United States Environmental Protection Agency
CBP/TRS 18/88
March 1988
Arsenic Transport and
Impact in Chesapeake Bay
Food Webs
Chesapeake
Bay
Program
-------
ATTACHMENT A
June, 1987
ARSENIC IMPACT ON GROWTH, FECUNDITY, SPECIES COMPOSITION AND
SUBSEQUENT TRANSPORT OF ARSENIC IN ESTUARINE
FOOD WEBS
by
James G. Sanders
Richard W. Osman
Kevin G. Sellner
The Academy of Natural Sciences
Benedict Estuarine Research Laboratory
Benedict, MD 20612
Project Officer: Clyde C. Bishop
Office of Research and Development
U.S. Environmental Protection Agency
Washington, D.C. 20460
-------
ABSTRACT
Estuarine organisms respond differently when exposed to dissolved arsenic.
Phytoplankton, particularly centric diatoms, exhibit large changes in growth rate of
dominant species, leading to changes in the species composition of the phytoplankton
community. Zooplankton and benthic organisms, on the other hand, were quite
tolerant of high concentrations of arsenic. Only the barnacle, Balanus improvisus,
was significantly affected by concentrations of arsenate as high as 56 pg 1-1. Balanus
was also the only organism to incorporate significant concentrations of arsenic.
However, this incorporation may have been in shell material and not in animal
tissue. Other organisms did not incorporate significant quantitites of arsenic. The
concentrations of arsenic used in these experiments, 1-56 pgl"1, effectively bracket
concentrations in estuaries, even those receiving considerable impact from man.
Thus, the lack of direct response to arsenic seen in these experiments, except within
the phytoplankton community, can be considered to be relevant to arsenic impacts to
estuarine and coastal marine systems.
However, there are other pathways for impact of a toxic substance within an
aquatic food web. Arsenic, because of its large impact upon phytoplankton species
composition and community structure, is a prime candidate for such indirect
impacts. Even though direct effects of arsenic are limited to phytoplankton and
direct impacts to other trophic levels are minor, potential exists for indirect effects
associated with changes in trophic structure or the ingestion of arsenic through food.
Our experiments have shown that changes in dominant species can drastically alter
an herbivore's ability to procure enough food to successfully reproduce. In addition,
although arsenic dissolved in the water may be unavailable and nontoxic to higher
trophic levels, arsenic incorporated in their food may be quite toxic.
ii
-------
CONTENTS
ABSTRACT ii
INTRODUCTION 1
METHODS 5
Experimental Procedures 5
Phytoplankton experiments 5
Zooplankton experiments 6
Benthic experiments 9
Arsenic analyses 15
RESULTS 16
Phytoplankton 16
Zooplankton 24
Benthic Organisms 31
Arsenic Incorporation by Benthic Organisms 42
DISCUSSION 46
ACKNOWLEDGEMENTS 51
REFERENCES 52
iii
-------
INTRODUCTION
The pollution of an estuary is often the initial and greatest impact borne by the
marine system as a whole. Estuaries serve as principal conduits for man's inputs to
the oceans. Many industrial and municipal activities, such as power generating
stations, wastewater treatment facilities, and industrial plants of all types, are
located on estuaries. These activities generate liquid and solid wastes, some of
which deliberately or accidentally are discharged into the nearby water. In addition,
the concentration of boat traffic in these same waterways results in many small
spills and leakages which add pollutants. Therefore, estuaries are necessary places
to investigate potential impacts of toxic compounds.
Zero discharge of toxic wastes into marine and estuarine environments,
although an enviable goal, may never be a realistic alternative for modern man.
Even if deliberate introduction of toxic compounds does not occur, the oceans will
always face the inevitable, accidental discharge of these materials. Many
compounds within these discharges are of extreme interest to environmental
planners because:
1. they are greatly elevated in affected waters relative to natural water
concentrations,
2. they are toxic to biota, and
3. they are actively accumulated by biota in aquatic systems.
We do not know to what extent chronic or acute discharges of industrial wastes
will alter aquatic ecosystems. The literature suggests that the trace materials likely
to be contained within these wastes are toxic to some organisms. However, the
results from these single-species bioassays are difficult to extrapolate to estuaries
which have complex hydrography and food webs. Sessile organisms in the vicinity of
an effluent have the greatest potential for harm, but plankton contained within the
1
-------
effluent plume also should be affected. Additionally, if the compound is bioactive,
uptake and transformation by organisms will greatly affect its toxicity and transfer
to other organisms within the estuary.
Arsenic is an example of this type of compound. It is a classic poison, with its
primary commercial use as a pesticide (Mackenzie et al., 1979). It is considered a
"priority pollutant" by EPA, and was present in 48% of industrial effluents tested by
the Agency (Keith & Telliard, 1979). Arsenic is present in all aquatic systems,
principally in the form of an inorganic ion, arsenate (Waslenchuk, 1978; Sanders,
1980). Reduced arsenic (arsenite) is generally considered more toxic and is used as a
pesticide (Peoples, 1975). However, at low levels, arsenate is likely to have the
greatest impact on aquatic primary producers. It is chemically similar to phosphate,
a necessary nutrient for plant growth, and it is readily taken up by phytoplankton
(Sanders & Windom, 1980). Arsenate uptake can be followed by accumulation
within the cell, or it can be chemically altered (reduced to arsenite and methylated)
and released into the surrounding water. As much as 80% of the dissolved arsenic
present in Chesapeake Bay is taken up and released in this fashion (Sanders, 1980,
1983, 1985).
Arsenic concentrations within aquatic ecosystems vary widely. Fresh waters
free from anthropogenic contamination generally contain 1 to 10 pg M (Andreae,
1978; Forstner & Wittmann, 1983) with a world-wide, average concentration of 1.7
pg I-1 (Martin & Whitfield, 1983). Oceans are much less variable, ranging between
1.0 and 1.5 pg'l-i (Andreae, 1978; Waslenchuk, 1978; Sanders, 1980). Estuaries fall
somewhere in between. In the Chesapeake Bay, concentrations are lowest at the
head of the Bay, with concentrations less than O.Syg-l"1. Average concentrations
peak in the mid-Bay region around an average of 1.5-2.0 pg'I"1, falling slowly
offshore (Sanders, 1985). Higher concentrations in the mid reaches of the Bay imply
inputs from man's activities, with annual additions recently estimated at 36 metric
2
-------
tons (Sanders, 1985). Occasionally, very high concentrations can be found, with the
highest known concentration, 60 pg'l"1, being found in the Nanticoke River near the
location of an abandoned fly ash pile (Sanders, unpublished data). Presumably,
other "hotspots" occur.
All organisms do not react similarly to a given pollutant. Although
phytoplankton are most affected by arsenate, the least toxic form, invertebrates are
more susceptible to arsenite and methylated forms (Peoples, 1975). Therefore, they
could be harmed by the biogeochemical alterations discussed above. Even within a
particular taxon, all species will not react to the same levels of arsenic. For example,
all phytoplankton species are not equally sensitive to arsenate; some are inhibited at
levels just exceeding ambient concentrations of arsenate in the oceans
(1.0-1.5 pg'l1), while others are resistant to arsenate concentrations two orders of
magnitude higher (Sanders & Vermersch, 1982). Within the phytoplankton
community, where rapid growth and species succession are normal occurrences,
arsenic can cause a shift in the dominant species toward resistant forms and a
concurrent decrease in productivity (Sanders & Vermersch, 1982; Sanders & Cibik,
1985). Therefore, the addition of arsenic to the water column not only can have a
direct toxic effect, but also can result in dramatic shifts in both the dominance within
the phytoplankton community and possibly the flux of carbon and nitrogen between
trophic levels. Coupled with this is the alteration of chemical form, which will affect
(generally increase) arsenic toxicity to higher trophic levels.
Beyond the single trophic level, the effects of a toxic substance (i.e., arsenic) on
an ecosystem such as the Chesapeake Bay are potentially quite complex. Besides the
direct effects postulated above, individuals of heterotrophic species can be influenced
by either changes in the abundances and sizes of prey species or the contamination of
these food organisms. For example, an arsenic-induced shift in phytoplankton
composition to small flagellates or small centric diatoms could reduce the ingestion
3
-------
of phytoplankton by copepods and therefore their fecundity. Small cells are not
captured as efficiently as larger diatoms (e.g., Nival & Nival, 1976; Bartram, 1980).
Thus changes in phytoplankton size may ultimately result in an abundance of small,
noncrustacean grazers (ciliates, rotifers) that can effectively feed on flagellates and a
concomitant decrease in larger zooplankton. This shift, extrapolated to the next
trophic level, predicts that the same aquatic system will support a lowered density of
harvestable fish (see Ryther, 1969; Parsons, 1976; Landry, 1977; Hendrickson et al.,
1982).
Little is known about how such primary and secondary effects of a toxin
combine to change a community or ecosystem. Except for simple systems with two or
three species or communities dominated by a "keystone species" (e.g., Paine, 1969;
Dayton, 1975), we have no models, empirical or theoretical, which could be used to
predict how (or if) changes to a single species might alter the whole system.
Therefore, it is important that the contributions of the various pathways to an
overall toxin-effect be determined for aquatic ecosystems.
Previous research has documented that toxic compounds, including arsenic, are
present in the Chesapeake Bay (Bieri et al., 1982; Sanders, 1985, 1986; Sanders &
Riedel, 1987a), perhaps in quantities large enough to adversely affect water quality,
sediment quality, and the biota of the Chesapeake Bay. The annual loadings of toxic
substances will increase in coming years, even though immediate control of some
point sources may mitigate contamination in specific sub-estuaries. In order to
implement the most effective management practices and to evaluate their success,
we must be able to predict not only which compounds are likely to cause serious
impacts to various biota but also how these impacts affect the dynamics of the
ecosystem.
In an earlier study funded by the U.S. Environmental Protection Agency's
Office of Research and Development (Sanders et al., 1987; Attachment A), we
4
-------
examined the effects of arsenic on three trophic levels within a representative
estuarine food web. Our purpose was to distinguish and quantify those changes in
species abundance, mortality, growth rates, and reproduction that are caused
directly by the dissolved arsenic and those that could be a function of trophic
relationships. The three trophic levels we investigated were the phytoplankton
assemblage, zooplanktonic herbivores, and benthic suspension feeders. We chose
this food web because 1) trophic relationships were fairly simple and not confounded
by such factors as the complex behavior of highly motile species or the biological,
physical, and chemical variability found in sedimentary environments, 2) all species
could be easily manipulated for experimentation, and 3) many species were
economically important within the Chesapeake Bay (e.g., the oyster, Crassostrea
virginica) or were important in estuarine food webs (e.g., phytoplankton and
zooplankton).
Estuarine organisms responded differently when exposed to dissolved arsenate
(Sanders et al., 1987). Phytoplankton, particularly centric diatoms, exhibited large
changes in growth rate of dominant species, leading to changes in the species
composition of the phytoplankton community. Zooplankton and benthic organisms,
on the other hand, were quite tolerant of high concentrations of arsenic. Over a
range of arsenate concentrations of 1 to 56 pg 1-1, only the barnacle, Balanus
improvisus, was affected by arsenate which significantly reduced the barnacle's
growth rate by 6 %. Although a statistically significant result, such a small decrease
is unlikely to affect the organism in a natural system. Balanus was also the only
organism to incorporate significant concentrations of arsenic. However, this
incorporation may have been in shell material and not in animal tissue. Other
organisms did not incorporate significant quantities of arsenic. The concentrations
of arsenic used in these experiments were chosen to effectively bracket
concentrations in estuaries, even those receiving considerable impact from man.
5
-------
Thus, the lack of direct response to arsenic seen in these experiments, except within
the phytoplankton community, indicates that at present levels of arsenic input, this
pathway will have little impact on estuarine and coastal marine systems.
However, there are other pathways for impact of a toxic substance within an
aquatic food web. Arsenic, because of its large impact upon phytoplankton species
composition and community structure, is a prime candidate for such indirect
impacts. Even though direct effects of arsenic are limited to phytoplankton and
direct impacts to other trophic levels are minor, potential exists for indirect effects
associated with changes in trophic structure or the ingestion of arsenic through food.
Experiments performed during the first phase of the study indicated that changes in
dominant species can drastically alter an herbivore's ability to procure enough food
to successfully reproduce (Sanders, 1986; Sanders et al., 1987). In addition, although
arsenic dissolved in the water may be unavailable and nontoxic to higher trophic
levels, arsenic incorporated in their food may be quite toxic. This research program
was designed to follow up on the earlier study, and has investigated the effects of
arsenic-induced changes in phytoplankton dominance on higher trophic levels and
the impact of ingestion of arsenic-contaminated food items. We have compared the
growth, mortality, and fecundity of dominant zooplankton species and important
benthic species when fed arsenic-sensitive, arsenic-resistant, and arsenic-
contaminated foods. These experiments were conducted both in the laboratory,
using cultured species, and under natural conditions utilizing natural
phytoplankton assemblages that had been exposed to arsenic. In the latter
experiments, the response of natural microzooplankton assemblages to arsenic and
arsenic-induced changes in phytoplankton composition were also investigated.
Many different kinds of test systems have been developed or promoted for use
in pollutant assessment at the community level. In addition, several volumes have
been released recently which detail and contrast many systems (e.g..White, 1984;
6
-------
Cairns, 1986). Perhaps one of the most persuasive approaches has been the
development of microcosms, or miniature ecosystems. Early attempts at enclosing
oceanic plankton communities met with limited success (Strickland & Terhune,
1961; McAllister et al., 1961; Antia et al., 1963) but captured the immediate
attention and interest of aquatic scientists. Since then, the use and sophistication of
such enclosures have increased steadily (Menzel & Case, 1977; Grice & Reeve, 1982).
Microcosms generally are defined as confined pieces of a natural ecosystem
maintained under controlled (or known) environmental conditions, which are
representative of the portion of the ecosystem from which they were taken, and
which cannot be readily duplicated in the laboratory by the assembling of component
parts (e.g., Pritchard & Bourquin, 1984). They are not, however, perfect simulations
of ecosystems, nor do they contain all of the important biogeochemical processes that
control the ecosystem.
Enclosures offer many advantages:
1. Two or more trophic levels, with their representative species, can be
maintained for relatively long (weeks) periods of time.
2. The same populations can be repeatedly sampled.
3. Communities can be manipulated.
4. The systems are amenable to mathematical modeling (Grice, 1984).
There are disadvantages as well; Grice (1984) lists 4 major ones:
1. Horizontal and vertical mixing are reduced or eliminated.
2. Wall effects can become severe.
3. The structures can be fragile and difficult to maintain.
4. They can be relatively expensive.
However, they are well-suited and perhaps the best way for us to experiment
with whole ecosystems, to allow the field of ecotoxicology to move beyond the
descriptive stage (Oviatt, 1984; Cairns, 1986).
7
-------
For this study, we have developed a series of separate and linked experiments
using cultured and natural communities that combine the control and flexibility of
laboratory experimentation with the realism of microcosm studies. Our approach
has not been to enclose and manipulate the complex components of an actual
ecosystem. Rather than attempting to model completely such a system we chose to
work with a small subset of species and examine discrete processes (e.g., trophic
transport of a pollutant) that would be common to most systems. An added
advantage to this approach is that it is less expensive to large-scale microcosm
studies. Our design is described in more detail in the following sections.
8
-------
METHODS
EXPERIMENTAL PROCEDURES
Experiments were conducted either in closed-system laboratory tanks or in
outdoor flow-through microcosms. The outdoor systems were designed for studies
utilizing natural phytoplankton. Laboratory studies were designed to mimic the
changes in phytoplankton community structure that occur when natural
Chesapeake Bay communities are exposed to arsenic.
Natural phytoplankton and zooplankton communities were sampled locally
from the Patuxent River estuary, a subestuary of the Chesapeake Bay (Figure 1).
Zooplankton were collected using hand-held nets from the laboratory pier; dominant
species (Acartia tonsa, Eurytemora affinis) were separated from the remaining
organisms and cultured in the laboratory using standard techniques. Balanus was
collected by exposing artificial substrates on which larvae attached (fouling panels
similar to those used by Osman, 1977,1982). Crassoairea larvae and juveniles were
obtained from an oyster hatchery on the eastern shore of Maryland (Flo-Max
Industries, Crisfield). The experimental designs are described below in detail.
Arsenic levels in these experiments ranged from ambient (0.5-1.0 pg"ll) to 25
[jg-l-i. Treatments receiving arsenic inputs (10-25 pg-l"1) were designed to contain
enough arsenic to greatly increase the cellular arsenic content of cultured
phytoplankton or to significantly alter the species composition and succession of
dominant species within natural phytoplankton assemblages without harming any
of the invertebrate species studied. Earlier work (Sanders, 1986; Sanders et al.,
1987) had shown that this range would achieve this result.
9
-------
\\
Chesapeake
Region
V
SCALE
nautical miles
STATUTE MILES
Figure 1. The Chesapeake Bay region.
10
-------
Benthic Experiments
Two different benthic species were chosen for study: the oyster, Crassostrea
virginica, and the barnacle, Balanus improuisus. These organisms were selected as
representative and important members of the filter feeding community of
Chesapeake Bay. Separate experiments were performed with each.
The general experimental design utilized three feeding treatments for each test
species: 1) algal species sensitive to arsenic, 2) algal species resistant to arsenic, and
3) resistant species contaminated by arsenic.
Using this design we hypothesized:
1. If the uptake and transformation of arsenic by a food organism had a
deleterious effect on the benthic species, this would result in a significant difference
in growth and/or survivorship between resistant species treatments with and
without arsenic.
2. If the replacement of a normally dominant sensitive species by a resistant
species affected the benthic herbivore, then a significant difference in growth or
survivorship between the feeding treatments (without arsenic) would result.
3. If the processes in 1 and 2 above interacted then the greatest differences
should be found between individuals fed sensitive species and individuals in the
treatment with resistant species and arsenic.
The first experiment was conducted with newly attached oysters (Crassostrea
virginica). Cultured oyster larvae were exposed to sixty 115 cm2PVC panels in an
oyster hatchery on the eastern shore of Maryland. The panels with attached
juveniles were placed in coolers filled with Bay water and transported to the
laboratory. At the laboratory panels were sorted into three groups based on oyster
density, and equal numbers from each of these groups were randomly assigned to one
of twelve 80 liter treatment tanks. In the tanks, panels were placed in racks which
separated them and held them vertically. Six tanks were then assigned randomly to
11
-------
each of the 2 arsenic treatments (0, 22 ng H). As in previous experiments, a
continuous flow design was used and panels were removed to 12 separate, 40 liter
feeding chambers for 4 h per day. Feeding chambers were aerated and filled with
static cultures of either [sochrysis sp. (sensitive species) or Thalassiosira pseudonana
(resistant species). Half of the tanks within each of the 2 arsenic treatments were
randomly assigned to the Isochrysis feeding treatment and the other half were
assigned to the Thalassiosira treatment. Algal concentrations ranged between 1 x
104 and 2 x 104 cells'ml'1 during the course of the experiment. These densities were
based on estimates of oyster nutritional requirements (Kennedy & Breisch, 1981)
and estimates of algal carbon content.
The experiment was continued for 4 weeks and each experimental panel was
photographed at 0, 2, and 4 weeks. The area covered by each individual oyster was
used as a measure of its size and growth. These measurements were made by
analyzing the photographs using the laboratory's image analysis system. Oysters
showing no growth (75% of all oysters, with no significant differences between
treatments) were considered unsuccessful recruits and were assumed to have died
before the beginning of the experiment. They were not included in any growth rate
analyses.
The second experiment was performed with newly settled barnacles. PVC
panels (115 cm '-) were placed in the Patuxent River at Solomons, Maryland for 2
wks; the result was a heavy set of new individuals with little recruitment of
competing species. All panels were returned to the laboratory and carefully
inspected. The density of barnacle settlement varied greatly among the panels and
60 with approximately equal densities were chosen from the 180 panels originally
exposed. These panels were gently washed, photographed and 5 panels were
randomly chosen and assigned to each of twelve 20 liter chambers. These chambers
were then assigned to 1 of 6 algal food treatments combined with 2 arsenic
12
-------
treatments. Table 1 shows the overall design. Two species of algae sensitive to
arsenic (Cerataulina pelagica and Isochrysia sp.) and two resistant species
(Thalassiosira pseudonana and Dunaliella sp.) were used. Regardless of treatment,
each tank received an equivalent amount of food (7.8 mg C'd'1). This amount was
based on estimates of barnacle biomass and nutritional requirements. Observations
indicated that preferred algal species were cleared in 4-6 hours but that
nonpreferred species remained even after the 24 h feeding cycle. In each of 4
treatment tanks, barnacles were fed one of the four algal species. In addition, each of
two other tanks received one of the resistant species and 25 pgl-1 arsenic per day.
The remaining 6 tanks were assigned randomly to one of three treatments: both
sensitive species (Cerataulina pelagica and Isochrysis sp.), both resistant species
(Thalassiosira pseudonana and Dunaliella sp.), and both resistant species +
dissolved arsenic. In each of these treatments, equal amounts of each of the two
species were used with the total amount fed equivalent to the single species
treatments.
All chambers were aerated and maintained as static cultures. Water and algae
were completely replaced every 24 hours. All panels were photographed after 2
weeks and at the end of the experiment (4 weeks). Photographs were then analyzed
using the laboratory's image analysis system. This analysis consisted of measuring
the surface area covered by individual barnacles and recording whether they were
alive or dead. The majority of panels had more than 1000 barnacles, all of which
were of similar age and size. Because of this similarity in size, a sample size of 100
barnacles per panel was judged to be more than sufficient for estimating growth. To
eliminate any bias in choosing those barnacles measured, panels were divided into 7
equal-sized sectors and one sector was chosen randomly for analysis. If < 100
barnacles were found in the sector, additional sectors were analyzed until this
13
-------
Table 1. Experimental design for the barnacle experiment. Shown are the species
of algae fed to barnacles in each type of treatment, the number of
experimental chambers used for each treatment, and whether the algae
were cultured with or without arsenic. Algal species: Cerataulina pelagica
(CPEL), Isochrysis galbana (TISO), Thalassiosira pseudonana (SWAN1),
Dunaliella tertiolecta (DUN).
SPECIES OF ALGAE USED FOR
FEEDING
NO. OF
TREATMENT CHAMBERS ARSENIC cp^ T|S0 SWAN1 DUN
1
1
-
X
2
1
-
X
3
1
-
X
4
1
-
X
5
2
-
X
X
6
2
-
X
X
7
1
+
X
8
1
+
X
9
2
+
X
X
14
-------
minimum was reached. In order to measure growth unbiased by sampling different
individuals, the same sectors were analyzed at the different sampling times.
At the end of the experiment, barnacles were removed from the panels for
arsenic analyses (see below).
Zooplankton Experiments
The effects of arsenic on trophic relationships between the dominant copepod in
the Chesapeake Bay, Acartia Lonsa, and phytoplankton species were investigated.
Both the direct effect of ingesting arsenic-laden phytoplankton and the indirect
effect of changes in food types on the fecundity and development of this important
zooplankton species were studied. In a previous phase of these studies, dissolved
arsenic had a negligible effect on copepods below a concentration of 100 pg'l-1
(Sanders, 1986).
As in the barnacle experiment, the same two phytoplankton species known to
be sensitive to arsenic (Ceratauhna petagica and Isochrysis sp.) and another pair
known to be resistant to arsenic (Thalassiosira pseudonana and Dunaliella sp.) were
used as food for Acartia tansa. The resistant algae were offered with and without
incorporated arsenic. Arsenic loading of the phytoplankton cells was accomplished
by incubating the cells in 25 pgl-1 arsenate for at least 2 days and not more than 4
days before offering it to the copepods, a procedure designed to maximize the cellular
arsenic content. Thus, three treatments, arsenic-sensitive algae, arsenic-resistant
algae and resistant algae contaminated with arsenic were run in triplicate for both
the fecundity and development studies described below. Initially and at each
transfer, cell densities were set to provide a surplus of food (greater than 0.67 mg C 1-
'). The algae were kept in suspension by gentle swirling of the dishes at least once
per day. The concentrations of arsenic carried over in the culture medium and in the
arsenic loaded cells were measured in each of the experiments (see below).
15
-------
All experiments were conducted in acid-washed 500 ml polymethylpentene
containers (11 cm diameter, 7 cm depth) incubated at 25° C using a light:dark cycle of
12:12 hr. For the experiments, Patuxent River water (ll-13%o) was filtered (1 pm
nominal pore size) and autoclaved (121° C, 15 PSI, 3 min). 300 ml of this water was
used in each experimental vessel. Acartia tonsa used in the experiments were
cultured at25°C on a mix of Thalassiosira weissflogu and [sochrysis sp. on a 12:12 hr
light:dark cycle. Several days prior to the experiments, the copepods were switched
to a diet that contained a combination of all four of the experimental algae.
Fecundity--Adult female copepods were isolated from healthy stock cultures
and 20 were placed in each experimental vessel. Algae were added according to the
three treatments described above. Every two days, the adult females were
transferred using large-bore pipettes to clean chambers to which the prescribed
concentrations of algae were added. The remaining eggs and nauplii were removed
using a 20 pm mesh net and preserved in buffered formaldehyde (2% final
concentration) for future counting. Males were not added for the following reasons:
1) it was assumed that the females had been fertilized in the stock cultures, 2) males
would increase grazing pressure on eggs and nauplii, and 3) Parrish & Wilson (1978)
found that egg laying in Acartia tonsa, following a general increase over the first 2-3
days after fertilization and removal of males, showed a relatively high daily egg
production for the next 10 to 14 days.
Development--Qne hundred early stage (Nl or N2) nauplii were added to each
experimental vessel and the appropriate algae added for each treatment. The early
stage nauplii were obtained by isolating adult females in fresh food for 24 hours then
removing them and allowing the eggs to incubate for another 24 hours. The nauplii
produced were then transferred to glass depressions, inspected microscopically for
eggs (which were removed), and then rinsed into the experimental chambers. At 2
and 4 days into the experiment, the copepods were gently collected on a 54 pm screen
16
-------
and transferred to clean dishes with fresh food. On the 6th day, the experiment was
terminated and the copepods were preserved in buffered formaldehyde for later
counting and determination of developmental stage-
Arsenic Uptake--Adult Eurytemora affinia were collected from the Patuxent
River in large numbers by towing a zooplankton net (202 pm mesh) along the
laboratory pier. The animals were preconditioned on 6 species of phytoplankton,
Tkalassiosira pseudonana, Dunaliella sp., Rh.izosolen.ia fragilissima, Isochrysis
galbana, Prorocentrum mariae-lebouriae, and Skeletonema costatum, for 42 hr, then
rinsed free of algal cultures and kept without food for an additional 24 hr. The adults
were then exposed to unialgal cultures containing one of the six phytoplankton
species above, both clean cultures and cultures which had elevated arsenic
concentrations in their cells, and allowed to graze for 24 hr. Approximately 5,000
individuals were placed in each culture, with culture densities equivalent to 5 pg wet
weight of algae per individual. After grazing, individuals were removed from the
culture, rinsed, and resuspended in filtered water for 1 hr to allow them to clear their
guts. After this period, several subsamples of each group of Eurytemora were taken
for enumeration, and the remainder were dried and prepared for arsenic analysis.
Subsamples of the algal cultures were also taken for arsenic analysis.
Integrative Experiment
A final experiment was performed that combined all direct arsenic and indirect
trophic response effects. The natural phytoplankton assemblage was exposed to
elevated arsenic concentrations, then was used as food for a variety of filter-feeding
organisms. A system of large-volume, outdoor tanks was utilized for this
experiment. Natural phytoplankton from the Patuxent River were cultured in
cylindrical fiberglass tanks 76 cm in diameter and 112 cm in height, containing a
volume of 500 1. The tanks are submerged in a raceway through which running
water was circulated to control and mairftain the temperature of the water in the
17
-------
tanks to within 1° C of the ambient water temperature. The concept and objective
was to operate the tanks as continuous, flow-through phytoplankton cultures using
the mesohaline river water without nutrient enrichment as the diluent. The tanks
were initially filled with Patuxent River water containing the natural
phytoplankton assemblage after passage through 35-pm nylon mesh to remove large
herbivores. This initial screening did not remove a significant fraction of the
phytoplankton. After Tilling, each tank was sampled and the phytoplankton species
composition and abundance compared between tanks as a measure of tank-to-tank
variability (Sanders et al., 1981). Filtered (1-pm) water was then continuously
supplied to each tank at a nominal dilution rate of 50% per day (Sanders & Cibik,
1985; D'Elia et al., L986) and continuous infusions of the toxic substances were
begun. Because assemblages were maintained as flow-through cultures, adverse
chemical changes associated with photosynthesis and respiration (low O2, high
NH4 *" concentrations, wide variations in pH) were minimized within the tanks.
Three tanks received no arsenic additions, three received arsenic additions.
The cultures were maintained for a period of 4 weeks. Arsenic was added to the
cultures at one level, 10 pg'l-1, chosen after earlier experimentation had shown that
this concentration was sufficient to inhibit the growth of several dominant species
without affecting overall biomass levels or productivity (Sanders & Vermersch,
1982; Sanders & Cibik, 1985). After the experiment was begun, doses of two cultures
were inadvertently altered (one control, one arsenic-dosed); thus results of these two
cultures were not included in this analysis.
The phytoplankton assemblage in each tank was monitored daily for in vivo
fluorescence, a rapid measurement of plant biomass (D'Elia et al., 1986), and was
sampled every other day for phytoplankton species composition and total cell density
(Sanders & Cibik, 1985; Sanders et al., 1987).
18
-------
Microzooplankton were sampled from the tanks every 2-3 days from initiation
to termination of the experiment. Whole water samples (500 ml) were preserved by
pouring them onto 12.5 ml of formalin {2% formaldehyde final concentration).
Subsamples, 25 ml for the initial time period and 50 ml for all other time periods,
were settled and enumerated using the Utermohl technique (Lund et al., 1958).
Taxonomic identification was made to the lowest taxonomic level possible using the
above techniques.
The overflow from each tank was split into three parts and was pumped into
separate chambers containing either barnacles on plates, adult oysters, or Acartia.
individuals. In each chamber organisms fed on the natural phytoplankton that were
present in the overflow. The separate feeding tanks were started 5 d after arsenic
flows began thus allowing the phytoplankton assemblage a period of time to respond
to the arsenic stress.
Feeding Cultures - Water from each of the control and arsenic-treated
microcosms was continuously pumped (50 ml min1, flushing rate = 1.9 times per
day) into eighteen 38 liter aquaria which were located in a separate raceway. The
aquaria used for zooplankton experiments contained two plexiglass enclosures (11.5
x 11.5 x 35 cm) with four circular windows (5.7 cm diameter, 25.5 cm2), two each on
opposing sides, covered with 35 pm mesh netting. The volume of the enclosures,
determined by the cross-sectional dimensions of the chambers and the height of the
aquaria (30 cm), was 3.9 liters. The experimental organisms, A. tonsa adults for the
fecundity study and nauplii for the development study, could thus be contained
within the enclosures and exposed to the phytoplankton populations which
developed in the individual microcosms.
Six days after the start of the experiment, organisms were added to the
aquaria. In each of the 6 zooplankton aquaria, 20 adult, female copepods were added
to one enclosure. Adult survival was very poor after two days so this first attempt
19
-------
was aborted. The aquaria and enclosures were set up again on day 11. Thereafter,
every two days for six days total, the enclosures were gently drained and the
contents rinsed into a 500 ml polymethylpentene container. The contents of the
containers were scanned using a dissecting microscope and the viable adults
enumerated and transferred to 35 ym filtered water from the appropriate chamber.
Care was taken not to transfer any eggs or nauplii with the adults. The remaining
concentrate from the enclosure which contained the eggs, nauplii and expired adults
was poured into a net cup fitted with 20pm mesh nytex netting. The contents of the
cup were then rinsed into a sample jar and preserved with formaldehyde (2% final
concentration). The number of eggs, nauplii and adults in the samples were
enumerated in a Wildco plexiglass sorting wheel using a dissecting microscope.
The development study was initiated on the seventh day of the integrative
experiment. The second enclosure in each aquaria received 150 nauplii (stages 1 &
2) which were obtained as in the laboratory study described above. After eight days
of incubation in the enclosures, the entire contents of the chambers were gently
drained and rinsed into sample bottles and fixed with formaldehyde as above. The
total number of eggs, nauplii, copepodites, and adults in the sample were determined
microscopically as well as the developmental stage (i.e. six naupliar, 5 copepodite,
and adult males and females). Subsampling was necessary for some of the samples
which had very high densities of eggs and nauplii.
Each of the aquaria with oysters contained 10 individuals placed haphazardly
on the bottom. Oysters were cleaned, and each had an edge of its shell filed straight
to create a reference point for measurements. Using calipers, each oyster was
measured perpendicular to the filed edge and then photographed. After two weeks
and at the end of the experiment, oysters were remeasured to determine the extent of
shell growth, and rephotographed for image analysis.
20
-------
Each of the remaining 6 aquaria were assigned 5 panels with attached
barnacles. Panels were held on a rack which oriented them perpendicular to the
bottom. As in the earlier barnacle experiment, individual plates were photographed
before and after the growth period, and individual growth rates were determined
using image analysis.
Because of their small size, enough zooplankton could not be collected for
analysis of arsenic content. Tissues of both barnacles and oysters were sampled for
arsenic analysis. Some barnacles were too small to allow for separation of tissues
from shell material. For oysters tissues, newly formed shell and old shell were
analyzed separately for arsenic content. Phytoplankton samples were also taken for
arsenic content.
ARSENIC ANALYSIS
The concentration and chemical form of arsenic within each experiment were
monitored within the water column and organisms. Water samples were collected in
rigorously cleaned (Boyle & Huestedt, 1983) plastic bottles and analyzed by hydride
generation (Braman et al., 1977). This method of analysis permits determination of
the total concentration of arsenic and also its chemical form. Limits of detection in
our laboratory are about 20 ng 1-1.
Solids were dried, weighed, and ashed at 500°C for 24 hr in the presence of an
ashing aid [Mg(N03) and MgO] to prevent loss of arsenic (Uthe et al., 1974). After
ashing, the residue was dissolved in IN HC1 and analyzed as above.
Technique accuracy was assessed through the use of standard reference
materials, NBS #1566, oyster tissue, and NRC NASS-1, a seawater standard.
Recoveries of these materials averaged 95% and 91%, respectively.
21
-------
RESULTS
BENTHIC EXPERIMENTS
Juvenile Oysters
This experiment employed a nested design and was analyzed using nested
analysis of variance. Oysters were nested on panels which were nested within
replicate treatment tanks which, in turn, were nested within the 4 treatments. This
analysis allowed us to identify any potential differences between panels within
chambers and chambers within treatments that may have contributed to observed
differences.
At the beginning of the experiment over 20,000 newly settled oysters were
found on the experimental panels. Of these more than 5,500 increased in size over
the course of the experiment and were judged to have successfully recruited prior to
the study. The sizes of these recruits after 4 weeks were quite variable (Figure 2),
but no significant differences were found between any of the treatments (Table 2). In
fact, the combined effects of arsenic and algal species explained less than 4% of the
variation in oyster size.
Because it was not possible to distinguish living and dead oysters on the
photographs, much of the size variation found among the oysters probably resulted
from mortality during the study. Small individuals were likely to be ones that had
died early in the experiment. To examine whether growth differences existed among
individuals that survived the length of the experiment, we reanalyzed growth using
only individuals that reached sizes larger than 10 mm2. This analysis assumed that
those individuals that had attained this arbitrary size (< 10% of successful recruits)
were still alive (they had clearly grown) and examined whether differences existed
among treatments for this group of oysters. As in the previous analysis, there was no
significant effect resulting from arsenic or algal food (Table 3).
22
-------
m
Cvl
i
K
IX
C
EX
EX
ESS
EM
KXXXX
EXXXXXX
GXXXXXXXX
IXXXXXXXXXXXXXXX
mmmmmnm
mmmmmmmmmmmumu
i—i—i—1—i—1—i—1—i—'—i 1 r
o o o o o o o
o m o m o in o
•<*- ro ro c\l CN t- t-
O
-
o
-in
o
m
s|DnpiAipu| )o jaquunfyj
Figure 2. Variability in size of juvenile oysters at the end of the feeding experiment.
23
-------
Table 2. Nested Analysis of Variance of Juvenile Oyster growth data. Data were
analyzed at the end of the experiment. *MS for Chamber used as the error
term.
Juvenile Oysters
Source
ss
DF
F
P
Arsenic
1.1887
1
0.04*
0.8419
Alyae
98 2979
1
3.51 *
0.0979
A rsemc* Algae
3 5919
1
0 22
0 6364
Chamber in Treatment
224 0476
8
1 74
0.0835
Panel in Chamber
3223.3201
46
4.36
0.0001
Krror
87901.3163
5471
24
-------
Table 3. Nested Analysis of Variance of Juvenile Oyster growth data. Data were
analyzed at the end of the experiment. Only individuals larger than 10
mm2 were analyzed. *MS for Chamber used as the error term.
Juvenile Oysters
Source
ss
DF
F
P
Arsenic
216 0456
1
2.49*
0.1781
Alyae
37.7710
1
0 44*
0.5417
Arsenic* Algae
1 1508
1
0.05
0.8259
Chamber in Treatment
694.7792
8
3.65
0.0003
Krror
36309.4393
1528
25
-------
Again, assuming that size and length of survivorship had a positive
relationship, size distributions were used to determine whether survivorship differed
among treatments. Figure 3 compares the number of individuals that attained
successively larger size classes. These "survivorship" curves obviously do not differ
between treatments.
Given the large number of juvenile oysters measured, the results of this
experiment are unambiguous. Even though we might expect these juvenile life
stages to be fairly intolerant of stress, they did not demonstrate any reductions in
growth or survivorship that could be related to the experimental treatments.
Juvenile Barnacles
Both the growth and survivorship of barnacles were analyzed using analysis of
variance and treatments were compared using either Duncan's or Bonferoni's (when
sample sizes were unequal) a posteriori tests. The percent survivorship was
computed for each panel and transformed for analysis using an arcsine square root
transformation. A one-way ANOVA was used with treatment tank as the main
effect. Growth was analyzed using a nested ANOVA model with panels nested
within tank and tanks nested with treatment.
As can be seen in Table 4, survivorship was clearly affected by species of algal
food, but not by the presence of arsenic. Barnacle survivorship was reduced by as
much as 30% when fed Dunahella (a species resistant to arsenic) and 25% when fed
Isochrysis (a sensitive species). The survivorship of barnacles in chambers fed a
mixture of Dunahella and Thalassiosira (both resistant species) was also lower than
other treatments, but not significantly (p>0.05) different.
An analysis of barnacle sizes at the beginning of the experiment showed no
significant differences between those assigned to the different treatments. More
than 90% of the barnacles in each treatment were <0.5 mm2. After 4 weeks there
26
-------
*1
M •
¦9
~I
rt>
CO
CO
c
<
o
~1
w
cr
t-* •
o
o
<<
in
r+
fD
~1
(Si
tr
a>
<
p
•-!
M •
o
c
w
3
rt-
W
Q.
X
CO
q:
o
>
>
c£
Z>
CO
U)
D
D
~U
• I
>
• MM
*o
_C
0
>
• MM
-M
3
E
o
100
•—• Sensitive + As
»-« Sensitive Ctrl
h—a Resistant + As
~ Resistant Ctrl
—I > 1 > 1 1 1 1 1 ' 1 1 1
4 6 8 10 12 14 16
OYSTER SIZE CLASS (mm2)
-------
Table 4. A one-way analysis of barnacle survivorship. Data were transformed
using an arcsine square root transformation and means show percent
survivorship based on that transformation. Lines connect means with no
significant differences (based on Duncan's a posteriori test).
Source DF SS F p
Treatment 11 1.5002 4.05 0.0003
Error 48 1.6149
TREATMENT
DUNCAN GROUPING MEAN ALGAE ARSENIC
97.2
SWANl
-
96.9
CPEL
-
96.9
CPEL + TISO
-
95.6
CPEL + TISO
-
94.9
SWANl
+ DUN
+
93.5
SWANl
+
90.5
SWANl
+ DUN
-
89.9
SWANl
+ DUN
-
86.9
SWANl
+ DUN
+
72.3
TISO
-
68.3
DUN
+
67.8
DUN
-
28
-------
were significant differences in the mean size of barnacles within the various
treatments (Table 5) despite the large variability in the growth (size) of barnacles
within treatments (Figure 4). This variability can be seen in the significant
differences found in mean barnacle size between tanks within treatments and panels
within the same tank. Nevertheless, several patterns can be seen in the a posteriori
comparisons:
1. Barnacles fed the sensitive species Cerataulina grew to a significantly
larger mean size (12.3 mm2) than those in other treatments.
2. Barnacles fed the resistant species Thalassiosira grew to a mean size
significantly smaller (10.1 mm2) than those fed Cerataulina, but significantly larger
than those in all other treatments.
3. Barnacles fed the second sensitive species Isochrysis grew to a larger size
(8.9 mm2) than those fed the resistant species Dunaliella (7.8 mm2), but less than
those fed the other two species.
4. Because barnacles grew well when fed 1 of the 2 sensitive species and when
fed 1 of the 2 resistant species, mixed cultures produced intermediate growth, not
different from one another.
5. Barnacles fed Thalassiosira with arsenic grew significantly more slowly
(8.5 mm2) than those fed clean Thalassiosira (10.1 mm2).
6. Finally, as was found with Thalassiosira, the mixed culture of the
resistant species, Thalassiosira and Dunaliella, with arsenic resulted in
significantly lower barnacle growth than the same treatment without arsenic.
29
-------
Table 5. A nested ANOVA of barnacle growth. Growth was measured as area
covered (mm2) by each individual after 4 weeks. Lines connect means with
no significant differences (based on Bonferoni's a posteriori test).
Source
DF
SS
Main Variable
Treatment
Nested Variables
Tank in Treatment
Pane! in Tank
3
48
13872
3535
24681
37.31
28.98
13.70
<0.0001
0.0001
<0.0001
Error
7370
326387
TREATMENT
BONEERON I GROUPS
MEAN
ALGAE
ARSENIC
12.27 OPEL
10.08 SWAN1
8.93 T1SO
8.88 SWAN 1 + DUN
8.84 CPEL + TISO
8.53 DUN
8.50 DUN
7.96 SWAN 1
7.83 SWANl + DUN
+
+
+
30
-------
IlUUUUUllL
Vllllll
i—1 i 1 i 1 i r-i i 1 i 1 i 1 i 1 1 i 1 i
?a°ss°?so? a ° s ?
(p;oi p %) A3N3f103dJ
Figure 4. Size distributions of barnacles in each of the 9 treatments used in the
feeding experiment. Shown are the distributions at the end of the
experiment. Size categories are in 0.5 mm2 intervals. Feeding treatments
with and without arsenic are displayed together.
31
-------
ZOOPLANKTON EXPERIMENTS
Fecundity
The results of the fecundity study are given in Table 6 and Figure 5. The
results indicate inhibitory effects of both ingestion of arsenic-laden food and algal
species composition.
After two days, the copepods fed sensitive algae had produced approximately
the same numbers of offspring (eggs + nauplii) as those fed the resistant algae
without arsenic loading (Table 6). However, those fed the resistant algae with
arsenic produced significantly fewer offspring, approximately 25% of control
production. Production in both treatments fed resistant algae (both with and
without incorporated arsenic) continued to decline relative to controls as the
experiment progressed. By the end of the experiment, copepods fed either the
resistant algae with or without arsenic loading produced less than one
offspring copepod1 d-1 while those fed the sensitive algae had continued to produce
26.9 offspring copepod-^d-1 (Table 6).
Development
The results of the development study are presented in Figure 6. As with the
fecundity study, both the ingestion of arsenic-laden phytoplankton and the change in
species composition had an inhibitory effect. Though survival was low (8-40%
overall), differences between treatments could be seen in percent survival and in the
advancement of juveniles though the developmental stages. The low survival may
have been caused by handling of the copepods during transfers to clean culture
vessels and fresh food or perhaps to overcrowding. Heinle (1966) found decreased
development to adults at concentrations above 40 copepods l-1. Our experiments
were performed with 333 naupliH-1. as this density was within the range of natural
densities encountered in Chesapeake Bay (up to 500-1-1; Brownlee & Jacobs, 1987).
32
-------
Table 6. Summary of Acartia tonsa fecundity study. Egg and naupliar production
by females fed sensitive, resistant, or resistant, arsenic-laden
phytoplankton.
Production, eggs and naupliifemale -l-d-1 (xiSE)
Algal Type Day 0-2 Day 2-4 Day 4-6 Overall
Sensitive
Resistant
Resistant, arsenic laden
34.8 ± 1.4 26.8 ±1.9
32.6 ±3.9 8.2 ±4.9
8.8 ±0.7 0.4 ±0.3
26.9 ±6.3 29.5 ±2.3
0.4 ±0.05 13.7 ±5.2
0.4 ±0.3 3.2 ±1.4
33
-------
.« .« m
n m c
QJ
-------
¦3
d-3
. «o
z
m
n-2
o
K)
T-
O
04
-T~
O
• VVv;.;Oc ' "«*•-:£/•
>-5
V- *2
r-
s
"T~
O
M
o
I—
8
-T"
O
C4
-s
s
N
LJ
O
<
I—
U1
U
CL
O
_J
LJ
>
LJ
Q
30NVQNnaV
Figure 6. Development of Acartia tonsa over a six-day period in response to
phytoplankton species composition and arsenic-laden phytoplankton.
Number of occurrences of each developmental stage totaled for the
replicates. N1-N6, copepod naupliar stages 1-6; C1-C4, copepod copepodite
stages 1-4; C5-M, copepodites stage 5 males; A-M, adult males; C5 + A-F,
copepodite stage 5 females and adult females. A. Arsenic-sensitive algae,
B. Arsenic-resistant algae without arsenic loading of the phytoplankton,
and C. Arsenic-resistant algae with arsenic loading. See text for species
composition of algae.
-------
The average survival over the six day period was 40,19 and 8% for those
copepods fed sensitive algae, resistant algae without arsenic, and resistant algae
with arsenic, respectively. In the sensitive algae treatment, many of the nauplii had
developed into early and late stage copepodites and a few had reached maturity with
some eggs having been produced. In the treatment with resistant algae without
arsenic, development never proceeded beyond the early copepodite stages (Cl and
C2) and in the arsenic treatment development was restricted to the naupliar stages.
In both experiments, arsenic in the water carried over with the phytoplankton
offered as food was considered insignificant. Sanders (1986) found that levels in
water must be greater than 100 pg H to affect the survival of adult and juvenile
Eurytemora affinis, another important copepod in the Chesapeake Bay. Arsenic
analyses showed that <7 pg 1-1 arsenic was present in the water of the experimental
vessels containing the arsenic-laden phytoplankton.
The egg laying rates (fecundity) obtained with the sensitive algae throughout
the experiments and with the resistant algae without arsenic during the first 2 day
period were within the normal ranges reported for Acartia tonsa by Parrish & Wilson
(1978, means between 26 and 38 eggs copepod-l-d-U. The six day time period for
development from N1-N2 stage nauplii to adult, which was found for at least a few
individuals fed the sensitive algae, agrees well with the results of Heinle (1966).
Thus, the sensitive algae treatment was considered to reflect reasonably natural
conditions.
INTEGRATIVE EXPERIMENT
Changes in Phytoplankton Species Composition
The natural phytoplankton assemblage responded rapidly to arsenic additions.
Species composition and cell densities underwent very large changes in response to
arsenate. Cell densities rapidly increased in arsenic-dosed cultures, peaking at
36
-------
levels greater than 4 times that of controls (Figure 7). This increase was largely
attributable to an increase in the growth of one small centric diatom, Thalassiosira
pseudonana. By day 9, average densities of T. pseudonana in arsenic-dosed tanks
exceeded 9 times the density in controls. During this time, its growth rate averaged
1.1 div d-i in arsenic-dosed tanks and 0.7 div d-i in controls. After this period, this
species did not exhibit differences in growth rate between treatments. The species
persisted in arsenic-dosed tanks 2 d longer than in controls, then experienced a
precipitous drop in density (Figure 8).
After the T. pseudonana bloom, arsenic-dosed tanks were dominated by small
flagellates, a succession which was not reflected in control tanks. These changes to
dominant species in response to arsenic stress resulted in markedly altered
assemblages. This trend continued in that the growth rate of another important
large diatom, Ceratauhna pelagica, was reduced 33% by arsenic and this species did
not succeed in arsenic-dosed tanks as it did in controls during the latter part of the
experiment.
In contrast, phytoplankton biomass was not altered by arsenic dosing; all tanks
exhibited similar levels of biomass throughout the experiment (Figure 9). The
biomass of the assemblages remained relatively constant because, as a consequence
of the increase in Thalassiosira sp. within arsenic-dosed tanks, there was
corresponding decline in the growth rate of larger centric diatoms such as
Cerataulina pelagica and to a lesser extent, Rhizosolenia fragdissima, relative to
controls.
The large increase in Thalassiosira and small flagellates in the first half of the
experiment caused large differences in the taxonomic composition of the two
phytoplankton arsenic assemblages on a numerical basis; however, after the initial
bloom declined, there was little difference between treatments (Figure 10).
37
-------
70-1
60-
50-
40-
O 30"
20-
X
10-
in
30
35
40
_l
Ld
O
- 70-,
£ ]
(75 60"
TOTAL
» -¦ <10 fim
LlJ 50-
Q
40-
30-
20-
10-
30
40
25
DAYS
Figure 7. Phytoplankton total cell densities and densities of cells < 10pm in size in
control and arsenic-dosed cultures during the Integrative experiment. A.
Control tanks, B. arsenic-dosed tanks.
38
-------
UJ
K0
<2
2
(/)
(D
(A
O
?r
S"
c/5
cc
o"
c/:
f>" .
"1
O
w
T3
a-
CD
3
CO
w
O-
C
•n
~—• •
3
crq
C*
tr
0>
HH
3
O)
p
ci-
<
rt>
X
T3
a>
¦-«
60-i
T 50H
o
X
(/)
_l
_l
Ld
o
£
CO
40-
30-
20-
LjJ 10-1
O
0-
0
rt!
3
•—• Arsenic
¦- -m Control
¦ -~f 8 >1—*—«~i
15 20 25 30
DAYS
-------
<§¦
~-j
rD
CD
HH
3 ET
<9
o
e» "H-
p
3
?r
<
rD
n>
X
*T3
re
•-«
o
3
cr
§3
0 W
c/)
cT
t/>
O)
p
w
p
-t
a>
o-
cr
*<
5
G
C
O
13
c
o
*1
fD
W
O
a>
3
o
a-
c
•-j
M •
3
crq
c+
XT
CO
CO
400-
350-
300-
LjJ
> 250H
I—
<;
Lj 200-j
cr
co~ 150^
CO
<
^ iooH
o
00 50-]
0-
Arsenic
Control
0
'/\\
' \"
* 11
\ i,
~
i \
~
~~I I I I I I
10 15 20 25 30 35
DAYS
-------
I
o
70-
60-
50H
40-
30-
20-
10-
cn
_i
LlJ
O
£
CO
z
UJ
a
0-
701
60-
50-
40-
30-
20-
10-
0-
Centric3
E39 Dinoflagellates
I I Cryptophytes
Other
1*
I
%
i
I
8 10 13 15 17 20 22
B
I
553
SSS3
%
I
I
I
%
SS
8 10 13 15 17 20 22
DAYS
Figure 10. Composition of phytoplankton during the Integrative experiment. A.
Control tanks. B. Arsenic-dosed tanks.
41
-------
There was also considerable difference in the size structure of the community.
The success of Tkalassiosira sp. and the decline in growth of larger diatoms led to a
large increase in the proportion of the population that was < 10pm in size (Figure 7).
Microzooplankton Species Composition
On the average over all time periods, microzooplankton densities were higher
in the arsenic (mean = 6800 l-1) than in the control (mean = 4200 H) culture tanks
(Table 7). Average total phytoplankton density (Figure 7) also showed the same
pattern. Analysis of the phytoplankton by size (< or > 10 ^m) indicated that the
difference between the two treatments was due to an increase in the < 10 pm size
category in the arsenic-dosed cultures. As discussed in the Introduction, and as
hypothesized, the arsenic stress led to an increase in the numbers of small
phytoplankton (Figure 7) in the treatment tanks which correlates with the relative
increase in microzooplankton densities.
The microzooplankton tended to respond more to the changes in size
distribution and species composition of the phytoplankton than to the levels of
arsenic in the tanks. Total microzooplankton (Figure 11) in the control cultures,
maintained moderate (900-4000 H) abundances during the first 17 days of the
experiment. The control populations peaked on day 24 reaching about 15,000
organisms H and returned to about 1,000 l1 at the end of the experiment. Between
culture variability of total microzooplankton abundance for the two controls was low
(mean CV over all time periods = 10%) relative to that of the arsenic treatments
(mean CV over all time periods = 46%). Phytoplankton variability was also greater
in the arsenic treatments with the mean CV over all time periods equal to 10 and
16% for the controls and arsenic treated populations, respectively.
In both arsenic-dosed cultures abundance maxima occurred on day 10 reaching
between 7,000 and 9,000 1-'. By day 17, microzooplankton numbers had decreased to
230.and 2600 H for the arsenic-dosed tanks 1 and 2, respectively. By the 24th day,
42
-------
Table 7. Microzooplankton densities (individuals !"1) during the integrative
experiment.
Day
Control
Arsenic-dosed
Tank 3
Tank 4
Mean
Tank 1
Tank 2
Mean
1
1260
2226
1743
1344
2520
1932
5
3129
3696
3413
4368
2877
3623
10
2730
2226
2478
8883
7140
8012
17
903
1029
966
231
2625
1428
•24
15204
14763
14984
49098
1239
25169
33
1302
1407
1355
420
840
630
Average
4088
4225
4156
10724
2874
6799
43
-------
eg-
CD
S
n'
o
N
O
O
TJ
ft)
X
fl>
•
3
ft)
13
50-i
•—• Control 1
- ¦—¦ Control 2
a- -a Arsenic 1
« -« Arsenic 2
30-
20-
0—
0
DAYS
-------
microzooplankton densities in tank 1 had increased dramatically to 49,000 1-1, while
the densities in tank 2 had decreased to 1,000 l-i. Just preceding and during this
period of peak abundance in tank 1, the phytoplankton populations were different in
abundance and species composition. For example, on the 17th day of the experiment
when microzooplankton densities in the arsenic tanks were relatively low, tank 1
had high densities of phytoplankton mostly (83%) composed of small (< 10 pm)
forms. Though small phytoplar.kters were also dominant (81%) in tank 2 at this
time period, the total densities were less than half those in tank 1.
The microzooplankton were composed of ciliated protozoa, rotifers, sarcodinids
and copepod nauplii. Those species which obtained abundances greater than
1,000 H were the oligotrich ciliates Lohmaniella sp. and Laboea sp, the tintinnine
ciliate, Eutintinnus sp., and the rotifer Synchaeta sp. In the first half of the
experiment, Synchaeta sp. densities tracked very closely the distribution of T.
pseudonana. During the second half of the experiment, ciliates (mostly Lohmaniella
and Eutintinnus sp. dominated the microzooplankton in the controls and arsenic
tank 1 and appeared to be associated with relatively high abundances of the < 10 pm
phytoplankton. Arsenic tank 2, which had the lowest microzooplankton densities,
also had the lowest concentrations of small phytoplankton.
Feeding Cultures
Zouplankton Fecundity - Adult survival in the fecundity enclosures was low in
the arsenic tanks relative to the controls (Figure 12). Average survival was 56% for
the controls, and 38% for the arsenic treatments over the three time periods.
Conversely, fecundity (eggs + nauplii) on a per female per day basis was higher for
the copepods exposed to arsenic, especially during the last two time periods (Figure
13, Table 8). Mean offspring per female per day averaged 27 for controls as compared
to 50 in the arsenic treatments over the same time periods as above. By the end of
45
-------
i-ID
.y o
£ o
< o
CO
£
o
sunav jo d^a^riN
Figure 12. Adult Acartia tonsa survival during the fecundity study.
46
-------
aOd3dOO/llldnVN S333
SAiivinwno
Figure 13. Fecundity (cumulative eggs + nauplii copepod*1 d-1) of Acartia tonsa.
47
-------
Table 8. Summary of Acartia tonsa fecundity. Egg and naupliar production by
females fed from control and arsenic-dosed phytoplankton cultures.
Treatment
Production, eggs and nauplii female -i-d-l
Day 0-2 Day 2-4 Day 4-6 Overall
Control
Kggs 12.4 8 1 5.8 8.8
Nauplii 19 0 17.9 18.0 18.3
Arsenic-dosed
Kggs 10.4 21.1 13.8 15.1
Nauplii 30.2 43.4 30.2 34.6
48
-------
the fecundity study, the accumulative number of offspring per female copepod per
day was 149 for the arsenic treatments and 81 for the controls.
During the fecundity study, phytoplankton biomass was similar for both the
control and arsenic treatments. However, the species composition and dominance of
the phytoplankton differed between the two treatments with the arsenic treatment
dominated by < 10 pm phytoplankton and the controls dominated by > 10 pm
phytoplankton. As adult copepods prefer cells > 10 pm in diameter (Nival & Nival,
1976; Berman & Heinle, 1980; Ryther & Sanders 1980), this could explain the poor
survival of the adult copepods in the arsenic treated cultures. Based on the same
argument, reduced fecundity in the arsenic treatments would be expected but the
opposite was found. Average microzooplankton densities were over 2 ^ times higher
in the arsenic relative to control tanks. As microzooplankton are within the
preferred size range for copepod feeding and have been shown to be prey for copepods
(Berk et al., 1977; Robertson, 1983; Stoecker & Sanders, 1985; Stoecker & Egloff,
1987), their greater presence might explain the higher fecundity in the arsenic
treatments, but would also imply that adult survival rates should also be higher. It
has been argued that microzooplankton are a better quality food than phytoplankton
and it has been shown that their inclusion in the diet of Acartia tonsa results in
enhanced egg production (Stoecker & Egloff, 1987). In an attempt to explain the
discrepancy between the adult survival results and fecundity, it is possible that the
higher quality food provided by the microzooplankton could have resulted in greater
fecundity but perhaps lacked something essential for adult survival.
Zooplankton Development - Copepod development was rapid in the integrative
experiment relative to the laboratory studies described earlier. Very few adults
were produced in the laboratory study after six days. The integrative experiment
was thus allowed to continue for eight days. The results (Table 9, Figure 14) suggest
that most of the naupliar stages were second generation nauplii,.i.e. produced from
49
-------
Table 9. Summary of Acartia tonsa development after 8 days when fed from control
or arsenic-dosed phytoplankton cultures.
Treatment
Developmental Stage
Control
Arsenic-dosed
Adult-female
Adult-male
CV-femule
CV-male
CIV
Clll
C1I
CI
N VI
NV
NIV
Nil!
NI&1I
Eggs
Eggs-female-1
10.5
17
0.5
0.5
0
0
1
0
0
74.6
146.7
427 4
811.3
1034
94.4
5
9
1.5
1.5
0
0 5
0.5
0
0
4.3
23.0
123.9
667.7
357.5
57.9
50
-------
Copepodite Stages Adults
Figure 14. Acartia tunsa development in the two arsenic-dosed tanks and the two
control tanks. A. Control tank 3. B. Control tank 4. C. Arsenic tank
2. D. Arsenic tank 1.
51
-------
adults which had developed from the original nauplii. This suggestion is supported
by the greater than 10 fold increase in the number of nauplii in two of the
treatments, the paucity of copepodites, and the number of adults present in most
treatments.
The average number of eggs per adult female copepod was depressed in the
arsenic enclosures (58) relative to the controls (94). However, for the other life
history stages, the copepods receiving effluent from control tanks and those
receiving effluent from arsenic tank 2 developed well, whereas the populations
receiving effluent from arsenic tank 1 developed poorly. Thalassiosira pseudonana,
one of the arsenic resistant species which resulted in poor fecundity and development
when offered as food to copepods in the laboratory studies, dominated the arsenic
treated assemblages. The high densities of this species, particularly in arsenic tank
1, could explain the poor development found for this treatment.
Oyster Growth - Oyster growth was analyzed in terms of shell size and
measured as the change in length perpendicular to the marked edge and as a change
in shell area measured from photographs. Growth was analyzed separately for the
first and second halves of the experiment and for the whole experiment.
Although there were some differences in results based on linear and areal
estimates of growth, the major patterns were the same. Differences probably
resulted from the more variable and less accurate linear measurements. The results
of these analyses based on area are shown in Table 10, and several patterns can be
seen:
1. No differences were found in mean growth between oysters feeding on
effluent from the control tanks. This was true over all time periods and using both
types of growth estimates.
52
-------
Table 10. One-way ANOVA of oyster growth with growth measured as percent
change in size (based on shell area measured from photographs). Data
were transformed (arcsine square root) for analysis and Bonferoni a
posteriori tests were used to compare means.
Sampling
Period
Source
DF
SS
WeeksO-2 Chamber 3
Error 34
Weeks 2 - 4 Chamber 3
Error 36
Weeks 0- 4 Chamber 3
lirror 34
0.1493 5.09 0.0051
0 3323
0 0878 6.97 0.0008
0 0126
0.0129 0.90 0.4489
0.0143
Weeks 0 - 2
Weeks 2 - 4
Weeks 0-4
Treatment
Arsenic 2
Control 2
Control I
Arsenic 1
Mean
7.35
4.77
1.98
1.29
Grp Treatment
Mean
Grp Treatment
Mean
Arsenic 1
12.0
Conti oi 2
14.1
Control 1
10.4
Arsenic 1
13.9
Control 2
9.7
Control 1
12.2
Arsenic 2
2.2
1 Arsenic 2
9.1
53
-------
2. Oyster feeding on effluent from tanks treated with arsenic exhibited
significant differences between tanks and from control tanks during both sampling
periods. These differences resulted principally from the effects of the effluent from
arsenic tank 2. During the first half of the experiment oysters fed from this tank
grew at the highest rate. However, during the second half of the experiment these
same oysters grew at a rate significantly less than those in all other tanks.
3. Because growth rates switched between the two time periods, no
significant differences were found between any of the treatments over the course of
the whole experiment.
Barnacle Growth - The results for barnacle growth were similar to those found
for oysters. As in the barnacle feeding experiment, barnacle growth was analyzed at
the end of the experiment as a difference in mean area covered. A nested ANOVA
was used with tanks nested within treatments.
As can be seen in Table 11 the mean growth of barnacles was reduced when fed
the effluent from the arsenic-dosed cultures. However, as we found for the oysters,
the reduced growth resulted mostly from the strong effects in arsenic tank 2 (Figure
15). Growth was severely reduced to approximately 30% of that found for barnacles
feeding from other tanks.
It is interesting to note that oyster and barnacle growth, and densities of
microzooplankton and the < 10 pm phytoplankton were lowest in arsenic tank 2
during the second half of the integrated experiment. The low numbers of small
phytoplankton might explain the poor growth of all the other organisms examined.
Alternatively, if either the oysters or barnacles were depending on the
microzooplankton for an important part of their daily ration, then the low densities
of microzooplankton might explain the lower growth in these benthic filter feeders.
54
-------
Table 11. Nested ANOVA of barnacle growth in the integrative experiment.
Bonferoni's a posteriori test was used to compare mean growth in mm2
among treatments and among chambers.
Source DF1 SS F p
Main Variable
Live/Dead at Wk 4 1 11172 63.19 0.0001
Treatment 1 5584 31.58 0.0001
Nested Variables
Tank in Treatment 2 20375 57 67 0.0001
lirror 1234 218188
By Treatment By Chamber
Source Mean Grp Source Mean Grp
Control
14.74 1
Arsenic 1
16.83
Arsenic
9.97
J Control 2
16.23
Control 1
13.80
Arsenic 2
5.01
55
-------
cn
crv
»— •
<2
a>
t—1
cn
o
ro
>—I
p
c+-
a>
cd
m •
CTQ
P
O
• •
3
P
n>
(D
D
cn
3
a
P
c-t-
in
O)
a
3
£L
a ^
^ <-*¦
o P
Cn
N*
a>
g
*1
a-
a>
~-• •
ro
P
Ci
to
f+
¦-J
XT
5
cy-
cr
c
CO
<-*-
<
P
P
M
o"
D
C/3
c/T
o
P
c-r
cr
c-r
£>
P
•-j
£>
p
p
CO
o
tr
Of
o
in
?
3
D
v>
a>
<—r-
ET
TJ
a>
Co
~—i
D
P
rf-
c~*
(D
a> era
•-<
«<*
P
c/x
H— «
N
a>
ft)
25—1
15-
5-
o
-t-1
o
I—
O
y~
o
z
Ld
Z)
o
Ld
ct:
CS3 WEEK 4
I 1 JNTTW.
-5-
-15-
-25—1
23-i
15-
5-
l§RRr
Control 1
-5 —
-15-
-25-J
S WCIK 4
~ NTVM
TJ
25 —|
15 —
5 —H
-5 —
-15 —
-25-1
25-i
15-
5-
CS WEEK 4
INITIAL
Control 2
iSr
Arsenic 1
-5-
-15 -
-25 —'
TJ
CSJ WEEX 4
CZZJ WITW_
iSa-
Arsenic 2
T
2
-1—I—1 I 1—I
6 8 10
"I 1 I
8 10
BARNACLE SIZE (mm2)
-------
ARSENIC INCORPORATION AND TRANSFER
As noted in previous research (Sanders et al., 1987), there was little uptake of
dissolved arsenic by trophic levels other than phytoplankton; however, the potential
exists for arsenic uptake through trophic relationships. The arsenic content of the
biota in the barnacle and adult oyster experiments was determined at the end of each
experiment. In addition, the potential for arsenic transfer during zooplankton
feeding was also examined.
Zooplankton
Adult E. affinis showed a small incorporation of arsenic when fed arsenic-
contaminated phytoplankton. Overall, the arsenic content of E. affinis significantly
increased to an average of 11.18 ± 0.86 pg g-1 in organisms fed contaminated food,
as compared to an arsenic content of 8.91 ± 0.53 pgg-1 in organisms fed control
algae. In all but one instance (those fed Iaochrysis sp.), an increase was noted
(Figure 16). The average increase, 25%, was far less than the average increase in
arsenic content of the phytoplankton, an increase of 213%, from 5.66 to 17.69 pg g-K
All species of phytoplankton exhibited an increase of at least 93% (Figure 16).
Barnacles
Barnacles also exhibited significant increases in arsenic content when fed
arsenic-contaminated phytoplankton, larger in magnitude than the increases noted
when they were exposed to elevated dissolved arsenic concentrations (Sanders et al.,
1987). It was noted in this previous work that barnacle analyses were performed on
the total organism, shell and tissue; thus, it was not possible to separate tissue
incorporation from adsorption to shell. Additional samples from an earlier
experiment with adult barnacles (Sanders et al., 1987) were analyzed using both an
acid digestion to remove shell-associated arsenic and a dry ash technique to
determine total arsenic. When these additional analyses were performed, it was
clear that the arsenic uptake that occurred as a result of elevated dissolved arsenic
57
-------
15n
o>
c
0}
C
o
o
3
ID-
S'
i
i
SWAN DUN RFRAG T1S0 PROWL SCOST
Aa CONTENT OF £. affinia
GROWN ON DIFFERENT ALGAL CLONES
30-i
o>
j 2°"
c
Q)
C
o
O 10-
3
B
I
I
reres Araonic
' i Control
I
I
SWAN DUN RFRAG T1SO PROML SCOST
As CONTENT OF MARINE PHYTOPLANKTON
Figure 16. Arsenic content of zooplankton fed control and arsenic-contaminated
phytoplankton. A. Zooplankton B. Arsenic content of phytoplankton.
58
-------
concentrations was associated with shell material; no arsenic was incorporated in
tissue (Figure 17).
The young barnacles fed clean and arsenic-contaminated phytoplankton
exhibited a large increase in arsenic concentration, from 0.34 ± 0.11 pg g"1 to 1.73 ±
0.09 pgg"1, with barnacles feeding on individual species and combined algal species
showing similar arsenic incorporation (Figure 18). As noted in zooplankton uptake
studies, this increase was much less than the increase in arsenic content noted in the
two species of phytoplankton on which they were fed (Figure 18). Because of the very
small sample size, we were not able to distinguish between incorporation in tissue or
shell material; concentrations shown are for the total organism.
Barnacles grown on effluent from the outdoor phytoplankton cultures also
exhibited a similar increase in arsenic content: an increase from 0.34 ± 0.12 pg g_1
in controls to 2.11 ± 0.7 pg g-'.in those fed arsenic-contaminated foods (Figure 19).
In this experiment, we were able to separate arsenic incorporated in tissue from that
associated with shell material. Although the arsenic in the shell increased slightly,
from 0.04 to 0.16 pg'g-1, most of the increase was in tissue content of arsenic, which
increased from 0.30 pg'g"1 in controls to 1.95 pg'g"1 in barnacles feeding on arsenic-
dosed effluent (Figure 19).
Oysters
Adult oysters exposed to arsenic-contaminated food also exhibited significant
increases in arsenic content of tissues from 5.30 ± 0.43 pgg'1 in oysters fed control
phytoplankton to 8.16 ± 0.44 pg g-1 in oysters fed contaminated phytoplankton
(Figure 20). Analysis of shell material did not indicate large arsenic uptake. Old
and new shell material averaged 0.38 and 0.19 pg g-i,respectively, in arsenic-dosed
tanks and 0.13 and 0.12 pgg-1, respectively, in control tanks. Although slightly
higher in arsenic-dosed tanks, the variability of arsenic in all samples was quite
high and the samples were near the limit of detection of the method.
59
-------
3
aq'
e
a>
a-.>
,7
w wj
o n
<2.
(T> O
P- O
* S
W *?
(D 2
3 5.
O
n
o
3
o
ft)
3
3"
o>
(/>
3"
a>
-<
P .—
ET.P
O 3
3 O.
5" ^
•— •
(/)
(A
c
CO
cr
p
-i
3
P
o
a>
CD
X
>C
o
V)
a>
O-
C-t-
o
<
D5
¦i
>—• •
o
c
C/>
As in Barnacles
cn
•
cn
S.
(/)
<
3.0n
2.5-
2.0-
c 1.5—1
(L)
¦+->
C
o
O 1.0-j
0.5-
0.0-
7777A Tissue
CZZ] Shell
0.4 23.8 55.8
As Content in Medium, /xg*l~1
-------
rrn Araanlc
I I Control
Dunallella
ThoL+Oun.
Thalaaaloaira
nm ArMnlo
CZJ Control
Thalammon'im Dunolialla
Species of Phytoplankton
Used as Feeding Cultures
Figure 18. Arsenic content of barnacles fed control and arsenic-contaminated
phytoplankton. A. Arsenic content of barnacles. B. Arsenic content of
phytoplankton.
61
-------
crq
c
•-J
o>
h-»
CD
-o>
<< w
£ 13
>-•
¦— n
f»
3 13
*" S
c-r O
O "l
o O
O- T
C P
"j jr
5' °"
crq 3
ci- cr
® cr
tr1 s»
3
p
o
P (T>
W
rt>
<
m
0)
fD
(D Q_
X j..
T3 P
0> (Ki
2.0
3 O
fD C
c-r
O
P
D
O-
P
w
53
K3 TISSUE
IZZ1 SHELL
2.0-
cn
«
en
=1
c
0
-4—'
c
O
o
w
<
1.0-
0.0-
Control
Q-
o
c/i
a-
Arsenic — Dosed
Treatment
-------
X
TJ
a>
~i
^3
•
¦s
•-J
CD
K>
O
«• 8 >
ET
2-5.
p 2.
2 3
. Q- o
p o
a>§
ft) o
2.3
" s-
§¦§
2? cr
3 -
E n
™ »—<
13 -
ffq P
tr$
" cr
H-1
3 :r
c* -—'
^
CD
- CD
P O-
•
5*3
^ Oq
o
3
*)
CP
CD
C
0
-M
c
o
o
w
<
10.0-1
8.0-
6.0-
4.0-
2.0-
0.0-
IX2 TISSUE
~=~ OLD SHELL
NEW SHELL
Control
Arsenic — Dosed
Treatment
-------
DISCUSSION
Throughout this study and its predecessor (Sanders et al., 1987), the response
of estuarine organisms has been largely as originally hypothesized. Phytoplankton,
particularly centric diatoms, exhibited declines in growth rate upon exposure to
arsenic, leading to changes in the species composition of the assemblage. In different
experiments, different centric diatoms were sensitive to arsenate: Cerataulina
pelagica, Chaetoceros debile, and Rhizosolenia fragilissima are most notable. In all
cases, however, the growth of one small centric, Thalassiosira sp., has been greatly
accelerated by arsenate dosing. The increase in growth rate and strong dominance
exhibited by Thalassiosira sp. is not likely to be a reaction to the arsenic itself, but
rather a reaction to reduced competition with other, inhibited dominant species.
This particular pattern of replacement of large centric diatoms by the small
Thalassiosira sp. has been observed throughout experiments with arsenate and
Chesapeake Bay phytoplankton (Sanders & Cibik, 1985; Sanders, 1986; Sanders et
al., 1987). In the experiment discussed here and in previous studies, the increase in
Thalassiosira sp. did not lead to elevated biomass levels (Figure 9); rather the
species replaced other dominants.
Other trophic levels, however, exhibited relative resistance to dissolved
arsenic. Only the barnacle, Balaam> improvisus, exhibited any sensitivity to
dissolved arsenic in earlier studies, and although significant, the reduction in
growth rate was extremely small, less than 6%. Such a reduction is unlikely to have
ecological significance.
Arsenic incorporation by the various trophic levels followed a similar pattern:
with the exception of the phytoplankton there was no incorporation of dissolved
arsenic by the tissues of estuarine organisms. Again, these findings were
strengthened by earlier work in the United States with Crassostrea (Zaroogian &
64
-------
Hoffman, 1982) and in the Baltic Sea in experimental microcosms containing a large
variety of intertidal and subtidal organisms (Notini & Rosemarin, 1986). In both the
Baltic Sea experiments and in our studies, the only arsenic incorporation was
sorption to calcareous shell material. Our conclusion at the end of the previous study
was that direct exposure of animals to dissolved arsenic, even at high levels
sometimes found in heavily impacted areas, was not likely to cause harm; only the
plants within the estuary are at risk (Sanders et al., 1987). The results of this study
support this conclusion. However, as we theorized, there are other, less direct
pathways for impact within the estuarine ecosystem. We hypothesized that, because
of the high degree of arsenic incorporation within the phytoplankton and the
potential for significant shifts in phytoplankton species composition, estuarine
animals may be affected considerably, either from ingestion of arsenic indirectly
from their food and/or from alteration of feeding relationships. The results of this
study support this hypothesis, but, at the same time, underscore the considerable
complexity of the system (and make us realize how little we really understand).
Arsenic contained within food, unlike dissolved arsenic, appeared to be incorporated
by all organisms. Although not concentrated to the extent seen in marine algae
(Sanders, 1978; 1979; 1980), all species tested had elevated arsenic concentrations.
Such a result has been seen in an earlier, integrated study of arsenic impact in the
intertidal Baltic Sea (Rosemarin et al., 1985). In those experiments, organisms
feeding on arsenic-contaminated macroscopic algae (Fucus) had elevated arsenic
content of similar magnitude to our results. The levels of arsenic incorporation are
not large, however. Only the juvenile barnacles exhibited high levels of arsenic
incorporation; other species increased their arsenic content by only 25-50%.
In addition to arsenic incorporation, we have shown that some estuarine
organisms can be inhibited by arsenic within the food chain. Zooplankton showed
strong reductions in survival and fecundity when fed an altered phytoplankton
65
-------
assemblage and even larger reductions when fed an altered, arsenic-contaminated
assemblage. Barnacles showed a similar reduction in growth when fed arsenic-
contaminated algae; however, their response to altered species composition was
considerably more complex and less well understood. Indeed, it appears that
zooplankton and barnacles react quite differently to the same dominant alga,
Thalassiosira sp.: zooplankton were strongly inhibited, barnacles did quite well.
Clearly, we do not yet understand feeding relationships for these important species.
In contrast, oysters were not at all affected in these experiments. Oysters,
although they incorporated significant quantities of arsenic contained in food
species, showed no change in growth, either in response to altered diet or to arsenic-
contaminated food.
The large differences in response of the various organisms is quite interesting.
It is our belief that the varying response is caused by the variable ability of these
organisms to select food. The mechanisms, degree, and importance of particle
selection by oysters are still poorly understood and under debate (Kennedy &
Breisch, 1981). It appears that oysters can grow successfully on a wide variety of
different diets (assuming that concentrations are sufficient), being able to collect and
assimilate small phytoplankton species; perhaps down to 3 pm in size (Roger Newell,
University of Maryland, personal communication), as well as microzooplankton over
100 pm in size (Krsinic, 1987). In this sense, then, oysters may be relatively non-
selective feeders. Copepods, on the other hand, have been shown to select prey based
upon both size (Parsons et al.,1969; Berman & Heinle, 1980; Ryther & Sanders,
1980) and shape or species composition (Provasoli, 1977). In addition, they appear to
be much less efficient at capturing small phytoplankton less than 7-10 pm in size
(Nival &. Nival, 1976; Berman & Heinle, 1980; Ryther &. Sanders, 1980). Thus,
arsenic-induced changes in phytoplankton species composition might be expected to
affect Acartia to a greater extent than Crassostrea.
66
-------
Microzooplankton densities and species composition also appeared to be
affected more by changes in phytoplankton size distribution and species composition
than by the presence of arsenic in the culture tank experiments. Most
microzooplankton prefer small (< 10 pm) phytoplankton cells (Heinbokel, 1978;
Capriulo, 1982; Rassoulzadegan, 1982), and thus responded in general with
increased numbers in the arsenic treatments which contained increased numbers of
small phytoplankton. In fact, the large discrepancy between microzooplankton
densities in the two arsenic tanks during the second half of the experiment might be
explained by the difference in abundance of small phytoplankton, because one
arsenic tank contained elevated phytoplankton densities as well.
Balanus remains somewhat of an enigma. Barnacles, as oysters, can capture
and ingest a wide range of particle sizes. Although Crisp & Southward (1961) found
that 30 pm particles were the smallest that could be retained by barnacle cirral nets,
fine particles as small as 1-2 pm can be captured and ingested (Southward 1955a, b;
Barnes, 1959). On the other hand, barnacles are capable of capturing larger
particles such as nauplii and are likely to expand the size range of particles they
capture as they grow. This potential shift in preferred food makes the effects of
arsenic-induced changes in algal species composition quite complex for this species.
However, we used newly attached Balanus in our experiments, and our results
suggest that these newly recruited individuals can ingest and may even prefer small
diatoms over flagellate species of any size. In addition, the high densities of
microzooplankton, a potential food item, in one arsenic treatment tank help explain
the unusual barnacle results. It is interesting that arsenic tank 1 contained higher
densities of small phytoplankton which may have stimulated microzooplankton
growth, leading to elevated microzooplankton densities. These elevated densities, in
turn, may have stimulated barnacle growth in this tank.
67
-------
The potential complexity of the effects of a pollutant within even a simple
trophic system, is demonstrated by the integrative experiment. The arsenic-induced
bloom of Thalassiosira during the first two weeks resulted in increased oyster
growth, increased microzooplankton densities, but reduced zooplankton survivorship
and fecundity. In the second two weeks, no differences in phytoplankton species
were obvious but microzooplankton densities clearly diverged in the two arsenic
tanks while the controls remained intermediate. Both barnacles and oysters showed
similar reduced growth in arsenic tank 2. Additionally, as with the
microzooplankton, barnacles exhibited increased growth in arsenic tank 1. Thus, it
would appear that the benthic suspension feeders were responding to arsenic-
induced changes in the phytoplankton and the microzooplankton which were
responding in part to changes in the phytoplankton. However, changes were not
simple and the two arsenic-dosed trophic systems did not respond in the same way,
but within each system the changes were consistent between trophic levels.
A number of investigators have been studying the intertidal, Fucus
ue.sicu/o.su.s-based ecosystem of the Baltic Sea (Rosemarin et al., 1985; Notini &
Rosemarin, 1986; Notini et al., 1988; Rosemarin & Notini, 1988). These mesocosm
studies demonstrated that Fucus, as with phytoplankton, is greatly inhibited by low
levels of arsenic (7 pg-1-') and completely eliminated by higher levels (75 y g* I-1)-
Inhibition and elimination of the Fucus caused major structural shifts in algal and
invertebrate communities and dramatically reduced net ecosystem production.
In contrast to our studies of a plankton-based ecosystem in which trophic
pathways are quite complex, the Baltic Sea ecosystem is a straight-forward food
chain, based upon the macroalga, Fucus. Thus, the large, significant effects seen in
such a system are predictable. In our experiments, the effects at the primary trophic
level were exactly as in the Baltic: loss of sensitive species and replacement by less
desirable species. However, at secondary levels, the response in our studies was less
68
-------
clear-cut, presumably caused by higher tropic levels switching between a variety of
food types. The mechanisms of trophic transfer and carbon cycling and feeding
relationships is an area receiving a great amount of attention at present (see Verity,
1987; Lessard et al., 1987 for examples) in the Chesapeake Bay and in other coastal
ecosystems. As our understanding of such feeding relationships improves in coming
years, our ability to predict indirect toxic effects involving food web alterations will
improve as well.
Clearly, alteration of trophic pathways is potentially a very important
component of pollution. The arsenic-induced shifts in the abundances of dominant
phytoplankton species had significant effects on higher trophic levels, as did the
ingestion of arsenic-contaminated food. That these effects were very complex and
differed greatly between the species investigated is important. Equally important is
the potential ability to predict these effects from existing knowledge of the biology
and feeding relationships of important species. In the case of the three tested
species, their relative abilities to select and utilize different algal species would seem
to be the principal cause of the observed differences in the effects of arsenic. If such
an event can be generalized to other species, to the extent that we understand the
above selectivity, then we can model and predict arsenic impact. However, the
results of the integrative experiment underscore both the complexity and variability
of the natural system: seemingly small changes in phytoplankton species
composition may result in larger fluctuations in microzooplankton densities, leading
to large differences in impact in larger animals. We are still not at the point of
confident predictions at the ecosystem level. However, the model systems tested
here provide an important first step toward such predictions. We were able to
perform only one integrative experiment because of fiscal constraints, yet we were
able to predict many of the results. With refinements, and as our understanding of
Carbon transfer within the Chesapeake Bay improves, our predictions will improve.
69
-------
In addition, the use of such systems points out clearly where further information
about the ecosystem needs to be gathered. This aspect is of great value in identifying
future research needs.
The use of regulated, model systems such as these should be continued. The
type of experiments performed here, using natural phytoplankton assemblages
maintained under realistic environmental conditions, can provide community-level
information. However, it is important that experiments such as these are carried out
over the annual cycle because of seasonal variability of phytoplankton communities
and variability in environmental characteristics of the estuary. In addition, other
contaminants will behave differently from arsenic. Arsenic was chosen as a model
contaminant for this research because, as a nutrient analogue, it reacts with the
lowest trophic levels. Other elements and organic compounds will have differing
geochemistries and necessarily different biological reactivity (Sanders & Riedel,
1987b). However, the approach and methods that are used here are applicable to
virtually any dissolved contaminant. We recommend that several more experiments
be performed using the same techniques as the integrative experiment. These
experiments would build on the knowledge and understanding gained within our
first experiment. At least one experiment should be performed with arsenic; others
should be performed with other contaminants for which we have adequate
background information (e.g., copper).
In addition, we need to continue to develop new ways to examine how
communities and ecosystems respond to chemical stress. Another method for
determining disturbance in an algal community may be a concept termed the
pollution induced community tolerance (PICT) (Blanck et al., 1987). This suggests
that a phytoplankton community, if disturbed by a pollutant, will respond by
developing an increased tolerance to that pollutant through the replacement of
sensitive species by more resistant ones. This can result either through selection for
70
-------
tolerant genotypes of the same species or a shift in species composition toward
resistant forms of other taxa. By itself, this concept allows simple, inexpensive
determinations of community disturbance and may be very useful in future
assessment of pollutant impact. Used in conjunction with the methods employed in
this study, PICT may provide important resolution into the mechanisms of
community shifts in the phytoplankton.
These two approaches by no means are the sole mechanisms available for
determining pollution-induced stress. There are others, and more will be introduced
with time. However, these techniques cannot stand alone. After a determination of
disturbance has been made, it is critical that we evaluate its effects and consider
whether in the context of the ecosystem the disturbance is likely to have a negative
impact. The application of new and innovative research tools permits such
evaluations to rely more on scientific judgement and less on guesswork.
71
-------
REFERENCES
Andreae, M. 0. 1978. Distribution and speciation of arsenic in natural waters and
some marine algae. Deep-Sea Res. 25:391-402.
Antia, N.J., C.D. McAllister, T.R. Parsons, K. Stevens and J.D.H. Strickland. 1963.
Further measurements on primary production using a large volume plastic
sphere. Limnol. Oceanogr. 8:166-184.
Barnes, H. 1959. Stomach content and micro-feeding of some common cirripedes.
Can. J.Zool.: 37:231-236.
Bartram, W. C. 1980. Experimental development of a model for the feeding of
neritic copepods on phytoplankton. J. Plankton Res. 3:25-51.
Berk, S.G., D.C. Brownlee, D.R. Heinle, H.J. Kling and R.R. Colwell. 1977. Ciliates
as a food source for marine planktonic copepods. Microb. Ecol. 4:27-40.
Berman, M. S. and D. R. Heinle. 1980. Modifications of the feeding behavior of
marine copepods by sub-lethal concentrations of water-accommodated fuel oil.
Mar. Biol. 56:59-64.
Bieri, R., O. Bricker, R. Byrne, R. Diaz, G. Helz, J. Hill, R. Huggett, R. Kerhin,
M. Nichols, E. Reinharz, L. Schaffner, D. Wildung and C. Strobel. 1982. Toxic
substances, p. 263-375. In E. G. Macalaster, D. A. Barker, and M. Kasper
(eds.), Chesapeake Bay Program technical studies: a synthesis. EPA,
Washington, D.C.
Blanck, H., S-A. Wangberg, and S. Molander. 1987. Pollution-induced community
tolerance (PICT) - a new ecotoxicological tool. In Tenth symposium on aquatic
toxicology, ASTM, in press.
Boyle, E. A. and S. Huestedt. 1983. Aspects of the surface distributions of copper,
72
-------
nickel, cadmium, and lead in the North Atlantic and North Pacific, p. 379-394.
In C. S. Wong, E. Boyle, K. W. Bruland, J. D. Burton, and E. D. Goldberg (eds.).
Trace metals in sea water. Plenum Press, New York.
Braman, R. S., D. L. Johnson, C. C. Foreback, J. M. Ammons and J. L. Bricker. 1977.
Separation and determination of nanogram amounts of inorganic arsenic and
methylarsenic compounds. Anal. Chem. 49:621-625.
Brownlee, D.C. and F. Jacobs. 1987. Mesozooplankton and microzooplankton in the
Chesapeake Bay, p. 217-269. In S.K. Majumdar, L.W. Hall, Jr., and H.M.
Austin (eds.), Contaminant problems and management of living Chesapeake
Bay resources. Pennsylvania Acad. Sci., Easton, PA.
Cairns, J. 1986. Community toxicity testing. ASTM publication 920, Philadelphia,
PA.
Capriulo, G. M. 1982. Feeding of field collected tintinnid microzooplankton on
natural food. Mar. Biol. 71:73-86.
Crisp, D. J. and A. J. Southward. 1961. Different types of cirral activity of
barnacles. Phil. Trans. Roy. Soc. B 243:271-308.
Dayton, P. K. 1975. Experimental evaluation of ecological dominance in a rocky
intertidal algal community. Ecol. Monog. 45:137-159.
D'Elia, C. F., J. G. Sanders, and W. R. Boynton. 1986. Nutrient enrichment studies
in a coastal plain estuary: phytoplankton growth in large-scale, continuous
cultures. Can. J. Fish. Aquat. Sci. 43:397-406.
Forstner, U. and G.T.W. Wittmann. 1983. Metal pollution in the aquatic
environment. Springer-Verlag, Berlin.
Grice, G.D. 1984. Use of enclosures in studying stress on plankton communities, p.
563-574. In H.H. White (ed.), Concepts in marine pollution measurements.
MD Sea Grant publication UM-SG-TS-84-03.
Grice, G. D. and M. R. Reeve. 1982. Introduction and description of experimental
73
-------
ecosystems, p. 1-10. In G.D. Grice and M.R. Reeve (eds.), Marine mesocosms.
Springer-Verlag, New York.
Heinbokel, J. R. 1978. Studies on the functional role of tintinnids in the Southern
California Bight. I. Grazing and growth rate in laboratory cultures. Mar. Biol.
47: 191-197.
Heinle, D. R. 1966. Production of a calanoid copepod, Acartia tonsa, in the Patuxent
River estuary. Chesapeake Sci. 7:59-74.
Hendrickson, P., K. G. Sellner, B. Rojas de Mendiola, N. Ochoa and R. Zimmermann.
1982. The composition of particulate organic matter and biomass in the
Peruvian upwelling region during ICANE 1977 (Nov. 14-Dec. 2). J. Plankton
Res. 4:163-186.
Keith, L. H. and W. A. Telliard. 1979. Priority pollutants. I-A perspective view.
Environ. Sci. Technol. 13:416-423.
Kennedy, V. S. and L. L. Breisch. 1981. Maryland's oysters: research and
management. Univ. of Maryland Sea Grant Publication TS-81-04. 286 pp.
Krsinic, F. 1987. On the ecology of tinitnnines in the Bay of Mali Ston (eastern
Adriatic). Est. Coast. Shelf Sci. 24:401-418.
Landry, M. R. 1977. A review of important concepts in the trophic organization of
pelagic ecosystems. Helgol. wiss. Meeres. 30:8-17.
Lessard, E.J., D.A. Caron, K. Ho and M.A. Voytek. 1987. Grazing impact and food
selection of nano-,micro-, and macrozooplankton in natural estuarine
communities. Trans. Amer. Geophys. Union 68:1782.
Lund, J. W. G., C. Kipling, and E. D. LeCren. 1958. The inverted microscope method
of estimating algal numbers and the statistical basis of estimations by
counting. Hydrobiologia 11:143-170.
McAllister, C.D., T.R. Parsons, K. Stevens and J.D.H. Strickland. 1961.
74
-------
Measurements of primary production in coastal sea water using a large volume
plastic sphere. Limnol. Oceanogr. 6:237-259.
Mackenzie, F. T., R. J. Lantzy, and V. Paterson. 1979. Global trace cycles and
predictions. Math. Geol. 11:99-144.
Martin, J-M. and M. Whitfield. 1983. The significance of the river input of chemical
elements to the ocean, p. 265-296. In C. S. Wong, E. Boyle, K. W. Bruland, J.
D. Burton, and E. D. Goldberg (eds.). Trace metals in sea water. Plenum Press,
New York.
Menzel, D.W. and J. Case. 1977. Concept and design: controlled ecosystem
pollution experiment. Bull. Mar. Sci. 27:1-7.
Nival, P. and S. Nival. 1976. Particle retention efficiencies of an herbivorous
cope pod, Acartia clausi (adult and copepodite stages): Effects on grazing.
Limnol. Oceanogr. 21:24-28.
Notini, M. and A. Rosemarin. 1986. Fate and effects of arsenic in a mesoscale model
ecosystem. Status report prepared for the ESTHER Project, Swedish
Environmental Research Group, Karlskrona, Sweden, January.
Notini, M., K. Holmgren and A. Rosemarin. 1988. Long-term fate and effects of low-
levels of arsenate on the Baltic Sea Fucus vesiculosus ecosystem. In
Proceedings, 21st european marine biology symposium, in press.
Osman, R. W. 1977. The establishment and development of a marine epifaunal
community. Ecol. Monog. 47:37-63.
Osman, R. W. 1982. Artificial substrates as ecological islands, pp. 71-1 i4. In J.
Cairns (ed.), Artificial Substrates. Ann Arbor Science Publ., Massachusetts.
Oviatt, C.A. 1984. Introduction: ecology as an experimental science and
management tool, p. 539-548. In H. H. White (ed.), Concepts in marine
pollution measurements. MD Sea Grant publication UM-SG-TS-84-03.
Paine, R. T. 1969. A note on trophic complexity and community stability. Am. Nat.
75
-------
100:65-75.
Parrish, K. K. and D. F. Wilson. 1978. Fecundity studies on Acartia tonsa
(Copepoda:Calanoida) in standardized culture. Mar. Biol. 46:65-81.
Parsons, T. R. 1976. The structure of life in the sea, pp. 81-97. In D. H. Cushing
and J. J. Walsh (eds.), The ecology of the seas. W. B. Saunders Co.,
Philadelphia.
Parsons, T. R., R. J. LeBrasseur, J. D. Fulton, and 0. D. Kennedy. 1969. Production
studies in the Straits of Georgia. EL Secondary production under the Fraser
River plume, February to May, 1967. J. Exp. Mar. Biol. Ecol. 3:39-50.
Peoples, S. A. 1975. Review of arsenical pesticides, p. 1-12. In E. A. Woolson (ed.),
Arsenical pesticides. ACS Symp. Ser. 7, American Chemical Society.
Pritchard, P.H. and A.W. Bourquin. 1984. A perspective on the role of microcosms
in environmental fate and effects assessments, p. 117-138. In H.H. White (ed.),
Concepts in marine pollution measurements. MD Sea Grant publication UM-
SG-TS-84-03.
Provasoli, L. 1977. Cultivation of animals, pp. 1295-1319. In Kinne, O. (ed.)
Marine ecology, Vol.HQ, Cultivation, Part 3. Wiley, London.
Rassoulzadegan, F. 1982. Dependence of grazing rate, gross growth efficiency and
food size range on temperature in a pelagic oligotrichous ciliate Lohmanniella
spiralis Leeg., fed on naturally occurring particulate matter. Ann. Inst.
Oceanogr. 58(2):177-184.
Robertson, J.R. 1983. Predation by estuarine zooplankton on tintinnid ciliates. Est.
Coast. Shelf Sci. 16:27-36.
Rosemarin, A. and M. Notini. 1988. Structural and functional effects of low levels of
arsenic in a Baltic Sea littoral model ecosystem. Unpublished manuscript.
Rosemarin, A., M. Notini, and K. Holmgren. 1985. The fate of arsenic in the Baltic
Sea Fucus vesiculosus ecosystem. Ambio 14:342-345.
76
-------
Ryther, J. H. 1969. Photosynthesis and fish production in the sea. Science 166:72-
76.
Ryther, J. H. and J. G. Sanders. 1980. Experimental evidence of zooplankton control
of the species composition and size distribution of marine phytoplankton. Mar.
Ecol. Prog. Ser. 3:279-283.
Sanders, J. G. 1978. Interactions between arsenic and marine algae. Ph.D.
Dissertation, University of North Carolina, Chapel Hill, 84 p.
Sanders. J. G. 1979. The concentration and speciation of arsenic in marine
macroalgae.Est. Coast. Mar. Sci. 9:95-99.
Sanders, J. G. 1980. Arsenic cycling in marine systems. Mar. Environ. Res. 3:257-
266.
Sanders, J. G. 1983. Role of marine phytoplankton in determining the chemical
speciation and biogeochemical cycling of arsenic. Can. J. Fi.sh. AquSt. Sci.
40:192-196.
Sanders, J. G. 1985. Arsenic geochemistry in Chesapeake Bay: dependence upon
anthropogenic inputs and phytoplankton species composition. Mar. Chem.
17:329-340.
Sanders, J. G. 1986. Direct and indirect effects of arsenic on the survival and
fecundity of estuarine zooplankton. Can. J. Fish. Aquat. Sci. 43:694-699.
Sanders, J.G. and S.J. Cibik. 1985. Adaptive behavior of euryhaline phytoplankton
communities to arsenic stress. Mar. Ecol. Prog. Ser. 22:199-205.
Sanders, J. G. and G. F. Riedel. 1987a. Control of trace element toxicity by
phytoplankton, p. 131-149. In J. A. Saunders, L. Kosak-Channing, and E. E.
Conn (eds.), Recent advances in phytochemistry, vol 21. Plenum Press.
Sanders, J.G. and G.F. Riedel. 1987b. Chemical and physical processes influencing
77
-------
bioavailability of toxics in estuaries, p. 87-106. MM.P. Lynch andE.C. Krome
(eds.), Perspectives on the Chesapeake Bay: advances in estuarine sciences.
Chesapeake Research Consortium Publ. No. 127.
Sanders, J. G. and P. S. Vermersch. 1982. Response of marine phytoplankton to low
levels of arsenate. J. Plankton Res. 4:881-893.
Sanders, J. G. and H. L. Windom. 1980. The uptake and reduction of arsenic species
by marine algae. Est. Coast. Mar. Sci. 10:555-567.
Sanders, J. G., J. H. Ryther, and J. H. Batchelder. 1981. Effects of copper, chlorine,
and thermal addition on the species composition of marine phytoplankton. J.
Exp. Mar. Biol. Ecol. 49:81-102.
Sanders, J. G., R. W. Osman, and K. G. Sellner. 1987. Arsenic impact on growth,
fecundity, species composition and subsequent transport of arsenic in estuarine
food webs. Final report to EPA Office of Research and Development under
grant R810680-01. April 1987.
Southward, A. J. 1955a. Feeding of barnacles. Nature 175:1124.
Southward, A. J. 1955b. On the behaviour of barnacles. I. The relation of cirral and
other activities to temperature. J. Mar. Biol. Assoc. U. K. 34:403-422.
Stoecker, D.K. and D.A. EglofT. 1987. Predation by Acartia tonsa Dana on
planktonic ciliates and rotifers. J. Exp. Mar. Biol. Ecol. 110:53-68.
Stoecker, D.K. and N.K. Sanders. 1985. Differential grazing by Acartia tonsa on a
dinoflagellate and a tintinnid. J. Plank. Res. 7:85-100.
Strickland, J.D.H. and L.D.B. Terhune. 1961. The study of in situ marine
photosynthesis using a large plastic bag. Limnol. Oceanogr. 6:93-96.
Uthe, J. F., H. C. Freeman, J. R. Johnston, and P. Michalik. 1974. Comparison of
wet ashing and dry ashing for the determination of arsenic in marine
organisms, using methylated arsenicals for standards. J. Assoc. Off. Anal.
Chem. 57:1363-1365.
78
-------
Verity, P.G. 1987. Factors driving changes in the pelagic trophic structure of
estuaries, with implications for the Chesapeake Bay, p. 35-56. In M.P. Lynch
and E.C. Krome (eds.), Perspectives on the Chesapeake Bay: advances in
estuarine sciences. Chesapeake Research Consortium Publ. No. 127.
Waslenchuk, D. G. 1978. The budget and geochemistry of arsenic in a continental
shelf environment. Mar. Chem. 7:39-52.
White, H. H. 1984. Concepts in marine pollution measurements. MD Sea Grant
publication UM-SG-TS-84-03, College Park, MD.
Zaroogian, G. E. and G. L. Hoffman. 1982. Arsenic uptake and loss in the American
oyster, Crassostrea virginica. Environ. Monit. Assess. 1:345-358.
79
-------
INTRODUCTION
Estuaries serve as principal conduits for man's inputs to the oceans. Many
industrial and municipal activities, such as power generating stations, wastewater
treatment facilities, and industrial plants of all types, are located on estuaries.
These activities generate liquid and solid wastes, some of which deliberately or
accidentally are discharged into the nearby water. In addition, the concentration of
boat traffic in these same waterways results in many small spills and leakages
which add pollutants. Consequently, pollution of an estuary is often the initial and
greatest impact borne by the marine system as a whole. Therefore, estuaries are
necessary places to investigate potential impacts of toxic compounds.
Zero discharge of toxic wastes into marine and estuarine environments,
although an enviable goal, may never be a realistic alternative for modern man.
Even if deliberate introduction of toxic compounds does not occur, the oceans will
always face the inevitable, accidental discharge of these materials. Many
compounds within these discharges are of extreme interest to environmental
planners because:
1. they are greatly elevated in affected waters relative to natural water
concentrations,
2. they are toxic to biota,
3. they are actively accumulated by biota in aquatic systems.
We do not know to what extent chronic or acute discharges of industrial wastes
will alter aquatic ecosystems. The literature suggests that the trace materials likely
to be contained within these wastes are toxic to some organisms. However, the
results from these single-species bioassays are difficult to extrapolate to estuaries
which have complex hydrography and food webs. Sessile organisms in the vicinity of
1
-------
an effluent have the greatest potential for harm, but plankton contained within the
effluent plume also should be affected. Additionally, if the compound is bioactive,
uptake and transformation by organisms will greatly affect its toxicity and transfer
to other organisms within the estuary.
Arsenic is an example of this type of compound. It is a classic poison, with its
primary commercial use as a pesticide (Mackenzie et al., 1979). It is considered a
"priority pollutant" by EPA, and was present in 48% of industrial effluents tested by
the Agency (Keith & Telliard, 1979). Arsenic is present in all aquatic systems,
principally in the form of an inorganic ion, arsenate (Waslenchuk, 1978; Sanders,
1980). Reduced arsenic (arsenite) is generally considered more toxic and is used as a
pesticide (Peoples, 1975). However, at low levels, arsenate is likely to have the
greatest impact on aquatic primary producers. It is chemically similar to phosphate,
a necessary nutrient for plant growth, and it is readily taken up by phytoplankton
(Sanders & Windom, 1980). Arsenate uptake can be followed by accumulation
within the cell, or it can be chemically altered (reduced to arsenite and methylated)
and released into the surrounding water. As much as 80% of the dissolved arsenic
present in coastal marine systems is taken up and released in this fashion (Sanders,
1980,1983, 19bo).
All organisms do not react similarly to a given pollutant. Although
phytoplankton are most affected by arsenate, the least toxic form, invertebrates are
more susceptible to arsenite and methylated forms (Peoples, 1975). Therefore, they
could be harmed by the biogeochemical alterations discussed above. Even within a
particular taxon, all species will not react to the same levels of arsenic. For example,
all phytoplankton species are not equally sensitive to arsenate; some are inhibited at
levels just exceeding ambient concentrations of arsenate in the oceans (1.0-1.5 pg l"1;
Andreae, 1978; Waslenchuk, 1978; Sanders, 1980), while others are resistant to
arsenate concentrations two orders of magnitude higher (Sanders and Vermersch,
2
-------
1982). Within the phytoplankton community, where rapid growth and species
succession are normal occurrences, arsenic can cause a shift in the dominant species
toward resistant forms and a concurrent decrease in productivity (Sanders and
Vermersch, 1982; Sanders and Cibik, 1985). Therefore, the addition of arsenic to the
water column not only can have a direct toxic effect, but also can result in dramatic
shifts in both the dominance within the phytoplankton community and possibly the
flux of carbon and nitrogen between trophic levels. Coupled with this is the
alteration of chemical form, which will affect (generally increase) arsenic toxicity to
higher trophic levels.
Beyond the single trophic level, the effects of a toxic substance (i.e., arsenic) on
an ecosystem are potentially quite complex. Besides the direct effects postulated
above, individuals of heterotrophic species can be influenced by either changes in the
abundances and sizes of prey species or the contamination of these food organisms.
*
For example, an arsenic-induced shift in phytoplankton composition to small
flagellates or small centric diatoms could reduce the ingestion of phytoplankton by
copepods and therefore their fecundity. Small cells are not captured as efficiently as
larger diatoms (e.g., Nival and Nival, 1976; Bartram, 1980). Thus changes in
phytoplankton size may ultimately result in an abundance of small, noncrustacean
grazers (tintinnids, rotifers) that can effectively feed on flagellates and a
concomitant decrease in larger zooplankton. This shift, extrapolated to the next
trophic level, predicts that the same aquatic system will support a lowered density of
harvestable fish (see Ryther, 1969; Parsons, 1976; Landry, 1977; Hendrikson et al.,
1982).
Little is known about how such primary and secondary effects of a toxin
combine to change a community or ecosystem. Except for simple systems with two or
three species or communities dominated by a "keystone species" (e.g., Paine, 1969;
Dayton, 1975), we have no models, empirical or theoretical, which could be used to
3
-------
predict how (or if) changes to a single species might alter the whole system.
Therefore, it is important that the contributions of the various pathways to an
overall toxin-effect be determined for aquatic ecosystems.
We have examined the effects of arsenic on three trophic levels within a
representative estuarine food web. Our purpose was to distinguish and quantify
those changes in species abundance, mortality, growth rates, and reproduction that
are caused directly by the dissolved arsenic and those that could be a function of
trophic relationships. The three trophic levels we investigated were the
phytoplankton assemblage, zooplanktonic herbivores, and benthic supension-feeders
ingesting phytoplankton and/or zooplankton. We have chosen this food web because
1) trophic relationships are straight-forward, 2) all species can be easily manipulated
for experimentation, and 3) many species are economically important within the
estuary (e.g., the oyster, Crassustrea virginica) or are important in many estuarine
food webs (e.g., phytoplankton and zooplankton, Sellner, 1983,1987).
4
-------
METHODS
EXPERIMENTAL PROCEDURES
Experiments were conducted either in closed-system laboratory tanks or in
outdoor flow-through microcosms. The outdoor systems were ideal for studies
utilizing natural phytoplankton. Experiments with representative species were
performed indoors in continuous flow systems. All treatments in each experiment
(including controls) were replicated three times.
Natural phytoplankton and zooplankton communities were sampled locally
from the Patuxent River estuary, a subestuary of the Chesapeake Bay. Benthic
species were either harvested from local populations or collected by exposing
artificial substrates on which larvae of the chosen species attached (fouling panels
similar to those used by Osman, 1977, 1982). The three trophic levels and
experimental design are described below in detail.
Phytoplankton Experiments
The phytoplankton are dominated by diatoms and a variety of flagellates at
different times of year. By timing experiments to this temporal change in species
composition, we studied the reaction of diatom and microflagellate dominated
communities. Algal densities in the Patuxent River are high, approximately 106 -
107 cells H, similar to densities found in other mesohaline regions of the
Chesapeake Bay (e.g., Van Valkenburg et al., 1978; Sanders, 1985; Sellner, 1987).
Natural assemblages from the river were cultured outdoors in large-volume
microscosms under ambient nutrient concentrations and natural light and
temperature conditions. The tanks were operated as flow-through cultures (see
Sanders et al., 1981; Sanders and Cibik, 1985 for details of design). Under these
conditions, the natural assemblage can be maintained for weeks at cell densities
5
-------
similar to those in the river (Sanders et al., 1981; Sanders and Vermersch, 1982).
The advantage of this approach was that we could use actual multi-species
assemblages (rather than laboratory clones), with replication, that had not been
exposed to high or artificial light and nutrient conditions.
Zooplankton Experiments
The zooplankton community of the Patuxent River is dominated by high
densities of copepods, with Acartia tonsa forming >70% of total numbers in
mesohaline waters in spring-summer-fall and Eurytemora affinis the principal
winter copepod (Heinle, 1966; Heinle and Flemer, 1975; ANS, 1981; Sellner and
Horwitz, 1982; Brownlee and Jacobs, 1987). Maximum densities and highest growth
rates for A. tonsa occur in July and August (Heinle, 1966; ANS, 1981; Sellner and
Horwitz, 1982) while E. affinis densities in the lower estuary peak in March (Heinle,
1969). Although copepods form the largest component of the zooplankton biomass,
rotifers, ciliates, and bivalve larvae also contribute significantly to the community,
the latter group in late spring (Sellner and Horwitz, 1982; Brownlee and Jacobs,
1987).
Two sets of experiments (one with Acartia, one with Eurytemora) were
conducted with the initiation of each dependent on seasonal abundances of the two
species. The direct effects of sublethal concentrations of dissolved arsenic on these
copepods were determined in a series of experiments using long-term culture
procedures and short-term grazing techniques. Culture experiments were
undertaken to determine the effects of dissolved inorganic arsenic on copepod
survival and egg and nauplii production, e.g, whether exposure over the life of adult
copepods resulted in significant alterations in longevity or reproductive potential.
Grazing experiments were conducted to determine whether chronic, low-level
arsenic exposures resulted in decreased feeding ability of the copepod. If arsenic
6
-------
reduced feeding and carbon intake, energy reserves possibly used for reproduction
could decline resulting in lower fecundity and eventually population mortality.
Eurytemora—Laboratory cultures of E. affinis were acclimated to 15°C and 11-
14%r> salinity. Cultures were fed saturating levels of Thalassiosira weisflogii and
Isochrysis sp. After isolation of egg-bearing females, all nauplii produced over a 2-d
period were collected and sorted into 4 groups, a group for arsenic levels of 0,15, 35
and 50 pg H. Approximately 200 nauplii were placed in individual 1.5 liter flasks
containing phytoplankton noted above and after acclimation overnight, >63 pm
nauplii were transferred to control and arsenic solutions without phytoplankton.
Approximately 6 h later, animals were returned to solutions containing equal
portions of T. weisflogii and Isochrysis sp.
Direct effects of arsenic on the copepod populations were determined using a
diel feeding and exposure schedule. Thirty-five adult females and 7 adult males (Fo
generation) were placed in individual 500 ml containers in a water bath.
Populations were transferred to solutions containing saturating mixtures of the two
phytoplankton species for overnight feeding in darkness. Each morning, the
populations were removed from each container and gently rinsed free of cells and
placed in solutions free of phytoplankton but containing the four concentrations of
dissolved arsenic. The solutions from the overnight feeding period were passed
through 53 pm mesh; trapped material (nauplii and eggs) was washed into small
vials and preserved with buffered formalin. Numbers of dead adults were noted and
the adults were transferred to the vials. In vivo fluorescence measurements were
made on the phytoplankton cultures from the overnight feeding period and after
addition of fresh phytoplankton mixtures to initial prefeeding saturating levels.
Copepods in phytoplankton-free arsenic solutions were placed in a constant
temperature water bath under low light for the day. In the evening, copepods were
returned to phytoplankton cultures without arsenic with the nauplii and eggs
7
-------
produced in the day added to the vials containing preserved nauplii and eggs from
the morning.
In order to estimate long-term effects on reproductive potential of Eurytemora,
nauplii from adults reared in arsenic were transferred to four arsenic levels and
cultured for 19 d. At the end of this period, these Fj populations were sacrificed and
preserved with buffered formalin. Numbers of females (egg-bearing and total), adult
males and immature stages were determined. In addition, offspring from this
population were hatched (F2 generation) and reared for 17 d to determine numbers of
mature and immature individuals.
Grazing estimates for the initial populations (Fo generation) were estimated
using two techniques. Daily changes in in vivo fluorescence were determined from
the differences between initial and final fluorescence values for each nocturnal
feeding period (see above). Radioisotopes were also employed where phytoplankton
were pre-labelled with '4C-bicarbonate and offered to replicate subsample
populations from each of the arsenic concentrations 4 d after exposure to arsenic.
Adult Eurytemora were allowed to feed for 2.9-3.1 h, screened, held over fuming,
concentrated HC1, and transferred to scintillation vials. Following digestion,
activities of phytoplankton and copepods were estimated using liquid scintillation
counting procedures (Sellner and Olson, 1985).
Acarfia-Experiments with Acartia tonsa were conducted in a similar but
simpler manner. Populations were collected from Chesapeake Bay in August, 1985
and returned to the laboratory. Nauplii were collected from the populations and
transferred as 4 distinct groups to 25°C, 15 %o solutions containing 0, 15, 35 and 50
yg l-l of arsenate. Feeding was undertaken as described above for Eurytemora.
After 1-10 d, three groups of 35 females and 7 males were isolated from each arsenic
concentration and transferred to fresh solutions. Using the protocols for long-term
culture experiments for estimating nauplii produced per day described above,
8
-------
nauplii numbers, fluorescence changes and grazing rates were determined over a 3-
week period.
The effects of arsenic concentrations and day on each of these parameters were
determined using ANOVA procedures.
Benthic Experiments
Four species of benthic suspension-feeding invertebrates were used in the
study. Species were chosen to represent both a broad range of feeding types and a
diversity of taxonomic groups. Each of the four species can be locally dominant,
several are classified as representative important species, and one, the oyster,
Crassostrea virginica, is economically important.
The experiments with benthic invertebrates were designed to examine the
direct effects of dissolved arsenic on a variety of common estuarine species. A
general experimental design was used in all experiments with this design modified
to accommodate differences between the species examined.
In general, two sets of environmental chambers were used in each experiment.
These were experimental chambers in which the organisms were kept for most of an
experiment and exposed to a particular level of arsenic and the feeding chambers in
which the organisms were placed to feed in isolation from the experimental
conditions. Using this design, the organisms were exposed to dissolved arsenic only,
and could not assimilate arsenic by ingestion of contaminated food organisms. In
each experiment, the experimental animals were rinsed in flowing filtered water
before being placed in either the experimental or feeding chambers.
Four species were used in a total of six experiments. Three experiments were
conducted with the oyster Crassostrea virginica. Separate experiments were
conducted with larval oysters, newly settled juvenile oysters, and adults. These
experiments were designed to examine whether this common and important species
9
-------
was sensitive in a particular life-history stage. In the remaining three experiments
the barnacle Balanus improuisus, the colonial bryozoan Victorella pavida, and the
anemone Diadumene leucolena were investigated. All four species are very common
within the Chesapeake Bay and together are representative of the phylogenetic,
morphologic, behavioral, and trophic groups common to estuaries. The species were
all from very different phyla; the bryozoan was colonial while the remaining species
were solitary, Diadumene had limited motility compared to the others which were
permanently attached to substrates, and the oysters and Victorella fed exclusively on
phytoplankton, Diadumene fed on zooplankton, and Balanus was omnivorous.
Barnacles--Experimental populations of the barnacle, Balanus improuisus,
were collected on 115 cm2 PVC panels. Approximately 200 panels were exposed in
the lower Patuxent River estuary 45 days prior to the beginning of the experiment.
These panels were collected, cleaned of any debris, and all organisms other than
Balanus were removed. In particular, the predatory flatworm Stylochus elhpticus
was carefully removed from all panels. After being cleaned, each panel was blotted
dry, numbered, and weighed. Twenty randomly chosen panels were then assigned to
one of nine 80 1 environmental chambers. Chambers were then assigned randomly
to one of three arsenic treatments, resulting in 3 replicate chambers per treatment.
Treatments consisted of control, low, and high concentrations of arsenate (0,
12, and 42 pg H, respectively). Arsenate levels were maintained using the
continuous flow design discussed previously. The experiment continued for 22 days
during which time the water temperature fluctuated between 10 and 12°C and
salinity was 8-9%o. For 3-4 h each day panels were placed in the feeding chambers.
These chambers were supplied with a continuous flow of raw (unfiltered) Patuxent
River water.
10
-------
As a control for the abbreviated feeding time, 6 panels were kept in a
continuous flow of raw water for the duration of the experiment. Also another 6
panels were selected haphazardly at the beginning of the experiment and frozen for
later analysis.
After 22 d of exposure to arsenic, panels were placed in continuously-flowing
seawater for an additional 10 d. This allowed us to test for any residual effects of
exposure after this short recovery period. At the end of this 10 d period, all panels
were dried and weighed. All barnacles on each panel were sampled and frozen for
later analysis.
Bryozoans--The methods used for the bryozoan experiment were very similar to
those used for the barnacle experiment. Two hundred previously numbered PVC
panels were suspended in the Patuxent River for 45 days prior to the experiment.
After this time each panel was colonized and completely covered by the bryozoan,
Victorella pauida. Panels were recovered two days before the beginning of the
experiment. Victorella was completely removed from one surface of each panel and
the remaining surface was designated the test surface. A border 1-2 cm wide was
cleared of all Victorella, leaving a central area of approximately 50-75 cm2 covered
by the bryozoan.
After these manipulations each panel was photographed to record the
abundance of Victorella in terms of area covered. At the end of the experiment each
panel was again photographed to record the change in cover during the experiment.
The clean border and back surface of the panel were designed to supply sufficient
space for colony growth.
As in the barnacle experiment, 20 panels were randomly assigned to one of 9
environmental chambers. Three replicate chambers were assigned to each of three
arsenic treatments (0,13, and 36 pg H). Panels were kept in the experimental
11
-------
chambers for 20 h per day. For 4 h each day panels were removed and placed in the
feeding chambers. In these chambers Victorella colonies were exposed to unfiltered
river water, allowing them to feed on their normal diet of phytoplankton.
Anemone.s--The anemone, Diadumene leucolena, was collected from oyster reefs
in the Patuxent River and Chesapeake Bay. Individual anemones were carefully
removed from the oyster shell. Each individual was placed on a separate 115 cm2
PVC panel. Panels were held in flowing water in a laboratory seatable. After
approximately 24 h most anemones had reattached to the experimental substrates.
Anemones were held for 2 weeks in the laboratory seatable and any unhealthy or
damaged individuals were removed. Each individual was fed daily with an
abundance of nauplii of cultured brine shrimp, Artemia sp.
At the beginning of the experiment, panels with attached anemones were
photographed, blotted dry, and weighed. Panels were then assigned randomly to one
of 9 environmental chambers resulting in 6 - 7 panels in each chamber. As in the
previous experiments the 9 chambers were randomly assigned to one of three arsenic
treatments (0, 13, and 36 pg H).
Panels were kept in the experimental chambers for 20 h per day. During the
remaining 4 h panels were placed in feeding chambers, each with an equal volume of
water. The feeding chambers were filled with filtered river water immediately
before the anemones were added and were constantly aerated during the feeding
period. An equal volume of Artemia nauplii were added to each chamber with this
volume dependent on the productivity of laboratory cultures.
At the end of the experiment each panel was photographed, blotted dry, and
weighed. Anemones were then removed for arsenic analyses and the panel was
reweighed to determine its weight alone. The growth of the anemones over the 30
day experiment was determined both as a change in wet weight and as a change in
12
-------
area of the basal disc. Areas were measured on photographs using the laboratory's
image analysis system.
Adult Oy.siers--Approximately 400 cultured oysters were obtained from a local
hatchery and acclimated to the Patuxent River for two weeks prior to the beginning
of the experiment. Each oyster was cleaned and 180 were numbered and weighed at
the beginning of the experiment. Weights were measured using the technique of
Andrews (1961). Twenty numbered and twenty unnumbered oysters were randomly
assigned to each of the 9 treatment chambers and chambers were then randomly
assigned to one of 3 arsenic treatments (0,13, and 35 pg 1-1). An additional 20
oysters were sacrificed at the beginning of the experiment to measure initial tissue
levels of arsenic. Tissue was carefully removed, weighed, and dried for analysis.
Finally, 10 oysters were maintained in constantly flowing river water for the
duration of the experiment. These were used to evaluate the reductions in growth
resulting from the limited feeding schedule used during the experiment.
The experiment ran for a period of four weeks. Oysters were kept in the
experimental chambers for 20 h per day and moved to the feeding chambers for the
remaining 4 h. In the feeding chambers the oysters were exposed to a continuous
flow of unfiltered river water. After 2 weeks 10 of the unnumbered oysters were
chosen haphazardly, and the tissue removed and dried for analysis. At the end of the
experiment the numbered oysters were weighed and the tissue of all oysters was
collected and prepared for analysis.
Larval Oysters--ln this experiment, conducted simultaneously with the
Juvenile Oyster Experiment, 3 replicate chambers were used for each treatment.
Oyster larvae, approximately 4 d old were obtained from a hatchery on the eastern
shore of Maryland. After transport to the laboratory an equal volume of larvae were
13
-------
added to each of 6 environmental chambers. Random counts indicated that each
chamber contained a mean of 187 ±20 larvae ml-l. Chambers were randomly
assigned to one of two treatments (0, 25 pg H).
As in previous experiments, larvae were fed in isolation from dissolved arsenic.
However, to prevent stress and injury resulting from daily transport between
chambers, separate feeding chambers were not used and larvae were fed in the
experimental chambers. Prior to feeding water was gently siphoned from the
chambers using plankton net filters to prevent removal of larvae. After removal of
the treatment water, Isochrysis sp. was added to each chamber to bring cell densities
to 104 cells ml-l. Cell counts were made daily to ensure that densities were the same
in all treatments. These experiments were designed to continue for 10 -15 days. The
number of larvae successfully settling on substrates in each chamber were used to
estimate the effects of each treatment.
Juvenile Oysters--This experiment was conducted using newly attached
oysters. Cultured oyster larvae were exposed to 60 115 cm2 panels in an oyster
hatchery on the eastern shore of Maryland. The panels with attached larvae were
then transported to the laboratory, placed in one of three groups depending on the
density of oysters, and an equal number from each of these groups was randomly
assigned to one of 12 treatment tanks. Of the 12 chambers, only 6 were relevant to
the present study. These were assigned randomly to one of two arsenic treatments
(0, 25 pg-H). As in previous experiments, the continuous flow design was used and
panels were removed to separate feeding chambers for 4 h per day. Feeding
chambers were aerated and filled with static cultures of Isochrysis sp. Algal
concentrations ranged between 1 x 104 and 2 x 104 cellsmH during the course of the
experiment.
14
-------
The experiment was continued for 4 weeks and each experimental panel was
photographed at 0, 2, and 4 weeks. The area covered by each individual oyster was
used as a measure of its size and growth. These measurements were made by
analyzing the photographs using the laboratory's image analysis system. Oysters
showing no growth were considered unsuccessful recruits and were assumed to have
died before the beginning of the experiment. They were not included in any growth
rate analyses.
Arsenic Analyses
The concentration and chemical form of arsenic within each experiment were
monitored within the water column and organisms. Water samples were collected in
rigorously cleaned (Boyle and Huested, 1983) plastic bottles and analyzed by hydride
generation and detection of specific arsenic hydrides using atomic absorption
spectrometry (Braman et al., 1977). This method of analysis permits determination
of the total concentration of arsenic and also its chemical form. This technique is
very sensitive; limits of detection in our laboratory are about 20 ng 1-1. Solids were
dried, weighed, and ashed at 500°C for 24 hr in the presence of an ashing aid
[Mg(N(>3) and MgO] to prevent loss of arsenic (Uthe et al., 1974). After ashing, the
residue was dissolved in IN HC1 and analyzed as above.
Technique accuracy was assessed through the use of standard reference
materials, NBS #1566, oyster tissue, and NRC NASS-1, a seawater standard.
Recoveries of these materials averaged 95% and 91% for the oyster tissue and
seawater, respectively.
15
-------
RESULTS
PHYTOPLANKTON
Two experiments were performed, timed to coincide with different dominant
algal forms within the phytoplankton community: one during spring, to coincide
with the diatom/dinoflagellate bloom, and one during late summer, a time of
microflagellate dominance.
Spring Experiment—The spring experiment was conducted during 18 April and
17 May. Water temperatures rose during the experiment, and ranged between 16-
25°C. Salinity varied slightly, ranging between 11.0-13.5%o. Arsenate was added
continuously at three estimated concentrations: 3, 6, and 21 pg H, plus controls.
Each concentration of arsenate was added to triplicate sets of tanks. Actual
measured arsenate concentrations were very close to predicted values (Table 1) with
the low and medium treatments receiving approximately 30% more arsenic than
predicted. Arsenic concentrations in control tanks and the Patuxent River (source of
seawater for the experiments) were similar, approximately 0.3 pg l1 (Table 1).
Arsenate had little effect on the overall growth rate or total cell densities of the
assemblages, even at the highest concentration. In fact, tanks exposed to the various
levels of arsenate contained slightly higher cell densities after the first 12 d,
resulting from the elevated growth of a small centric diatom, Thalassiosira sp. (see
below). However, the higher densities were not significantly different from control
densities (Figure 1A).
Diversity also remained constant between the controls and various treatments.
Number of species, and measurements of species richness (Margalef, 1951) and
16
-------
Table 1. Dissolved arsenic concentration in phytoplankton experiments.
Values shown are mean ± S.E., in pg-H. All arsenic present was
in the form of arsenate, no reduced or methylated forms were
present.
Added As,
pgl-i
Actual As,
yg H ± SE
0 (source)
0.32
±
0.06
0 (control tanks)
0.34
±
0.03
3
4.25
±
0.29
6
8.60
±
0.62
21
19.8
±
0.79
0 (source)
1.22
±
0.13
0 (control tanks)
1.29
±
0.15
3
11.0
±
1.36
6
22.3
±
2.43
21
62.0
±
9.14
Spring Experiment
Summer Experiment
17
-------
Figure 1A.
Changes in total cell density at various arsenate concentrations
through time, in the spring experiment. ® ® = control (range
indicates 2 SE), o a = 3 pg-H- A A = 6 pg-H, ^^ = 21 pg-1'1.
NUMBER OF DAYS
18
-------
Figure IB. Changes in density of Cerataulina pelagica as a % of total cell density at
various arsenate concentrations through time, in the spring
experiment. • • = control, n a = 3 pg l-l. A A = 6 pg H,
«—» = 21pgM.
100-
80-
_ 60-
40-
20-
10 15 20
NUMBER OF DAYS
19
-------
Figure 1C. Changes in density of Thalassiosira sp. as a % of total cell density at
various arsenate concentrations through time, in the spring
experiment. « « — control, o a — 3 pg-l-i.A A = Gpg-l-1,
~ # = 21pg.H.
a
1/1
o
* CO
o
N
'en
c
a;
a
-------
species evenness (Pielou, 1966) showed no significant differences between
treatments and controls (ANCOVA, p>0.05).
A number of dominant species, however, were greatly affected by arsenate. All
three arsenate levels significantly inhibited growth of the most dominant algal
species, Cerataulina pelagica, a relatively large centric diatom. This species
comprised 40-80% of the total cell densities in control tanks throughout the
experiment. It rapidly disappeared, however, in arsenate-treated tanks; after only 7
d, it was present at densities less than 5% of total (Figure IB). After 18 d, cell
densities of this species rose slightly in treated tanks, but never approached the
densities seen in control tanks.
On the other hand, another important centric diatom, Thalassiosira sp., was
not affected by arsenate treatment; its growth was accelerated in treated tanks.
Thalassiosira sp. bloomed only in arsenate-treated tanks; densities of this species in
control tanks were quite low (Figure 1C). Another diatom, Rhizoaolenia
fragilissima, was not affected by arsenate: The combined density of these two species
more than offset the loss of C. pelagica in control assemblages; therefore, the total
number of centric diatoms remained essentially unchanged due to arsenate, even
though the most important dominant was almost eliminated in treatment tanks.
Other species, of lesser importance numerically, were similar in both controls
and treatments. The relative abundance of dinoflagellates and small, unidentified
flagellates increased slightly in arsenate-treated assemblages. Other species
showed no response to arsenate.
Late Summer Experiment-This experiment was performed between 5 and 29
September. Water temperatures were high at the beginning and fell slowly
throughout the experiment, ranging between 29.0-2l.0°C. Salinities varied between
11.5-14.5%o. Arsenate was added at three concentrations: 3, 6, and 21 pg l-l. Actual
21
-------
arsenate concentrations, however, were approximately three times these levels
(Table 1).
As in the spring, arsenate had little effect on the overall growth rate of the
assemblage or the total cell densities, at least for the first 15 d (Figure 2A). After
this period, cell densities declined somewhat relative to controls in the tanks
receiving the highest arsenic concentration. At the end of the experiment, cell
densities in these tanks were only 33% of those in controls (Figure 2A). This decline
was due solely to the decline of one sensitive diatom species (discussed below). As in
the earlier experiment, species diversity in the various treatments did not vary
significantly (ANCOVA, p>0.05).
Species composition during this experiment was quite different than during the
spring experiment. The spring experiment was dominated by centric diatoms, which
comprised an average of 79.6% of total cell density. In the summer experiment,
diatoms were much less important, averaging 42.1%. The assemblages in the
summer experiment were dominated by a number of small, flagellated species,
particularly cryptophytes such as Chroomonas sp., and other small, unidentified
flagellates. During the last 7 d of the experiment, a centric diatom, Chaetoceros
debile, bloomed, dominating until the end of the experiment. C. debile exhibited the
greatest response to arsenate. Its growth rate was reduced from 1.44 div d-i in
control assemblages to 1.28 div d-1 in the 6 pg-H arsenate treatment, and finally to
1.10 div-d"1 in the 21 yg H arsenate treatment, a 24% reduction in growth rate from
controls to the highest arsenic concentration (Figure 2B). Because this species
strongly dominated the phytoplankton for the last 10 d of the experiment, the overall
assemblage showed a significant reduction in total cell density, as discussed above
(Figure 2).
22
-------
Figure 2. Changes in total cell density and density of Chaetoceros debile as a % of
total cell density at various arsenate concentrations through time in the
summer experiment. • • = control, a a = 3 pg-l-l, ~ A =
6pg-H, +—^ = 21 pg-l-l.
10S
107-
106
0
100-1
S 80-
oo
c
CD
Q
_ 60-
-------
Other dominant spectes showed little, if any, response to arsenate. Growth of
cryptophytes was slightly reduced by arsenate; small, unidentified flagellates grew
slightly better in arsenate treatments; both differences were slight.
ZOOPLANKTON
Eurytemora affinis-Eurytemora were maintained at four estimated arsenic
concentrations, 0,15, 35, and 50 yg H. Actual arsenic concentrations were quite
similar to predicted levels; the two highest concentrations were slightly lower than
predicted (Table 2). Data collected from replicate cultures of Eurytemora affinis
indicated that dissolved arsenic concentrations ranging from 0-38 yg 1-' had no
demonstrable effect on the copepod, even over the long term. Average daily
reproduction rates, measured as the number of nauplii produced per female, for an 8-
day period were similar (F = 0.66, p>0.5) at 25.5, 27.3, 24.6 and 26.6, respectively,
for populations maintained in 0,13, 30 and 38 yg H (Figure 3). Grazing rates were
also similar over the four arsenic concentrations. Copepod feeding resulted in
average in vivo fluorescence decreases of 4.30-5.05 relative fluorescence units
copepod-1 h1 over 10 days, indicating similar feeding rates for control and arsenic-
exposed populations (F = 1.56, p>0.25) (Figure 4). Carbon incorporation rates
calculated from experiments with 14C-labeled phytoplankton were also identical
(F = 3.11, p>0.1) and averaged 0.018-0.026 ml of culture cleared copepod-1 h-l
(Figure 5).
Data were also collected for survivorship and development in the three
generations. The results are qualitative due to funding limitations for collecting,
preserving and counting additional samples. Over an 11-d period, 86%, 77%, 81%
and 74% of initial female densities in the Fo generation were still viable in the four
arsenic levels (0,15, 35 and 50 yg H).
24
-------
Table 2. Dissolved arsenic concentration in zooplankton experiments.
Values shown are mean ± S.E., in pg H. All arsenic present was
in the form of arsenate.
rwo Added As, Actual As,
Organism H ± gE
Eurytemora affinis
0
0.15 ± 0.02
15
13.4 ± 0.69
35
29.7 ± 3.03
50
38.4 ± 4.73
Acartia tonsa
0
1.01 ± 0.37
15
15.6 ± 1.09
35
30.8 ± 1.92
50
42.8 ± 3.70
25
-------
Figure 3. Nauplii production by female Eurytemora affinis, in nauplii-cH, at
various arsenate concentrations. Values shown are mean ± SE.
30-.
O
25
jr
-O 20H
o ~
Z> O 15 H
Q
o
(Z
Q_
Cl
3
O
10-
5-
0
\
5
10 15 20 25 30
ARSENATE (fig ¦ I"1)
-r~
35
40
26
-------
Figure 4.
Grazing by Eurytemora affinis, measured as change (A) in relative
fluorescence units copepod1 h-1, at various arsenic concentrations.
Values shown are mean ± SE.
5-
< 1-
0 i i i i i i i 1 1
0 5 10 15 20 25 30 35 40
ARSENATE (fig • I"')
27
-------
Figure 5. Grazing rate of Eurytemora affinis measured with 14Q-Iabelled
phytoplankton (ml clearedcopepod-1!!-1) at various arsenic
concentrations. Values shown are means ± SE.
0.03-1
O
2
NJ
<
cr
o
i
"O
o
a
-------
In contrast to other results, the distributions of sexually immature CIV (fourth
copepodite stage) copepodites and mature adults in Fi and F2 generations indicate
that sublethal arsenic levels have detrimental effects on population success for
copepods. The relative abundances of copepods reaching maturity were higher in
control cultures versus cultures containing 50 pg H of arsenate; in the F1
generation, 92% and 49% of individuals reached maturity in control and arsenic
solutions, respectively (Table 3). In the F2 populations, mature individuals reached
72% and 41% of total numbers in the two solutions (Table 3). The female:male ratio
fluctuated widely in cultures dosed with arsenate; in contrast, both generations of E.
affinis grown under control conditions exhibited a very stable ratio of 1.5-1.7 (F:M).
In both treatments, the number of egg-bearing females declined with each successive
generation, albeit more rapidly in arsenate-treated cultures.
Other qualitative information also supports the concept of long-term arsenic
exposure reducing population success in Eurytemora. Daily nauplii production was
estimated for 22 additional days beyond the data presented above and data
generated from these collections suggest that egg production was significantly lower
in high arsenic solutions than in control or low arsenic levels. Unfortunately, these
results are equivocal because of copepod contamination of the phytoplankton
cultures (T1. weisfloggii or Isochrysis) used as food in the study. Assuming random
transfer of contaminating copepods and offspring from the food into the experimental
containers, significantly lower egg production was observed in the highest arsenic
solutions over the study period (F = 8.77, p < 0.05).
Acartia tonsa--Acartia were maintained at the same estimated arsenic
concentrations as were Eurytemora. Actual arsenic concentrations were quite
similar to predicted (Table 2). As noted in Eurytemora, there was no detectable effect
of increasing arsenic levels on nauplii production in Acartia tonsa over a 10-d period.
29
-------
Table 3. Number (percent) of juvenile copepods that reach maturity and
breakdown by sex in successive generations o(Eurytemora affinis exposed
to 0 (control) and 50 pg H added arsenic.
Control 50 pg.1-1
Generation Male(%) Female (%) Immature (%) Male (%) Female (%) Immature (%)
F, 31(36%) 48(56%) 7(8%) 53(39%) 14(10%) 70(51%)
F2 63(27%) 106(45%) 67(28%) 19( 9%) 69(32%) 126(59%)
30
-------
Similar nauplii production rates (F = 0.24, p = 0.87) were obtained for the four
arsenic concentrations, with 17.8,19.0,18.1 and 20.5 nauplii produced female*1 d1,
respectively (Figure 6).
Grazing rates for Acartia ranged from 1.19-1.53 relative fluorescence units
copepod1 h-i with highest rates observed in cultures containing 50 yg l-l of arsenate;
however, there were no significant differences between the rates. Rates were 1.33,
1.19, 1.26 and 1.53 relative units copepod ' h-' suggesting grazing rate fluctuations
independent of arsenate level (Figure 7).
BENTHIC ORGANISMS
The results of the benthic invertebrate experiments were fairly consistent.
Although a diversity of taxa and trophic groups were examined, different life-history
stages used, and a variety of measurements made, only barnacles seem to have been
affected significantly by the presence of arsenate in the water.
Each experiment employed a standard nested design. Individual organisms or
groups of organisms (i.e. panels) were nested or grouped within experimental
chambers. The experimental chambers were then grouped by treatment. This
design allowed us to identify any potential differences between chambers within
treatments that may have contributed to observed differences. Because of this
standard design, all experiments were analyzed using the same analysis model:
nested analysis of variance with the observed variable nested within chamber and
chamber nested within arsenic treatment.
Barnacles— Barnacles were assigned to estimated arsenic concentrations of 0
(control), 12, and 42 yg-H. Measured arsenic concentrations were higher;
concentrations averaged 23.8 and 55.8 yg-H in the two arsenic treatments (Table 4).
The barnacles on each panel were considered to be a small discrete population and
31
-------
Figure 6. Nauplii production by Acartia tonsa females, as nauplii d-l, at various
arsenate concentrations. Values shown are mean ± SE.
20-
o" 15-
"O
10-
5-
0-
10 15 20 25 30 35
ARSENATE (fig • I"')
40 45
32
-------
Figure 7. Algal ingestion by Acartia tonsa measured as change (A) in relative
fluorescence units copepod-l hr-1, at various arsenic concentrations.
Values shown are means ± SE.
2.0-i
I
_c
1.5-
O
z
N
<
cn
o
T>
o
a
aj
CL
o
o
1.0-
0.5-
0.0-
0
5
10 15 20 25 30
ARSENATE (jug * I*')
35
40
i
45
33
-------
Table 4. Dissolved arsenic concentration in benthic experiments. Values
shown are mean ± S.E., in pg l1. All arsenic present was in the
form of arsenate.
Organism
Added As,
pgl-i
Actual As,
pgM ± SE
Barnacles
Bryozoans
Anemones
Oysters (adult)
Oysters (juvenile)
0 (source)
0 (control tanks)
12
42
0 (source)
0 (control tanks)
13
36
0 (source)
0 (control tanks)
13
36
0 (source)
0 (control tanks)
12
35
0
25
0.27 ± 0.05
0.38 ± 0.04
23.8 ± 2.99
55.8 ± 11.66
0.70
0.66 ± 0.21
7.06 ± 0.51
28.1 ± 4.95
1.07 ± 0.11
1.03 ± 0.07
12.9 ± 1.29
35.5 ± 2.06
0.44 ± 0.11
0.51 ± 0.05
8.97 ± 0.89
31.4 ± 2.23
0.17 ± 0.09
23.1 ± 3.28
34
-------
the growth of these populations was measured as a change in wet weight. Panel
weights were measured after 2 weeks and at the end of the experiment. The change
in weight of the barnacle populations was compared over both time periods. The
original weight of each panel was also included in the analysis model as a covariate
to correct for initial differences among panels in the sizes or numbers of barnacles.
Table 5 shows the results of these analyses. After 2 weeks neither the high nor
the low arsenic treatments had any measurable effect on the barnacles. However, by
the end of the experiment there was a small but significant reduction in the growth
rates in both arsenic treatments relative to growth in the control chambers (Figure
8). Both analyses also indicate that there were no significant differences between
chambers within treatments, indicating that conditions not controlled (e.g. chamber
location) had no observable effect.
Growth rates of barnacles in the experimental treatments were also compared
to those kept in constantly flowing raw seawater (Table 5). It is clear that the
reduction in feeding time associated with the experiment had a much greater effect
on barnacle growth rates than the presence of arsenic in the water. However, the
significant reduction in growth in arsenic treatments during the second half of the
experiment (which included 10 d of no arsenic exposure and continuous feeding)
indicated a residual effect of arsenic exposure on growth.
Bryozoan.s-Bryozoans were maintained in estimated arsenic concentrations of
0 (control, 13, and 36 pg W. Measured concentrations were somewhat lower,
averaging 7.06 and 28.1 pg-H (Table 4). As in the barnacle experiment, population
growth was used to estimate the effect of arsenic on Victorella colonies. However,
growth was measured as the change in area occupied rather than a change in weight.
When observed after 1 week, it was clear that Victorella was growing onto
clean panel surfaces but generally at a slow rate. However, when observed and
35
-------
Table 5. Nested analysis of variance (ANOVA) of barnacle growth data. Data
were analyzed after two weeks and at the end of the experiment. A
posteriori tests used Duncan's Multiple Range test. Means are mean
panel weight in grams. The second a posteriori test for the two-week data
was based on a similar analysis that included the data for panels kept in a
continuous flow of raw water. *MS for Chamber used as the error term.
Balanus - Growth after 2 Weeks
Source SS DF F
As Treatment
21.2952
2
4.65*
0.0646
Chamber in Treatment
13.7249
6
0.62
0.7138
Weight in Chamber
137.4066
9
4.14
0.0001
Error
597.3182
162
A POSTERIORI TESTS:
Main Experiment
Low Control High
2.63 2.36 1.86
Main Experiment Compared to Continuous Flow Raw Water
Raw Low Control High
11.62 2.63 2.36 1.86
Balanus - Growth through End of Experiment
Source SS DF F
As Treatment
41.6794
2
11.25*
0.0095
Chamber in Treatment
11.0492
6
0.32
0.9256
Weight in Chamber
95.5604
9
1.83
0.0766
Error
416.8746
72
A POSTERIORI TEST:
Main Experiment
Control Low High
6.01 5.69 5.66
36
-------
Figure 8. Growth rates, expressed as % of control growth, of Balanus, Diadumene,
and Crassostrea exposed to various concentrations of arsenate. Open
symbols indicate growth rate of populations maintained in flowing,
unfiltered seawater.
500
100-
50-
10-
5-
0
10 20 30 40
ARSENATE (/xg • I"1)
Bornocle 4- wk
Juv. Oyster
Anemone
Adult Oyster
Barnacle 2 wk
"" 1 r_
50
37
-------
photographed after 2 weeks, populations had declined markedly in all tanks.
Detailed observation of randomly selected panels did not reveal anything abnormal.
Colonies continued to feed when placed in the feeding chambers, but abundances
declined greatly and by the end of the experiment the area covered on panels was 10 -
25% of that observed at the start of the experiment. Because these reductions
occurred in all chambers and all treatments, no significant effect was found between
treatments.
The reductions appear to have resulted from predation by species such as the
polychaete worm, Nereis succinea. These worms were able to colonize panels when
they were exposed to unfiltered water in the feeding chambers. With little other food
available within the experimental chamber, bryozoans became their principal food.
Therefore, if arsenic had an effect on bryozoan growth it was not detectable in the
presence of such extreme predation pressure.
Sea Anemones-Anemones were maintained in similar arsenic concentrations.
Arsenic averaged 12.9 and 35.5 pg H in the two treatments, essentially as predicted
(Table 4). Unlike the previous 2 experiments, individual anemones rather than
populations were measured. The two growth measures, basal area and wet weight
are compared in Table 6. As in the barnacle experiment, no differences were found
between experimental chambers within treatments. Also it is clear that significant
differences in Diaclumene growth were not found among the three treatments.
Although the anemones exhibited rapid and significant growth, their growth rate
was not measurably affected by the presence of dissolved arsenic (Figure 8).
Adult Oysters-- Arsenic concentrations in the treatments averajed 8.97 and
31.4 pg l"1 (Table 4). In terms of shell weight, no significant differences were found
among the three arsenic treatments (Figure 8; Table 7). Also, as in the previous
38
-------
Table 6. Nested analysis of variance (ANOVA) of sea anemone growth data. Data
were collected at the end of the experiment. A posteriori tests used
Duncan's Multiple Range test. Means are for individual anemones and
are in square cm or grams. *MS for Chamber used as the error term.
Diadumene - Basal Area
Source SS DF F p
As Treatment 0.0736 2 0.48* 0.6387
Chamber in Treatment 0.4568 6 0.79 0.5824
Error 4.6278 48
A POSTERIORI TEST:
Low Control High
0.71 0.71 0.64
Diadumene - Weight
Source SS DF F p
As Treatment 0.0367 2 0.27* 0.7744
Chamber in Treatment 0.4128 6 1.72 0.1373
Error 1.9231 48
A POSTERIORI TEST:
High Low Control
~0M (L35 0.32
39
-------
Table 7. Nested analysis of variance (ANOVA) of oyster growth data. Data were
analyzed at the end of the experiment. A posteriori tests used Duncan's
Multiple Range test. Means are mean weight in grams. The second a
posteriori test was based on a similar analysis that included the data for
oysters kept in a continuous flow of raw water. *MS for Chamber used as
the error term.
Adult Oysters
Source SS DF F p
As Treatment
0.3765
2
1.57*
0.2985
Chamber in Treatment
0.7173
6
1.38
0.2255
Weight in Chamber
0.6803
9
0.87
0.5510
Error
14.0323
162
A POSTERIORI TESTS:
Main Experiment
Control High Low
0.51 olo oil
Main Experiment Compared to Continuous Flow Raw Water
Raw Control High Low
1.54 0.51 0.50 0.31
40
-------
experiments no effects attributable to environmental chamber were found. In
general the oysters added 0.3 - 0.6 g of shell during the experiment. Increases in
shell mass were not influenced by dissolved arsenic. As might be expected growth
rates were proportional to the size of the oyster. However, because individual
oysters were of similar size, no significant differences in weight within chambers
were found (Table 7).
Because some oysters were maintained in flowing, unfiltered seawater during
the experiment, the effect of the experimental reductions in feeding times were also
measured. As can be seen in Table 7, oyster growth was significantly reduced in the
experimental treatments and was approximately 30% of that observed for those
individuals exposed continuously to food.
Larval Ov.sters--When panels in the larval oyster chambers were examined
after 10 days, no successful recruits were found. Many empty larval shells were
found adhering to the panel surface, but these individuals appeared to have died as
larvae and fell to the bottom of the chamber. Water samples indicated that no larvae
remained in the water column.
The cause for the complete mortality of larvae in all treatments has not been
determined. However, a calculation error resulted in much lower quantities of algal
food being used than planned in the experimental design. The amounts available
may have been insufficient for growth and development.
Juvenile Oysters—Measured arsenic levels in this experiment were very close to
predicted (Table 4). At the beginning of the experiment over 10,000 newly settled
oysters were found on the experimental panels. Of these more than 2,500 increased
in size over the course of the experiment and were judged to have successfully
recruited prior to the study. As in the previous experiments the growth of these
41
-------
juveniles was not affected by arsenic (Table 8). The analysis did demonstrate that
the juvenile oysters were quite variable in their growth rates and that within
chambers growth rates were significantly different among panels. However, there
were no differences among chambers within treatments in mean growth rate and no
differences among treatments.
Given the large number of juvenile oysters measured, the results of this
experiment are unambiguous. Even though we might expect these much smaller
and younger life stages to be less tolerant of stress than adults, the results were the
same as observed in the adult experiment.
ARSENIC INCORPORATION BY BENTHIC ORGANISMS
Arsenic concentrations within tissues was measured at the completion of
experiments with barnacles, anemones, and adult oysters. Barnacles exhibited
significantly higher (p<0.01) body burdens of arsenic that were proportional to the
concentration of arsenic in the treatment (Table 9; Figure 9). Barnacles exposed to
23.8 pg I*1 contained approximately twice the amount of arsenic, on a weight basis,
as controls; barnacles exposed to 55.8 pg 1-1 contained over 3 times the arsenic.
Other benthic organisms did not incorporate significant quantities of arsenic.
Tissue levels remained relatively constant, regardless of treatment arsenic
concentration (Table 9; Figure 9).
42
-------
Table 8. Nested analysis of variance (ANOVA) of juvenile oyster growth data.
Data were analyzed at the end of the experiment. A posteriori test used
Duncan's Multiple Range test. Means are mean area covered by an oyster
in square mm. *MS for Chamber used as the error term. **MS for Panel
used as the error term.
Juvenile Oysters
Source §S DF F
As Treatment
Chamber in Treatment
Panel in Chamber
Error
0.3305 1
177.5131 4
2070.8024 23
41059.7918 2603
0.01* 0.9354
0.49** 0.6984
5.71 0.0001
A POSTERIORI TEST:
Arsenic Control
4.23 3.46
43
-------
Table 9. Arsenic incorporation by benthic organisms. Values are in pgg-1
dry weight. Balanus values include shell and tissue
concentrations, others include only tissue concentrations.
s added As incorporation,
pg-l-l pg.g-i ± SE
Balanus sp. 0 0.98 ± 0.09
23.8 1.89 ± 0.24
55.8 3.33 ± 0.52
Anemones 0 9.07 ± 0.38
12.9 8.93 ± 0.18
35.5 9.84 ± 0.23
Oysters 0 6.42 ± 0.38
9.0 5.24 ± 0.14
31.4 6.97 ± 0.34
44
-------
Figure 9. Arsenic content, in pgg"1, of benthic organisms maintained in various
arsenic concentrations. Values shown are means ± SE. • = Barnacles,
Q = anemones, A = adult oysters.
en
8-
cr>
=L
O
U
h-
<
Cd
O 4-
Q_
cr
o
(J
z ?-
(/)
<
0 5 10 15 20 25 30 35 40 45 50 55
As ADDED (/iq • I-')
45
-------
DISCUSSION
The response of estuarine organisms to arsenate was largely as hypothesized.
Phytoplankton, particularly centric diatoms, exhibited large changes in growth rate
of dominant species upon exposure to arsenic, leading to changes in the species
composition of the assemblage. In each phytoplankton experiment, different centric
diatoms were sensitive to arsenate, Cerataulina pelagica during the late spring, and
Chaetoceros debile during the summer. In the spring experiment, Thalassiosira
sp.was able to fluorish in the absence of C. pelagica\ it was not present in large
numbers in the summer experiment. Similar changes have been observed in earlier
experiments with arsenate and Chesapeake Bay phytoplankton (Sanders and Cibik,
1985; Sanders, 1986). In every experiment in which Tkalassiosira sp. was a
dominant, its growth rate in arsenic-treated assemblages exceeded its growth in
controls. This increase in arsenic treatments is probably not triggered by the arsenic
itself; rather, this species likely is responding to increased availability of light or
nutrients caused by the decline of another centric diatom, in this case, Cerataulina
pelagica. Note that Thalas'siusira sp. did not bloom until C. pelagica had essentially
disappeared from arsenic-treated tanks (Figures IB, 1C).
The flagellate-dominated assemblage in the summer experiment was not
greatly affected by arsenate. Only the diatom, Chaetocerus debile, exhibited
significant growth inhibition (Figure 2B). Therefore, arsenate impacts to
phytoplankton communities in temperate ecosystems are likely to be seasonal,
causing alteration to spring and fall diatom blooms and perhaps having little effect
during summer months.
Earlier work had indicated that zooplankton are quite resistant to arsenic
(Biesinger and Christensen, 1972; Passino and Novak, 1984); initial studies with
Eurytemora affinis had yielded similar results (Sanders, 1986). Therefore, we
46
-------
designed a series of experiments to study the impact of arsenate to all life stages of
the copepods, E. affinis and Acartia tonsa, and to follow the success of the population
through several generations. Our long term studies largely upheld earlier results;
there was no effect of arsenate on algal incorporation by either species, nor did
arsenic impair production of nauplii. Thus, relatively high arsenic concentrations
appear to be necessary to alter crustacean metabolic activities, as suggested by
earlier studies. There was, however, an apparent impact of chronic arsenic exposure
on the ability of nauplii in the F i and F2 generations to mature to adult; in each
generation, far fewer nauplii matured to adulthood when exposed to arsenic (Table
3). In addition, female to male ratios varied considerably in arsenate-dosed
generations while ratios in control populations remained constant. These results
may be important to long-term success of the population if copepod populations are
exposed to chronic elevation of arsenic concentrations. Residence times of effluents
as well as dilution within the system become increasingly important factors in the
assessment of a chronic effect. However, the importance of lower population success
over two to three generations, as observed here, may be masked by the relatively
large potential for indirect effects from the ingestion and assimilation of arsenic-rich
phytoplankton species or the food web effects resulting from arsenic-induced shifts in
phytoplankton species or sizes (e.g., Sanders, 1986).
Benthic organisms also exhibited general tolerance to arsenic. With the
exception of Balanus, none of the organisms tested showed reduced growth rates in
response to arsenic treatments. With Balanus, there was a small, nonsignificant
reduction in growth rate over the first two weeks of the experiment in response to the
highest arsenic treatment, approximately 56 pg 1-1. By the end of the experiment,
both of the arsenic-treated populations had significantly reduced growth, but the
reduction was small, approximately 6% (Table 5). The importance of such a small
decrease to the overall population is unknown.
47
-------
Balanus was also the only organism to incorporate significant concentrations of
arsenic during the experiments (Figure 9, Table 9), doubling and tripling its arsenic
content at the two levels of arsenic treatment. However, because of its small size,
Balanus was also the only organism in which shell material was tested for arsenic
content along with muscle tissue; thus, the increased arsenic content could be caused
simply by adsorption of arsenate to the carbonate shell matrix. Studies of arsenic
flux through Baltic Sea microcosms have indicated that shells of organisms (Lymnea
peregra, Mytilus eduhs, Cardium sp.) exhibited high uptake of arsenic (1.7-4.6 times
control) when exposed to 7.5 pg H arsenate (Rosemarin etal., 1985). Therefore,
further analyses need to be performed to determine whether barnacles actually
incorporated arsenic into body tissues.
A similar study of arsenic incorporation by Crassostrea virgimca also
demonstrated that chronic exposure to low arsenic concentrations did not lead to
significant incorporation of arsenic (Zaroogian and HofTman, 1982). In addition,
studies of estuarine organisms (Lymnea peregra, Gammarus oceanicus, Idotea
baltica) have indicated low to insignificant uptake from exposure to elevated arsenic
concentrations in water (Notini and Rosemarin, 1986).
Other benthic organisms exhibited little response to arsenate. Although
experiments with larval oysters were not successful, experiments with newly-settled
juveniles demonstrated that even these fragile organisms were not affected by the
arsenic concentrations presented. In addition, none of the other tested species
incorporated significant quantities of arsenic during the 4-week test periods, further
evidence that arsenate does not cause direct harm to trophic levels above
phytoplankton.
The concentrations of arsenate used in these experiments were designed to
assess direct potential impact at the highest possible concentrations that are
environmentally realistic. Arsenic concentrations within the open ocean average
48
-------
1.0-1.5 pg l"1 (Andreae, 1978; Waslenchuk, 1978; Sanders, 1980). Within estuaries,
the arsenic concentration is much more variable; however, the usual range is
between 0.1-5 pg H (Andreae, 1978; Martin and Whitfield, 1983). Even in impacted
estuaries, arsenic concentrations rarely rise above 10 pg H. The highest
concentration that we have measured in the Chesapeake Bay was in the vicinity of
an abandoned fly ash dump in the Nanticoke River; concentrations ranged between
4.15 - 60.6 pg l"1 in a localized area around the dump (Sanders, unpublished data).
To our knowledge, the concentrations used in this study cover the range of possible
concentrations in natural systems, even those receiving considerable impact from
man. Therefore, the lack of direct response to arsenic seen in these experiments,
except with the phytoplankton community, can be considered to be relevant to
arsenic impacts to estuarine and coastal marine systems.
However, as outlined in the Introduction, there are other pathways for impact
of a toxic substance within an aquatic food web. Arsenic, because of its large impact
upon phytoplankton species composition and community structure, is a prime
candidate for such indirect impacts. We hypothesized that direct effects of arsenic
would be limited to phytoplankton and that direct impacts to other trophic levels
would be minor relative to potential indirect effects associated changes in trophic
structure or the ingestion of arsenic through food. Our experiments have shown that
changes in dominant species can drastically alter an herbivore's ability to procure
enough food to successfully reproduce (Sanders, 1986). In addition, although arsenic
dissolved in the water may be unavailable and nontoxic to higher trophic levels,
arsenic incorporated in their food may be quite toxic.
For example, a shift to smaller phytoplankton taxa could lead to reductions in
preferred phytoplankton food items and favor prduction of smaller microzooplankton
species. Enhancement of the microbial loop theoretically reduces carbon transfer to
highest trophic levels, thereby reducing production in fish and shellfish stocks.
49
-------
Transfer of arsenic incorporated in resistant phytoplankton to planktonic and
benthic suspension feeders could also prove more of an immediate danger for the food
web than exposures to dilute dissolved arsenic levels. Ingestion and subsequent
assimilation of arsenic in phytoplankton might conceivably result in accumulation
to levels that might alter normal cellular metabolism in metazoans, potentially
reducing viability, diversion of energy to reproduction and/or offspring or slower
growth.
Another potential indirect effect is the chemical transformation of arsenate to
arsenite and methylarsonate after algal uptake. This transformation, which readily
occurs in productive ecosystems (Sanders, 1985,1986), effectively increases the
potential for direct arsenic toxicity, for the transformed arsenic species are far more
toxic to organisms than arsenate (Nissen and Benson, 1982).
Such studies of the indirect impacts of arsenic and other toxics largely remain
to be done. Although this research project originally included these studies, funding
reductions precluded their completion. We are currently proceeding with a limited
set of experiments through funding from another branch of EPA (the Chesapeake
Bay Program); if those funds continue, we will be able to complete the entire project.
50
-------
ACKNOWLEDGEMENTS
A number of people have contributed to this work. S. Cibik, J. Bianchi, M.
Olson, and L. Currence contributed greatly to the design and implementation of
experiments. G. Riedel was instrumental in the coordination and modification of
analysis techniques to meet the needs of this project. R. Batiuk critiqued the draft of
this report. We thank them for their contributions, suggestions, and cooperation. A
portion of this work was funded through the Environmental Protection Agency's
Chesapeake Bay Program, grant # X-003312.
51
-------
REFERENCES
Andreae, M. O. 1978. Distribution and speciation of arsenic in natural waters and
some marine algae. Deep-Sea Res. 25:391-402.
Andrews, J. D. 1961. Measurement of shell growth in oysters by weighing in water.
Proc. Nat. Shellfisheries Assoc. 52:1-11.
ANS (Academy of Natural Sciences, Philadelphia). 1981. Zooplankton, pp. V-H.l to
V-H.117. In Acad, Nat. Sci. Phila., ed. Vol. ID: Chalk Point 316
demonstration of thermal entrainment and impingement impacts on the
Patuxent River in accordance with COMAR 08.05.04.13. Prepared for the
Potomac Electric Power Company by The Academy of Natural Sciences. Acad.
Nat. Sci. Phila. Report No. 81-5.
Bartram, W. C. 1980. Experimental development of a model for the feeding of
neritic copepods on phytoplankton. J. Plankton Res. 3:25-51.
Biesinger, K. E. and G. M. Christensen. 1972. Effects of various metals on survival,
growth, reproduction, and metabolism of Daphnia magna. J. Fish. Res. Bd.
Canada 29:1691-1700.
Boyle, E. A. and S. Huestedt. 1983. Aspects of the surface distributions of copper,
nickel, cadmium, and lead in the North Atlantic and North Pacific, p. 379-394.
In: C. S. Wong, E. Boyle, K. W. Bruland, J. D. Burton, and E. D. Goldberg
(eds.). Trace metals in sea water. Plenum Press, New York.
Braman, R. S., D. L. Johnson, C. C. Foreback, J. M. Ammons and J. L. Bricker. 1977.
Separation and determination of nanogram amounts of inorganic arsenic and
methylarsenic compounds. Anal. Chem. 49:621-625.
52
-------
Brownlee, D. C. and F. Jacobs. 1987. Mesozooplankton and microzooplankton in the
Chesapeake Bay. In: S. K. Majumdar, L. W. Hall, Jr., and H. M. Austin (eds.),
Contaminant problems and management of living Chesapeake Bay resources.
Pa. Academy Sci., Philadelphia, in press.
Dayton, P. K. 1975. Experimental evaluation of ecological dominance in a rocky
intertidal algal community. Ecol. Monog. 45:137-159.
Heinle, D. R. 1966. Production of a calanoid copepod, Acartia tonsa, in the Patuxent
River estuary. Chesapeake Sci. 7:59-74.
Heinle, D. R. 1969. Temperature and zooplankton. Chesapeake Sci. 10:186-209.
Heinle, D. R. and D. A. Flemer. 1975. Carbon requirements of a population of the
estuarine copepod Eurytemora affinis. Mar. Biol. 31:235-247.
Hendrickson, P., K. G. Sellner, B. Rojas de Mendiola, N. Ochoa and R. Zimmermann.
1982. The composition of particulate organic matter and biomass in the
Peruvian upwelling region during ICANE 1977 (Nov. 14-Dec. 2). J. Plankton
Res. 4:163-186.
Keith, L. H. and W. A. Telliard. 1979. Priority pollutants. I-A perspective view.
Environ. Sci. Technol. 13:416-423.
Landry, M. R. 1977. A review of important concepts in the trophic organization of
pelagic ecosystems. Helgol. wiss. Meeres. 30:8-17.
Mackenzie, F. T., R. J. Lantzy, and V. Paterson. 1979. Global trace cycles and
predictions. Math. Geol. 11:99-144.
Margalef, R. 1951. Diversidad de especies en les communidades naturales. Publnes.
Inst. Biol. apl. Barcelona 9:5-27.
Martin, J-M. and M. Whitfield. 1983. The significance of the river input of chemical
elements to the ocean, p. 265-296. In: C. S. Wong, E. Boyle, K. W. Bruland, J.
D. Burton, and E. D. Goldberg (eds.), Trace metals in sea water. Plenum Press,
NY.
53
-------
Nissen, P. and A. A. Benson. 1982. Arsenic metabolism in freshwater and
terrestrial plants. Physiol. Plant. 54:446-450.
Nival, P. and S. Nival. 1976. Particle retention efficiencies of an herbivorous
copepod, Acartia clausi (adult and copepodite stages): Effects on grazing.
Limnol. Oceanogr. 21:24-28.
Notini, M. and A. Rosemarin. 1986. Fate and effects of arsenic in a mesoscale model
ecosystem. Status Report prepared for the ESTHER Project, Swedish
Environmental Research Group, Karlskrona, Sweden, January.
Osman, R. W. 1977. The establishment and development of a marine epifaunal
community. Ecol. Monog. 47:37-63.
Osman, R. W. 1982. Artificial substrates as ecological islands, pp. 71-114. In: J.
Cairns (ed.), Artificial substrates. Ann Arbor Science Publ., Massachusetts.
Paine, R. T. 1969. A note on trophic complexity and community stability. Am. Nat.
100:65-75.
Parsons, T. R. 1976. The structure of life in the sea, pp. 81-97. In: D. H. Cushing
and J. J. Walsh (eds.), The ecology of the seas. W. B. Saunders Co.,
Philadelphia.
Passino, D. R. M. and A. J. Novak. 1984. Toxicity of arsenate and DDT to the
cladoceran Busmina longirustris. Bull. Environ. Contam. Toxicol. 33:325-329.
Peoples, S. A. 1975. Review of arsenical pesticides, p. 1-12. In: E. A. Woolson (ed.),
Arsenical pesticides. ACS Symp. Ser. 7, American Chemical Society,
Washington, D.C.
Pielou, E. C. 1966. The measurement of diversity in different types of biological
collections. J. Theor. Biol. 13:131-144.
Rosemarin, A., M. Notini, and K. Holmgren. 1985. The fate of arsenic in the Baltic
Sea Fucus vesiculosus ecosystem. Ambio 14:342-345.
54
-------
Ryther, J. H. 1969. Photosynthesis and fish production in the sea. Science 166:72-
76.
Sanders, J. G. 1980. Arsenic cycling in marine systems. Mar. Environ. Res. 3:257-
266.
Sanders, J. G. 1983. Role of marine phytoplankton in determining the chemical
speciation and biogeochemical cycling of arsenic. Can. J. Fish. Aquat. Sci.
40:192-196.
Sanders, J. G. 1985. Arsenic geochemistry in Chesapeake Bay: dependence upon
anthropogenic inputs and phytoplankton species composition. Mar. Chem.
17:329-340.
Sanders, J. G. 1986. Direct and indirect effects of arsenic on the survival and
fecundity of estuarine zooplankton. Can. J. Fish. Aquat. Sci. 43:694-699.
Sanders, J. G. and S. J. Cibik. 1985. Adaptive behavior of euryhaline phytoplankton
communities to arsenic stress. Mar. Ecol. Prog. Ser. 22:199-205.
Sanders, J. G. and P. S. Vermersch. 1982. Response of marine phytoplankton to low
levels of arsenate. J. Plankton Res. 4:881-893.
Sanders, J. G. and H. L. Windom. 1980. The uptake and reduction of arsenic species
by marine algae. Estuarine Coastal Mar. Sci. 10:555-567.
Sanders, J. G., J. H. Ryther, and J. H. Batchelder. 1981. Effects of copper, chlorine,
and thermal addition on the species composition of marine phytoplankton. J.
Exp. Mar. Biol. Ecol. 49:81-102.
Sellner, K. G. 1983. Plankton productivity and biomass in a tributary of the upper
Chesapeake Bay. I. Importance of size-fractionated phytoplankton
productivity, biomass and species composition in carbon export. Est. Coastal
Shelf Sci., 17:197-206.
Sellner, K. G. 1987. Phytoplankton in Chesapeake Bay: role in carbon, oxygen and
55
-------
nutrient dynamics. In: S. K. Majumdar, L. W. Hall, Jr., and H. M. Austin
(eds.), Contaminant problems and management of living Chesapeake Bay
resources. Pa. Academy Sci., Philadelphia, in press.
Sellner, K. G. and R. J. Horwitz. 1982. Plankton interactions in the Patuxent River
estuary: Field studies of community composition and density, with a
deterministic model of the effects of zooplankton grazing on phytoplankton
carbon, production of fecal matter, sediment oxygen demand and nutrient
regeneration. Acad. Nat. Sci., Philadelphia, PA. Rept. No. 82-14D. 111pp.
Sellner, K. G. and M. M. Olson. 1985. Copepod grazing on red tides of Chesapeake
Bay, pp 245-250. In: D. M. Anderson, A. W. White and D. G. Baden (eds.),
Toxic dinoflagellates. Elsevier, NY.
Uthe, J. F., H. C. Freeman, J. R. Johnston, and P. Michalik. 1974. Comparison of
wet ashing and dry ashing for the determination of arsenic in marine
organisms, using methylated arsenicals for standards. J. Assoc. Off. Anal.
Chem. 57:1363-1365.
Van Valkenburg, S. D., J. K. Jones, and D. R. Heinle. 1978. A comparison by size
and volume of detritus versus phytoplankton in Chesapeake Bay. Est. Coastal
Mar. Sci. 6:569-582.
Waslenchuk, D. G. 1978. The budget and geochemistry of arsenic in a continental
shelf environment. Mar. Chem. 7:39-52.
Zaroogian, G. E. and G. L. Hoffman. 1982. Arsenic uptake and loss in the American
oyster, Crassotrea virginica. Environ. Monit. Assess. 1:345-358.
56
------- |