrient conditions prevail in a given area, the
greater is the probability of hypoxia.
0.140
[_ crop TIP/DSi = 0.0000136X217, R2 = 0.65, P = 0.02
forest TIP/DSi = 0.0000452X1'46, R2 = 0.45, P = 0.04
0.120 1—
'cJo 0.100 i—
O [_
£3
£3, 0.080 :—
O l_
1
^ 0.060 i
O
"P
O
<; 0.040 j—
r » * ^---^.
0.020 i— ---"""
. • «
10 15 20 25 30 35 40 45
Precipitation (cm/spring)
Figure 5. Atomic ratios of total inorganic phosphorus to dissolved silicate in the
integrated spring discharges from a Rhode River, Maryland cropland watershed
(round dots and dashed line) and a Rhode River, Maryland forested watershed
(square symbols and solid line) as a function of depth of precipitation per spring.
Hypoxia in Chesapeake Bay normally occurs in two distinctly different spatial/temporal
patterns. One is in the deep water down the middle of the Bay. Here the water is usually
vertically stratified and conditions are ripe for dissolved oxygen consumption to exceed inputs.
The mid-Bay zone is too deep for photosynthetic oxygen production and the diffusion and
advection of dissolved oxygen from the surface water layer are slow processes. A rain of
organic matter from over-enriched surface waters triggers hypoxia. This hypoxia is usually of
long duration. The second pattern is in shallow waters around the edges of the Bay and in the
upper reaches of the tidal tributary rivers. These areas are often so enriched with nutrients that
light is limiting. Here, in warm, calm weather very high rates of primary production result in
supersaturated dissolved oxygen levels in the afternoon, but hypoxic conditions in the late night
to early morning period. These diurnally hypoxic shallow waters are located near major nutrient
inputs around the periphery of the Bay.
318
-------
SUMMARY
Hypoxia in receiving waters results from excessive nutrient inputs. These excessive
nutrients often originate as diffuse sources from the watershed as the result of natural variations
in precipitation and changes in land use by humans. Higher precipitation in unusually wet
seasons brings about both increased water yield from the watershed and increases in the
concentration of nitrate, organic nitrogen, inorganic phosphate, and organic phosphorus in the
watershed discharges. The resulting increases in nutrient discharges due to precipitation
increases are usually non-linear, often approximating either a power or exponential function of
precipitation volume. Conversion of native vegetation to intensive agriculture also causes large
increases in nitrogen, phosphorus, and organic carbon discharges, but little or no increase in
dissolved silicate discharges to receiving waters. The combination of land use change and
variations in weather can bring about at least three orders of magnitude change in some nutrient
fluxes to receiving waters.
The atomic ratios of nutrients in watershed discharges also change with land use and
precipitation volume. Nitrogen to phosphorus ratios in cropland discharges decline from ratios
above the Redfield ratio in very dry seasons to well below the Redfield ratio in very wet seasons.
Ratios of TIN and TIP to dissolved silicate increase with the conversion from native vegetation
to agriculture and with increasing precipitation, but usually remain below the Redfield ratio.
Human conversion of native watershed vegetation to agricultural land uses greatly
increases the discharge of nitrogen and phosphorus, especially the inorganic fractions. This
increased nutrient loading to receiving water results in a much greater risk of hypoxia over much
larger areas. Interannual variation in seasonal precipitation controls the timing and spatial extent
of these risks from hypoxia.
ACKNOWLEDGEMENTS
Portions of this research were supported by the U.S. Environmental Protection Agency,
the U.S. National Science Foundation, the U.S. National Oceanic and Atmospheric
Administration Coastal Ocean Program, and the Smithsonian Environmental Sciences Program.
REFERENCES
Boynton, W.R. 2000. Impact of nutrient inflows on Chesapeake Bay. In: Agriculture and
Phosphorus Management, A.N. Sharpley (Ed.). Lewis, New York. pp. 23-40.
Boynton, W.R., J.H. Garber, R. Summers, and W.M. Kemp. 1995. Inputs, transformations, and
transport of nitrogen and phosphorus in Chesapeake Bay and selected tributaries.
Estuaries 18:285-314.
319
-------
Boynton, W.R., and W.M. Kemp. 2000. Influence of river flow and nutrient loads on selected
ecosystem processes. In: Estuarine Science, a Synthetic Approach to Research and
Practice, I.E. Hobbie (Ed.). Island Press, Washington, DC. pp. 269-298.
Correll, D.L. 1977. An overview of the Rhode River watershed program. In: Watershed
Research in Eastern North America, D.L. Correll (Ed.). Smithsonian Press, Washington
DC. pp. 105-124.
Correll, D.L. 1981. Nutrient mass balances for the watershed, headwaters intertidal zone, and
basin of the Rhode River estuary. Limnology and Oceanography 26:1142-1149.
Correll, D.L. 1987. Nutrients in Chesapeake Bay. In: Contaminant Problems and Management
of Living Chesapeake Bay Resources, S.K. Majumdar, L.W. Hall, Jr. and H.M. Austin
(Eds.). Pennsylvania Academy of Science, Philadelphia, Pennsylvania, pp. 298-320
Correll, D.L., I.E. Jordan, and D.E. Weller. 1999a. Effects of interannual variation of
precipitation on stream discharge from Rhode River sub watersheds. Journal of the
American Water Resources Association 35:73-82.
Correll, D.L., T.E. Jordan, and D.E. Weller. 1999b. Transport of nitrogen and phosphorus from
Rhode River watersheds during storm events. Water Resource Research 35:2513-1521.
Correll, D.L., T.E. Jordan, and D.E. Weller. 1999c. Effects of precipitation and air temperature
on phosphorus fluxes from Rhode River watersheds. Journal of Environmental Quality
28:144-154.
Correll, D.L., T.E Jordan, and D.E. Weller. 1999d. Effects of precipitation and air temperature
on nitrogen discharges from Rhode River watersheds. Water, Air and Soil Pollution
115:547-575.
Correll, D.L., T.E. Jordan, and D.E. Weller. 2000. Dissolved silicate dynamics of the Rhode
River watershed and estuary. Estuaries 23:188-198.
Correll, D.L., T.E. Jordan, and D.E. Weller. 2001. Effects of precipitation, air temperature, and
land use on organic carbon discharges from Rhode River watersheds. Water, Air and Soil
Pollution 128:139-159.
Havens, K.E., N.G. Aumen, R.T. James, and V.H. Smith. 1996. Rapid ecological changes in a
large, subtropical lake undergoing cultural eutrophication. Ambio 25:150-155.
Kemp, W.M., E.M. Smith, M. Marvin-DePasquale, and W.R. Boynton. 1997. Organic carbon
balance and net ecosystem metabolism in Chesapeake Bay. Marine Ecology Progress
Series 150:229-248.
320
-------
Peterjohn, W.T., and D.L. Correll. 1984. Nutrient dynamics in an agricultural watershed:
Observations on the role of a riparian forest. Ecology 65:1466-1475.
Rabalais, N.N., R.E. Turner, D. Justic, Q. Dortch, WJ. Wiseman, Jr., and B.K. Sen Gupta.
2000. In: Estuarine Science, a Synthetic Approach to Research and Practice, I.E. Hobbie
(Ed.). Island Press, Washington, DC. pp. 241-268.
321
-------
322
-------
IMPLICATIONS OF GLOBAL CLIMATE CHANGE FOR COASTAL AND
ESTUARINE HYPOXIA: HYPOTHESES, OBSERVATIONS AND MODELS FOR THE
NORTHERN GULF OF MEXICO
Dubravko Justic1, Nancy N. Rabalais2, and R. Eugene Turner1
ABSTRACT
A large-scale hypoxic zone (< 2 mg O2/L) in the coastal waters of the northern Gulf of
Mexico, recently exceeding 20,000 km2, overlaps with habitat and fishing grounds of
commercial fish and shrimp species. We have developed a simple eutrophication model that
accurately describes temporal variability in both surface and bottom water oxygen concentrations
at a station within the core of the Gulfs hypoxic zone. A sensitivity analysis revealed that the
model is highly sensitive to external forcing, yet sufficiently robust to withstand order of
magnitude changes in nitrate inputs from the Mississippi River. Model simulations suggest that
bottom water hypoxia in the northern Gulf of Mexico has intensified in recent historical time, as
a probable consequence of increased net productivity and subsequent increase in the vertical flux
of organic carbon. Apparently, the long-term increase in riverine nutrient fluxes has been the
primary factor controlling this historical decline in oxygen concentrations. Nevertheless, the
influence of climatic factors on hypoxia has been significant, and may increase further as a result
of global climate change. Projections from global circulation models suggest, for example, that
freshwater discharge from the Mississippi River to the coastal ocean would increase 20% if
atmospheric carbon dioxide (CC>2) concentration doubles. According to the models, the higher
Mississippi River runoff would be accompanied by an increase in winter and summer
temperatures over the Gulf Coast region of approximately 4°C and 2°C, respectively. For a
hypothetical 2 x CC>2 scenario, we estimated that the maximum monthly riverine nitrate input
would likely exceed 8xl06 kg/day. This value would be higher than any monthly nitrate input
value on record for the entire period 1954-2000. Thus, global climate change would likely have
a major impact on nutrient-enhanced productivity and dissolved oxygen dynamics in the northern
Gulf of Mexico. Model simulations suggest a close coupling between climate change and
hypoxia, and indicate a potential for future expansion of the Gulfs hypoxic zone as a result of
global warming. For example, in simulation experiments a 20% increase in annual runoff into
the Mississippi River relative to the 1985-1992 average resulted in a 50% increase in net primary
productivity of the upper water column (0-10 m) and a 30-60% decrease in summertime
subpycnoclinal (10-20 m) oxygen content within the present day hypoxic zone. These model
projections are in agreement with the observed increase in severity and areal extent of hypoxia
following the flood of 1993. Future expansion of the hypoxic zone in the northern Gulf of
Mexico would likely have important implications for coastal food webs.
1 Coastal Ecology Institute and Department of Oceanography and Coastal Sciences, Louisiana State University,
Baton Rouge, LA 70803, USA. 2 Louisiana Universities Marine Consortium, 8124 Hwy. 56, Chauvin, LA 70344,
USA.
323
-------
INTRODUCTION
There is a growing consensus among scientists that human activities, which have
increased atmospheric concentrations of carbon dioxide (CO2) by one-third during the last 100
years, may be responsible for an increase in the earth's temperatures. This so-called "global
warming" theory is not without challengers who argue that scientific proof is incomplete or
contradictory, and that there remain many uncertainties about the nature of climate variability
and climate change. Nevertheless, global temperature averages have increased by almost 1°C
during the last 150 years (Jones et al. 1999), and further temperature increase seems probable.
Global circulation models (GCMs), forced by enhanced greenhouse gas concentrations, have
projected the earth's temperature to rise 2 to 6°C over the next 100 years (IPCC 1996). Such an
increase in temperature would likely produce an enhanced global hydrologic cycle that would be
manifested in altered freshwater runoff. This hypothesis is supported by several lines of
evidence, including "paleofloods", decadal trends in the freshwater runoff and GCM scenarios.
In the United States, there is historic evidence suggesting that a change in climate
enhances the frequency of extreme flood events. An analysis of a 5,000-year-old geological
record for the southwestern United States (Ely et al. 1993) suggested that floods occurred more
frequently during transitions from cool to warm climate conditions. Apparently, even modest
changes in climate were able to produce large changes in the magnitude of floods. Additional
evidence in support of the above hypothesis comes from a 7,000-year-old record of over-bank
floods in the upper Mississippi River tributaries (Knox 1993). Approximately 3,300 years ago,
an abrupt shift in flood behavior occurred, producing frequent floods of magnitudes that now
only recur every 500 years or longer. Also, an analysis of data collected by the U.S. Geological
Survey indicates statistically significant increasing trends in monthly streamflow during the past
five decades across most of the conterminous United States (Lins and Michaels 1994). These
results seem to support the hypothesis that enhanced greenhouse forcing produces an enhanced
hydrologic cycle. One of the GCM studies (Miller and Russell 1992) examined the impact of
global warming on the annual runoff in 33 of the world's largest rivers. For the 2 x CO2
scenario, runoff increases were projected in all studied rivers in high, northern latitudes with a
maximum increase of 47 %. At lower latitudes, there were both projected increases and
decreases, ranging from +96% to -43%. Significantly, the model projected an increase in the
annual runoff in 25 of the 33 simulated rivers.
The northern Gulf of Mexico (Figure 1), which receives inflows from the Mississippi
River, the sixth largest river in the world (Milliman and Meade 1983), is one of the coastal areas
that may experience increased freshwater and nutrient inputs in the future. According to the
GCM study referenced above (Miller and Russell 1992), the annual Mississippi River runoff
would increase about 20% if the concentrations of atmospheric CO2 doubles (Figure 2). This
hydrologic change would be accompanied by an estimated increase in summer and winter
temperatures over the Gulf Coast region of 2°C and 4°C, respectively (Giorgi etal. 1994). The
higher runoff was projected for the May-August period, with an annual maximum most likely
occurring in May. While there are no other GCM estimates of projected Mississippi River
runoff, this result is in general agreement with a separate projected 2 x CO2 scenario increase in
rainfall over the Mississippi River drainage basin (Ibid.).
324
-------
In this work, we review probable implications of climate change for the Gulf of Mexico
hypoxic zone, focusing on two areas. First, we examine the role of climatic and anthropogenic
factors in the historical increase of hypoxia, and, second, we discuss model scenarios linking
hypoxia to global climate change. In this analysis, we use our previously published physical-
biological model (Justic etal. 1996, 1997) and extensive, long-term data sets collected at a fixed
station within the core of the Gulf of Mexico hypoxic zone (C6, Figure 1).
94°
93°
92
Figure 1. Map of the northern Gulf of Mexico showing station grid and location of station C6.
Shaded areas represent the distribution of hypoxic (< 2 mg O2/L) bottom waters during
August 1988 and July 1993.
Sensitivity of the Northern Gulf of Mexico Coastal Waters to Global Climate Change
The combined discharges of the Mississippi and Atchafalaya Rivers account for 98% of the total
freshwater inflow into the northern Gulf of Mexico (Dinnel and Wiseman 1986). The nutrient-
rich plumes of these two rivers rapidly form the Louisiana Coastal Current that flows
predominantly westward along the Louisiana coast, and then southward along the Texas coast.
Riverine nutrients are confined within the upper 10 m by a strong seasonal pycnocline (Aat= 4 -
10 kg/m3), which persists from April through October (Rabalais et al. 1991). Given this physical
setting, it is not surprising that biological processes in the northern Gulf of Mexico are strongly
influenced by the pattern and relative magnitude of riverine freshwater runoff (Justic et al. 1993).
Changes in the areal extent of hypoxic (< 2 mg O2/L) bottom waters provide a representative
example of the riverine influence on coastal productivity processes (Figure 1). The northern
Gulf of Mexico is presently the site of the largest (> 20,000 km2) and most severe coastal
hypoxic zone in the western Atlantic Ocean (Rabalais et al. 1999). Hypoxia normally occurs
325
-------
M urray
Orange
Nile
M agdalena
Kolym a
Indus
San Francisco
Colorado
Am ur
Zaire (Congo)
Yangtze
Yellow
M ekong
Yukon
Tigris-Euphrates
St. Lawrence
Orinoco
LaPlata
Ob
Niger
M ississippi
Lena
M ackenzie
Irrawaddy
Hsi Chiang
Fraser
Danube
Colum bia
Brahm a-Ganges
Am azon
Y^^7A/\
Y///////A7A7,
Y///////A
'/////A\
'//////A
'///////A
'/////////A
^////A -
'//////A
'/////A
Figure 2. Average annual discharges and projected 2 x CC>2 discharges for 33 of the world's
major rivers (data obtained from Miller and Russell 1992). Mississippi River is indicated
by an arrow.
from March through October in waters below the pycnocline, and extends between 5 and 60 km
offshore (Rabalais et al. 1996). However, during the drought of 1988 (a 52-year low discharge
record of the Mississippi River), bottom water oxygen concentrations were significantly higher
than normal and formation of a continuous hypoxic zone along the coast did not occur in
midsummer (Figure 1). The opposite behavior was observed following the flood of 1993 (a 62-
year maximum discharge for August and September) when the areal extent of summertime
hypoxia doubled with respect to the average hydrologic year (Rabalais et al. 1998). Hypoxia in
the coastal bottom waters of the northern Gulf of Mexico develops as a result of the synergistic
326
-------
interaction of high surface water primary productivity yielding a high carbon flux to the
sediments, and the high stability of the water column. The 1993 extreme hypoxia event was
associated with both an increased stability of the water column and nutrient-enhanced primary
productivity, caused by the greatly increased nutrient concentrations and phytoplankton biomass
in the coastal waters due to the Mississippi River flood (Dortch 1994, Rabalais et al. 1998).
Coupling Between Climate and Hypoxia
Climate change, if manifested by increasing riverine freshwater inflow, may affect
coastal and estuarine ecosystems in several ways. First, changes in freshwater inflow will affect
the stability of the water column, and this effect may be enhanced due to increases in sea surface
water temperatures. Vertical density gradients are likely to increase that could further decrease
vertical oxygen transport and create conditions in the bottom water more favorable for the
development of severe hypoxia or anoxia (Justic et al. 1996). Second, the concentrations of
nitrogen (N), phosphorus (P), and silicon (Si) in riverine freshwater inflows are typically an
order of magnitude higher than those in coastal waters (Justic et al. 1995a, 1995b). The mass
fluxes of riverine nutrients are generally well correlated with integrated runoff values (Turner
and Rabalais 1991, Goolsby et al. 1999). Consequently, nutrient inputs to the coastal oceans are
expected to increase as a result of increasing riverine runoff. This could have an immediate
effect on the productivity of coastal phytoplankton. Third, the stoichiometric ratios of riverine
nutrients, Si:N, N:P and Si:P, may differ from those in the coastal oceans (Justic et al. 1995a).
Increased freshwater inflow, therefore, may also affect coastal phytoplankton communities by
increasing or decreasing the potential for single nutrient limitation and overall nutrient balance
(Smayda 1990, Dortch and Whitledge 1992, Justic etal. 1995a, 1995b, Turner etal. 1998).
Thus, it appears that there is a plausible link between projected global climate change and the
productivity of river-dominated coastal waters.
METHODS
The Study Site
Our study area encompasses the Louisiana coastal waters (Figure 1). Station C6, located
in the inner section of the hypoxic zone, was used as a reference site for our studies of the
potential impacts of climate change. This site was chosen because it possesses the longest and
most consistent oceanographic data records (1985 - present) available for the northern Gulf of
Mexico. Three distinct oceanographic features of this region facilitated the application of a two-
box modeling scheme. First, from the beginning of April to the end of October, a strong
pycnocline (Aat= 4-10 kg/m3) is typically found at the average depth of 10 m (Rabalais et al.
1991). Because the total water depth is only about 20 m, the pycnocline virtually divides the
upper and the lower water column into two distinct water bodies of approximately equal
volumes. Second, the horizontal oxygen transport in the inner section of the hypoxic zone
appears to be of lesser importance than the vertical oxygen transport. This is suggested by a high
coherence between changes in vertical temperature gradients and changes in bottom oxygen
concentration. In contrast, a strong tidal signal, which would indicate horizontal transport, is not
327
-------
present in the periodograms of oxygen data series from station C6 (Rabalais etal. 1994). Also,
maximum lateral displacement of water parcels that can be expected due to diurnal and
semidiurnal currents is only about 3 km (Ibid.), which is not likely to affect the inner section of a
60 km wide hypoxic zone. Third, because of the high turbidity of the continental shelf waters
near the Mississippi River, primary productivity below the depth of 10 m is low (Lohrenz et al.
1990), and may be considered insignificant when compared to vertical oxygen transport.
Approach to Modeling
For this study, we adopted our previously published two-box model, which assumes
uniform properties for the water layers above and below the average depth of the pycnocline
(Justic et al. 1996). The model includes relevant physical and biological processes that affect
oxygen cycling in shallow, river-dominated, coastal waters. A brief presentation of the most
important theoretical formulations is given below.
The net productivity of the upper water column (NP, g (VnWday, 0 - 10 m) is described
by the expression
NP = Fot + D0 + INTS
S
where Fot is the total air-sea oxygen input flux (g O2/m2/day), D0 is the diffusive oxygen flux
through the pycnocline to the bottom layer (g (VnWday), and INTS is the rate of change in the
oxygen content of the upper water column (g O2/m2/day), given as
INT, =
f(dO/dt) dz. ( 2 )
0
We computed the air-sea oxygen influx (F0) from a formulation proposed by Stigebrandt (1991)
that takes into account the effect of gas transfer due to bubbles:
Fa = V(O2-1.025 O2'). (3)
In the above expression, V denotes transfer velocity (m/day), O2 is the surface water dissolved
oxygen concentration (g (Vm3, 0-1 m), and O/ is the surface water dissolved oxygen saturation
value (g (Vm3, 0 - 1 m). Negative Fa values indicate that the oxygen flux is directed towards
the water column. The transfer velocity was computed from a formula given by Liss and
Merlivat (1986). The vertical diffusive flux of oxygen (D0) from the upper water column to the
lower water column was estimated from the equation:
Do = - Kz (cO2/dz) I z = depth of pycnocline ( 4 )
where Kz is the vertical eddy diffusivity (m2/second), O2 is dissolved oxygen concentration (g
(Vm3), and z is depth (m). We assumed that the only properties of the stratified water column
328
-------
controlling Kz are the turbulent kinetic energy dissipation rate (s) and the buoyancy frequency (=
Brunt- Vaisala frequency) (TV):
Buoyancy frequency N (1 /second) was calculated directly from the conductivity -temperature-
depth profiles (CTD) using the expression:
where g is the acceleration due to gravity (9.81 m/second2), /?„ is the average density of the water
column (kg/m3), and dp/dz is the vertical density gradient (kg/m4).
The total oxygen uptake in the lower water column (TR, g (VnWday, 10 - 20 m) was described
by the expression
TR = -INTb + D0 (7)
where D0 is the diffusive oxygen flux from the surface layer through the pycnocline
(g O2/m2/day), and INTb is the rate of change in the oxygen content in the lower water column
(g O2/m2/day), given as
20
INTb = f(dO2/dt) dz. ( 8 )
10
We assumed that the estimated TR value also accounted for a large portion of anaerobic
respiration. Most of the dissolved sulfide produced under anoxic conditions, for example, is
oxydized by free dissolved oxygen (e.g. Berner 1982). Conversion of oxygen to carbon
equivalents was accomplished using a ratio of 0.288 by weight (mol C : mol 62 = 106 : 138, RQ
= 0.77; Redfield etal. 1963).
Data
The data on temperature, salinity and dissolved oxygen concentration were obtained from
a series of monitoring cruises conducted during the period June 1985 - October 1993. Our
sampling station (C6, Figure 1) was occupied on a biweekly to monthly basis. Standard water
column profile data were obtained from a Hydrolab® Surveyor or a SeaBird® CTD system with
an SEE 13-01 (S/N 106) dissolved oxygen meter. The dissolved oxygen measurements were
calibrated with Winkler titrations (Parsons et al. 1984) that were periodically carried-out during
deployment of the instruments.
The daily-averaged discharge values for the lower Mississippi River at Tarbert Landing
(August 1954-May 2000) were provided by the U.S. Army Corps of Engineers. Those daily-
averaged discharges were inferred from data-adaptive models of discharge versus water level,
329
-------
whose accuracy is normally greater than 90% (Bratkovich et al. 1994). Monthly discharge
averages, used in the nitrate input calculations, were computed from the daily-averaged
discharge values. The monitoring station at Tarbert Landing is located in Mississippi, 13 km
downstream from the inlet channel to the Old River control structure, where one-third of the
Mississippi River is diverted to the Atchafalaya River. The discharge at Tarbert Landing,
therefore, accounts for only about 70% of the total Mississippi and Atchafalaya River discharge.
In this analysis, we used the monthly records (August 1954-May 2000) of nitrate
concentration at St. Francisville. St. Francisville is located in Louisiana, approximately 430 km
upstream from the Mississippi River Delta. The average monthly nitrate loads to the northern
Gulf were computed by multiplying the average monthly nitrate concentrations by the respective
monthly river discharge averages. Data sources and analytical methods used to determine nitrate
concentrations are discussed in Turner and Rabalais (1991) and Goolsby etal. (1999). Nitrogen
is often considered to be the limiting nutrient for growth of the estuarine and coastal water
phytoplankton (e.g. D'Elia et al. 1986). The data from the northern Gulf of Mexico indicated
that the frequency of stoichiometric nitrogen limitation was on the order of 30% (Justic et al.
1995a).
The data were subdivided into 1985-1992 and 1993 subsets. The 1985-1992 subset
included 2 years with above average annual discharge of the Mississippi River (1990 and 1991),
3 years with below average discharge (1987, 1988 and 1992), and 3 average hydrologic years
(1985, 1986 and 1989). Given the time-span of the data, we considered the 1985-1992 data
subset to be representative of the present day climate. The flood of 1993, in contrast, provided
us with the opportunity to examine conditions that may occur as a result of future climate
change. In this respect, the "natural experiment" of 1993 was used to validate the model-
generated scenarios.
RESULTS
Climatic Influences on Riverine Nitrate Discharge
The average nitrate discharge in the lower Mississippi River increased about 3-fold
between 1967-1982 and by about a factor of 2 between 1987-1992 (Figure 3). From 1964-2000,
the average nitrate concentration also increased about 3-fold while the average discharge
increased about 40%. Partitioning the observed trend in nitrate discharge between the two
components, nitrate concentration and river flow rate, revealed that about 80% of the observed
increase in nitrate discharge is due to the increase in nitrate concentration (Figure 4). This
indicates that the historical increase in the anthropogenic nutrient inputs has had a far greater
impact on the lower Mississippi River nitrate discharge than has a change in climate.
Nevertheless, the influence of climatic factors on nitrate flux has been significant and may
increase further as a result of global climate change. This argument is supported by two lines of
evidence: First, the residual component of nitrate discharge, obtained by removing trends from
the time-series, is controlled primarily by the variability in
330
-------
o
30000
20000
10000
0
O)
O
3.0
2.5
2.0
1.5
1.0
0.5
0.0
•o
^
D
!^
o"
7x10
6x106
5x106
4x106
3x106
2x106
1x106
Figure 3. Monthly averages (1954-2000) of the lower Mississippi River discharge (Q), nitrate
concentration (N-NOs), and nitrate discharge (N-NOs). Smoothed curves are estimated
third order polynomial fits on 12-month weighted averages.
331
-------
13 100
Z 80
Z
CD
£ 60
2
c
•B 40
&
2
CD
DC.
20
1955 1960 1965 1970 1975 1980 1985 1990 1995 2000
Years
Figure 4. Partitioning of observed trends in the N-NOs discharge in the lower Mississippi River
into anthropogenic (N-NOs concentration) and climatic (discharge, Q), components.
Symbols indicate relative contributions (%) of the two components based on deviations
from averages of the base period 1954-1967.
discharge, i.e. climatic factors. Also, there is a highly significant relationship between river
discharge and nitrate concentration at the low end of the river flow rate spectrum; i.e. up to about
13,000 m3/second (Figure 5). Thus, the variations in nitrate discharge between flood and
drought years are significantly larger than the variations in river flow rate. This makes the lower
Mississippi River nitrate discharge potentially sensitive to future changes in the frequency of
extreme climatic events.
Our estimated nitrate discharge sensitivity value for the lower Mississippi River, i.e., the
percent increase in nitrate discharge that corresponds to a 1% increase in the average river flow
rate, was 1.16%. As suggested by Alexander etal. (1996), flux sensitivity values above 1% are
indicative of rivers where runoff from agricultural, urban, and forested lands is the main source
of nitrate. Because of the amplified influence of river flow rate on nitrate discharge in the
Mississippi River Basin (Figure 5), nutrient management efforts there in the future (Brezonik et
al. 1999, Goolsby et al. 1999) may be more challenging. Projections of the global circulation
models (GCMs) suggest that freshwater discharge from the Mississippi River to the coastal
ocean may increase 20% if atmospheric CC>2 concentration doubles (Miller and Russell 1992).
This scenario is in agreement with a projected increase in precipitation over the Mississippi
River Basin (e.g., Giorgi et al. 1994). Thus, a hypothetical 2 x CC>2 climate change scenario-
projected flow rate in the lower Mississippi River (19,000 m3/second) would be comparable to
that of the flood of 1993 (21,800 m3/second).
332
-------
6x10°
5x10
4x10°
x 3x106
O™
Z 2x106
1x10°
246 81012141618202224262830323436
2.0
1.5
1.0
0.5
0.0
246 81012141618202224262830323436
Figure 5. Relationships between river flow rate (Q), nitrate concentration (N-NO3), and nitrate
discharge (N-NOs) in the lower Mississippi River during 1983-2000. To demonstrate a
specific relationship between river flow rate and nitrate concentration, the flow rate
regime was subdivided into 18 categories, each corresponding to a range in flow of 2,000
mVsecond. Symbols denote average nitrate concentrations and discharges corresponding
to the respective river flow rate categories. Vertical bars represent + 1 standard error.
The horizontal lines in the lower panel denote the range of monthly averaged flow rates
during the drought of 1988 and the flood of 1993.
While a detailed discussion of complex watershed processes affecting nitrate
concentration and discharge remains beyond the scope of this paper, there are several ways in
which anticipated changes in precipitation and resulting increases in river flow rate may
333
-------
influence nitrate discharge. First, the increased precipitation will leach more nitrate from soils
into the tributaries and mainstream of the Mississippi River (Goolsby et al. 1999). Second,
unless riverine nitrate concentrations are reduced, the higher river flow will necessarily lead to
an increased nitrate discharges (Figure 5). Finally, the higher river flows will also decrease the
water residence times in canals, lakes and small streams in the upper parts of the watershed.
This would substantially reduce the possibility of nitrogen losses due to denitrification (Howarth
et al. 1996, Alexander et al. 2000), and ultimately result in a higher nitrate concentration in the
mainstem of the Mississippi River.
It is impossible at this time to predict with reliability future trends in nitrate
concentrations in the lower Mississippi River. If nutrient control programs within the
Mississippi River Basin are implemented (Brezonik et al. 1999, Goolsby et al. 1999), nitrate
concentration may decrease. In contrast, as indicated previously, nitrate concentrations can
increase in response to an increase in discharge without such controls (Figure 5). Thus, for the
purpose of this analysis, we assumed that the nitrate concentration will not change relative to
data for the period 1983-2000. Using the hypothetical 2 x CO2 climate scenario flow rate of
19,000 m3/second, we obtained an estimate of 2.47xl06 kg/day for the average annual 2 x CC>2
scenario nitrate discharge. If a 20% increase in discharge variability, relative to 1983-2000, is
also assumed, then the maximum nitrate discharge could exceed 8xl06 kg/day (Figure 6). This
latter value would be higher than any monthly nitrate discharge value on record for the entire
period 1954-2000 (Figure 3).
0)
8x10°
7x1 06
6x1 06
5x1 06
4x1 06
3x1 06
2x1 06
1x106
0
I
T
T
1954-1967 1968-1982 1983-2000
Period
2XCCL
Figure 6. Box-plots showing nitrate discharge (N-NOs) statistics for 1954-1967, 1968-1982, and
1983-2000, as well as model projections for a 2 x CC>2 climate.
334
-------
Model Calibration and Sensitivity Analysis
Our water column dissolved oxygen model was calibrated using the 1985-1993 data set
for the Mississippi River and Northern Gulf of Mexico. Given the time-span of this data set, we
considered it suitable for model calibration. A sensitivity analysis revealed that the model is
highly sensitive to external forcing, yet sufficiently robust to withstand an order of magnitude
change in nitrate loading between successive months, such as those encountered during the flood
of 1993.
Calibration results for surface and bottom water layer oxygen concentrations are
illustrated in Figure 7. For the bottom layer (10-20 m), the model agreed exceptionally well with
the observed values, both in terms of the annual and interannual variability. The agreement was
also very good for the surface layer (0-10 m), with the exception of 1990, 1991, and 1993, for
which the predicted (calibrated) summertime oxygen concentrations were somewhat higher than
observed.
O)
o
s
-I—'
c
o
o
o
o"
12
10
8
Surface
10
Bottom
1985 1986 1987 1988 1989 1990 1991 1992 1993
Years
Figure 7. Observed and predicted (calibration) monthly averages of surface (0-10 m) and bottom
(10-20 m) oxygen concentrations at station C6 for the period June 1985-November 1993.
335
-------
Simulated Decadal Changes in Surface and Bottom Oxygen Concentrations
Model simulations for a station within the core of the present day hypoxic zone (C6;
Figure 1) indicated a decadal trend of a slight increase in the annual average oxygen
concentrations in the upper water column (0-10 m) and a decrease in the lower water column
(10-20 m) (Figure 8). Specifically, the annual average oxygen concentration at 10-20 m depth
decreased from 6.6 mg/L in 1955-1965 to 6.2 mg/L in 1965-1975, to 4.9 mg/L in 1975-1985, to
4.2 mg/L during 1990-2000. As expected, the differences in summertime oxygen concentrations
for these periods were even greater. For example, the average oxygen concentration in the lower
water column during August decreased from 5.8 mg/L in 1955-1965 to 4.2 mg/L in 1965-1975,
to 1.9 mg/L in 1975-1985, to 0.9 mg/L during 1990-2000. The model identified the mid 1970s
as the start of the recurring hypoxia (< 2 mg O2/L) in the lower water column. Our results
should, however, be interpreted with caution, because the model only predicts the average
oxygen concentration for the entire lower water column. It is probable that hypoxia in the near
bottom waters was sporadically present during the late 1960s and early 1970s, and perhaps even
earlier. Our model results also suggest that the annual average oxygen concentration in the upper
water column (0-10 m) increased from 7.0 mg/L in 1955-1965 to 7.3 mg/L in 1965-1975, to 7.9
mg/L in 1975-1985, to 8.4 mg/L during 1990-2000.
Future Scenarios for the Northern Gulf of Mexico
In a series of modeling studies (Justic et al. 1996, 1997), we used a coupled physical-
biological model with climate forcing to examine the possible impacts of climate change on the
Gulf of Mexico hypoxic zone. Model simulations suggested that increased riverine freshwater
runoff (20%) and increased surface water zone temperatures (2-4°C) would significantly affect
the stability of the Gulf water column. Vertical density gradients between the upper (0-10 m)
and the lower (10-20 m) water column would increase, and would likely exceed values observed
during the peak of the flood of 1993 (Justic et al. 1996). Increased riverine nitrogen discharges
during the late spring would enhance the net productivity (NP) of the upper water column.
Following a 20% increase in the annual Mississippi River runoff, the annual NP value at a
station within the core of the hypoxic zone would increase about 53%, from 122 gC/m2/year
(1985-1992) to 187 gC/m2/year (Justic et al. 1997). This later value is 21% higher than the
annual NP value following the flood of 1993. Model results also suggested that summertime
subpycnoclinal (10-20 m) oxygen content would decrease 30-60%, relative to the 1985-1992
average. This would cause almost total oxygen depletion in the lower water column, which
could persist for several weeks (Figure 9). It is unlikely, however, that increased carbon
deposition would further enhance benthic and epibenthic respiration within the present day
hypoxic zone, since bottom waters are already severely depleted in oxygen. More likely, a
significant portion of the sedimented organic matter resulting from increased production will be
buried or, perhaps, exported from the area, leading to an expanded hypoxic zone.
336
-------
o
1955 1960 1965 1970 1975 1980 1985 1990 1995 2000
Years
Figure 8. Simulated changes in the annual average surface (0-10 m) and bottom (10-20 m) water
layer oxygen concentrations at station C6 for the period January 1955-May 2000. Shaded
area in the lower chart denotes hypoxic conditions (< 2 mg O2/L) in bottom waters.
Freshwater runoff, via its negative effect on salinity, is a critical parameter governing
biological processes in the northern Gulf of Mexico estuaries and coastal waters. The annual
yield of penaeid shrimp (Farfantepenaeus aztecus - brown shrimp, and Litopenaeus setiferus -
white shrimp) in the Gulf of Mexico is inversely related to the annual discharge of the
Mississippi River, perhaps because of the reduced estuarine salinities at high river flows (Turner
1992, Mulholland et al. 1997). Penaeid shrimp postlarvae require estuarine habitats with
salinities greater than 10 %o. In the case of Louisiana, salinities are primarily influenced by river
flow and precipitation. Mississippi River discharge affects the lower estuaries, while rainfall
affects the upper bays and estuaries. With heavy rains and/or high river flow, salinities in the
marshes are reduced. If salinities are reduced beyond acceptable conditions, the shrimp
postlarvae do not move as high up into the marshes, ultimately influencing the adult stock. This
is important because if the freshwater runoff increases further as a result of global warming,
estuarine salinities may decrease, possibly leading to reduced yields of shrimp and other species
favoring higher estuarine salinities. Temperature also influences shrimp growth. Growth is
inhibited in waters with a temperature below 20°C. Global warming, therefore, may expand the
region of high shrimp yield northwards, increasing shrimp harvest throughout the region,
assuming that salt marsh nursery areas are not negatively affected by other factors, such as water
level salinity changes.
337
-------
12
Figure 9. Seasonal changes in the integrated subpycnoclinal oxygen content (10-20 m) at station
C6 in the core of the hypoxic zone. Observed monthly averages for 1985-1992 and 1993
are compared to a Monte-Carlo simulation for a 2 x CC>2 climate change scenario. The
2 x CC>2 scenario probability plot is comprised of 2880 points.
The effects of hypoxia on demersal and benthic communities will likely intensify as
hypoxia stress worsens, due to either increase in areal extent, severity, or duration (see Rabalais
2002). Catches in trawls are negligible when the bottom water dissolved oxygen concentration
338
-------
falls below 2 mg/L (Pavela et al. 1983, Renaud 1986). Motile fishes and invertebrates migrate
from the bottom water hypoxic area or into the upper water column. Mass mortalities are likely,
however, if these animals are trapped against the shore by a large, anoxic water mass. Such a
scenario could become a serious occurrence in the northern Gulf of Mexico, if the areal extent of
hypoxia increases. Heavy mortalities already occur in the benthic infauna, and species diversity
is drastically reduced when ambient oxygen concentrations decrease below 0.5 mg/L (Gaston
1985, Boesch and Rabalais 1991, Rabalais et al. 1993). An increase in areal extent and severity
of hypoxia will also decrease recovery rates and reduce food resources (infauna) for recolonizing
demersal groups, such as the commercially important penaeid shrimps. In addition, alterations in
benthic community structure will have implications for sedimentary processes, benthic pelagic
coupling, and energy flow. Major alterations in benthic communities due to hypoxia stress,
especially a reduction in diversity and biomass, will certainly alter the productivity base that
leads to healthy fishery stocks.
CONCLUSIONS
The average nitrate discharge in the lower Mississippi River has increased from about 0.6
x 106 kg/day in 1954-1967 to about 2 x 106 kg/day in 1983-2000, which is a 3.3-fold increase.
During the same time period, the average nitrate concentration in the river increased about
2.3-fold (from 0.61 mg N-NOs/L to 1.37 mg N-NOs/L), while the average river flow rate
increased 40% (from 11,381 m3/second to 15,874 m3/second). Partitioning of the total increase
in riverine nitrate discharge revealed that about 80% of the observed increase can be explained
by the increase in nitrate concentration. Nevertheless, the residual component of nitrate
discharge is controlled primarily by the variability in runoff, i.e., climatic factors. Nitrate
concentration is also highly correlated with river flow rate at the low end of the flow rate
spectrum, up to about 13,000 m3/second. This particular relationship clearly affects nitrate
discharge, primarily by amplifying the variations in discharge between flood and drought years.
Consequently, future changes in the frequency of droughts or floods, or an overall change in
freshwater discharge, may substantially alter the input of nitrate into the northern Gulf of
Mexico.
We have developed a simple eutrophication model that accurately describes changes in
surface and bottom water layer oxygen concentrations for a station within the core of the Gulf of
Mexico hypoxic zone. A sensitivity analysis revealed that the model is highly sensitive to
external forcing, yet sufficiently robust to withstand order of magnitude changes in the nitrate
discharge of the Mississippi River. Model simulations indicated that bottom water hypoxia in
the northern Gulf of Mexico has intensified in recent historical time, as a probable consequence
of increased net productivity and subsequent increase in the vertical flux (deposition) of the
organic carbon. Apparently, the long-term increase in riverine nutrient discharge has been the
primary factor driving this historical decline in dissolved oxygen concentration. Our modeling
study supports the hypothesis that riverine nutrient inputs, via their influence on net productivity
in the upper water column, play a major role in controlling the development of bottom water
hypoxia and accumulation of organic carbon in coastal sediments.
339
-------
Projections of global circulation models (GCMs) suggest that freshwater discharge from
the Mississippi River to the coastal ocean would increase about 20% if atmospheric CC>2
concentration doubles. The higher Mississippi River runoff would also be accompanied by an
increase in winter and summer temperatures over the Gulf Coast region of 4.2°C and 2.2°C,
respectively according to the GCMs. For a hypothetical 2 x CC>2 global climate change scenario,
we estimated that the maximum monthly nitrate riverine discharge would likely exceed 8xl06
kg/day. This value would be higher than any monthly nitrate riverine discharge value on record
for the entire period 1954-2000. Thus, global climate change would likely have a major impact
on nutrient-enhanced productivity in the northern Gulf of Mexico. This, in turn, would likely
affect the oxygen cycling in this coastal ecosystem, which is presently the site of the largest (>
20,000 km2) and the most severe coastal hypoxic zone (< 2 mg O2/L) in the western Atlantic
Ocean. Model simulations suggest a close coupling between climate change and hypoxia, and
indicate a potential for future expansion of the Gulfs hypoxic zone as a result of global
warming. In simulation experiments, a 20% increase in annual runoff of the Mississippi River,
relative to a 1985-1992 average, resulted in a 50% increase in net primary productivity of the
upper water column (0-10 m) and a 30-60% decrease in summertime subpycnoclinal (10-20 m)
oxygen content within the present day hypoxic zone. Those model projections are in agreement
with the observed increase in severity and areal extent of hypoxia following the Great Flood of
1993.
ACKNOWLEDGMENTS
This research was funded in part by the U. S. Department of Energy's National Institute
for Global Environmental Change (NIGEC) through the NIGEC South Central Regional Center.
Other sources of support came from the U.S. National Oceanographic and Atmospheric
Administration (NOAA) Coastal Ocean Program grants for hypoxia studies (N-GOMEX 2000),
the U.S. Environmental Protection Agency Gulf Coast Regional Climate Change Assessment
Program, and the NOAA Nutrient Enhanced Coastal Ocean Productivity (NECOP) Program.
REFERENCES
Alexander, R.B., P.S. Murdoch, and R.A. Smith. 1996. Streamflow-induced variations in nitrate
flux in tributaries to the Atlantic Coastal Zone. Biogeochemistry 33:149-177.
Alexander, R.B., P.S. Murdoch, R.A. Smith, and G.E. Schwarz. 2000. Effect of stream channel
size on the delivery of nitrogen to the Gulf of Mexico. Nature 403:758-761.
Boesch, D.F., andN.N. Rabalais. 1991. Effects of hypoxia on continental shelf benthos:
comparison between the New York Bight and the northern Gulf of Mexico. In, Modern
and Ancient Continental Shelf Hypoxia, R.V. Tyson and T.H. Pearson (eds.).Geological
Society Special Publication No 58, London, pp. 27-34
340
-------
Berner, R.A. 1982. Burial of organic carbon and pyrite sulfur in the modern ocean: its
geochemical and environmental significance. American Journal of Science. 282: 451-
473.
Bratkovich, A., S.P. Dinnel, and D.A. Goolsby. 1994. Variability and prediction of freshwater
and nitrate fluxes for the Louisiana-Texas shelf: Mississippi and Atchafalaya River
source functions. Estuaries 17:766-778.
Brezonik, P.L., VJ. Bierman, Jr., R. Alexander, J. Anderson, J. Barko, M. Dortch, L. Hatch,
G.L. Hitchcock, D. Keeney, D. Mulla, V. Smith, C. Walker, T. Whitledge, W.J.
Wiseman, Jr. 1999. Effects of reducing nutrient loads to surface waters within the
Mississippi River Basin and the Gulf of Mexico. Topic 4 Report for the Integrated
Assessment of Hypoxia in the Gulf of Mexico. NOAA Coastal Ocean Program Decision
Analysis Series No. 18. NOAA Coastal Ocean Program, Silver Spring, Maryland.
D'Elia, C.J., J.G. Sanders, and W.R. Boynton. 1986. Nutrient enrichment studies in a coastal
plain estuary: phytoplankton growth in large-scale, continuous cultures. Canadian
Journal of Fisheries and Aquatic Science 43: 397-406.
Dinnel, S., and W.J. Wiseman, Jr. 1986. Freshwater on the Louisiana shelf. Continental Shelf
Research 6: 765-784.
Dortch, Q. 1994. Changes in phytoplankton number and species composition In Coastal
Oceanographic Effects of Summer 1993 Mississippi River Flooding, M. J. Dowgiallo
(ed.), National Oceanic and Atmospheric Administration, Coastal Ocean Program, Silver
Spring, Maryland, pp. 46-49.
Dortch, Q., and T.E. Whitledge. 1992. Does nitrogen or silicon limit phytoplankton production
in the Mississippi River plume and nearby regions? Continental Shelf Research 12:
1293-1309.
Ely, L.L., Y. Enzel, V.R. Baker, and D.R. Cayan. 1993. A 5000-year record of extreme floods
and climate change in the southwestern United States. Nature 262: 410-412.
Gaston, G.R. 1985. Effects of hypoxia on macrobenthos of the inner shelf off Cameron,
Louisiana. Estuarine, Coastal and Shelf Science 20: 603-613.
Giorgi, F., C. Shields-Brodeur, and G.T. Bates. 1994. Regional climate change scenarios
produced with a nested regional climate model. Journal of Climate 7: 375-399.
Goolsby, D.A., W.A. Battaglin, G.B. Lawrence, R.S. Artz, B.T. Aulenbach, R.P. Hooper, D.R.
Keeney, and G.J. Stensland. 1999. Flux and Sources of Nutrients in the Mississippi-
Atchafalaya River Basin: Topic 3 Report for the Integrated Assessment on Hypoxia in
the Gulf of Mexico. NOAA Coastal Ocean Program Decision Analysis Series No. 15.
NOAA Coastal Ocean Office, Silver Spring, Maryland.
341
-------
Howarth, R.W., G. Billen, D. Swaney, A. Townsend, N. Jaworski, K. Lajtha, J.A. Downing, R.E.
Elmgren, N. Caraco, T. Jordan, F. Berendse, J. Freney, V. Kudeyarov, P. Murdoch, and
Z-L Zhu. 1996. Regional nitrogen budgets and riverine N & P fluxes for the drainages
to the North Atlantic Ocean: natural and human influences. Biogeochemistry 35:75-139.
IPCC (Intergovernmental Panel on Climate Change). 1996. Climate Change 1995. The Science
of Climate Change. Contribution of Working Group I. Cambridge University Press,
Cambridge.
Jones, P.D., M. New, D.E. Parker, S. Martin, and I.G. Rigor. 1999. Surface air temperature and
its changes over the past 150 years. Reviews of Geophysics 37: 173-199.
Justic, D., N.N. Rabalais, R.E. Turner, and W.J. Wiseman, Jr. 1993. Seasonal coupling between
riverborne nutrients, net productivity and hypoxia. Marine Pollution Bulletin 26: 184-
189.
Justic, D., N.N. Rabalais, and R.E. Turner. 1995a. Changes in nutrient structure of river-
dominated coastal waters: Stoichiometric nutrient balance and its consequences.
Estuarine, Coastal and Shelf Science 40: 339-356.
Justic, D., N.N. Rabalais, R.E. Turner, and Q. Dortch. 1995b. Stoichiometric nutrient balance
and origin of coastal eutrophication. Marine Pollution Bulletin 30: 41-46.
Justic, D., N.N. Rabalais, and R.E. Turner. 1996. Effects of climate change on hypoxia in
coastal waters: A doubled CC>2 scenario for the northern Gulf of Mexico. Limnology and
Oceanography 41: 992-1003.
Justic, D., N.N. Rabalais, and R.E. Turner. 1997. Impacts of climate change on net productivity
of coastal waters: implications for carbon budgets and hypoxia. Climate Research 8:
225-237.
Knox, J.C. 1993. Large increases in flood magnitude in response to modest changes in climate.
Nature 361: 430-432.
Lins, H.F., and P.J. Michaels. 1994. Increasing U.S. streamflow linked to greenhouse forcing.
Eos Transactions, American Geophysical Union, 75(25).
Liss, P.S., and L. Merlivat. 1986. Air-sea gas exchange rates: Introduction and synthesis. In:
The Role of Air-Sea Exchange in Geochemical Cycling, Buat-Menard, P., (ed.),Reidel,
pp.113-127.
Lohrenz, S.E., M.J. Dagg, and T.E. Whitledge. 1990. Enhanced primary production in the
plume/oceanic interface of the Mississippi River. Continental Shelf Research. 10: 639-
664.
342
-------
Miller , J.R., and G.L. Russell. 1992. The impact of global warming on river runoff. Journal of
Geophysical Research 97: 2757-2764.
Milliman, J.D., and R.H. Meade. 1983. World-wide delivery of river sediment to the ocean.
Journal of Geology 91: 1-21.
Mulholland, P.J., G.R. Best, C.C. Coutant, G.M. Hornberger, J.L. Meyer, P.J. Robinson, J.R.
Stenberg, R.E. Turner, F. Vera-Herrera, and R.G. Wetzel. 1997. Effects of climate
change on freshwater ecosystems of the southeastern United States and the Gulf of
Mexico coast. Hydrological Processes 11: 949 - 970.
Parsons, T.R., Y. Maita, and M. Lalli. 1984. A Manual of Chemical and Biological Methods for
Seawater Analyses, Pergamon Press, New York, pp. 173.
Pavela, J.S., J.L. Ross, and M.E. Chittenden, Jr. 1983. Sharp reduction in abundance of fishes
and benthic macroinvertebrates in the Gulf of Mexico off Texas associated with hypoxia.
Northeast Gulf Science 6: 167-173.
Rabalais, N.N. 2002. The effects of seasonally severe hypoxia on continental shelf fauna.
Proceedings of the Sixth International Symposium on Fish Physiology, Toxicology, and
Water Quality. LaPaz, Mexico, January 2001. R.V. Thurston (Ed.). U.S. Environmental
Protection Agency, Ecosystems Research Division, Athens, Georgia, USA.
Rabalais, N.N., R.E. Turner, W.J. Wiseman, Jr., andD.F. Boesch. 1991. A brief summary of
hypoxia on the northern Gulf of Mexico continental shelf: 1985-1988. In:, Modern and
Ancient Continental Shelf Anoxia, R. V. Tyson and T. H. Pearson (eds).Geological
Society Special Publication No. 58, London, pp. 35-47.
Rabalais, N.N., L.E. Smith, E.B. Overton, and A.L. Zoeller. 1993. Influence of hypoxia on the
interpretation of effects of petroleum production activities. OCS Study/MMS 93-0022.
U.S. Dept. of the Interior, Minerals Management Service, Gulf of Mexico OCS Region,
New Orleans, Louisiana.
Rabalais, N.N., W.J. Wiseman, and R.E. Turner. 1994. Comparison of continuous records of
near-bottom dissolved oxygen from the hypoxia zone along the Louisiana coast.
Estuaries 17: 850-861.
Rabalais, N.N., R.E. Turner, D. Justic, Q. Dortch, W.J. Wiseman, Jr., and Sen B.K. Gupta. 1996.
Nutrient changes in the Mississippi River and system responses on the adjacent
continental shelf. Estuaries 19: 386 - 407.
Rabalais, N.N., R.E. Turner, W.J. Wiseman, Jr., and Q. Dortch. 1998. Consequences of the
1993 Mississippi River flood in the Gulf of Mexico. Regulated Rivers: Research &
Management 14: 161-177.
343
-------
Rabalais, N.N., R.E. Turner, D. Justic, Q. Dortch, and WJ. Wiseman, Jr. 1999.
Characterization of Hypoxia: Topic 1. Report for the Integrated Assessment on Hypoxia
in the Gulf of Mexico. NOAA Coastal Ocean Program Decision Analysis Series No. 15.
NOAA Coastal Ocean Office, Silver Spring, Maryland.
Redfield, A.C., B.H. Ketchum, and F.A. Richards. 1963. The influence of organisms on the
composition of seawater. In: The Sea, Vol 2. M.N. Hill, (ed.) John Wiley, New York,
pp. 26-77.
Renaud, M.L. 1986. Hypoxia in Louisiana coastal waters during 1983: implications for
fisheries. Fishery Bulletin 84: 19-26.
Smayda, T.J. 1990. Novel and nuisance phytoplankton blooms in the sea: evidence for global
epidemic. In: Toxic Marine Phytoplankton. E. Graveli, B. Sundstrom, R. Edler and D.M.
Anderson (eds.), Elsevier Science, New York. pp. 29-40
Stigebrandt, A. 1991. Computation of oxygen fluxes through the sea surface and net production
of organic matter with application to the Baltic and adjacent seas. Limnology and
Oceanography 36: 444-454.
Turner, R.E., and N.N. Rabalais. 1991. Changes in the Mississippi River water quality this
century - Implications for coastal food webs. BioScience 41: 140-147.
Turner, R.E. 1992. Coastal wetlands and penaeid shrimp habitat.. In Stemming the Tide of
Coastal Fish Habitat Loss, R. H.Stroud. (ed.), Proceedings of the 14th Annual Marine
Recreational Fisheries Symposium, Baltimore, Maryland, National Coalition for Marine
Conservation, Savannah, Georgia, pp 97 - 104
Turner, R.E., N. Qureshi, N.N. Rabalais, Q. Dortch, D. Justic, R.F. Shaw, and J. Cope. 1998.
Fluctuating silicate:nitrate ratios and coastal plankton food webs. Proceedings of the
National Academy of Science. USA 95: 13048-13051.
344
-------
EUTROPHICATION MODELING CAPABILITIES FOR ASSESSING WATER
QUALITY AND ECOLOGICAL ENDPOINTS
Robert F. Carousel1 and Rosemarie C. Russo1
ABSTRACT
A primary environmental focus for the use of mathematical models is for characterization
of the sources of nutrients and sediments and their relative loadings from large river basins, and
to assess the impact of alternative land uses and their changes from smaller sub-basins on
aggregate water quality in rivers, lakes, and estuaries. To assess from these modeling results,
additional models can be used to assess the probability of excessive algal blooms, low dissolved
oxygen (hypoxia), and related fish kills. For such a comprehensive evaluation, a linked
modeling system is required consisting of models that will simulate constituents and processes
necessary to evaluate nutrient budgets and cycles on land and in the aquatic environment.
Modeled processes include: hydrodynamics, sediment erosion and transport, water temperature,
oxygen and BOD dynamics, inorganic and organic nutrients, and growth/respiration of algae,
specified species within each trophic level, and toxicity of pollutants to modeled organisms,
indirect effects produced by consequent changes in grazing and predation pressures, changes in
decay rates and detritus and nutrient cycling, and dissolved oxygen. A pilot study was conducted
in the Tensas watershed, located in the northeast corner of the State of Louisiana. An evaluation
was conducted of using a watershed model linked to a water quality model linked to an
ecological model that provided full capabilities for simulating eutrophication for nutrients for use
in determining indications of ecological impairment with fish health as an endpoint. Results
indicate little sensitivity of chlorophyll a and Secchi to phosphate inputs except at peaks;
however, under a 50% reduction in total dissolved solids (TSS) these organisms became more
sensitive to phosphate inputs. Benthic detritivores (chironomids) and benthic fishes were
impacted by nutrients. Nutrient sensitivity appears at reduced TSS. High TSS appears to limit
algae.
Ecosystems Research Division, National Exposure Research Laboratory, Office of Research and Development,
United States Environmental Protection Agency, Athens, GA 30605 USA.
345
-------
INTRODUCTION
Nutrients and sediments discharged into the Gulf of Mexico, largely from the Mississippi
River Basin, are contributing to the formation of a zone of low dissolved oxygen (hypoxia) along
the coasts of Louisiana and Texas. It is estimated that over 90% of the phosphorus and Kjeldahl
nitrogen loads to the Gulf originate from the Mississippi Basin (Burkart and James 1999). A
major environmental concern is the impacts of these nutrient/sediment discharges and the likely
effects of land management alternatives on the formation of the hypoxia zone. The use of
mathematical models is one type of assessment that can be used for evaluating the impacts of
nutrient loadings from the Mississippi Basin/Region and subsequent interaction in the Gulf of
Mexico through the linking of watershed hydrologic and water quality simulation models within
the Mississippi Region. One such application would consist of watershed hydrologic and water
quality components (e.g., the U.S. EPA Hydrological Simulation Program - FORTRAN (HSPF)
(Bicknell et al. 1997)), lake/impoundment hydraulic and water quality components (e.g., U.S.
Army Corps of Engineers CE-QUAL-W2 program (Cole and Buchak 1995) and ecological
interactions (e.g., EPA AQUATOX model (Park 1998)).
In order to develop management strategies for reducing nutrient transport to the Gulf, it is
necessary to evaluate and quantify the point and nonpoint nutrient sources delivered to the Gulf
from individual watersheds in the Mississippi Basin. The specific uses that the management
strategies are designed for are the following:
1) Evaluation of methods to reduce sediment and nutrient loads within specific tributary
river basins in the Region.
2) Performance of TMDLs and study of local water quality impacts of sediment and nutrient
load reductions in tributary river basins.
3) Estimation of total nutrient loads to the Mississippi River and the Gulf of Mexico to
allow assessment of the relative benefits of reductions in tributary river basins on the total
loads.
The objective of this work is to demonstrate how linking watershed and water quality and
ecological models together can be used for evaluating the impacts of nutrient/sediment
discharges from individual watersheds within the Mississippi Region. The proposed watershed
modeling system is designed to enable researchers and regulators to evaluate the impacts of such
things as human activities, agricultural practices, land use changes/policies, regulatory actions,
and wetland conversion and restoration. This modeling system can be used to facilitate the
following assessments: 1) estimation of basin-wide nutrient loads, 2) prediction of changes in
nutrient loading from changes in the watershed, and 3) prediction of local water quality and
biological responses to nutrient inputs.
346
-------
OVERVIEW OF MODELING METHODOLOGY AND BASIN REPRESENTATION
The methodology and approach has similarities to the basin modeling effort in the
Chesapeake Bay Region (i.e., the Chesapeake Bay Watershed Model (CBWM) (NVPDC, 1983;
Linker et al. 1993; Donigian et al. 1994). While this project's long term extent and objectives
are similar to those for the Chesapeake Model, the basin modeling demonstrated in this effort has
two major differences including evaluation of local water quality (i.e., in upstream areas) and an
ecological endpoint (i.e., biological interactions) instead of traditional water quality constituents
(e.g., DO, BOD/COD, etc.).
This section contains an overview of the approach to developing models for use in a
demonstration example for the Tensas watershed located in the Lower Mississippi, including the
models selected to represent parts of the system, the physical domains of the models, model
linkages, and the constituents and processes required. A conceptual approach is illustrated in
Figure 1. For this demonstration three models, the U.S. EPA HSPF watershed/river water
quality modeling program linked to the Army Corps' CE-QUAL-W2 reservoir water quality
model linked to the EPA AQUATOX ecological model will be used to evaluate sediment and
nutrient loadings for the Tensas watershed. The modeling methodology has been designed to
provide deterministic estimates, on a continuous time basis, of river basin sediment and nutrient
loads, and the impacts of transport and transformation (water quality and selected ecosystem
impacts) in rivers and lakes.
for Water Quality
l.*k
IHl
&
r * l|r%r r* ssif
l,-.,\\K.ims-^
XV O \}i!iK-l
« .4 l-'yfil
Figure 1. Conceptual modeling framework for water quality and ecological impacts.
347
-------
Watershed/land area modeling for the Tensas is done on a land use and land cover basis, with
detailed nutrient cycling and budget simulation for land types that are the principal sources of
nutrients; i.e., major crops, pasture, and other agricultural activities such as animal confinement
and feeding areas. Forest, wetland, and urban land categories will also be included in the model.
Land segments, which will experience the same meteorologic inputs and have the same
parameter set, will be relatively large; i.e., on the order of the U.S. Geological Survey (USGS)
cataloging unit or smaller. Stream segments (and the tributary land areas, delineated by land
category) will be smaller to resolve the locations of nonpoint and point sources of pollutant
loading and resulting water quality impacts better. Each river basin will be modeled as a
separate entity, but region-wide databases will be developed to support the modeling of all river
basins. This includes long time series of rainfall, evaporation, other meteorologic data inputs,
discharge and water monitoring data for rivers and lakes, crop-specific nutrient application rates
and timing, and other agricultural activities.
The models will simulate the constituents and processes necessary to evaluate nutrient
budgets and cycles on land and in the aquatic environment. These include flow, sediment
erosion and transport, water temperature, oxygen and BOD dynamics, inorganic and organic
nutrients, and growth/respiration of algae.
COMPONENT MODELS
U.S. EPA HSPF
Overview
The Hydrological Simulation Program-FORTRAN, known as HSPF, is a mathematical
model developed under EPA sponsorship for use on digital computers to simulate hydrologic and
water quality processes in natural and man-made water systems. It is an analytical tool which
has application in the planning, design, and operation of water resources systems. The model
enables the use of probabilistic analysis in the fields of hydrology and water quality
management. HSPF uses such information as the time history of rainfall, temperature,
evaporation, and parameters related to land use patterns, soil characteristics, and agricultural
practices to simulate the processes that occur in a watershed. The initial result of an HSPF
simulation is a time history of the quantity and quality of water transported over the land surface
and through various soil zones down to the groundwater aquifers. Runoff flow rate, sediment
loads, nutrients, pesticides, toxic chemicals, and other quality constituent concentrations can also
be predicted. The model uses these results and stream channel information to simulate instream
processes. From this information, HSPF produces a time history of water quantity and quality at
any point in the watershed that can be used by other models to infer the impacts of land
management activities on water quality and/or ecological endpoints.
CE-QUAL-W2
The primary sources for the following description are the CE-QUAL-W2 User's Manual
(Cole and Buchak 1995) and Edinger and Wu (1999).
348
-------
Overview
CE-QUAL-W2 is a longitudinal-vertical hydrodynamic and transport model built for
long-term, time-varying water quality simulations of lakes, reservoirs, and estuaries. Because
the model assumes lateral homogeneity, it is best suited for relatively long and narrow water
bodies exhibiting longitudinal and vertical water quality gradients. CE-QUAL-W2 accurately
reproduces vertical and longitudinal water quality gradients and is capable of multi-decade
simulations. CE-QUAL-W2 currently includes water quality routines for 22 parameters. CE-
QUAL-W2 can be used to infer changes in circulation and water quality as well as provide
boundary condition data to embedded 3-D models or to near-field models such as PLUMES or
CORMIX. CE-QUAL-W2 has become a popular tool for simulation and analysis of water
quality problems in U.S. reservoirs. It is commonly used for reservoir studies by the U.S. Army
Corps of Engineers, U.S. Geological Survey, U.S. Bureau of Reclamation, Tennessee Valley
Authority, and U.S. EPA.
US EPA AQUATOX
Overview
AQUATOX (Park 1998) is an aquatic ecosystem model whose purpose is to represent the
fate of common pollutants (nutrients and sediment) and toxic contaminants and their effect on
the system biomass. There is a decided emphasis in AQUATOX on assessing the impact of a
trace toxic contaminant on the ecosystem. AQUATOX simulates aquatic ecosystems in streams
(a single reach), ponds, lakes, reservoirs, and artificial enclosures. AQUATOX is designed as a
non-dimensional point model and includes higher trophic levels (primary consumers, secondary
consumers, tertiary consumers). It simulates the effects of nutrient or sediment pollution with or
without concurrent contamination by a toxic compound. (The "without toxic" simulation is
generated by default and is treated as an experimental control.) Results are expressed in terms of
biomass concentration over time for user-selected species within each trophic level and provide
for completely mixed or horizontally stratified systems (two zones: epilimnion and hypolimnion
separated by a constant-depth thermocline). The model allows constant or time-variable loading
of nutrients, toxic contaminants, and sediment from a point-source, nonpoint source, or
precipitation. Input files containing time-variable loading of nutrients, sediments, and toxic
contaminant can be imported. AQUATOX simulates effects on user-specified species within
trophic levels including: acute toxicity of contaminant to modeled organisms, indirect effects
produced by consequent changes in grazing and predation pressures, changes in pollutant decay
rates and detritus and nutrient cycling, and changes in dissolved oxygen.
MODEL LINKAGES
There are effectively three models to be considered when developing the linkage issues.
The three models are HSPF(land), CE-QUAL-W2 (for lakes) and AQUATOX (for biological
interactions). The three models simulate different processes/constituents at different time and
space scales. Their physical domains and resulting transfer of information must be appropriately
349
-------
integrated to allow efficient operation and effective representation of the sub-basins of the
Mississippi region. These process/constituent, space, and time linkage issues and approaches for
the three possible interfaces (i.e., HSPF(land)-CE-QUAL-W2, HSPF(stream) - CE-QUAL-W2,
and CE-QUAL-W2 - AQUATOX) are discussed below.
HSPF(land) - CE-QUAL-W2 Linkage
In this linkage, nonpoint source loadings of flow, sediment, heat, and constituents are
generated by HSPF and transferred to water bodies (lakes) simulated by CE-QUAL-W2. For
modeling applications at this scale, an appropriate spatial linkage scheme consists of allocating
the total load generated from the area tributary to the lake to CE-QUAL-W2 "segments" by
prorating based on the length of shoreline in each segment.
Although HSPF allows the user to select among various intervals for the timestep of the
simulation; i.e., the interval of the internal model calculations, for most watersheds a one-hour
timestep is normally used. In addition to watershed size and hydrologic response times, the
appropriate timestep is also controlled by the availability of representative precipitation data.
For most watersheds, there are sufficient hourly precipitation stations in and near the watershed
to allow an hourly simulation interval. Since CE-QUAL-W2 operates on a variable timestep, it
will read the hourly inputs from HSPF and interpolate/adjust the values to match its internal
timestep.
Since the water quality state variables in CE-QUAL-W2 are very similar to those in
HSPF RCHRES, the corresponding data used for the HSPF(land) - HSPF(stream) linkage can be
used to complete this linkage.
HSPF(stream) - CE-QUAL-W2 Linkage
In most situations, this linkage will be implemented in both directions, since HSPF
RCHRES will provide the upstream inflows (and sometimes tributary inflows) to CE-QUAL-
W2, and then at the downstream end of a lake, HSPF will receive the CE-QUAL-W2 outflows
for transport and transformation in downstream reaches. The spatial linkage is straightforward,
and the temporal issues are relatively straightforward; they are complicated only by the variable
timestep in CE-QUAL-W2. While the constituents in the two water quality models are similar,
there are some differences in definition of organic material constituents. The principal difference
to be resolved is that HSPF includes separate refractory organic nitrogen, phosphorus, and
carbon variables, while CE-QUAL-W2 has refractory and labile organic matter variables, each of
which contains nitrogen, phosphorus, and carbon.
CE-QUAL-W2 - AQUATOX Linkage
In most situations, this linkage will be implemented in both directions, since CE-QUAL-
W2 will provide outflows for transport and transformation in downstream reaches and
350
-------
AQUATOX will provide updating changes in decay rates and detritus and nutrient recycling, and
changes in dissolved oxygen. The spatial linkage is straightforward, and the temporal issues are
relatively straightforward; they are complicated only by the variable timestep in CE-QUAL-W2.
POTENTIAL LINKAGE STRATEGIES
Linkage of HSPF with the candidate receiving water eutrophication models will result in
a modeling system that can be used to specify total maximum daily loads in complex watersheds
that include a wide variety of hydrologic and hydrodynamic conditions. The linked modeling
system will have many more capabilities than any of its components. It will enable a detailed
simulation of very complex watershed systems.
Choice of a linkage strategy involves making tradeoffs among ease of use for end users,
cost of developing the linked system and cost of supporting the linked system.
There are two steps that are common to all linkage strategies. First, parameters and other
input data required by the models must be evaluated. This evaluation includes broad topics such
as definition of parameters and unit systems along with detailed topics such as specification of
fields and records in a model's input file(s). Second, output data produced by the models must
be reviewed. Again, broad topics must be addressed. These topics include definition of output
variables and their units as well as detailed topics such as specification of fields and records in a
model's output file(s).
There are three potential techniques that could be used to link HSPF with the candidate
receiving water eutrophication models: (a) the eutrophication models could be loosely coupled
to HSPF; (b) the eutrophication models could be coupled to HSPF through an automated
modeling supervisor; (c) the eutrophication models could be tightly coupled to HSPF.
Loose coupling involves use of the output from one model to satisfy the input
requirements of the second model. Each model is available as originally developed. If the output
from one model is not in the format needed by the second model (as is most often the case), then
either a simple conversion utility may be supplied by the developers or the user may be directed
to import the data into a spreadsheet or database program, reformat it, and export it in the format
needed by the second model.
This approach was selected for the Tensas watershed demonstration because it has
relatively low development costs, moderate user costs, and moderate support costs. Lower
development costs are due to the fact that very little code is needed beyond the models
themselves. Moderate user costs are the result of the requirement that users must understand
many details of the models and their interaction with each other. Analysis and comparison of
results must be done using the model's native tools, if available. Support costs are mixed.
Support of the individual models can be done by the original developers or other experienced
351
-------
users. User groups may exist for the component models. Problems can be framed in terms that
the developer or other experienced user understands because the original model code is being
used. Resolution of inconsistencies between the model's conventions and world view can raise
questions. Support of the linkages between models can be more significant due to the ad hoc
nature of the connections. Details of file naming and location conventions can cause problems.
DEMONSTRATION SITE - TENSAS WATERSHED
The Tensas River flows through the upper northeast part of Louisiana and empties into
the Red River just upstream of the Red River's confluence with the Atchafalaya/Mississippi
Rivers. This study focuses on the basin above its confluence with Bayou Macon at River Mile
46 near Coolers Point, Louisiana. Figure 2 shows the location of the basin in Louisiana, and
Figure 3 shows the river and the extent of the basin as considered in this study. The Tensas
Basin is located primarily in the parishes of East Carroll, Franklin, Madison, and Tensas.
Historically, the predominant landform was of bottomland hardwoods that have been cleared and
drained for conversion to crop production. The resulting loss of wetlands and the application of
nitrogen in the form of fertilizers have resulted in significant water quality degradation. The
agricultural nutrients from this basin and others in the lower Mississippi basin are suspected as
primary causes in the formation of a zone of hypoxia along the inner continental shelf of the
Louisiana and Texas coasts in the Gulf of Mexico. The Tensas Basin is being evaluated as a
representative watershed to investigate the impact of these nutrients and to evaluate alternative
management practices.
Figure 2. Location of Tensas watershed.
352
-------
Figure 3. Map of Tensas watershed and river system.
353
-------
Hydrologic and meteorologic data
Precipitation
Within and near the Tensas Basin, the National Weather Service maintains numerous
hydrometeorologic data collection stations that have been in operation for more than 25 years.
Five hourly stations and six daily stations were identified that have extended periods of record
and no more than 10% missing data over their time spans. These stations, their periods of
record, and their locations relative to the basin are shown in Figure 4.
inuriy I
^nffusifi
TftlUill B*i:-fl
Figure 4. Locations of Meteorologic and Streamflow Stations.
354
-------
HSPF generally uses measured pan evaporation to derive an estimate of lake evaporation, which
is assumed to be the potential evapotranspiration (PET); i.e., PET = (pan evap) X (pan
coefficient.) The simulated actual evapotranspiration is computed by the program based on the
model algorithms, ET parameters, and the input PET data.
Pan evaporation data are available at three locations near the Tensas Basin. These
stations are shown on a map of the area in Figure 4. Pan evaporation data are less variable than
rainfall; therefore, a basin of this size generally requires only one or two records.
Examination of mean annual pan evaporation (60 inches) and lake evaporation (45
inches) data for the region indicates that a pan coefficient of 0.75 should be used (Environmental
Data Service 1979).
Streamflow
To calibrate the model, a reliable, long-term record of measured streamflow data is
compared with simulated streamflow. Examination of the USGS surface water data for
Louisiana provided numerous gauging stations in the basin; however, only one station has long-
term streamflow data. The continuous streamflow station at Tendal, Louisiana (see Figure 4)
provides a reasonably long-term (> 50 years) record. The record is described by the USGS as
"fair", and it represents only 27% (282 sq. mi.) of the total basin area. Therefore, the calibration
is based only on this portion of the basin. The other gages in the basin were either very short-
term, or measured only river stage. During a field trip to the basin, evidence of other streamflow
stations was investigated, and it was determined that data at additional locations (most likely) do
not exist.
Other data
Other data types often required for hydrologic simulation are point source inflows
(sources of water) and diversions (removal of water). Since land use in the Tensas Basin is
predominantly agricultural, irrigation withdrawal and return flows are likely present in the
watershed; however, based on information obtained during a field trip to the watershed and
discussions with local and state agency representatives, irrigation was determined to be an
insignificant loss mechanism. Also, no significant point sources (e.g., sewage treatment plants)
or diversions (e.g., municipal water supply) were identified.
Measured water levels in wetlands are generally needed for comparison with simulated
values if the high water table/wetland version of HSPF-PERLND (Version 12) is to be calibrated
for the wetland portion of the basin. While three long-term USGS records of groundwater levels,
containing data at an approximate interval of one quarter (3 months) were identified as existing
within the basin (R. Seanor, personal communication), these data were determined to be
inadequate for wetland calibration because of their location and the scarcity of data.
355
-------
Segmentation and characterization of the watershed Basin and river segmentation
Land Segmentation
The purpose of segmenting the watershed is to divide the study area into individual land
segments that are assumed to produce a homogeneous hydrologic and water quality response.
The segmentation then allows the user to assign identical model parameter values to all parts of
the watershed that produce the same unit response of runoff (and other quantities such as
chemical constituents) for a uniform set of meteorologic conditions. Where the weather patterns
vary across a watershed, it is necessary also to divide the land segments by meteorology to
accurately reflect spatial meteorologic variability and its effect on the hydrology and water
quality of the watershed. For a watershed the size of the Tensas Basin, the meteorologic
variability is usually reflected in the difference in annual precipitation totals at various locations
in the watershed.
The Tensas Basin above Bayou Macon was segmented into 24 similarly sized areas based
primarily on drainage boundaries and stream reach locations (i.e., major tributary inflows and the
flow station at Tendal). These segments are shown in Figure 5. A level of meteorological
segmentation (shown in Figure 6) was also imposed on the basin, by assigning the 24 segments
to five of the final rainfall records, based on distance from the rainfall station to the center of the
segment. Differences in other characteristics, such as soils and topography (e.g., slope and
elevation) were negligible, and were not considered in the segmentation.
The drainage boundaries and resulting segmentation summarized above was performed
by analyzing the basin outline and the EPA River Reach (RF1 and RF3) hydrography coverages.
All RF3 stream segments were assigned tributary area using a "Euclidean allocation" procedure,
which assumes that water drains to the nearest stream segment. The resulting 1,400 segments
were "simplified;" i.e., combined to produce (eventually) the final 24 segments by imposing
segment boundaries at major tributary inflows and other locations to create segments with stream
lengths on the order of 5-25 miles. Errors in these segments were corrected by using maps,
information obtained during a field trip to the basin, and the USGS ground-truthed drainage area
data for Louisiana (USGS 1971).
The land segmentation was performed iteratively with the stream segmentation, until an
appropriate level of segmentation was achieved. This task was made more difficult than usual
due to errors in the hydrography coverage and the extremely low slopes in the basin. As a result
of the low slopes, agricultural leveling, channelization, and other activities, most drainage
boundaries have been modified from their natural location, and now follow roads and manmade
levees. Furthermore, the actual flow direction and channel network was not always possible to
determine from the hydrography coverage, and no information could be derived from the Digital
Elevation Map (DEM) because of the flat topography.
356
-------
w
L/-'
Figures. Base land segmentation.
357
-------
' -MJlMyKpmn;
LAKE P'ftOvlp=r*C:F'
WIMNS6CRO55
Figure 6. Meteorologic segmentation.
358
-------
The river channel network in the Tensas Basin is the major pathway by which pollutants
are transported from the watershed. As such, it is important to accurately represent or
characterize the channel system in the HSPF model of the watershed. The river reach
segmentation requires consideration of river travel time, riverbed slope continuity, and entry
points of major tributaries. As stated above, this task was difficult due to the flat terrain, poorly
defined bayou drainage patterns near the main channel, and the manmade channelization that has
occurred in the watershed. Since the land segmentation was based largely on the channel
network (slopes and tributary inflow locations), the river segmentation was effectively
accomplished in conjunction with the land segmentation. Due to the questions about the flow
patterns, and lack of data, it was determined that only a limited set of tributary reaches could be
characterized and modeled. Therefore, fourteen river reaches were selected for simulation.
These reaches, shown in Figure 7, consist of nine reaches along the main stem of the Tensas
River and five tributaries. This Tensas River model represents approximately 120 miles of the
river channel from Swan Lake in the north to the confluence with Bayou Macon in the south.
Figure 7. River segments in the Tensas rive basin model.
Land use
359
-------
Land use affects the hydrologic response of a watershed by influencing infiltration,
surface runoff, and water losses from evaporation or transpiration by vegetation. The movement
of water through the system, and subsequent erosion and chemical transport, are all affected
significantly by the vegetation (i.e., crops, pasture, or forest).
The Tensas Basin is predominantly agricultural, with cropland comprising approximately
65% of the total area. The balance of the basin is comprised of forest (27%) and forested
wetland (5%). A minor fraction is covered by urban development, lakes, and other categories.
These data were obtained from the land use coverage contained in the BASINS (Lahlou et al.
1998) database for EPA Region 6. Based on this breakdown, the following four preliminary land
uses were selected for explicit representation in the model:
1) Forest
2) Wetland
3) Agriculture
4) Other
Once the land segmentation was finalized, the land use coverage was overlaid onto the
segmentation, and a data set consisting of the land use distribution (based on these four
categories) for each segment was generated.
The second step in deriving the final land use data was to break down the agricultural
category into major crops. The Natural Resources Inventory (NRI) (USDA 1992) database
provided detailed percentages of land uses for the entire Tensas Basin in 1982, 1987, and 1992.
Although these data did not have any spatial resolution, the agricultural land use category
resolution was very detailed, including percentages often different crops and other activities.
The NRI basinwide crop-specific percentages for 1987 and 1992 were averaged to produce
basinwide percentages for the following four crops:
1) Soybeans
2) Cotton
3) Corn/Sorghum
4) Other crops
Soils
Soils have a large influence on basin discharge behavior because their properties
determine the rates of infiltration and interflow, which in turn affect the timing of surface runoff.
Variability of soils characteristics within a watershed can produce different hydrologic responses
from different parts of the basin. However, information obtained during a field trip to the basin
confirmed the expectation that there is a uniform distribution of soils in the basin. Also, soils-
related hydrology parameters are typically adjusted during calibration. No further effort was
expended on soils data analysis.
360
-------
Surface physiography
The Tensas watershed is representative of the entire Lower Mississippi River Alluvial
Floodplain. It is relatively flat, containing bottomland hardwood swamps. These wetlands are
largely concentrated near channels (rivers and bayous), and are associated with abandoned river
channels, point-bar swales, backswamps, and natural levees. Most of the original forest and
wetlands have been cleared, leveled, and converted to cropland, and also channelized, resulting
in major changes in water movement patterns and loss of riparian areas.
CALIBRATION AND VERIFICATION
Calibration
The principal time series data needed for hydrologic calibration (rainfall, evaporation,
and observed flow) indicate that long-term simulations are possible at the Tendal, Louisiana
gage. The six final generated hourly precipitation data sets support simulations from 1969
through 1994. Since either evaporation station is adequate, the available time span of
evaporation data is 1961 through October 1994, and the flow record covers the period 1935
through October 1993. Therefore, the common period of record for all of the data is 1969
through October 1993, or approximately 25 years. The land use data most closely reflects the
state of the watershed in the late 1970's to the mid 1980's. Since recent land use changes have
been slow, the coverage is probably applicable to periods from 1975 until the present time.
Based on this information, we decided to calibrate the model over the eight-year period 1985-
1992, and verify using the five-year period 1980-1984.
The procedures and parameter adjustments involved in these phases are more completely
described in Donigian etal. (1984), and the HSPF hydrologic calibration expert system
(HSPEXP) (Lumb et al. 1994). HSPEXP produces a standard set of mass balance, statistical,
and hydrograph comparisons that greatly facilitate calibration. It also provides advice on
parameter adjustments and enforces various error criteria (user-defined) for deciding whether
each phase of calibration is satisfactory. HSPEXP was used in the calibration of the Tensas
Basin.
Wetland portions of the basin are modeled using the new "high water table" version of
PERLND (Bicknell et al. 1999), which requires calibration based on comparison of simulated
and observed groundwater levels in the wetlands. Groundwater level data in wetlands could not
be found for the Tensas Basin; however, the system can be adjusted to generate reasonable water
levels in wetlands based on expected behavior. Given the very small amount of land categorized
as wetland in the basin, this has no significant impact on the overall hydrologic calibration. In
addition to the standard PERLND parameters, the main variables for calibrating these levels
include the soil porosities (PCW, PGW, UPGW) and the base elevation for groundwater outflow
(BELV), which corresponds to the bottom of nearby channels.
361
-------
Verification
Verification is an evaluation of the final calibration parameter values using a second
meteorologic time period different from that used in calibration. The evaluation is done using
the final calibration with a new period of record and then evaluating how well the simulated
results match the recorded information for this new time period. A poor verification may result
in need for re-calibration of the watershed with additional data.
Verification of the simulation follows conclusion of the simulation evaluation.
Verification is dependent on the availability of additional hydrometeorological time series data
beyond that used for the calibration period. Verification was performed over a five-year period
from 1980 - 1984.
A limited amount of observed water quality data for Bayou Macon was identified to
provide calibration points for support of that effort, and the data consist of: water quality
measurements at the Tendal station during the period 1979 - 1997, water quality measurements
at Lake Providence at Tensas Bayou during 1984 - 1990, and water quality measurements at
Clayton (Tensas River) during 1988 - 1997.
We could not find any data for either sediment or nutrients for the Tensas watershed.
Sediment and nutrient data were taken from BASINS (Lahlou et al. 1998), and generalized input
parameters were developed for HSPF.
SIMULATIONS
Simulations were performed from 1974 through 1990, for a total of 16 years. The land
use data most closely reflects the state of the watershed in the late 1970's to the early 1990's.
Since recent changes have been slow, the coverage is probably applicable to periods from 1975
until the present time. HSPF was run to produce an output time series for flow, sediment
erosion, and transport of nitrogen that were input into CE-QUAL-W2. The CE-QUAL-W2
provided outflows for transport and transformation in the Bayou Macon endpoint including
temperature and BOD dynamics. These were input into AQUATOX which provided changes in
decay rates and detritus and nutrient recycling on biological endpoints.
RESULTS
Calibration was carried out as described above, and the results are shown in Table 1 and
Figures 8 through 12. The overall annual water balance, in Table 1, indicates that a fairly good
water balance has been achieved, both on an annual basis and an overall basis. This conclusion
is reinforced by the monthly averaged flows over the calibration period, in Figure 8. The model
is generally oversimulating some extremely large storms, as is indicated in the calibration flow
duration curve shown in Figure 8 and particularly in the 1991 daily flow comparison in Figure
12. Examination of the rainfall that generated the large oversimulated storms in December 1990
and May 1991 suggests that there is a problem with the record, or it is not representative of the
average rainfall that fell on the upper portion of the basin in those periods.
362
-------
4Oi < ,-*
During calibration, it was necessary to adjust some of the FTABLEs to store a significant
fraction of the peak flood flow in order to reduce the peaks to what is shown in these results. It
is difficult to represent surface flooding with HSPF RCHRES, since the water cannot be easily
routed back to the tributary land areas (i.e.., PERLNDs) as occurs during large floods in flat areas
such as the Tensas.
,.„.
1000™
j 1Oil
li- i
t€is,-
1 , . '. , , ,
t ? ri 10 ?ft TO 7i> "flfl 98 S9
Pe-rce'nf citrinee t! OW excfMfdud
DM ¥ Mr AN Ft OW M IFNDAi
Figure 8. Calibration Period Flow Duration.
•ItffiO
?>.«< S; 1M>';-f S'-fyij l*''Ji
MONTH! ¥ r I C'W at 1'FWlA!
Figure 9. Calibration Period - Mean Monthly Flow.
363
-------
r
^
_
>
1200-
<
4(10 :
OL~
OBSfcRVfcD
1
0 N D J I" M A M J
1987
at
J A
Figure 10. Calibration - Water Year 1987 Daily Flow.
r-
/iti-fiy .
3,700 _
/i-OQ .......
°i ......
-
f?00 -
......
400 • .....
fi L
O N D J f y A M >J ,t
'
DAILY ML M4 I IOW a! I F.NDAI,
Figure 11. Calibration - Water Year 1989 Daily Flow.
364
-------
4500 i- — OBRtttvr O
4000 ;-
:,
;* 3000 v. ......
S 2500 K-
u
:- ......
15(10;
mnru- •,-.,' -^
'
u N L) J I- MAM J J A.
I- L O W at I L N DAL
Figure 12. Calibration - Water Year 1991 Daily Flow.
The verification results shown in Table 1 and Figures 13 to 16 are reasonably good. Both the
flow duration and time series plots suggest oversimulation of the medium-sized storms. In both
the calibration and verification periods, the storm hydrograph shape and seasonal flow
distribution are well replicated, indicating the recession and evapotranspiration parameter sets
are well calibrated.
The CE-QUAL-W2 and AQUATOX models were run based on historical (1980's) data
and applied over the length of the simulation period. A series of runs to evaluate
uncertainly/sensitivity analyses were conducted to determine the most important driving factors
and to determine what conditions would correspond to targets for chlorophyll a or Secchi depth
as shown in Figures 17 through 21. Response variables were not well correlated with nutrients;
only sizes of chlorophyll a peaks are sensitive. Secchi depth was not sensitive. The system is
light-limited due to high TSS. A drastic (90%) reduction of TSS may initially lead to increase in
algal productivity. When a (50%) reduction in TSS was imposed, chlorophyll a and Secchi
depth became more sensitive to nutrients. With the addition of reductions of organics and
nutrients to represent a land management practice, reduced levels of chlorophyll a and effects
throughout the benthic food chain were observed, including benthic fishes.
365
-------
Table 1. Annual Flow Comparison - Runoff (inches).
Year
Verification
1980
1981
1982
1983
1984
Total
Calibration
1985
1986
1987
1988
1989
1990
1991
1992
Total
Grand Total
Observed
21.7
5.8
30.8
33.2
25.8
117.2
14.7
12.5
21.1
10.6
24.5
20.5
34.0
16.1
153.9
271.1
Simulated
20.0
13.3
32.0
38.2
23.7
127.1
14.1
10.7
21.6
10.7
22.0
28.2
42.4
13.9
163.6
290.6
Ratio (Sim/Obs)
0.92
2.29
1.04
1.15
0.92
1.08
0.96
0.86
1.02
1.01
0.90
1.38
1.24
0.86
1.06
1.07
ffCVift .
10 '2$ M ;0 90
Vicsnl ti:,;ir»i.<;> Tl OW »K.v-i:-4-iAI
99
Figure 13. Verification Period Flow Duration.
366
-------
—
-
_
7OTO
li
:.
12(1(1 -
Bflll —
w %*' fef ^^
400 :'
0_
OBSLMVLU
1981
I f tt Y Mi' A N: F L O W i« I :L NO A I.
Figure 14. Verification Period - Mean Monthly Flow.
u.
.......
3KIJO
3200 -
, ......
2-1 DO -
^ .. , .
16011,
......
800 • ......
4(lfl ,
0-
M
198'
1 I
J f M A M ,1 J
lit
Figure 15. Verification - Water Year 1982 Daily Flow.
367
-------
3500
,?•
j 23QO
1500.™
1
«•=->'
N L) i J I- M A M J
1982 ' 19*3
DAILY FLOW a! TFNOAL.
A
Figure 16. Verification - Water Year 1983 Daily Flow.
DISCUSSION
As Figures 17 through 20 indicate, there are many potential impacts on nutrient cycling
associated with the life cycle processes of the higher trophic organisms represented in ecological
models. The consequences for nutrient cycling are also complicated because large zooplankton
have a lower rate of nutrient excretion per unit mass. A zooplankton assemblage dominated by
larger-sized individuals should thus be characterized by slower nutrient recycling and lower
primary production rates than an equivalent biomass of small-sized individuals (Carpenter and
Kitchell 1993).
In studies that have the greatest relevance to the relationship and importance of higher
trophic levels in mediation of nutrient inputs and eutrophication, Carpenter etal. (1996) and
Schindler etal. (1996) provide evidence that pulses of phosphorus entering lakes whose food-
web is structured by large piscivores are not as likely to result in algal blooms as those where
planktivorous fishes form the highest trophic level.
An analysis of relevant studies presented by Carpenter and Cottingham (1997) concludes
that resilience to phosphorus pulses in a lake ecosystem is intimately connected with the
community of fishes. If piscivorous fishes are abundant, the impact of phosphorus pulses is
damped because of the slower recycle rate of phosphorus incorporated in fish. In this scenario,
phosphorus inputs contribute to increased zooplankton biomass. In the absence of piscivorous
fishes, the impact of phosphorus pulses is more immediately felt and phosphorus inputs
contribute to increased algal biomass and eutrophication.
368
-------
Although these studies support the idea that higher trophic levels should be included in
models of eutrophication processes in lake ecosystems, the identification of the most appropriate
process algorithms remains as a future research need. There is sufficient evidence to suggest that
the science in the water quality based models should start moving towards an ecological end
point.
Figure 17. Sensitivity to chlorophyll a to phosphate inputs.
Figure 18. Sensitivity of Secchi depths to phosphate inputs.
369
-------
Figure 19. Benthic fish sensitivity to nutrients.
(U|
I
•i i
I '.4
laLM^^-::.Jx
I
I
Figure 20. Chlorohphyll a and nutrients under 50% reduction in TSS.
370
-------
REFERENCES
Bicknell, B.R., J.C. Imhoff, J.L. Kittle, Jr., A.S. Donigian, Jr., and R.C. Johanson. 1997.
Hydrological Simulation Program - FORTRAN. User's Manual for Release 11.
EPA/600/R-97/080. U.S. EPA Environmental Research Laboratory, Athens, GA. 763.
Bicknell, B.R., J.C. Imhoff, J.L. Kittle Jr., A.S. Donigian, Jr., T.H Jobes, and R.C. Johanson.
1999. (Draft), Hydrological Simulation Program - FORTRAN, User's Manual for Version
12. U.S. EPA, National Exposure Research Laboratory, Athens, GA.
Carpenter, S.R., and J.F. Kitchell. 1993. Cascading trophic interactions. Chapter 1 in The
Trophic Cascade in Lakes, S.R. Carpenter and J.F. Kitchell (eds.), Cambridge University
Press, Cambridge, England.
Carpenter, S.R., J.F. Kitchell, K.L. Cottingham, D.E. Schindler, D.L. Christensen, D.M. Post,
and N. Voichick. 1996. Chlorophyll variability, nutrient input, and grazing: evidence
from whole-lake experiments. Ecology. 77(3), 725-735.
Carpenter, S.R., and K.L. Cottingham. 1997. Resilience and restoration of lakes. Conservation
Ecology (Online). 1(2), 2-19.
Cole, T.M., and E.M.Buchak. 1995. "CE-QUAL-W2: A Two-Dimensional, Laterally
Averaged, Hydrodynamic and Water Quality Model, Version 2.0: User Manual,"
Instruction Report EL-95-1, US Army Engineer Waterways Experiment Station,
Vicksburg, MS.
Donigian, Jr., A.S., B.R. Bicknell, A.S. Patwardhan, L.C. Linker, C.H. Chang, and R. Reynolds.
1994. Chesapeake Bay Program Watershed Model Application to Calculate Bay Nutrient
Loadings: Final Findings and Recommendations. U.S. EPA Chesapeake Bay Program,
Annapolis, MD.
Donigian, A.S., Jr., J.C. Imhoff, B.R. Bicknell, and J.L. Kittle, Jr. 1984. Application Guide for
the Hydrological Simulation Program - FORTRAN EPA 600/3-84-066, Environmental
Research Laboratory, U.S. EPA, Athens, GA. 30613.
Donigian, A.S., Jr., D.W. Meier, and P.P. Jowise. 1986. Stream Transport and Agricultural
Runoff for Exposure Assessment: A Methodology. EPA/600/3-86-011, Environmental
Research Laboratory, U.S. EPA, Athens, GA. 30613.
Edinger, J.E., and J. Wu. 1999. Description of CE-QUAL-W2's Water Quality Model. J.E.
Edinger Associates, Inc., Wayne, PA.
Environmental Data Service. 1979. Climatic Atlas of the United States, NOAA, National
Climatic Data Center, Asheville, NC.
371
-------
Lahlou, M., L. Shoemaker, S. Choudhury, R. Elmer., A. Hu, H. Manguerra, and A. Parker.
1998. Better Assessment Science Integrating Point and Nonpoint Sources. BASINS,
Version 2.0 User's Manual. EPA-823-B-98-006. U.S. EPA, Washington, D.C. 65 p.
Lumb, A.M., R.B. McCammon, and J.L. Kittle, Jr. 1994. Users Manual for an Expert System
(HSPEXP) for Calibration of the Hydrological Simulation Program - FORTRAN. Water-
Resources Investigations Report 94-4168, U.S. Geological Survey, Reston, VA. 102 p.
Linker, L.C., G.E. Stigall, C.H. Chang, and A.S. Donigian, Jr. 1993. The Chesapeake Bay
Watershed Model. Prepared by U.S. EPA, Chesapeake Bay Program Office, Annapolis,
MD.
NVPDC 1983. Chesapeake Bay Basin Model, FINAL REPORT, prepared for U.S. EPA
Chesapeake Bay Program, Annapolis, MD by Northern Virginia Planning District
Commission, Annandale, VA.
Park, R.A. 1998. AQUATOX for Windows- A Modular Toxic Effects Model for Aquatic
Ecosystems (DRAFT REPORT). Prepared for U.S. EPA Office of Science and
Technology and Office of Pollution Prevention and Toxics, Washington, DC.
Schindler, D.E., S.R. Carpenter, K.L. Cottingham, X. He, J.R. Hodgson, J.F. Kitchell, andP.A.
Sorrano. 1996. Food web structure and littoral coupling to pelagic trophic cascades. In:
Food Webs: Integration of Pattern and Dynamics, G.A. Pollis and K.O. Winemiller
(eds.), Chapman and Hall, New York.
USDA. 1992. National Resources Inventory. NRCS.
USGS. 1971. Drainage Area of Louisiana Streams. U.S. Geological Survey, Louisiana District,
Baton Rouge, LA
372
------- |