84

       Maryland Univ.,
       P**pated  for
U.S. Environmental Protection Ageiujt
I.VW.Q III Information Resource
Cinier (3PM52)
£11 Chednut Street
H-udelpfcia, PA  1910?
•fin
     EPA Report Collection
     information Resource.Center
     US EPA Region 3
     Philadelphia, PA  19107

-------
i ;
* '                                                                         EPA-600/3-84-015
                                                                           January 1984
                                      SUBMERGED AQUATIC VEGETAT-WN  IN
                                       UPPER CHESAU-EAKE BAY:   STUDIES
                                       RELATED TO POSSIBLE  CAUSES  OF
                                      THE  RECENT  DECLINE I« ABUNDANCE3
                            r .         —  W-. MichaeV Kemp,b
                                            Walter R.  Boynton,0
                             :,->             J,  Court  Stevenson,b
 I           .                                  Jay C.  Means,^   ,
,!  "                                       Robert  R. Tw1lley,b  and
                                            Thomas-Wv-0oaes,b
•-                                         _._..   Editors
ij"
        :*-,',                 ' .
              .  r  .    ...    •••   -    .'c-
            ,,aUn1Vers1ty of Maryland Center  for Environmental
               and EstuaMne Studies  Contribution No«  1431

              bHorn Point Environmental  Laboratories
               Cambridge, Maryland
              cChesapeak« Bloldgltal >Laboratory
               Solomons,  Maryland   20688 '  ^
                                    Grant  Nos.  R8059320IO and X003248010
                                              Project Officer
                                              Pr.  David Flemer
                                    U.S.  Environmental  Protection Agency
                                              2083 West Street
                                           Annapolis, MD   21401
                                                IT
                                           NATIONAL TECHNICAL
                                          INFORMATION SERVICE
                                              1.1 OIPMTWNI OF COiCEdCl
                                                »«*«nio, n. mil

-------
                                            TECHNICAL REPORT DATA
                                     (neatt rtati Instructions on the rei/ene before completing)
          1. R6POTT NO.
            EPA-600/3-84-015
                                                                    3. RECIP
          14. TITLE AND SUBTITLE    „,...,.     _.       ,
           Submerged Aquatic  Vegetation in Upper Chesapeake
           Bay:  Studies  Related  to Possible Causes of the
           Recent Decline in  Abundance.
                                                         5. REPORT DATE
                                                           January 1984
                                                         8. PERFORMING ORGANIZATION CODE
          . AUTHOR(S)              „ •   • •
           W. M. Kemp, W.  R.  Boynton, J. C. Stevenson,
           J. C. Means,  R.  R.  Twilley and T, W. Jones
                                                         I. PERFORMING ORGANIZATION REPORT NO.
t
          9. PERFORMING ORGANIZATION NAME AND ADDRESS
           Center  for  Environmental  and Estuarine Studies
           University  of  Maryland
           Cambridge,  MD    21613
                                                         10. PROGRAM ELEMENT NO.
                                                                    G RANT NO.
                                                                    and
                                                                     X003248010
          12. SPONSORING AGENCY NAME AND ADDRESS
           Chesapeake  Bay Program
           U.S. EPA
           Central Regional  Laboratory
           Annapolis.  MD   21401	
                                                         13. TYPE OF REPORT AND PERIOD COVERED
                                                          Technical	
                                                         14. SPONSORING AGENCY CODE
                                                          EPA/600/05
          15. SUPPLEMENTARY NOTES
          To. ABSTRACT
               This paper  provides a synthesis of research conducted on possible causes of
          the decline in abundance of submerged aquatic  vegetation (SAV) in upper Chesapeake
          Bay beginning  in the late 1960's.  Three factors potentially were emphasized in
          this study:  runoff of agricultural) herbicides; erosional  inputs of fine-grain
          sediments;  nutrient enrichment/and associated  algal  growth.  Widespread use of
          herbicides  in  the estuarine watershed occurred contemporaneous with the SAV
          loss; however, extensive sampling of estuarine water and sediments during 1980-81
          revealed that  typical  bay concentrations of herbicides  (primarily atrazine)
          rarely exceeded  2ppb.   However, normal concentrations (< 5 ppb) were shown experi-
          mently to have little  measurable effect .on plants.   Historical Increases in
          turbidity have been documented for some bay tributaries  since the 1940's.  Light
          (PAR) attenuation by suspended fine-grain sediments  contributed more to total
          turbidity in bay shallows (< 1.5m) than did phytoplankton  chlo'ophyll  a.  Evidence
          Indicated that plant photosynthesis was I1ght-lImited for  much of the ?ay.
          Effects of the continual increase in nutrient  enrichment of the bay (documented
          since 1930) were tested by experimentally fertilizing pond roesocosms at levels
          common to the  upper estuary.  Moderate to high nutrient  loadings resulted in
          significant  increases  in growth of epiphytic and planktonic algae and decreases
          in ^iV prrtHiirfr ir\n
          17.
                                         KEY WORDS AND DOCUMENT ANALYSIS
                           DESCRIPTORS
          18 DISTRIBUTION STATEMENT
Release to  Public
                                                       b.lDENTIFIERS/OPEN ENDED TERMS
                                              U.S. EPA Region 111
                                              Regional Center for Envi
                                                Information           •
                                              1880 Arch Street (3PM52)
                                              Philadelphia, PA 19103 r
                                                       19. SECUniTY CLASS (ThisRepn:!/
so: sEciMTV CLASS n
 Unclassified
                                                                                 c.  COSATI Field/Croup
                                                                                  onmental
                                                                                  2). NO. OF PAGES
                                                                      22. PRICE
          EPA P*m 2220.1 (••«. 4-77)   PREVIOUS KOITION i« OMOLBTC

-------
fl
                                                NOTICE

                          This document has been reviewed in accordance with
                          U.S. Environmental Protection Agency policy and
                          approved for publication.  Mention of trade names
                          or commercial products does not constitute endorse-
                          ment or recommendation for use.
                                                 ii

-------
                                                          19103
                                 CONTENTS

                                                                    Page


SUMMARY AND CONCLUSIONS ............................................. vi

CHAPTER I.  Concentrations of the Herbicides Atrazlne and Linuron in
            Agricultural  Drainage, Tributary Estuaries and Littoral
            Zones of Upper Chesapeake Bay

     Acknowledgements........... ....... ......... .................... I-ii
     Introduction.
     Methods.
     Results and Discussion.
     References	
     Figures	
     Tables	
     Appendices	
-1
 2
-6
-15
-18
-29
-34
CHAPTER II.  Temporal Responses of a Submerged Vascular Plant and
             its Associated Autotrophic Community to Atrazine
             Stress in Estuarlne Microcosms

     Acknowl edgements	  Il-i i
     Abstract	  II-l
     Int roduct 1 on	  11-2
     Materials and Methods	  II-3
     Results and Discussion	  11-6
     References	  11-22

CHAPTER III.   Effects of Atrazine and Linuron on Photosynthesis
              and Growth of Potamogeton perfoliatus and Myriophyllum
              splcatum 1n Estuarlne Microcosms

     Introducti on	  111-1
     Materials and Methods	  III-2
     Results  and Discussion	  111-9
     Implications:  Herbicides and Estuarlne Vascular Plants	  111-29
     References	  111-37
     Appendi ces	  II1-41

CHAPTER IV.  Atrazine Uptake, Phytotoxlcity, Release, and Short-
             Term Recovry for the Submerged Aquatic Plant,
             Potamogeton perfoliatus

     Abstract	 IV-1
     Introduction	 IV-2
     Materials and Methods	 IV-3
     Results  and Discussion	 IV-5
                             111

-------
                                                                     Page

     Conclusions	 IV-15
     References	 IV-16
CHAPTER V.  Uptake and Phytotoxicity of So1l-Sorbed Atrazlne for
            the Submerged Aquatic Plant, Potamogeton perfoliatus

     Abstract	  V-l
     Introductl on	  V-2
     Materials and Methods	  V-3
     Results and Discussion	  V-4
     References	  V-10

CHAPTER VI.  Degradation of Atrazine in Estuarine Water/Sediment
             Systems and Selected Soils

     Acknowledgements	  VI -11
     Abstract	  VI-1
     Introduction	  VI-2
     Materials and Methods	  VI-3
     Resul ts and Di scussion	  V1-6
     References	  VI-19

CHAPTER VII.  Uptake and Photosynthetic Inhibition of Atrazlne
              and its Degradation Products on Four Species of
              Submerged Vascular Plants

     Acknowledgements.	  VII-11
     Abstract	  VII-1
     Introduction	  VI1-2
     Materials and Methods	  VII-3
     Results and Discussion	  VI1-4
     References.	  VI1-14

CHAPTER VIII.  Effects of Nutrient Enrichment in Experimental Ponds
               Containing Submerged Vascular Plant Communities

     Table of Contents	  VIII-11
     List of Figures	  VIII-lv
     Introduction	  VIII-1
     Materials and Methods	  VIII-2
     Results	  VI11-8
     Discussion	  VIII-30
     References	  VII1-54
     Appendices	  VII1-61
                              iv

-------
                                                                         Page

CHAPTER  IX.  The  Decline of Submerged Vascular  Plants  1n Upper
             Chesapeake Bay:  Summary of Results  Concerning
             Possible Causes

     Acknowl edgements	     IX-ii
     Abstract	     IX-1
     Int roductl on	     IX-2
     Approach to  Problen	     IX-4
     Herbicide  Inputs, Fate and Phytotox1c1ty	     IX-4
     Suspendable  Sediments and SAV Light Responses	     IX-7
     Nutrient Enrichment and Algal-SAV Relations	     IX-11
     Synthesis  of Findings and Implications	     IX-14
     References	     IX-19

                                           Regional Center lor hnuronniniral Information

                                                  US f-P\ Region Ml

                                                   165(1 AK h Si

                                                 Philadelphia P4 1<)I»3
         . EPA Region III
      Regional Center for Environmental
       Information
      1650 Arch Street (3PM52)
      Philadelphia, PA 19103

-------
                            SUMMARY AND CONCLUSIONS


     This report presents results of research conducted at University of Mary-
land's Center for Environmental and Estuarine Studies concerning factors poten-
tially involved in the decline of submerged aquatic vegetation (SAV) in upper
Chesapeake Bay.  This research examines three main factors in relation to SAV
growth and production:  agricultural  herbicides; suspended sediments and associ-
ated light attenuation; nutrient enrichment and resulting algal  growth and
light attenuation.  A few ongoing studies (e.g., Goldsborough 1983,  Staver
1983) are not included here; however, in this discussion some reference is
made to their findings which will be available soon.   The results of other
relevant studies are summarized elsewhere (Boynton et al. 1983;  Wetzel  et al.
1983).

     Considerable effort was expended to investigate the potential  importance
of herbicides in contributing to the overall  stress of the estuary's SAV popula-
tions (Chapters I-VII).  This research emphasized two specific compounds.  The
first of these, atrazine, which is closely associated with corn  crops,  has
been the most widely used herbicide in the region, while the second  compound,
linuron, is commonly employed in weed control  for soybeans.  Concentrations of
these two herbicides were monitored in water and sediments throughout the
upper bay over the period 1980-81 (Chapter I).  A hierarchically designed
stratified sampling scheme revealed typical aqueous concentrations of both
compounds to be about 0-3 ppb in the main bay, 0-5 ppb in a major eastern
shore tributary, and 0-40 ppb in a creek connecting a small estuarlne cove to
surrounding agricultural fields.  Concentrations In the creek and small  cove
were measured at l-4h intervals before, during and after all runoff  events,
and values above 5 ppb never persisted for more than 6-8 h.  Atrazine concen-
trations associated with suspended or deposited sediments were less  than 5 ppb
for > 95% of samples and never exceeded 20 ppb.

     Initial studies indicated a wide range of physiological and morphological
responses of one common SAV species, Potamocjeton perfoliatus, in response to
herbicide treatment, including photosynthetic depression, stem elongation,
reduction 1n stem weight per unit length, and increased chlorophyll  a per unit
leaf area (Chapter II).  Several of these effects are analogous  to observed
adaptations of this and related species to reduced light intensity (e.g.,
Goldsborough 1983).

     At atrazine or linuorn concentrations between 5-100 ppb significant photo-
synthetic inhibition was observed for both _P. perfoliatus and Myriophyllum
spicatum in microcosms, followed by strong recovery (toward untreated control
plants) within 1-3 wk, eventhough herbicide levels remained within 5-10% of
initial values throughout (Chapters II and HI).  Plant biomass  decreased
significantly after 5 wk of treatment at herbicide concentrations >  50 ppb for
                                      vi

-------
P. perfoil atus and > 500 ppb for H. spciatum (Chapter III).  Overall, the
effects of the two herbicides were statistically Identical, while some differ-
ences between plant specie* were observed (M. spicatum being more tolerant).
Estimates of I^Q (herbicide concentration at which 1% loss of photosynthesis
(Pa) is predicted) were 2-4 ppb for P_. perfqliatus and 8-11 ppb for M.
spicatum, and values of 150 (concentration for 50% loss of Pa) range"? from
45-55 ppb and 80-117 ppb respectively.  Similar phytotoxicities were observed
for Zannichellia palustris and Ruppla maritlma (Chapter VII).

     Rapid uptake of ^C-labelled atrazine was demonstrated for P. perfoil atus,
with equilibrium between internal and external concentrations beTng achieved
within about 1 h (Chapter IV).  A direct relation between atrazine uptake and
photosynthetic depression was observed for this plant; however, disproportion-
ately high apparent uptake at low external herb'dde concentrations suggests a
two-step uptake process with simple sorptlon (without inhibition of photosyn-
thesis) dominating at low concentrations.  Root uptake of atrazine appears  to
be of little Importance for these plants.  Initial photosynthetic recovery of
atrazine-treated plants was effected by release of sorbed herbicide within 2 h
after rinsing in atrazine-free water.  Some (~ 5%) loss of photosynthesis
was evident after 3 d of wash; however, this difference was not statistically
significant.  Short-term (2 h) experimental  exposures to atrazine revealed
reductions in £. perfoliatus photosynthesis which were similar to those observed
over 2-6 wk 1n~~microco3ms, with values of 159 being about 80 ppb.

     Atrazine is readily sorbed to soil and sediment particles, with a parti-
tion coefficient (sorbediaqueous) greater than 1.0 (Chapter V).  However, the
potential importance for plant uptake of atrazine sorbed to overlying sediments
(resting on SAV leaves) seems to be remote.   Experiments with ll>C-label1ed
atrazine showed negligible plant uptake of herbicide sorbed to soils concentra-
tions of at about 120 ppb.  In addition, the presence of epiphytic sediments
significantly retarded leaf uptake of aqueous atrazine, although such sediments
themselves Inhibited photosynthesis presumably by attenuation of light and
reduction of CO? utpake.

     The degradation of 1<4C-1abelled atrazine was observed under simulated
field conditions for upper and middle bay sediment-water systems and for two
common agricultural soils In the Maryland coastal plain.  The distributions of
atrazine and two categories of metabolites or degradation products (hydroxy-
atrazine and dealkylated atrazine) were followed over an 80 d period.  The
half-life (time for 50% degradation to metabolites) for atrazine was markedly
shorter for estuarine systems (15-20 d) than for soils (330-385 d).  The accum-
ulation of hydroxyatrazine in experimental estuarine water and sediments raised
questions concerning the potential phytotoxicity of these compounds.  Bioassay
experiments were performed with 4 species of SAV to examine uptake and photo-
synthetic depression for uC-labelled atrazine and 3 metabolites (Chapter
VII).  Overall, the Inhibitory effect of the metabolite, hydroxyatrazine, on
plant photosynthesis was negligible compared to that for atrazine, with no
significant inhibition even at 1500 ppb.  Some significant loss of Pa was
observed for deethylated atrazine at 500 ppb; however, this metabolite has  a
short half-life in the estuary, being similar to that for atrazine.

     The effects of nutrient enrichment on algal (plankton and epiphytes)
                                        vii

-------
growth and SAV production and abundance were Investigated by fertilizing 8
(duplicates at 4 levels) experimental ponds (500m2) during June-August
1981 (Chapter VIII).  These ponds, which were seeded with sediment, water and
plants from the Choptank River estuary, were maintained In batch-mode for
sequential periods of 7-10 d punctuated by complete exchange of water followed
by retreatment prior to the next batch period.  Maximum fertilization rates
were typical of nutrient loading In areas of upper Chesapeake Bay receiving
direct agricultural runoff.  Nutrient concentrations 1n treated ponds *ere
reduced rapidly to control levels within 1-3 d,  and plant tissue nutrient
contents were directly related to treatment.  Initial  growth of the two domin-
ant SAV species (P. perfollatus and £. maritime) was enhanced 1n fertilized
ponds; however, pTani abundance 1n August was inversely related to treatment,
with SAV virtually eliminated at the highest dosage.

     Planktonic and epiphytic algal biomass (as  chlorophyll a) increased sig-
nificantly with treatment.  Light (PAR) attenuation by mlcroTlgae was suffi-
cient to account for the reduction in SAV production and abundance in August.
Epiphytic growth accounted for most of the light reduction although attenuation
in the water column was also necessary to reduce PAR below plant compensation
levels.  Direct measurements of epiphyte effects on both PAR attenuation (by
leaf scrapings 1n petri dishes) and plant photosynthesis (with llfC-labelled
bicarbonate) confirmed this relationship.  Preliminary evidence suggests that
a shortening of SAV growing season, as observed  here 1n response to fertiliza-
tion, may ultimately lead to decimation of these plant populations by dis-
rupting plant reproduction.  Light attenuation by microalgae and suspendable
sediments may affect the normal balance between  SAV production and respiration
leading to premature-flowering and/or insufficient translocatlon to underground
propagates,I both of which would reduce the viability of regrowth 1n the fol-
lowing spring.  It 1s concluded that further research  is needed to understand
the reproductive capacities and strategies for these plants.

     The results of 2 recent studies continuing  this line of research are not
Included in this report but will be available soon.  A second year (1982) of
fertilization in the experimental ponds provided a more detailed examination
of the nutrient-algal-SAV relationships (Staver  1983).  Problems encountered
In the batch-mode approach in 1981 were alleviated with a continuous flow
system and more frequent treatment.  In this 1982 study only 4 ponds were
used, and SAV communities in these were essentially mono-specific stands of _P.
perfollatus, thus eliminating the complicating problems of differential epi-
phytic colonization on 2 SAV species.  The general patterns observed 1n 1981
were more pronounced and less equivocal in the 1982 research.  Detailed studies
of the responses and adaptations of £. perfpliatus to (high 100X, medium 34%,
low 6X) light were also done 1n 1982 (Goldsborough 1983).  Numerous morpho-
logical and physiological changes in this plant  were observed in response to
reduced (moderate and low) light, Including stem elongation, Increased pigmen-
tation, increased specific leaf area, as well as Increased Initial slope for
photosynthesis versus irradlance relations.  Most of these adjustments appear
to confer adaptive advantage on shaded plants; however, after 2 wk of exposure
to low light significant reductions in stem density, flowering and underground
reproductive propagules were observed.

     A summary and synthesis of the research Included 1n this report 1s pro-



                                      viii

-------
vided 1n the final chapter (IX).  Here, we consider the relative contributions
of herbicide runoff, sediment loading <>nd nutrient enrichment to the environ-
mental stress experienced by SAV in upper Chesapeake Bay.  Combining these
research findings in a simple conceptual  framework as well  as a numerical
simulation model suggested that the relative importance of effects on SAV
associated with these 3 inputs is as follows, nutrients > sediments » herbi-
cides.  A narrative scenario is developed to interpret spatial and temporal
aspects of the SAV decline in light of this research, and potential  resource
management strategies are discussed.
                                       ix

-------
                                                          UMCEES 63-63 CBL
                                   CHAPTER I
                       CONCENTRATIONS OF THE  HERBICIDES
                     ATRAZINE AND IINURON IN AGRICULTURAL
                  DRAINAGE,  TRIBUTARY ESTUARIES AND LITTORAL
                        ZONES OF UPPER CHESAPEAKE  BAY*
                                  April,  1983
Contribution No.  1431   Center for Environmental  and EstuaMne Studies,
 University of Maryland.

tHorn Point Environmental Laboratories, Box 775, Cambridge,  MD  21613.

^Chesapeake Biological Laboratory, Box 38, Solomons,  MD  20688.
                                      1-1
                                 W.R.  Boyntontf                                               I
                        J.C.  Means# and J.C.  Stevensont                                      '
                                     with                                                    i
                                  W.M. Kemp-;                                                  1
                                  R.  Twllleyt
                                                                                             J

-------
                               ACKNOWLEDGEMENTS

     Many individuals contributed to the successful completion of this survey
and to these people we express our thanks.  In particular we acknowledge the
enthusiastic assistance of M. Meteyer, N. Kaumeyer and K. Kaumeyer in
conducting our field work and laboratory analyses and K. Staver for help in
sampling of agricultural drainage during storm events.  Dr.  Randy Sperry
provided statistical advice with support from Md. Dept. Natural Resources,
Power Plant Siting Program.  The R.V.'s Orion and Venus were operated by
Capts. W.C. Keefe and R. Younger, respectively.  This research was supported
by a grant from EPA's Chesapeake Bay Program (Grant No. CR 805932-01-02 Task
7).
                                      I-ii

-------
                                 INTRODUCTION

     Although herbicides have provided effective weed control  resulting in
higher crop yields and are a necessity for modern no-till  agriculture
(Phillips et al. 1980), they may also have deleterious environmental  effects
upon non-target species (Galston 1979).  Recently there has been concern that
herbicide runoff from agricultural  fields may have played  a role in the
decline of submerged aquatic vegetation (SAV) in Chesapeake Bay (Stevenson and
Confer 1978).  Early changes in SAV populations coincided  with increased
herbicide usage in the Chesapeake watershed in the 1960's  (Bayley et al.
1978).  During that period the principle herbicide was atrazine
(2-chloro-4-ethylamino-6-isopropyl-amino-s-triazine) which was applied to
large acreages of corn.  Later, other herbicides such as linuron
[3-(3,4-dich1orophenyl)-l-methoxy-l-methylurea], were also used widely because
of a shift to soybeans which were planted in rotation with corn every other
year.

     Previous reports of the presence of atrazine residues in aquatic systems
receiving runoff from agricultural  areas showed concentrations could  reach 20
pg JT1 (Waldron 1974, Richard et al. 1975, Miur et al. 1978, Schepers et al.
1978, Frank et al. 1979, Roberts et al. 1970).   However, these studies were
all done in freshwater systems which differ in many ways from estuaries such
as Chesapeake Bay.  The possibility of high herbicide levels in Chesapeake Bay
waters was first suggested by Correll et al. (1978) who reported dissolved
atrazine and linuron concentrations in the range of 1-10 ppb in several upper
bay tributaries and up to the ppm range in sediments.  However subsequent
measurements by Austin (see Stevenson and Confer, 1978) and Newby et  al.
(1978) in surface waters of the open bay showed concentrations rf atrazine
only as high as 1.0 pg jr* and concentrations declined rapidly in a
non-conservative manner to undetectable levels (<0.1 vg JT*) in a seaward
direction.  Similarly, Zahnow and Riggleman (1980) reported that
concentrations of the pheuylurea herbicide linuron were non-detectable (<.01
ug 1~1) in either the water column or sediments in several  tributary  rivers of
the bay, including the Choptank and Rhode River estuaries.  The variability
between these studies may have resulted for a number of reasons, including:
the proximity of sampling times to run-off events; the persistance of
different herbicides in estuarine waters; the location of  sampling stations
relative to herbicide source area (littoral zone vs. open  bay waters); and
differences in herbicide loadings due to annual variations in precipitation
and resultant runoff.

     In view of these uncertainties, and the limited amount of herbicide data
available, we conducted a three-phase investigation of selected herbicide
(atrazine and linuron) concentrations in the upper Chesapeake Bay (Maryland
portion of bay).  The first phase of this investigation documented herbicide
concentrations in water and sediments at 23 littoral zone  sites throughout the
                                       1-1

-------
upper bay where SAV communities were: (1) still present; (2) now absent but
had been present in the recent past or; (3) had not been present for at least
two decades.  The second phase involved monthly measurements of herbicide
concentrations along the longitudinal axis of a major upper-bay tributary
(Choptank River estuary) receiving drainage from an agriculturally dominated
watershed.  The third phase involved detailed monitoring of herbicide
discharge from an experimental agricultural watershed where application rates
were known.   Our objectives were to develop a preliminary description of the
time-space characteristics of herbicide concentrations in the upper bay and to
relate these field observations to in situ distributions of SAV and to
experimentally derived relationships between SAV growth and herbicide dose.
                                    METHODS

                                Sampling Design
     Herbicide concentrations in water (dissolved) and sediments were
monitored at twenty-three shallow water (0.5 - 1.0 m depth) sites in the upper
Chesapeake Bay (Fig. 1; Table 1) on 4 occasions (23-28 April; 11-16 June;
24-29 July; 15-22 September in 1980).  Sampling areas were chosen along the
open bay shoreline throughout the upper bay, in tributary rivers and in
hydrodynamically quiescent coves to provide a representative sampling of SAV
habitats.  In each habitat area several stations were identified and included:
(1) those having SAV communities (Status I as indicated by the 1978 aerial
surveys of Anderson and Macomber, 1979); (2) sites at which SAV had been
present prior to 1969 (Stotts, pers. comm) but had not been present within  the
last decade (Status III); and locations which had not supported SAV
communities at least since 1959-61 (Status IV).  During the present study
there were 3 sites where vegetation had been present in 1978 but failed to
appear in 1980; these stations were designed as being Status II.

     Atrazi ne and linuron were also measured monthly from April through
September at 4 to 7 locations along the longitudinal  axis of the Choptank
River estuary, a major eastern shore tributary of Chesapeake Bay (Fig.  2).
Sampling sites were selected based on salinity (S) conditions in the estuary
during each cruise such that there was equal increment in surface salinity
among stations from tidal-freshwater (5=1 ppt) to the bay (S=9-17 ppt).  In
addition, one station was routinely sampled at a site mid-way from the main
channel and littoral zone station SI-14 (Todds Cove).  A single water sample
was collected at 0.5 m at each station for dissolved residue analyses.   In
addition one sediment grab was also collected at each station and a sub-sample
of surficial  material (top 5 cm) was analyzed for residues.  On an occasional
basis, 10-20* of near-surface water was filtered and the particulate fraction
analyzed for residues.

     A controlled agricultural watershed (Fig. 3) was monitored to estimate
herbicide delivery to the Choptank River estuary.   Runoff and tidal influences
were monitored using aim width Pars hell flume (Lomax et al. 1979).
                                       1-2

-------
tT
              Downstream from the flume, four additional stations were also sampled, the
              last being adjacent to the Choptank River.  Runoff events were monitored prior
              to application of agricultural fertilizers and herbicides, immediately
              following application and during the summer period after the emergence of
              field crops.  Herbicide runoff was also monitored on a weekly basis during
              1981 at the U.S. Geological Survey gauging station at Beaverdam (located on
              Kings Creek which enters the Choptank River 55.5 km upstream of the mouth)
              which drains a larger watershed.

                   All water samples for herbicide analyses were collected in pre-cleaned
              and solvent washed 2 liter glass jars with aluminum foil-lined caps.
              Immediately after sampling, a 1 liter aliquot of each sample was filtered
              through precombusted Whatman GFC glass fiber filters (1.2 \im) and pumped
              through a pr
-------
     The ShP-PAK containing water column residues were eluted with 5 m* of DIG
methanol  and the eluent was evaporated under a stream of nitrogen and taken up
in 1 m* of DIG methanol containing a known quantity of an internal standard
(diphenylamine) for a quantitative analysis.

     Atrazine residues were quantified using a Hewlett-Packard 5840 gas
chromatograph equipped with a 2 mm ID x ? M all-glass column packed with 10
percent Carbowax 20 M on 80/100 Supelcoport.  The column was maintained at a
temperature of 160 C and the injector and detector were held at 225 C and 300
C, respectively.  An alkali-flame nitrogen-phosphorus specific detector was
used to measure the amount of atrazine and internal standard in a 4 ji£ sample.
Quantification was achieved by comparing the atrazine/internal standard ratio
(AT/IS) to a calibration curve of AT/IS vs. atrazine concentration.

     Linuron residues were analyzed using a Waters Associates 6000A liquid
chromatograph equipped with a \i Bondapak-Phenyl  (Trade name) column, a model
440 fixed wavelength ultraviolet detector (254 pm), a Perkin-Elmer M-2
electronic integrator and a Linear 159 chart recorder.  Twenty microliter
sanples were introduced onto the column using a Rheodyne 7125 injection valve
and eluted using an isocratic solvent system of 35 percent acetonitrile/water.
Areas under the sample peaks were integrated and quantified by comparison with
a calibration curve.

             Quality Assurance Program for Analysis of Herbicides

     As an integral part of the analytical program for this study, a quality
assurance (QA) program was developed which included both frequent internal
checks and periodic external checks on the precision and accuracy of the
analytical methods employed for trace analyses of herbicides in water,
suspended particulates and sediments of Chesapeake Bay.

     The internal program for quality assurance started In the field and
included the establishment of sampling protocols and storage procedures which
ensured that herbicide residues present in the samples at the time of sapling
would not degrade or volatilize from the samples.  During the first year of
this study storage experiments were performed both internally and as part of
an external  QA audit to assure that the integrity of the samples was
preserved.  These studies indicated that no detectable losses of sample at
environmentally realistic levels (1-2 ppb) occurred under the conditions used.
The details of the sampling and storage techniques are presented in the
methods section of this report.

     With the large number of samples being collected and analysed, record
keeping was recognized as an important component of our QA program.  The
following steps were taken to insure continuity of sample Identification and
data reporting.

     1.   All samples were labeled at sampling with the date, site
          of origin and sample type.   These data were also entered
          1n a continuous logbook at the time of sampling.  Field
          work sheets documenting the sampling data and site
                                       1-4

-------
-
 .
                         information were also filled out at each station.

                    2.   A second book contained all of the written laboratory
                         methods employed for sampling, storage, extraction,
                         concentration and analysis.

                    3.   All weights and measures used during the preparation of
                         the samples for analyses were recorded in a laboratory
                         record book along with the appropriate sample
                         identification information.

                    4.   All analytical data relating to the quantitation  of the
                         herbicide residue were kept on the chromatograms.

                    This last step was the most convenient for us since: 1)  the integration
               data and quantifications were printed on the chromatograms  in one case or were
               printed on a tape and stapled to the corresponding chromatogram and 2) this
               allowed us to keep the standard chromatograms and sample chromatograms together
               for future reference.  Each working standard curve used for quantitation was
               also retained in the file.  Our methods were set up such that a minimum of
               calculation was necessary.  Peak areas or area/area ratios  (internal
               standard/sample) were divided by the slope of the working calibration curve and
               corrected for concentration factors or extraction weights.   Quantitation data
               summaries for logical groups of samples (i.e. 1979 field HgO samples; 1980
               linuron f^O samples; etc.) were then prepared.
                    At the analysis stage, the Internal  QA program was centered upon
               maintaining a consistantly high level  of analytical results.   The program was
               designed to insure:  1) quantitative recovery of herbicide residues  from water
               and sediment samples, and 2) accurate, reproducible and precise determination
               of the concentrations of herbicide residues in water,  suspended particulates
               and sediment extracts.  The program included the following steps:

                    1.   Preparat"'"!! of fresh standards from 99.5+% pure analytical
                         grade compounds and intercomparisons of old  standards with
                         new standards every two weeks.

                    2.   Daily three-point checks of working standard curves  and
                         recalibration of instruments before sample analysis  and
                         preparation of new standard curves when instrument
                         parameters were changed (i.e., new column, new detector
                         bead, etc.).

                    3.   Method blanks (water, solvent, SEP-Pack) were run with
                         each batch of samples.   All  blanks were below our
                         detection limit.

                    4.   Blind samples with known residues were periodically
                         inserted in the sample  analysis schedule.  No sample
                         varied more than 10$ from the known amount.
                                                      1-5

-------
                                   •*-*—
                                   \
     b.   Une In 10-12 samples was analyzed In duplicate.  No
          variations of more than 10% were observed.

     6.   All samples were injected in duplicate.  If results varied
          by more than 5%, the samples were reanalyzed.

     7.   Samples which were analyzed were reanalyzed with a known
          spike.  Recoveries of 90-100% were considered acceptable.

     8.   Mass spectral confirmation using a three ion/selected ion
          monitoring technique was employed on field samples with
          significant residues (>0.6 ppb) to confirm the identity of
          the herbicide residue and to cross-check the quantitation.

     In addition to the internal program for QA the laboratory was evaluated
three times by external QA auditors.  In each case, blinded samples were
received, spiked in duplicate into estuarine water, and analyzed by the normal
laboratory protocols.  In each case, the results of our analyses were within
the acceptable limits of accuracy, reproducibility and precision.  (Appendix
Tables 10 and 11).

     The following methodological parameters were determined in these external
audits.

                     ATRAZINE                     LINURON
Accuracy*                +_ 15
Reproducibility (1 ppb}  + 3.0%                      ±2.9%
Precision                + 0.016                     +0.006
Detection Limit          0.05 ppb (water)            (J.05 ppb (water)
                         0.1 ppb (sediment)          0.1 ppb (sediment)
* includes extraction efficiency

     As stated earlier, the first external QA audit Included a check on sample
storage procedures.  No significant losses were observed using our storage
protocols.
                            RESULTS AND DISCUSSION

                    Upper Bay Littoral Zone Investigations

     Measurements of dissolved atrazine and linuron concentrations,  as well  as
SAV biomass measurements, associated with the littoral zone investigations  are
listed in Appendix Table I.  During the April, June, July and September
cruises, atrazine concentrations ranged from
£0.05-0.3 ppb, £0.05-1.14 ppb, £0.05-0.4 ppb and £0.05-0.93 ppb,
resprectively.  Maximum dissolved atrazine concentrations were consistently
less than 2 ppb and the majority of samples yielded concentrations below the
detection limit (0.05 ppb).  During the April  and June sampling periods
                                       1-6

-------
linuron residues were below the detection limit (0.05 ppb) at all 23 stations
while in July and September concentrations ranged from ^0.05-1.3 ppb and
^0.05-0.75 ppb, respectively.  The lack of detectable linuron residues in
April and June was attributed to the fact that large quantities of this
herbicide are usually not applied until July when soybeans are planted.

     To provide a broad view of dissolved herbicide distribution in the upper
bay we have summarized our data by averaging concentration measurements from
similar geographical areas during each of the 4 sampling periods (Table 2).
Atrazine residues above detectable concentrations were observed earlier in the
year and within more geographical locations in the bay than was the case for
linuron.  For example, dissolved atrazine X).l ppb occurred at 32-36% of our
stations from April-July while such concentrations of linuron were exceeded at
only 13-20% of stations in July and September.  Finally, for both herbicides,
averaged concentrations were generally higher in either the upper bay or tidal-
fresh tributary areas than in other geographical groupings, although some
exceptions occurred (e.g. mid arid lower bay in April).

     While averaged herbicide concentration data is useful in providing a
general view of conditions in the upper bay, the time-space occurrence of
maximum concentrations is also important (Table 3) particularly since it
appears that short-term (1 hr) exposure of seagrasses to atrazine can depress
photosynthetic rates (Jones et al., 1983).  In April maximum dissolved
atrazine concentrations were relatively low in all  areas (jCO.05-0.30 ppb) but
tended to be somewhat higher in the mid and lower bay groups (Si's 1 and 24).
However, maximum concentrations in June, July and September tended to occur in
the upper bay and/or tidal fresh tributaries, paralleling the trend in the
averaged data presented in Table 2.  For reasons that are not clear at this
time, stations in the vicinity of the Patuxent River estuary (Si's 1, 2 and
27) consistently had ths highest concentration in the mid-bay region and in
the tidal freshwater category on 2 of 4 occasions (SI-27).  Except that
detectable concentrations of dissolved linuron were not observed in the April
and June surveys, the spatial pattern of maximum concentrations closely
paralleled that observed for atrazine.

     Sediment concentrations of atrazine and linuron were measured at each of
the 23 littoral zone sites during each survey.  Detectable concentrations are
given in Table 4 and the results of all analyses are contained in Appendix 2.
Detectable concentrations in surficial sediments were generally lower and much
less frequently observed than in the water column.  Detectable levels of
atrazine (>0.1 pg kg"*) were found in bottom sediments only 5 times during the
entire survey (total sample number = 82) and none exceeded 1.0 ug kg-1.
Atrazine residues were most frequently encountered in June and all  occurred in
the upper bay (Si's 5 and 7) or tidal-fresh tributaries (SI-18).  Linuron
concentrations in sediments exceeded limits of detectability at only 1 site
(SI-23; 0.51 pg kg"*) near the town of Crisfield in September.  It appears
that little of either atrazine or linuron remains associated with sediments in
the estuary.  Cunninghan (1980) showed that over 90% of atrazine adsorbed to
sediment is released when placed in estuarine water.  In addition, Jones et
al.  (1982) reported that atrazine in the water column degrades rapidly with a
half-life of 1-2 weeks, suggesting that estuarine concentrations may be

-------
ephenierel.  Linuron has an even faster degredation rate in estuarine waters
(Means et al., 1983) which may account for its even lower frequency of
occurrance in the study area.

                   Choptank River Longitudinal Survey. 1980

     To obtain a more detailed characterization of herbicide distributions in
one area of the bay, monthly samples were taken along the longitudinal axis
(channel stations) of the Choptank River estuary, t .e watershed of which is
agriculturally dominated.  Station locations were given in Figure 2, and all
measurements of dissolved atrazine and linuron concentrations are listed in
Appendix Table 3 and summarized in Figure 4.  The general pattern that emerged
for both compounds was one in which concentrations were near or below
deto^Mon levels in the early spring (April and May)  and late summer
(September) and in the range of £0.05-1.0 ppb  luring  the months of June, July
and August.  There was some indication that highest residue concentrations
occu-red in the tidal-fresh region as might be expected since the ratio of
river volume to drainage area is minimal and hence the potential for dilution
is limited.  However, on several occasions elevated concentrations were found
in other regions of the estuary suggesting that inputs from adjacent
agricultural fields might oe of sufficient magnitude  to elevate residue
concentrations in the mainstan channel areas.  In any case, dissolved
concentrations were generally below 1.0 ppb throughout the study area and
never exceeded 2.0 ppb.  Of the 38 stations sampled,  53% exhibited residue
concentration of atrazine below detection levels and  only 26% had
concentrations X10 ppb.  Only 16% of the stations sampled showed significant
concentrations of lir.uron.

     Consistent with the results of the upper bay littoral zone survey,
sediment concentrations of atrazine and linuron were  even lower and detectable
concentrations (X). 1 ug kgr^) were rarely encountered (Appendix Table 4).  In
fact, of the 38 stations sampled only 2 had detectable concentrations of
atrazine or linuron.  While considerably fewer samples were analyzed for
herbicide residues associated with the suspended particulate fraction, we did
not detect significant concentrations (M).l pg k(fl)  in any of samples
analyzed (Appendix Table 5).
                          Agricultural Runoff Studies

     In 1980 relatively low spring precipitation caused very low superficial
water tables in the Choptank River region where our intensive runoff studies
were centered.   The silty-clay Mattapax soils of this region have a high
moisture holding capacity and can absorb large quantities of water when dry.
As a result of the precipitation-soil conditions evidenced in 1980, there was
virtually no measurable runoff prior to July and hence it is doubtful  that
data collected during this period are representative of more normal years.  In
any case, concentrations of dissolved atrazine and linuron measured during
1980 at several locations on the Horn Point Experimental  watershed are listed
in Appendix Table 6 (see Fig. 3 for station locations).  During the monitoring
                                       1-8

-------
f>
               period, concentrations of linuron in agricultural base-flow drainage ranged
               from below our level of detection (^0.05 ppb) to 2.61 ppb in a small embayment
               (pond) adjacent to  the Choptank River.  Peak atrazine concentrations ranged
               from 8.2 ppb in the flume to 18.1 ppb immediately downstream adjacent to a
               cornfield.  Atrazine concentrations were more frequently in excess of 1.0 ppb
               earlier in the year, suggesting that even when baseflow runoff is the
               predominant means of transport, maximum concentrations still occur shortly
               after application of these compounds.  During dry conditions we also found
               that elevated herbicide concentrations were largely restricted to drainage
               ditches connected to agricultural fields.  As indicated in Appendix Table 6
               all samples taken from the Choptank adjacent to agricultural fields exhibited
               residue concentrations below detection levels, even when appreciable
               concentrations were observed in drainage creeks.  Apparently degradation rates
               for these compounds are sufficiently rapid (Jones et al. 1982) coupled with
               dilution in estuarine waters such that elevated concentrations were not
               observed in open waters of the Choptank adjacent to the experimental watershed
               i n 1980.

                    Of the 67 determinations of atrazine and linuron residues associated with
               suspended particulates in the water column, only one sample exhibited a
               significant concentration (Appendix Table 7).  Similarly, atrazine and linuron
               concentrations in soils collected from cropland and drainage chennels were
               generally below detection levels (_<0.1 ugkg'M and never exceeded 1.2 ppb
               (Appendix Table 8).  Hence 1t appears that during 1980 there was little mass
               transfer of herbicide in any form (dissolved or adsorbed) between fields and
               estuary nor was there a significant accumulation of residues in cropland or
               aquatic sediments.

                    However, in 1981 precipitation in the spring and early summer months was
               near average and runoff events at the Horn Point Experimental watershed were
               recorded up until mid-summer when a protracted dry period began.  Also, flows
               at Beaver Dam (Fig. 2) measured at the U.S. Geological Survey station on Md.
               Rt. 331 were substantial following spring planting.

                    Atrazine is usually applied at a rate of 2.2 kg ha"1 during final  disking
               and seeding operations in conventional tillage of corn used by over 90% of the
               fanners in the Choptank region (R. Wade, Agricultural Extension Agent,
               Dorchester Co., MD).  Normally, planting begins in late April on excessively
               drained sandy soils (Gallstown) in this area; followed by planting in
               moderately well drained soils (e.g. sassafrass) and finally proceeds to heavy
               clay soils (e.g. Mattapex) in May when evapotranspiration dries up excessive
               moistjre.  Therefore in a large (1500 ha) watershed such as that around
               Beaverdam Branch, there is a progression of application on various soils over
               a number of weeks, which began in late April in 1981.

                    Despite its much larger size the Beaverdam watershed behaved similarly to
               the Horn Point Experimental watershed except that concentrations were
               detectable a little earlier than at Horn Point due to earlier application.
               However, since the peak concentrations at Beaverdam was lower (9 ppb) during
               the Intensive runoff period of May 11-15 (Figs. 5a and 5c) total delivery of
               atrazine to the Choptank River was almost identical per watershed (Appendix
                                                      1-9

-------
Table 9).  During the highest runoff month (May) an estimated 4.6 kg atrazine
was transported from the watershed, while only 0.2 kg and 0.07 kg were
transported in June and July, respectively.  Therefore the total  1981 atrazine
leakage rate was in the range of Horn Point about 0.6% of the input.   This
latter loading was derived from an estimate of corn acreage in the Beaverdam
watershed from a small plane and the recommended rate ot application for this
area (2.2 kg ha'1).

     One of the reasons for this relatively low rate of atrazine leakage may
be that delivery occurred after the study period but Wauchope and Leonard's
(1980) formula suggests little leakage that late in the season.  A more likely
reason is that relatively low pH which ranged from 4.5 to 6.8 during this
period reduced atrazine mobility.  Runoff losses of atrazine have been
reported to be much lower in -inlimed soils and leaching increases with
increased pH (Gaynor and Volk, 1981).  Another important factor is that the
relatively flat eastern shore landscape may preclude as much herbicide runoff
as that found on the western shore (1%) by Wu (1980).  Therefore, herbicide
application on Maryland's eastern shore may present potential fewer
environmental problems than originally anticipated by Correl 1 et al.  (1977).

     In contrast to the extended period of herbicide application at the
Beaverdam watershed, 18 ha of corn was planted in only two days at the Horn
Point Experimental watershed beginning on May 4, 1981.  Initial  atrazine
concentration in ditches on this date was undetectable, demonstrating little
spray drift during application (Appendix Table 9).

     A week after application, the largest runoff events of 1981 resulted from
a series of thunderstorms which passed over the region (Fig. 5c).  At Horn
Point, 12.2 x 10^ nH of water passed through the flume in the agricultural
field over the period of May 11-15.  Atrazine concentrations ranged from 1.6
to a peak of 13.7 ppb in samples taken over this period.  An estimated 197 g
atrazine was lost downstream or about 0.5% of the total application (39.6 kg).
It appears that runoff throughout the basin was significant enough to increase
atrazine concentrations in the estuary proper (F1g. 5b) over those observed in
1980 (Fig. 4).

     Although there were several small precipitation events later 1n the
month, none was large enough to promote runoff.  In succeeding months, due to
declining atrazine concentrations, much less residue was delivered to the Bay.
In June an estimated 34 g was lost and 16 g in July.  The total  delivery
during this 3 month period accounted for 0.6% of the total application.  Since
»i prolonged drought period began in late July,  very little further herbicide
leakage occurred in 1981 from the Horn Point watershed.  Also, although
soybeans were planted in July very low concentrations (<2 ppb) of linuron plus
little flow of water produced unestimatable low leakage of linuron.  We
discuss the temporal coupling of runoff and herbicide application in more
detail in a later section of this report.

                          SAV-Herbicide Relationships

     We had anticipated finding SAV communities at 10 of the 23 littoral  zone
                                     1-10

-------
r"

'
        stations  saripled  in  1980 based on the results of the 1978 aerial surveys of
        Anderson  and Macomber  (1979).  However, at 3 sites (Si's 15, 20 and 23)
        aboveground bicmass  never appeared during 1980 and at all upper bay locations
        where  SAV was  present  in 1980 (Si's 4, 9 and 28), biomass disappeared prior to
        the  September  sampling period.   In addition, qualitative observations taken
        since  1980 indicate  that SAV has declined at other of our stations including
        Si's 10,  13 and 14 (Staver, pers. romm.; K?umeyer, pers cornm.).  In contrast,
        SAV  communities around Smith Island (SI-21) in the lower bay appeared robust
        in late 1982 (Kemp,  pers conm.).  It appears that some declines in
        distribution and  abundance have  occurred post-1978 and continued through 1982,
        at least  at the stations we sampled.  The results of SAV biomass sampling are
        listed in Appendix Table 1 and summarized in Table 5.  In general SAV appeared
        to be  more ephemeral in the upper bay region than in the mid and lower bay
        although  the highest mean biomass observation was obtained at a location in
        the  upper bay  (SI-28).  With few exceptions (e.g. SI-28 in June; SI-14 in
        July)  SAV standing stocks were low compared to those observed in other SAV
        studies in the bay (Kaumeyer et  a!., 1980; Wetzel et al., 1980).

            To explore our  data for possible relationships between herbicides and SAV
        abundance, dissolved atrazine and linuron concentrations measured during field
        surveys were plotted against SAV biomass estimates (Fig. 6).  The resultant
        scatter plots  suggest  little in  the way of strong patterns except that a broad
        range  in  SAV biomass was associated with herbicide concentrations below our
        level  of  detection.  An alternative approach is to consider SAV status (i.e.
        present,  early disappearance in  1980, not present since 1969 and not present
        since  1959-61) relative to observed herbicide concentrations.  Specifically,
        we calculated  the percent occurrance of atrazine above levels of detection for
        all  four  bay surveys for each SAV status group.  Results indicated that
        atrazine  concentrations were above detection levels, 45%, 50%, 42% and 66% of
        measurements in Status I, la (a  sub-class of I where SAV disappeared early in
        1980), III and IV areas, respectively.  Again, strong trends did not emerge
        although  there was some indication that detectable herbicides were more
        frequently associated  with thosa areas where SAV had not been observed since
        the  1959-61 period and in areas  where SAV disappeared early in the 1980
        growing season.   The frequency of atrazine occurrance was similar in areas
        with normal SAV in 1980 and in those areas where SAV had not been observed
        since  1969.  However,  if the above calculation is done using 0.5 ppb as a
        criteria  for significant herbicide presence, a stronger pattern appears
        wherein this criteria  is equalled or exceeded in 7%, 20%, 13% and 42% of
        observations in areas  having SAV Status I, la, III and IV, respectively.
        While  this is  more suggestive of herbicide influence the pattern may be the
        result of some other factor co-varying with herbicide presence (nutrient
        and/ >r sediment effects on light attenuation).   In the case of SAV it has been
        postulated that nutrient enrichment and light attenuation by sediments, in
        addition  to herbicides, may be stressing or limiting the distribution of these
        communities in the bay (Stevenson and Confer, 1978).  The results of other
        experiments and field  studies conducted as part  of our overall SAV program
        were syntheized into an ecosystem simulation model, the results of which
        suggested the  relative potential SAV stress of 3 factors to be
        nutr1ents>sed1ments>herbicides (Kemp et al., 1983).
                                              1-11

-------
     While these studies indicated a secondary role for herbicides as an
influence on SAV distributions in the upper Chesapeake, it is not possible to
completely discount herbicides as a factor in some areas of the bay for
several  reasons.  For example, in the field studies reported here our sampling
schedule was sufficiently discontinuous ( 6 wk intervals)  such that we could
not consistently detect short-term elevations or depressions in herbicide
concentration due to runoff events in all littoral zones of the bay and hence
our herbicide values only represent instantaneous concentrations.  However,
elevated, but apparently ephemeral, herbicide concentrations occur and may
represent a stress to SAV in some areas of the bay.  Jones et al. (1982), for
example, has shown that SAV photosynthesis was depressed even when exposure
times were short (1 hr) and recovery took 4-8 hr in herbicide-free medium.  In
the course of experimental  dose-response studies (Kemp et  al., 1983) we did
not test for the cumulative effects of repeated pulse exposures of SAV to
herbicides and thus the effects of this type of exposure remain unknown.
Additionally, the frequency of field measurements of herbicides were not
sufficiently intense to determine the maximum levels of herbicide to which
natural  SAV communities were exposed.

     An alternative procedure for evaluating possible impacts of  herbicides on
SAV is to compare in situ herbicide concentrations with the results of
herbicide dose-response experiments conducted in conjunction with this study.
A summary of the results of laboratory work reported by Kemp et al. (1983) is
presented in Fig. 7.  Herbicide concentrations causing 50% inhibition (
SAV net photosynthesis (Pa) ranged between 55-117 ppb and  45-80 ppb for
atrazine and linuron, respectively.  Concentrations causing incipient (1%)
inhibition of Pa ranged from 2-11 ppb, consistently higher than concentrations
observed in littoral zones throughout the upper bay.  Values for IJ^Q and IJQ
for SAV biomass were similar to those reported for macrophyte photosynthesis.

     As indicated earlier in this report, maximum concentrations  of atrazine
and linuron approached 50 and 20 ppb, respectively, during a strong runoff
event in May, 1981.  More frequently concentrations were in the range of 1-10
ppb and existed for only short periods of time (2-8 h) being rapidly
dissipated by dilution and degradation.  Although some species may be more or
less sensitive to herbicide stress than those we studied (e.g. Walsh et al.,
i982; Correll and Wu, 1982) our in situ field observations of herbicide
concentration coupled with laboratory dose-response work suggests that while
herbicides may stress plants in some estuarlne areas it is difficult to
conclude that these compounds were primarily responsible for the  bay-wide
decline of SAV communities.

                Annual Variability in Herbicide Concentrations

     In this section we summarize some observations concerning the potential
for significant annual variations in herbicide concentrations in the bay.
Necessarily, much of what can be said is circumstantial but may still  be
useful in judging the representativeness of the data collected in 1980 and
1981.

     In Fig. 8 we have summarized several features of freshwater flow from the
                                      1-12

-------
Susquehanna River for purposes of placing the results of the 1980 studies in
perspective relative to this potential  herbicide source.  Considerable
variation was evident among annual  average flows as well as intra-annual  flows
between 1976-1982 ranging from about 52,000 cfs in 1979 to about 29,000 cfs in
1980 (Fig. 8).  In this view 1980 was a low-flow year, as was 1981 and 1982.

     Concentrations of atrazine detected in the Susquehanna River at Conowingo
Dam (U.S.  Geological Survey, 1982)  were plotted against river flow and are
shown in Fig. 9.  With few exceptions detectable herbicide residues  wer
associated with low river flow.  Of the 14 occasions when detectable residues
were observed, 12 occurred at river discharge rates below 40,000 cfs and  the
majority occurred at flows less than 20,000 cfs.  Most of these lower flows
occurred during May-September suggesting that herbicides entering the bay were
largely derived from compounds applied to croplands in the spring or summer of
the same year.  The lack of detectable residues during the winter-spring
freshet further supports this contention.  Finally, there was little if any
pattern in atrazine concentration among the years for which data were
available, even though there were considerable differences in river  discharge
characteristics.  In general, concentrations ranged from <0.1 ppb to 1.2  ppb,
with values commonly in the range of 0.2-0.6 ppb between 1978 and 1981.

     Atrazine concentrations measured at upper bay stations (e.g. Si's 4, 5,
6, 7, 8 and 28) in 1980 were generally comparable to those observed  in
discharge over the period 1978-1981.  In fact, atrazine and linuron
concentrations along the mainstem of the bay in 1977 and 1980 were generally
comparable (Fig. 10).  From the data available, it seems reasonable  to
tentatively conclude that dissolved concentrations of atrazine in the upper
bay and along the mainstem are generally below 1 ppb and do not exhibit  strong
annual  variabilities.

     However the situation may be somewhat different in tributary rivers.  In
areas dominated by agriculture it appears that higher (10 ppb) herbicide
concentrations can occur, even if only for brief periods of time. These
events  appear to be related more to local rainfall patterns and herbicide
application times than to more general  features of climate and river flow.
Kemp et al. (1981) summarized atrazine runoff data reported in several  studies
and found that the total amount of  residue transported into waterbodies  was a
function of the timing of the first runoff event after herbicide application;
the longer the time interval between application and initial  runoff  event the
lower the prcentage of applied herbicide which was removed.  Uauchope and
Leonard (1980) found an exponential decrease in concentration of atrazine with
increasing time interval between application and initial runoff event.   These
results are consistent with those reported earlier in this paper. Since  a
strong  runoff event occurred at the experimental watershed at Horn Point  only
3 days  after herbicide application in 1981, resultant concentrations may
approximate maximum values to be expected in such tributary rivers (Fig.  5b).
IP any  case, it appears that considerable short-term variability is  to be
expected in herbicide concentrations in waters derived from agricultural  areas
and tributary waters due to the interaction of application date and  subsequent
runoff  events and the limited dilution potential of small  creeks and rivers.
Because of these transient events,  it may well be worth investigating the
                                      1-13

-------
cummulative response of SAV to repeated short-term exposure to elevated
herbicide concentrations.
                                     1-14

-------
i
                                              LITtRATUkh CITED

             Anderson, R.R. and R.T. Macomber.  1980.  Distribution of submerged vascular
                  plants, Chesapeake Bay, Maryland.  Rept. to USEPA, Chesapeake Bay Program
                  by Aero Eco, Reston, VA.

             Bayley, S. , V.D. Stotts, P.P. Springer and J. Steenis.  1978.  Changes in
                  submerged aquatic macrophyte populations at the head of the Chesapeake
                  Bay, 1958-1975.  Estuarinss 1:74-85.

             Cornell, D.L., T. Wu, J.W. Pierce and M.A. Faust.  1977.   Rural non-point
                  pollution studies in Maryland (non-point pollution studies on
                  agricultural land use types prevalent in the Coastal Plain zone of
                  Maryland).  EPA 904/9-77-001.  Washington, DC.  361 p.

             Correll , D.L. , J.U. Pierce and T.L. Wu.  1978.  Herbicides and submerged
                  plants in Chesapeake Bay, p. 858-877.  jn:  Am. Soc. Civil Engineering.
                  Coastal Zone-78.

             Correll, D.L. and T.L. Wu.  1982.  Atrazi ne torir.ity to submerged vascular
                  plants in simulated estuarine microcosms.  Aquat. Bot. 14:151-158.

             Cunningham, J.J.  1980.  Effects of herbicide stress on the structure and
                  metabolism of the submerged macrophyte, Potamogetpn perfoliatus, in
                  estuarine microcosms.  M.S. Thesis, Univ. MD, College Park, MD.

             Frank, R. and G.J. Sirons.  1979.  Atrazine:  Its use in corn production and
                  its losses to stream waters in Southern Ontario, 1975-1977.  The Science
                  of the Total Environment. 12:223-239.

             Galston, A.  1979.  Herbicides:  a mixed blessing.  Bioscience 29:85-90.

             Gaynor, J.D. and V.V. Volk.  1981.  Runoff losses of atrazine and terbutryn
                  from unlimed and limed soils.  Environ. Sci. and Technol. 15:440-443.
f             Jones, T.W., W.M. Kemp, J.C. Stevenson and J.C. Means.  1982.  Degradation of
                  atrazine in estuarine water/ sediment systems and selected soils.  J.
                  Environ. Qual. (In press).

             Jones, T.U., W.M. Kemp. P.S. Estes and J.C. Stevenson.  1983.  Atrazine
                  uptake, phytotoxicity, release and short-term recovery for the submerged
                  aquatic plant, Potamogeton perfoliatus.  horn Pt. Environ. Laboratories,
                  Box 775, Cambridge, MD.

            Kaumeyer, K., W.R. Boynton, L. Lubbers, K, Staver, S. Bunker, W.M. Kemp and
                  J.C. Means.  1981.  Metabolism and biomass of submerged macrophyte
                  communities 1n northern Chesapeake Bay, p. 353-400.  In:  Kemp et al .
                  (eds).  Submerged Aquatic Vegetation in Chesapeake Bay:  It's Ecological
                  Role in the Bay Ecosystems and Factors Leading to Its Decline.  Horn Pt.
                  Environ. Laboratories, Box 775, Cambridge, MD.

-------
r
il
               Kemp, W.M., J.C. Means, T.W. Jones and J.C. Stevenson.  1982.  herbicides in
                    Chesapeake Bay and their effects on submerged aquatic vegetation:  A
                    synthesis of research supported by U.S.E.P.A., Chesapeake Bay Program, p.
                    503-567.  In:  Chesapeake Bay Program Technical Studies:  A Synthesis.
                    U.S.E.P.A., Washington, DC.

               Kemp, W.M., W.R. Boynton, J.C. Stevenson, R.R. Twilley and J.C. Means.  1983.
                    The decline of submerged vascular plants in Chesapeake Bay:  Summary of
                    results concerning possible causes.  Mar. Tech. J. (In press).

               .omax, K., J.C. Stevenson, M.S. Christy and J.R. Todd.  1979.  Nonpoint
                    pollution data collection problems on low flat coastal topography, p.
                    46-52.  _[n:  K. Flynn (ed.) Nonpoint Pollution Control - tools and
                    techniques for the future.  Interstate Commission on the Potomac River
                    Basin, Rockville, MD.

               Means, J.C., Wijayarante and W.R. Boynton.  1983.  Fate and transport of
                    selected pesticides in estuarine environments:  Implications for ocean
                    pollution.  Can. J. Fish. Aquat. Sci.  (In press).

               Muir, D.C.G., J.Y. Yoo and B.E. Baker.  1978.  Residues of atrazine and
                    N-deethylated atrazine in water from five agricultural in Quebec, Arch.
                    Eriv. Contam. Toxicol. 7:221-235.

               Newby, L.C., R.A. Kahrs, K. Adams and M. Szolics.  1978.  Atrazine residues in
                    the Chesapeake Bay.  Proc. N.E. Weed. Sci. Soc.  32:339 (Abstract).

               Richard, J.J., 6.A. Junk, M.J. Avery, N.L. Nehring, J.S. Fritz and H.J. Svec.
                    1975.  Analysis of various Iowa waters for selected pesticides:
                    Atrazine, DDE and Dieldrin-1974.  Pesticides Monitor. J. 9:117-123.

               Roberts, 6.C., G.J. Sirons, R. Frank and H.E. Collins.  1979.  Triazine
                    residues in a watershed in southwestern Ontario (1973-1975).  Jour. Great
                    Lakes Research 5:246-255.

               Schepers, J.S., D.R. Anderson, G.E. Schuman, E.J. Vavricka and H.D. Wittmuss.
                    1978.  Agricultural runoff in tha midwest.  Nebraska Agric. Exp. Station
                    Public No. 5643.

               Stevenson, J.C. and N.M. Confer.  1978.  Summary of available information on
                    Chesapeake Bay submerged vegetation.  U.S. Dept. Inter.  IWS/OBS-78/66.
                    NTIS, Springfield, VA.  333 pp.

               United States Geological Survey.  1983.  Estimated streamflow entering
                    Chesapeake Bay.  U.S.G.S., 208 Carroll Blvd., LaSalle Road, Towson, MD.

               Waldron, A.C. and C.W. Bailey.  1974.  Pesticide movement from cropland into
                    Lake Erie.  Environmental Protection Agency Technical Series.
                    EPA-660/2-74-033, Washington, DC.
                                                    1-16

-------
Walsh, t.t. , U.L. Hansen and D.A. Lawrence.  1982.  A flow-through system for
     exposure of seagrass to pollutants.  Mar. Envir. Res. 7:1-11.

Walsh, J.J.  1976.  Herbivory as a factor in patterns of nutrient utilization
     in the sea.  Limnol. Oceanogr. 21:1-13.

Wauchope, R.D. and R.A. Leonard.  1980.  Maximum pesticide concentrations in
     agricultural runoff:  a semi-empirical formula.  J. Environ. Qual.
     9:665-672.

Wetzel, R.L., P.A. Penhale and K.L. Webb.  1981.  Plant community structure
     and physical-chemical regimes at the Vaucluse Shores study site.  Chap.
     2, Sect. A.  Virginia Inst. Mar. Sci., Gloucester Pt., VA.

Wu, T.L.  1980.  Dissipation of the herbicides atrazine and alachlor in a
     Maryland corn field.  J. Environ. Qual. 9:459-465.

Zahnow, E.W. and J.C. Riggleman.  1980.  Search for linuron residues in
     tributaries of the Chesapeake Bay.  J. Agric. Food Chem. 28:974-978.
                                      1-17

-------
                                                             SCALE Of MILES
                                                           10    0    10    20
 UPPER CHESAPEAKE SITE

 INDEXING STATION LOCATIONS

     • SAV present in 1980

     O SAV lost since 1976

     A SAV lost since 1969

     X SAV not present since 1959-61
Figure 1.   Location and coding of sites Investigated during the littoral zone
           survey conducted 1n upper Chesapeake Bay during 1980.  Letitude,
           longitude and other station characteristics are given 1n Table 1.

-------
                                                                                                              oca*
                                                                                                              _J  (O t-
at z
a. a
< o
S-


0>
                                                         1-19

-------
                                                        Marsh Watershed

                                                                  J
                s. Agricultural Watershed '•',
Figure 3.   Configuration and  sampling  station  locations at the Horn Point
           experimental  watershed  located adjacent to the Choptank River
           estuary.
                                       1-20

-------
o
u
o
O
o
•o
o
>
"o
m
1.0


0.5

  0
     1.0
     0.5
         0
       mouth
                    25 Apr 80
                                    • Atrazine
                                    o Linuron
              16  Sep 80
             10
20
30      40
     tidal-fresh
                    River Mile, nautical
 Figure 4.  Concentrations of dissolved atrarlne and llnuron 1n
           surface waters of the main channel  of the Choptank River
           estuary 1n 1980.

-------
r
                   »
                   xT
o

|

»-
2

U
z
o
u

UJ
z
IM

o;
»-
<
                      10-
                       10-
                           (0)
                   *_
                   jo"
                   <
                   tr
                   u

                   I
                   u
                   o
                   u
                   ffi
                   tr
                       lO-i
                       8-
                       6-
4-
2-
     INITIAL

     HERBICIDE
     APPLICATION

     W0ler*h*d
          MPEL
                                                   -RUNOFF FROM
                                                   AGRICULTURAL
                                                   FIELD   HPEL
                                                  FLOW OVER
                                                  BEAVER
                                               \  DAM
                                                \
                                          MAY
                                            JUNE
    (b)
-
-
•^




W///////////A

^•B



mmmm
\
1


J

D™
I | J] J
                             75      65     60      35   v  25     20

                                         DISTANCE  FROM  MOUTH  (km)
                                                      10
                Figure 5.   Temporal patterns of atrazlne concentration sin  runoff from 2
                           agricultural watersheds in the Choptank River basin during the
                           spring of  1981 (a) spatial distribution of herbicides along the
                           longitudinal axis of the Choptank River estuary from May 10-13,
                           1981  (b) and details of rainfall  at the Horn Point watershed,
                           runoff from two sites and dissolved atrazlne concentrations measured
                           at  the flume at HOrn Point and at Beaverdam  (c).  Data are presented
                           in  Appendix Table 9.
                                                 1-22

-------
  E   «•
  u
  *-   3-


  a?   2-
  U
  UJ   1-
  cc
  0.
                           ll II 1.111
  X ioo
  o
     soH
 £

 u.
         Horn  Pt.
.  I UU
    «oo4  Beaver  Dam
.o
 a.
    13-
    10-
 UJ  s-
         Horn  Pt.
OC


<  10-
8  '^
o
                          13.7
                                10.0
                                   A
                                     -
        Beaver  Dam /v**0
                   I          Ii_^I          I

            APR     MAY      JUN      JUL      AUG
 Flgurs 5c.
                           1-23

-------
            80
         I  60
         i
         CD

         W
            40
            20
                '  1
                •  1
                •  I
                  1
                *'.:!
                                     Sampling Periods:
                                 • Apr
                                 o Jun«
                                 u M
                                 A Sep
• Herbicide cone, below level of detection
         (S0.05 ppb)
                           Apr data from 3 atationa
                                            e
                0.05       0.2           0.4          0.6
                                  Diaaorved Atrazin*. ppb
                                           —0~-
                                            0.8 1.0
         N
            80
            60
            40
            20
               O  i

               A  !
               O  i
                0.05
              Sampling Periods:
             .Apr
             o June
             • Jo)
             * Sep
 Herbicide cone, below level of detection
          ($0.05 ppb)
                       . Apr data from 5 stations

                      - Sep data from 3 stutions
                           0.2
04           0.6
   Lint won. IH*(I
                                             08
Figure  6.  Scatter diagram of  SAV blomass versus  dissolved atrazlne (a)  and
            dissolved  Unuron (b)  concentrations observed  at Httoral zone
            stations 1n  upper Chesapeake  Bay 1n  1980.
                                            1-24

-------
  O
  o
  CD
 CL

  CD

 i
      100
       80
       60
       40.
       20.
                                ATRAZfNE
                                           M. spicatum
                                    Y= 147.4 - 47.0X Q"
                                   45 ppb
                                          lJ5 80 ppb
                             10
                                    50    100
                       Log  Herbicide Concentration (ppb)

Figure  7.  Regressions of mean apparent 02 production (Pa; expressed at % of
          control value) versus the log of herbicide concentration for P.
          perfollatus and  M. spicatum treated with (a) atrazlne and (b)~
          llnuron.TTlso sRbwn are I50 values (herbicide concentration ore-
          dieted to cause  50% photosynthetlc Inhibition) for each species.
          Diagram was adapted from Kemp et al.  (1983).
                                       1-25

-------
                               CM
                               00
                               O>
                               o
                               CO
                               O> €0
                               0>
                               CO
                               N.
                               0>
o
CD
                           o
                           CM
sp   '
                  |enuuv '
                                                                                            CM
                                                                                            CO
                                                                                            0)
                                                                                            CO
                                                                                            00
                                                                                            O
                                                                                            co
                                                                                            o
                                                                                            0»
                                                                                            N
                                                                                            O
                                                                                          CO
                                                                                          N.
                                                                                          0)
                                                                                            O)
                                                                                                           O)
                                                                                                      O)  ID
                                                                                                      10 -M
                                                                                                      (/I  «O
                                                                                                      O> O

                                                                                                      Q.
                                                                                                      X   •
                                                                                                      (!) CM
                                                                                                        00
                                                                                                        c
                                                                                                       f- T3
                                                                                                        2 O
                                                                                                        O ••-
                                                                                                        C J-
                                                                                                        O QJ
                                                                                                       U £^
                                                                                                     •O
                                                                                                      01 S-
                                                                                                      t- O
                                                                                                      3 <4-

                                                                                                      ra X
                                                                                                      O] O

                                                                                                 CO  — £

                                                                                                £   ». >»
                                                                                                 (S   o> .—

                                                                                                 o   > ^
                                                                                                     E5 c
                                                                                                        o
                                                                                                     c o>
                                                                                                     ID O>
                                                                                                     ^ ID
                                                                                                     Ol i.
                                                                                                     3 Ol
                                                                                                     CT >
                                                                                                     tn RI .
                                                                                                        -a oo
                                                                                                        c a»
                                i— 
                             0) 
                              c
                             •»-   »—
                                                                                                     f  10 •>-
                                                                                                     W    Q.
00


i.
                                                Sp 'MOJJ
                                                                            'AV
                                            1-26

-------






5




• * ,
* *

. . •
•/•
•i
* * ,
* • »

• *
I * •"
r ""
<



c
*• o o> co o
CO CO N- h- ',5
O> O) O> O) U
T- t- T- T- Q>
• X O -4 g
n
o

«••
tt

CD
V.
*•".<
••y
.*,
. • .
.* .-
[•/
*jC
» *•
;Q
* • #
• * «
• *• *
* • i
• * *

;c







0
o
CM




o
(O
^"




o
CM







0
to













CO
o
T*»
X
CO
"o
o
u_

^*fc
^^
*TJJJ

o
CO
0)
CO
^
(O
(. •<->
0) IO
> 0
oE
 (0
S- E
C 10
Ol C
U C
c m
O f
U O)
3
O) O-
C (0
•r- 3 •

IO CO
i- O> CO
*-> x: w
IO 4-> t— 1
•o  r— •
qdd 'auj
        U)
        o


P3A|OSSIQ
                                                    CD    42
                                                    >    T3


                                                           O

                                                           E

                                                           i.

                                                           10

                                                          •O


                                                           O)
                                                          4->
                                                          4->
                                                           (O

                                                          in


                                                            •
                                                          CTi



                                                           I
                                                           Ol
                                          «o o
                                          •O L.
                                            <*-
                                          O)
                                          C«T>
                                          IO Q>
                                          i. t—
                                          41 -i-
                                          > Q.
                                          •o E

                                          (/) 
-------
,"
CO
DO

CD
X.
CO
Q)
a
CO
0}
0)
£
O
                                                               *£
                                                 O
                                                 CO
                                                 O>
                                                          •*
                                             i      i      i
                                             w    o     *>
                                             *-'    *-'     o
      <*- Q> C
      O S- O
         10 O
      I/)
      i- o -a
      ai oo c
      +j a> 10
      
      o <*- c
      to   ai
      >*- 10 >
      J- 4-> O)
      3 IO 4->
      v> o 

      c   a)
      f-  • 0)
           §
           —
      O O)
      J. ~* C

      c -o *J
      T3 r»
      c r^
      IO 91
        *—«
      a,

      = *S
      N
      „>,

      *• 3
 >,   »^

S    13 -O
 C    V C
—    > <0

 CO    *o »
                                                                                   J- t-l
            <0i—
                                                                             <4- O
                                                                              o<*-
                                                                                      a>
                                                                              «/> --, « —
                                                                              C <0 3  $
                                                                              o CQ -t->  ai
                                                                              ^-    Ifl Z
                                                                                0)
       j.  0> ^I <0
       C Q--4->
     .  Q) «0    —»
       t> wi  e oo
       c: a>  or*-
         5.c  i. en
         (_) M- r-<
       ai
               qdd  'suoijejjuaouoo  eppiqjeH
                                                   1-28

-------
Table 1.  Station coding, location and site descriptors associated with
          Upper Bay site indexing study, 1980.  Units are:  silt-clay
          content, %; median diameter, mm; light extinction, nr1
          (k=extinction coefficient).
                                               SEDIMENT TYPt
SAV STATION
STATUS CODt
1 SI-4
SAV
Present ,, Q
loan ai-y
i'OU
Sl-10

si-n

SI-M

SI-21

SI-28

II Sl-15
SAV
Lost 51-20
Since
1978 51-23

III SI-1
SAV
Lost SI-2
Since
1969 SI-S

Sl-6

51-16

SI-17

Sl-19

SI-22

51-24

51-25

SI-27

IV 51-7
SAV Not
Present 51-8
(«t leist)
Since 51-18
1959-61)
.
LOCATION
Susquehanna Mouth
Concord Point
Magothy River
James Pond
Chester ft. Mouth
E. Neck Narrows
Chester River
Cliffs Bight
Choptank Diver
Todds Cove
S»1th Island
Otter Is.
Gunpowder R. Mouth
Lower Is. Pt.
ChopUnk Mver
Benonl Pt.
South Marsh Is.
Johruon Pt.
Big Annemessex R.
Wear Point
Patuxent R. Mouth
Solomons Is.
Mid-Chesapeake
Western Shores
Susqoehanna Mouth
Furnace Bay
Susquehanna Flats

Chortank River
Leconpte Cr.
Choptark River
Horn Point
Choptank River
Frailer Neck
Smith Island
Horse Hmwch
819 AnneMSsex R.
Moon Bay
Barren Island
Sand Pt.
Patuxent River
Buena Vista
Elk River
Cabin John Cr.
Elk River
Veazey Neck
Choptank River
Kingston Ldg.
LATITUDE/
LONGITUDE
39* 32' 08"
76*05 '17"
39" 05 '33"
76*26'07"
39*02'54"
76*13'42"
39*05 ' 17"
76*09' 30"
38*36 '38"
76*1' '00"
37*58'16"
75*59'42*
39* 18' 34"
76*20'40"
»*41'03"
76*12'54"
38*06 ' 37*
76*03 '18*
38*02 '33-
75.49.49.
38*19'07"
76*27 '09*
38*29 '26"
76*29'45"
39'33'17"
76*02 '38"
39*30'42"
76*04 '16"
38'36'H"
76*10'35"
38*35 '38"
78*07 '52"
38*44-00"
76*00 '09"
37*56'57-
75*59-42"
38*04-12"
75*47-42"
38*18-45*
76*15-00*
38*30'46"
76*39'47"
3f V34"
75 -/'ll*
39*28-11*
75*57 -01 "
38*46-26*
75*57'57*
PERCENT
SILT/CLAY
55.8124.4

17.011.3

9.014.7

32.0*23.5

34 9i24.4

54.617.7

45.2H1.0

75.1114.3

63.5124.5

42.4H4.0

0.910.5

0.810.7

60.H33.4

1.4H. 9

31.8110. 9

52.010.0

2.7*0.5

80.U8. 2

29.3124.3

52.4130.3

0.610.2

28.81). 2

2.712.8

51.8112.9

MEDIAN
DIAMETER
0.2010.08

0.20i0.02

0.1310.01

0.2010.03

0.1410.01

0.1210.0

0.2010.03

0.1310.0

O.UiO.Oi

0.3710.06

0.2St0.09

.201.01

.171.03

.231.02

.191.01

l.Oi.OO

.381.05

.141.02

.391.03

.16-. 03

2.01.00

.141.01

.721.86

.291.15

LIGHT
EXTINCTION
1.5t0.52

2.5U.4

1.4i0.3

1.7t0.2

1.7i0.6

1.9H.O

1.110.6

1.310.5

2.310.8

1.710.9

1.510.9

1.810.6

2.310.6

1.8:0.7

2. Oil. 7

l.OiO.O

4.4i2.7

3.9ili

2.4lO£

2.611JB

2.910.2

4.010.9

2.610.4

4. 111. 3


-------
1
             Table 2.
Summary of maximum and average dissolved atrazine (A)  and
Mnuron (B) concentrations measured in various geographical
 re          UPP6r Chesapeake Ba> durin9 4 sampling periods
                                                           Sampling Periods
      A. Atrazine
         • Maximum  Observed
              Concentration

         • Percent  Observations
                    >0.5  ppb
                    £0.1  ppb
         • Mean Concentrations
           by Geographical Area3
                    Upper Bay
                    Upper-Mid Bay
                    Mid Bay
                    Lower Bay
                    Tidal Fresh Tribs.
                    Stations with SAV
                 0.3
                 0
                36
                0.09
               <0.05
                0.11
                0.11
               <0.05
                0.08
                                                         June
                              1.1-1.2
 27
 32
 0.78
 0.18
 0.13
<0.05
<0.05
 0.29
                                                 July
               0.4
  0
 35
 0.18
<0.05
<0.05
 0.07
 0.12
<0.05
                              September
               0.9
  9
 17
£0.05
<0.05
~0.16
<0.05
 0.30
<0.05
                                                           Sampling Periods
      B. Linuron
         • Maximum Observed
              Concentration
         • Percent Observations
                   £0.50  ppb
                   >0.10  ppb
         • Mean Concentrations .
           by Geographical Area0
                   Upper  Bay
                   Upper-Mid Bay
                   Mid Bay
                   Lower  Bay
                   Tidal  Fresh Tribs.
                   Stations with SAV
                                          April
                   <0.05

                   0
                   0
                   <0\05
                   <0.05
                   <0.05
                   <0.05
                   <0.05
                   <0.05
                                  June
  <0.05

   0
   0
  <0.05
  "0.05
  "0.05
  "0.05
  ^0.05
  <0.05
1.3

 15
 20
 0.42
<0.05
"0.05
<0.05
~0.20
 0.13
           aStations were grouped according to geographical  areas  as  follows:
                Upper Bay:  Si's 4, 5, 6,  7 and 8.
                Mid-Upper Bay:  Si's 9, 10, 13 and 28.
                Mid Bay:  Si's 1, 2, 14, 15, 16 and 15.
                Lower Bay:  Si's 20, 21, 22, 23 and 24.
                Tidal Fresh Tributaries:
                Stations with SAV:
                  locations.
                              September
  0.75

   9
  13
   0.12
  <0.05
  <0.05
  70.05
  ~0.25
  <0.05
             ._.   SI'S 18, 19 and 27.
             SI'S  4,  9, 10, 13, 14, 21  and 28;  Refer  to  F1g. 1 for
                                                 1-30

-------
       Table 3.  Station locations at which maximum dissolved atrazine U)
                 and linuron (B) concentrations (ppb) were observed during
                 1980 littoral zone surveys.
A. Atrazine

   Locations
    Upper Bay
    Upper-Mid Bay
    Mid Bay
    Lower Bay
    Tidal Fresh Tribs.
    Stations with SAV
B. Linuron

   Locations a
    Upper Bay
    Upper-Mid Bay
    Mid Bay
    Lower Bay
    Tidal Fresh Tribs.
    Stations with SAV
                                                  Sampling Periods
April
0.20(SI-7)
<0.05
7.30(51-1)
0.30(SI-24)
0.10(51-18)
0.20(SI-21)
June
1.14(SI-6)b
0.45(51-13)
0.65(51-2)
0.10(51-21)
0.10(51-19)
1.09(51-4)
July
0.4(51-8)
<0.05
D".13(SI-1)
1.20(51-22)
0.25 (SI-27)
0.20(51-4)
September
0.11(51-8
0.10(SI-13)
0.93(51-1
<0.05
ff.91(SI-27)
0.10(51-13)
                                                  Sampling Periods
April
<0.05
<0.05
TO. 05
TO. 05
<0.05
TO. 05
June
<0.05
TO. 05
TO. 05
TO. 05
TO. 05
TO. 05
July
1.30(SI-7)C
<0.05
7.25(51-1)
<0.05
7.55(51-27)
0.80(SI-4)
September
0.56(51-8)
<0.05
7.3l(SI-l)
<0.05
7.75(51-27
<0.05
aSee Table 2 for station groupings and Fig.l for locations.

 At other upper bay stations atrazine concentrations (ppb)  were
 1.0?(SI-4), <0.05(51-5), 1.08(51-7) and 0.7(SI-8).

cLinuron concentration was 0.8 ppb at Station SI-4 (upper bay area).
                                            1-31

-------
Table 4.  Stations and dates at which detectable UO.l vg kg"1)
          concentrations of atrazine and linuron were found in surficial
          sediments collected from littoral zone sites in upper
          Chesapeake Bay, 1980.
                     t

                                Herbicide Concentration, yq kg"1

  Date           Station            Atrazine         Linuron
April 28          SI-23               0.37            <0.10
June 13
June 13
June 12
Sept. 17
Sept. 15
SI-5
SI-7
SI-18
SI-28
SI-23
0.65
0.43
0.64
0.93
<0.10
<0.10
SO. 10
SO. 10
<0.10
0.51

-------
-*«
 f
         Table 5.  Summary of mean SAV aboveground biomass (gdw/m2)
                   measured on 4 occasions in upper Chesapeake Bay, 1980.


         DATE       UPPER BAY                   MID-BAY               LOWER BAY

Apr i 1
June
Jul y
Sept
SI-4
0
24.8
0.0
0.0
SI-28
0
91.2
17.2
0.0
51-9 51-10
0 0
10.3
0.8 2.0
0.0 19.2
51-13
0
4.6
5.6
12.0
51-14
0
33.2
66.0
49.2
51-21
0
14.4
35.6
29.2
                                           1-33

-------
Appendix Table 1.  Listing  of dissolved herbicide concentrations (ppb), SAV
                   biomase.  and other selected variables measured at 23 littoral
                   zone statiuis in upper Chesapeake Bay,  1980.  See text for
                   explanation of SAV status.  Temperature in °C, salinity  as
                   ppt, and SAV biontass as gdw/nr.  All water quality samples
                   were taken 0.5m below the surface.


 SAV    STATIONS   EftTE   TEMP  fiAT.TMTTy    HEBBTCTne CCNC fPCb)    	SAV
5TATDS
  II
 III
  IV
April 1980 Cruise
SI-4
SI-9
SI-10
SI-13
SI-14
SI-21
SI-28
0JL **W
SI-15
SI-20
SI-23
SI-1
SI-2
SI-5
SI-6
SI-16
SI-19
SI-22
SI-24
SI-25
SI-27
SI-7
SI-8
SI-18

24/4
24/4
24/4
25/4
25/4
28/4

25/4
28/4
28/4
23/4
23/4
24/4
24/4
25/4
25/4
28/4
28/4
28/4
1/5
25/4
24/4
25/4

13.2
18.7
19.2
16.2
17.8
16.3

18.8
15.9
17.0
15.6
17.8
13.6
13.6
19.3
18.0
15.8
17.5
15.3
15.4
17.3
16.8
18.0

0.0
3.3
5.2
5.5
9.6
11.7

9.8
11.9
n.i
8.7
7.6
0.0
0.0
9.1
0.9
11.6
10.9
8.7
3.2
0.0
0.0
0.4

A. tr 92 JflS Lipuroo
0.07
10.05
10.05
10.05
0.15
0.20

0.10
10.05
10.05
0.30
10.05
0.15
10.05
10.05
10.05
10.05
0.30
0.10
10.05
0.20
10.05
0.10
1-34
10.05
10.05
10.05
10.05
10.05
10.05

10.05
10.05
10.05
10.05
10.05
10.05
10.05
10.05
10.05
10.05
10.05
10.05
10.05
10.05
10.05
10.05

Biomasa
0.0
0.0
0.0
0.0
0.0
0.0

0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0

Species
-
-
~
-
-
-

-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-


-------
r
   Appendix Table 1 (cont.)
   STATUS
                                                                                    SAV
    II
   III
    IV
]
SI-4
SI-9
SI-10
SI-13
SI-14
SI-21
SI-28
SI-15
SI-20
SI-23
SI-1
SI-2
SI-5
SI-6
SI-16
SI-19
SI-22
SI-24
SI-25
SI-27
SI-7
SI-8
SI-18
3une 1980 Cruise
13/6 20.5 0.1
12/6
12/6
12/6
11/6
13/6
12/6
11/6
11/6
16/6
16/6
15/6
13/6
12/6
12/6
1V6
11/6
11/6
16/6
13/6
13/6
12/6
24.8
25.5
23.8
22.4
24.0
24.9
24.5
21.1
23.4
21.9
19.5
22.5
21.8
-
21.5
20.7
22.4
24.7
22.0
22.0
21.1
9.3
6.5
10.0
12.8
2.3
9.9
12.2
12.9
13.0
12.3
0.1
0.1
9.5
-
13.0
13.2
11.2
7.7
0.3
0.2
0.6
Atrazine
1.09
._. — -MTYP '
pi^uron
10.05
10.05 10.05
0.45
10.05
0.10
0.09
10.05
10.05
10.05
10.05
0.65
10.05
1.14
10.05
0.10
10.05
10.05
10.05
10.05
1.08
0.70
10.05
10.05
10.05
10.05
10.05
10.05
10.05
10.05
10.05
10.05
10.05
10.05
10.05
10.05
10.05
10.05
10.05
10.05
10.05
10.05
10.05
Biomass
24.8
10.3
4.6
33.2
14.4
91.2
0.0
0.0
0.0
0.0
o.n
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
Species
H^ spieatutn
mixed9
Bi. flaritina
£^ perfoliatus
E^ perfoliatus
^BBEina
E*. perfoliatus
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
                                                  1-35

-------
'II
 ••
 •  Appendix Table 1 (cont.)
 SAV    STATIONS   DATE  TEMP   SALINITY   HERBICIDE CQNC^DDb)
STATUS
                                                                                  SAV
     II
    III
     IV
July I960 Cruise
SI-4
SI-9
SI-10
SI-13
SI-14
SI-21
SI-28
SI-15
SI-20
SI-23
SI-1
SI-2
SI-5
SI-6
SI-16
SI-19
SI-22
SI-24
SI-25
SI-27
SI-7
SI-8
SI-18
26/7
26/7
25/7
25/7
24/7
24/7
25/7
24/7
24/7
24/7
29/7
26/7
26/7
26/7
24/7
25/7
24/7
24/7
24/7
29/7
26/7
26/7
25/7
27.3
31.9
-
31.8
28.6
26.7
26.7
29.6
28.9
25.7
28.3
28.5
26.2
30.1
28.4
28.7
26.2
25.5
2P.O
28.2
31.2
29.7
28.4
0.2
8.4
-
7.3
9.8
14.7
3.8
10.5
12.8
14.1
14.0
12.6
0.3
0.1
10.4
3.1
15.2
13.7
12.9
10.0
0.2
0.2
1.7
Atrazine
0.20
10.05
10.05
10.05
10.05
Linuron Bionass
0.80
10.05
10.05
10.05
10.05
NOT SAMPLED
10.05
10.05
NOT
NOT
0.13
10.05
10.05
0.20
10.05
0.10
0.20
10.05
10.05
0.25
0.20
0.40
10.05
10.05
10.05
SAMPLED
SAMPLED
0.25
10.05
10.05
10.05
10.05
10.05
10.05
10.05
10.05
0.55
1.30
10.05
10.05
0.0
0.8
2.0
5.6
66.0
35.6
17.2
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
Species
-
Vj. americana
£*. perfoliatus
SM. perfoliatus
B«. maritima
E*. perfoliatus
Z. marina
Pj. perfoliatus
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
—
                                                 1-36

-------
                                                /
Appendix Table 1 (cont.)



 SAV    STATIONS   DKTE  TEMP   SALINITY   HERBICIDE CCNC  (ppbj
                                  SAV
STATOS Septe|riper 1980 Cruise
I SI-4
SI-9
SI-10
SI-13
SI-14
SI-21
SI-28
II SI-15
SI-20
SI-23
III SI-1
SI-2
SI-5
SI-6
SI-16
SI-19
SI-22
SI-24
SI-25
SI-27
IV SI-7
SI-8
SI-18
18/9
17/9
17/9
17/9
16/9
15/9
17/9
16/9
15/9
15/9
22/9
18/9
12/9
17/9
16/9
16/9
15/9
15/9
17/9
22/9
17/9
17/9
16/3
20.5
25.2
23.4
22.2
23.1
26.2
25.2
23.8
24.9
25.8
26.7
25.9
20.0
23.2
26.5
25.7
25.9
25.6
21.5
26.0
25.8
26.0
25.7
aSarnples contained a mixture of
canadensis and ^ americana.
0.2
12.4
12.5
10.9
15.2
20.9
7.9
14.9
18.8
19.6
19.0
17.0
0.2
0.2
14.0
5.5
20.9
19.0
17.4
13.1
2.4
3.1
3.5
the following
Atrazine
10.05
10.05
10.05
OJO
10.05
10.05
0.07
10.05
10.05
10.05
0.93
10.05
10.05
10.05
10.05
10.05
10.05
10.05
10.05
0.91
0.06
0.11
10.05
species:
Lin or on
10.05
10.05
10.05
10.05
10.05
10.05
10.05
10.05
10.05
10.05
0.31
10.05
10.05
10.05
10.05
10.05
10.05
10.05
10.05
0.75
10.05
0.56
10.05
HL. spicatumaf \
Biomass
0.0
0.0
19.2
12.0
49.2
29.2
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
PA perfol
Species
-
mixed9
PA perfoliatus
PA maritima
PA perfoliatus
Z. mgffirfp
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
iatajs,. EA
1-37

-------
Appendix Table 2.  Listing  of atrazine and linuron concentrations (ppb) in
                   surf icial sediment samples (upper 5 cm) collected from 23
                   littoral zone  locations in upper Chesapeake Bay,  1980.

                                                           Sediment
April
Station
SI-4
SI-9
SI-10
SI-13
SI-14
CT— 91
OJ. ZJ.
CT— 7ft
OJ. £O
CT—1 ^
OX AD
SI-20
SI-23
SI-1
SI-2
SI-5
CTW;
Al.^)
SI-16
SI-19
SI-22
SI-24
SI-25
SI-27
SI-7
SI -8
SI-18
1980 Cruise
Date
24/4
24/4
24/4
24/4
25/4



28/4
28/4
23/4
23/4
24/4

25/4
25/4




28/4
1/5




25/4
Herbicide
Atrazine
10.1
iO.l
iO.l
10.1
10.1



10.1
0.37
10.1
10.1
10.1

10.1
10.1


___— — KTYT JIM71T VfVTi.

10.10
10.10


"""" ntiTUT VTBTkp

10.10
Concentration fpob)
Linuron
10.1
10.1
10.1
10.1
10.1



10.1
10.1
10.1
10.1
10.1

10.1
10.1




10.10
10.10




10.10
                                       1-38

-------
                  Sediment
JUne 1980
Station
SX-4
SI -9
•JJL y
SI-10
SI-13
SI-14
SI-21
SI-28
SI-15
SI-20
SI-23
SI-1
SI-2
SI-5
SI -6
SI-16
SI-19
SI-22
51-24
SI-25
51-21
SI-7
SI-8
SI-18
Cruise
Date
13/6

12/6
12/6
12/6
11/6
13/6
12/6
11/6
1V6
16/6
16/6
13/6
13/6
12/6
12/6
11/6
1V6
11/6
16/6
13/6
18/6
12/6
Herbicitfe
Atrazine
10.1

£0.1
iO.l
10.1
iO.l
10.1
£0.1
10.1
10.1
10.1
10.1
0.65
10.1
10.1
10.1
10.1
10.1
10.1
10.1
0.43
10.1
0.64
Concentration >fpb)
Linuron
10.1

10.1
10.1
10.1
10.1
10.1
10.1
10.1
10.1
10.1
10.1
10.1
10.1
10.1
10.1
10.1
10.1
10.1
10.1
10.1
10.1
10.1
1-39

-------
                   Sediment
July
Station
SI -4
SI-9
SI-10
SI-13
SI-14
SI-21
SI-28
SI-15
SI-20
SI-23
SL-1
SE-2
SI-5
SI-6
SI-16
SI-19
SI-22
SI-24
SI-25
SI-27
SI-7
SI -8
SI-18
1980 Cruise
Date
26/7
26/7
25/7
25/7
24/7
24/7
25/7
24/7
24/7
24/7
29/7
26/7
26/7
26/7
24/7
25/7
23/7
24/7
24/7
29/7


26/7
26/7
Herbicitfe
Atrazine
10.1
10.1
10.1
10.1
10.1
10.1
10.1
10.1
10.1
10.1
10.1
10.1
10.1
10.1
10.1
10.1
10.1
10.1
10.1
10.1


10.1
10.1
Concentration (rob)
jAmron
10.1
10.1
10.1
10.1
10.1
10.1
10.1
10.1
10.1
10.1
10.1
10.1
10.1
10.1
10.1
10.1
10.1
10.1
10.1
10.1


10.1
10.1
1-40

-------
                    Sediment
Septenber
Station
SI-4
SI-9
SI-10
SI-13
SI-14
SI-21
SI-28
SI-15
SI-20
SI-23
SI-1
SI-2
SI-5
SI-6
SI-16
SI-19
SI-22
SI-24
SI-25
SI-27
SI-7
SI-8
SI-18
1980 Cruise
Date
17/9
17/9
17/9
17/9
16/9
15/9
17/9
16/9
15/9
15/9
22/9
18/9
18/9
17/9
16/9
16/9
15/9
15/9
16/9
22/9
17/9
17/9
16/9
Herbicid^
Atrazine
10.1
10.1
10.1
10.1
10.1
iO.1
0.93
10.1
10.1
10.1
10.1
10.1
10.1
10.1
10.1
10.1
10.1
10.1
10.1
10.1
10.1
10.1
10.1
Concentration (Epbl
Linuron
10.1
10.1
10.1
10.1
10.1
10.1
10.1
10.1
10.1
0.51
10.1
10.1
10.1
10.1
10.1
10.1
10.1
10.1
10.1
10.1
10.1
10.1
10.1
1-41

-------
•t!
        Appendix  Table 3.
         27/5/80
         12/6/80
         25/7/80
         18/8/80
         16/9/80
         Summary of station locations, physical characteristics and dissolved
         herbicide concentrations (ppb) measured during monthly water quality
         surveys in the Choptank River Estuary, 1980.  All samples were taken
         0.5m from the surface. Station locations are shown in Fig. 2.
    LOCATION

Kingston Ldg.
Frazer Neck
Goost Pt.
Choptank Mouth
Lloyds Ldg.
Blinkhorn Cr.
Jamaica Pt.
Hambrook Bar
LeCompte Bay
Todds Cove
Choptank Mouth
Kingston Ldg.
Frazer Neck
Blinkhorn Cr.
Goose Pt.
LeCompte Cr.
Todds Cove
Choptank Mouth
Crowberry Cr.
Frazer Neck
Cabin Cr.
Goose Pt.
Howell Pt.
Todds Cove
Choptank Mouth
Marker 60
Marker 55
Marker 41
Marker 31
Chlora Pt.
Todds Cove
Choptank Mouth
Marker 64
Kingston Ldg.
Frazer Neck
Cabin Cr.
Todds Cove
Choptank Mouth
                                    RIVER"
                                     MILE
33.
29.
17,
 0.0
26.8
22.
19.
10.8
 8.4
 3.5
 0.0
33.
29.
22.
17.
 8.4
 3.5
 0.0
35.
29.
                                       .3
                                       .5
                                       .7
                                       .4
                                       .4
.3
.5
.4
.7
.9
.5
                                     21.0
17.
10.
 3.5
 0.0
35.
30.
23.
16.
 7.6
 3.5
 0.0
38.
33.
29.
21.0
 3.5
 0.0
                                       .7
                                       .6
  .9
  .3
  .7
  .2
  .4
  .3
  .5
                              SALINITY
 0.5
 1.7
 6.7
 8.9
 1.2
 3.1
 5.9
 8.1
 9.1
 9.4
 9.8
 1.4
 2.6
 4.8
 7.0
 9.5
 9.9
10
 1
 4.0
 6.3
 7.8
10
11
11.8
 1.2
 4.0
 6.6
 9.9
12.2
12.8
14.2
 1.3
 3.4
 5.7
10.1
15.5
               .4
               .5
               .3
               .1
                                                  17.0
                         TEMP
                         TO"

                         18.0
                         17.6
                         17.1
                         14.5
                         22.8
                         22.2
                         21.7
                         20.8
                                                                20.
                                                                21.
21.5
22.3
22.0
21.7
21.5
21.4
21.
21.
                                                                  .2
                                                                  .2
27.6
28,4
28.1
                         28.
                         27.
                         27.
                         28.
                         27.9
                         27.4
                         27.0
                         26.7
                         26.7
                         26.*
                         25.9
                         25.2
                         25.4
                         25.6
                         24.6
                         24.3
                         24.1
                                         HEKBICIDE CONC.
Atrazine
I ppb)
0.08
0.07
<0.05
~0.06
0.06
0.10
0.06
<0.05
TO. 05
TO. 05
30.05
TO. 05
~0.30
0.50
<0.05
~0.55
0.06
0.35
0.72
<0.05
~0.23
<0.05
TO. 05
TO. 05
TO. 05
~0.48
0.16
0.08
<0.05
TO. 05
"0.39
<0.05
TO. 05
"0.06
<0.05
TO. 05
TO. 05
TO. 05
Linuron
Ippb)
<0.05
TO. 05
TO. 05
TO. 05
TO. 05
TO. 05
TO. 05
TO. 05
TO. 05
TO. 05
TO. 05
TO. 05
TO. 05
TO. 05
TO. 05
TO. 05
TO. 05
TO. 05
TO. 05
TO. 05
TO. 05
"0.08
<0.05
TO. 05
TO. 05
"1.14
0.78
0.06
<0.05
"0.20
0.95
<0.05
"0.05
<0.05
TO. 05
TO. 05
TO. 05
TO. 05
         aRiver mile  0.0  (nautical)  taken  as  being  between Blackwalnut and Cook  Pts.

-------
   f
                        Appendix Table 4.  Concentrat'ions of atrazine and linuron found in
                                  surficial  sediments (upper 5 cm) from samples collected
                                  along the  longitudinal  axis of the Choptank River estuary,
                                  1980.  Station locations are given in Figure 2.
                      DATE

                    25/4/80
                    12/5/80
                    25/7/80
                    18/8/80
\  ;
\ i
16/9/80
    LOCATION

Kingston Ldg.
Frazer Neck
Goose Pt.
Choptank Mouth

Kingston Ldg.
Frazer Neck
Blinkhorn Cr.
Goose Pt.
LeCompte Cr.

Crowberry Cr.
Frazer Neck
Cabin Cr.
Goose Pt.
Howe11 Pt.
Todds Pt.
Choptank Mouth

Crowberry Cr.
Marker "55"
Marker "41"
Marker "31"
Chlora Pt.
Todds Pt.
Choptank Mouth

Marker "64"
Frazer Neck
Choptank Mouth
                            RIVER
                             MILE
                            33,
                            29,
                            17,
                             0.0

                            33.3
                            29.5
                            22.4
                            17.7
                             8.4

                            35.9
                                                29,
                                                21.
                                                17.
                                                10.6
                                                 3.5
                                                 0.0
                            35.
                            30.
                            23.
                            16.
                             7.
                             3.5
                             0.0
  .9
  .3
  .7
  .2
  .6
38.4
29.5
 0.0
                     SEDIMENT
           HERBICIDE CONCENTRATIONS (ppb)
             AtrazineLinuron
                                                                                  0.1
                £0.1

                <0.1
                ~0.13
                <0.1
                <&.!
                                                                                  0.38
                                                              0.1
                                                                                  0.1
                                                   1-43

-------
Appendix Table 5.  Listing of atrazine and linuron concentrations  (ppb)
          measured in the participate fraction from water samples  collected
          along the longitudinal  axis of the Choptank Estuary,  1980.
          Station locations are given in Figure 2.


                                                PARTICULATE
DATE
25/4
27/5
27/5
12/6
11/6
27/7
27/7
18/8
18/8
18/8
16/9
16/9
16/9
RIVER
LOCATION MILE
Frazer Neck 29.5
Jamaica Point 19.4
Choptank Mouth 0.0
Kingston Ldg. 33.3
Choptank Mouth 0.0
Crowberry Cr. 35.9
Choptank Mo'Jth 0.0
Marker "60" 35.9
Marker "31" 16.2
Choptank Mouth 0.0
Marker "64" 38.4
Choptank Mouth 0.0
Frazer Neck 29.5
VOLUME
FILTERED
20£
20£
30£
104
20£
184
20£
20£
204
204
20£
20£
204
HERBICIDE CONCENTRATION (ppb)
Atrazine
£0.1
£0.1
£0.1
£0.1
£0.1
£0.1
£0.1
£0.1
£C.l
£0.1
£0.1
£0.1
<0.1
Linuron
£0.1
£0.1
£0.1
£0.1
£0.1
£0.1
£0.1
<0.1
<0.1
£0.1
£0.1
<0.1
<0.l
                                     1-44

-------
                                                                      •^ooooooooooooojpg
                                                                                                                                                      
  3 -M f-
  V)  (O U.
  rtj TJ
  Ol      Of
  S -O  Ol
      C OO
  C  (0
  O
  s-  
  fl3 ^~  ^

  O) •—
  C ITS r—
 •r-  L-  (O
  M O) •*->
  ID >  C
  S_ Ol Ol
 •u M e

     E T
 •o o 01
 oi s- a.
 > «»- x
 •—     LJ
 o -a
 VI O) +J   .
 10 -t->  C  Ul
 1- U ••-  C
 •a 01 o  o
     r- Q. -r-

 o o  c  10
     o  t-  u
 in      o  o
 c  u> z i—
 o  ai
 i- •—  Ol  C
 •<->  Q.JT  o
 «  E  +•> t-
 t-  *O     4->
 •»->   •*->  nj

 ai  t-     co
    ^(At/ivOkO*/>*o«o^**i/ii/»*/>mO*in^*i/>4/>4j>*/>4rtr*.
                                 OOOOOOOOOOOOOOOOOO   000   O   O   O   O   OOOO   OOOtOOOtM^OOOOOOCOOtOOOOOOO
                                                                                            Swj   w>   *n   «n «; *o tn   >o o» *nisj ^^ o o*
                                                                                            O   O   O   OOOO   r«» P>J *O */»*» ^»
                                                )^ocD<.->»n^t(^OM«
                                                                                                                                                                vlvfvi   vi
UJ


 0)

                                                                                                          ^^ «SJ CM IN*   ISt CM CM *S* O O O f* C
                                                          !CZCZ&o.f   30-^   3   a   x   a   T "w a. 5   5 5 a. a. :
                                                                                                                                       > o o o o a. a. a.  oo
                                                                                                                                          0000
                                                                                                                                                     .. a.
                                                                                                                                                                isii??
                                                                                                                                                   pope c	- —p p
                                                                                                                                                                         a.

-------
 "•* "•
rii
                Appendix Table   7,
Concentrations of  atrazine  and linuron  in the  suspended
particulate  fraction  of  water  samples  collected from
several  monitoring  stations at the  Horn Point  Experimental
watershed,  1980.    See Fig.  3 for  station  locations.
NR=not  recorded.
                                       UAH



                                       1/5
                                                 LOCATION
                                       2/5




                                       18/5

                                       27/5

                                       1/6

                                       3/6

                                       22/6

                                       2/7

                                       8/7

                                       16/7

                                       17/7





                                       22/7



                                       23/7
                                                  Bridge
       Pund


      Choptank

       Uier
       Pond

      Choptank

       Uier

      Choptank

       Uier

       Mier

       Uier



       Uier
1IHI


0815
1000
1300
1845
 NR
 NR
0845
1100
1445
1805
 NR
1200
1430
1500
1130

1605
1500
 NR
1400

0730

 NR

1230

1800

 NR

 NR

 NR

1300
                                                                          SUSPENDED PART1CULATI

                                                                        HERBICIDE CONCENTRATION,  ppl
       Uier      0010
       Uier      2215
      Bridge      2200
       Pond      2115
      Chopunk     2100

       Wter      2275
       Pond      2315
       Pond      2330
                                                  Uier
                                                  Uier
                                                  Uier
                                                  Uier
                                                  Pond
                                                  Pond
                                                Pond (mid)
                                                  Pond
                                                Pond (mouth)
                                                Pond (Hid)
                                                  Pond
                                                  Pond
                                                Pond (mouth)
                                                Pond (stream)
                 0100
                 0145
                 0800
                 1215
                 1742
                 0015

                 0030
                 0045
                 1315
                 1330
                 1400
                 1832
                 1832
                 1843
Atraiine
 <0.10
 70.10
 70.10
 70.10
 70 10
 70.10
 .10
 
-------
Appendix Table  8.
      Concentrations of atrazine  and  linuron  in  soil  samples
      collected  at the Horn Point  Experimental  watershed,  1980.
DATE
6/7
8/7
4/8
12/8
21/8
26/8
5/9
10/9
19/9
26/9
3/10
24/10
31/10
14/11
26/11
    LOCATION
                 HERBICIDE CONCENTRATION,  ppb

                 Atrazine             Linuron
crop
crop
crop
crop
crop
crop
crop
crop
crop
crop
crop
crop
crop
crop
crop
soil,
soil,
soil,
soil,
soil,
soil,
soil,
soil,
soil,
soil,
soil,
soil,
soil,
soil,
soil,
HPEL°
HPEL
HPEL
HPEL
HPEL
HPEL
Wier
Wier
HPEL
HPEL
Wier
Wier
HPEL
Wier
HPEL
<0.10
<0.10
~0.26
 1.20
 0.79
 1.10
<0.10
TO. 10
TO. 10
TO. 10
TO. 10
TO. 10
TO. 10
TO. 10
<0.10
<0.10
<0.10
TO. 10
TO. 10
<0.10
TO. 10
TO. 10
TO. 10
TO. 10
TO. 10
TO. 10
TO. 10
TO. 10
TO. 10
TO. 10
aHPEL referens to the Horn Point Experimental watershed.
                                     1-47

-------
fi
                 Appendix  Table   9.
Concentrations of  dissolved  atrazine and  linuron  in  water
collected  from several  locations  at  the Horn Point
Experimental  watershed  and  from several  stations  along  the
Choptank River estuary  in 1981.   See Figs.  1-3  for station
locations.
                                     LOCATION
                                                    DATE
                                                                TIME
                                       DlibULVll)

                               HERBICIDE COUCtNJJ(ATJCIN_._ppti

                               AtraZine           Linuron
                                     Horn Point Experimental Matershed
                                    Bridge
                                                    5/10
                                                    5/11
       5/15
       5/29
       6/2
       6/8
       6/10
       6/15
       6/19
       6/20
       6/Z2
       7/13
       7/20
       7/22

       S/15
       5/15

       5/11
                                    Pond («f
-------
                          QUALITY ASSURANCE  DATA


Appendix Table id   Summary of duplicate  field analyses of water.
                        1980            HERBICIDE CONC. (ppb)
STATION                 DATE          ATRAZINE          LINURON

SI-5                    24/4           0.15             <0.05
                                       0.20             <0.05
SI-1                    23/4           0.30             <0.05
                                       0.34             <0.05
SI-4                    13/6           1.09             <0.05
                                       1.05             <0.05
SI-?9                   12/6           0.10             <0.05
                                       0.13             <0.05
SI-4                    26/7           0.20              0.80
                                       0.21              0.84
SI-1                    29/7           0.13              0.25
                                       0.10              0.20
SI-1                    22/9           0.93              0.31
                                       0.89              0.34
SI-4                    18/9          <0.05             <0.05
                                      <0.05             <0.05
Jamaica Pt              27/5           0.06             <0.05
                                       0.05             <0.05
Marker 41               18/8           0.08              0.06
                                       0.08              0.05
Patuxent                13/7           0.81             <0.05
                                       0.81             <0.05
                                  1-49

-------
f:
                                 QUALITY ASSURANCE DATA


Appendix Table 11.  Summary of spiked sample analyses and percent recovery.
         STATION
                1980
                DATE
         SI-1            23/4
         +2 ppb spike    23/4
         SI-1            16/6
         +2 ppb spike    16/6
         SI-1            29/7
         +2 ppb spike    29/7
         SI-1            22/9
         +2 ppb spike    22/9
         Wier         22/7 (2225)
         +2 ppb spike 22/7 (2225)
         Wier         4/8 (1430)
         +2 ppb spike 4/8 (1430)
         Wier         3/10 (1000)
         +2 ppb spike 3/10 (1000)
 HERBICIDE CONC. (ppb)         % RECOVERY

ATRAZINE       LINURON       AT.
                             0.15
                             2.09

                            <0.05
                             2.01

                             0.13
                             2.01

                             0.93
                             3.01

                             0.76
                             2.59

                            <0.05
                             2.05

                            <0.05
                             1.91
               <0.05
                1.98

               <0.05
                2.03

                0.25
                2.08

                0.3]
                2.18

               <0.05
                1.93

               <0.05
                1.97

               <0.05
                2.03
 97.2


100.5


 94.4


102.7


 93.8


102.5


 95.5
 LIN.



 99.0


101.5


 92.4


 94.4


 96.5


 98.5


101.5
                                                   1-50

-------
                                CHAPTER II
             TEMPORAL RESPONSES OF A SUBMERGED VASCULAR PLANT
                 AND ITS ASSOCIATED AUTOTROPHIC COMMUNITY
                TO ATRAZINE STRESS IN ESTUARINE MICROCOSMS
                        J.J.  Cunningham1,  W.M.  Kemp1,
                       M.R. Lewis1»z and J.C. Stevenson1
      Point Environmental Laboratories, University of Maryland,
 CEES, P.O. Box 775, Cambridge, MD.  21613   U.S.A.

2Present address:  Department of Biology, Dalhousle University,
 Halifax, Nova Scotia, Canada B3H 401

-------
•f
 r
                                               ACKNOWLEDGEMENTS


                    Numerous individuals contributed to the conduct  of this research project,
               including:   S. Slacum and S.  Long for laboratory assistance; L.  Douglas,  R.
               Brinsfield  and S. Bellinger for consultation on statistical  analyses; J.
               Gilliard for typing;  D.  Kennedy and F. Younger for drafting; and R.  Galloway,
               T. Jones, R. Twilley, J. Means, W. Boynton,  T. Fisher, D.  Correll  and R.
               Sjoblad for advice and review of manuscripts.  Atrazine analyses were performed
               by C. Ganz  of EnCase  Laboratories in cooperation with L. Newby and L. Bal-
               lantine of  Ciba-Geigy Corp.  Computational work was supported by the Univer-
               sity of Maryland Computer Science Center.  This work  was funded  in part  by a
               grant from  the U.S. Environmental Protection Agency's Chesapeake Bay Program
               No. R805932010.
                                                      11-11

-------
                                                                                         f
                                    ABSTRACT


     Experiments were conducted over eight weeks in large (700 1)  estuarine
microcosms to test a variety of physiological  and morphological  responses of
the submerged vascular plant, Potamogeton perfoliatus.  to the herbicide,  atra-
zine.  At atrazine concentrations of about 0.13 ppm (equivalent  to highest
values occurring in the estuary) significant decreases  in apparent 03 produ-
ction (Pa) were exhibited immediately following treatment; however,  a signifi-
cant trend of photosynthetic recovery was evident within 2 wk, eventhough
atrazine levels remained relatively constant.   Since herbicide treatment  exerted
negligible effect on dark respiration (RM)t the metabolic ratio  Pa=Rn responded
in much the same way as did Pa.  Integrated community production and consump-
tion of 03 followed patterns similar to those  for P. perfoliatus,  and on
hypothesized atrazine-induced shift in relative contributions of each auto-
trophic group (macrophytes, phytoplankton, epiphytes benthic microalgae)  was
not borne out.  While areal densities of plant shoots for the experimental
populations were unaffected by this level of treatment,  total  biomass decreased
significantly, and a lag of 2-4 wk was evident following initial losses of Pa.
Morphology of individual shoots was markedly influenced  by atrazine, with
significant increases in mean shoot length and decreases in weight per unit
length.  Furthermore, chlorophyll a per unit leaf area  Increased 5-fold with
atrazine treatment, and Such effecTs are analogous to previously reported
shade adaptations of this and other submerged  plants.  The implications of
these results are discussed in terms of design for further dose-response  studies
of herbicides and SAV.
                                      II-l
I

-------
                                INTRODUCTION


     Structural and functional attributes of submerged vascular plants provide
a basis for diverse ecosystems which dominate the littoral zones of many aqua-
tic environments.  These plants, which are among the most productive In the
world  (Westlake, 1968; Mann 1982), supply a food-base for variegated detrital
trophic chains.  Their physical structure creates a unique habitat for epi-
phytic organisms (Nagel, 1968; Marsh, 1973; Nelson, 1979) and a refuge for
small fish and invertebrates from predation (Adams, 1976; Coen et al., 1981).
Submerged aquatic vegetation (SAV) markedly influences several aspects of its
environment, including:  the local physical circulation and related sedimento-
logical processes (Ginsburg and Lowenstam; Ginsburg, 1958; Fonseca et al.
1982;  Harlin et al.  1982).  The cycling of nutrients (McRoy et al. 1972,
Howard-Williams, 1981), and the geochemistry of sediment pore-waters (lizumi
et al. 1980, Kenworthy et al. 1982).

     While the ecological importance of SAV is widely recognized, in many
aquatic systems this role is being diminished or lost as plant populations
undergo major declines 1n abundance (e.g. Den Hartog and Poldermann, 1975;
Peres and Picard, 1975).  In one large North American estuary, Chesapeake
Bay, a continuing decrease in SAV distribution and abundance has been apparent
since the mid-1960's (Bayley et al. 1978, Orth and Moore 1981).  Numerous
factors have been suggested as contributing to this and other SAV declines,
Including decreased water clarity; (e.g. Kullberg, 1974; Jupp and Spence 1977);
nutrient enrichment and associated algal growth (Phillips et al., 1978; Sand-
Jensen and S0ndergaard, 1981); and Intense, localized phenomenon such as
grazing and foraging by waterfowl  and fish (Orth 1975; Kijirboe 1980).

     Che recent study has suggested that Inadvertent runoff of agricultural
herbicides (primarily the widely used compound, atrazlne) may have contributed
to the loss of SAV 1n Chesapeake Bay (Correll  and Wu 1982).  Moreover, 1n some
watercourses dense populations of these plants have been perceived as deleteri-
ous to human Interests, and active application of herbicides has been used to
control SAV growth (e.g., Brooker and Edwards 1975, Robson and Barrett 1977).
Thus, there 1s a need to understand the nature and magnitude of herbicide
effects on SAV.  Several studies have examined herbicide effect on SAV by
monitoring either loss of photosynthesis in shortterm (2h-4d) exposures (Sutton
et al. 1969, Bieleckl and Skrabka 1976) or changes in growth, mortality and/or
photosynthesis after longer (3-6 wk) exposures (Forney and Davis 1981, Correll
and Wu 1982).  However, few Investigations have considered temooral patterns
of response, and other measures of SAV physiology (such as pigmentation and
morphology) have largely been overlooked in relation to herbicides.  Some
researchers have measured herbicide effects on photosynthesis and respiration
of Integrated SAV communities (Brooker and Edwards 1973 a,b, Strange 1976,
Strange and Shreck 1976), but none have measured effects at levels of Individual

                                   II-2

-------
plants, plant populations, and associated communities.

     The present study was initiated to investigate temporal trends in photo-
synthesis, respiration, biomass and abundance (as well as other morphological
and physiological variables) for the submerged plant, Potamogeton perfoliatus.
in response to atrazine stress over several weeks.  Broader, ecological effects
of atrazine were also considered in terms of oxygen production and respiration
for whole experimental communities and for major autotrophic groups comprising
those communities.  The overall goal of this study was to characterize atrazine
effects on a spectrum of plant and community physiological and ecological
variables and to determine temporal patterns of these effects.  This study was
not intended to develop detailed dose-response relationships for atrazine and
£. perfoliatus; these are provided elsewhere (Kemp et al. 1982).


                             MATERIALS AND METHODS


Experimental Design

     We hypothesized that indications of stress in P_. perfoliatus would be
evident within 1-3 weeks following atrazine treatment in terms of several
physiological variables, including photosynthesis, respiration, pigmentation,
stem and leaf weight and meristic relations.  We also postulated that some
recovery would occur within weeks, and that other autotrophic groups would
exhibit responses inverse to those for SAV due to competitive interaction.
Since the objectives of this study were to discern the qualitative nature of
stress responses, we utilized relatively high concentrations of atrazine (ap-
proximating extreme in situ conditions) to increase the likelihood of observ-
able effects.

     Time-series observations of SAV responses required experimental popula-
tions large enough to allow weekly removal of several plants without markedly
disturbing the population (<5% removal).  Plant growth (shoot length and
density) was measured non-destructively at weekly Intervals to minimize
sampling effects; however, periodic removal of plants was also necessary to
establish length-weight relations.  The large experimental microcosms used
here provided sufficient plant material for the above measurements,and ade-
quate water volume and surface area were also available in these systems for
growth of associated micro-algal populations.                                             i~

     A split-plot, 2-way analysis of variance was employed for statistical
interpretation, with herbicide dosage as the main plot and time and replicates
as subplots.  In this factorial experiment significant differences (p<0.05)
among means were isolated with the Student-Newman-Kuels multiple range test
(Snedecor and Cochran 1967).

Experimental Systems

     Microcosm communities were established 1n 6 large glass aquaria (1.8 x
0.6 x 0.8m), each containing approximately 700 t filtered (5u) water and
12 cm depth of sediments from the Choptank River estuary, a tributary on the

                                       II-3

-------
eastern shore of Chesapeake Bay.  The sandy-silt sediment,  which was obtained
from an existing SAV bed, had sand/silt/clay proportions of 76/20/4 and con-
tained about 0.8% organic matter.  Microcosms were maintained in an air-
conditioned laboratory and water temperature remained relatively constant
throughout the study (22-26 C), with typical diel  ranges of 2 C.  An initial
salinity of 9.5°/oo increased about l°/oo from beginning to end of experiment
due to evaporation.  The microcosms were illuminated by banks of "cool  white"
fluorescent lights on a 14 h photoperiod.  Water in the microcosms was  contin-
uously circulated using centrifugal pumps (Little Giant Model 2 MD, rated flow
of 17 i min'1), with inlet and outlet diffusers separated at opposite
ends of the tanks.  Each microcosm was seeded initially with 75 individual
plants (shoots 6-50 cm, rhizomes 2-5 cm) of the submerged vascular plant,
Potamogeton perfoliatus. obtained in July 1978 from a population 1n the Chop-
tank River estuary.

Atrazine Addition and Analysis

     After establishing the microcosm communities, a period of 3 weeks  was
allowed for these systems to equilibrate prior to additon of atrazine.   In  an
effort to simulate actual runoff conditions, atrazine was added to microcosms
adsorbed onto sediments.  Technical grade (96.4% purity) atrazine was dissolved
in 100 mi. acetone and dried onto a 1.0 kg (dry weight) sediment sample.
Aquaria were randomly paired and selected for each of 2 treatments:  high
dosage (860 mg atrazine); low dosage (86 mg atrazine) and control  (sediment
addition only).  The herbicide and/or sediment dosages were Introduced  by
vigorously mixing into the microcosm water.  Water, sediment and suspended
solids (>0.3 y) were collected for analysis of atrazine residues at 3 times
during the experiment (1 d prior to treatment, and 36 h and 30 d following
treatment).  Atrazine residues were extracted in methanol (a Soxhlet apparatus
was used for sediments) and analyzed using gas chromatographic techniques
(Purkayastha and Cochrane 1973).  Analyses were performed at EnCase Labora-
tories (Greensboro, NC, USA) through the cooperation of C1ba-Geigy Corporation.

Photosynthesis and Respiration

     Photosynthesis and respiration of P. perfoliatus were measured in  each
microcosm at 6-8 d intervals beginning T wk prior to treatment.  Individual
plants were carefully removed from sediments with minimum disruption, and a
single shoot with about 5 cm length of rhizome (roots intact) was washed free
of sediments and epiphytes and placed Into a BOD bottle (300 ml) containing
ambient water from the respective microcosm.  Apparent photosynthesis (Pa)
and dark respiration (Rn) were measured as oxygen (Og) changes during incuba-
tions of 2 h in 6 replicate clear and opaque bottles suspended at the depth of
the plant canopy in each system  (Buesa 1975).  Epiphytes removed from individual
plants were incubated separately in light and cu.rk (L and D) BOD bottles, and
triplicate L and D bottles containing only water and suspended material were
incubated to estimate plankton production and respiration (used also as a
"blank" to correct plant and epiphyte measurements).  Metabolism of the benthic
community was determined using  L and D acrylic cylinders (800 mi) inserted
2 cm into sediments.   In all cases rates were estimated as changes in 03
measured at 2-3 h Intervals over a 9-12 h Incubation.  Bottles were suspended
in front of the circulation pump outlet diffuser to gently and continually  mix

                                       II-4

-------
them, and bottles were shaken by hand at 30min Intervals to insure that gas
diffusion gradients would be minimized (e.g., Conover 1967, Westlake 1978).

     At the end of each incubation, plants were placed in aluminum foil  dishes
and dried at 60 C to constant weight, ^standardize metabolic rates to  dry
weight of plant material (g 0- (g d.w.)  h ).  Plant and epiphyte
production and respiration rates were also extrapolated to g Oo m"^"1  by
multiplying specific rates by plant density and photoperiod, while water depth
(or height under chamber) was used to obtain areal  rates for plankton and
benthos.  Polarographic oxygen electrodes (Orbisphere Model 2709) were used to
measure 03 concentrations, and air calibration was  performed daily.  The
potential problems of lacunal storage of 02 in submerged vascular plants
(e.g., Zieman and Wetzel 1980) were not encountered in these well  circulated
systems (Westlake 1978), and periodic contemporaneous measurements of uC-uptake
(Lewis et al., 1982) yielded photosynthetic quotients (02 production:1"^
incorporation) of about 1.5 (Lewis, 1980).

     Community metabolism (Pa and Rn) was also monitored weekly by measuring
mean 03 concentrations in the open water of each microcosm at the beginning
and end of the pho'«.operiod.  Production and respiration of microcosm communi-
ties were taken as the increase and decrease, respectively, in 03 concentration
between successive dawn and dusk measurements (Odum, 1956).  Concentrations of
62 were maintained above 5.5 ppm throughout the experiment by periodic  aeration
using standard aquarium bubblers.  Corrections for  gas transfer across the
air-water interface were made by applying a diffusion coefficient to the ob-
served mean saturation deficit (Odum and Hoskins, 1958).  Coefficients  for 02
diffusion (averaging 0.15 g Q£ irr^^atm"1) were obtained by
monitoring the rate of 02 invasion into microcosms  which had been previously
deoxygenated by bubbling with nitrogen gas (Kemp et al., 1983).

Plant Abundance, Biomass and Morphometry

     Plant density (shoots nr2) and shoot length were determined each
week beginning 2 wk prior to treatment.  Quadrats in each microcosm were care-
fully marked, and all stems in each were counted and measured for length during
a given sampling.  Plant biomass In the aquaria was estimated with the following
non-destructive technique.  At the beginning and end of the experiment,  and
twice during the study, 20-30 plants representative of the entire spectrum of
size-classes, were selected from each experimental  treatment.  Plants were
measured individually for length and number of leaves and were dried to  constant
weight (at 60 C).  Biomass of each microcosm was estimated for each week using
linear regressions of weight versus length and applying measurements of  mean
and total length to obtain total plant weight.  At  the conclusion of the experi-
ment, 3 replicate cores (9 cm dia) wtre taken from  microcosm communities to
determine representative ratios of above-ground:below-ground plant biomass,
and total aboveground (shoot) material was harvested, dried and weighed.

Other Variables

     Chlorophyll a concentrations 1n P. oerfoliatus were measured at the end
of the experiment~by taking leaf sectTons (18 mm*)  in duplicate from the
uppermost leaves of 3 plants in microcosms.  Each leaf disc was extracted in

                                       II-5

-------
tl

               the dark at 5 C for 24 h in 90% acetone after tissue grinding (Vernon,  1960).
               Absorbance values were read at 665 and 745 nm on  a Beckman (Model  88) spec-
               trophotometer, and chlorophyll  concentrations were calculated using the tri-
               chromatic equations given by Strickland and Parsons (1972).   Plankton and
               benthic algal pigments were estimated according to the methods of  Strickland
               and Parsons (1972).

                    Photosynthetically active radiation (PAR)  was measured  weekly in each
               microcosm at mid-water depth using a LICOR (Model  LI-192SB)  cosine-corrected
               sensor.  Total suspended solids were also measured by filtering triplicate 250
               mi water samples through pre-weighted, dried Whatman GFC (1.2 y) filters,
               which were dried to constant weight at 90 C, cooled in dessicator  and reweighed
               (Strickland and Parsons, 1972).  The filtrate was  frozen for subsequent analyses
               of nitrate, nitrite and orthophosphate using methods of Strickland and  Parsons
               (1972), while ammonium was analyzed as in Grasshoff and Johannsen  (1972).
               Temperature and salinity were monitored at each measurement  of microcosm 02
               concentrations (weekly at dawn-dusk-dawn) using a  Beckman (Model RS5-3) induc-
               tion salinometer and thermistor.
                                            RESULTS AND DISCUSSION


               Atrazine Concentrations and Partitioning

                    Aqueous concentrations of atrazine were relatively  constant  throughout
               the experiment following treatments, with values  of about  0.13 ppm  for  "low"
               dose systems and 1.20 ppm for "high" dose.   The lower concentrations  correspond
               to the maximum concentrations of atrazine observed  in Chesapeake  Bay  (Kemp et
               al.» 1982).  Atrazine concentrations were similar at 3 and 30  d following
               treatment in both high and low dose experimental  systems,  and  atrazine  budgets
               developed for these microcosm, indicate about 15-20% loss  of the  herbicide
               during the experiment.  Assuming a first-order degradation process, this implies
               a half-life of about 3-4 mo,  which is 3-6 times that estimated by Jones et al.
               (1982) under different experimental conditions (exposed  to natural  ultraviolet
               radiation).  By far, the vast majority of the compound remained in  solution,
               with about 11, 6 and 0.03% of the total  mass contained in  sediments,  plants
               and suspended solids, respectively (Fig. 1).  The estimated plant uptake of
               atrazine was similar to that  reported by previous investigators (Funderbunk
               and Lawrence, 1963; Negi et al., 1963; Sutton et  al., 1969).

               Environmental Variables

                    During the course of the experiment there were several notable changes in
               the milieu of these microcosms.  Some of these changes are evident  in control
               systems, and probably reflect natural processes in  the development  of microcosm
               communities.  For example, total seston (suspended  particulates)  underwent a
               significant decrease in concentration from the beginning to end of  the  study
               but showed no response to treatment (Table 1).  While cylces of fluctuating
               concentrations of dissolved nutrients may be evident on  longer time-scales
               (Kemp et al., 1981), during the 7-8 wk of this study control and  low  dose
               microcosms experienced only small changes in nutrients (Table  1).  Under high

                                                     11-6

-------
       o
CO    «


CO    <
o    °
u
o
cc

o
                                               v'•:•••.  £co
                                               V-- •. »^   •.
                                                      • c
                                                      • 0>
                                                      ye

                                                      :-o
                                                                        ^
                                                                ~.   o> "~^

                                                                o   E  S?

                                                                T3  CD  |*5


                                                                £  fO  »0

                                                                O»  —  —
                                                                O      *-*

                                                                O
LJ

O   ,0
CD

 I

CO

CO
LJJ

Z


N

<    -


^    >
*-    <

<    O











. •


^
0>
o
^^
• .
•
.

• .
r
£

u
4
a

.

•

•
» «•
: 5

i >
> i*
> 2

i
£
X
1

.



<
} .
I> • .

°**o
10*
89
62

•
.
•


, -
•
•
• .
• .
' •
•
. /
. •
.
•

(0
c
o

•

• .
v • ~: ;
V.- ;
V- . •'
V': .'.


Cs..

\\::\-
v ••;./.
«k.:;;.-
• ••.•'-
• °c
• ^»
• ^r
• K*
. * *.
•./CO
»c
;!• E

.'CO

J.'.;
^**
••*
• •
o:1.
. •
?1
'•'i:.
** .-
'.::'.

                                                                o


                                                                o
                                                                    O

                                                                     <
).
u
)*
)°
]*
)'.'
)•:



•


•
)

M
•o
"o
(O
a.'
4O
3
(0
•
•
•
. •
o>"5
ES^
°S
/n O

•
.

. • .
' • .
.
. .'
• • •
•

c
o
OL
• .'
. • •
t ; • ; :
**:'".
^•'.•'.~
V-y :
>»'.:'..:

0
kV« •" .'•
*.'•'.' '••'
V." •'".'.'
v .'. ••
*".••'.'
• .**•",

:/6;
'i .•:.':••••.•.•
iments.
..^
•• «
v.w


• ..
^ *•*"
:/;"•
'•'•.t
• .
                                                                o


                                                                o
                                                                               ID ai
                                                                                 •a
                                                                               V) -f-
                                                                               E O
                                                                               «/) -t-

                                                                               OJP
                                                                               u s~
                                                                               O 0)
                                                                               i. f.
                                                                               u

                                                                              1*5
                                                                              C (I)
                                                                              •i- CO
                                                                              S- O
                                                                              « -o


                                                                              *-> 2
                                                                               O
                                                                              ..- 00
                                                                              N
                                                                              10 M-
                                                                              s- o
                                                                              4J
                                                                              10 •»->


                                                                              «- O)
                                                                              
                                                                              10
                                                                              •— o>
                                                                               i—

                                                                              e"o
                                                                              a*
                                                                               I O
                                                                               : ro
                                                                                 c
                                                                                 
                                                                              •^
                                                                              u.
                                  II-7

-------
TABLE 1.  Summary of selected environmental variables for microcosm communities
          treated with the herbicide, atrazine.*
Variable
Total Seston
(mg/Ji)
PAR
(pEin m"2s~ l)
NO;
( uM)
NO;
NHj
U)

Control
Pre Post
27.8 19.2
±0.2 ±5.1
122
±4
0.46 0.65
±0.11
0.22 0.12
±0.05
0.21 0.93
±0.15
0.17 0.14
±0.02
Treatment
Low
Pre Post
27.5 21.3
±1.9 ± 5.5
131
±10
0.35 1.13
±0.46
0.35 0.23
±0.10
2.32 2.07
±0.14
0.32 0.08
±0.02

High
Pre Post
29.4 20.3
±1.4 ±5.0
131
±8
0.41 11.2
±1.30
0.12 1.44
±0.41
0.30 40.21
±11.8
0.191 0.63
±0.21
*6iven are means ± standard deviation for 2 pre-treatment samplings  and for
 samplings in the final 2 weeks of post treatment observation.   For
 replication, n=4 for seston, nutrients and post-treatment PAR,  n=l  for
 pretreatment nutrients.
#Photosynthet1cally active radiation measured at mid depth of water  column
 (ca. 35 cm below water surface).
                                     II-8

-------
                herbicide treatments there were marked Increases 1n concentration for all
  rj            nutrient species measured (NO^, NO?, NHt, P0=4), resulting
                from plant excretion and decomposition (e.g., Strange, 1976).  Although light
                available for plant photosynthesis (PAR)  was below saturation values (Golds-
                borough, 1983), there were no significant differences among treatments (Table
                !)•

                Plant Photosynthesis and Respiration

                      Over the first 4 wk of measurements apparent photosynthesis (Pa) for
                control  plants (P. perfoliatusl exhibited a continual  Increase (0.41 to 1.11
                gOgifZd"* ), witn" a significant decline (to 0.67 gOom^d"1 )
                in che last week (F1g. 2a).  Values of Pa normalized per unit weight of
                plant did not change significantly through week 6 (mean = 2.79 mg 02 g dry
                wt."1^"1), with a sharp decline again in  week 7 (1.75).  Under low
                atrazine dosage Pq decreased significantly with treatment in week 4 followed
                by a steady and significant recovery (Fig. 2a).  While Pa for low dose micro-
                cosms remained below that for controls throughout post-treatment  observations,
                by week 7 it was statistically indistinguishable from pre-treatment values
                (week 3).  Furthermore, Pa per unit weight was not significantly  different
                for control and low dosed plants 1n the final week of observation.   At high
                herbicide dosage Pa remaine-l negative throughout.  This recovery  of Pa at
                low dose is remarkable in view of the fact that atrazine concentrations re-
                mained relatively constant throughout.  To our knowledge, there have been  no
                previous reports of this behavior, and the mechanism whereby 1t might occur  1s
                uncertain (Kemp et al.  1982).

                     The mean losses of Pa  over the 4 wk following treatment for plants
                exposed in this study to Tow (0.13 ppm) and high (1.2 ppm) dosage of atrazine
                were 100% and about 50% compared to control plants, respectively.  These results
                are comparable to previous reports on effects of atrazine and related herbicides
                on SAV photosynthesis.  In their Investigation of simazine effects on 3 fresh-
                water vascular plants, Sutton et al. (1969) observed a maximum of 40% reduction
/ -              1n 0? concentration (rather than rate of production) after 4 d at 0.12 ppm.
                Similarly, Bielecki and Skrabka (1976) reported about 35% loss of carbon diox-
                ide fixation by the macrophyte, Spirodela polyrrhiza, exposed to  0.2 ppm sima-
                zine for 24 h, while Fowler (1977) observed about 50% decrease 1n Pa of Myrio-
                phyllum sp. at 0.5 ppm of another ^-triazine (DPX 2674) after 10  d.  In a
                study of longer duration, Correll and Wu (1982) reported Pa for 4 SAV species
                (9_. pectlnatus, Z. palustris, _V.  americana and Zostera marina) after 3-6  wk
                exposures to 0.075 and 0.65 ppm atrazine.  .*t the lower concentration effects
 -r              on Pa ranged from a 48% stimulation to a 39% depression compared  to controls,
                while the higher concentration resulted in 71-100% loss of Pa. There was  no
  '              evidence of metabolic recovery 1n any of the short-term experiments, and data
  \              provided 1n the longer study of Correll and Wu (1982) was Insufficient for any
 ,\              such Interpretation.

                     The pronounced effects of atrazine and other £-tr1az1ne herbicides on SAV
                photosynthesis 1s to be expected in view of the fact that the mode of toxic
                action Involves H111 Reaction blockage on the reducing side of photosystem II
                (Souza-Machado et al., 1978).  However, the anticipated effects on plant respir-
                ation are, perhaps, less obvious.  Dark respiration (Rn) was unaffected by

                                                       II-9

-------
                         POTAMOGETON  PERFOLIATUS
                   LO
                
                O
                m
                 UJ 0.5
                 <  +1
                 Or
                    -I
                         .•.'.'.'a) PRODUCTION,  P0
.'•'•   • .|b)RESPIRATION,Rn
                       • '. •'. .[c) RATIO, P0: Rn
                          34567

                              WEEK  OF  EXPERIMENT
F1g.  2.  Metabolic  responses of Potamogeton perfollatus to low  (L = 0.13 ppm)
        and high (H « 1.20 ppm) dosages of atrazlne as compared to controls
        1C), In terms of a) apparent 02 production (Pa), b) dark respiration
        (Rn;, and  c) the ratio P^Rn-  Points and vertical bars Indicate
        mean ± standard error.  Shaded portion of figure represents pretreat-
        ment period.

                                  11-10

-------
low atrazine treatment, remaining relatively constant throughout the study.
Under high dosage Rp showed a steady and significant decline through week 7
(Fig. 2b).  The ratio of Pa to dark respiration provides an index of the
metabolic energy balance for the plant population (e.g., Drew 1979).  Since Rn
changed very little, patterns of Pa:Rn (on a diel  basis) were similar to the
temporal and treatment trends for Pa alone (Fig.  2c).  Thus, for the recovery
period (weeks 5-7) low dosed plants were able to undergo minimal net growth,
as Pa:Rn remained near 1.0 for weeks 5-6, increasing somewhat t.o 1.6 in week
/ •

Plant Abundance and Selected Morphological  and Physiological  Variables

     Responses of plant photosynthesis and respiration to herbicide treatment
were eventually manifested as changes in abundance and biomass of these SAV
populations, as well as changes in plant morphology.  Density (#/m2 ) of
shoots (stems with leaves) increased in a nearly linear manner between v.eeks 2
and 6 for both control and low dose microcosms, with no change occurring from
week 6 to 7 (Fig. 3a).  There was no significant difference between these two
treatments.  At high atrazine dosage (1.2 ppm) a significant decline was ob-
served from week 5 to 7, with final shoot density only about 20% of control
values (Table 2).  While changes in shoot numbers provided only weak indication
of herbicide response here for P. perfoliatus, Correll and Wu (1982) found
shoot mortality to be a reasonable measure of stress for V. americana treated
with relatively low concentrations of atrazine (0.012 ppmj.

     At the beginning and throughout the experiment, representative samples of
plant shoots were obtained from microcosms to establish length (L,  cm) versus
weight (B, gm) relationships, so that stem density data could be translated
into biomass.  Significant linear regressions were obtained throughout; however,
significant changes in slope (p<0.05, analysis of covariance) were  also ob-
served.  For pre-treatment weeks the regression for plants in all microcosms
was

               B * 0.0032 L + 0.0003                    (r2 « 0.89).

The slope of this relation for the pooled control  system data for weeks 5 and
7 (F1£. 4) had increased sign;Ticantly (indicating that a unit length of plant
was heavier). The slope of the L-B regression for low dosed plants  was signifi-
cantly less than that for controls in the final 3 wk (F1g.  4), and the mean
L:B ratio was significantly Increased with treatment (2.4, 4.3 and  6.1, re-
spectively, Table 2).

     Applying the regressions 1n Fig. 4 to the stem density data 1n Fig. 3a,
we obtain ar estimate of plant biomass for f_. perfoliatus populations 1n experi-
mental microcosms (F1g. 3b).  In control microcosms through week 6, net accre-
tion of plant biomass proceeded at a consistent rate (0.7 g d.w. m^d'1)
which is identical to the mean value of post-treatment Pa (assuming that g
Og:g d.w. * 1.0 as in Odum, 1971).  Significant effects of atrazine dosage
on total plant biomass were detectable 2 wk following dosage for both treat-
ments (F1g. 3b).  Thus, there was a 1 wk lag between the Initial Indication of
metabolic decline and 1nc1p1ertly observable losses of plant material.  By
week 7 biomass 1n the high dose microcosms had been virtually eliminated, as a

                                     11-11

-------
15,
                                  30
                              CO  20
                              CO
                              o   »o
                              CD
                                 400
*,
*~



W200
                                 300
                               E
                               o

                               I
                               H
                               O
     10
1
oJPLANT
High Dose
• Low Dos0v \
v •*•
ControU j\

1^*1
tl?l


'_•


•';
>>
'.",
'•
'"•




BIOMASS

-



*

•
•
.
[T
1





;?
1
^]
0
y-
1






fc
\ •

'•

'•'•

•*•
•••





B
1
i



^



•".
•

*
.••
'•
".•




&
*>






m


*•



>.




*|

b) SHOOT
.

,&!

"fill
1 I * - *s!

1
•

;'/

1

s
jf;
f
1



1
DENSITY
-
^




1
\
5

.
.•
X-

• *

'/;
i
1
*v
j-l
!»
^
p.






.'•

-'
, *
J
:
i.
:•.

..*

Ji
1
s;
1
. *
I
|
YJ






fl

"•



••
i
•'.•



* .
\


$
7*
J;
t^
>>,
is
1





rX-
.
;•
•'

•
• •
J]
'•:
•^
>
_. .








1
i
c)MEAN SHOOT LENGTH
-
.




.



»





.-



i









1
'rt
1
!^
t?
•i>
1
r-5
K*
^










*









T.
1

• .•
.'.'.
**

'.•'.



:S
'i
1
'?'
si-

1


1









f
X-
*
*



•



1


.*
. •

'•



S]
Vt
J*
**'.,
,'**

;/j
iff

*!•'.

?<










rT-
'Ai



•



1





'.-'


• '
£
ic
1
i§
Cy

i

•fj











^







a





."•


* *
1
k:
1
8
&'

1


1










z-

"





1
." •
'.'









S
1
f-3

1


?j








                                          234567
                                                WEEK  OF  EXPERIMENT

                   F1g. 3.   Temporal  responses  of a) plant blomass, b) shoot density, and c)  mean
                            shoot  length  for  Potamogeton perfollatus exposed to low (0.13 pom)
                            and high  (1.20  ppm) doses of atrazlne compared to controls.   Vertical
                            bar Indicate  ±  standard error, and dashed lines with arrows  represents
                            the Initiation  of treatment.
                                                     11-12

-------
TABLE 2.  Summary of selected physiological  and morphological  characteristics
treated with the Herbicide, atrazine.*
Treatment
Measurement
Shoot Biomass (Ba)
(g d.w. m'2)
Rhizobial Biomass (Bb)
(g d.w. m'2)
Ratio, Bh:B.
u a
Shoot Density
(no. m'2)
Mean length of Shoots
(cm plant'1)
Length: Weight Ratio**
(m g-1)
Chlorophyll a
(mg m-2(leafT)
Internodal Length
(cm)
Control
44.3 ±
40.0 ±
0.93 ±
458 t
15.5 ±
2.4 ±
28 ±
1.37 ±
17.1
12.9
0.22
13
2.9
0.3
8
0.06
Low
24.3 ±
20.0 ±
0.94 ±
477 ±
20.2 ±
4.3 ±
158 ±
1.11 ±

8.7
8.6
0.54
41
2.5
0.6
16
0.09
High
1.4 ± 0.1
-f
_t
107 ± 6
19.6 ± 1.9
6.1 ± 2.3
114 ± 5
1.89 i 0.82
 *Data are from samples taken in last week of experiment,  except  in the case of
  Shoot Lengths which are averaged over the last 3 wk.  Given are mean values ±
  standard deviation, where n = 12 for Chlorophyll, n - 6 for Biomass and Shoot
  Length, and n=20-30 for LengthrWeight and Internodal Length.
**Rat1o for individual shoots from weeks 4, 7 and 8.
 tRhizobial biomass is not reported for high dose systems  because plant material
  appeared to be dead and decaying.
                                     11-13

-------
    0.15
 o»
- 0.10
LU
(T
Q

H

<  0.05

_l
£L
                     LENGTH  VERSUS   WEIGHT
                      (Potomogeton  perf o\ iotus)
          10
                               Low  Dose
                              0.0027X-0.0059

                              (r2«0.75)
                                      High Dose
                                    Y*0.0039X-0.0438
                                     (r2 *0.76)
15       20       25       30

     STEM   LENGTH.cm
35
     F1g. 4.  Dry weight of plant stems versus stem length for Potamogeton perfollatus
            1n microcosms treated with atrazlne at low  (solid circles, 0.13 ppm)
            and high (circles with dots, 1.20 ppm) doses and for untreated plants
            1n control (open circle) microcosms.  Plant samples taken 2 and 4 wk
            after Initial treatment.
                                  11-14

-------
r i
t
* t
               consequence of negative values  for Pa  over the  previous  4  wk.   This  4 wk
               lapse between full  metabolic  and  biomass  response  is  consistent with the
               respiratory turn-over time (T = biomass:respiration)  just  prior to treatment,
               where biomass was at about 15 g m~2 and respiration was  0.4 g  m~2d~l  (T =
               37 d).  Plant biomass in low  dose microcosms  remained constant after week  5 at
               low dose, suggesting again a  recovery  of  metabolic balance (c.f., Fig. 2).
               while both herbicide dosages  caused significant  reduction  in total standing
               stocks, the ratio of below-ground to above-ground  biomass  was  unaffected by
               atrazine treatment  (Table 2).  Other investigators (Brooker and Edwards, 1973;
               Strange, 1976) have reported  marked losses of macrophytic  biomass after prolonged
               (30 d) exposures to high concentrations of herbicides (>0.5 ppm); however,  the
               temporal correspondence between reductions in Pa and  biomass have not been
               examined previously.

                    The size-distribution of £.  perfoliatus  shoots was  unaffected by
               herbicide treatment in our experiments, with  plants in the 6-10 cm class
               dominating throughout the experiment at all atrazine  levels (Cunningham,
               1979).  Average length of individual shoots,  however,  was  significantly
               increased for both  low and high treatments within  2 wk following atrazine
               amendments (Fig. 3c, Table 2).   In contrast,  Forney and  Davis  (1981)  re-
               ported the opposite response, with 26-51% reduction in growth  of shoots
               for £. perfoliatus  exposed to 0.1-0.32 ppm atrazine for  4  wk.   This differ-
               ence may be due to  the relatively low  light intensities  (75 wEin m^s"1)
               under which their experiments were conducted.   The elongation  effect  which
               we observed is consistent with  the change in  lengthrweight ratio discussed
               above.

                    There is some  evidence that  the photosynthetic stress imposed by
               atrazine could elicit this increase in stem length, where  for  instance,
               Young and Evans (1978) found  that leaf elongation  and etiolation were an
               indication of atrazine stress and tolerance in  perennial grasses.  Stem
               and leaf elongation have been suggested by numerous Investigators as  indices
               of shade adaptation in submerged  vascular plants.  Spence  (1975) has  re-
               viewed the substantial literature documenting this effect  for  various
               Potamogetonaceae under reduced  Intensity  and  quality  of  light  in field and
               laboratory experiments.  While  the control mechanism  for this  response of
               submerged aquatic plants to decreased  light is  uncertain,  it is of physio-
               logical Interest that atrazine  exposure produced a similar effect.

                    Another Index  of shade response by Pptamogeton spp. and other SAV
               species with distinct leaf and  stem structure (Spence 1975) 1s "Internodal
               length" (I.e., the  mean distance  along the stem between  leaf nodes).   For
               atrazine treated £. perfoliatus,  the effect on  Internodal  length was  less
               clear than for mean stem length and for L:B ratio. Although a significant
               (38%) Increase In this index  was  evident  for  high  dose plants  compared to
               controls, there was a slight  decrease  for low atrazine treated plants.  No
               pre-treatment differences were  evident among  microcosm plant populations.
               For this analysis,  special effort was  made to select  plants representative*
               of size distributions of respective experimental populations,  since  a
               significant relationship was  found between Internodal  length (I, cm)  and
               total stem length (L, cm),


                                                    11-15

-------
                   I = 0.025 L + 1.002                 (r2 =0.83).

Thus, this index (internodal length) is subject to the vagueries of visually
determined "representative samples".

     Concentrations of chlorophyll a in£. perfoliatus leaves also exhibited
a reaction to atrazine treatment whTch is analogous to that expected under
extreme light limitaticn, where significantly elevated levels of chlorophyll
£ were measured for herbicide treated plants (Table 2).  An inverse rela-
tionship between ambient light and chlorophyll £ has been previously
reported for various submerged plants (e.g., Bowes et al., 1977; Wiginton
and McMillan, 1979; Drew, 1979).

     Several Investigators have observed this "greening effect" of increased
chlorophyll content under atrazine stress for algae and terrestrial plants;
however, we have found no previous reference to effects of atrazine on pigment
composition in aquatic vascular plants (Ebert and Dumford, 1976).  The physio-
logical mechanisms whereby these reactions (elongation and increased chloro-
phyll) to atrazine stress might be effected are presently unclear (Ebert and
Dumford, 1976; Moreland, 1980), as is the relationship (if any) to modes of
shade adaptation.

Community Photosynthesis and Respiration

     Apparent production (Pa), night respiration (Rn) and the metabolic
ratio (Pa-'Rn) f°r *ne integrated microcosm communities are provided in Fig.
5.  The general response of these metabolic variables was similar to that for
P_. perfoliatus alone (Fig. 2), with significant effects on Pa at low and high
dose and recovery of low dose Pa toward the end of the experiment.  A distinct
difference, however, was the pronounced decrease in community Pa following
treatment, even for control microcosms.  This may be partially attributable to
the addition of sediments as a vehicle control; however, seston levels in
post-treatment weeks actually decreased for all microcosms steadily over the
course of the experiment (Table 1 and Cunningham, 1979).  Probably, a more
reasonable explanation lies in the sharp reduction in dissolved nutrients and
associated phytoplankton populations.  In control microcosms by week 4, concen-
trations of NH^, NOj and P0=4 decreased to 0.10, 0.15 and 0.14 uM
respectively, and between weeks 3 and 4 algal Pa decreased by more than 50%.
However, without untreated controls this remains an open question.

     The response of community R,, to treatment (F1g. 5b) was almost identical
to that for £. perfoliatus respiration, with significant effects evident only
at high dosage.  Community Pa:Rn ratios also followed similar pattern to
those for SAV (Fig. 5c), although ratios were generally below 1.0 (even for
controls) indicating a net consumption of previously stored (probably sediment)
detrltal organic matter (Odum, 1956).  Presumably, the lack of a strong response
in community Rn to herbicide treatment may further reflect the relative im-
portance of detrltal (rather than direct grazing) trophic chains 1n these
microcosms, since the loss of Pa did not produce a comparable reduction in
Rn.

     Few studies have reported effects of atrazine treatment on community Pa

                                     11-16

-------
                                 COMMUNITY
                            a) PRODUCTION.?.
                             WRESPIRATION.R
              o
              -   o
              a:  -I
                 -2
                                    10, R.:R.
                       34567

                            WEEK OF  EXPERIMENT

F1g.  5.  Metabolic responses of microcosm communities containing Potamogeton
        perfollatus to low (L = 0.13 ppm) and high (H = 1.20 ppm) dosages of
        atrazlne as compared to controls (C), 1n terms of a)  apparent 0?
        production (Pa),  b) night respiration (Rn), and c) the ratio Pa:Rn.
        Points and vertical bars Indicate mean ± range and shaded portion of
        figure represents pretreatment  period.

                                  11-17

-------
and Rn.  However, two investigations have measured effects of other herbicides
on community oxygen metabolism.  Brooker and Edwards (1973 a,b) found that
community metabolism in a reservoir treated with paraquat to control  the SAV,
Myriophyllum sp., exhibited an initial loss of Pa, followed by a return to
pre-treatment levels within 1-2 wk.  However, most of this recovery was attrib-
uted to increased phytoplankton production, with an eventual recolonization
after 1-2 mo by macrophytic algae (Chara globularis) rather than SAV.  Strange
(1976) reported virtual loss of community Pa (as CCfc uptake) in microcosms
containing the SAV, Egeria densa, within 30 d following treatment with 0.5 ppm
mixed herbicides (diquat and endothall).  Daily measurements of Pa indicated
little or no recovery after treatment.  Here as in our study, the effect on
respiration was less marked, so that Pa:Rn followed a pattern similar to
that for Pa.  In contrast, Brooker and Edwards (1973 a,b) observed losses of
Pa and Rn to be similar, with little effect on Pa:Rn, probably because of
the mechanism of paraquat action involving formation of generally reactive
free-radicals.

     We had hypothesized that the relative balance among major autotrophic
groups (SAV, epiphytes, phytoplankton, and benthic microalgae) would  shift in
response to differential effects of herbicide treatment and resulting changes
in competitive advantage for light and nutrients.  However, our observations
failed to corroborate this hypothesis.  The relative contributions of these 4
groups *.o total  gross 0? production (operationally defined as Pa plus Rn)
are given in Fig. 6.   In control and low dose microcosms P. £erfoliatu£ contri-
butions increased steadily from 41 and 44% in week 3 to 77 and 74% in week 7,
while the phytoplankton proportions decreased from 41 and 43% to 8 and 7% in
control and low dose, respectively.  Phytoplankton and epiphytes did comprise
a slightly greater portion of total production under high dosage; however, the
rates were all so low that it 1s hard to ascribe any particular Importance to
this.  The absence of herbicide induced changes in the balance among  autotrophs
is probably the consequence of two major factors.  First, herbicide toxicity
may be similar for all pnotosynthetic organisms in the microcosm communities.
Secondly, the particularly low concentrations of dissolved nutrients  in micro-
cosm water columns may have conferred overwhelming advantage to P. perfoliatus
which was able to draw on higher concentrations available in sedTment pore-
waters.

Final Comments
     This experiment has demonstrated the ability of a submerged vascular
plant, P_. perfoliatus, to recover photosynthetic 03 production within  several
weeks following continuous treatment with moderately high concentrations  of
the herbicide atrazine.  Such observations emphasize the importance of examining
temporal patterns over weeks or months for understanding the effects of herbi-
cides on SAV.  More subtle manifestations of effects such as changes in repro-
ductive success would require even longer term experiments.   We report here
several other indices of atrazine effects such as stem elongation,  reduced
stem weight per unit length, increased internodal  length and increased levels
of chlorophyll £, all of which are parallel to tht previously observed response
of this and related species to reduced light intensities.  Such indices could
be used routinely as measures of atrazine stress;  however, care must be taken
not to confuse plant responses to herbicide and shade.  The effects of atrazine

                                   11-18

-------
         \ t-
TABLE 3.  Coefficients cf variation (CV)  for apparent  production in microcosm
          communities treated with the herbicide,  atrazi.ne.*
Treatment
Component
CONTROL
Macrophyte
Plankton
LOU DOSE
Macrophyte
Plankton
HIGH DOSE
Macrophyte
Plankton
Week of Experiment
3t
40. 2*
37.8t
58. 6f
27.0*
26.3t
19.3f
4
21.3
12.4
50.3
34.2
174.4
171.4
5
20.8
65.4
38.3
122.8
115.6
85.3
6
40.5
64.3
23.4
64.3
136.1
137.1
7
17.2
95.4
40.2
111.9
80.0
166.7
*CV is defined as (standard deviation/mean)  x 100.
tWeek 3 is prior to treatment; weeks 4-7 are post-treatment.
                                     11-19

-------
0SAV
BSPHYTO PLANKTON
                                           BENTHIC ALGAE
                                           EPIPHYTES
                               4567

                            WEEK OF  EXPERIMENT
Fig.  6.  Relative contributions (compared to  control microcosm  in week 6) of
        Potamogeton perfpii atus (SAV),  phytoplankton, benthic  microalgae and
        epiphytic algae to total gross  0? production (defined  as Pa plus
        Rn) of experimental communities in a) control, b)  low  (0.13 ppm)
        atrazine dosed, and c) high (1.20 ppm) atrazine dosed  microcosms.
        Treatment was  initiated 1 d prior to week 4 measurements.

                                 11-20

-------
on competitive interactions among primary producers in SAV communities may
not involve compensatory shifts in relative balance.  However,  this question
needs to be addressed in experimental systems where both sediment and aqueous
nutrients are continuously available.  By characterizing a broad range of
atrazine effects on P_. perfoliatus physiology and morphology, this study has
provided a basis for further examination of SAV responses to herbicides at
lower concentrations typical of those occurring frequently in estuaries such
as Chesapeake Bay.
                                      11-21

-------
                                   REFERENCES


 1.  Adams, S.M.  (1976).  The ecology of eelgrass, Zostera marina (L.),
     fish communities.  I.  Structural analysis.  J. exp. mar. Biol.  Ecol.
     22:269-291.

 2.  Bayley, S., Stotts, V.D., Springer, P.P., Steenis, J.  (1978).  Changes
     In submerged aquatic macrophyte populations at the head of the Chesapeake
     Bay, 1958-1975.  Estuaries 1:74-75.

 3.  Bielecki, K., Skrabka, H.  (1976).  Effect of some herbicides on photo-
     synthesis of Splrodela polyrhiza (Lemnaceae).  Acta Agrobot. 29: 59-68.

 4.  Bowes, G., Van, T.K., Garrard, L.A., Haller, W.T.  (1977).  Adaptation to
     low light levels by Hydrilla.  J. Aquat. Plant Mgmt. 15:32-35.

 5.  Brooker, M.P., Edwards, R.W.   (1973a).  Effects of the herbicide paraquat
     on the ecology of a reservoir.  I.  Botanical and chemical aspects.
     Freshwat. Biol. 3:157-75.

 6.  Brooker, M.P., Edwards, R.W.   (1973b).  Effects of the herbicide paraquat
     on ecology of a reservoir.  II.  Community metabolism.  Freshwat.  Biol.
     3:383-89.

 7.  Brooker, M.P., Edwards, R.W.   (1975).  Aquatic herbicides and the control
     of water weeds.  Water Res. 9:1-15.

 8.  Buesa, R.J.  (1975).  Population biomass and metabolic rates of  marine
     angiosperms on the northwestern Cuban shelf.  Aquatic Bot. 1:11-23.

 9.  Coen, L.D., Heck, K.L., Abele, L.G.  (1981).  Experiments on competition
     and predation among shrimps of seagrass meadows.  Ecology 62:1484-1493.

10.  Conover, J.T.  (1968).  The Importance of natural diffusion gradients 1n
     the metabolism of benthic marine plants.  Botanica Mar. 11:1-9.

11.  Cornell, D.L., Wu, T.L.  (1982).  Atrazine toxidty to submersed vascu'ar
     plants in simulated estuarine microcosms.  Aquatic Bot. (In press).

12.  Cunningham, J.J.  (1979).  Responses of microcosm communities containing
     submerged aquatic vegetation  to herbicide stress.  MS Thesis, Univ. of
     Maryland, College Park.  66 pp.
                                     11-22

-------
t!
               13.   Den  Hartog,  C.,  Polderman,  P.J.G.   (1975).  Changes  in the  seagrass  popu-
                    lations  of the Dutch  Waddenzee.  Aquatic  Bot.  1:141-147.

               14.   Drew,  E.A.   (1979).   Physiological  aspects of  primary production  in  sea-
                    grasses.  Aquatic  Bot.  7:139-50.

               15.   Ebert,  E., Dumford, S.W.   (1976).   Effects of  triazine herbicides  on the
                    physiology of plants.   Residue  Reviews 65:1-103.

               16.   Fonseca,  M.S., Fisher,  J.S.,  Zieman, J.C., Thayer, 6.W.   (1982).   Influence
                    of the  seagrass, Zostera marina  L, on current flow.  Estuar. Coast.
                    Shelf  Sci. 15:351- 364.

               17.   Forney,  D.R., Davis,  D.E.   (1981).  Effects of low concentrations  if
                    herbicides on submersed aquatic  plants.   Weed  Sci. 29:677-85.

               18.   Fowler,  M.C.  (1977).   Laboratory trials  of a  new triazine  herbicide (DPX
                    3674)  on  various aquatic species of macrophytes and algae.  Weed Res. 17,
                    191-195.

               19.   Funderbunk,  H.H.,  Lawrence, J.M.  (1963).  Absorption and translocation of
                    radioactive  herbicides  in submersed and emersed aquatic weeds.  Weed Res.
                    3:304-11.

               20.   Ginsburg, R.N.,  Lowenstam,  H.A.  (1958),  The  influence of marine  bottom
                    communities  on the depositional  environment of sediment.  J. Geol.  66:
                    310-318.

               21.   Goldsborough, W.J. 1983.   Light-shade adaptation for the submerged vascu-
                    lar  plant, Potamogeton  perfoliatus.  MS Thesis, Univ. Maryland, College
                    Park (In  press).

               22.   Grasshoff, K., Johannsen,  H.   (1972).  A  new sensitive and  direct  method
                    for  the automatic  determination  of  ammonia in seawater.  J. Cons.  lut.
                    Explor.  Mer. 34:516-521.

               23.   Harlin,  M.M., Thome-Miller,  B., Boothroyd, J.C.  (1982).   Seagrass-sediment
                    dynamics  of  a flood-tidal delta  in  Rhode  Island (USA).  Aquatic Bot.
                    14:127-138.

               24.   Howard-Williams, C.   (1981).   Studies on  the ability of a Potamogeton
                    pectinatus community  to remove dissolved  nitrogen and phosphorus compounds
                    from lake water.   J.  Appl.  Ecol. 18:619-637.

               25.   Hzumi,  H.,  Hattori,  A., McRoy,  C.P.  (1980).  Nitrate and  nitrite in the
                    interstitial waters of  eelgrass  beds in relation to the rhizosphere.  J.
                    exp. mar. Biol.  Ecol. 47:191-201.

               26.   Jones,  T.W., Kemp, W.M., Stevenson, J.C., Means, J.C.  (1982).  Degrada-
                    tion of atrazine in estuarine water/sediment systems and selected  soils.
                    J. Environ.  Qual.   (In  press).


                                                    11-23

-------
27.  Jupp, B.P., Spence, D.H.N.  (1977).  Limitations on macrophytes in a
     eutrophic lake, Loch Leven.  I.  Effects of phytoplankton.  J.  Ecol.
     65:175-186.

28.  Kemp, W.M., Lewis, M.R., Cunningham, J.J., Stevenson, J.C., Boynton, W.R.
     (1981).  Microcosms, macrophytes, and hierarchies:  Environmental  research
     in the Chesapeake Bay.  Microcosm Research in Ecology.  (Ed. by J. Giesy).
     ERDA Conf. Ser., NTIS, Springfield, Virginia.  911-35.

29.  Kemp, W.M., Means, J.C., Jones, T.W., Stevenson, J.C.  (1982).  Herbicides
     in Chesapeake Bay and their effects on submerged aquatic vegetation.
     Univ. Maryland Centr Environ.  Est. Stud. Ref. No. 1295-HPEL.  Horn Pt.
     Env. Labs, Cambridge, MD.  82 pp.

30.  Kenworthy, W.J., Zieman, J.C., Thayer, G.W.  (1982).  Evidence for the
     influence of seagrasses on the benthic nitrogen cycle in a coastal plain
     estuary near Beaufort, North Carolina (USA) Oecologia 54:152-158.

31.  Kiorboe, T.  (1980).  Distribution and production of submerged macrophytes
     in Tipper Grund (Ringkobing Fjord, Denmark), and the impact of waterfowl-
     grazing.  J. Appl. Ecol. 17:657-687.

32.  Kullberg, R.G.  (1974).  Distribution of aquatic macrophytes related to
     paper mill effluents in a Southern Michigan stream.  Amer. Midi. Nat.
     91:271-281.

33.  Lewis, M.R.  (1980).  An investigation of some homostatic properties of
     model ecosystems In terms of community metabolism and component interac-
     tions.  MS Thesis, Univ. of Maryland, College Park.  150 pp.

34.  Lewis, M.R., Kemp, W.M., Cunningham, J.J., Stevenson, J.C.  (1982).  A
     rapid technique for preparation of aquatic macrophyte samples for measuring
     lkC incorporation.  Aquatic Bot.  (In press).

35.  Mann, K.H.  (1982).  Ecology of coastal waters:  a system approach.  Univ.
     of California Press, Berkeley.

36.  Marsh, A.G.  (1973).  The Zostera epifauna! community in the York  River,
     Virginia.  Chesapeake Sci.14:87-97.

37.  McRoy, C.P., Barsdate, R.J., Nebert, M.  (1972).  Phosphorus cycling in
     an eel grass (Zostera marina L.) ecosystem.  Limnol. Oceanogr.  17:58-67.

38.  Moreland, D.E.  (1980).  Mechanisms of action of herbicides.  Ann. Rev.
     Plant Physiol. 31:597-638.

39.  Nagel, J.S.  (1968).  Distribution of the epibiota of macrobenthic plants.
     Contr. Mar. Sci., Univ. Texas. 13:105-144.

40.  Negi, N.S., Funderbunk, H.H.,  Davis, D.E.  (1964).  Metabolism of atrazine
     by susceptible and resistant plants.  Weeds 12:52-57.


                                     11-24

-------
•I  f
                 41.   Nelson,  W.G.   (1979).   An analysis  of  structural  pattern  In  an eelgrass
                      (Zostara marina  L.).   amphipod  community.   J. exp.  mar. Biol.  Ecol.
                      39:231-263:

                 42.   Odum,  E.P.   (1971).   Fundamentals of Ecology.   W.B. Saunders,  Philadelphia.
                      574 pp.

                 43.   Odum,  H.T.   (1956).   Primary  production  1n  flowing  waters.   Limnol.
                      Qceanogr.  1:102-17.

                 44.   Odum,  H.T.,  Hoskin, C.M.  (1958).   Comparative  studies on the  metabolism
                      of marine  waters.  Publ.  Inst.  Mar. Sci. (Texas)  8:23-55.

                 45.   Orth,  R.J.   (1975).   Destruction of eelgrass, Zpstera marina,  by  the
                      cownose  ray  Rhinoptera  bonasus. in  the Chesapeake Bay.CResapeake Sci.
                      16:205-208.

                 46.   Orth,  R.J.,  Moore, K.A.   (1981).  Submerged aquatic vegetation of the
                      Chesapeake Bay:   Past, present  and  future.  Trans.  N. Amer.  Wildl. Nat.
                      Res. Corf. 46, 271-83.

                 47.   Peres, J., Picard, J.   (1975).  Causes de la rarefaction  et  de la
                      disparition  des  herbiers  de Posidonia  oceanica  sur  les cotes Francaises
                      de La  Medi terra ne'e.   Aquatic  Bot. 1:133-39.

                 48.   Phillips,  G.L.,  Eminson,  D.F.,  Moss, B.  1978.  A mechanism  to account  for
                      macrophyte decline in progressively enriched freshwaters. Aquatic.  Bot.
                      4:103-126.

                 49.   Purkayastha,  R., Cochrane, U.P.  (1973).  A method  for analysis of atrazine.
                      J. Agric.  Food Chem.  2:93-102.

                 50.   Robson,  T.O., Barrett,  P.R.F.  (1977).  Review  of effects of aquatic
                      herbicides.   In  Perrlng,  F.H.,  Mellanby, K. (eds.).  Ecological Effects
                      of Pesticides.   Linnean Soc.  Symp.  Ser., No. 5, Academic  Press, New  York,
                      pp. 111-31.

                 51.   Sand-Jensen,  K., Sffndergaard, M.  (1981).   Phytoplankton  and epiphyte
                      development  and  their shading effect on  submerged macrophytes  in  lakes  of
                      different  nutrient status.   Int. Rev.  ges.  fydrobiol.  66:529-552.

                 52.   Snedecor,  G.W.,  Cochran,  W.G.  (1967).  Statistical Methods.  Iowa State
                      Univ.  Press.   Ames.

                 53.   Souza-Machado, V., Armtzen, C.J., Bandeen,  J.D.,  Stephenson, G.R. (1978).
                      Comparative  trlazlne  effects  upon system II photo-chemistry  in chloroplasts
                      of two common lambsquarters (Chenopodium album) blotypes. Weed Sc1.
                      26:318-22.
                                                      11-25

-------
54.  Spence, D.H.N.  (1975).  Light and plant response in freshwater.  In:
     Evans, G.C., Bainbndge, R., Rockham (eds.).  Light as an Ecological
     Factor.  Blackwell Sci. Publ., Oxford,  pp. 93-133.

55.  Strange, R.J.  (1976).  Nutrient release and community metabolism fol-
     lowing application of herbicide to macrophytes in microcosms.  J. Appl.
     Ecolo. 13:889-897.

56.  Strange, R.M., Schreck, C.B.  (1976).  Response of aerobic community
     metabolism to chemical treatment of aquatic macrophytes.  J. Aquatic
     Plant Mange.  14:45-50.

57.  Strickland, J.D.H., Parsons, T.R.  (1972).  A Practical Handbook of Sea-
     water Analysis.  Fish Res. Bd., Canada  Bull. 167.  Ottawa.  310 pp.

58.  Sutton, D.C., Durham, D.A., Bingham, S.W., Foy, C.L.  (1969).  Influence
     of simazine on apparent photosynthesis  of aquatic plants and herbicide
     residue removal from water.  Weed Sci.  17:56-59.

59.  Vernon, L.P.  (1960).  Spectrophotometric determination of chlorophylls
     and pheophytins in plant extracts.  Anal. Chem. 32:1144-50.

60.  Westlake, D.F.  (1963).  Comparisons of plant productivity. Biol. Rev.
     38:385-425.

61.  Westlake, D.F.  (1978).  Rapid exchange of oxygen between plant and water.
     Verh. Intern. Verein. Limnol. 20:2363-67.

62.  Wiginton, J.R., McMillan, C.  (1979).  Cholorophyll composition under
     controlled light conditions as related  to the distribution of seagrasses
     1n Texas and the U.S. Virgin Islands.  Aquatic Bot.  6:171-84.

63.  Young, J.A., Evans, R.A.  (1978).  Etiolated growth of range grasses for
     an indication of tolerance to atrazine.  Weed Sci. 26:480-83.

64.  Zieman, J.C., Wetzel, R.G.  (1980).  Productivity in Seagrasses:  Methods
     and Rates.   In:  Phillips, R.C., MeRoy, C.P. (eds.).  Handbook of Seagrass
     Biology:  An Ecosystem Perspective.  Garland STPM Press, New York, pp.
     87-118.
                                     11-26

-------
                                             CHAPTER  III


                     RESPONSE  OF POTAMOGETON PERFOLIATUS AND MYRIOPHYLLUM SPICATUM
                                      PHOTOSYNTHESIS  AND GROWTH
                        TO ATRAZINE AND  LINURON STRESS IN ESTUARINE MICROCOSMS*
                                           W. Michael Kempt
                                       Jeffrey J. Cunnlnghamt**
                                         J. Court Stevensont
                                          Walter R.  Boynton#
                                            Jay C. Means!
                                             August  1982
             Contribution  No.  1431 Center  for Environmental and Estuarlne Studies,
              Un1ver1sty  of Maryland

             tHorn  Point  Environmental  Laboratories, P.O. Box 775,
              Cambridge,  MD.    21613

:             fChesapeake  Biological Laboratory, P.O. Box 38,
:              Solomons, MD.   20688

            **Present  Address:  Department of Agricultural and Resource Economics,
i              University  of Maryland,  College Park, MD.   20742
j

-------
f.
                                                 INTRODUCTION


                    The  littoral  waters  of  Chesapeake Bay  historically  have  been dominated  by
               productive communities  of submerged  vascular  plants  (e.g.  Bayley et  al.  1978;
               Stevenson and Confer 1978).  This  submerged  aquatic vegetation  (SAV)  can  pro-
               vide food, habitat and  refuge  from predators  for many  fish and invertebrates
               during crucial  stages of  their growth  cycle (Adams 1976).  Moreover,  SAV  com-
               munities  stabilize surrounding estuarine  ecosystems  by trapping and  binding
               sediments (Orth 1977),  and by  modulating  pulses in estuarine  nutrient  cycles
               (Barsdate et al.  1974;  Howard-Williams 1981).

                    Over the last two  decades, the  numerous  species of  submerged macrophytes
               native to Chesapeake Bay  have  undergone a dramatic decline in  abundance  and
               distribution, so  that current  populations in  the upper and middle bay  are less
               than 5% of their  previous levels  (Stevenson and Confer 1978;  Orth and  Moore
               1981). Just prior to this reduction  of native SAV species, there was an
               unprecedented invasion  of the  exotic SAV, Myriophyllum spicatum, which grew
               to nuisance proportions in a period  of 3-5  years, after  which  it too decreased
               rapidly (Bayley et al.  1978).

                    The  cause of this  decline has been a subject of considerable debate over
               the last  several  years; however,  one of the primary  factors postulated as a
               causative agent is the  runoff  of  agricultural herbicides (Kemp et al.  1982).
               The vast  agricultural enterprise  in  Chesapeake Bay's watershed employs a
               variety of phytotoxic compounds for  weed  control. Two  of the  most widely used
               herbicides in this region are  atrazine and  linuron.  In some previous experi-
               ments the phytotoxic effects of atrazine  were tested for the  macrophyte,
               Potamogeton perfoliatus (Cunningham  1979).  This plant, which  is one  of the
               dominant  native species in middle and  upper Chesapeake Bay, exhibited  strong
               photosynthetic inhibition at herbicide levels of about 100 and 1000  ppb. How-
               ever, even the lower of these  conccentrations is slightly  greater than the
               highest levels measured in runoff water entering the estuary  (Hershner et al.
               1982). Other recent studies  have  reported variable results regarding growth
               and mortality responses of this genus  to  various concentrations of atrazine
               (Forney and Davis 1981; Cornell and  Wu 1982), although some suprassion of
               growth was generally evident at 10-30  ppb (Forney and  Davis 1981).

                    The  purpose  of the present study  was to  further Investigate herbicide
               phytotoxicity in  the estuarine environment, examining  additional compounds and
               plant species over a broader range of  herbicide concentrations. In addition to
               P. perfoliatus we selected M.  spicatum for  these experiments  to test the
               Tfypothesis that it is more Tolerant  to herbicide stress  than  the native  species,
               thus possibly explaining  Its Invasion  just  prior to  onset  of  the general SAV
               decline.  Experiments were conducted  using two herbicides,  atrazine and linuron,
               each of which is  considered  representative  of a whole  class of compounds (Kemp

                                                    III-l

-------
et al. 1982). The study design employed nominal  herbicide concentrations of
0, 5, 50, 100, 500 and 1000 ppb aqueous,  and plant  response was  measured in
terms of apparent production and respiration as  well  as growth and abundance.
Herbicide concentrations in these experiments were  compared to those  observed
in *n extensive estuarine monitoring program to  interpret these  experimental
results in terms of the potential role of herbicides  in the SAV  decline.


                             MATERIALS AND METHODS


                              Experimental  Systems

     Experimental microcosm communities were established in 42 glass  aquaria
(32x 35 x 48 cm), each containing about 50 I of  filtered (5 p) estuarine
water and about 8 cm depth of sediment. Microcosms  were seeded with 10
individual Potamogeton perfpliatus or Myriophyllum  spicatum plants of 6-14 cm
stem length (2-6 cm rhizome) inserted into the sediments. Sediments were a
sandy silt similar to that described in Cunningham  (1979).  Water,  sediments
and plants were all obtained from the brackish (8-12)  ppt portion  of  the
Choptank River estuary, a tributary of Chesapeake Bay.

     Microcosms were maintained in an air-conditioned  laboratory providing
relatively constant water temperature (21.5 ± 1.5 C) throughout  the study.
Salinity was kept at 9.0 ± 1.1 ppt by replacing  evaporative losses with
distilled water. Illumination was provided by banks of cool  white  fluorescent
lights on a 14 h photoperlod, yielding about 150-200  pE1n nr^s-l at the
water surface. Water was continuously recirculated  using centrifugal  submerged
pumps with Inlet and outlet diffusers at  opposite ends of each aquarium.  The
inner walls of these microcosms were scraped clean  of  epibota  at weekly
intervals.


                         Photosynthesis and Respiration

     Apparent photosynthesis (Pa) and night respiration (Rn) for each plant
community were estimated twice weekly on  successive days using measurements of
dissolved oxygen (DO) in the water column at dawn and  dusk (Odum and  Hoskins
1958; McConnel1 1962). Concentrations of DO were maintained above  3-4 ppm
throughout the study by periodic aeration with standard aquarium bubblers.

     Corrections for gas transfer across  the air-water Interface were made by
applying a diffusion coefficient to the observed mean  departure  from  satura-
tion (Odum and Hoskins 1958). Appropriate coefficients were obtained  by
measuring the rate of oxygen Invasion Into filtered (1.0 ym) aquarium water
(previously deoxygenated by bubbling with N? gas).  These aquaria contained
no sediment, and a control with DO near saturation  was utilized  to account
for any DO change due to microbial metabolism. In all, 5 such  diffusion  experi-
ments were conducted *ndi2 examples are provided 1n F1g. 1. Coefficients ranged
from 0.15-0.22 g02 m  h   atm   and averaged 0.18.  Periodically, Pg and  Rn were
determined by measuring DO at 2-4 h Intervals over  an  entire dlel  sequence,
and these experiments demonstrated that the error in  estimating  metabolic

                                     111-2

-------
    3.0
    2.5
 O)
»
 X
O
•g  2.0
w
(0
*-
a
x
UJ
    1.5
    1.0
                    GAS TRANSFER ACROSS
                    AIR-WATER INTERFACE
           Control
Experiment 1
  k-0.21
                                   Experiment 2
                                     k» 0.16
                       _L
                                    JL
                                                            9.5
    i_
     O>
9.0 5
     c
     0)
     O)
                                                            8.5
                          X
                          O
                          •o
                          0)
                          "o
                          CO
                          0)
                          5
                          o
                     8.0
                                                                 o
                                                                O
                                                            7.5
                   J
                       1            2
                        Time from Start (h)
 Figure 1.  Example results of two experiments  to measure the coefficient of 02
          diffusion across the air-water interface  in experimental microcosms.
                                III-3

-------
rates from DO at dawn and dusk only was small  (<5%).  Examples of 2 such diel
patterns for control  (Fig. 2a) and treated (Fig. 2b)  microcosms with _P.
perfoliatus are presented in Fig. 2. Reaeration (diffusion)  corrections were
either positive (Fig. 2a) or negative (Fig. 2b) but were always <10-15% of Pa.

     Contribution of plankton to the observed  DO changes in  microcosm water
were estimated during the first experiment by  incubating duplicate clear and
opaque BOD bottles containing aquarium water.  Again,  Pa and  Rn were inferred
from DO changes in light and dark bottles which were  incubated for 2-3 h
suspended in front of the circulation pump outlet diffuser to gently and con-
tinually mix them. Plankton contribution to total water DO changes was generally
neglibible (<5%), and therefore was not measured in the second experiment.


                          Plant Abundance and  Biomass

     Plant density (shoots/nr) and shoot lengtn were  determined in each micro-
cosm at 3-4 wk intervals for the atrazine experiment. At each sampling all
stems were counted and measured for length. At the beginning and end of
experiments several representative plants were selected from each microcosm,
and these were measured individually and dried to constant weight at 60C
(272h). Regressions of dry weight versus stem  length  were obtained and used to
infer biomass from the stem density and length surveys. Initial biomass (i.e.
weight of plants inserted in sediment) was estimated  more directly by obtaining
a blotted wet weight prior to planting and developing a regression of wet  to
dry weight on a subsample of 35-40 plants (Fig. 3). When experiments were
terminated, all plants were harvested from each microcosm and shoots were
dried at 60C to constant weight. Biomass of roots and rhizomes was not con-
sidered here, since it was found to be proportional to shoot biomass for all
herbicide treatments in an earlier study (Cunninghern  1979),  and since it was
<3Q% of total plant biomass in control microcosms (Fig. 4).


                    Herbicide Dosing, Sampling and Analysis

     After establishing the microcosm communities, a  period  of 7-8 wk was
allowed for development and equilibration prior to herbicide dosing.
Technical grade (96.4% purity) atrazine was dissolved 1n 100 ml of methanol
for delivery Into the aquaria at nominal concentrations (calculated on the
basis of microcosm water volume) of 0, 5, 50,  100, 500 and 1000 ppb, with both
untreated and vehicle (methanol only) controls. Duplicate microcosms were
randomly selected for each treatment.

     At the termination of these experiments,  samples of water, sediment and
plant material were obtained from each aquarium for herbicide analysis. Two
liters of water were collected in glass bottles and stored at 4C until analysis.
Approximately 1.0 kg of sediment was collected from each system using brass
corers, and these samples were frozen at -40C  1n glass bottles until analysis.
Samples of plant material were drained, wrapped 1n aluminum foil and frozen at
-40C until analysis. The extraction, clean-up  and concentration procedures for
atrazine and Unuron were similar and, therefore, will be presented together.
All glassware and apparatus used in the extraction and clean-up procedures were

                                      111-4

-------
            a) CONTROL
            Oo CONCENTRATION
                                           Reaeration
                                           Correc'fion
            RATE of ©2 CHANGE
            b) 100 ppb ATRAZINE
  cv
 O
  o>
       8

       7

       S
            O2 CONCENTRATION
CM
     -H)-2

 '«  +0.1

       o
  CM
 O
     +0.2
          Pa
                          Reaeration Correction
           RATE of O2 CHANGE
                8      12      16     20

                            Time (h)
                                             24
Figure 2.  Example die! patterns of 02 concentration and .-ate of chanqe
          in 02 with correction for gas exchange for a) control micro-
          cosm and b) 100 ppb atrazine treatment.  Apparent 0-,
          production (Pa) and night respiration (Rn) are indicated
                        III-5

-------
   0.4
   0.3
o>
O>
   0.2
   0.1
                       WATER CONTENT
          o  P. perfoliatus, control
          Q  F. perfoliatus. 1-5 ppb
          x  P. perfoliatus. 50 ppb
          •  M. spicatum. control
                   0 /
                   s
                  xo
      o  /
        /

    »/'°J^  ^All Plants
              Y - 0.072X + 0.037
V^  »       (r2 = 0.88)
                      •Potamogetan perfoliatus
                       Y- 0.098X - O.O05
                       (r2-0.92)
                            2          3

                           Wet Weight, g
Figure 3.  Water content (dry versus wet weight) for Potamogeton perfollatus
         and Myrlophynum spicatum us»ri in microcostn experiments:	^
                              III-6

-------
  -
K «•
              jn
               Q.
              •*»
               O)


              F"   3
               d>
               o
               o
              rr
                                    ROOT vs SHOOT WEIGHT

                                    (Potamogeton perfoliatus)


                                      /                  ®
                            »v

468

  Shoot Weight, g/plant
                                                                  10
               Figure 4.  Belowground (Root) versus aboveground (Shoot) weights  for

                        Potamogeton perfoliatus used in microcosm experiments.
                                            III-7

-------
washed with glass distilled delonlzed water,  ana pesticide grade methanol  prior
to use.

     '"xtractlon of frozen sediment and plant  tissue samples was  achieved by
weighing the frozen material  in a tared,  pre-extracted Soxhlet thimble.
Samples were exhaustively extr?cted in a  Soxhlet apparatus for 24 h  with
200 ml of d1st11led-1n-glass (DIG) methanol  (Burdick  and  Jackson,  Milwaukee,  MI).
Following extraction, sediment or tissue  samples were oven-dried over  night  at
105C.  Methanol extracts were dried over  anhydrous  sodium sulfate, concentrated
in a rotary vacuum evaporator to approximately  10 ml,  and taken  to dryness
under a stream of pure nitrogen. The residue  was dissolved in 1  ml of  DIG
met anol for analysis. Some samples required  clean-up prior to analysis, where
the entire 1 ml sample was passed through a preparative reverse-phase  liquid
chromatograpMcc column (Bondapak CIQ) "sing  a  solvent system of 50:50
methanolrwater.  The fractlon(s) for atrazlne and/or  Unuron were collected,
evaporated under nitrogen and dissolved 1n 1  ml  of  DIG methanol  for
quantitative analysis.

     Water samples (1-2 t) in glass bottles were filtered through  pre-
combusted Whatman GF/C glass-fiber filters and  then pumped at a  rate of
3 ml/min through a SEP-PAK CIQ (Waters Associates)  cartridge which was
pre-cleaned with 4 ml each of DIG acetonltrile,  DIG methanol and distilled-
deionized (DDI) water. After this procedure,  the bottle was washed with
50ml of DDI water, and this was also pumped  through  the  SEP-PAK.  Finally,
15 ml of DDI water was passed through the SEP-PAK to  remove salt.  The  SEP-PAK
was eluted with 5 ml of DIG methanol, and the eluent  was  evaporated  under a
stream of nitrogen and dissolved in 1 ml  of DIG methanol  containing  a  known
quantity of an internal standard (diphenylamine) for  quantitative  analysis.

     Quantitative analysis was performed  using  a combination of  gas  chromatography
and liquid chromatography. Atrazlne residues  were quantified using a Hewlett-
Packard 5840 gas chromatograph equipped with  2  m length of all-glass column
(2 mm ID) packed with 10% Carbowax 20 M on 80/100 Supelcoport. The column
was maintained at a temperature of 160C,  andd Injector and detector  were held
at 225C and 300C, respectively. An alkali-flame nitrogen-phosphorus  specific
detector was used to measure the amount of atrazlne and Internal  standard  in
a 4 pi sample. Quantification was achieved by comparing the atrazine/internal
standard ratio (AT/IS) to a calibration curve of AT/IS versus atrazlne
concentration.

     Unuron and atrazine residues were measured using a  Waters  Associates
6000A liquid chromatograph equipped with  a p-Bondapak-Phenyl (T.M.)  column,
a model 440 fixed wavelength ultraviolet  detector (254 ym), a Perkin-Elmer
M-2 electronic integrator and a Linear 159 chart recorder. Samples (20 yt)
were Introduced into the column using a Rheodyne 7125 Injection  valve  and
eluted using an 1socrat1c solvent system  of 35% accetonitrlle/water. Areas
under the sample peaks were Integrated and compared to a  calibration curve for
the compound.
                                     111-8

-------
                        Physical  and Chemical  Variables

     Dissolved oxygen and temperature measurements were obtained using polaro-
graphic oxygen electrodes (Orbisphere Model 2709) calibrated dally 1n moist
air. Salinity was measured at the beginning of each week using a Beckman
(Model RS5-3) Induction salinometer. Photosynthetlcally active radiation was
determined weekly In selected microcosms at mid-water depth using a LICOR
(Model-LI-192SB) cosine-corrected sensor. Concentrations of dissolved Inorganic
nitrogen species (NH*, NO? and N(E) were measured on microcosm water
passing a Whatman GFC (1.2 u) filter using a Tecnnicon Auto-Analyzer system.
Nitrate and nitrite were determined according to the cadmium-reduction method,
and ammonium analysis employed the Solarozano technique (Strickland and Parsons
1972).


                              Statistical  Approach

     Several statistical tests were utilized to identify significant effects
of herbicides in the microcosm experiments. All analyses were done using appar-
ent photosynthesis (Pa) and biomass as the variables indicating macrophyte
performance. Testing of effects on Pa was divided into two stages involving
pre-treatment and post-treatment data. In the first stage, pre-treatment data
were subjected to an analysis of co-variance to determine if differences 1n
the rate of Increase of Pa among microcosms developed during the 4-5 wk
period prior to herbicide treatment.  In addition, the means of pre-treatment
data were compared to determine if photosynthetic rates were equal  in all
treatments at the start of the experiment. This pre-treatment analysis was
essential in the present study because differences among microcosms could
develop in response to factors (such as sediment conditions) not related to
herbicide treatment, and such differences would invalidate post-treatment
comparisons among experimental systems. The second stage of our statistical
Interpretation consisted of one-way analysis of variance (AN OVA) testing,
where herbicide concentration was the treatment.  Tests were run Independently
for observations in each post-treatment week and significant differences among
means were Identified using the Student-Newman Kuels (SNK) multiple range
test. Analysis of post-treatment data was done both with and without correction
for pre-treatment differences in Pa.  Biomass data were tested using simple
one-way AN OVA with SNK.


                          RESULTS AND DISCUSSION


                           Atrazine Experiments

     For the sake of clarity, atrazlne and linuron experiments will be discussed
separately. However, these will be compared and the overall implications
considered in a final section.
                                     111-9

-------
«            Atrazlne Concentrations
                  Atrazlne concentrations remained relatively unchanged through the experi-
             mental period, with 84-89% of the original dose remaining after 4 wk (Table 1).
             Assuming first-order degradation kinetics, this Is equivalent to a half-life
             of about 3-4 mo. which Is comparable to values observed by Cunningham (1979)
             and Correll and Wu (1982) for similar microcosms. This degradation rate 1s
             considerably slower than the rate reported by Jones et al. (1982a) using smaller
             experimental systems exposed to ambient temperature and light. Concentrations
             of the herbicide taken up by either P. perfollatus or £. spicaturn were below
             our limits of detection (1.0 pg/kg) Tn all experiments, and concentrations
             were also low for the sediments. Thus, very little of the applied atrazlne was
             sorbed or degraded, so that concentrations were relatively constant.


             Atrazlne Effects on Photosynthesis and Respiration

                  Measurements of Pa and Rn for these SAV microcosms were Initiated 2 wk
             after planting of the communities, and they were continued for 5 wk prior to
             treatment. During this pretreatment phase, Pa and Rn (F1g. 5 and 6, Appendix
             Tables 1 A-D, 5 A-D) exhibited a general increase as the communities became
             established. Values for Pa were consistently higher in the M^. spicatum
             communities, and this may be attributable to the 30% greater quantum efficiency
             of this plant compared to £. perfollatus (Kemp et al. 1981). Since the analysis
             of experimental effects compares treatment level to control for all dates
             following initial dosing, some consideration was given to t.ie comparability
             of these communities prior to treatment. Analysis of covariance Indicated tv»at
             the slopes of increasing Pa versus time for all aquaria in a given experiment
             were not significantly different (Table 2). The mean pretreatment values of
             Pa were, however, statistically different (Table 2), so that all post-treatment
             values could be normalized to controls according to the pretreatment ratio of
             mean Pa for a given microcosm to mean Pa of controls (referred to as "intercept
             adjustment"). We present ANOVA results for these data both with and without
             this adjustment.

                  The effects of atrazlne on Pa of P. perfoliatus and M_. spicatum are sum-
             marized in Table 3. At atrazlne >50 ppF there was significant depression of
             Pa for £. perfollatus in all posT-treatment weeks (8-11), while at 5 ppb the
  t           effects were significant in weeks 9 and 10 but not 8 and 11. The vehicle con-
             trols exhibited a small but significant decrease in Pa for week 9, with an
             effect 1n week 8 only apparent without Intercept adjustment. Myriophyllum
  j           spicatum Pa was less sensitive to the negative effects of atrazlne, and In
  !           fact there was a significant enhancement of Pa over vehicle controls at
  I           5 ppb. Similar growth stimulation at low concentration of photosynthetic
  )           Inhibitor herbicides has been observed for various terrestrial weeds (Ries
  |           1976; Hoffman and Lavy 1978) and submerged plants (Forney and Davis 1981;
  i           Correll and Hu 1982).  The mechanism for this stimulation is unclear. However,
  i5           two possible explanations are: first, utilization of nitrogen released during
  i           degradation of atrazlne by these nutrient-limited plant populations (Kemp et
  •           al. 1980); and second, Increased chlorophyll £ production under atrazlne stress
  !           (Cunningham 1979). A significant loss of Pa was evident at atrazlne >50 ppb
  \           for the first 2 wk of treatment. However, in weeks 9 and 10 no negative effects
  i

  I                                                 111-10

-------
Table 1. Results of atrazine analysis of water, sediments and plant tissue at the end of
     the 1979 microcosm experiments*.  Given are means ± standard deviation.
Sample
50 ppb atrazine
100 ppb atrazine
500 ppb atrazine
1000 ppb atrazine

42
89
445
869
Water
.3 i 4.5
.6 ± 24.6
i 57.6
± 128
Sediment
TR*
TR
1.4 ± 0.8
2.4 ± 0.4
Plant Tissue
(wg/kg)
TR
TR
TR
TR
Percent
Recovery
84
89
89
87
#Control andbppD samples were contaminnated prior to analysis.
*TR ^0.1 ug/kg for sediments and 1.0 ug/kg for plant tissues.
                                            III-ll

-------
 CM
O
 o>
 J
•o
 O
0)
ffl
a
a
                 P. perfoliatus vs ATRAZINE
       -  CONTROL
         VEHICLE CONTROL
         5 ppb
         50 ppb
            J	L.
         100 ppb
         500 ppb
         1000 ppb
                                  -.P.
                                     treatment
                                     initiated
                                      -..T. i .T.T. T« -.T.-.T.-r-?
                                      •• •!••"• ••"•-t» *•• •••.."• •
                                  „**.».*.!.• i-.. r..-	'.-
                                  »•"••, • . », ••»  •••••»•••",«
                          6
                              7
                                              10    11
                     Week of Experiment
Figure 5.
         Mean of four weekly measurements  of  apparent 02 pro-
         duction for microcosm communities containing Potamogeton
         perfpllatus exposed to 5 levels of atrazlne  dosage with
         two kinds of controls (untreated  and treated with solvent
         only).  Shaded areas represent approximate departures
         from control levels.
                          111-12

-------
o
oZ
**
»
co
Q.
a
                       M. spicatum vs  ATRAZINE
-   8
 CM
O   6
O>
E   4
PH
6

4

2

6

4

2
             .  CONTROL
               5 ppb
              50  ppb
                 _l	I
              100 ppb
              500 ppb
              1000 ppb
                                         -treatment
                                          initiated
                                      .-*:*'••
                                          _•••?•
                                .**•?«?.•;£:•::
                                   ••
                            6
                                 8
                                  9    10   11
                         Week of Experiment
Figure  6.  Mean of four weekly measurements of apparent 0? pro-
          duction for microcosm communities containing Myriophyllum
          spicatum (see Fig. 5 caption) exposed to 5 levels of
          atrazine dosage with vehicle control (treated with solvents
                          111-13

-------
i •
             Table 2. Summary of one-way analysis of variance and covarlance for pretreat-
                  ment behavior in experiments to tsst herbicide effects on Potamogeton
                  perfoliatus and Myriophyllum splcatum apparent photosynthesis (Pa). Tests
                  are for equality of slopes as Pa increased during pre-treatment and
                  equality of mean pre-treatment values of Pa.t
Experiment
Potamogeton
vs. Atrazine
Myrophyl 1 urn
vs. Atrazine
Potamogeton
vs. tinuron
Myrophyl 1 urn
Equality of
Slopes I
0.3645
0.8064
0.2270
0.6561
Equality of
Cell Means*
0.0011*
0.0000*
0.0110*
0.0353*
               vs.  Linuron
             tTable entries are probabilities that slopes or means are different.
             lvalues <0.05 Indicate that there are significant differences (p<0.05) among
              pre-treatment slopes of Pa over time.
             *Va1ues <0.05 were taken to indicate that time-zero Intercepts (I.e., Initial
              Pa) were significantly different among Individual microcosms.
                                              111-14

-------
i!
             were seen at concentration <100 ppb.  Respiration  generally  followed  Pa  closely
             (Appendix Tables IB and D, 5B and D)  with  some  lag  1n  response time,  so that
             the ratio Pa:Rn provided a less sensitive  measure of  stress.

                  The effects observed 1n this experiment  at concentrations >50 ppb  are
             similar to those reported previously  for various  £-triazine herbicides  and sub-
             merged vascular plant species. For example, Sutton  et  al.  (1969) found  40%
             loss of Pa for £. canadensis and Myriophyllum sp. after  4 d at 120 ppb  atrazine,
             while Bielecki and Skrabka (1976) found similar effects  for Spirodela polyrhiza
             after 24 h exposure to 100 ppb simazine, and  Fowler (1977)  reported  50%
             decrease in Pa of Myriophyllum sp. at 500  ppb of  another triazine (DPX
             2674) after 10 d. Correll and Wu (1982) observed  a  range of effects  from 48%
             stimulation to 39% depression of Pa after  3-6 wk  exposure to  75-650 ppb atra-
             zine for 4 other SAV species (_P. pectinatus,  Zanichellia palustris, Vallisneria
             americana, Zostera marina). Recently, Walsh et  al.  (1982) has estimated the
             atrazine concentration for a 50% loss of Pa in  the  tropical seagrass, Thalassia
             testudinum, to be 320 ppb.

                  There is striking evidence of photosynthetic recovery  for both P_.
             perfoliatus and M^. spicatum ? wk after treatments of _<50 ppb  atrazine,  and
             for M. spTcatum even at 100 ppb (Figs. 5 and  6, Table  3).  This  recovery 1s
             simiTar to that reported by Cunningham et  al. (1982) for £. perfoliatus at
             100 ppb.  What is so remarkable about this recovery is that it occurred in the
             face of relatively constant external  aqueous  concentrations of the herbicide.
             Jones et al. (1982b) have demonstrated the ability  of  £. perfoliatus exposed
             to 10-100 ppb atrazine to recover Pa  with  sequential washes in herbicide-
             free water, and this response was attributable  to removal of  the herbicide
             from the chloroplasts. In the present case, however, the mechanism of
             recovery must involve some sort of persistent detoxification, but Us exact
             description awaits further Investigation.

                  Phytoplankton production comprised a  small to  negligible portion of total
             community photosynthesis, representing <5% of community  Pa  for all but
             one occasion (when 1t was <15%). In fact,  mean  plankton  Pa  was negative for
             25 of 42 occasions (Table "?). A possible explanation for this reduced phyto-
             plankton production is nutrient limitation (Kemp  et al.  1980). Concentrations
             of dissolved Inorganic nitrogen remained low  in all herbicide experiments
             (Table 5), with NHj generally below 3 yM and  NO*  below 5 MM for
             all but the high herbicide treatments (>500 ppbj. Nevertheless,  there was a
             significant trend of reduced Pa and Increased Rn  in the  plankton with Increasing
             atrazine, particularly above 50 ppb.  This  pattern 1s consistent  with  various
             studies of £-tr1az1ne toxlcity for algae (Walsh 1972;  Pruss and  Hlggins 1974;
             Valentine and Singham 1976; Hawxby et al.  1977).  Since the  walls of these
             experimental aquaria were maintained  clean of algal  periphyton,  while sediments
             (Cunningham 1979) and plankton (Table 4) were generally  heterotrophic,  dawn-dusk
             DO Increases 1n microcosm water probably represent  slight underestimates of
             vascular plant Pa.


             Atrazine Effects on Plant Biomass

                  Time-course patterns of £. perfoliatus biomass 1n response  to atrazine

                                                  111-15

-------
Table 3. Summary of one-way analysis of variance for post-treatment data  1n
     experiments to test atrazlne effects on apparent photosynthesis  (Pa)  of
     2 macrophyte species.*
  Experiment
                 Treatment
                                             Experimental Week Number
                                           nr
                                      11
Potamogeton
mogi
7~Ai
  vs. Atrazine
ttyrlophyllum
  vs. Atrazfne
                           No Intercept Adjustment
Control
Vehicle Control
5 ppb
50 ppb
100 ppb
500 ppb
1000 ppb

Vehicle Control
5 ppb
50 ppb
100 ppb
500 ppb
1000 ppb
 0.650a
 0.536b
 0.505b
 0.284c
 O.OOSd
-0.286e
-0.196e

 0.514a
 0.617a
 0.211b
-0.024c
-0.183d
-0.448e
 0.529a
 0.301b
 0.158C
-0.069d
-0.383e
-0.341e
-O.bl7f

 0.353a
 0.392a
-0.134b
-0.253b
-0.389b
-0.547b
 0.556a
 0.568a
 0.416b
 0.307C
 0.066d
-0.348e
•0.291e

 0.540a
 0.674b
 0.394a
 0.398a
-0.122c
-0.268d
 0.549a
 0.610a
 O.SlOa
 0.341b
 0.140c
-0 198d
                                                                    0.588a,b
                                                                    0.672a
                                                                    0.514b
                                                                    0.323c
                                                                   -0.083d
                                                                   -0.321e
                        Intercepts Adjusted to  Control
Potamogeton Cont^
vs. Atrazine Vehicle Control
5 ppb
50 ppb
100 ppb
500 ppb
1000 ppb
Myriophyllum Vehicle Control
vs. Atrazine 5 ppb
50 ppb
100 ppb
500 ppb
1000 ppb
0.602a
0.563a
0.505a
0.225b
-0.124c
-0.334d
-0.209c
0. 188a
0.399b
-0.060C
-0.294d
-0.270d
-0.734e
0.481a
0.328b
0.158C
-0.129d
-0.473e,f
-0.565f
-0.396e
0.027a
0.175a
-0.4055
-0.522b,c
-0.476 b,c
-0.833C
0.508a,b
0.595a
0.416b
0.248C
-0.066d
-0.396e
-0.304f
0.178a
0.457b
0.123a
0.128a
-0.209C
-0.554d
O.SOla
0.637b
0.510a
0.282c
O.OOSd
-0.246e
-0.173e
0.262a
0.455b
0.243a
0.053C
0.170d
0.607e
*ANOVA tested against treatment Tor each week after treatment Initieted.  Values
 given are mean P.  (g Oom'-^h"1)  , and letters following each entry  indicate
 significantly (p<0.05) different statistical group according to an SNK multiple
 range test (read vertically with treatment  for each experiment).
                                  111-16
                                                                                            i

-------
t

0 •- 0
«- C f
O O 4->
m f ID
cn E 01
C 3 TJ
f V)
c c -o
f O t-
10 0 <0
C CM C
o o 10
0 4->
4-> in
I/I Ol
IO I/I VI
4-> OJ C
4-> ID
E 0 Oj
in c E
O ^3 Ol
0 «-
L. VI (O
O Ol
f 3 C
•0 >
C > f
f 'S
IO
c c
f 3
f -O >
10 4->
L. ID CM
f f O
0. 0
in O 4->
01 v> Ol
fQ
C-— -. 01
ID 1 4->
"0 C i-
10 f- C
<00) — •*

o cy***"*
f Ol f
•o •
o « c
t. 41 O
O. C f
O 10 E
4J t- 3
C 10 C •
f JC 0 1
Q.4-) T3
4-> "» O 1
C ««
QJ -O  4J O
O. IO IO
CL O) o cn
If C C
O (rt "r* *^~
>> C CM
c- 10 oe 01
E "5. 01 1—
3 O f >
CO f 4->
• >, to 4->
a. o.*x
Ol
JO
•2





d
VI
c
1
4->
10
Ol
01
M
IO
£








0
o
o
fl

0
o
m

0
o
f— 1

0


in

-0
e c
£
fl O
1
VO 00
in CM
CM in
CM CM
1
CM 00
CO CM
44 4-1
cn CM
oo vo
fi fi
vo o
CM fl
4-1 44
r-» in
* CM
fl fl
1

00 CO
00
co oo
h- co

fl CM
fl fl
1
*»• m
««• CO
44 44
r» eo
CO *
0 O
10 C
CL 0£
cn vo
r-.
00 Ol
O fi
CM fl
4-1 4-1
CO CO
O 0
1
«— 4 f**
fi O
4-1 4-1
CM fl
CO*
O O
1
in CM
4-1 4-1
0 ft
0 0
00 O
fl CM
4-1 4-1
0 0
CM fl
o o
§VD
0

fi in
CM fl
oo
1
00 CM
O fi

CM m
CM O
00
O CM
* fl
44 44
00 VO
CO O
00
10 C
CL ee
cn 01
oo oo
OO CO
CO CO
in co
o o
fl CM
r- co
4-1 44
cn f-
CM *
0 O
1
r- in
* CM
4-1 4-1
VO VO
fl CO
0 O
1
CO CO
fl CO
4-1 44
CO *
CO fl
0 0
r«- co
CO CO

in vo
cn co
O 0
r-t CO
<— 1 fl

VO «O
O fi
0 O
fl O
r- *
4-1 44
in*
CM fl
00
c?af
cn in
00 01
U
"a
CM CO
CO O
CO
4-1 44
o in
f^ O
o vo
i
0 r~
CO fl
4-1 44
CO ft
co r-»
CM fl
i
cn CM
fi CO
4-1 44
VO VO
o in
CM f"
i
in en
44 44
vo r~-
oo *
c-» vo
CO CM

CO O
* 00
O 0
1
cn oo
eo *

Sg
1
* in
* CO
44 44
r> oo
CO *
O 0
IO C
CL oe
cn vo
00 Ol
CO ^-»
fl VO
CM CM
4-1 41
O PO
fi ro
0 O
i
oo r-
CM fl
44 4-1
* f-»
fl fl
00
1
* VO
CO ft
CO CO
fl (M
0 O
1
1— 1 ff
CM CO
4-1 44
CO VO
0 fi
o o
CM in
CM fl

in in
CM O
00
VO O
O fi

m r>»
i— i i—>
O 0
1
O CM
* fl
eo vo
CO O
O 0
•0 C
o. ae.
cnST
cn 01
t— i ~-*
                                            111-17

-------








"O
i.
*o
T3
CT> C.
c to
•P™ -fc-*
C IO
•V—
f0 41
4J
C I/)
0 C
O  at

g» ^
o at
§u
c
10 o
*-~ 0
«o i/>
c 3
0 0

+^ fr.
m 10
ti ^
C JC
Ol 4->
O t-
C X
o
U T3
at
c 
ai at
T- U
i- 4-*
4^
C I/I Z
01 3.
4-* 4J
C >»
10 O f
§f
g
o
H- C
l_ "*"
§10 C
01 O
3 O 4-»
to at IB
• IO ">
in at
CM T3
at
t—








































+,>
c
at
€

10
at
L.
r—

















































a
O.
a.

o
o
—^

.f.

0
a.
a.

^
o
n












o
a.
a.

o
o


4-

Q
CL
QL

=>
n





^^
O
<-
c
5
at -o
t— a
u a.

c o
at •
> in

*
^_
o

ij>
c
s



at

S


at
•o
•»•
o
•^
J3
l_
£



IO
01
U
at
a.
i/>





1 CO
o
35










4-<^
1C
^











ICO
o
z










•4*^
3C
Z









ten
o








^^-
.3C
Z




























^
co
t
ao

41

CO
CO
f
in


ao
^.
f
o

41

^^
in
•
•-•




in
CM
O

41

«— I
CO

o


o*
CM
«
*H

^|

in
in
*
CM


CM
•"*
CM
41
CM
(
<— 1


f^
in
o

41

CM
CM
•-4


0
00
^^^
IO
oo"


s
c.
3
c

~*

IO
J_*
**
•~
*5
(^
L.
£
*
a-
o o» m
o o c
• • •
0 O f~<

•H *l 4(

o r* vo
^ *-H f— (
• • •
^H C? ^H
C\J

*-H <— 1
^h ^D
9 ^
i-l CM

41 41
O
10 ^4 -3-
10 • in
• O •
CM V CO




in oo 9>
CM CM CM
O O CO

41 41 41

00 O CM
CO CO CM

O O CO


•-« f^
CO f-l
• e
i— 1 i— «

4' 4(
O
in «-• r*.
co » in
• o •
CM V CM


O» i-> CM
CM CM 4-
O O CM
4< 4) 4)
10 i»«. ro
•W CM CM
# * #
O O CM


10 co in

o o es

4( 41 41

in ^ o
O CM 00
CM O CM
"*• *~*
*-* CM f~
to o <-i 1-1 o
00 0 00

at F** at **^ at to
at ^*» at cy* at ^**»
2O2 *»»3t 00


o
L.
3
C

~*





•D
•^
U
Q.
IO
z
CO O
(^ O

Cf^ CO

41 41 1

ff> CO
1,^ CM

O C3
i— 1

r>- o
00 0

CO O

41 41 1

in o
10 o

*r o




CM ^>
co to
IO CM

41 1 41

oo oo


CO -H


to
00

0

41 1

i— 1 i— 1
o ao
9 v
CM i-l


«*• IO CM
CM »-i in
O O *f
41 41 41
<• oo in
CO CO f*
• • •
o o in


o •-•
co 
f^- O
0 Q

41 4f

O CM
to o
•4 o
*-* *-*
^+ tv r>*
tO O i-H I-l
oo O
JX -^.^ 00 Jf
ai t*> ai ••». at
203 ^»2


















111-18

-------
treatment are presented in Fig. 7. For the control  microcosms net production
of above-ground biomass exhibited a logistic growth pattern,  with an initial
lag phase of 0.6 gdw nr^d"*, an exponential phase of 2.1 gdw  nr^d'*,
and an incipient plateau of 1.0 gdw nr'd"1. These correspond  well (assuming
Q£ = dry weiaht = 1.0) to apparent photosynthesis levels of 1.2, 1.5 and
1.8 gO^  d  , respectively. The pattern of P. perfoliatus growth was
unaffected by 5 ppb atrazine, and while some effect is evident after 4 weeks of
exposure at 50 ppb, it was not significant )ANOVA;  SNK tests).  Atrazine MOO ppb
resulted in significant loss of biomass at week 11, although  the effect at
week 8 was significant only at concentrations >500 ppb. Forney and Davis
(1981) also found significant loss of biomass Tor P_. perfoliatus after 4 wk
exposure to 100 ppb atrazine, while Fcwler (1977) reported 20-80% loss of
plant vigor (visually discerned) for £. pectinatus  at 125 ppb of another s-triazine,
(DPX 3674), for 30-54 d.  The submersed aquatic plant, J/. americana is apparently
more sensitive to herbicide stress, where Correll and Wu (1982) observed significant
increases in mortality after 47 d at 12 ppb atrazine.
     Biomass production by M. spicatum (Fig. 8) exhibited more
trend in control systems, wTth highest rates during the final  i
1.7 gdw m~'d~l. The same growth sequence was exhibited in all  t
                                                               of an exponential
                                                              interval  being  about
    gdw m'^d"1. The same growth sequence was exhibited in all  treatments  through
100 ppb, with no significant treatment effect.  Only at 500 and 1000 ppb were
biomass levels significantly reduced compared to controls. Thus,  biomass  data
suggest (as did Pa measurements) that this exotic species of SAV  was more
tolerant of herbicide stress than was the native. However, unlike the Pa
data, there was no evidence in biomass data of  a growth stimulation at the  low
treatment (5 ppb).

     The biomass estimates in Figs. 7 and 8 are based on direct measurements
for the first and fourth dates (the first was based on wet weights of plants
inserted into aquarium sediment, and the fourth was from final harvests),
but indirect methods were used for the two intermediate observations. These
are based on length-weight relations developed  during the course  of the
experiment. In an earlier study (Cunningham et  al. 1982), a significant
decrease in slope of this relation was observed at atrazine treatments of
100-1000 ppb for £. perfoliatus. In the present study, the atrazine treated
plants similarly "exhibited a reduced slope (Fig. 9), although  the difference
was not significant, perhaps due to the shorter water columns  in  these micro-
cosms compared to those used by Cunningham et al. (1982).  These  data are from
experimental week 8, and the few plants taken from 500 and 1000 ppb atrazine
treatments followed the same general trend as those from 50-100 ppb.  A pro-
nounced difference in slope occurred for M. spicatum treated with atrazine,
and eventhough fewer samples were taken fb~r this species, the slopes were
significantly separated (p<0.05) (Fig. 10).  The mechanism for this stem  elong-
ation under atrazine stress is unclear, although it is similar to a well-
established shade response (Cunningham et al. 1982).


                              Linuron Experiments

Linuron Concentrations
     Concentrations of linuron were dramatically reduced during the 4 wk experi-

                                     111-19

-------
Figure 7.
                    P. perfrliatus vs  ATRAZ1NE
.
15-
10
5
15
10
1 5
8
«
E 15
«^
* 10
• 5
(O O
T- T-
3SBUJO|a 11
03 5
0.
^
(0
= 15
0 '**
CO
£ 10
5
15
10
5

VEHICLE
CONTROL
range
"s^^
mea0^r*xj

- 50 ppb
' _ P*
" 100 ppb
.
™. *!
500 ppb
*
1000 ppb

fx*!3
f ..i..i |:-:-:-:-xi




•n
Initiate
**
0)
«
CO
**





















4
•:':•'

^

:5:

•.V
•:•:•



:•:




bn
h

I
l:>:


I
•i-.-
x'S












•
;'




•".%•

:S:

r^
i









+— i
1

4~










-
"
-


-
•


                                      8
                                   11
              Week of Experiment
Time-course measurements of shoot biomass  for  Potamogeton
perfoliatus exposed to  5 levels  of atrazine  dosage with
vehicle control (solvent treatment).   Initial  biomass was
estimated by measuring  wet weight and  using  wet-dry weight
regression; middle two  values  were estimated by measuring
stem lengths and using  a length-weight regression; final
biomass estimated by liarvest.   Given are mean  and range of
two replicate microcosms.
                               II1-20

-------
                  M. soicatum vs ATRAZINE
     8


     4





     8
    CO
    o
    o
    o
    k.
    o
        8
        4
   T3   ^

   0)
 CO
 CO
 CO


 o

m

"c
 CO
       8


       4
JS   8

o
CO   y|
co   4
>
       8


       4h
            VEHICLE

            CONTROL
                    range
                 mean
            5 ppb
            50 ppb
            100 ppb
            500 ppb
            1000 ppb
               1        579

                    Week of Experiment
                                          12
Figure 8.
       Time-course measurements of shoot blomass  (see Fig. 7 caption)
       for Myriophynum spicatum exposed to 5 levels of atrazine
       dosagt with vehicle control (treated with  solvent only).
                           111-21

-------
      0.25
   o,


  £
   O>

  "55



   >,  0.15

  Q
   c
  £
  Q.
      0.10
      0.05
                       WEIGHT vs LENGTH

                           P. perfoliatuf,
Controls
                      '°° «/   o     .
      ~~....—        ~/T~^y
Y= 0.0036X - 0.01   o  /        xx

     (r2 = 0.83)       V        /
                     »^X •
                  /
              •  /o
          o   <>y
        O'v   V
      /•• •/


 0  ox*° °'°
                     oo
         rx_^f50-100 ppb Atrazine

                    .0028X - 0.02
           o  x^	f 50-100 pp

           o o'    W= 0.002
            X O    '»•    x_o _
           /g     o       v«—
                    20        40        60

                         Stem Length, cm
                                80
Figure 9.  Regressions of stem length versus weight for Potamogeton perfollatus
         in microcosms untreated and exposed to atrazine.
                         111-22

-------
I
                    0.20
                                             WEIGHT vs LENGTH
                                                 M. spicatum
                                  Controls
                            Y- 0.0051X -
                                          20       30       40

                                              Stem Length, cm
50
               Figure 10.  Regressions of stem length versus weight for Myriophyllum spicatum
                          in microcosms untreated and exposed to.atrazine^—	
                                                  in-23

-------
mental period, where only about 15% of the original dose remained to the end
(Table 6). This amounts to a half-life (time until 50% loss) of about 1-2 wk,
for a first-order decay process. The degradation of Unuron was about 10 times
more rapid than that observed for atrazine in comparable experiments. While
there have been no previous studies of Unuron degradation In estuarine systems,
the literature for agricultural soils confirms the relatively labile nature of
this compound (Kemp et al. 1982). The amount of Unuron sorbed to sediments
1n these aquaria was similar to that in the atrazine experiment, but partition
coefficients (soil concentrat1on:aqueous concentration) were about 10 times
higher for linuron, which is consistent again with the agronomy literature
(Kemp et al. 1982).

Linuron Effects on Photosynthesis and Respiration

     After an Initial drop from week 3 to 4, Pa and Rn for both P. perfollatus
and J^. spicatum controls steadily Increased until week 8, after which no change
was "apparent (Figs. 11 and 12). Values of Pa and Rn were similar for both
species (Appendix Tables 4 A-D and 6 A-D). As with the atr^zine experiment, we
considered the comparability of pretreatment values of Pa by  means of an
analysis of covariance to compare the slope of increasing pretreatment Pa and
the mean Pa at time of treatment. We determined that the pretreatment slopes
of Pa versus time were not significantly different, but mean values of Pa
were (Table 2). Therefore, we analyzed metabolic data both with and without
adjusting post-treatment means to account for this departure from pretreatment
controls (referred to as "intercept adjustment").

     The results of sequential one-way ANOVA (with SNK) tests for differences
in Pa among Unuron treatments are summarized in Table 7. No effect on _P.
perfoliatus photosynthesis was evident at 5 ppb linuron, while concentrations
2;50 ppb always resulted in significant depression of Pa compared to controls,
and even greater effect occured at Unuron >500 ppb. Intercept adjustment
did not alter statistical Interpretation. For M. splcatum significant loss of
Pa occurred at Unuron ^100 ppb and occasionally at 50 ppb as well. While there
was a slight stimulation of Pa for M. spicatum at 5 ppb, this effect was not
significant. Respiration generally Toll owed a pattern similar to that for Pa,
Indicating a close coupling of anabolic and catabolic processes (Appendix
Tables 4B and 0, 6B and D). However, effects on respiration were less severe,
and the ratio Pa:Rn indicated that growth would be depressed 1n response
to Unuron treatment, although the effect would be somewhat buffered.

     Photosynthesis for both species exhibited a significant recovery trend
(Figs. 11 and 12) at all treatments jClOO ppb (determined by significance of
the slope of Pa versus time). This pattern is similar to that observed for
atrazine treated plants. However, the mechanisms for recovery are probably
quite different since linuron degradation occurred rapidly compared to atrazine
in these microcosms. About 36% of the original dose had already been lost just
1 wk after treatment. Correspondingly, at the 50 ppb treatment about 30% recovery
of Pa was apparent after 1 wk of exposure. Thus, while recovery of Pa 1n the
atrazine experiments probably Involved some sort of enzymatic detoxification,
simple degradation of dissolved Unuron in these experimental systems may
account for the observed return of photosynthesis.


                                     111-24

-------
Table 6. Results of Unuron analysis of water, sediments and plant tissues at the end of
     the 1980 microcosm experiments.  Given are means ± standard deviation.
Sample
Control
Vehicle control
5 ppb linuron
50 ppb linuron
100 ppb Unuron
500 ppb linuron
1000 ppb linuron
Water
( M9/»)
0
0
1.1 ± 0.2
4.4 t 3.1
12.8 t 5.6
69.7 ± 38.7
100.0 ± 27.9
Sediment
(ug/kg)
0
0
TR*
TR
TR
1.0 ± 0.5
1.4 ± 0.8
Plant Tissue
(wg/kg)
0
0
TR
TR
TR
TR
TR
Percent
Recovery
—
—
22
9
13
14
11
*TR _< 0.1 pg/kg for sediments ana 1.0 wg/kg for plant tissue.
                                        111-25

-------
                         P. perfoliatus vs  LINURON
                  3456     789    10   11

                               Week of Experiment

Figure 11.   Mean of five weekly measurements  of  apparent 02 production for
            microcosm communities containing  Potamogetor, perfoliatus exposed
            to 5 levels of linuron dosage with one  control (see Fig. 5 caption).
                               111-26

-------
         CM
         O
         O>
         E
         o
         •o
         O
         0
         CD
         a
         a
                         M. spicatum vs LINURON
               •  MeOH
        7   6
•  5 ppb
               .  50 ppb
                 CONTROL
                 100 ppb
                                       treatment
                                       initiated
                                     10   11
              34     56     7    89
                            Week of Experiment
Figure 12.   Mean of five" weekly measurements of apparent 02 production  for
            microcosm communities containing Myriophyllum spicatum (see
            F1g. 5 caption) exposed to 5  levels of llnuron dosage with  two
            kinds of controls (untreated  and treated with solvent only).
                               111-27

-------
Table 7. Summary of one-way analysis of variance for post-treatment  data  from experiments
     to test linuron effects on apparent photosynthesis  (Pa)  of  2  submerged  macrophyte
     species.
 Experiment     Treatment
                                               Experimental Week  Dumber
                                                                 Iff
                                                                            II
                                  No  intercept Adjustment
Potamogeton
  vs
moge
7TT
inuron
Myriophyllum
  vs. Linuron
Potamogeton
Tiogj
» L
  vs. uTnuron
Control
Vehicle
5 ppb
50 ppb
100 ppb
500 ppb
1000 ppb

Control
Vehicle
5 ppb
50 ppb
100 ppb
500 ppb
1000 ppb
          Control
          Vehicle
          5 ppb
          50 ppb
          100 ppb
          500 ppb
          1000 ppb
Cont
                     Cont
  0.239a
,  0.218a
  0.122a
 -O.lllb
 -0.243b
 -0.264b
 -0.279b

  0.218a
.  0.493b
  0.384b
 -O.OSlc
 -0.179c
 -0.144c
 -0.156c
        Cont
 0.342a
 0.310a
 0.132a
-0.090b
-0.202b
-0.253b
-0.313b

 0.339a,b
 0.331a,b
 0.425a
 0.139b
 0.088c
-0.141c
-0.083c
 0.371a
 0.457a
 0.387a
 0.1655
 0.075b
-0.147c
-0.246C

 0.407a
 0.576b
 0.691b
 0.365a
-0.050c
 0.016C
-0.056C
                                Intercepts Adjusted to  Control
      0.282a
     , 0.394a
      0.272a
     -0.035b
     -0.213b
     -0.170b
     -0.1396
              0.385a
              0.486a
              0.282a
             -0.014b
             -0.173b
             -0.159b
             -0.174b
             0.414a
             0.628b
             0.536b
             0.241c
             0.102c
            -0.052d
            -0.107d
 0.371a
 0.488a
 0.431a
 0.192b
 O.lOSb
-0.135c
-0.216c

 0.519a
 0.587a
 0.763b
 0.374c
 0.136d
-O.OOSe
-0.946e
             0.414a
             0.665b
             0.581b
             0.267c
             0.135c
            -0.042d
            -0.077d
 0.412a
 0,519a
 0.391a
 0.242b
 0.164b
-O.lllc
-0.203C

 0.505a,b
 0.596a
 0.794c
 0.422b
 0.160d
 O.OOOe
-0.046e
              0.455a
              0.695b
              0.5407a
              0.318C
              0.193d
             -0.0177e
             -0.064e
Myriophyllum
vs. Linuron





Control
Vehicle Cont.
5 ppb
50 ppb
100 ppb
500 ppb
1000 ppb
0.179a
0.283a
0.143a
-0.112b
-0.232b,c
-0.339c
-0.354c
0.300a
0.121a
0.184a
O.lOSa
-0.141b
-0.336b
-0.282b
0.368a
0.367a
0.450a
0.335a
-0.003b
-0.211C
-0.255C
0.479a,b
0.377a,b
0.522a
0.343b
0.083c
-0.200d
-0.244d
0.467a,b
0.386b
0.553a
0.391b
0.107c
-0.195d
-0.246d
*ANOVA testecTagainst treatment  for  each  week  after treatment Initiated. Values given
 are mean P.  (g Oo m'^h"1) and letters  following  each  entry  Indicate significantly
 different *p<0.05) statistical  groups  according  to SNK multiple range test (read vertically
 with treatment for each experiment).
                                         111-28

-------
     Overall, the effects of linuron treatment on productivity of these two
plant species are remarkably similar to those observed for atrazine exposure.
Few previous studies have examined linuron toxicity for aquat'Ic plants.
However, Bielecki and Skrabka (1976) have shown that short-term (24 h) losses
of photosynthesis by Spirodela polyrrhiza were similar for linuron and sima-
zine (an s^triazine closely related to atrazine), although slightly less severe
for the latter compound. While the modes of action for these two herbicides
are somewhat different, as with atrazine both are photosynthetic inhibitors.
"he similarity of effects between linuron and the s-triazines is consistent with
a recent study by O'Brien and Prendeville (1979) wFich demonstrated that linuron
and simazine induced identical rates of electrolyte leakage from the cell membrane
of Lemna minor. Toxicities of linuron and atrazine for a wide range of terrestial
weeds have also shown to be relatively comparable (Stevenson and Confer 1978).

Linuron Effects on Plant Biomass
     At the end of this experiment total plant biomass was harvested from each
microcosm, and these data are summarized in Fig. 13. For £. perfpliatus there
was a clear trend of decreasing biomass with treatment at herbicide concentrations
2.50 ppb.  No effect was evident at 5 ppb, but significant loss of plant material
occurred at >50 ppb (p£0.05, 1-way ANOVA, SNK). In fact, treatment effects
separated inTo 3 groups (control and 5 ppb, 50 and 100 ppb, and 500 and 1000 ppb).
There was a small but insignificant increase in biomass for M. spicatum at 5 ppb,
although no such effect was apparrent in Pa data. A trend similar to that for
£. perfoliatus occurred for M^. spicatum, with mean biomass decreasing as treat-
ments increased from 50-1000 ppb, but significant losses (p<0.05) were seen
only at linuron 2.500 ppb. At 100 ppb the effect was marginaTly significant
(p^0.20). It appears that plant biomass is a less sensitive index of linuron
stress than Pa, possible because Pa and Rn tend to become adjusted, minimizing
effects on plant growth.


             IMPLICATIONS: HERBICIDES AND ESTUARINE VASCULAR PLANTS


                Relative Phytotoxicity of Herbicides for Plants

     The relationships between herbicide dosage and plant production or biomass
can be expressed in terms of an exponential dose-response model (BUss 1970;
Horowitz 1976).  In Fig. 14 we present mean Pa as a percent of vehicle control
plotted versus the logarithm of nominal herbicide concentration for the four
experiments performed in this study. This model provides a good descrip-
tion for the results of two experiments (atrazine/fl. spicatum and linuron/
P. perfoliatus), with r' _> 0.99. While some version of the logistic model might
Tfave been more appropriate for the other two experiments (Bliss 1970; Streibig
1980), the exponential model, nonetheless, accounted for > 94% of the variance
in those data (Fig. 14).

     Overall, the dose-response relations for all four experiments were simi-
lar with slopes and intercepts ranging from 36.8-48.9 and 110.5-147.4,
respectively. However, several differences are worth noting. The generally
greater sensitivity of £. perfoliatus to both herbicides is evident here by the

                                     111-29

-------
     10
 (0
 o
 o
 o
 i_
 O
•a
 o>
 w"
 o>
 CO

 o
m

o
Q.
O

O
CO
             (a)
        P. perfoiiatus

       (x  and  range)
    10
             (b)
        M. spicatum
          and range)
        control
50
100
500     1000
                            Linuron Dose, ppb
Figure 13.  Shoot biomass for a) P. perfoiiatus and b) M. spicatum harvested after
           5 weeks exposure to SHevels of linuron dosage with untreated controls
           Given are means and ranges for duplicate microcosms.  Asterisks indi-
           cate significant (p<0.05) differences  between treatments and consols
                                 111-30

-------
 I1
•
                        100
                        80
                        60
                     0)
                         40
                     o
                     O
                     JB
                     .O

                     0>
M   20
                                                              a) ATRAZINE
            E, perfoliatus
       .  Y- 121.3 - 41.3X
            (r2 - 0.95)
                                                                        M. spicatum
                                                                      Y-  14~7.4 -47.0X
                                                                        (r2 = 0.99)
                                                  I50 - 55 ppb
                                                            ol
o

to
o
CO
0.
c
CO
0>
                        100
                         80
                         60
                         40
                         20
                                                              b) LINURON
   P. perfoliatus
Y- 110.5 - 36.8X
    (r2 - 1.00)
  M. spicatum
Y-142.1 - 48.9X
   (r2 . 0.94)
                                               •so • 45
                                              10
                                        50   100
                                                  500   1000
                                          Log Herbicide Concentration (ppb)
                 Fig.  14. Regressions of mean apparent 02 production (as a % of  control) over
                         the first 4 weeks of treatment versus log of herbicide concentration
                         for £. perfoliatus and H. spicatum treated with a) atrazine and h)
                         linuron. Also shown are  values of ICQ (herbicide concentrations
                         anddhe"bicid """ 50% photosynthetlc inhibition) for  each species
                                                     111-31

-------
.$            fact that mean Pa's  for this  species  were  lower  than  for M.  spicatum  at  all
f            but the highest herbicide concentrations.  The  apparent  greater  tolerancce  of
             M.  spicatum to atrazine and Unuron exposure and  the  actual  stimulation  of its
:            production at low concentrations  might  lead to some speculations  concerning
             the proliferation of this species in  upper Chesapeake Bay  prior to  the general
             die-back of all submerged plants  (Bayley et al.  1978).  Using 150  (herbicide
             concentration producing 50% Inhibition) as an  index of  relative toxicity,
             we  also find that Unuron was  slightly  more effective than atrazine at reducing
             Pa  for both plant species. Combining  the results  of all four experiments in
             this statistical  model,  we find a reasonably strong regression  (rz  =  0.93),
             with ICQ = 72 ppb.  The  herbicide  concentration inducing incipient (1%) inhibi-
             tion (II.Q) can be inferred from  this model, and  values of IJ.Q range from
                                                                          .
             2-11 ppb.  Relationships  between  plant  biomass  (after  a  5 wk treatment) and
             herbicide  dosing concentration were  also  statistically  significant for all
             experiments,  with r2 _> 0»?3 for  all  but the  atrazine/M. spicatum study.  Values
             for It  ft and  Ig0 were similar for  Pa and  biomass  (TabTe 8) , wm en is  not
             surprising given the physiological relation  between production and final
             standing crop.

                  A  few previous  studies have used  the same  exponential dose-response model
             for herbicides  and submerged vascular  plants.  Forney  and Davis (1981) estimated
             the ranges of Ii g to be 1-14 ppb  and  that 159  to be  from  53-907 ppb  for
             E.  canadensis,  vl americana and  P_. perfoliatys  exposed  to  atrazine using a
             variety of growth parameters to  measure inhibition. They also found that N[.
             spicatum was  significantly  more  tolerant  of  herbicide stress  than the other
             species. Walsh  et al. (1982) reported  that Pa for ]_.  testudinum exhibited
             somewhat less sensitivity to atrazine  exposure  than observed  in the present
             study for  £.  perfoliatus and M.  spicatum, with  I£Q for  the seagrass being
             about 320  ppb.  The effects  of~atraz1ne on submerged plants reported by Correll
             et  al.  (1978) and Correll and Wu (1982) were not  treated in a dose-response
             model.  However, using a  linear Interpolation between  %  inhibition of  Pa
             observed at 2 test concentrations, we  estimate  159 values to  have ranged
             from about 150-450 ppb,  with the highest  value  for the  temperate seagrass, Z.
             marina. Yet,  at 12 ppb atrazine  mortality of _V. americana was estimated  to ¥e
             about 501  (I.e., ISQ »  12 ppb) 1n  a  later experiment  (Correll and Wu  1982).
                  In general  the effects of herbicide treatment  were more  severe  on Pa  than
             on biomass (B);  however,  the pattern  was similar  in both cases.   The  ratio
             Pa/3 provides a  measurement of normalized plant growth or  turnovr rate.   In  our
             study this ratio decreased with time  in  control microcosms, perhaps  indicating  a
             decrease in new  growth.   With herbicide  treatment the ratio Pa/B  exhibited a
             similar (but less pronounced) trend to that  for either Pa  or  B separately.


                                         Photosynthetic Recovery

                  In the above discussion we have  used an operational definition  of herbicide
             effect as the mean loss  of Pa compared to vehicle controls over a 5  wk treatment
             period. However, we noted earlier that recovery of  Pa was  evident for all
             4 herbicide/plant combinations at concentrations  ^  100 ppb.  In the case  of
             atrazine this recovery occurred even  in  the  face  of relatively constant  aqueous
             concentrations external  to the plant  tissue. Thus,  Integrated over an entire

                                                  III- 32

-------
J
Table 8. Summary of dose-response relations for herbicide treatment of submerged vascular
     plants.
Herbicide
Atrazine
Linuron
Combined
Plant
P. perfoliatus
M. spicatum
P. perfoliatus
M. spicatum
Combined
Apparent Photosynthesis (Pa)*#
r?
0.95
0.99
1.00
0.94
0.93
',.0 '50
4 55
11 117
2 45
8 80
5 72
Final
r2
0.96
0.53
0.95
0.93
0.72
Biomass
'1.0
1
3
1
16
3
(B)*
'50
30
91
25
137
61
         'Based on  linear  regression of  P* and B(express as % of vehicle controls) versus log
         nerbicide concentration, with  r" for least-square fit  line and Ij g and  I5Q being herbi-
         ci,,e concentration  (ppb) associated with  1% and  50% inhibition, respectively.
         #Mean values of Pa taken for the first A post-treatment weeks.
                                              111-33

-------
4 mo growing season, the effects might tend to be mitigated by regrowth. On
the other hand, experiments using short-term (ca. 2 h) exposures of submerged
plants to atrazine followed by sequential washes with herbicide-free water,
revealed that complete photosynthetic recovery was not apparent for 1-2 d
(Jones et al. 1982b). Whereas the former mode of recovery appears to involve
some sort of enzymatic detoxification within the plant, the latter results
simply from release of previously cell-bound atrazine. It seems that brief
exposures to atrazine of about 1 h may result in some photosynthetic inhibi-
tion for 1-2 d, while loss of Pa from extended (1-5 wk) bathing in herbicide
concentrations may be molified within a week. Che consequence of this
"buffering" effect is that both short and long-term dose-response experiments
yield similar results in terms of 150 and other measures of inhibition (c.f.
Jones et al. 1982b).


              Field Concentrations of Herbicides and Their Effects

     At the outset of this study the information on herbicide concentrations in
Chesapeake Bay was sparse, although the few existing data suggested that con-
centrations in shallow coves could approach 10 ppb (Correll et al. 1978).
Recently, Stevenson et al. (1983) have reported the results of an extensive survey
of atrazine and linuron concentrations in various regions of Chesapeake Bay
during the spring/summer seasons of 1980 and 1981. In this study herbicide con-
centrations (atrazine and linuron) were measured in water, suspended material
and sediments at 23 littoral zon3 stations in the main bay area; at 8-10 stations
along the salinity gradient of a major tributary (Choptank River); and at 4-5
stations along a runoff creek draining agricultural fields adjacent to the Choptank.
Bay-wide samples were taken bi-monthly from April-September, Choptank samples were
taken monthly during the same period and drainage from agricultural fields was
monitored prior to, during and for 12-24 h following runoff events.  Precipitation
patterns in 1980 were similar to the ten-year (1970-80) average, while 1981
was characterized by Intense rainfall in early May, with light sporadic precipi-
tation in June and July (Kemp et al. 1982).

     Tie ranges of atrazine and linuron concentrations observed in this survey
are summarized in Table 9. A transect down the main axis of the estuary revealed
that concentrations of either compound rarely exceeded 1 ppb, with highest values
in the brackish reaches. A similar transect along the main salinity gradient
of a major tributary (Choptank River/estuary), Indicated concentrations 1n 1980
similar to those 1n the open bay, while 1n 1981 maximum values were 9.3 and
3.0 ppb, respectively, for atrazine and linuron. The high atrazine concentra-
tion was actually from a freshwater station just above the head of tide. In a
creek and small estuarine cove connected to the Choptank and receiving direct
agricultural runoff, peak concentrations approached 50 ppb atrazine and 20 ppb
linuron, with highest values occurring during the May 1981 runoff events.
Generally, concentrations 1n May were on the order of 1-5 ppb, and those > 10 ppb
existed for only 2-8 h, being rapidly dissipated by dilution, adsorption and
degradation. These concentrations are similar to those reported elsewhere for
these coastal plain estuaries (Newby et al. 1978; Correll et al. 1978; Wu
1980; Zahnow and Riggleman 1980; Hershner et al. 1982).

     Using these herbicide data, a relatively extreme runoff scenario can be

                                      111-34

-------
Table 9. Summary of ranges of atrazine and linuron concentrations observed in
     water from Chesapeake Bay and'its primary and secondary tributaries.*

1.
2.

3.
SAMPLING
LOCATION
Horn Point
(Runoff Creek
and Cove)
Choptank Estuary
•Monthly Survey
Runoff Event
Chesapeake Bay
SAMPLING
PERIOD
May-Jul 1980
May 1981
May-Jul 1980
May 1981
Apr-Sep 1980
CONCENTRATION
ATRAZINE
0.1-18.3
0.7-46.0
0.0- 0.8
0.2- 9.3
0.0- 1.2
RANGE (ppb)
LINURON
0.0- 2.6
0.4-18.3
0.1- 1.2
0.3- 3.0
0.0- 1.3
*Data summarized from Stevenson et aT. (T983).
                                 111-35

-------
I!
             developed for submerged vascular plants growing 1n small coves and the overall
             effects of resulting herbicide concentrations on plant production can then be
             estimated using Fig. 14 and Table 8. In a wet spring such as that of 1981, we
             might find 2 major runoff events occurring in May spaced 2 wk apart and generating
             peak atrazine concentrations of 50 and 20 ppb, respectively, which last about
             6 h, followed by baseline concentrations of 1-5 ppb until July. A subsequent
             runoff event in July (after the planting of a second grain crop) might result
             in peak linuron concentrations of 10 ppb for 6 h followed by 1-3 ppb for the
             remainder of the month. Such environmental conditions could be expected to
             result in the following sequence of Pa inhibition: May -- 50% loss of Pa
             for 2 d, 5% loss for 12 d, 30% loss for 2 d, 5% loss for 15 d; June — 1% loss
             for 30 d; July -- 25% loss for 3d, 5% loss for 28 d. Integrated over the
             entire 3 mo period, this would amount to a mean of 6% inhibition of Pa using
             a simplistic additive approach.

                  It is of interest to note that two recent papers investigating atrazine
             effects on submerged plant growth (Forney and Davis 1981; Correll and Wu 1982)
             reached what appear to be opposite conclusions concerning the possible threat
             posed by this herbicide for submerged plant populations  in Chesapeake Bay.
             Whereas Forney and Davis (1981) stated that it is unlikely that atrazine "caused
             the die-off of plants in Chesapeake Bay," Correll  and Wu (1982) suggested that
             "there is every reason to be concerned that herbicides such as atrazine may be
             seriously impacting submersed vascular plant populations in estuaries." Despite
             the apparent difference in these two conclusions the experimental results from
             these two studies are, as we discussed previously, similar. However, these two
             statements may not be as contradictory as they seem.

                  The data and analyses which we have provided here indicate that 1-10% loss
             of submerged plant production might occur in certain estuarine regions as a result
             of runoff of herbicides such as atrazine and linuron over the ccurse of a growing
             season. Since these plant populations are stressed generally by reduced light and
             other perturbations (Kemp et al. 1982), such an herbicide-induced loss of pro-
             duction could add to the adversity of their environment, and therefore, the posi-
             tion of Correll and Wu (1982) would not be unreasonable. When these plants are
             existing at the margins of their adaptability range, small incremental stresses
             can undermine reproductive viability by shifting the physiological energy
             balance from net positive to negative.  However, as Indicated by Forney and
             Davis (1981), it is difficult to conclude from the results of this and other
             studies that herbicides, per se, caused the decline in Chesapeake Bay's
             submerged plants.  Further studies examining the short and long-term effects of
             herbicide stress on SAV reproduction may be warrented to resolve this issue.
                                                   111-36

-------
r
                                                   REFERENCES
                 1.  Adams, S.M. 1976, Feeding ecology of eelgrass fish communities. Trans.
                     Amer.  Fish. Soc. 105:514-519.

                 2.  Barsdate, R.J., M. Nebert, and C.P. McRoy. 1974. Lagoon contributions to
                     sediments and water of the Berino Sea, pp. 553-575. In D.W. Hood and
                     E.J. Kelly (eds.). Oceanography of the Bering Sea. UnTv. Alaska Press,
                     Fairbanks.

                 3.  Bayley, S., V.D. Stotts, P.F. Springer, J. Steenis. 1978. Changes in sub-
                     merged aquatic macrophyte populations at the head of the Chesapeake Bay,
                     1958-1975. Estuaries. 1:74-85.

                 4.  Bielecki, K. and H. Skrabka. 1976. Effect of some herbicides on photosyn-
                     thesis of Spirodela polyrhiza (Lemnaceae). Acta Agrobot. 29:59-68.

                 5.  Bliss, C.I. 1970. Statistics in biology, Vol. 2. McGraw-Hill,  London.

                 6.  Cartwright, P.M. 1976. General growth response of plant, pp. 55-82. Jii
                     L.J. Audus (ed.). Herbicides, physiology, biochemistry, ecology, Vol. 1.
                     Academic Press, London.

                 7.  Correll, D.L., J.W. Pierce and T-L Wu. 1978. Herbicides and submerged
                     plants in Chesapeake Bay, pp. 858-877. _In Amer. Soc. Civil Eng. (ed.).
                     Coastal Zone-78. ASCE Publ.

                 8.  Correll, D.L. and T.L. Wu. 1982. Atrazine toxicity to submersed vascular
                     plants in simulated estuanne microcosms. Aquatic Bot. (In press).

                 9.  Cunningham, J.J. 1979. Responses of microcosm communities containing sub-
                     merged aquatic vegetation to herbicide stress. MS Thesis. Univ. MD, Col-
                     lege Park.  66 pp.

                10.  Cunningham, J.J., W.M. Kemp, M.R. Lewis and J.C. Stevenson. 1982. Pat-
                     terns of ecological and physiological response and recovery of the sub-
                     merged macrophyte, Potamogeton perfoliatus, exposed to atrazine stress
                     in estuarine microcosms. (Unpubl. MS.).

                11.  Forney, D.R. and D.E. Davis. 1981. Effects of low concentrations of herbi-
                     cides on submersed aquatic plants. Weed Sci. 29:677-85.

                12.  Fowler, M.C. 1977. Laboratory trials of a new trlazine herbicide (DPX
                     3674) on various aquatic species of macrophytes and algae. Weed Res.
                     17:191-195.

                                                     111-37

-------
13.   Hawxby, K., B. Tubea,  J.  CWnby and E. Basler.  1977. Effects of various
     classes of herbicides  on  four species of algae. Pestic Biochem. Physiol.
     7:203-209.

14.   Hershner C., K. Ward,  J.  Illowsky.  1982.  The effects of atrazine on
     Zostera marina in the  Chesapeake Bay, Virginia.  Virginia Instit. Mar.
     Sci., Gloucester Pt.,  VA.  US Environ. Protect. Agency, Annapolis, MD.
     No. R805953.

15.   Hoffman, O.W. and T.L. Lavy.  1978. Plant competition for atrazine. Weed
     Sci.  26:94-99.

16.   Horowitz, M. 1976. Application of bioassay techniques to herbicide invest-
     igations. Weed Res. 16:209-215.

17.   Howard-Williams, C. 1981. Studies on the ability of a Potamogeton pectino-
     tus community to remove dissolved nitrogen and phosphorus compounds from
     TaTe water. J. Appl.  Ecol.  18:619-637.

18.   Jones, T.W., W.M. Kemp, J.C.  Stevenson and J.C. Means. 1982a.  Degradation
     of atrazinne in estuarine water/sediment systems and selected  soils. J.
     Environ. Qual. (In press).

19.   Jones, T.W., W.M. Kemp, P.S.  Estes and J.C. Stevenson. 1982b.  Atrazine up-
     take, phytotoxicity,  release  and short-term recovery for the submerged
     aquatic plant Potamogeton perfoliatus. (Unpublished MS).

20.   Kemp, W.M., M.L. Lewis, J.J.  Cunningham, J.C.  Stevenson, W.R.  Boynton.
     1980.  Microcosms, macrophytes, and hierarchies: Environmental  research
     in Chesapeake Bay, pp. 911-936. Jn^ J. Giesy (ed.).  Microcosm research in
     ecology. ERDA Conf. 781101.

21.   Kemp, W.M., J.J. Cunningham,  J. Perron, S. Garabrandt, M.R. Lewis, R.
     Hanson, R. Twilley, W.R.  Boynton, and J.C. Stevenson. 1981. Experimental
     observations of turbidity/light relations and their influence  on submerged
     macrophytes in northern Chessapeake Bay, pp. 2-29 _In_ W.M. Kemp  et al.
     (eds.).  Submerged aquatic vegetation in Chesapeake Bay. Univ.  Md. Cntr.
     Env. Est.  Res. Ref.  No.  HPEL 81-68, Cambridge, MD.

22.   Kemp, W.M., J.C. Means, T.W.  Jones and J.C. Stevenson. 1982. Herbicides
     in Chesapeake Bay and  their effects on submerged aquatic vegetation.
     Univ. Maryland Center Enviorn. Est. Stud. Ref. No.  1295-HPEL.  Horn Pt.
     Env. Labs, Cambridge,  MD. 82 pp.

23.   McConnell, W.J. 1962.  Productivity relationships in carboy microcosms.
     Limnol.  Oceanogr. 7:335-343.

24.   Newby, L.C., R.A. Kahrs,  K. Adams, and M. Szolics.   1978.  Atrazine resi-
     dues in the Chesapeake Bay.  Proc. N.E. Weed Sci. Soc. 32:339  (Abstract).
                                     111-38

-------
•I?
               25.   O'Brien, M.C. and G.N.  Prendeville. 1979. Effects of herbicides on cell
                    membrane permeability in Lemna minor. Weed Res. 19:331-334.

               26.   Odum, H.T. and C.M. Hoskin. 1958. Comparative studies on the metabolism of
                    marine waters. Publ. Inst. Mar. Sci. (Texas) 8:23-55.

               27.   Orth, R.J. 1977. The importance of sediment stability in seagrass communi-
                    ties, p. 281-300 In B.C. Coull (ed.) Ecology of marine benthos. Univ. S.
                    Carolina Press, Columbia.

               28.   Orth, R.J. and K.A. Moore. 1981.  Submerged aquatic vegetation of the Ches-
                    apeake Bay: Past, present and future. Trans. N. Amer. Widl. Nat. Res.
                    46:271-283.

               29.   Pruss, S.W. and E.R. Higgins. 1974. Effects of low levels of simazine on
                    plankton algae in a non-stratified lake. Proc. N.E. Weed Contr. Counc.
                    28:12^-131.

               30.   Ries, S.K. 1976. Subtoxic effects on plants, p. 313-344. In L.J. Audus
                    (ed.) Herbicides, physiology, biochemistry, ecology, vol ,~7. Academic
                    Press, London.

               31.   Stevenson, J.C. and N.M. Confer.  1978. Summary of available information on
                    Chesapeake Bay submerged vegetation. US Dept. Inter. FWS/OBS-78/66.
                    NTIS, Springfield,  VA.  333 pp.

               32.   Stevenson, J.C., W.R. Boynton, J.C. Means, W.M. Kemp, and R.R. Twilley.
                    1983.  Concentrations of the herbicides, atrazine and linuron, in agricul-
                    tural drainage, tributary estuaries and littoral zones of upper Chesapeake
                    Bay.  Report to US EPA, Annapolis, MD.

               33.   Streibig, J.C.  Models  for curve-fitting herbicides dose response data.
                    Acta Agric. Scandinav.  30:59-64.

               34.   Strickland, J.D.H.  and  T.R. Parsons. 1972. A Practical Handbook of Sea-
                    water Analysis. Fish. Res. Bd., Canada Bull. 167. Ottawa. 310 pp.

               35,   Sutton, D.L., D.A.  Durham, S.W. Bingham, C.L. Foy. 1969. Influence of sim-
                    azine on apparent photosynthesis  of aquatic plants and herbicide residue
                    re-noval from water. Weed. Sci. 17:56-59.

               36.   Vallentine, J.P. and S.W. Bingham. 1976. Influence of algae on amitrole
                    and atrzine residues in water. J. Can. Bot. 54:2100-2107.

               37.   Walsh, G.E. 1972. Effect of herbicides on photosynthesis and growth  of
                    marine unicellular algae. Hyacinth. Contr. J. 10:45-48.

               38.   Walsh, G.P., D.L. Hansen and D.A. Lawrence.  1982.  A flow-through system
                    for exposure of seagrass to pollutants.  Mar. Environ. Res. 7:1-11.
                                                    111-39

-------
39.  Wu, T.L. 1980. Dissipation of the herbicides atrazine and alachlor in a
     Maryland Cornfield. J. Environ. Qual. 9:459-465.

40.  Zahnow, E.W. and J.D. Riggleman. 1980. Search for linuron residues in
     tributaries of the Chesapeake Bay.J. Agric.  Food  Chem. 28:974-978.
                                      111-40

-------
APPENDIX
        111-41

-------
lAJlLt 1A  MSPONSC OF  AM-AKtNT  f MOTOSTN1HESIS Of PU1AMOGETO* TO ATKAZJNt
          CONClNThATlUNMr.F.fi. ).   1979 EXPt M«t NT?
          UNITS AK£ BhAHS  OxrUlN  UK CUBIC MlltH  Hk HUIJK
. TREATMENT
.CONTROL
.CONTROL
.CONTROL
.CONTROL
. flEAN
. ERROR
. VEH-CONTROL
.WEH-CONTROL
.UEH-CONTROL
•.UEH-CONTROL
. MEAN
. ERROR
.ATRHZINE 5
.ATRAZINE 3
.ATRAZINE S
.ATRAZINE S
. MEAN
. CRROft
.ATRAZINE SO
.ATRAZINE 30
.ATRAZINE 30
.ATRAZINE 30
. MEAN
. ERROR
IATRAZINE too
.ATRAZINE 100
.ATRAZINE 100
.ATRAZINE 100
. MEAN
. ERROR
.ATRAZINE 300
.ATRAZINE 300
.ATRAZINE 300
.ATRAZINE 500
•
. MEAN
. ERROR
.ATRAZINE 1000
.ATRAZINE 1000
.ATRAZINE 1000
.ATRAZINE 1000
. MEAN
. ERROR
.
1ANK.
11 .
11 .
22 .
22
.
•
If .
19 .
20 .
20 .
.
•
12 .
12 .
13 .
13 .
.
•
3 .
3 .
24 .
24 .
^
•
13 .
13 .
27 .
27 .
t
•
16 .
14 .
26 .
26 .
.
•
S .
3 .
10 .
10 .
.
•
j
.1*6
.244
.371
.428
.310
.107
.267
.293
.144
.233
.239
.064
.IBS
.246
.145
.187
.171
.041
.131
.229
.167
.242
.192
.032
.206
.216
.194
.282
.224
.039
.272
.392
.739
.083
.372
.273
.179
.233
.196
.287
.209
.030
4
.163
.173
.390
.319
.262
.111
.142
.119
.239
,19»
.180
.063
.203
.103
.143
.164
.133
.042
.130
.130
.297
.233
.197
.082
.092
.119
.206
.187
.131
.034
.270
.200
.201
.109
.193
.066
.130
.207
.187
.130
.184
.038
S
.396
.392
.303
.336
.357
.045
.307
.313
.338
.380
.339
.033
.270
.239
.279
.316
.276
.032
.303
.292
.357
.368
.430
.133
.297
.298
.297
.293
.297
.001
.313
.288
.440
.423
.366
.077
.230
.373
.258
.312
.301
.030
6
.404
.283
.260
.249
.299
.071
.247
.181
.429
.363
.305
.112
.322
.239
.433
.231
.316
.098
.307
.197
.543
.306
188
44
. .64
.049
.410
.221
.211
.191
.260
.218
.339
.338
.294
.066
.371
.196
.375
.236
.289
.083
uti •
7
.352
.391
.264
.372
.445
.162
.368
.346
.381
.641
.334
.117
.396
.404
.437
.426
.416
.019
.313
.483
.623
.639
.313
.131
.129
.171
.380
.431
.278
.ISO
.440
.338
.391
.364
.488
.086
.322
.303
.418
.483
.470
.041
B
.603
.643
.653
.701
.630
.040
.480
.539
.530
.393
.336
.047
.433
.331
.491
.345
.305
.042
• .221
.342
.235
.339
.284
.063
-.033
.043
-.032
-72
•>B
0
-.335
-.133
-.401
-.233
-.283
.107
-.191
-.107
-.323
-.164
-.183
.101
9
.416
.548
.?97
.514
.329
.039
.239
.237
.375
.333
.301
.064
.173
.142
.202
.111
.137
.040
-.143
-.077
-.017
-.039
-.069
.036
-.372
-.499
-.303
-.189
-.341
.130
-.593
-.491
-.494
-.489
-.317
.031
-.326
-.333
-.420
-.449
-.397
.068
10
.327
.333
.393
.369
.356
.031
.400
.593
.316
.361
.567
.038
.486
.346
.285
.348
.416
.121
.293
.313
.306
.314
.307
.010
.004
.081
.075
.105
.066
.043
-.342
-.298
-.335
-.395
-.347
.040
-.237
-.223
-.403
-.299
-.274
.069
li
.511
.341
.315
.629
.349
.035
.680
.737
.499
.325
.610
.116
.365
.618
.408
.449
.310
.098
.339
.346
.331
.348
.341
.008
.123
.138
.105
.173
.140
.031
-.218
-.155
-.212
-.208
-.198
.029
-.170
-.115
-.203
-.ISO
-.139
.031
                                           111-42

-------
TAbll Ifc  kFSFONSE OF  KLSMkAUUN OF  F UIAMOOE TC1N TO ATkA/INC  CONCtNTKAT UJNi
          (P.Kb.>.   197V  EXf CRIMf Nl .
          UNITS ARE GKAflB  OXYWN  F•t^  (.u^M,   METEk' Mk HQUk.
.
. TkiAlfltNT
.CONTROL
.CONTROL
.CONTROL
.CONTROL
. MEAN
. ERROR
.UEH-CONTROL
.VEH-CONTKOL
.VEH-CONTROL
.UEH-CONTROL
. HE AN
. ERROR
.ATRAZINE 5
.ATRAZINE 5
.ATRAZINE S
.ATRAZINE S
. HEAN
. ERROR
IATRAZINE so
.ATRAZINE SO
.ATRAZINE SO
.ATRAZINE SO
. HE AN
. ERROR
.ATRAZINE 100
.ATRAZINE 100
.ATRAZINE 100
.ATRAZINE 100
. HE AN
. ERROR
! ATRAZINE 300
.ATRAZINE 500
.ATRAZINE 500
.ATRAZINE 500
. MEAN
. ERROR
.ATRAZINE 1000
.ATRAZINE 1000
.ATRAZINE 1000
.ATRAZINE 1000
. MEAN
. ERROR
.
TANK.
11 .
11 .
22 .
22 .
.
•
19 .
19 .
20 .
20 .
t
•
12 .
12 .
13 .
13 .
t
•
3 .
3 .
24 .
24 .
,
•
13 .
13 .
27 .
27 .
%
•
16 .
16 .
26 .
26 .
.
•
5 .
S .
10 .
10 .
,
'

i
-.244
-.287
-.340
-.419
-.322
.075
-.375
-.406
-.206
-.265
-.313
.094
-.298
-.331
-.234
-.27?
-.284
.041
-.218
-.268
-.311
-.342
-.285
.054
-.277
-.294
-.264
-.343 -
-.294
.033
-.317
-.381
-.821
-.282
T.450
.231
-.207
-.261
-.336
-.371
-.*89
.058

4
-.273
-.271
-.400
-.375
-.330
.067
-.349
-.300
-.316
-.282
-.312
.028
-.384
-.321
-.288
-.286
-.320
.046
-.238
-.242
-.396
-.347
-.311
.073
-.214
-.224
-.330
-.313
-.275
.047
-.349
-.297
-.403
-.484
-.348
.057
-.230
-.255
-.342
-.303
-.302
.049

^
-.349
-.376
-.318
-.373
-.339
.027
-.403
-.407
-.341
-.370
-.380
.031
-.364
-.371
-.325
-.347
-.357
.021
-.332
-.333
-.508
-.549
-.431
.114
-.303
•-.313
-.349
-.371
-.334
.032
-.345
-.347
-.432
-.479
-.406
.041
-.282
-.375
-.397
r.402
-.377
.048

4,
-.382
-.331
-.340
-,321
-.343
.027
-.311
-.293
-.402
-.375
-.346
.051
-.377
-.366
-.451
-.319
-.378
.055
-.306
-.249
-.575
-.543
-.418
.165
-.133
-.088
-.406
-.317
-.241
.146
-.319
-.283
-.497
-.489
-.397
.112
-.357
-.267
-.421
-.384
-.334
.069
UtLK
T
-.361
-.443
-.401
-.492
-.424
.034
-.463
-.344
-.438
-.493
-.490
.040
-.408
-.418
-.430
-.419
-.419
.009
-.348
-.316
-.586
-.611
-.465
.155
-.141
-.179
-.386
--.401 -
-.282
.129
-.473
-.475
-.487
-.576
-.303
.049
-.461
-.472
-.464
-.496
-.436
.030

•a
-.677
-.677
-.774
-.774
-.725
.056
-.713
-.713
-.674
-.674
-.693
.023
-.703
-.703
-.620
-.620
-.661
.048
-.584
-.584
-.707
-.707
-.645
.071
-.463
-.463
-.485
-.485
-.474
.013
-.550
-.550
-.527
-.327
-.538
.013
-.311
-.309
-.531
-.531
-.520
.183

Cj
-.339
-.360
-.647
-.332
-.374
.049
-.740
-.724
-.668
-.710
-.710
.031
'.643
-.682
-.387
-.363
-.619
.054
-.612
-.503
-.548
-.611
-.368
.053
-.454
-.536
-.542
-.728
-.565
.116
-.476
-.523
-.476
-.513
-.497
.025
-.327
-.346
-.374
-.425
-.379
.049

PO
-.303
-.527
-.632
-.641
-.376
.071
-.627
-.642
-.584
-.619
~.AtB
.025
-.392
-.595
-.403
-.503
-.523
.091
-.411
-.430
-.389
-.393
-.506
.099
-.154
-.204
-.405
-.392 —
-.289
.128
-.330
-.326
-.493
-.494
-.416
.091
-.196
-.223
-.293
-.263
-.235
.038
t
1!
-.377
-.633
-.674
-.694
-.644
.032
-.721 !
-.749
-.376
-.393
-.445
.095
-.613 !
-.641
-.446
-.521
-.560
.081
-.437
-.457
-.794
-.800
-.422
.203
-.205 !
-.194
-.488
-.474
-.340
.143
-.189 I
-.235
-.374
-.379
-.294
.097
-.143 !
-.132
-.144
-.143 .
-.144 .
.007
                                             111-43

-------
TABLE 1C  KfStONSC OF GROSS fHOIUSTNTHtSIS  Of  F'OIAMOGETON TO ATKAZJW
          CONCLNTf>ATlONSU'.l .*. ) .   1171 EXfEhlHFNT.
          UNITS ARC OkAMS OXYGEN F'Ef. CUt^lC  METEk UK HOUR.
. TRLATnENI
.CONTROL
.CONTKOL
.CONTROL
.CONTKOL
. MEAN
. ERROR
.VEH-CONTROL
.VEH-CDNTKOL
.WEH-CONTRQL
. VEH-CONTROL
. MEAN
. ERROR
.ATRAZINE 3
.ATRAZINE S
.ATRAZINE S
.ATRAZINE S
. HE AN
. ERROR
.ATRAZINE 30
.ATKAZINE 30
.ATRAZINE 30
.ATRAZINE 30
. MEAN
. ERROR
TANK.
11 .
11 .
22 .
22 .
.
•
IV .
19 .
20 .
20 .
t
•
12 .
12 .
13 .
13 .
t
•
3 .
3 .
24 .
24 .
(
.
4
.140
.646
.84?
j.ois
.76:
.211
.79?
.842
.433
.624
.478
.191
.405
.709
.473
.567
.588
.098
.436
.404
.602
.721
.391
.117
4
.sso
.552
.950
.643
.724
.204
.630
.536
.701
.595
.616
.068
.740
.332
.546
.563
.600
.093
.492
.468
.852
.718
.632
.185
e>
.912
.918
.748
.861
.860
.079
.872
.683
.835
.898
.872
.027
.780
.738
. .734
.829
.J73
.040
.769
.761
1.269
1.336
1.034
.312
fe
.937
.746
.736
.699
.780
.107
.683
.595
.992
.887
.789
.182
.851
.731
1. 085
.697
.846
.172
.736
.346
1.349
1.267
.974
.394
UF.EK
-f
.858
1.211
.826
1.261
1.039
.229
1.017
1.306
1.222
1.333
1.220
.143
.967
.990
1.039
1.012
1.002
.031
.802
.923
1.443
1.494
1.166
.334
ft
1.512
1.391
1.736
1.784
1.666
.112
1.478
1.337
1.473
1.338
1.306
.036
1.438
1.316
1.360
1.414
1.432
.063
1.036
1.139
1.224
1.328
1.187
.121
4
1.210
1.333
1.503
1.288
1.333
.124
1.295
1.250
1.310
1.327
1.-J93
.033
1.073
1.097
1.023
.899
1.023
.089
.712
.427
.730
.817
.726
.079
10
1.232
1.273
1.477
1.467
1.362
.128
1.479
1.491
1.334
1.429
1.433
.071
1.313
1.379
.630
1.031
1.149
.243
.8*9
.918
1.130
1.143
1.013
.142
II
1.316
1.426
1.438
1.600
1.430
.116
1.690
1.813
1.306
1.355
1.541
.249
1.422
1.313
1.041
1.176
1.294
.211
.930
.985
1.444
1.468
1.212
.283
.ATRAZINE
.ATRAZINE
.ATRAZINE
.ATRAZINE
. MEAN
. ERROR
'.ATRAZINE
.ATRAZINE
.ATRAZINE
.ATRAZINE
•
. MEAN
. ERROR
.ATRAZINE
.ATRAZINE
.ATRAZINE
.ATRAZINE
'. MEAN
. ERROR
•
100
100
100
100


300
300
300
300


1000
1000
1000
1000


13
13
27
27


16
16
26
24


3
3
10
10


. .593
. .628
. .363
. .762
. .434
. .088
. .716
. .926
. 1.888
. .481
. 1.003
. .418
. .469
. .619
. .667
. .807
. .613
. .117
.392
.432
.696
.624
.336
.147
.738
.613
.743
.703
.710
.049
.452
.363
.666
.374
.406
.093
.721
.737
.786
.814
.744
.043
.824
.773
1.043
1.094
.934
.139
.444
.899
.813
.873
.828
.100
.379
.172
.978
.443
.348
.330
.704
.414
1.033
1.023
.830
.222
.871
.370
.943
.794
.737
.171
.334
.421
.919
.992
.471
.330
1.102
1.222
1.073
1.370
1.192
.135
1.168
1.164
1.068
1.180
1.109
.072
.393
.492
.447
.731
.471
.044
.433
.418
.338
.485
.449
.117
.243
.323
.420
.379
.344
.270
.243
.231
.454
.830
.430
.270
.073
.240
.173
.230
.179
.077
.131
.149
.103
.144
.134
.017
.219
.347
.442
.433
.470
.213
.148
.138
.337
.294
.233
.094
.037
.092
.010
.049
.034
.034
.412
.429
.788
.834
.414
.227
.044
.174
.312
.322
.213
.131
.031
.049
-.004
.030
.042
.028
                                            111-44

-------
 TAU.E !!•  kCSMIN'jL Of !-!> (-aim II' I OIAHOM TUN 10 AT(.«?INC CONCl NTMT lONi
           . HOU*.
   THATHLNT
                 TANK.
                                                        UCEK
.CU4UOL
.CONTKOL
                  It
                  II
                  22
                 19
                 19
                 30
                 20
 .CONTROL

 .  nt AM
 .  CkROh

 .VCH-CnNTROL
 .UCM-CON1ROI
 .vrri-CONTROL
 .VEM-CQNIROL

 .  fit AN
 .  ERROR
 .AT(M2iNE     9    11
 .nrftAtriNC     9    i;
 .AIR4ZINE     9    13
 .ATRAZINE     9    13

 . HE AN
 . ERROR

 .'ATRAZINE   90     3
 .ATRAZINC   SO     3
 .ATRAZINE   SO    24
 .ATRAZINC   SO    24

 . flEAN
 . ERROR

 .ATRAZINE  100    IS
 .ATRAZINE  100    IS
 .ATRAZINE  100    27
 .ATRAZINC  100    27


 '. ERROR
 •
 .ATRAZINC  SOO    14
 .ATRAZINE  SOO    16
 .ATRAZINC  SOO    26
.ATRAZINC  SOO    24

. MEAN
. ERROR
.811
.RIO
1.091
1.021
.943
.134
.712
.711
.499
.939
.772
.122
.&OO
.638
.975
.851
.744
.178
.407
.397
.fl?0
.70U
.582
.214
1.073
1.043
.953
.906
.994
.078
.742
.749
1.050
1.027
.902
.158
LOSS
.ess
.741
.776
.643
.134
.794
.414
1.047
.948
.841
.199
.975
1.334
.428
1.143
1.03?
.290
.79S
1.004
1.749
1.299
1.091
.237
,891
,990
.844
.904
.898
.044
.473
.794
.784
.883
.774
.087
.841
.979
.923
.931
.9?0
,osr
.350
.327
.941
.449
.427
.109
1.048
1 .019
.938
.eee
.972
.073
.997
.924
.884
.904
.918
.031
. BUA
.855
.744
.904
.853
.043
.943
.998
.844
.889
.913
.944
. .431
. .743
. .420
. .488
. .470
. .097
.401
. .894
. .537
. .708
. .479
. .139
. .744
. .739
. .733
. .822
. .799
. .042
. .838
. 1.029
. .900
, .301
, .772
, .322
.529
.321
.497
.373
.480
.110
.904
.937
.730
.471
.419
.119
.430
.931
.989
.397
.337
.077
.774
.473
.499
.237
.331
.224
.742
.444
.858
.841
.774
.104
.919
.872
1.094
1.039
.980
.103
.980
.932
.831
.793
.894
.084
.898
.830
1.019
.883
.897
.084
.894
.493
1.004
.787
.824
.144
1.003
.791
.944
.932
.917
.090
1.072
.337
1.010
.497
.834
.247
.813
.770
.722
.491
.749
.034
.971
.947
1.014
1.017
.993
.027
.90S
1.328
1.043
1.044
1.145
.271
.801
.933
.984
1.07S
.934
.114
.930
1.179
.803
.979
.972
.134
.444
.739
.792
.879
.747
.097
.378
.584
.332
.479
.444
.113
-.114
.097
-.044
.148
.014
.124
-.409
-.278
-.741
-.480
-.532
.203
.272
.208
.344
.197
.293
.048
-.237
-.133
-.031
-.044
-.121
.093
-.819
-.931
-.339
-.240
-.442
.299
-1.244
-.939
-1.038
-.933
-1.044
.142
.821
.918
.707
.492
.784
.104
.713
.733
.320
.330
.424
.113
.024
.397
.189
.248
.219
.199
-.977
-.914
-.717
-.800
-.832
.114
.922
.944
.874
.842
.904
.044
.774
.737
.414
.439
.394
.197
.410
.814
.213
.349
.301
.243
-1.193
-.440
-.547
-.349
-.732
.283
.ATRAZINC 1000
.ATRAZINE 1000
.ATRAZINE 1000
.ATRAZINE 1000

.  MEAN
.  ERROR
                 3 .  .843     .943     .887   1.039   1.132
                 3 .  .949     .812    1.000    .734   1.044
                10 .  .983     .947     .490    .891    .901
                10 .  .774     .495     .774    .447    .978
                                 .414   -.997  -1.209  -1.189
                                 .344   -.948   -.991   -.871
                                 .408  -1.123  -1.344  -1.424
                                 .30°  -1.094  -1.137  -1.049
                      .733
                      .142
.414
.114
.804
.128
.837
.133
1.032
 .081
.383  -1.042  -1.142  -1.103
.204    .099    .143    .198
                                       111-45

-------



1
IO
4_>
cr
01
c
o
u

f—
IO
c_
Ol
Ol
0


TJ
O>
t/>
C
ex
X
Ol
in • o

ex vw
o •<-
t- o
U Ol
*o ex
E v>
CM Ol
(. 10
o c

•v>»
f~* I/I
E 01
V> "O
o
U V)
O Ol
t. in
u 01
*E 4->
Js
CJ) IO
•<— ^ Q^
in c
1/1 ••-
E «n
O «-
•^ Oi
ja .0
E
c jC
ID
"ex 01
c
o°£
10
3 O
t/)
in
• c
 0
in r>-
o o



•— s.5^
CO CM
x-^^^
^T ^*
vo r-
• •
O O



^~.^—
CM CO
>_^>— '
CM ^>
in in
9 9
oo


^^*~*
o^ o
^ CM
*^_ J-*-_J
r"«» r^t
^ «o
0 c
oo



*
I-H oo CM ao
PH i-H CM CM

o i— * v—t in
10 r*^ 10 vo
0 0 O O




o> o
^».x
r^ 01
si



3
IO
•pv
^_
o

t_
ai
ex
•
«»-l


r-ieo cM«*-en IP^CM CM-~OO to^en ooo loo
*•» O CM CO in JO "* r». VO in Icn l/> OOO lo O
oo COCOCM |roo vo^-co («»--H ooo |oo
•H -fl -ft +1





O OOO Ito- Cft«-H |O t-l CM (CM
« r^co }in r—o t-i oo lo
• • • 1 • •• 1 • •• I*
o c\j^- (m ^^r (^ oo (o





^ ^-r^ KD f-io H-H 00*0 KVJ
to r**. t-H to- csj oo kn ro csj Ico
O *-Jro (c\j cor*! (in roin (^






rn locsi |a^ o^r tesj ^oo |ro
*** l^> *cfr \&\ ^^ ^3 tf^ ^o ^J |O^
* •* ]• •• • •• [•
o cviif) (r^ r**ir> •-* oo O* (oo
rH i-H





ro ^-t co |O ^ r*«» Ko »™« ^J |r*««
LO vo ^-< Kft ir> r** Ko vo n M-
• • * (• •• I • •• 1 •
O ^H ^1 |p^ ^^ C^ (00 CO CO |CO
*^ f" ^ 1^™^





-» oo vo |r* co CM • loo in co ta-
ft cni-i En coo Ko co en Ko
* •• 1 • •• I * • * 1 •
O c\)^*- (co om |c\j ^j-o Ir^
^H rH I^H ^H CM «"H





«± in ro ^- CT> «—» ^— * r*- co o co vo |£* ^t- 10 ^f CM o |cp vo
iO ^D CO CD r^ *^ ko co co co 10 CM jco i/) o> r*** ^H co •*"* CD
DO CMCNJCMCM ICMO cn^-c^vo (T^CM 004*0*00 (r^ c\J
4*t ^1 4^1 i I



S C9> ^T O ^** C3 GA Cn C2
!/l *^.^ t/7 !"«• ^ IO *v.^ (/)
00 Ol "»-. Ol Ch Ol
•fl -^Ol +1 C7>OI +1 "•v.OI -H
in 2 "-v.3 •— i 3
X ^*— x V0«^ x CM«^ x











I I 1-46

-------
•
                         a
                         3
                          Ol

                         S
                                   Ol
   o
   o
   o
   o
   o
   in
                                        o
                                        in
                                        m
^o
 o  u


 II
                                        6
                                        Ol


                                        IB
   Ol

   o
   01
   o,
   vo
                                                 CM en
                                                 co co
o co


o o
                                                             0
                                                             in
                                                                         «.: co
                                                       O    CM -ll
O Oil
O Ol


o cal
                                                 O O
                                                 o cri
                                                 in *»•
                                                             in in
                                                       o    -H ^H
                                                                         in
                                                                         CVJ (Vl
                                                                                     oo uo
                                                                                     r^ co
                                         R
                                                                                     0 0(0
                                                 O 0
                                                   I VO

                                                   I CO
                                                       O    CM «-tl
                        oo *»•
                        CO CM
                                   CO CO t-H    Cft '


                                   to coin    en •
                                                 O O|O    CO r-l
                        *t vot
                        (M .-I I
O CO
co r»»
                                                 in O
                                                 co «»•
                                                 O 0
                                                 CO CO
            O^ *^ JT"**    ^O ^J1
            ^- Olcvj    «l- vo


            CM C«J|cM    «f ^
                                                                                    oo
                                               vo a>|co
                                               in evifer
                                                 •   •   •
                                               in intf>
                                                 O 0
                                                       oo    cr> mlcsi    co vo
                                                             C3 r^ Ic^    c? co
                                     I  VO|
                                     i  col
                                                                                    co
                                                        n

                                                       i x
                                                      n


                                                     I X
                                                 o> o
            f^ Ol

            >-. Ol
            VO 2
            oo 
            m 3B
                                   en oo
                                   r-.

                                   en  01
f-

o 01


C0"s-*
o

"o.
v>
                                                                                      111-47

-------
f  *
                                                                                       PO I-H
                                                                                       CO CO
                                                                                       O O
                                                                                      ro
       CM   O in
                                                                                      PO ro
                                                                                       •  •
                                                                                      O O
            PO vokn
            vo OOKVI
                                                                                            o    -H
                                                                                       j oo
                                                                                       • PO
CO PO

do
                                                                                                PO vokn
                                                                                           O   •-• ml
                                                                                    O O
                                                                                    o o
                                                                                               vo
                                                                                   o o
                                                                                              00 OOt
                                                                                              CM VO
                                                                                              «M CM
                                                                                                 00
                    CL
                    (0
                                                                                                         O)
                                                                                                     X
                                                                                                         VI
                                                                                                         0)
                                                                                                         Ol
                                                                                                         a
                                                                                   E

                                                                                  •4->
                                                                                  
-------
TAULf 4A  fcCSf-OWSf  Of  AWAMNT PHOTOSrWTHfSIS Or KlIanlllJClON  10 LINIIKON
          CONCENTKflrIDNSlC.f .b.).  1 9BO  fXfEK1MENI.
          UNITS ARC  GRAMS OXYGEN FTK  CU&IC  ME UK PER MOUIi.
...
... .
. TREATMENT
! CONTROL
.CONTROL
.CONTROL
.CONTROL
.CONTROL
.CONTROL
.CONTROL
.CONTROL
. HEAN
. ERROR










.VEH-CONTROL
.VEH-CONTROL
.VEH-CONTROL
.VEH-CONTROL
. MEAN
. ERROR
.LINURON
.LINURON
.LINURON
.LINURON
. HEAN
. ERROR
.LINURON
.LINURON
.LINURON
.LINURON
. MEAN
. ERROR
I LINURON
.LINURON
.LINURON
.LINURON
. HEAN
. ERROR
.LINURON
.LINURON
.LINURON
.LINURON
. HEAN
. ERROR
.LINURON
.LINURON
.LINURON
.LINURON
. MEAN
. ERROR


S
3
S
S


50
50
30
SO


100
100
100
100


300
500
300
500


1000
1000
1000
1000


'
TANK.
IV .
IV .
6 .
6 .
IS .
IS .
26 .
26 .
.
•
22 .
22 .
24 .
24 .
,
•
17 .
17 .
IB .
IB .
t
•
1 .
1 .
4 .
4 .
t
•
a .
8 .
21 .
21 .
9
'
2 .
2 .
7 .
7 .
t
•
13 .
13 .
20 .
20 .
t
*

3
.183
.169
.212
.164
.112
.092
.176
.161
.13V
.03V
.31V
.314
.266
.237
.28V
.032
.225
.218
.236
.240
.230
.010
.237
.241
.233
.163
.218
.037
.330
.292
.198
.137
.239
.088
.180
.146
.309
.296
.233
.082
.170
.172
.183
.149
.168
.014

4
.027
.013
.08V
.093
.030
.039
.024
-.037
.033
.042
.047
.047
.12V
.134
.08V
.04V
.003
.016
.087
.072
.044
.041
.043
.143
.035
.043
.066
.052
.105
.128
.063
.080
.094
.028
.037
.047
.031
.083
.054
.020
.039
.004
.074
.086
.051
.037

S
.260
.237
.191
.238
.134
.136
.234
.233
.213
.043
.282
.283
.340
.308
.304
.027
.189
.170
.235
.266
.215
.044
.189
.22V
.127
.122
.167
.051
.166
.208
.247
.213
.208
.033
.132
.153
.192
.243
.183
.043
.076
.085
.293
.236
.178
.113

fe
.140
.163
.260
.290
.123
.161
.377
.413
.244
.110
.324
.336
.411
.418
.372
.04V
.212
.201
.273
.330
.234
.060
.183
.187
.232
.242
.211
.030
.233
.271
.173
.224
.226
.040
.154
.168
.259
.267
.212
.039
.139
.187
.320
.318
.246
.083
UEEK
7
.173
.081
.243
.179
.223
.163
.428
.420
.239
.124
.299
.072
.348
.134
.216
.128
.104
.037
.213
.134
.122
.073
-.042
-.123
-.096
-.182
-.111
.038
-.189
-.235
-.213
-.314
-.243
.035
-.187
-.243
-.246
-.379
-.264
.081
-.358
-.175
-.375
-.206
-.278
.103

8
.210
avv
.301
.2VS
.227
.293
.602
.607
.342
.167
.246
.371
.247
.377
.310
.074
-.143
-.003
.316
.360
.132
.246
-.136
.012
-.205
-.032
-.090
.099
-.273
-.110
-.324
-.099
-.201
.114
-.217
-.166
-.370
-.235
-.252
.086
-.389
-.369
-.238
-.237
-.313
.077

9
.226
.238
.348
.333
.293
.307
.392
.606
.371
.146
.437
.473
.440
.441
.433
.016
.318
.347
.438
.443
.386
.064
.196
.198
.120
.147
.163
.038
.098
.113
.014
.066
.073
.044
-.137
-.146
-.135
-.141
-.143
.008
-.283
-.306
-.191
-.202
-.246
.038


-------
TAfcLE 4k  RESfONSC  Of M. 6MI.AT ION Or  PDTAMOGCON TO IINUKUN CflNCtNTKATlONS
          (P.P.b.).  1VUO EXKtMMLNl
          UNITS AKC OK«rlS OXTGLN ItK CUB'C   Mt.TEK I£K HOUK.
t
(
. TKIATMCNT
.CONTROL
.CONTROL
.CONTROL
.CONTKOL
.CONTROL
.CONTROL
.CONTROL
.CONTROL
. HE AN
. ERROR










luEH-CONTROL
.WEH-CONTROL
.UEH-CON1ROL
.VEH-CONTROL
. HEAN
. ERROR
.LINURON
.LINURON
.LINURON
.LINURON
. HEAN
. ERROR
.LINURON
.LINURON
.LINURON
.LINURON
. HEAN
. ERROR
.LINURON
.LINURON
.LINURON
.LINURON
. MEAN
. ERROR
.LINURUN
.LINURON
.LINURON
.LINURON
'. HEAN
. ERROR
! LINURON
.LINURON
.LINURON
.LINURON
! HEAN
. ERROR


5
5
3
5


SO
SO
so
so


100
100
100
100


500
500
500
500


1000
1000
1000
1000



TANK.
19 .
19 .
6 .
6 .
IS .
IS .
26 .
26 .
t
•
22 .
22 .
24 .
24 .
^
•
17 .
17 .
18 .
IB .
.
•
1 .
1 .
4 .
4 .
.
•
8 .
a .
21 .
21 .
.
•
2 .
2 .
7 .
7 .
t
•
13 .
13 .
20 .
20 .
,
*

3
-.313
-.290
-.249
-.227
-.201
-.193
-.404
-.390
-.283
.081
-.352
-.354
-.337
-.350
-.353
.003
-.287
-.260
-.348
-.331
-.306
.040
-.112
-.336
-.301
-.235
-.296
.043
-.248
-.232
-.256
-.230
-.241
.013
-.242
-.218
-.332
-.313
-.276
.055
-.317
-.317
-.319
-.307
-.315
.005

A
-.331
-.309
-.240
-.205
-.214
-.165
-.472
-.446
-.298
.113
-.308
-.320
-.360
-.355
-.336
.026
-.309
-.290
-.321
-.307
-.307
.013
-.245
-.317
-.287
-.262
-.278
.031
-.250
-.234
-.232
-.221
-.239
.015
-.188
-.172
-.314
-.280
-.238
.069
-.372
-.208
-.326
-.307
-.303
.069

"5
-.366
-.374
-.275
-.282
-.224
-.226
-.508
-.522
-.347
.118
-.400
-.410
-.426
-.433
-.417
.013
-.330
-.330
-.358
-.354
-.343
.015
-.32C
-.334
-.253
-.246
-.290
.047
-.262
-.258
-.291
-.280
-.273
.015
-.214
-.235
-.322
-.325
-.274
.038
-.335
-.306
-.333
-.353
-.338
.023

If
-.322
-.347
-.279
-.308
-.205
-.219
-.496
-.522
-.337
.117
-.403
-.459
-.429
-.464
-.439
.026
-.330
-.360
-.390
-.443
-.381
.048
-.296
-.336
-.285
-.303
-.305
.022
-.260
-.306
-.268
-.281
-.279
.020
-.207
-.230
-.317
-.330
-.271
.062
-.297
-.299
-.339
-.395
-.332
.046
UEEK.
T
-.435
-.444
-.346
-.377
-.242
-.279
-.689
-.565
-.425
.145
-.661
-.752
-.594
-.683
-.672
.065
-.447
-.557
-.616
-.611
-.538
.079
-.348
-.387
-.309
-.378
-.355
.035
-.316
-.393
-.430
-.399
-.385
.049
-.252
-.218
-.371
-.362
-.301
.077
-.522
-.522
-.451
-.451
-.486
.041

e
-.385
-.297
-.331
-.247
-.247
-.206
-.526
-.460
-.337
.112
-.551
-.483
-.454
-.379
-.467
.071
-.570
-.376
-.452
-.410
-.452
.085
-.248
-.218
-.377
-.187
-.257
.083
-.372
-.235
-.314
-.237
-.294
.061
-.187
-.163
-.260
-.232
-.211
.044
-.376
-.337
-.261
-.2 4
-.297
.073

9
-.377
-.355
-.332
-.298
-.270
-.253
-.556
-.539
-.372
.116
-.544
-.520
-.423
-.417
-.476
.065
-.346
-.346
-.500
-.483
-.419
.085
-.300
-.290
-.244
-.231
-.266
.034
-.223
-.212
-.231
-.253
-.233
.020
-.168
-.139
-.176
-.143
-.162
.013
-.314
-.308
-.208
-.193
-.236
.063

10
-.308
-.323
-.291
-.313
-.264
-.278
-.369
-.563
-.364
.126
-.486
-.499
-.408
-.427
-.455
.044
-.263
-.J07
-.454
-.480
-.376
.107
-.274
-.290
-.268
-.282
-.276
.010
-.217
-.246
-.298
-.315
-.269
.045
-.111
-.133
-.180
-.203
-.137
.042
-.269
-.281
-.173
-.189
-.228
.033

II
-.356
-.351
-.339
-.342
-.294
-.279
-.571
-.581
-.389
.119
-.486
-.479
-.415
-.393
-.443
.046
-.368
-.367
-.478
-.493
-.427
.069
-.350
-.344
-.336
-.294
-.331
.0-3
-.221
-.223
-.296
-.283
-.237
.039
-.103
-.101
-.207
-.177
-.147
.033
-.248
-.237
-.204
-.191
-.223
.032
                                          111-50

-------
ti?
                            TA&LE  4C   RESPONSE  OF  OROSS PHOTOSYNTHESIS Of  POTAMUC.E TUN TO UNUNUN
                                      CONCENTkATIONS(P.F>.k. > .   ]?HO EXPERIMENT.
                                      UNITS  AKE GKAMS OXYGEN  F'CK  CUBIC   METER ftK HOUK.
                                                                                  •It If,
                              TKEAJMENT
                                            TANf,.
.CONTROL
.CONTROL
.CONTROL
.CONTROL
.CONTROL
.CONTROL
.CONTROL
.CONTROL
. HE AN
. ERROR










.VEH-CONTROL
.VEH-CONTROL
.WEH-CONTROL
.WEH-CONTROL
. MEAN
. ERROR
.LINURON
.LINURDN
.LINURON
.LINURON
. MEAN
. ERROR
.LINURON
.LINURON
.LINURON
.LINURON
. MEAN
. ERROR
.LINURON
.LINURON
.LINURON
.LINURON
•
. MEAN
. ERROR
.LINURON
.LINURON
.LINURON
.LINURON
. MEAN
. ERROR
.LINURON
.LINURON
.LINURON
.LINURON
. MEAN
. ERROR


S
s
3
3


30
30
30
30


100
100
100
100


300
300
300
300


1000
1000
1000
1000


19 .
19 .
6 .
6 .
IS .
13 .
24 .
26 .
.
•
22 .
22 .
24 .
24 .
.
•
17 .
17 .
IB .
18 .
t
•
1 .
1 .
4 .
4 .
•
•
S .
8 .
21 .
21 .
t
•
2 .
2 .
7 .
7 .
9
•
13 .
13 .
20 .
20 .
.
•
.622
.571
.561
.4e:
.393
.362
.742
.707
.335
.237
.812
.809
.763
.747
.783
.03?
.626
.381
.723
.703
.638
.066
.674
.711
.634
492
.633
.097
.677
.617
.337
.460
.378
.093
.318
.432
.774
.733
.619
.138
.614
.617
.630
.378
.610
.022
,4v:
.447
.423
.382
.330
.270
.686
.588
.43?
.133
.479
.496
.634
.630
.360
.084
.436
.421
.336
.302
.474
.034
.387
.390
.438
.409
.436
.0*2
.436
.453
.413
.390
.429
.032
.300
.288
.490
.474
.388
.109
.3*0
.294
.330
.313
.473
.122
.772
.781
.575
.632
.466
.472
.945
.964
.701
.193
.842
.838
.937
.914
.888
.045
.631
.632
.736
.742
.695
.063
.648
.696
.482
.447
.373
.116
.333
.369
.634
.606
.390
.032
.431
.482
.443
.4*8
.348
.121
.343
.314
.7*0
.732
.430
.141
.391
.669
.651
.721
.410
.472
1.071
1.144
.716
.263
.889
.978
1.012
1.068
.987
.075
.674
.70S
.819
.931
.787
.126
.398
.638
.631
.666
.638
.031
.398
.699
.330
.418
.414
.062
.444
.490
.7C-»
.72*
.991
.143
.373
.403
.796
.871
.712
.144
.782
.703
.728
.707
.392
.334
1.393
1.212
.834
.302
1.224
1.124
1.180
1.109
1.159
.033
.729
.816
1.073
.990
.902
.158
.446
.418
.336
.347
.387
.034
.232
.298
.388
.243
.296
.064
.144
.062
.274
.128
.137
.089
.373
.333
.237
.426
.403
.124
.749
.613
.764
.641
.372
.381
1.338
1.231
.814
.306
1.017
1.048
.883
.907
.964
.081
.634
.321
.931
.934
.763
.212
.211
.316
.323
.230
.270
.038
.248
.247
.116
.233
.211
.064
.045
.060
-.005
.070
.042
.033
.137
.103
.106
.062
.102
.031
.734
.733
.813
.769
.673
.661
1.370
1.361
.892
.296
1.219
1.201
1.032
1.024
1.119
.103
.803
.832
1.137
1.122
.973
.181
.616
.604
.461
.470
.538
.084
.411
.413
.366
.421
.403
.023
.098
.077
.091
.062
.082
.014
.134
.123
.101
.071
.113
.033
.638
.681
.783
.842
.641
.633
1.417
1.393
.880
.336
1.212
1.226
1.022
1.043
1.126
.108
.774
.836
1.093
1.123
.937
.178
.334
.374
.613
.383
.381
.023
.436
.430
.308
.536
.482
.047
.064
.083
.079
.112
.084
.020
.109
.119
.084
.104
.104
.013
.906
.874
.832
.861
.687
.677
1.376
1.420
.937
.286
1.197
1.207
1.078
1.073
1.139
.073
.833
.849
1.088
1.182
.988
.174
.666
.674
.780
.4*7
.703
.032
.471
.322
.348
.332
.323
.037
.034
.079
.134
.113
.093
.034
.113
.11*
.111
.108
.113
.003
                                                                       111-51

-------
TAbLF 4I>  MSfUNSE  Of f—h RATIO Of POTAMOOETON  TO  LINIIKON CONCENTRATIONS
          iF.i-.b.t.  iveo EXPEMHENT.
          UNITS AR£  GKAflS OXYGEN PE* CU0IC  HFTER  PL*  MUUK.
  TREATMENT
                TANK.
                                                      UEl.v
                                                                                  10
                                                                                          II
.CONTROL
.CONTROL
.CONTROL
.CONTROL
.CONTROL
.CONTROL
.CONTROL
.CONTROL
. MEAN
. ERROR










,'VEH-CONTROL
.UEH-CONTROL
.WEH-CONTROL
.VEH-CONTROL
. MEAN
. ERROR
.LINURON
.LINURQM
.LINURON
.LINURON
. MEAN
. ERROR
.LINURON
.LINURON
.LINURON
.LINURON
•
. BEAN
. ERROR
.LINURON
.LINURON
.LINURON
.LINURON
•
. flEAN
. ERROR
.LINURON
.LINURON
.LINURON
.LINURON
. MEAN
. ERROR
! LINURON
.LINURON
.LINURON
.LINURON
. MEAN
. ERROR


3
S
s
3


30
30
30
30


100
100
100
100


300
300
300
300


1000
1000
1000
1000


19 .
19 .
6 .
6 .
IS .
13 .
26 .
26 .
t
•
22 .
22 .
24 .
24 .
f
•
17 .
17 .
16 .
IB .
*
•
1 .
1 .
4 .
4 .
•
•
8 .
8 .
21 .
21 .
/
•
2 .
2 .
7 .
7 .
*
•
13 .
13 .
20 .
20 .
^
•
.385
.363
.831
.722
.537
.477
.434
.413
.378
.148
.904
.887
.743
.734
.818
.091
.784
.838
.478
.723
.736
.070
.760
.717
.774
.694
.736
.037
1.331
1.2S9
.773
.396
.990
.361
.744
.670
.931
.946
.823
.137
.336
.343
.374
.483
.334
.037
.082
.049
.371
.463
.140
.236
.031
-.083
.164
.182
.153
.147
.338
.377
.239
.126
.010
.033
.271
.233
.143
.129
.176
.437
.122
.164
.230
.133
.420
.347
.230
.362
.393
.124
.197
.273
.162
.296
.232
.063
.103
.019
.227
.280
.138
.118
.•MO
.687
.693
.844
.687
.690
.441
.446
.632
.134
.703
.693
.798
.711
.727
.048
.373
.313
.656
.751
.624
.103
.376
.666
.302
.496
.363
.088
.634
.806
.849
.761
.762
.093
.710
.651
.396
.746
.676
.067
.2*7
.278
.823
.721
.313
.C34
.435
.527
.932
.942
.600
.753
.760
.791
.717
.183
.804
.732
.938
.901
.849
.100
.642
.338
.700
.745
.661
.081
.618
.SS7
.814
.799
.697
.129
.904
.886
.633
.797
.810
.113
.744
.730
.817
.809
.773
.044
.333
.623
.944
.80S
.727
.183
.398
.18?
.702
.475
.839
.584
.621
.743
.570
.213
.432
.096
.386
.223
.340
.220
.233
.066
.346
.219
.216
.113
-.121
-.318
-.311
-.481
-.308
.147
-.398
-.646
-.493
-.787
-.631
.121
-.742
-1.113
-.663
-1.047
-.892
.223
-.686
-.333
-.831
-.437
-.377
.223
.345
.670
.909
1.194
.919
1.422
1,144
1.320
1.013
.309
.446
.768
.344
.995
.688
.245
-.234
-.013
.704
.878
.329
.347
-.348
.035
-.344
-.171
-.302
.296
-.734
-.431
-1.032
-.418
-.634
.291
-1.160
-1.031
-1.423
-1.099
-1 . 1 78
.171
-1.033
-1.093
-.989
-1.107
-1.036
.033
.399
.670
1.048
1.183
1.093
1.213
1.065
1.124
1.000
.233
.840
.910
1.040
1.038
.962
.103
.9)9
1.003
.876
.913
.928
.034
.633
.683
.492
.636
.616
.083
.439
.342
.036
.261
.324
.213
-.813
-.918
-.881
-.972
-.896
.066
-.908
-.994
-.918
-1.036
-.964
.061
.672
.693
1.299
1.291
1.027
.883
1.107
1.080
1.007
.241
1.093
1.036
1.103
1.042
1.074
.030
1.340
1.322
1.013
.942
1.204
.278
.620
.579
.888
.674
.690
.137
.408
.435
.303
.298
.411
.143
-.820
-.782
-.936
-.832
-.832
.073
-.996
-.975
-.719
-.832
-.933
.064
1.146
1.088
1.113
1.117
.939
1.023
1.011
1.043
1.061
.069
1.064
1.121
1.198
1.333
1.179
.114
.867
.916
.877
.988
.912
.033
.306
.364
.923
.949
.740
.239
.733
.914
.433
.337
.660
.207
-.874
-.604
-.734
-.768
-.730
.111
-.944
-.938
-.838
-.838
-.894
.034
                                             III-5Z

-------
l»bLC 3A  kESPONbE OF AtfAK'lNl fHOTOSTNl ME B16 Or HrMOPHYUUM TO ATXAZ1NC
          CONCENTkATlONblP.P.H.I.   1979 EUERIHINT
          UNI1S AKC GhAMS OXYGIN fEK Cu&lC  flETEK PEK HUUK.
                                                      WEEIv
  TKEATMENT
                TANK,
                                                                                 10
.VEH-CONTROL
.WEH-CONTROL
.VEH-CONTR01.
.VEH-CONTROL
. MEAN
. EftftOfc
.ATKAZINE
.ATRA21NE
.ATKAZINE
.AfRAZINE
. MEAN
. ERROR
.ATRAZINE
.ATRAZINE
.AlfcAZINE
.ATRAZINE
. MEAN
. ERROR
.ATRAZINE
.ATRAZINE
.ATRAZINE
.ATRAZINE
. MEAN
. ERROR
.ATRAZINE
.ATRAZINE
.ATRAZINE
.ATRAZINE
. MEAN
. ERROR
.ATRAZINE
.ATRAZINE
.ATRAZINE
.ATRAZINE
.' MEAN
. ERROR


3
S
3
S


SO
50
30
30


100
100
100
100


300
300
300
300


10OO
1000
1000
1000


37 .
37 .
39 .
39 .
.
•
35 .
33 .
40 .
40 .
.
•
31 .
31 .
26 .
26 .
,
•
33 .
33 .
34 .
34 .
f
•
41 .
41 .
42 .
42 .
f
•
32 .
32 .
38 .
38 .
.
•
.134
.222
.216
.284
.214
.06?
.IBB
.297
.436
.393
.383
.176
.200
.288
.000
.000
.366
.131
.178
.337
.239
.270
.261
.075
.093
.310
.189
.222
.204
.089
.408
.432
.320
.523
.421
.083
.135
.095
.227
.239
.174
.070
.380
.325
.187
.230
.280
.088
.231
.091
.000
.000
.349
.227
.340
.313
.294
.213
.290
.054
.103
.149
.138
.084
.118
.030
.432
.316
.421
.369
.384
.053
.454
.454
.445
.437
.447
.008
.473
.333
.346
.317
.472
.083
.314
.541
.OOO
.000
.482
.054
.532
.518
.400
.371
.455
.082
.226
.210
.250
.229
.229
.016
.437
.463
.377
.399
.424 '
.043
.423
.264
.434
.272
.348
.093
.323
.126
.499
.292
.310
.153
.603
.248
.000
.000
.424
.149
.461
.297
.393
.7^6
.339
.083
.181
.102
.249
.191
.181
.060 .
.337
.423
.417
.291
.422
.109
.446
.462
.437
.327
.468
.041
.351
.319
.546
.591
.302
.105
.568
.729
.000
.000
.538
.146
.398
.480
.383
.472
.433
.050
.177
.222
.233
.294
.231
.048
.629
.758
.321
.422
.532
.198
.440
.489
.534
.591
.513
.064
.454
.503
.722
.788
.617
.163
.236
.326
.000
.000
.211
.098
-.033
.055
-.133
.013
-.024
.081
-.253
-.136
-.2BO
-.103
-.183
.067
-.361
-.342
-.317
-.373
-.448
.107
.229
.207
.316
.460
.353
.158
.264
.230
.636
.436
.392
.187
.087
.148
.000
.000
-.134
.313
-.037
-.003
-.483
-.488
-.253
.269
-.331
-.235
-.500
-.488
-.388
.128
-.707
-.491
-.301
-.490
-.347
.107
.473
.440
.355
.347
.304
.036
.563
.342
.791
.798
.674
.139
.392
.344
.000
.000
.393
.060
.263
.314
.332
.482
.396
.139
-.112
-.103
-.170
-.104
-.122
.032'
-.321
-,31»
-.271
-.160
-.268
.073
.450
.671
.539
.691
.588
.114
.348
.765
.640
.735
.672
.098
.391
.512
.000
.000
.314
.097
.298
.422
.268
.303
.323
.068
-.103
-.133
-.040
-.035
-.083
.043
-.376
-.379
-.272
-.233
-.320
.066
                                             111-53

-------
1AH1C  S..'  KlSfUNKl  OF  RESf'UATION OF  nTRIOPHrLLUn TO ATfcA/INC CONCENlkATIONE,
          (f-.P.fc.).   J?7» EXt-EMm.N1
          UNlTb  Aht  OKAMS OXTCCN  PL ft  fuBlC  HETER PER HOUfc.
  TRtATntNl
                TANK.
                                                      WEEK
                                                                  a
.VEH-CCJNTROL
.VEH-CONTROL
.VEH-CONTROL
.VEH-CONTROL
. MEAN
. ERROR
37
37
39
39


. -.185
. -.301
. -.269
. -.351
. -.276
. .070
-.166
-.125
-.310
-.316
-.229
.096
-.443
-.494
-.443
-.436
-.434
.027
-.436
-.360
-.446
-.343
-.396
.052
-.443
-.464
-.436
-.493
-.464
.021
-.660
-.660
-.695
-.695
-.677
.020
-.691
-.684
-.548
-.601
-.631
.069
-.576
-.542
-.588
-.583
-.373
.021
-.393
-.621
-.389
-.611
-.604
..015
•
t
.
•
§
•
                                                               .623   -.363
                                                               .628   -.495
                                                               .873  -1.118
                                                               .875   -.833
                         -.620
                         -.622
                         -.798
                         -.774
                        -.634
                        -.693
                        -.682
                        -.632
 .ATRAZINE    3   33  . -.247   -.432    -.439    -.320   -.409
 .ATRAZINE    3   33  . -.334   -.419    -.376    -.107   -.492
 .ATRAZINE    3   40  . -.406   -.334    -.534    -.521   -.328
 .ATRAZINE    S   40  . -.526   -.343    -.379    -.447   -.574

 . MEAN               . -.383   -.382    -.487    -.349   -.501   -.751    -.733   -.703   -.666
 . ERROR              .  .116    .031     .096     .181    .070    .143     .284    .096    .028

 .ATRAZINE   SO   31  . -.231   -.210    -.467    -.599   -.563   -.704    -.686   -.328   -.397
 .ATRAZINE   SO   31  . -.304   -.122    -.515    -.236   -.624   -.704    -.437   -.524   -.583
 .ATRAZINE   SO   26  .  .000    .000     .000     .000    .000    .000     .000    .000    .000
 .ATRAZINE   SO   26  .  .000    .000     .000     .000    .000    .000     .000    .000    .000

 . MEAN               . -.447   -.395    -.567    -.318   -.599   -.736    -.533   -.347   -.610
 . ERROR              .  .222    .267     .091     .189    .030    .037     .102    .035    .024

 .ATRAZINE  100   33  . -.339   -.443    -.501    -.488   -.469   -.493    -.339   -.434   -.313
 .ATRAZINE  100   33  . -.461   -.434    -.333    -.401    -.498   -.493    -.776   -.498   -.313
 .ATRAZINE  100   34  . -.437   -.487    -.507    -.516   -.471   -.663    -.491   -.624   -.561
 .ATRAZINE  100   34  . -.490   -.462    -.311    -.431    -.300   -.663    -.514   -.615   -.537

 . MLAN               . -.432   -.4Lfc    -.513    -.464    -.484   -.578    -.380   -.543   -.332
 . EKROR              .  .066    .023     .014     .050    .017    .098     .132    .092    .022

 ...TRAZINE  300   41  . -.254   -..'82    -.297    -.233    -.207   -.360    -.303   -.134   -.263
 .ATRAZINE  300   41  . -.394   -.284   -.301    -.178    -.319   -.360    -.254   -.139   -.260
 .ATRAZINE  500   42  . -.:.!*   -.347    -.333    -.339   -.313   -.448    -.478   -.244   -.327
 ...IRAZINE  300   42  . -.407   -.333    -.331    -.279    -.332   -.448    -.SIS   -.217   -.302

 . 1CAN              . -.343   -.312
 . IRfcOk             .  .071    .034

 .AIRAZINE 1000   32 . -.374   -.33'
 .ATRAZINE 1000   32 . -.483   -.495
 .AlhAZINE 1000   38 .-.'?*   -.'76
.AIRAZINE 1000   36 . -.17?   -,48V

. MEAN              . -.466   -.'»(•   -,4C    -.518   -.484   -.i'3:     .492    -.363   -.398
. ERROR             .  .084    .025    .('6     .081     .089     .f:'     .I'?*     .157    .069
                                      .320   -.262
                                      .026    .067
-.293
 .038
.404   -.388
.031    .12*
                                      .313   -.379   -.322
                                      .SI."   -.327   -.391
                                      .41.-   -.563   -.393
                                      .'30   -.402   -.431
         .546
         .546
         .327
         .M7
                                                                     -.47C
                .1?*
                .747
-.288
 .032

-.497
-.387
-.372
-.337
                                          111-54

-------
TAhlE 3C  KtStONbl Or GKOSS  f MOTOS>rNTMLSlS OF HTk'lOF HTLLUn  TO  ATKAZINC
          CONCENTRATIONS^^ .11. >.   1V79 EXI'tRlHFN'
          UNITS ARE GKAnS  OxrCiEN  fFF  CoBlC  METER fEK HOUN.
. TREATMENT
.VEH-CONTROL
.VEH-CONTROL
.VEH-CONTROL
.VEH-CONTROL
. HE AN
. ERROR
.ATRAZINE S
.ATRAZINE 5
.ATRAZINE S
.ATRAZINE S
. MEAN
. ERROR
.ATRAZINE SO
.ATRAZINE SO
.ATRAZINE SO
.A1RAZINE SO
. MEAN
. ERROR
.ATRAZINE 100
.ATRAZINE 100
.ATRAZINE 100
.ATRAZINE 100
. MEAN
. ERROR
.ATRAZIPC SOO
.ATRAZINE 500
.ATRAZINE SOO
.ATRAZINE SOO
. MEAN
. ERROR
.ATRAZINE 1000
.ATRAZINE 1000
.ATRAZINE 1000
1 .ATRAZINE 1000
. MEAN
. ERROR
TANK.
37 .
37 .
39 .
39 .
.
35 .
35 .
40 .
40 .
.
31 .
31 .
26 .
26 .
.
33 .
33 .
34 .
34 .
.
41 .
41 .
42 .
42 .
.
32 .
32 .
38 .
38 .
•
3
.393
.644
.593
.776
.601
.159
.534
.792
1.024
1.329
.338
.524
.713
.000
.000
.992
.460
.653
1.003
.850
.956
.865
.155
.450
.861
.636
.792
.683
.183
.932
1.108
.927
1.323
1.072
.187
<
.367
.271
.661
.682
.493
.207
.985
.912
.654
.710
.815
.158
.325
.262
.000
.000
.901
. .597
.960
.920
.976
.861
.929
.031
.497
.347
.623
.353
.353
.052
1.179
1.099
1.087
1.033
1.082
.072
S
1.0. 7
!.!• i
1.064
1.047
1.083
.043
1 .086
.880
1.321
1.328
1.154
.214
1.168
1.261
.000
.000
1.276
.082
1.233
1.265
1.110
1.086
1.173
.089
.642
.632
.716
.720
.677
.047
1.178
1.180
.960
1.070
1.097
.105
4
1.033
.768
1.058
.751
.902
.166
.772
.275
1.228
.917
.79B
.397
1.442
.579
.000
.000
1.150
.392
1.144
.859
1.115
.917
1.009
.142
.533
.351
.724
.581
.548
.134
1.367
1.162
1.209
.834
1.148
.215
WEEK
T
1.066
1.111
1.075
1.217
1.117
.069
.924
1.208
1.285
1.395
1.203
.201
1.360
1.603
.000
.000
1.378
.164
1.055
1.177
1.043
1.172
1.112
.073
.467
.669
.674
.738
.642
.124
1.361
1.383
.871
1.023
1.210
.323
&
1.164
1.413
1.507
1.564
1.462
.090
1.333
1.383
1.947
2.013
1.669
.361
1.222
1.312
.000
.000
1.241
.063
.658
.746
.795
.942
.785
.119
.251
.348
.407
.324
.382
.114
.203
.422
.207
.351
.296
.109
9
1.197
1.166
1.284
1.301
1.237
.066
1.055
.922
2.203
1.602
1.446
.584
1.046
.760
.000
.000
.640
.344
.717
1.083
.205
.231
.559
.421
.096
.121
.170
.233
.139
.060
-.050
.229
.166
.228
.143
.132
>0
1.279
1.199
1.378
1.366
1.303
.084
1.432
1.413
1.909
1.882
1.639
.273
1.131
1.078
.000
.000
1.160
.069
.873
1.011
1.406
1.343
1.158
.257
.104
.120
.171
.201
.149
.043
.277
.462
.030
.186
.244
.173
II
1.281
1.541
1.364
1.546
1.433
.132
1.464
1.737
1.394
1.620
1.604
.112
1.226
1.329
.000
.000
1.368
.118
1.018
1.143
1.053
1.054
1.067
.053
.263
.232
.417
.367
.320
.087
.319
.162
.249
.217
.237
.066
.
•
.
*
'
•
.
•
•
•
I
•

                                   111-55

-------
TAhLE !H  KfSf'ONSr Of f-K RATIO  OF  MTh lOnirLLUM TO ATkAZINf  CONCENTRATIONS
          O'.KIi. ). 197V EXIERIMENI.
          UNITS AkE GKAMS OXYliLN  fEK  CU
-------
 TAbLE  6*   hlSf-ONSE  Of  APPARENT PHOTOSYNTHESIS Of MYRIOHMYLLUr TO LJNUKIIK
           CQNCENTRATJONSCF-.F . H. ) .   1980 FXF'ERIHENT.
           UNITS  ARE GRAMS OXTDLN  F'tK Cuft'C   METER FEk HUUK.

                                                       UEtK
 .  TREATMENT      TANK.     3        45       b       f       8       9       lO'l


 .CONTKOL          29 .   .398    .148    .1OI    .149     .274     .370    .360     .*S7     .31)
 .CONTROL          29 .   .343    .103    .099    .196     .211     .311    .401     .530     .33:
 .CONTROL          30 .   .262    .069    .Ii6    .106     .166     .274    .334     .412     .431
 .CONTROL          30 .   .317    .003    .121    .199     .063     .246    .333     .419     .373

 •  HE AN              ..   .330    .081    .109    .162     .179     .300    .367     .479     .467
 •  ERROR              .   .037    .061    .011    .044     .090     .034    .022     .075     .080
.WEN-CONTROL
.WEN-CONTROL
. WEH-CONTROL
.WEN-CONTROL
. MEAN
. ERROR
.LINURON 5
.LINURON 3
.LINURON 5
.LINURON 5
. MEAN
. ERROR
.LINURON SO
.LINURON 50
.LINURON SO
.LINURON SO
. MEAN
. ERROfc
32 .
3* .
36 .
36 .
.
•
31 .
31 .
41 .
41 .

•
37 .
37 .
40 .
40 .
.
.
.322
.242
.393
.343
.32S
.063
.546
.467
.603
.547
.541
.056
.788
.724
.381
.340
.558
.230
-.020
.011
.080
.072
.036
.048
.049
.038
-.014
-.009
.016
.032
.146
.159
.120
.111
.134
.022
.053
.022
.193
.238
.126
.103
.235
.201
.136
.096
.167
.063
.336
.257
.143
.149
.221
.093
.132
.238
.355
.318
.266
.090
.323
.319
.226
.236
.276
.052
.273
.275
.322
.323
.299
.027
.296
.243
.347
.246
.283
.049
.265
.187
.139
-.020
.143
.120
-.113
-.253
-.008
-.072
-.111
.104
-.066
.133
.147
.270
.121
.139
.391
.430
-.123
.039
.184
.270
.143
.200
.041
.049
.108
.077
.339
.326
.409
.393
.367
.040
.537
.579
.334
.350
.450
.126
.446
.448
.214
.231
.335
.130
.357
.352
.394
.405
.377
.026
.601
.383
.467
.436
.322
.082
.474
.481
.216
.202
.343
.153
.402
.363
.389
.391
.386
.017
.371
.652
.476
.514
.553
.077
.473
.535
.277
.280
.391
.133
.LINURON   100   42
.LINURON   100   42
.LINURON   100   38
.LINURON   100   38

. MEAN
. ERROR

.LINURON   300   39
.LINURON   500   39
.LINURON   500   33
.LINURON   300   33
•
. MEAN
. ERROR
•
.LINURON  1000   34
.LINURON  1000   34
.LINURON  1000   35
.LINURON  1000   35

. MEAN
. ERROP
.273
.297
.275
.270
.279
.012
.598
.464
.359
.295
.429
.132
.098
.060
.156
.133
.112
.042
.071
.084
.103
.101
.090
.013
.125
.113
.381
.407
.256
.139
.334
.275
.172
.172
.238
.080
.104
.148
.456
.455
.291
.191
.367
.419
.324
.367
.369
.039
-.173
-.367
-.163
-.223
-.232
.094
-.435
-.263
-.287
-.352
-.339
.086
-.194
-.168
-.160
-.040
-.140
.069
-.329
-.325
-.413
-.274
-.336
.058
-.085
-.062
.054
.082
-.003
.083
-.250
-.245
-.171
-.176
-.210
.043
.014
.020
.148
.150
.083
.076
-.249
-.264
-.148
-.139
-.200
.066
.028
.010
.198
.193
.107
.102
-.228
-.213
-.125
-.214
-.195
.047
.440
.340
.429
.285
.373
.074
.003
.023
.136
.152
.078
.076
.164
.159
.352
.334
.252
.103
.185
.277
.475
.492
.357
.151
-.347
-.205
-.333
-.333
-.354
.133
-.273
-.273
-.297
-.282
-.282
.011
-.244
-.258
-.266
-.253
-.253
.009
-.226
-.237
-.239
-.232
-.243
.015
-.239
-.229
-.266
-.249
-.246
.016
                                            111-57

-------
T«m.f 6*  RESPONSE OF RESFIfcATION (ir HrMUKHniUM TO LINUhON CONCENTKAT IONS
          (P.f.Fi.).  )9BO  EXPERIMENT.
          UNITS ARC GhftttS  OJCT'JtN F'EK CUBIC   METER f-Ef. HOUF
                                                                                                                 4
,
t
. TREATMENT
.CONTROL
.CONTROL
.CONtfiiL
.CONTROL
. MEAN
. ERROfc






.WEN-CONTROL
.VEH-CONTROL
.VEH-CONTROL
.VEH-CONTROL
. MEAN
. ERROR
.LI HURON
.LINURON
.LINURON
.LINURON
. MEAN
. ERROR
.LINURON
.LINURON
.LINURON
.LINURON
. MEAN
. ERROR
.LINURON
.LINURON
.LINURON
.LINURON
. MEAN
. ERROR
.LINURON
.LINURON
.LINURON
.LINURON
. HEM*
. ERROR
.LINURON
.LINURON
.LINURON
.LINURON
! MEAN
. ERROR
•


5
5
3
5


50
SO
50
50


100
100
100
100


300
SOO
500
SOO


1000
.000
1000
1000


f
TANK.
29 .
29 .
30 .
30 .
,
.
32 .
32 .
36 .
3* .
t
•
31 .
31 .
41 .
41 .
,
.
37 .
3? .
40 .
40 .
(
•
42 .
42 .
36 .
38 .
(
•
3V .
3* .
33 .
33 .
•
•
34 .
34 .
33 .
39 .
t
•

1
-.IBB
-.249
-.294
-.274
-.276
.020
-.284
-.238
-.322
-.30o
-.293
.028
-.431
-.415
-.399
-.380
-.411
.030
-.330
-.625
-.288
-.283
-.436
.177
-.23>
-.227
-.239
-.233
-.233
.005
-.411
-.33*
-.284
-.240
-.323
.073
-.336
-.307
-.330
-.373
-.342
.028

4
-.331
-.307
-.326
-.311
-.319
.012
-.253
-.234
-.314
-.317
-.284
.036
-.321
-.291
-.326
-.301
-.310
.017
-.483
-.489
-.290
-.273
-.385
.118
-.2*0
-.281
-.273
-.248
-.273
.018
-.377
-.349
-.237
-.263
-.312
.060
-.347
-.363
-.423
-.343
-.369
.037

5
-.294
-.270
-.316
-.312
-.298
.021
-.279
-.239
-.304
-.293
-.284
.020
-.363
-.340
-.365
-.345
-.333
.013
-.363
-.338
-.299
-.263
-.416
.137
-.286
-.274
-.300
-.303
-.291
.013
-.406
-.390
-.280
-.266
-.333
.073
-.331
-.314
-.436
-.441
-.383
.073

•
-.278
-.289
-.310
-.327
-.301
.022
-.283
-.332
-.373
-.390
-.343
.047
-.421
-.441
-.378
-.400
-.410
.027
-.327
-.339
-.323
-.367
-.444
.116
-.280
-.326
-.347
-.384
-.334
.043
-.447
-.437
-.298
-.308
-.377
.086
-.346
-.378
-.476
-.514
-.428
.079
UEEf
7
-.324
-.363
-.428
-.328
-.368
.043
-.330
-.407
-.449
-.443
-.407
.035
-.366
-.609
-.322
-.620
-.379
.043
-.441
-•«0i
- S«2
• .303
-.397
. .139
-.323
-.497
-.341
-.341
-.376
.081
-.322
-.32i
-.339
-.374
-.439
.097
-.399
-.399
-.493
-.493
-.446
.034

I
-.329
-.233
-.310
-.242
-.27?
.047
- .389
-.280
-.476
-.302
-.362
.090
-.463
-.420
-.501
-.413
-.453
.045
-.474
-.400
-.205
-.149
-.307
.133
-.289
-.231
-.238
-.193
-.243
.041
-.416
-.363
-.302
-.262
-.336
.068
-.293
-.237
-.322
-.277
-.282
.033

9
-.306
-.286
-.276
-.243
-.278
.023
-.342
-.316
-.341
-.319
-.329
.014
-.306
-.484
-.438
-.426
-.463
.038
-.580
-.362
-.293
-.277
-.428
.163
-.262
-.239
-.226
-.223
-.242
.021
-.332
-.331
-.219
-.196
-.279
.084
-.273
-.264
-.291
-.282
-.277
.012

10
-.239
-.269
-.239
-.272
-.233
.018
-.287
-.327
-.319
-.359
-.323
.030
-.434
-.471
-.377
-.406
-.427
.043
-.338
-.366
-.301
-.322
-.432
.140
-.228
-.232
-.214
-.253
-.237
.020
-.324
-.368
-.163
-.250
-.277
.089
-.216
-.247
-.242
-.245
-.237
.014

II
-.313
-.3or
-.323
-.291
-.308
.014
-.369
-.327
-.391
-.360
-.362
.027
-.332
-.343
-.434
-.433
-.495
.061
-.368
-.390
-.343
-.310
-.433
.146
-.293
-.278
-.300
-.272
-.286
.013
-.368
-.334
-.161
-.083
-.242
.141
-.238
-.238
-.230
-.247
-.243
.006
                                           111-53

-------
 TAMl  6C   HSIIiNM  Of  OkObS f MOIOSTNIHLSIS OF NlHOHirtlUn 10 LINUMIN
           ClINi [NTKATIONMI .( .h. » ,   IVHO IXIFMhNlI.
           UNITS Af>E CKAKS OX'ULN fl> CuB't  HETEK FEK HUUk.
.
. TREATMENT
.CONTROL
.CONTROL
.CONTROL
.CONTROL
. MEAN
. ERROR
.VCH-CONTROL
.VCH-CONTROL
.VEM-CONTROL
.VCH-CONTROL
. MEAN
. ERROR
.LINURON 3
.LINURON S
.LINURON 3
.LINURON 3
. HE AN
. ERROR
.
TANK.
2* .
2V .
30 .
30 .
m
•
32 .
32 .
3* .
3* .
t
•
31 .
31 .
41 .
41 .
B
.

)
.801
.4*2
.474
.700
.717
.037
.720
.603
.844
.771
.734
.101
1.178
1.048
1.1*2
1.07?
1.117
.043

4
.613
.534
.324
.43?
.328
.071
.334
.366
.31?
.317
.434
.ova
.4??
.443
.442
.413
.430
.03*

i
.312
.477
.318
.338
.326
.03?
.443
.383
.41?
.430
.324
.130
./44
.477
.444
.37?
.441
.04?

fc
.33?
.400
.340
.437
.384
.036
.550
.702
.878
.864
.748
.133
,?12
.V37
.734
.7?3
.830
.088
HEEr
T
.730
.720
.762
.a6S
.695
.08?
.738
.813
,?7S
.844
.833
.0*3
1.038
1.03?
.84?
.848
.?33
.110

•
.811
.64)
.707
. jB4
.4?!
.104
.47?
.323
.813
.4?3
.427
.134
1.070
1.018
.378
.417
.821
.23*

a
.78?
.800
.740
.4?8
.737
.047
.81*
.76*
.884
.83*
.828
.048
1.244
1.234
.*47
. ?44
1.0**
.174

10
.8*2
.?06
.747
.800
.836
.074
.738
.810
.841
.*07
.82*
.042
1.23*
1.242
.?»4
1.004
1.11*
.13*

II
.»32
.»73
.884
.77*
.8*7
.088
.»!»
,821
,»37
.8*4
.8*3
.031
1.344
1.412
1.111
1.120
1.247
.134
 .LINURO:.    30   37  .  1.33*
 .LINURON    30   37  .  1.3**
 .LINURON    30   40  .   .784
 .LINURON    SO   40  .   .737
           .824   1.128   1.013    .304
           .843   1.011   1.038    .387
           .324    .342    .774    .331
           .4*7    .317    .834    .334
                                 .804   1.238   1.227   1.248
                                 .740   1.234   1.273   1.341
                                 .32*    .427    .438    .73*
                                 .238    .418    .433    .714
  MEAN
  ERROR
.  1.170
.   .473
.LINURON   100   42 .   .3*4
.LINURON   100   42 .   .413
.LINURON   100   38 .   .410
.LINURON   100   38 .   .400
.473
.187

.303
.433
.338
.480
.804
.310
.*21
.134
.444
.122
                   .324    .4*3    .280
                   .4*7    .403    .32*
                   .802    .?41    .314
                   .831    .»*2 •  .233
.338
.283

.210
.134
.202
.231
.*33
.341

.282
.301
.370
.3*4
 .*48
 .330

 .333
 .373
 .447
 .307
1.023
 .334

 .43*
 .3**
 .418
 .3/3
  MEAN
  ERROR
   .403
   .00*
.LINURON   300   3* . 1.174
.LINURON   300   3* .  .»42
.LINURON   300   13 .  .73*
.LINURON   300   33 .  .431
.4*3
.034

.3**
.372
.444
.472
                                       .177
        .738
        .244
                   .»02    .»*4
                   .821   1.C40
                   .343    .741
                   .344    .7*8
        .2*4
        .033
                                                       .274
                .IN
                .172
        .200
        .032

        .234
        .183
        .008
        .0*2
        .337
        .084

        .242
        .247
        .133
        .0*8
. MEAN
. ERROR
   .••1
   .238
.LINURON  1000   34 .  ,?13
.LINURON  1000   34 .  .770
.LINURON  1000   33 .  .*!*
.LINURON  1000   33 .  .811
.327
.04*

.48*
.334
.728
.431
.708
.181
                                               .133
                   .427     .44*
                   .400     .807
                   .**0    1.142
                   .*32    1.211
        .274
        .134

        .211
        .333
        .137
        .337
        .134
        .107

        .133
        .03*
        .134
        .104
        .180
        .073

        .138
        .111
        .140
        .141
        .419
        .077

        .203
       ..232
        .084
        .211

        .188
        .072

        .077
        .lOf
        .080
        .0*1
. MEAN
. ERROR
   .833
   .074
.3*3
.104
.7*2
.207
.»37
.241
.24*
.101
.113
.041
.132
.014
.08*-
.014
         .308
         .103

         .288
         .282
         .101
       -.0*4

         .144
         .181

         .0*3
         .104
         .084
                                                                                       .008
                                         111-59

-------
)»HLl 6l>  KtBHINSf Of P-K kftln OF  flYKIOPHYLLUn TO LJNUKON CONCENTRATIONS,
          (P.F.K.). 1VBO EXf EMHCN1 .
          UNITE AHE CKAnS OXTOLN P£*  CofcU   HETEfc fEh MOUK.
.
.
. TkCATnENT
.CONTKOL
.CONTROL
.CONTROL
.CONTROL
. HE AN
. ERROR






.VEH-CONTROL
.VEH-CONTROL
.VEH-CONTROL
.VCH-CONTROL
. MEAN
. ERROR
.LINURON
.LINURON
.LINURON
.LINURON
. MEAN
. ERROR
.LINURON
.LINURON
.LINURON
.LINURON
. MEAN
. ERROR
.LINURON
.LINURON
.LINURON
.LINURON
. MEAN
. ERROK
.LINURON
.LINURON
.LINURON
.LINURON
. MEAN
. ERROR
.LINURON
.LINURON
.LINURON
.LINURON
. MEAN
. ERROR


3
3
3
3


30
30
30
30


100
too
too
100


300
300
300
300


1000
1000
1000
1000


,
TANK.
2V .
29 .
30 .
30 .
(
•
32 .
32 .
36 .
36 .
t
•
31 .
31 .
41 .
41 .
^
•
37 .
37 .
40 .
40 .
,
•
42 .
42 .
38 .
38 .
t
•
39 .
39 .
33 .
33 .
t
•
34 .
34 .
33 .
33 .
,
*

*
1.382
1.378
.891
1.157
1.202
.232
1.134
.938
1.220
1.121
1.103
.11V
1.211
1.123
1.311
1.43V
1.321
.183
1.433
1.138
1.323
1.201
1.27V
.124
1.182
1.308
1.131
1.14V
1.197
.073
1.433
1.303
1.233
1.22V
1.310
.101
1.302
1.107
1.226
.760
LOW
.240

4
.446
.334
.212
.010
.231
.187
-.079
.043
.235
.227
.111
.138
.133
.131
-.043
-.030
.033
.104
.301
.323
.414
.404
.361
.036
.338
.214
.371
.336
.413
.16V
.188
.241
.401
.381
.303
.104
.009
.063
.322
.443
.20V
.207

ft
.344
.367
.347
.388
.366
.018
.190
.083
.635
.807
.42V
.347
.647
.5V1
.373
.278
.472
.173
.393
.478
.478
.367
.330
.061
.437
.412
1.270
1.343
.863
.310
.823
.703
.614
.647
.497
.OV2
.493
.306
.772
.737
.632
.133

b
.334
.678
.34?
.609
.341
.143
.333
.717
.V32
.815
.734
.174
.767
.723
.598
.590
.66V
.089
.322
.492
.VV1
.880
.721
.232
.371
.434
1.314
1.183
.831
.487
..821
.V17
1.087
1.1 V2
1.004
.167
.333
.733
.998
.V37
.806
.213
UUK
7
.832
.381
.388
.176
.499
.288
.897
.397
.773
.333
.70S
.13V
.468
.307
.266
-.032
.232
.20V
-.236
-.421
-.033
-.236
-.236
.13V
-.338
-.738
-.478
-.634
-.602
.116
-.872
-.304
-.847
-.V41
-.7V1
.1V3
-.870
-.314
-1.081
-.673
-.783
.243

6
1.123
1.323
.884
1.017
1.087
.184
-.170
.473
.30V
.894
.377
.440
.806
1.024
-.246
.094
.420
.3V3
.302
.300
.200
.32V
.333
.123
-.471
-.727
-.620
-.207
-.336
.237
-.791
-.8V3
-1.374
-1.044
-1.027
.234
-.V3V
-1.132
-.V22
-1.018
-1.008
.103

«
1.176
1.402
1.283
1.449
1.327
.123
.Wl
1.032
1.1 VV
1.232
1.113
.120
1.061
1.194
.763
.822
.960
.203
.76V
.7V7
.723
.834
.781
.046
-.324
-.23V
.23V
.348
.011
.344
-.710
-.4V8
-.781
-.898
-.772
.092
-.8V4
-.V77
-.V14
-.8V7
-.V20
.03V

IO
2. 331
1.V70
1.724
1.340
1.BV1
.342
1.244
1.076
1.233
1.128
1.171
.002
1.324
1.238
1.23V
1.074
1.21V
.103
.881
.830
.718
.427
.74V
.118
.041
.07V
.6V2
.388
.333
.331
-,74f
-.717
-.1*97
-.334
-.733
.141
-1.044
-.960
-1.070
-1.02V
-1,026
.047

II
1.422
1.828
1.334
1.282
1.314
. .236
1.08V
1.110
.VVS
1.086
1.070
.031
1.034
1.201
1.048
1.187
1.117
.08V
.833
.V07
.803
.903
.861
.032
.096
.034
.660
.710
.373
.33?
-.620 .
-.402
-.774
-2.318
-1.12V
.929
-1.004
-.V42
-1.064
-1.008
-1.00V
.042
                                          111-60

-------
»
                                                CHAPTER IV
                                 ATRAZINE UPTAKE, PHYTOTOXICITY, RELEASE
                         AND SHORT-TERM RECOVERY FOR THE SUBMERGED AQUATIC PLANT,
                                         POTAMOGETON PERFOLIATUS
                                                T.W. JonesJ'2
                                                 W.M. Kemp2
                                                P.S. Estes3
                                              J.C.  Stevenson2
                 Salisbury State College
                 Department of Biological Science
                 Salisbury, MD   21801
                2
                 Horn Point Environmental Laboratories
                 Center for Environmental and Estuarine Studies
                 Cambridge, MD   21613

                 Oregon State University
                 Department of Zoology
                 CorvaMs, Oregon   97331


                                                         IV-i

-------
                                    ABSTRACT


     Atrazine uptake, phytoxicity, release, and short-term photosynthetic
recovery in the submerged vascular plant, Potamogeton perfoliatus, were investi-
gated in laboratory studies.  The processes of atrazine uptake and release
were rapid, approaching equilibrium within 1 n.  At low concentrations (<0.10
ppm), atrazine is accumulated into plant material  at concentrations greater
than those external to the plant.  At the lowest concentration tested (0.01
ppm), the ratio of atrazine in the shoot to the incubation solution was 9:1.
It is hypothesized that atrazine sorbs to plant tissue in addition to estab-
lishing an equilibrium between internal and external aqueous concentrations.
Thus, at lower external atrazine solution concentrations, the sorbed herbicide
exerts a yreater influence on the ratio of shoot to solution atrazine causing
the ratio to increase.  The uptake of atrazine by roots of £. perfoliatus was
found to be small in comparison to shoot uptake on a whole plant basis.The
150 (the concentration inhibiting photosynthesis by 50%) for atrazine in
solution was determined to be 0.08 ppm with the maximum observed photosynthetic
reduction (87%) at a solution concentration of 0.65 ppm.  Initial  photosynthetic
recovery of P_. perfoliatus following exposure to atrazine was rapid with oxygen
evolution from treated plants (0.05, 0.03, and 0.10 ppm) being statistically
indistinguishable (p<0.05) from control plants after 2 h of atrazine-free
wash.  However, there was an indication of residual photosynthetic depression
in dosed plants, even after a 77 h recovery period.  The ecological consequences
of atrazine runoff into areas containing submerged aquatic vegetation must be
evaluated 1n light of the ephemeral nature of the presence of atrazine in the
water and the kinetics of uptake and release by the plants.
                                     IV-1

-------
 r
•
                                                  INTRODUCTION


                     A small  but measurable percentage of the  herbicides  used  for  agricultural
                weed control  is lost from the croplands where  applied,  and some  eventually
                enters contiguous watercourses (Muir et al.  1978,  Triplett et  al.  1978,  Wu
                1980). If significant concentrations of these  herbicides  result  in adjacent
                aquatic systems, a potential  stress  exists for the indigenous  aquatic  vegetation
                (Stevenson and Confer 1978).   In many aquatic  environments such  as lakes and
                estuaries, a  conspicuous  and  productive community  is  that  of the submerged
                vascular plants. Over the last two decades,  one such  estuarine ecosystem,
                Chesapeake Bay, has experienced a marked decline in the abundance  of this
                submerged aquatic vegetation  (SAV).   This decline  occurred during  a period of
                expanding use of herbicides,  particularly atrazine (2-chloro-4-ethylamino-6-
                isopropylamino-s-triazine)  in the extensively  cropped watershed  (Bayley  et al.
                1978; Stevenson and Confer 1978). This coincidence of widespread herbicide use
                and SAV decline has been  interpreted to suggest a  possible causal  connection
                between the two trends (Stevenson and Confer 1978; Correll  et  al.  1978).

                     Several  recent studies have examined the  effect  of atrazine on SAV  species
                in Chesapeake Bay (Forney and Davis  1981; Correll  and Wu  1982; Cunningham et
                al. 1982). In each of these studies  consistent significant losses  of SAV growth
                and production were observed  at atrazine concentrations of .05-.10 ppm.   Con-
                centrations of .01-.012 ppm resulted in significant effects on SAV in  many
                (but not all) experiments.  In nature, these  submerged plants may be exposed
                ocassionally  to atrazine  concentrations on the order  of .005-.050  ppm  for brief
                periods during runoff events; however, dilution, adsorption, and degradation of
                the herbicide tend to reduce  concentrations  to 0-.005 ppm  within 6-24  h  (Correll
                et al. 1978;  Kemp et al.  1982). Since SAV exposure to phytotoxic herbicide
                concentrations tends to be short term, it becomes  important in assessing
                herbicide effects to know how rapidly the compounds are taken-up by
                plants, how uptake and plant  response are related, and  whether effects persist
                after herbicides are removed  from the water.

                     The purpose of this  study was to examine  the  kinetic  relationships  between
                SAV uptake of atrazine and phytotoxic response. Experiments conducted  here involve
                the plant, Potamogetpn perfoliatus,  a dominant SAV species in  upper Chesapeake
                Bay, and the  herbicide, atrazine, which is the most widely used  weed control
                chemical in the bay's watershed (Stevenson and Confer 1978). Specific  questions
                investigated  in this study include:  1) the rate of atrazine uptake; 2) the
                relative importance of root versus shoot uptake; 3) the relation between uptake
                and reduction in carbon assimilation; 4) the time  required for photosynthetic
                recovery following brief  (2 h) exposure to atrazine.
                                                      IV-2

-------
                             MATERIALS AND METHODS


Collection of Experimental Materials

     Experimental plants (_P. perfoliatus) were collected from shallow (0.5 to
1.5 m) areas of the Choptank River estuary, a tributary of Chesapeake Bay during
July 1981 and 1982. Plants were obtained 1 d prior to experimental  incubations
by gently removing roots and rhizomes from their substrate and washing leaves
and roots free of sediments and epiphytic material. Plants were held overnight
in filtered estuarine water enriched with Provasr'. i's ES medium at  25 C under
constant fluorescent illumination of 150 yEin • nr2 • s*1. All  experiments
were conducted either in a Percival environmental  chamber or a microcosm system
of 20 £ aquaria under the same light and temperature conditions, using filtered
(0.45 urn) estuarine water of salinity similar to that at the site of collec-
tion (9-11 ppt).

Atrazine Uptake and Phytotpxicity

     Matched sets of plant material (ca. 5 g wet wt) were selected  for use in a
series of 2 parallel treatments. The first treatment employed uniformly ring-
labelled [l4C]-atrazine (50 pci/mg, CIBA-GEIGY Corp ) dissolved in  methanol
to ascertain atrazine uptake, while the other used [i^C]sodium bicarbonate
(53 uCi/mmol) plus non-labelled atrazine dissolved in methanol to determine
the amount of C^-fixation at a given herbicide dosage. Plants were incubated
1n 2 I Erlenmeyer flasks at atrazine concentrations of 0.01,  0.05,  0.10, 0.45,
and 0,65 ppm.  A methanol vehicle control containing [*4C] sodium bicarbonate
plus the same amount of methanol as used with the atrazine and a standard
control with [l^C] sodium bicarbonate only were also performed.  The total
volume in each flask was adjusted to 1 t with ES medium, and all 4  flasks were
simultaneously incubated for 4 h with periodic (hourly) mixing by hand swirling.
At the end of the experiment plants were radioassayed as described  below.

^ime-Course Atrazine Uptake

     Clean plants (terminal 20 cm) were placed in replicate flasks  containing
either 0.02, 0.05 or 0.10 ppm 14C-atrazine (ring-labelled, 50 yci/mg) dis-
solved in 1.0 i. The flasks were incubated for a total of 24 h at constant
temperature and light. Plants were removed from the flasks at 0, 0.25, 0.50,
0.75, 1, 2, 4, 8, 12, and 24 h from the initial treatment, washed,  and processed
for radioassay as described below.

Photosynthetic Recovery

     Experiments to test photosynthetic recovery were Initiated by  placing
individually tagged plants in 300 ml BOD bottles containing filtered (1 urn)

                                      IV-3

-------
estuarine water. A control and 3 atrazine treatments (.005, .025,  and .100 ppm)
were used, with 5 replicate measurements of photosynthesis for each incubation.
Bottles containing plants were suspended in 55 i aquaria, with water circu-
lated vigorously using submersible pumps. Apparent photosynthesis  (Pa)  was
estimated by changes in dissolved oxygen (DO) during 2 h incubations, where DO
measurements were made with a polarographic oxygen electrode (Orbisphere Model
(2709).

     Two sets of experiments were conducted to Investigate photosynthetic
recovery. In the first of these, dosed plants were transferred to  atrazine-free
filtered estuarine water in 300ml BOD bottles at 2 h after initial atrazine
dosing, and Pa was measured again in a second 2 h incubation.  This was  repeated
two more times in new atrazine-free water,  and then the tagged plants were
returned to aquaria (one for each treatment) containing 20 i of filtered water
and maintained in the li^ht overnight. A final 2 h incubation  '-/as  conducted on
the following day, 24 h after initial washing to estimate Pa.  The  second
experiment involved washing plants in a larger volume of water and continued for
more than 3 d. Plants were returned to the  pre-incubation aquaria  following the
initial 2 h dosing period and allowed to flush in atrazine-free water (20 i)
for 3 h before being incubated in BOD bottles for 2 h to monitor photosynthesis.
This procedure was repeated at 29 and 77 h. At the end of each time interval
(3 h, 29 h, 77 h) the wash water in each aquarium was replaced with new filtered
water.

Atrazine Release

     Atrazine release was monitored for 0.10 ppm herbicide treated plants in
the first of the recovery experiments above. Here, 10 replicate plants  were
treated for 2 h with 0.10 ppm ring-labelled l^C-atrazine (50 uCi/mg). Five
of these plants were immediately processed  for radioassay as described  below.
The other 5 plants were placed in BOD bottles and treated as above. At
the end of each incubation in atrazine-free water, 1 ml  aliquots were taken
from each BOD bottle and placed in 10 ml of Aquasol-2 (New England Nuclear)
for subsequent radioassay. At the end of the entire series of  incubations and
washes, the plants were removed and dried for analysis of remaining 14C
content as described below.

     The release of l^C-atrazine Was also measured in a wash series of  various
solvents. Plants incubated for 2 h in 0.01, 0.05, 0.10, 0.45 and 0.65 ppm
14c-atrazine were subjected to a 4-step wash series. Four 1.0 t beakers were
arranged in the following order: (1) 500ml of estuarine water, (2) 0.1 %
Tween 80 (surfactant) in 500ml of estuarine water, (3) 500ml  of  estuarine
water, and (4) 500 ml of a 2% HC1 solution  in estuarine water.  The plants were
sequentially placed and agitated briefly in each of the beakers, blotted dry,
weighed, and dried at 80 C. Water from each of the 4 beakers was sampled (in
triplicate) after each wash for l^C-atrazine by placing a 1.0ml aliquot
into 10 ml Aquasol-2.
                                      IV-4

-------
*4C-Analysis of Plant Material

     All plant samples were dried at 80 C until constant weight (usually 24 h),
separated into root/rhizome (referred to as roots) and stem/leaf (referred to
as shoots) sections, weighed,  and ground to powder with mortar and pestle. An
aliquot (0.1-0.2 g) of the ground, dried plant material was placed in tubes and
digested with nitric acid according to the method of Lewis et al.  (1982). A
portion (1.0ml) of the digested plant material was placed in \  il  of Aquasol-
2. All counts were performed using a Packard Tri-Carb 460C Liq,. -,  Scintillation
Spectrometer. Quench corrections were made using the external standard channels
ratio method.
                             RESULTS AND DISCUSSION
Plant Uptake of Atrazine

     The movement of atrazine into £. perfoliatus is quite rapid, coming to
near equilibrium with the surrounding solution within 15 min (Fig. la). The
atrazine concentrations within the plant remained quite stable from the first
hour of incubation through the 24th hour (Fig. Ib).  A one-way ANOVA with
Student-Newman-Keuls (SNK) test of the means shows no significant difference
(p<0.05) between plant atrazine concentrations past 15 minutes of incubation.
Furthermore, varying solution concentrations of atrazine from .02 to .10 ppb
had little or no effect in the time required to attain equilibrium.

     There are few comparable studies involving herbicide uptake and subsequent
release by aquatic plant species. The most similar studies are ones involving
single-celled algae or bacteria. Vallentine and Bingham (1976) found that for
several algae (Scenedesmus. Chlamydomonas, Euglena) the major uptake occurred
within the first minute of incubation, with internal concentrations staying
relatively constant for the remaining experimental period (24 h).  Bohm and
Muller (1976), also working with Scenedesmus, found that cellular atrazine
uptake reached equilibrium with the culture solution in 4-8 h.

     The relative importance of root versus shoot uptake of atrazine appears
reasonably balanced (root uptake ranging from 16-40% of shoot uptake) when the
data are normalized on a per gram tissue basis (Table 1). However, when the
relatively low biomass of roots compared to shoots per plant is considered,
uptake by the root portion of the plant is far less significant.  In Table 1
uptake on a per gram root or shoot basis is compared to that on a total atrazine
per plant part basis, and it is apparent that actual uptake of atrazine per
plant by roots relative to shoots was very low. This estimate of root uptake
is probably lower than that occurring in the natural environment due to possible
injury to root systems during plant collection but still indicates a lesser
role of the root system in uptake of atrazine.  This is the same conclusion
reached by Forney and Davis (1981) for both P. perfoliatus and El odea canadensis
where they partitioned roots and shoots, dosTng one or the other or boTKI

     Since the roots and shoots of the plants were suspended in the same con-
centration of atrazine and the shoots contained a relatively high concentration

                                       IV-5

-------
l
                •H
                                    0.20
0.60
0.7B
                                                    TIME  (HR)
1.00
                Fig. 1. Atrazine uptake by  P. perfoliatus in (a) 24 hours at both 0.02 and
                       and 0.05 ppm and  (bj 1 hour at 0.10 ppm (x ± S.D.).

                                                     IV-6

-------
TABLE 1. Root:shoot ratio of ^C-atrazine uptake in P,  nerfoliatus.
  Atrazine
  Solution
Concentration
    (ppra)
 Root uptake per
  gram dry wt.
Shoot uptake per
  gram dry wt.
 Root uptake per plant
Shoot uptake per plant
     0.01

     0.05

     0.10

     0.45

     0.65
       0 '

       0.17

       0.16

       0.40

       0.39
         0.04

         0.07

         0.07

         0.16

         0.06
                                      IV-7

-------
i
of atrazine, translocate on of the compound from the roots to the shoots might
have been inhibited.   It is also not certain from these experiments whether
there was any translocation at all.  It is conceivable that some atrazine
diffused initially from roots to shoots. However, the low final atrazine con-
centrations in the roots indicate that this was ultimately unimportant.
Funderburk and Lawrence (1963) observed a movement of simazine from isolated
roots to shoots in Heteranthera dubia (waterstargrass), but this was over a 5 d
period compared to the 4 h incubation in the present study.  Thomas and Seaman
(1968) found no movement of simazine or atrazine when applied to the roots of
Potamogeton nodosus and similarly no translocation of atrazine was observed
from the roots of Potampgeton crispus (Sutton and Bingham 1968).  Therefore,
it might be concluded that translocation of significant quantities of atrazine
by Potamogeton is doubtful.

     Stem/leaf (shoot) uptake of atrazine (Fig. 2a) over the concentrations
tested exhibited a generally linear pattern reaching 0.69 vg atrazir.e/gdw
shoot (ppm) at 0.65 ppm aqueous concentration. Although plant dry weight is
generally considered to be a more reliable estimate of plant biomass, wet
weight was used in this study when considering atrazine uptake on a ppm basis.
The wet weight to dry weight ratio of shoots in _P. perfoliatus was determined
to be 10.3 ± 1.2 (n = 25) and for roots was 8.1 ± 0.6 (n = 25). The water
content, obviously a major portion of the biomass in a SAV species, must be
considered when discussing diffusion of atrazine.  Rhizome/root (root) uptake
was also linear with the internal concentration reaching 0.27 ppm in a treatment
of 0.65 ppm atrazine.

Atrazine Phytotoxicity

      Inhibition of photosynthesis with increasing atrazine concentrations
follows saturation kinetics with the maximum inhibitions (87 and 88%) occurring
at 0.45 and 0.65 ppm, respectively (Fig. 2a).  Dark ^-fixation for these
plants was determined to be 2% of the total C02 fixed in the light, and
uptake of atrazine was the same in the light as 1n the dark (this uptake
insensitivity to light for atrazine uptake has also been reported by Vallentine
and Bingham (1976) working with the alga Scenedesmus).  The 159 (atrazine
concentration at which photosynthesis is inhibited 50%) f or _P.  perfoliatus in
these experiments was 0.08 ppm. This is in relatively good agreement to the
ISQ found by Cunningham et al. (1982) who worked with plants from the same
_P. perfoliatus population as in this study. They reported a 50% reduction in
apparent photosynthesis (Pa) at an atrazine concentration of 0.13 ppm.  Of
interest is the fact that their experiments were long term (3 wks) and used
oxygen evolution to assess photosynthesis as compared to the experimental  time
(4 h) and methodology (14C) of our study.

     Outside of the work by Cunningham et al. (1982), it is difficult to compare
the results of this study on £. perfoliatus with others. Forney and Davis
(1981) did use as one of their SAV plants P_. perfoliatus but measured changes
in plant length, dry weight and mortality as indices of atrazine toxicity.
They estimated the I$Q in one experiment to be 0.474 ppm for length and
0.091 ppm for dry weight; however, in another experiment of longer duration
(4 wk) a 51% reduction in elongation rate occurred at 0.032 ppm. With El odea
canadensis, values of I$Q were obtained at 0.075 ppm and 0.123 ppm for elonga-

                                      IV-8

-------
   100
   80
   60
o
U
   40
   20
CO

«
UJ
I
CO
o
H
O
X
a.
                                                (A)
                                • Inhibl tion of


                            P hotosynthesis
                                     Atrazine  Uptake
                                   *

                                   *»
                                   0



                                   £
                                   o
                                                                  0.8
                                                                  0.6
                                                                  0.4
                                                                  0.2
            O.I      0.2    0.3     0.4    0.5    0.6   0.7


       ATRAZINE CONCENTRATION  IN THE  WATErt(ppm)
                                   o
                                   o
                                   -C
                                   CO


                                   UJ





                                   a.



                                   ui



                                   Nl


                                   ft

                                   »-
   100
   so
o
- 60
CD
   40
   20
               (B)
                    j_
                            _L
_L
_L
J.
            0.1     0.2    0.3     0.4     0.5    0.6    0.7

      ATRAZINE  CONCENTRATION  IN  THE SHOOTS

                          ^ng /gm   wet wt.

 Fig.  2.  (a) Atrazine uptake and  associated photosynthetic inhibition  in £.

         perfoliatus at atrazine  concentrations of 0.01 to 0.65 ppm.     ~

         (b) Internal atrazine concentration and associated photosynthetic

         inhibition in P. perfoliatus.
                                   IV-9

-------
tion and weight increment, respectively. While Correll and Wu (1982)  observed
a slight stimulation of apparent photosynthesis for P. pectinatus at  0.075 ppm
atrazine, reductions of 50% and 20% were found for Zahnichellia palustrus and
Vallisneria americana, respectively, at that concentration. Thus, it  would
appear that atrazine concentrations on the order of 0.100 ppm would effect an
    for most SAV species studied.
     The relationship of internal atrazine concentration in the shoot to
photosynthetic inhibition shows that the maximum photosynthetic inhibition
observed occurs when the internal atrazine concentration of the shoots reaches
about 0.45 ppm (Fig. 2b). A striking relationship is evident between the
external and internal atrazine concentrations in the shoots and the degree of
photosynthetic inhibition. An equilibrium appears to exist between internal
and external atrazine concentration only at the point in which maximum photo-
synthetic inhibition was observed. When the other atrazine solution concentra-
tions tested are examined, a slightly different case is found.  In Fig.  3 it
can be seen that the ratio of atrazine in the shoot to atrazine in solution  is
near unity at 0.45 ppm and increases (shoot accumulation) with lower concentra-
tions.   If atrazine uptake in plants is a passive diffusion process, one would
expect (barring significant degradation in a 4 h period) an equilibrium to be
established between the total internal  and external 14C-activity producing a
ratio of unity. However, the ratio for the shoots is always greater than one
indicating a possible atrazine accumulation against a concentration gradient.
This relative accumulation increased as the solution atrazine concentration
decreased.  The plant:solution ratios for roots versus atrazine concentrations
(Fig. 2) are all less than unity (with the exception of the roots in the 0.01 ppm
solutions).  Thus reflecting the relatively poor absorption of atrazine by
the P.  perfoliatus roots in these experiments.

     This apparent accumulation of atrazine at low concentrations has been
observed in algae by direct measurement (Bohm and Muller 1976) and indirectly
by measurement of the % atrazine remaining in solution (Vallentine and Bingham
1976). Atrazine is known to have a high binding affinity for organic matter
(Weber et al. 1969 ). The apparent accumulation at low atrazine concentrations
seen in this study and those of Bohm and Muller (1976) and Vallentine and
Bingham  (1976) could simply be a result of the fixed amount of atrazine sorbed
per gram dry weight plant material having a greater influence on the ratio of
internal to external atrazine as the external atrazine concentration decreases.
In support of this is the report by Geller (1979) who found a good correlation
between bacterial surface area and atrazine accumulation, i.e., increased
surface area yielded higher atrazine accumulation.

     Another possible explanation for this apparent accumulation of atrazine
in the shoots of £. perfoliatus at low atrazine concentrations may be found  in
the fact that photosynthetic inhibitors like atrazine bind to protein at their
sites of activity in the chloroplast (Pfister et al., 1980).  The total  number
of binding sites for atrazine in the chloroplasts of the plant would be a
finite number relative to the mass of photosynthetic units in the plant (the
number of sites is estimated to be between several hundred to twenty-five
hundred per chlorophyll molecule ( Izawa and Good, 1965; Tischer and Strotmann
et al. 1974)).  As atrazine diffuses into the plant chloroplasts and is bound
to the sites it would be removed from the solution phase thus reducing the

                                       IV- 10

-------
                                                                            •I
  Z B
  o
  J
  O 6
  0)
  a

  bl
  Z

  N
                    ATRAZINE  UPTAKE   EQUILIBRIA
                Shoots
   O.I      0.2     0.3      0.4     0.5     0.6      0.7
             SOLUTION  CONCENTRATION,ppm
                                                                 0.8
Fig. 3.
Ratio of atrazine uptake in £. perfoliatus to atrazine,concentration
in  the medium.
                                IV-11

-------
*
plants incubated for 4 hr in a concentration gradient of atrazine is shown in
Table 2. The purpose of using this wash sequence was to see the degree of
sorption of ^C-atrazine in the plant material. It was assumed that first
wash would remove the loosely associated surface film containing the !4C-atraz1ne
from the incubation solution, while the surfactant in the second wash would
remove physically bound chemicals. The third wash (filtered water only) was
simply to remove the surfactant from the plant and as a check on the amount of
14C-atrazine associated with the plant still removable by a water wash
alone. This would provide a somewhat clearer picture of the absolute effect of
the 4th wash, the 21 HC1 solution which was intended to shock the membrane
electrical potential allowing materials in the cytoplasm to flow out.

     The first water wash removed the most ^C-atrazine.  jhe surfactant
wash removed very little additonal herbicide and may just represent a carry-over
phenomenon from the first wash water. However, the acid wash did remove a
substantial amount of **C-atrazine.  The percentages 1n parentheses 1n Table 2
after the atrazine quantities indicate herbicide removal relative to the total
internal concentration (no percentages are given for the first wash due to
the variability of the amount of carry-over water from the incubation solutions).
The low percentage removed by wash |3 (just water) over the range of atrazine
concentrations tested indicates the thoroughness of the previous washes in
removing externally sorbed atrazine.

     The exact physio-chemical mechanisms of atrazine removal brought about
by the acid wash and the ^ocdMon of the atrazine within the plant material
are open questions.  One explanation is that atrazine is hydrolyzed to hydroxy-
atrazine by acid pH, facilitating the dissolution of this compound, which is
relatively more polar than atrazine. Therefore, the observed release of 14C-
atrazlne (ring-labelled) could have been an observation of l^C-hydroxyatrazine
going into solution only. Nevertheless, the high percentage of ^C-atrazlne
removed by the acid wash (371) does indicate that a substantial quantity of
atrazine sorbed by the plant material 1s located superficially.

Photosynthetic Recovery

     The photosynthetlc recovery of £. perfpliatus occurs as rapidly as uptake
and release of atrazine.  In both experimental procedures, the plants recovered
to greater than 751 of their maximum photosynthetlc potential within the
measured time of 2-3 h (Fig. 4a,b).  A rapid rate of recovery was apparent
within the first 2 h, with a slower recovery occurring over the next 10-80 h.
Although a SNK test shows no significant difference (p<.05) between the means
after 2-3 h, a slow but steady Increase in photosynthesis is evident (the
reason for the drop in photosynthesis seen in the 0.025 ppm treatment (Fig. 4a)
is unknown.  However, as seen in Fig. 4a, photosynthesis is still depressed
in the low volume wash experiment after 24 hours (a significant difference at
p<0.25 at all concentrations).  Even after 77 h of flushing in the large
volume experiment photosynthesis is still less than control plants in the
.10 ppm treated plants (Fig. 4b).  However, analysis of the plants in the
first recovery procedure exposed to 0.10 ppm of ^4C-atrazine indicated that
they contained no significant quantity of *4C at the end of 24 h of flushing.

     Cunningham et al. (1982) observed a partial recovery of photosynthesis by

                                     IV-12

-------
TABLE 2. The quantity of 14C-atrazine  (pg) removed  from_P.  perfoliatus
by the various washes following the four-hour Incubation period. ™
Atrazlne Solution
Concentration
0.01
0.05
0.10
0.45
0.65
"Numbers in parentheses
11
Water
0.14
0.10
0.26
1.14
1.54
are the pen
#2
Surfactant
0.01 (3)a
0.03 (5)
0.09 (11)
0.38 (16)
0.41 (13)
:ent of total f^a
#2
Water
0.00 (0)
0.00 (0)
0.00 (8)
0.19 (8)
0.20 (8)
nt 14C-atrazine
14
2% HC1
0.02 (5)
0.18 (29)
0.29 (37)
0.45 (19)
0.60 (19)
removed.
                                      IV-13

-------
.c ' w
»
^
o- 8
9
E
in
» 6
U)
X
H
TOSYN
*
o
X
* Z
»-
z
u
K
a) Experiment I

o
a: E
I * _
o *> t
0 0 a
0 tt






-


• n
CM
p E
• O -
' '
r
* ' L
;/ j •
X ]
11 \


P-I
* *
/
/ :
^
/
/' ^



ik
" i r
' '
'• '.
• '.i
;





r i






- ,

• ; '
^

1


j

•••



•;
'i
.'.



1
.

• .
/
/
/
/
/ :
^:
           a.
           a.
                 INITIAL
24
               TREATMENT TIME  SINCE INIT.AL TREATMENT (h)
r 10
^
T
«! s
(V
E
~*
to

Ul
X
^
> 4
CO
O
»-
0


"• z
.
z
e
IL

b.) Expert ment H



o
or _
jl
O
u





.




i *
i
|




s*

5
;!
::
N
O
r E
TT £X


S
/.
'A
y

/
a.

O
O
•
4.
i
\





••:
;!•;'



/•
:

:


1
T1
i
^

% .
>

r
/
^ j
/ .

/
n. 3
* INITIAL L

.














i .
a 1
U
n
/
j
r
r
f /

J
y
/
^
. /












1 .
1
.














4
,K
x
>
/
x
/
/
f
/

y
^
>
/
/

/
















29 77
I
                TREATMENT  TIME SINCE INITIAL TREATMENT(h)

Fig. 4. Photosynthetic recovery by P. perfoliatus following a 2 hour

        dosing period with 0.005, 07025, and 0.100 ppm atrazine determined
        over (a) 24 hours, and (b) 77 hours (x ± S.D.).


                                     IV-14

-------
P. perfoliatus after the initial atrazine dosing in their microcosm study. In
Their study, however, the recovery (Pa as measured by oxygen changes in the
microcosms) occurred 1 wk after dosing (0.13 ppm) while atrazine was still
present in the microcosm water (the atrazine concentration declined only 15%
during the 4 wk post-treatment period).  Although the experimental conditions
of their system and those of the present study were quite different, the combined
results do indicate the resilience of_P. perfoliatus to atrazine exposure.


                                  CONCLUSIONS


     The results obtained in this study indicate that Potamogeton perfoliatus
is a relatively sensitive plant to atrazine (150 = 0.075 ppm) when grown at 10 ppt
salinity. Root uptake of atrazine was small in relation to shoot uptake (~6
percent of shoot), and therefore, when considering short term photosynthetic
depression in SAV species, it is the atrazine in solution around the plant
shoots which is of most concern.  Uptake and release of atrazine is rapid
(ca. 1 h), but there is an indication of some residual  photosynthetic inhibi-
tion as long as 77 h following dosing eventhough no detectable atrazine could
be found in the plants.

     The sorption of atrazine to SAV surfaces (as well  as algal surfaces) leads
to amounts of atrazine associated with aquatic plant materials in higher
concentrations than that observed in the adjacent waters. This conceivably
could serve as a mechanism for bioaccumulation of atrazine in aquatic food
webs. However, Davis et al. (1979) in a study of the fate of atrazine in a salt
marsh ecosystem found no accumulation of atrazine in the animals (e.g., crabs,
snails) allowed to graze the dosed systems.

      In regard to the role of atrazine in the decline of SAV in the Chesapeake
Bay, it would appear from these data and the observed concentrations of atra-
zine in the Bay that photosynthetic losses would be ephemeral allowing the
plants to recover. The rapid degradation of atrazine in estuarine environments
(Jones et al. 1982) along with sorption to sediment and dilution through tidal
actions would keep SAV species from being exposed to lethal concentrations of
the herbicide. However, the ultimate effects of the presence of atrazine in
the water and the associated short-term photosynthetic losses on the biology of
tine plants is unknown. Especially of concern is any effect on reproductive
success.
                                      IV-15

-------
                                   REFERENCES


 1.   Bayley,  S., Stotts, V.D.,  Springer, P.P.,  and J.  Steenis. 1978. Changes
     in submerged aquatic macrophyte populations at the head of the Chesapeake
     Bay, 1958-1975. Estuaries  1:74-75.

 2.   Bohn,  H.H. and H.  Muller.  1976. Model  studies on  the accumulation of herbi-
     cides  by microalgae. Naturiwissenschaften  63:296.

 3.   Correll, D.L., Pierce, J.W., and T.L.  Wu.  1978. Herbicides and submerged
     plants in the Chesapeake Bay. Pages 858-877 j_n Proc. Symp. on Technicol.,
     Environmental, Socio-economical, and Regulatory Aspects of Coastal  Zone
     Management. Am. Soc. Cw. Eng.

 4.   Correll, D.L. and  T.L. Wu. 1982. Atrazine  toxicity to submersed vascular
     plants in simulated estuarine microcosms.  Aquatic Bot. (In press).

 5.   Cunningham, J.J.,  Kemp, W.M., Stevenson, J.C., and M.R. Lewis. 1982.  Pat-
     terns  of ecological and physiological  response and recovery of the  sub-
     merged macrophyte, Potamogeton perfoiliatus exposed to atrazine stress in
     estuarine microcosms^Report to U.S.  EPA, Annapolis, MD.


 6.   Davis, D.E., Weete, J.D.,  Pillai, C.G.P.,  Plumley, F.G., McEnerney, J.T.,
     Everest, J.W., Truelove, B., and A.M.  Diner. 1979. Atrazine fate and
     effecgts in a salt marsh.  EPA report no. 600/3-79-111. 84 pp.

 7.   Forney,  D.R., and  D.E. Daves. 1981. Effectgs of low concentrations  of herbi-
     cides  on submersed aquatic plants.  Weed Sci. 29:677-85.

 8.   Funderburk, H.H. Jr., and  J.M. Lawrence. 1963. Absorption and translocation
     of radioactive herbicides  in submersed and emersed aquatic weeds. Weed
     Res.,  3:304-311.

 9.   Geller,  A. 1979. Sorption  and desorption of atrazine by three bactgerial
     species  isolated from aquatic systr.ms. Arch. Environ. Contam. Toxicol.
     8:713-720.

10.   Izawa, S. and N.E. Good. 1965. The number  of sites sensitive to 3-(3,4-
     dichlorophenyl)-l,l-dimethyurea,3-(4-chlorophenyl)-l,l-dimethylurea
     and 2-chloro-4-(2-propylamino)-6-ethylamino-s-triazine in isolated
     chloroplasts. Biochim. Biophys. Acta 102:20-78.
                                     IV-16

-------
11.  Jones, T.W., Kemp, W.M., Stevenson, J.C. and J.C. Means.  1982.  Degrada-
     tion of atrazine in estuarine water/sediment systems and soils. J. Environ.
     Qual. 11:632-638.

12.  Kemp, W.M., Means, J.C., Jones, T.W.,  and J.C. Stevenson.  1982. Herbicides
     in Chesapeake Bay and their effects on submerged aquatic vegetation.
     Univ. Maryland Centr. Environ. Est. Stud. Ref. No. 1295-HPEL.
     Horn Pt. Env. Labs., Cambridge, Md. 82 pp.

13.  Lewis, M.R., Kemp, W.M., Cunningham, J.J., and J.C. Stevenson.  1982. A
     rapid technique for preparation of aquatic macrophyte samples for measuring
     14C incorporation. Aquatic Bot. 13:203-207.

14.  Muir, D.C.G., Yoo, J.Y., and B.E. Baker. 1978. Residues of atrazine and N-
     de-ethylated atrazine in water from five agricultural watersheds in Quebec.
     Arch. Environ. Contam. Toxicol. 7:221-224.

15.  Pfister, K., Steinback, K.E., Gardner, G., and C.J. Arntzen. 1980. Identi-
     fication of the photosystem II Herbicide binding protein.  PI. Physiol
     65(6): Supplement, Abstr. No. 48.

16.  Stevenson, J.C. and N.M. Confer. 1978. Summary of available information on
     Chesapeake Bay submerged vegetation. U.S. Department of Interior, Fish
     and Wildlife Service. (FWS/OBS-78/66). Washington, D.C. 335 pp.

17.  Sutton, D.L. and S.W. Bingham. 1968. Translocation patterns of simazine in
     Potamogeton crispus L. Proceed. N.E. Weed Control Conf. 22:357-361.

18.  Tischer, W. and H. Strotmann. 1977. Relationship between inhibitor binding
     by chloroplasts and inhibition of photosynthetic electron  transport.
     Biochem. Biophys. Acta 460:113-125.

19.  Thomas, T.M. and D.E. Seaman. 1968. Translocation studies  with Endothal-14C
     in Potamogeton nodosus Poir. Weed Res. 8:321-326.

20.  Triplett, G.B., Conner, B.J., and W.M. Edwards. 1978. Transport of atrazine
     and simazine in runoff from conventional and no-tillage crops.  J.
     Environ. Qual. 7:77-84.

21.  Vallentine, J.P. and S.W. Bingham. 1976. Influence of algae on amitrole and
     atrazine residues in water. J. Can. Bot. 54:2100-2107.

22.  Weber, J.B., S.B. Weed, and T.M. Ward. 1969. Adsorption of ^-triazines by
     soil organic matter. Weed Sci. 17:417-421.

23.  Wu, T.L. 1980. Dissipation of the herbicides atrazine and alachlor in a
     Maryland corn field. J. Environ. Qual. 9:459-465.
                                     IV-17

-------
                                   CHAPTER V
                UPTAKE AND °HYTOTOXICITY OF SOIL-SORBED ATRAZINE
           FOR THE SUBMERGED AQUATIC PLANT, POTAHOGETON PERFOLIATUS.
                                 T.W. Jones1'2

                                  P.S. Estes2
Salisbury State College
 Department of Biological Sciences
 Salisbury, MD.   21801

20regon State University
 Department of Zoology
 Corvalis, Oregon   97331
                                       V-i

-------
                                    ABSTRACT
     The effect of atrazine-sorbed soil on photosynthesis in Potamogeton per-
foliatus was investigated under laboratory conditions.  Leaves simultaneously
exposed to atrazine both in solution and sorbed to soil  sedimented on the
leaves exhibited a similar uptake of atrazine and associated photosynthetic
reduction as did leaves exposed to the same concentration of atrazine in solu-
tion only.  A small quantity of atrazine (0.19 ug/gdw leaf)  was found in
leaves treated with atrazine-sorbed soil at 120 ppb whereas  a significantly
larger amount (3.57 yg/gdw leaf) was present in leaves treated with dissolved
atrazine at a concentration of 100 ppb.  Based on these results, it is con-
cluded that atrazine sorbed to soil is less available for uptake by aquatic
plants than atrazine in solution.  Of greater physiological  concern is the
physical presence of the soil  on the leaves and the resultant attenuation of
licht.
                                      V-l

-------
                                  INTRODUCTION


     The ecological effects of herbicide runoff into estuarine aquatic environ-
ments like the Chesapeake Bay have received considerable attention recently
with a special emphasis toward the impctct on submerged aquatic vegetation
(SAV) (Stevenson and Confer 1978, Cornell et al. 1978).  Triazine herbicides
such as atrazine (2-chloro-4-ethylamino-6-isopropylamino-s-triazine) are
relatively mobile in the soil (Wauchope and Leonard, 1980J and do move from
agricultural fields. It has been observed that approximately 1% of the atrazine
applied to a field may ultimately enter nearby aquatic systems (Muir et al.,
1978; Triplett et al., 1978; Wu, 1980), most of which occurs during the first
major storm event following application (Muir et al., 1978; Triplett et al.,
1978; Wauchope, 1978).  This runoff can result in dissolved atrazine concentra-
tions ranging from less than 1 to 100 ppb in adjacent waters (Frank et al.,
1979; Frank and Sirons, 1979; Wu, 1980; Hershner et al. 1981).

     The partitioning of atrazine between the dissolved phase and the soil-
sorbed phase in the run-off component may be of key importance as to the avail-
ability of the herbicide for uptake by SAV species. Although the partition
coefficient (Kd = sorbed concentration * dissolved concentration at ecutlibrium)
for atrazine is variable due to soil parameters (most notably organic matter,
pH, and clay content), average values for many agricultural soil types are
usually between 1-5 (Talbert and Fletchall, 1965). Partition coefficients of
1-4 have also been reported for estuarine sediments (Means et al., 1980).
However, substantially higher Kd values (5-260) for suspended sediment in
run-off have been calculated (Correll and Wu, 1982). Since the Kd values are
routinely greater than 1, there is the potential for concentration of atrazine
within the suspended particulate fraction of field run-off material.  The
accumulation of this atrazine-sorbed sediment produces a microenvironment on
SAV leaf surfaces that theoretically could result in a high concentration of
atrazine in the interstitial water of that sediment.  Therefore, it has been
suggested that atrazine sorbed to suspended particulates may, in effect,
result in SAV exposure to elevated herbicide concentrations by sedimentation
of this material onto leaf surfaces (Correll and Wu, 1982).

     This research was initiated to examine the extent of uptake of atrazine
in P_. perfoliatus when the soil-sorbed herbicide is placed on plant leaves.
Atrazine uptake and the resulting photosynthetic depression in £. perfoliatus
leaves werrt investigated under the following conditions: 1) atrazine in solu-
tion and on the overlying soil, 2) atrazine in solution only, 3) atrazine on
the overlying soil  only.
                                      V-2

-------
                             MATERIALS AND METHODS


Sorption - Desorption of Atrazine on Soil

     The soil used throughout this study was Mattapex silt-clay collected from
an agricultural field that had not been tilled for four years. Two particle
size classes were tested for their sorptive properties. Che class contained
particles which passed through a 105 u sieve, but not through a 74 urn sieve.
The other contained particles which would pass through a 74ym, but not a
53ym sieve. These sizes were chosen because they could be suspended in a column
of water and yet would settle out within five minutes, suggesting that when
suspended in an aquatic environment, these particles might easily settle on
the leaves of submerged macrophytes.

     Sorption of atrazine was examined at a water:soil ratio of 5:1 using five
concentrations of uniformly ring-labelled 14C-atrazine in methanol (10, 50,
100, 500, 1000 ppb). Two-gram soil samples were added to 16 x 150 mm glass
screw cap tubes to which 10 ml of atrazine solution was then added. The soil
samples were allowed to equilibrate for 6 hours on a mechanical shaker at room
temperature after which they were centrifuged.  One-mi supernatant water samples
were taken from each tube and placed into 10 ml of Aquasol-2 (New England
Nuclear) for counting on a Packard Tri-Carb Model 460C Liquid Scintillation
spectrometer. The difference between the amount of atrazine originally in
solution and the amount remaining in solution after incubation was assumed to
be the amount sorbed to the soil.

     After water samples were removed from the tubes, the remaining supernatant
was aspirated off and replaced with 10 ml of deionized water.  The tubes were
shaken for 2 hours, centrifuged, and sampled as before to determine the amount
of atrazine which had desorbed from the soil. This entire procedure was repeated
to determine second-degree desorption. Wet weights and then dry weights of the
soil samples were taken to determine the interstitial water volume which was
used to correct desorption values.

 Plant Preparation

      Potampgeton perfoliatus (L) plants were collected from shallow waters of
the Choptank River estuary just prior to each experiment. Epiphytes and sedi-
ments were removed from the leaves manually and selected leaves from the ter-
minal 20 cm of at least 10 different plants were removed from the stems and
placed in filtered (.45 urn) Choptank water.

     Leaves were arranged in four rows on a rack consisting of a ring of PVC
pipe (150 mm inside diameter, 43 mm high) with a black plastic mesh (2.75 mm)
bottom with three supporting legs (15 mm high).  Four strands of monofilament

                                      V-3

-------
     In Treatment #2 leaves were exposed to 120 ppb atrazine on the soil with
no atrazine in solution. The other two bowls were identical  to those in Treat-
ment #1.  In Treatment #3, soil was applied without sorbed atrazine (me:hanol
served as a vehicle control), however, the water did contain 100 ppb atrazine.
Again the other bowls were identical to Treatments #1 and #2.


                              RESULTS & DISCUSSION


     The two particle size classes of Mattapex silt-clay examined for use in
this study exhibited similar adsorptive (Fig. 1) and desorptive (Fig. 2) char-
acteristics.  Ultimately, the larger size class (<105ym, >74ym) was chosen
over the smaller (<74um, >53ym) because the former has a greater tendency
to remain settled on £. perfoliatus leaves during the experimental procedures.
This larger size class has a partition coefficient (K/) for atrazine of 1.2
(rz = .99).

     The importance of achieving a uniform distribution of soil over the leaves
of a plant and the necessity of being able to apply and recover known quantities
of soil was addressed by the use of detached leaf experiments. Atrazine uptake
rates (in dissolved phase) by attached and detached leaves of £. perfoliatus
were not significantly (p<0.05) different.  The mean uptake of atrazine at a
dissolved concentration of 120 ppb (ug atrazine/gdw leaf) for attached
leaves was 6.31  ± 1.32 (x ± S.D.) and 5.97 ± 1.51 for detached leaves.
However, attached leaves did exhibit a higher mean photosynthetic rate (10.74 ±
2.01 mgC/gdw/hr) than did detached leaves (6.94 ± 1.93 mgC/gdw/hr).  Since
only relative photosynthetic differences among treatments was considered to be
of importance here, this photosynthetic rate difference was of no consequence.

     The results of experiments on atrazine uptake by leaves of !P. perfoliatijis
from sorbed-soil versus aqueous solution indicate the relative low availability
of soil-sorbed atrazine for plant uptake (Fig. 3). In Treatment 1, leaves that
were exposed to atrazine-sorbed soil and dissolved atrazine simultaneously
(at the proper K^ = 1.2) exhibited no significant difference (p<.01) in
uptake from leaves exposed to dissolved atrazine only (3.32 ± 0.31 versus
3.58 ± 0.21 u9 atrazine/gdw leaf, respectively). When leaves were exposed
to atrazine-sorbed soil only (Treatment 2), uptake was minimal (0.19 ± .03 ug/gdw)
as compared with uptake by leaves without soil exposed to an equal concentra-
tion of dissolved atrazine (3.57 ± .11 ug/gdw). The physical presence of
soil overlying the leaves did not impede or promote the uptake of atrazine
from solution (Treatment 3). The uptake by leaves with applied soil (no atra-
zine) and exposed to dissolved atrazine alone did not significantly differ
(P<.01) from the atrazine uptake by leaves without soil exposed to issolved
atrazine (4.55 ± .013 and 4.31 ± .06 wg atrazine/gdw, respectively).

     The results obtained in the atrazine-sorbed soil treatments above are
consistant with reported atrazine adsorption-desorption kinetics in relation
to waterrsediment ratios. As this ratio decreases, atrazine sorption to the
same sedimert (same K^) increases (Laskowski et al. 1980),  The waterrsediment
ratio in the microenvironment of pore-waters between adjacent sediment particles
accumulates on SAV leaf surfaces would be extremely low causing the atrazine

                                      V-4

-------
                                                                      GO

                                                                      6
                                                                          LU

                                                                          Z

                                                                          M
                                                                          <
                                                                          IT
                                                                          H

                                                                          <

                                                                          CO
                                                                          Z>
                                                                          o
                                                                          UJ
                                                                          Z>
                                                                          o
                                                                                    o
                                                                                    i
                                                                                   x
                                                                                   Ol
                                                                                   a.
                                                                      CM

                                                                      6
                                                                                   in
                                                                                   O
                                                                                    i
                                                                                   in
 i
 o
 ID
 10

 0)
 N
 0)

 (J
 s_
 (0
 a.
 C i-l
 o ••
   IT)

 O 4-
•r- O
 i. •!-
 O O
 (/I (/)
•o ..
 10 i-
   O)
 
-------
     1.0
o>
UJ
E  0.8
N

cr
I-
<

o

CD
cr
o
    0.6
u.
o
    0.4
    O.I
Z
UJ
U
Z
o
o
   0.05
                  O-
                                    1.00 ppm
                                    0.10 ppm
                                         0.01 ppm
           ORIGINAL        FIRST          SECOND

                           DESORPTION    DESORPTION

     Figure 2.  Atrazine desorption from Mattapex  silt-clay at a water:soil
              of 5:'.  Error bars represent range between the particle sizes uses
             (53-74 y and 75-105 u).
                               V-6

-------
8
*
o
o


bl


N
<
e
                                     i
             SOIL 8 WATER    SOIL ONLY     WATER ONLY


                      ATRAZINE  APPLICATION



                                      I    I LEAVES  W/ SOIL

                                      r^l LEAVES  W/0  SOIL


     Figure 3.  Atrazine  uptake in leaves of P. perfoliatus from atrazine-

              sorbed soil and from solut1on~(* ±  S.D.).
                              V-7

-------
J
to remain sorbed to the sediment. However, the upper surfaces of the accumu-
lated sediment would be exposed to the larger volume of water in the system
and the atrazine would desorb accordingly.

     The photosyr.thetic response of £. perfpliatus leaves to shading by varied
amounts of applied soil (0 to 9.93 g soil/gdw leaf) was quite variable (Fig.
4) and non-linear, appearing not to increase past about 4 g soil/gdw leaf at
which point the mean photosynthetic reduction was 28%. When atrazine and sur-
face soil are present in a system, the resulting photosynthetic reduction is
due to both physical shading of the leaves by the soil and to the presence of
atrazine (dissolved and/or sorbed to the soil). The mean value of 28% photosyn-
thetic reduction brought about by surface soil in the range of 4-10 g soil/gdw
leaf Fig. 4) was subtracted from the total photosynthetic reduction derived
experimentally to obtain the reduction due to atrazine alone (Table 1).  When
atrazine was present in solution, the resulting photosynthetic reduction
ranged from 52 to 69%. Additional atrazine sorbed to the surface soil did not
seem to result in any additive photosynthetic reduction. Also, when atrazine
was introduced on the soil only, very little photosynthetic reduction resulted,
again indicating that atrazine-sorbed soil was not taken up by the leaves of
_P. perfoliatus.

     It appears that the most important effect of soils (with or without atra-
zine) on plant leaves is attributable to the physical presence of sediment on
leaf upper surfaces. Published data on the effects of sediment on SAV leaf
surfaces on photosynthesis is wanting. However, data on epiphyte biomass on
SAV leaves are available. Epiphytes exert a similar influence on SAV species
as does settled soil in that they attenuate light.   It has been shown that
light to the leaf surface of Zostera maximum can be reduced by 90% by natural
epiphytic growth (Phillips et al. 1978, Borum and Wium-Andersen 1980) and
photosynthesis in Zostera can be reduced by as much as 31% (Sand-Jensen
1977) by overlying epiphytes.  Twilley et al.  (1982) found a 27% reduction 1n
SAV photosynthesis with epiphyte communities of 6.13 grams dry wt epiphyte
per gram dry wt SAV.  These values are surprisingly close to the maximum
amounts of soil (4-9 grams per gram dry wt leaf) applied to the leaves of the
present study and the average SAV photosynthetic reduction calculated (27%,
Table 1).  Apparently densities of naturally occurring epiphytes of the same
weight as soil on leaf surfaces correlate well in relation to photosynthetic
reduction.

     In conclusion, it would appear from the data of this study that soil-
sorbed atrazine is relatively unavailable for uptake by £. perfoliatus.  De-
sorption of atrazine from soil is rapid, and therefore, 7t is likely that the
soil deposited on SAV leaves would have an atrazine concentration in equili-
brium (K^) with the water. Forney and Davis (1981) came to a similar conclusion
following their study of 4 Chesapeake Bay SAV species and atrazine. They too
found that atrazine uptake from the water was tht main mode of entry of the
herbicide into the plant.  Furthermoie, atrazine degrades more rapidly to
hydroxyatrazine (nonphytotoxic) when it is in close proximity to soil surfaces
(Armstrong and Chesters 1968) so that the soil on leaf surfaces could even
serve to detoxify atrazine.
                                                      V-8

-------
cs>
o
tr
o
o

LJ
   40
    30
(/>
O  20

O

O.

Z  10
0
UJ

-------

                                   REFERENCES


 1.   Armstrong, D.E. and G.  Chesters. 1968. Adsorption catalyzed chemical
     hydrolysis of atrazine.  Environ. Sci.  Technol.  2:683-689.

 2.   Borum, J.  and S. Wiurn-Andersen. 1980.  Biomass  and production of epiphytes
     on eelgrass (Zostera marinum L.) in the presund,  Denmark.   Ophelia
     1:57-64.

 3.   Correll,  D.L., Pierce,  J.W., and T.L.  Wu.  1978. Herbicides and submerged
     olants in  the Chesapeake Bay. Pages 858-877 jji Proc.  Symp. on Technical,
     Environmental, Socio-economical, and Regulatory Aspects of Coastal Zone
     Management. Am. Soc. Av. Eng.

 4.   Correll,  D.L. and T. Wu. 1982. Aquatic toxicity to submerged vascular
     plants in  simulated estuarine microcosms.  Aquatic Botany (In press).

 5.   Frank, R., and G.J. Sirons.  1979. Atrazine: It's  use  in corn production
     and its loss to stream waters in southern  Chtario, 1975-1977. The Science
     of the Total Environment. 12:223-239.

 6.   Frank, R., G.J. Sirons,  R.L., Thomas,  K. McMillan. 1979. Triazine residues
     in suspended soilds (1974-1976) and water  (1977)  from the  mouths of Cana-
     dian streams flowing into the Great Lakes. Internat.  Assoc. Great Lakes
     Res. 5(2):131-138.

 7.   Hershner,  C., Ward, K.,  and  J. Illowsky. 1981.  The effects of atrazine on
     Zostera marina 1n the Chesapeake Bay,  Virginia. Interior report, EPA
     contract #R805953, Annapolis, MD.  169 pp.

 8.   Laskowski, D.A., and P.O. McCall. 1980. Determination of sorption constants
     in soil and sediments.  (Unpublished manuscript).

 9.   Means, J.C., T.W., Jones, T.S. Pait, R.D.  Wijayaratne. 1981. Adsorption
     of atrazine on Chesapeake Bay sediments and selected  soils. In: Submerged
     aquatic vegetation in Chesapeake Bay:  Its  ecological  role  inTay ecosystems
     and factors leading to its decline. Kemp,  W.M., J.C.  Stevenson, W. Boynton,
     and J.C.  Means (Eds.) Submitted to Chesapeake  Bay Program  USEPA, Annapolis,
     Maryland.  Ref. #UMCEES 81-28 HPEL.

10.   Muir, D.C.G., J.Y. Yoo,  B.E. Baker. 1978.  Residues of atrazine and N-
     deethylated atrazine in water from five agricultural  watersheds in Quebec.
     Arch. Environm. Contarn.  Toxicol. 7:221-235.
                                      V-10

-------
11.  Phillips, G.L., Eminson, D., and B. Moss. 1978. A mechanism to account for
     macorphyte decline in progressively eutrophicated freshwters.  Aquatic
     Bot. 4:103-126.

12.  Sand-Jen-.en, K. 1977. Effect of epiphytes on eelgrass photosynthesis.
     Aquatic Bot. 3:55-63.

13.  Stevenson, J.C. and N.M. Confer. 1978. Summary of available information
     on Chesapeake Bay submerged vegetation. U.S. Dept. of Interior.

14.  Talbert, R.E., and C.H. Fletchall.  1965. The adsorption of some s-atrazines
     in soils. Weeds 13:46-52.                                       ~

15.  Triplett, G.B., Jr., B.J. Conner, W.M. Edwards. 1978. Transport of atra-
     zine and simazien in runoff from conventional  and no-tillage corn.  J.
     Environ. Qual. 7(l}:77-84.

16.  Trfilley, R.R., W.M. Kemp, K.W. Staver, W.R. Boynton, J.C. Stevenson.
     1982. Effects of nutrient enrichment in experimental estuarine ponds
     containing submerged vascular plant communities. EPA Report, Grant No.
     R805932-01-1, Annapolis, Md.

17.  Wauchope, R.D. 1978. The pesticide content of surface water draining from
     agricultural fields - A review. J. Environ. Qual. 7(4):459-472.

18.  Wauchope, R.D., and R.A. Leonard. 1980. Maximum pesticide concentrations
     in agricultural runoff: A semi empirical prediction formula. J.  Environ.
     Qual. 9(4):665-672.

19.  Wu, T.L. 1980. Dissipation of the herbicides atrazine and alachlor in a
     Maryland corn field. J. Environ. Qual. 9(3):459-465.
                                      V-ll

-------
              CHAPTER VI
      DEGRADATION OF ATRAZINE IN
       ESTUARINE WATER/SEDIMENT
          SYSTEMS AND SOILS1
             T.  W. Jones2
              W. M. Kemp3
           J. C. Stevenson3
             J.  C. Means4
     Contribution No.  1341 of the
  University of Maryland Center for
 Environmental  and Estuarine Studies
          2Assistant Professor
       Salisbury State College
      Salisbury, Maryland   21801
                 and
     Visiting Assistant Professor
Horn Point Environmental  Laboratories
      Cambridge, Maryland   21613

          3Assistant Professor
Horn Point Environmental  Laboratories
      Cambridge', Maryland   21613

          **Ass1stant Professor
   Chesapeake Biological  Laboratory
      Solomons,  Maryland   20688
                 Vl-i

-------
                                ACKNOWLEDGEMENTS


     The authors acknowledge J. Metz and N. Kaumeyer for assistance 1n sample
analysis, and Drs. Larry Ballantine and Lloy Newby of the CIBA-GEIGY Corpora-
tion, Agricultural Division, Greensboro, N.C. 27409, for analytical  advice and
atrazine compounds.  This work was supported by the US Environmental  Protection
Agency grant #R 805932-01-1 through the Chesapake Bay Program.
                                       VI-11

-------
i
                                                ABSTRACT
                    Herbicides have been postulated as a cause of the disappearance of sub-
               merged aquatic vegetation in the Chesapeake Bay. This research was undertaken
               to determine the longevity of 2-chloro-4-ethylamino-6-isopropylamino-£-
               triazine (atrazine) in two estuarine water/sediment microcosm systems and two
               agricultural soil systems over an 80-day period under aerobic and low-oxygen
               conditions. Atrazine degradation proceeded more rapidly in the estuarine sys-
               tems than in the soil systems. The disappearance of atr^p.^ from the estua-
               rine water was relatively rapid with the half-life (50% remaining in the
               wat^r column) of the parent compound ranging from 3-12 days. Atrazine half-
               life was determined to be 15 and 20 days for the two estuarine sediments and
               330 and 385 days for the two agricultural soils. Hydroxyatrazine (2-hydroxy-
               4-ethylamino-6-isopropylamino-_s-triazine) was the major short-term metabolite
               in both the estuarine and terrestrial systems. By the 21st day of the experi-
               ment, the % rjf total extracted residues corresponding to atrazine, and hy-
               drcxyatrazine, were: 65, 20 and 10, 85, for the two estuarine systems; and
               66, 29, and 93, 5 for the two soil systems. Decreased oxygen levels had
               little effect on atrazine degradation in the experimental  systems. The rapid
               degradation of atrazine to hydroxyatrazine in estuarine water and sediment
               indicates a low probability for the accumulation of atrazine in the estuary.

               Additional index words: Herbicide, estuarine, decomposition, sorption, half-
                                       life, metabolites
                                                    VI-1

-------
                               INTRODUCTION


     Over the last several decades agricultural  production in the United
States increased largely through the development of efficient cropping
methods (26). Among the more important factors contributing to improved agri-
cultural yields, has been the extensive use of herbicides for weed control.
Atrazine (2-chloro-4-ethylamino-6-isopropylamino-js-triazine) a compound used
to control broadleaf weeds, has found widespread application for corn and
sorghum crops since its introduction in 1958, and it is currently being
tested for pine plantations (6).

     Invariably a fraction of the herbicide applied to agricultural fields is
transported to watercourses via surface runoff and groundwater interflow. Atr-
azine and its metabolites have been observed in freshwater streams contiguous
to agricultural land where it was found that from about 0.1% to 3.0% of the
atrazine placed on the fields is lost to the aquatic environment (9, 10, 18).
Most research concerning herbicide losses from agricultural soils has been
conducted at inland sites; however, a few recent studies have focused on the
coastal plain. For example, Wu (28) has reported a leakage rate of 1% for
atrazine in a Chesapeake Bay watershed, producing aqueous concentrations
averaging 17 ppb.

     The effect of these herbicides on non-target species of aquatic organ-
isms is of ecological concern. Generally, concentrations of atrazine in the
range of those observed in runoff are potentially toxic to a variety of estua-
rine plants (3, 11, 24). A key factor contributing to the phytotoxicity of a
herbicide such as atrazine is its longevity or rate of degradation. Atrazine
sorption and degradation is influenced by pH, temperature, and soil moisture,
as well as clay and organic content of soils (1, 14, 25). Apparently, atra-
zine persists in most soil environments for a relatively long duration, with
a half-life exceeding one year under some conditions.

     While Goswarni and Green (8) had suggested that the relatively high pH of
estuarine and coastal environments (> 7.0) would cause atrazine degradation
to be slowed, little empirical evidence has been available to substantiate
this hypothesis. Some preliminary results indicate that degradation of atra-
zine may proceed quite rapidly in a slightly brackish environment (2). How-
ever, their experimental system did not make direct comparison to soil
systems.

     The purpose of this study was to investigate atrazine sorption and degra-
dation in estuarine water and sediments, as well as terrestrial soils, under
a range of environmental conditions in the same experimental design.
                                     VI-2

-------
                          MATERIALS AND METHODS
Sediment/Soil Samples

     Water and surface sediments (top 30 cm) were collected from inshore loca-
tions in early July 1979, to represent two salinity and geographical areas of
the Upper Chesapeake Bay region. The Tangier Sound sample, 15 parts per
thousand salinity, represented an open bay site while the Choptank Estuary
sample, 8 parts per thousand salinity, was hydrographically more protected.
The estuarine waters were filtered (0.45 ym) and the estuarine sediments
sieved (2 mm) and kept moist at 5 C.

     Two soils common to the Eastern Shore watershed of the Bay were col-
lected at the same time for direct comparison of degradation rates in terres-
trial environments. Sassafras sandy loam (Typic Hapludult), a well drained
upland soil and Mattapex silty loam (Aquic Hapludult), a poorly drained low-
land soil, were obtained (top 30 cm) from sites where atrazine had not been
applied for over 3 yrs. The physical characteristics of the sediments and
soils in this study are listed in Table 1. The soils were also sieved (2 mm)
and held at 5 C at their ambient moisture content.

Atrazine Applications and Sampling

     For the estuarine samples, 600 g of wet-sediment (40% water, by weight)
and 500 ml of water from the respective sampling areas were added to each of
6 flasks (2 «,), 3 of which remained open to the atmosphere via a cotton-
stoppered glass tube through a rubber stopper (aerobic) and the other 3
sealed with rubber stoppers (low-oxygen). The sediments were treated with uni-
formly ring-labeled atrazine (50 yCi/mg, 95% purity) by the introduction of a
slurry containing 1 g of the respective sediment type and 5 yCi (100 mg) of
  C-atrazine in methanol yielding an initial atrazine concentration in each
flask of 0.10 ppm (water and sediment). All the flasks were maintained in a
water bath at ambient Choptank River temperatures (12-35 C) in full sunlight
starting in mid-July 1979 and ending in late October 1979. At each sampling
period (days 1, 7, 14, 21, 28, 45, 59, and 80), 1 ml of water from each flask
was passed through a 0.45  m syringe filter directly into a scintillation ,.
vial containing 10 ml of Aquasol-2  (New England Nuclear) to determine the   C
activity in the aqueous fraction. The filter was.then placed in another scin-
tillation vial with Aquasol-2 to determine the   C present in the suspended
particulate fraction.

     A 5 ml water sample was also taken from each flask at each sampling
period and extracted twice with a total of 10 ml of chloroform. The chloro-
form (non-polar) extract and the residual 5 ml of water (polar) were evapo-
rated at room temperature and stored at -b C until analyzed for the parent
compound and its major metabolites.


                                    VI-3

-------
f:
               TABLE 1.   Physical  and chemical characteristics of sediments  and  soils.t


Sediments
CHOPTANK
TANGIER
_So11s_
SASSAFRAS
MATT APEX
Organic
pH Carbon Sand Silt
%
5.4 0.55 49 34
4.4 0.85 59 28

5.5 0.91 69 14
6.4 0.92 16 52
Clay

17
13

17
32
               tSediments and  soils were seived to 2 mm prior to analysis.
                                                    VI-4

-------
     Sediment (2 ± 0.5 g) was removed from each flask at each sampling period
using a 2 cm diameter glass corer. The sediment was placed into a glass vial
containing 5 ml of 10 % aqueous acetonitrile and stored at -5 C until ex-
tracted. Following sampling, each flask was lightly swirled to resuspend the
surface sediments.

     A small quantity (3 mg) of uniformly ring-labeled   C-atrazine (150 yci)
dissolved in 100 ml of methanol was applied to 600 g subsamples each of both
Sassafras and Mattapex soils and the methanol allowed to evaporate while the
soils were thoroughly mixed. From these subsamples, 100 g aliquots of each
soil was placed into flasks (2 I) containing 300 g of untreated soil, and the
flasks shaken vigorously. To simulate natural soil moisture conditions for
these two soils, water was added to the samples of Sassafras soil to bring
them to 60 percent water-holding capacity (17% water by weight) and to the
Mattapex soil to bring them to 100 percent water-holding capacity (50% water,
by weight). The atrazine concentrations in the flask were 6.4 and 5.0 ppm
(water and soil) for the Sassafras and Mattapex soils, respectively. Three of
the flasks for each soil type were maintained under aerobic conditions and
three under low-oxygen conditions as described for the estuarine flasks. All
flasks were held in full sunlight at ambient air temperature over the same
period as the estuarine flasks. The soil temperature reached a maximum of 6 C
higher than the external air temperature on the hottest day recorded -- 36 C.
At each sampling period (same as estuarine samples), a total of 2 t 0.5 g of
soil were removed from three different loc?tions in each flask and collec-
tively placed in a glass vial containing 5 ml of 10 % aqueous acetonitrile
and maintained at -5 C. (Flasks were weighed at each sampling period and water
added as necessary. The   C loss due to evolution of   C0« from metabolized
atrazine was not assayed.

Extraction From Sediment and Soil
     The sediment and soil samples were refluxed in 20 ml of acetonitrile for
4 h. After refluxing, 1 ml aliquot from each extract was radioassayed for   C
activity as described previously. The remaining acetonitrile was refrigerated
at -5 C until further analysis by thin layer chromatography (TLC). To deter-
mine extractability, selected sediment/soil samples previously, extracted by
acetonitrile were combusted usina.a Harvey oxidizer and the   C0~ trapped in
Oxifluor (New England Nuclear).   C-glucose standards were combusted to estab-
lish instrument efficiency.

Thin Layer Chromatographic Analysis of Residues

     The dried chloroform and water extract residues were dissolved in 0.2 ml
of chloroform and 0.2 ml of methanol, respectively, and spotted on 20 x 20
glass plates with a 250 ym layer of silica gel 60 F-254. The plates were de-
veloped to a 15 cm-solvent front in benzene-acetic acid-water (60:40:3.
v/v/v). Atrazine, and three of its major degradation products, 2-chToro-4-
amino-6-i sopropylamino-s-tri azine (N-de-ethylated), 2-ch1oro-4-ethylamino-6-
amino-£-triazine (N-de-Tsopropylated), and 2-hydroxy-4-etnylamino-6-
isopropylamino-£-triazine (hydroxyatrazine), were co-chromatogramed on each
plate. Location of the standard compounds was determined with an ultraviolet
lamp (254 nm) and location of the extracted compounds by autoradiography


                                     VI-5

-------
using Kodak X-Omat R XR-5 film with an exposure time of 14 days. Typical RF
values of the standards were 0.90, 0.78, 0.74, and 0.25 for atrazine, N-de-
ethylated, N-de-isopropylated, and hydroxyatrazine,  respectively (the ex-
tracted samples had Rf values slightly less than the standards possibly due
to retardation by salts in the extracted samples).
         14
     The   C present in each spot appearing on the film was determined by re-
moving the silica gel from the plate (2 cm diameter) and placing it into
10 ml of Aquasol-2. The areas between the origin and hydroxyatrazine and be-
tween hydroxyatrazine and the N-dealkylat'id compounds were also removed and
counted separately. Since the RF values for the N-de-ethylated and N-de-
istpropylated daughter compounds were close (Rp of 0.78 and 0.74,  respec-
tively), the silica gel from those areas was combined and termed the "dealky-
lated products." The acetonitrile extracts of the sediments and soils were
concentrated under low heat and spotted onto the TLC plates and processed as
above.
                          RESULTS AND DISCUSSION

^C-Atrazine Disappearance from the Mater

     Atrazine removal from the water can occur via both sorption    to  the
sediment  and  degradation  in  the  water  to  metabolic products which may
be sorbed on the sediment (Fig. 1). After atrazine sorption to sediments, the
compound could be degraded to metabolic compounds and subsequently desorbed
into the water column. This mechanism may be of some importance, since atra-
zine degrada-tion in sterile, filtered water systems has been shown to be ex-
tremely slow P, 2, 7). The dashed lines in Fig. 1 indicate the suspected
lesser significance of atrazine degradation in the water.  Considering the
relatively short duration of our study period (80 days) and the reported slow
ring-cleavage of atrazine (5, 29), the evolution of   CO- by biological  degra-
dation was not mon-'tored.
                    14
     The removal of   C-activity from the water in both estuarine systems was
a relatively rapid process following first-order kinetics (Fig. 2). The
aerobic systems were found to have a significantly faster disappearance of
  C activity from the water than the low-oxygen treatments (p<0.01). This sig-
nificance difference comes about mainly because of the slower removal of the
  C-atrazine metabolites in the anaerobic systems after 21 d. This apparent
retardation of atrazine residue sorption to sediments due to low-oxygen
levels may be of significance in estuarine environments where the redox poten-
tial of surface sediments is low. Slowed sorption of atrazine residues out of
the water would increase the residence time of these compounds in the aqueous
environment permitting an increased uptake by planktonic organisms.
                                   14
     In addition, disappearance of   C from the water in the Tangier systems
was significantly more rapid than in the Choptank systems (p<0.01). The ini-
tial removal rates (through day 21) for the 4 systems were 1.05, 1.04, 1.10,
and 1.07 %/d for the Choptank aerobic, Choptank low-oxygen, Tangier aerobic,
and Tangier low-oxygen systems, respectively. The half-lives of total atra-
zine residues in the water (based only on   C activity) were 6 and 8 d for
the Tangier aerobic and Tangier low-oxygen systems, respectively, and 17 d


                                   VI-6

-------
  WATER
                   ATRAZINE
    DEALKYLATED
      ATRAZINE
                                      Degradation
HYDROXY-
ATRAZINE
                                    cb	Sor ption
                    ATRAZiNE
                                     -Degradation
    DEALKYLATED
     ATRAZINE
 SEDIMENT.
HYDRJXY-
ATRAZINE
                   rJON-EXTRACTABLEV
                      ATRAZINE     )
                      RESIDUES   J.[
Figure 1.  A conceptual model suggesting the fate of atrazine in the estuarine
       water/sediment systems.
                         Ml-7

-------
                        CHOPTANK   ESTUARY
LJ
H
    80
    60
    40
>   20
!-
\
             (a)
         AEROB I C
  \
\
                                                  LOW-OXYGEN
                                                               •i

             20      40      60        0       20      40     60
                         TANG  IE R   SOUN  D
                     (c)
                  AEROBIC
               Total  Residues
          1   i
                      l    1
                                              (d )
                                        LOW-OXYgcN
                                                                1    1
              20      40
                     60
    20     40     60
              DAYS   FROM  START  OF EXPERIMENT
    Figure 2.  The disappearance of total radioactivity and atrazine from estuarine
             water.
                                     VI-8

-------
t!
                   for the Choptank  systems.

                        These  rates  of disappearance from the water were much more rapid than
                   those  observed  by Ballantine  et  al.  (2)  in their microcosn study, where they
                   found  a half-life of  75 d for total   C-atrazine residues in the water. The
                   periodic  resuspension of the  sediment at  sampling times and the different
                   water  to  sediment ratio of the two  studies (10:1 in Ballantine  et al. vs.
                   2:1 in this study) may have produced faster disappearance rates. Decreases in
                   the water to soil ratio have  been found to lead to increased sorption of atra-
                   zine  (4).

                   Atrazine  Degradation  in Estuarine Water

                        A strong substrate-product  relationship between atrazine and hydroxyatra-
                   zine was  evident  witii a low but  relaJ'ively constant percentage of the   C
                   activity  associated with the  dealkylated  compounds in both salinity systems
                   (Fig.  3). Unlike  the  situation discussed  above for the total residues
                   (Fig.  2), no significant difference (p<0.01) was observed for atrazine degra-
                   dation under different oxygen levels within salinity systems, but degradation
                   in the Tangier  systems occurred  significantly faster (p<0.01) than in the
                   Choptank  systems. The average coefficient of variation for triplicate observa-
                   tions  of  %  atrazine at each sampling time were 19.9, 25.7, 30.3, and 25.0%
                   for the Choptank  aerobic, Choptank  low-oxygen, Tangier aerobic, and Tangier
                   low-oxygen  systems, respectively.

                        The  composition  of the total extract only (both chloroform and water
                   residual) is,represented by Fig. 3,  and does not take into account the total
                   decrease  of  C activity in the  water over time (Fig. 2). Therefore, the
                   half-life of atrazine in the  water  (including both sorption and degradation)
                   was 9  d for the aerobic and 12 d for the  low-oxygen Choptank systems, while
                   being  about 3 d for both Tangier systems. Again, these rates represent a more
                   rapid  disappearance of atrazine  from the water than observed by Ballantine
                   et al. (2),  where they found  the half-life of atrazine in the water to be
                   30 d.

                                                                      14
                        The  observed kinetics of the removal of total   C from the water and the
                   degradation rate  of atrazine  in  the estuarine systems may result from several
                   associated  mechanisms: (1) The rapid hydrolysis of atrazine to hydroxyatra-
                   zine in the water linked to the  greater sorptive properties of the relatively
                   polar  hydroxyatrazine to soil (sediment)  (5) would accelerate removal of   C
                   from the  water; and/or (2)    C-labeled atrazine may be sorbed to sediments
                   initiating  a rapid hydrolysis to hydroxyatrazine, whereupon the metabolite
                   desorbs back to the water (see Fig.  1). Catalysis of j>-triazines to hydroxy-
                   triazine  has been shown to occur once the herbicide has been sorbed to clay
                   and organic matter surfaces (1,  20,  22). For example, Armstrong et al.  (1)
                   found  a 10-fold increase in the  rate of atrazine hydrolysis when sterilized
                   soil was  added  to a solution  containing atrazine, while Ballantine et al. (2)
                   detected  essentially  no atrazine degradation over a 4 mo.period in synthetic
                   seawater  which  contained no sediment. Geller (7) using   C-atrazine in
                   sterile 0.02 M  phosphate buffer  (pH 7.2) found  less than 5% of the atrazine
                   had degraded after 3  mo.
                                                       VI-9

-------
                            WATER

100
80
CO
; uj 60
• ID
5 40
1 CO
£ 20
-! o
a
CHOPTANK,+ 02
" •• IH KB 9B F8 01 §•

-





j
7~J

"
*


^
•. :• O.KXXXXXX^S



I?
•
*
1
f:!



m f



J

* •
7-7


• i I i i i




b) CHOPTANK ,-02
-
~
-
••
2
kX\\\\\\XXXX\



P'V
XXXXXXXXXXXXI



/


**•
:.;.[xx\\xxxxxx


•'.
fcsX^C^s^. •••.•:•••
|

™
7


a
*
" •



i i i i v i i
J 7 14 21 28 45 59 80 7 14 21 28 45 59 80
• O
H
! U_
I 0
j K 100
z
u 80
O
o:
UJ 60
0.
40

20
n
c) TANG
-
-
»

•
-
-
-
-
-
T-
.'_
.
".
• .
5?
z
/;
y
I










m

• *
*
•.
• •

. *
•
i









-
,
•



•
*
.
Y,










ER, + 02
« *

•'/
. •
\>
','
•".
*
f *
r
-------
                  Atrazine degradation was most rapid in tb* Tangier water (Fig. 3) con-
             committant with the most rapid removal of total   C from the water (Fig. 2).
             Conversely, the Choptank water displayed a slower atrazine degradation and
             sorption rate with a greater percentage of dealkylated products being
             present. The distribution .-oefficients [K. = (yg/g adsorbed)/(yg/ml solu-
             tion)] of the two sediment types used in She present study were 1.75 and 2.56
             for the Choptank and Tangier sediments, respectively (16). The greater adsorp-
             tive properties of the Tangier sediment correlates with the observed shorter
             half-life of atrazine in the Tangier water in relation to the Choptank
             tending to support the sorption-degradation-desorption mechanism.

             Atrazine Degradation in Estuarine Sediment
                                 14
                  The % of total   C-atrazine associated with suspended particulate matter
             (0.45 urn) in the water proved to be insignificant. The flasks remained undis-
             turbed following sampling allowing,particulate matter to settle. Only  . very
             small (initially 10%) quantity of   C was retained by the filters themselves,
             and the magnitude of this residue diminished over time as the % of the more
             polar hydroxyatrazine metabolite increased.

                  The rate of atrazine degradation and the appearance of the metabolic
             products in the Tangier and Choptank sediments followed simple,  first-order
             kinetics, with the linearized (log^transformed) data producing highly signifi-
             cant (p<0.01) regressions having r  values in excess of 0.80. The distribu-
             tion of residues over time followed a pattern similiar to the water (Fig. 4);
             however, this data represents only the extractable portion of the total radio-
             activity in the sediment. Extractability decreased over time (Table 2) paral-
             lel to the kinetics of atrazine degradation to hydroxyatrazine for each sedi-
             ment type. For example, the rate of hydroxyatrazine production increased with
             the % of non-extractable   C more rapidly in the Tangier than,the Choptank
             sediment. By day 80 of the experiment, only 29% of the total   C was extract-
             able from the Tangier sediment, of which 90%.was hydroxyatrazine (Table 2,
;             Fig. 4). On the same day, 37% of the total   C was extractable from the,Chop-
             tank sediment and of this 70% was hydroxyatrazine. The non-extractable   C
             atrazine activity most likely consists largely of further degraded metabo-
,;             lites, as well as conjugates with soil organic matter (15). This view is sup-
j             ported by the decrease in % of hydroxyatrazine in the Tangier aerobic system
•             from day 14 (30%) to day 45 (16%) and the concommittant increases in non-
t             extractable activity (57 and 75%, respectively). This suggests that the rela-
|             tive availability of metabolic products for further degradation was greater
|             earlier in the experiment for the Tangier system than for the Choptank. In
f             the combined water/sediment Choptank system, the hydroxyatrazine % increased
             from day 14 to day 45 (35 to 61%, respectively), re-emphasizing the rela-
             tively slower degradation rates of that system.

                  A similar pattern was reported by Dao et al. (5) for 3 soils held at 0.1
             bar moisture content at 30 C. They found that the extraction efficiency
             ranged from 20 to 40% after 2 mo and that the major portion of the extract
             was comprised of polar compounds (most likely hydroxyanalogs of atrazine).
             Their extraction efficiencies compare well to the 30 to 40% extraction effi-
             ciency found after 2.5 mo for the sediments in our study (Table 2). Ballan-
             tine et al. (2), using a similar extraction procedure as in our study, found
             that 53% of the radioactivity in the sediment was non-extractable after 1 mo.


                                                  VI-11

-------
r
•' SED
CO
UJ
3 100
o
CO on
UJ B°
cr
.60
UJ
CO 40
0 2°
^ 0
M E N T S
a)CHOPTANK, + 02
-
-
•i
\
H B • • B 9





£
^ «
•
/<

*
*
• •
•
/

i

* *
*>
*
•v

i i i i i i i



b) CHOPTANK,-02
:H M • • • • a

/

1
/
r

* •
^ •
* •
\
/

:
\

IBIBH
^

r
7
/

•
/

i i i i * 1*1
f~ 7 14 21 28 45 59 80 7 14 2 28 45 59 80
UJ
i—
o 100
1- •
u. 8°
U-
O
60
z
ui 40
O
cr
uj 20
Q.
0
c) TANGI ER , + 0
.
-

—
-
-
-
-
™
-
™

*
.

?
/,
X
*o
''O
R B fl R H H















*
X
/
















•.•
' y









.

.
.
•
•
•
.
V









.

*.
*
•
*
.^
•I

^™JP
^









•
t






"7









•
«
• .
*
*
*
• »%
*
•
• •
..









i i i i i i i











d) TANGI ER ,-o2
- H m H H M hi M
-

~
-
-
_
-
•
-
;





$
/
/














*
?
^
^
i i










•^


•*
^ •
V









.
.
m
•
•
.
;

*/
















".
• •
/









• .
•

•
.*.
*
*
•
<-r

















7—









l i i l >
7 14 2 28 45 59 80 7 14 21 28 45 59 80
          DAYS   FROM START  OF  EXPERIMENT

            EZ2  Atrazine  Pa ren t
            l-'-'-'-J  Hydroxya t razi ne
            H  Oealkylated Atrazine
Figure 4. The percent composition of 14C-atrazine metabolites In the total extracted
        residues from the estuarlne sediments.
                               VI-12

-------
TABLE 2.  Extraction of ll»C-atrazine and Its metabolites with acetonitrile
          from sediments and solls.t

Days
1
14
45
80
108
Sediment
Choptank
60
45
28
37
21
Source
Tangier
50
24
19
29
22
Soil
Sassafras
94
83
74
65
23

Mattapex
87
66
35
20
15
tData from aerobic samples only, determined by combustion.
                                         VI-13

-------
     A consideration in comparing the results of the Choptank and Tangier sys-
tems is that they may contain very different populations of microorganisms
with different atrazine degradation potential. The effect that environmental
factors might have on the distribution of fungi and bacteria, the major bio-
logical degradsrs of atrazine in the soi? (12, 13, 29), is not clear at this
time. Several fungal species isolated from the Chesapeake Bay may degrade
atrazine (Speedy, pers. com). Never-the-less, the low concentrations of deal-
kylated atrazine metabolites and the rapid appearance of hydroxyatrazine
would suggest a minor role for microorganisms in the degradation of atrazine
(23).

Atrazine Degradation in Soils

     The apparent kinetics of atrazine degradation (considering only extract-
able residues) in the Sassafras sandy loam and Mattapex silty loam soils were
similar (Fig. 5) but markedly different from the estuarine sediments. Less
atrazine was degraded in the soils compared to the estuarine sediments after
80 days. The extractable percentage of   C in the soils over time (Table 2)
shows that the atrazine metabolites on the Sassafras soils were more loosely
bound tl.an those on the Mattapex soil (65% extractable versus 20% extractable
at day 80, respectively). If one assumes that the non-extractable activity is
something other than the parent compound, atrazine degradation in the Matta-
pex soil was much more rapid than in the Sassafras.

     The two major physical differences between the Sassafras and Mattapex
systems were the moisture content and the percent silt and clay (60, 14, 10%,
and 100, 52, 32%, respectively). Degradation of atrazine to hydroxyatrazine
in the Matttapex soil most probably would have been slowed by the high mois-
ture content. High soil moisture content may promote ring-cleavage of atra-
zine (19) but may also retard the conversion of atrazine to hydroxyatrazine
due to reduced adsorption onto the soil (4) and subsequent hydrolysis (1).
Conversely, the high clay content in the Mattapex soil may have accelerated
atrazine hydrolysis due to sorption on the highly acidic clay surfaces (21,
27). However, the exact source for the difference in degradation rates be-
tween the Sassafras and Mattapex soils cannot be determined from the current
data.

     The issue of extractability has unfortunately received too little atten-
tion in the herbicide literature. The fate of atrazine can only be accurately
determined when the exact compositon of the non-extractable residues is
known. The degradation rate of atrazine can vary markedly depending on how
the non-extractable portion is considered. The two extreme cases are where
(1) the non-extractable portion is considered to be other than the parent com-
pound, and (2) the non-extractable portion contains the same compounds on a
percentage basis as the extractable portion (Table 3). Therefore, the range
of half-lives for atrazine calculated from the slopes of the first-order
plots in the 4 experimental systems of this study were 13-15, 16-20, 110-330,
and 36-385 days for the Tangier, Choptank, Sassafras, and Mattapex systems,
respectively. The effect is most severe in the case for the agricultural
soils, where half-lives may differ by a factor of 10. Regardless of which
assumption is used, however, as long as it is employed consistently for all
experimental systems, the degradation rates (indicated by the slope of the
linearized, first-order decay) were dramatically greater for the estuarine
systems than for the agricultural soils.


                                    VI-14

-------
                             SOILS
C/)
UJ
3 100
Q
05
UJ 8C
or
60
UJ
0 40
H 20
tr o
a) SASSAFRAS, +02
. •• mm mm

-
• • •



;'•







mm
. •




_





(H
.
•
1







Tli ll ii
*- 7 14 2 28 45 59 80
X
Ul
<
0 '00
80
u.
O
60
UJ 40
O
tr
UJ 20
a.
«
c) MATTAPEX, + 02
-
-
••
\



7
^
5

LXX\\X\\XXXXXX

71


•^



^



n




-
b
I
I
) J

JA
'/
I
#
\
SS

>A
•
p
'i
\
\\
F

R
*
1
\
AS

» i
\
I
\

0
p
:
•
7
2

•
•
• •


                                        14  21  28  45  59  80
       14  21   28  45  59  80     7   14


          DAYS   FROM  START OF  EX
 21  28  45  59   80


PERIMENT
-
I

d
) N
/
11
•
1
/
iT
1 •- •.•:.k\X\XXXX\\X\l
AP
E
I
X
i
^
°2
KXXXXXXXXXXXM 1

            EZ3 Atraz ine   Paren t

            EH3 Hydroxyotrazine

            • Dealkylated  Atrazine

Figure 5. The percent composition of 14C-atrazine metabolites in the total extracted
        residues from the agricultural soils.

                              VI-15

-------
TABLE 3.  Atrazine degradation in the aerobic systems  with estuarine sediment/
          water and with agricultural  soils under two  different  assumptions
          concerning the composition of the non-extractable fraction.
                               ATRAZINE REMAINING IN TOTAL SYSTEMt
                           Non-Extractablett
                           As Non-Atrazine
                  Kon^ExtractabTettt
                   As Al1 Compounds
Esttarine
    Choptank
        Day  1
            14
            45
            80

    Tangier
        Day  1
            14
            45
            80
86
49
 9
 3
71
 5
 2
 1
98
75
25
 7
90
13
10
 2
Soil
    Mattapex
        Day  1
            14
            45
            80
                                 88
                                 73
                                 56
                                 53
85
66
31
19
                         93
                         86
                         75
                         79
96
96
82
86
t   Includes both water and sediment for estuarine systems.
ft  Non-extractable radioactivity considered to be compounds other than
         atrazine.
ttt Non-extractable radioactivity considered to contain the  same  per-
         centage of atrazine as the extractable portion.
                                      VI-16
                                                                                         f

-------
i
                     The half-life of atrazine in agricultural soils seems to be quite vari-
                able witn a range of 37 d to as long as 35 y (5, 1, respectively). Less re-
                search has been conducted on atrazine degradation in aquatic environments,
                and therefore, few comparative values are available. However, Armstrong et
                al. (1) found an atrazine half-life of 145 d in a Wisconsin lake sediment and
                Ballantine et al. (2) reported a half-life of approximately 30 d for Chesa-
                peake Bay sediment.  A general pattern emerges from these data, where degrada-
                tion rates tend to be substantially slower in agricultural soils (where the
                half-life generally ranges from 1-6 mo) than in estuarine systems (1-4 wk).

                     Rapid removal and degradation of atrazine to hydroxyatrazine in estua-
                rine water and sediment as demonstrated in the present study indicates a rela-
                tively reduced role of residual phytotoxicity in the estuary for the parent
                compound. However, the N-dealkylated metabolites (principally N-de-ethylated
                atrazine) may persist and accumulate in the estuary as has been observed in
                agricultural fields (5, 17, 22}. Although no studies have been conducted on
                the phytotoxicity of atrazine metabolites to aquatic plants, it is known that
                N-de-ethylated atrazine is toxic to agricultural crops such as oats while
                hydroxyatrazine is not (12). Therefore, the ecological impact of these atra-
                zine metabolites in aquatic systems should be examined further.
                                         14
                     The non-extractable   C atrazine activity most likely consists largely
                of further degraded daughter compounds, as well as conjugates with soil
                organic matter (20). This view is supported by the decrease in percentage of
                hydroxyatrazine in the Tangier system from day 14 (30 percent) to day 45 (16
                percent) and the concomittant increases in non-extractable activity (57 and
                75 percent, respectively). The Tangier system also exhibited more rapid rates
                of both atrazine removal from the water and parent molecule degradation (Fig.
                2). This suggests that the relative availability of daughter products for
                further degradation was greater earlier on in the experiment in the Tangier
                system than in the Choptank system. In the combined water/sediment Choptank
                system, the hydroxyatrazine percentage increased from day 14 to day 45 (35 to
                61  percent, respectively), reemphasizing the relatively slower degradation
                rates of that system.

                     The half-lives (TQ 5) of atrazine applied to a variety of soils reported
                in the literature are provided in Table 4 along with the estuarine and soil
                systems of the present study for comparative purposes. The problem of non-
                extractability of atrazine metabolites from soils creates a situation whereby
                many of the values presented in this table must be considered only approxi-
                mate. The atrazine half-lives in the soils of the present study are compar-
                able with the reported values from the literature. A general pattern emerges
                from these data, where degradation rates tend to be substantially slower in
                agricultural soils (where TQ 5 generally ranges from 1-6 months) than in estu-
                arine systems (where TQ 5 ranges from about 1-4 weeks).

                     In conclusion,  it appears that atrazine is generally more labile in the
                estuarine environment than on agricultural fields. As such, this trend tends
                to benefit all concerned, since 1t enhances weed control for the farmer,
                while diminishing the chances of effects to non-target species 1n the estu-
                ary. However, the N-dealkylated metabolites (principally N-de-ethylated atra-
                zine) may persist and accumulate 1n the estuary. Moreover, there is some
                reason to suspect that this compound may retain some phytotoxicity. For


                                                   VI-17

-------
example, Sirons et al. (26) found that N-de-ethylated atrazine was more per-
sistent than atrazine in soils, and because of its phytctoxic nature may ac-
count for the carry-over of phytotoxicity observed from one year to the next
in soils following atrazine applications. Dao et al.  (6) found the levels of
N-de-ethylated atrazine to remain almost constant in atrazine-treated soils
for two years following application. Rapid removal and degradation of atra-
zine to hydroxyatrazine in estuarine water and sediment as demonstrated in
the present study indicates a relatively reduced role of residual phc^o-
toxicity in the estuary for the parent compound. However, the relative per-
sistence of the dealkylated daughter products and their unknown phototoxicity
to estuarine plants should be examined further.
                                    VI-18

-------

                             LITERATURE CITED
 1.  Armstrong, D.E., C. Chesters and R.F. Harris. 1967. Atrazine hydrolysis
     in soil. Soil Sci. Soc. Am. Proc. 31:61-66.

 2.  Ballantine, L.G., L.C. Newby and B.J. Simcneaux. 1978. Fate of atrazine
     in a marine environment. Fourth International Congress of Pesticide Chem-
     istry, IUPAC, Zurich, Switzerland.

 3.  Cunningham, J.J. 1979. Responses of microcosm communities containing sub-
     merged aquatic vegetation to herbicide stress. Master's thesis, Univ. of
     Maryland. 63 pp.

 4.  Dao, T.H. and T.L. Lavy. 1978. Atrazine adsorption on soil  as influenced
     by temperature, moisture content, and electrolyte concentration. Weed
     Sci. 26:303-308.

 5.  Dao, T.H., T.L. Lavy and R.C. Sorensen. 1979. Atrazine degradation and
     residue distribution in soil. Soil Sci. Soc. Am. J. 43:1129-1134.

 6.  Eckert, R.E. 1979. Establishment of pine (Pinus spp.) transplants in
     perennial grass stands with atrazine. Weed Sci. 27:253-257.

 7.  Geller, A. 1980. Studies on the degradation of atrazine by bacterial com-
     munities enriched from various biotopes. Arch. Environ. Contam. Toxicol.
     9:289-305.

 8.  Goswarni, K.P. and R.E. Green. 1971. Microbial degradation of the herbi-
     cide atrazine and its 2-hydroxy analog in submerged soils.  Environ. S~i.
     and Tech. 5:426-429.

 9.  Hall, J.K., M. Pawlus and E.R. Higgins. 1972. Losses of atrazine 1n run-
     off water and soil sediment. J. Environ. Qual. 3:172-176.

10.  Hermann, W.D., J.C. Tournayre and H. Egli. 1979. Triazine herbicide
     residues in Central European Streams. J. Pest Monitoring. 13:128-131.

11.  Jones, T.W., J.C. Means, J.C. Stevenson, W.M. Kemp. 1981. Uptake and phy-
     totoxicity of atrazine in Potampgeton perfoliatus. In: Submerged aquatic
     vegetation in Chesapeake Bay: Its ecological role in~"bay ecosystems and
     factors leading to its decline." Kemp, W.M., J.C. Stevenson, W. Boynton,
     and J.C. Means, (Eds.) Submitted to Chesapeake Bay Program USEPA,
     Annapolis, Maryland, Ref. #UMCEES 81-28 HPEL.


                                   VI-19

-------
12.  Kaufman, D.D. and J. Blake. 1970. Degradation of atrazine by soil fungi.
     Soil Biol. Biochem. 2:73-80.

13.  Kaufman, D.D. and P.C. Kearney. 1970. Microbial  degradation of triazine
     herbicides. Residue Rev. 32:235-265.

14.  Kells, J.J., C.E. Rieck, R.L.  Blevins and W.M. Muir.  1980.  Atrazine dis-
     sipation as affected by surface pH and tillage.  Weed  Sci. 28:101-104.

15.  Lamoureau, G.L., I.E. Stafford, R.H. Shimabukuro and  R.G. Zaylski. 1973.
     Catabolism of the glutathione  conjugate of atrazine.  J.  Agr.  Food Chem.
     21:1020-1025.

16.  Means, J.C., T.W. Jones, T.S.  Pait,  R.O. Wijayaratne.  1981. Adsorption
     of atrazine on Chesapeake Bay sediments and selected  soils. In: Sub-
     merged aquatic vegetation in Chesapeake Bay:  Its ecological ro^e in bay
     ecosystems and factors leading to its decline."  Kemp,  W.M., J.C. Steven-
     son, W. Boynton, and J.C. Means (Eds.) Submitted to Chesapeake Bay Pro-
     gram USEPA, Annapolis, Maryland. Ref. #UMCEES 81-28 HPEL.

17.  Muir, D.C.G. and B.E. Baker. 1978. The disappearance  and movement of
     thres triazine herbicides and  several of their degradation  products in
     soil under field conditions. Weed Res. 18:111-120.

18.  Muir, D.C.G., J.Y. Yoo and B.E. Baker. 1978.  Residues  of atrazine and
     N-de-ethylated atrazine in water from five agricultural  watersheds in
     Quebec. Arch. Environ. Contam. Toxicol. 7:221-224.

19.  Obien, S.R. and R.E. Green. 1969. Degradation of atrazine in four
     Hawaiian soils. Weed Sci. 17:509-514.

20.  Roeth, F.W., Lavy, T.L., and O.C. Burnside. 1969. Atrazine  degradation
     in two roil profiles. Wesd Sci. 17:2u2-205.

21.  Russell, J.D., M. Cruz, J.L. White,  G.W. Bailey, W.R.  Payne,  Jr., J.D.
     Pope, Jr., and J.J. Teasley. 1968. Mode of chemical degradation of _s-
     triazine by moritimorillonite.  Science. 160:1340-1342.

22.  Sirons, G.R., R. Frank and T.  Sawyer. 1973. Residues  of  atrazine cyana-
     zine and their phytotoxic metabolites in a clay loam  soil.  J. Agric.
     Food Chem. 21:1016-1020.

23.  Skipper, H.D. and V.V. Volk. 1972. Biological and chemical  degradation
     of atrazine in soils. Weed Sci. 20:344-347.

24.  Stevenson, J.C. and N.M. Confer. 1978. Summary of available information
     on Chesapeake Bay submerged vegetation. Fish and Wildlife Service, Publ.
     OBS-78/66. U.S. Dept. of the Interior.

25.  Swanson, R.A. and G.R. Dutt. 1973. Chemical and physical processes that
     affect atrazine and distribution in sotfl systems. Soil Sci. Soc. Amer.
     Proc. 37:872-876.


                                   VI-20

-------
                   26.  U.S. Department of Commerce. 1974. Census of agriculture. GPO, Wash-
                        ington, D.C.

                   27.  White, J.L. 1975. Determination of susceptibility of ^-triazine herbi-
                        cides to protonation and hydrolysis by mineral surfaces. Arch. Environ.
                        Contam. Toxicol. 3:461-469.

                   28.  Wu, T.L. 1980. Dissipation of the herbicides atrazine and alachlor in a
                        Maryland corn field. J. Environ. Qual. 9:459-465.
                                                                                          14
                   29.  Wolf, D.C. and J.P. Martin. 1975. Microbial decomposition of ring-  C
                        atrazine, cyanuric acid, and 2-chloro-4,6-diamino^s-triazine. J.
                        Environ. Qual. 4:134-139.
•I
                                                      VI-21

-------
F  *
ri!
                                                   CHAPTER VII
                                     UPTAKE AND PHOTOSYNTHETIC INHIBITION OF
                             ATRAZINE AND  ITS DEGRADATION PRODUCTS ON FOUR SPECIES OF
                                            SUBMERGED VASCULAR PLANTS1
                                                   T. W. Jones2'3


                                                   L Wlnchell2
                                                  November 1982
                 University of Maryland Center for Environmental and Estuarine
                  Studies Ref. No. HPEL 83-251
                 2Horn Point Environmental Laboratories, P.O. Box 775
                  Cambridge, MD.   21613
                 3 Biology Department, Salisbury State College, Salisbury, MD.   21801

-------
                                ACKNOWLEDGEMENTS


     This work was supported by grants from CIBA-GEIGY Corporation and the
U.S. Environmental Protection Agency Chesapeake Bay Program.  We would like to
thank W. M. Kemp, H. M. LeBaron, J. C. Stevenson for review and comment on
various aspects of this work, P. Estes for laboratory assistance and K. Staver
for logistic support.
                                         Vll-ii

-------
                                    ABSTRACT
     The photosynthetic inhibitory effects of atrazlne and three of its major
metabolites, deethylatcd, deisopropylated, and hydroxyatrazine,  was determined
for four species of submerged macrophytes.  All  species showed a similar re-
sponse to varied dosages of the parent atrazlne compound with an average Ij
(concentration at which photosynthesis is Inhibited by 1%) for all  plants of
20 ppb and an average ISQ (concentration at which photosynthesis is inhibited
by 50%) for all plants of 95 ppb.  The degradation metabolites of atrazine
produced varying degrees of photosynthetic inhibition 1n the four species, but
generally the order of toxicity was deethylated > deisopropylated > hydroxy-
atrazine with hydroxyatrazine causing an apparent stimulation of photosynthesis
in several species.  Of four species tested Myriophyllurn spicatum was  the most
resistant to atrazine and its metabolites.  The magnitude of the actual  uptake
of the compounds (ug compound/gdw plant) by the plants correlated closely
with the photosynthetic inhibitory response, i.e., at the same concentration
the uptake of atrazine > deethylated > deisopropylated > hydroxyatrazine.
Considering that an extremely high environmental  concentration (0.5 ppm) of
deethylated atrazlne for an estuary only produced a photosynthetic inhibition
of from 20 to 40% in four major species of submerged macrophytes, 1t is con-
cluded that the degradation products of atrazine tested are not  of major sig-
nificance in the disappearance of the submerged grasses from the Chesapeake
Bay.
                                       VII-1

-------
                                  INTRODUCTION
     Atrazine (2-chloro-4-ethylamino-6-isopropylam1no-s_-triaz1ne),  a compound
used to control broadleaf weeds, has found widespread application for corn and
sorghum crops since its introduction in 1958.   Currently,  1t 1s the most  widely
used herbicide 1n the Chesapeake Bay watershed (Stevenson  and Confer 1978).
Recent studies concerning runoff from cropland in the Chesapeake Bay watershed
where atrazine has been applied have shown that a measurable percent (ca  It)
of the herbicide is ultimately transported from agricultural fields to the
aquatic environment (Wu 1980).  Thus, herbicides such as atrazine represent a
potential source of stress for the vegetation  in aquatic environments (e.g.,
estuaries) and these compounds have been suggested as a possible cause for the
recent decline of submerged aquatic vegetation (SAV) in Chesapeake  Bay (Correll
et al.  1978).  Dissolved concentrations of atrazine up to 17 ppb were found
by Correll and Wu (1981) and as high as 100 ppb by Kemp et al. (1982) in  agri-
cultural runoff, while concentrations of as low as 5-10 ppb have been shown to
significantly reduce photosynthesis in various SAV species (Kemp et al.  1982,
Forney and Davis 1981, Jones et al. 1982).

     A point of ecological concern In herbicide studies is the degradation
rate for these compounds 1n the environment.  Atrazine can persist  in soil
environments for a relatively long duration, with a half-life (time until 50%
of the parent compound remains) exceeding one  year under some conditions  (Arm-
strong et al. 1967).  Degradation has been found to occur  much more rapidly,
however, under the unique conditions found in  shallow estuarine environments,
where Jones et al. (1982) found the half-life  of atrazine  to be approximately
2 weeks.

     The degradation of atrazine in agricultural and estuarine environments
can occur by several pathways, Including (1) 2-hydroxylation forming 2-hydroxy-
4-etnylam1no-6-1sopropylam1no-sL-tr1az1ne(hydroxyatrazine), (2) N dealkylatlon
principally forming 2-chloro-4-am1no-6-1sopropylamino-striazine (deethylated
atrazine) or 2-chloro-4-ethylam1no-6-am1no-s_-tr1az1ne fdeisopropylated atrazine),
or 2-chloro-4,6-d1am1no-si-tr1az1ne, and (3) ring cleavage  releasing C0£ (Arm-
strong et al. 1967, Kaufman and Blake 1970, Skipper and Volk 1972,  Goswami and
Green 1971, Wolf and Martin 1975, and Jones et al. 1982).   The resulting  de-
al kylated degradation compounds show mixed phytotoxicities for terrestrial
plant species (Shimabukuro 1967, Kaufman and Blake 1970, Slrons et  al. 1973),
while hydroxyatrazine has been shown to have negligible phytotoxic  effects
(Gysin and Knusll 1960, Kaufman and Blake 1970).

     Studies on agricultural soils (Slrons et  al. 1973, Dao et al.  1979)  as
well as 1n estuarine environments (Jones et al. 1982) have Indicated that the
atrazine degradation metabolite, deethylated atrazine, may be more  persistent
than the parent compound and could account for carry-over toxlcity  In agricul-

                                     VH-2

-------
           tural soils from one year to the next.   However,  no data  exist on  the effects
           of this compound or other atrazlne metabolites  on any SAV species  nor on  the
           uptake of the compounds by these species.

                The purpose of this study was to Investigate the phytotoxidties of  atra-
           zlne and its 3 principle degradation  metabolites  (deethylated  atrazlne, deiso-
           propylated atrazine, and hydroxyatrazine),  and  the magnitude of their uptake
           by 4 important SAV species found in Chesapeake  Bay.


                                         METHODS & MATERIALS


           Phytotoxicity of Atrazina and its Metabolites

                Four submerged aquatic plant spec1«»s,  Potamogeton  perfoliatus, Ruppia
           maritima, Myripphyllum spicatum and Zant....heniapalus'tris were collected from
           the ChoptanV River estuary and its tributaries  (8-12 ppt  salinity).   Freshly
           collected plant shoots were rinsed in filtered  (1 urn)  Choptank  water  to
           remove epiphytes and sediment, and cut  to a specified length (10-15 cm de-
           pending on species).  Each plant was  placeJ in  an individual Wheaton  300  mi
           BOD bottle containing filtered Choptank writer and a predetermined  concentration
           of either atrazine or one of its 3 princi ..» degradation  products; deethylated
           atrazlne, delsopropylated atrazine, or  hydroxyatrazine.  Stock  solutions  of
           atrazine and it? metabolites were prepared  by dissolving  the compounds In
           methanol such that a small volume (1-3  mi)  could  be added to 1500 mi  of
           filtered water to achieve the desired concentrations and  this  water used  to
           fill the BOD bottles.

                The experimental  design consisted  of an array of 5 replicates each of
           atrazlne, deethylated, delsopropylated  and  hydroxyatrazine.  Five  replicate
           controls and dark bottles were Incubated simultaneously as well  as 5  methanol
           control bottles which contained a volume of methanol  equal to  that Introduced
           into the experimental  bottles with the  herbicide.  Duplicate 2 hour Incubations
           were carried out at herbicid" concentrations of 0.5 ppm,  1.0 ppm,  and 1.5 ppm
           1n water baths maintained at 23 ± 2°C.   Light was provided by  a fluorescent
           light bank yielding an Intensity of 180 uE1n/m2/sec at  the bottle  sur-
           faces.  Oxygen levels were measured with an Orbisphere  Laboratories Oxygen
           Indicator Model 2607.   Following the final  oxygen reading each plant  was  re-
           moved and dried to a constant weight at 60°C.   Oxygen production was  calculated
           as QZ PPro/S dry wt/hr.  The same experimental  design was  also  employed to test
           the effects of lower concentrations of atrazine only (10, 25,  50,  100, 250 ppb),
           in place of the atrazine metabolite compounds.   All other conditions  remained
           the same.
           Uptake of 14C Atrazlne Metabolites

                Plants of the same 4 species as above were collected 1n the previously
           described manner and placed in separate 2 * erhlenmyer flasks containing  1 t  of
           filtered river water and 0.5 ppm of either 1(*C ring-labelled atrazlne  (spe-
           cific activity (s.a.) 50 ud/mg), deethylated atrazlne (s.a.   21.3  uci/mg),

                                                VII-3
I

-------
deisopropylated atrazine (s.a. 13.7 wd/mg), or hydroxyatrazine (s.a.  33

     Incubations were conducted at 28°C 1n the light.  After 2 hours the
plants were removed, rinsed in filtered Choptank water, dried at 60°C  and
ground with a mortar and pestle.  Duplicate samples were processed basically
according to the method of Beer et al. (1982).  A 15 mg subsample was  taken
from each ground plant sample and placed in a vial  to which 0.1 ml water and
1 mi Beck,, "n BTS-450 tissue solubillzer was added.   The sample was digested
at 60°C for 24 hours, after which time 1 mi was transferred to a vial  con-
taining 10m* Beckman Ready-Solv NA pre-mixed liquid scintillation cocktail.
All samples were counted on a Packard Tri-Carb 460C liquid scintillation
counter.  The external standard ratios method was used to determine counting
efficiency.
                             RESULTS AND DISCUSSION
Atrazine Toxcity
     All four species of SAV responded in a similar manner to the dosage levels
(10-250 ppb) of atrazine tested, and analysis of covariance revealed no signif-
icant differences (P<.01) in dose-response pattern among species.  A repre-
sentative log plot is given in Figure 1, where the percent photosynthetic
inhibition caused by increasing atrazine concentrations is shown for Ruppia
maritima.  Although some photosynthetic inhibition was evident in Zannicnellia
and Myriophyllum at 25 ppb atrazine, significant effects did not appear until
50 ppb for all four species (Table 1).  Values of Ij (concentration at  which
photosynthesis is inhibited by IX) and ISQ (concentration at which photosyn-
thesis is inhibited by 50%) calculated from the linear regression equations
for the four species are presented 1n Table 2.  The species were very similar
in their Ij values (~ 20 ppb) indicating that they all have approximately
the same threshold sensitivity to atrazine.  However, the 150 values were
slightly more variable among the four species with Potamogeton being the most
sensitive (77 ppb) and Myriophyllum being the most resistant (104 ppb).

     These va:ues are in close agreement with the findings of two other studies
working with Potamogeton and Myriophyllum where oxygen or- the ^C-technique
was used to assess short-term (2-4 h) photosynthesis (K'-np et al. 1982, Jones
et al. 1982).  Kemp et al. (1982) in their utrazine study en Potamogeton also
monitored bionwss and found an I$Q of 130 ppb.  Although Forney and DavTs
(1981) using stem length or dry weight as the plant response to atrazine found
simillar Ij values as in our study, they found quite different 159 values.
For Potamogeton they reported an IJQ for length of 474 ppb and for Myriophyllum,
1104 ppb.  The variations in results of the effect of atrazine on such  growth
responses as length and dry weight in aquatic macrophytes 1s not surprising
when considering the difficulty 1n maintaining these plants under laboratory
conditions for extended periods of time (e.g., Correll et al. 1978).  It seems
that the most consistent results have been obtained using short-term experiments
monitoring photosynthetic parameters as measures of toxicity.  However, these
results may not be the most revealing as to the total biological effect of
atrazine on the plants populations in their environment, and hence there is  a

                                     VI1-4

-------
•t!
               (V.) NOI1I8IHNI OIJ.BHlNASO.LOHd
                                                            <~ ID
                                                            
-------
TABLE 1.  Apparent oxygen production (mg 02 gdw^h"1)  of plants
          exposed to atrazine.*
Plant Species
Atrazine
Treatment

Control
10 ppb
25 ppb
50 ppb
100 ppb
250 ppb
Zannichellia
palustris

6.46a
6.55a
5.97a
4.84b
2.53C
1.78C
Ruppia
maritima

8.93a
8.75a
8.79a
6.99b
5.61b
2.96C
Myriophyllum
spicatum

15.44a
15.43a
15.02a
li.88b
9.56b
4.44C
Potamoa.eton
perfoliatus

18.12a
17.09a
18.42a
13.01b
8.78C
3.19d
*6iven are mean values for 10 replicates followed by a letter designating
 statistically different groups for a given species (p<0.05) using  the
 Student-Neuman-Keuls test.
                                     VI1-6

-------
«r
             TABLE 2.  Calculated values of Ij and I$Q (ppb) as derived from linear
                       regression analysis of % photosynthetic inhibition versus log of
                       atrazine concentration.*
Plant
Species
Potamogeton perfoliatus
Myriophyllum s pi cat urn
Zannichellia palustris
Ruppia maritima

II
20
20
17
20
150
77
104
91
102
Corn
Coef
.93
.95
.95
.93
             *Il and  150  are defined as the atrazine concentration at which 1% and 501
               photosynthetic inhibition are apparent, respectively, (c.a., Forney and Davis
               1980).
                                                   VII-7

-------
need for the longer duration experiments.

Atrazine Metabolite Toxicities

     The repsonses of the four species of SAV tested to three levels (0.5,
1.0, 1.5 ppm) of atrazine and its 3 metabolites are shown in Figure 2.  All
four plants had similar photosynchetic responses to the degradation products;
however, several significant differences between species is evident.  Myrio-
phyl1 urn was less affected by the degradation products relative to the other
species.  Also, increasing concentrations of deisopropylated and deethylated
atrazine did not produce concommitant increases in photosynthetic inhibition
in Myriophyllum, possibly indicating the approach of saturation levels of
these compounds in reducing photosynthesis in this species.  These results
reinforce the observation that Myrlophyllum is relatively resistant to atrazine
as presented above (also see Kemp et al. 1982) and may indicate that this
species is able to detoxify atrazine to some extent.  Also the four species
displayed varying responses to hydroxyatrazine 1n  a range from photosynthetic
inhibition at all 3 concentrations (Zannichellia) to a stimulation of photosyn-
thesis at all concentrations (Potamogeton).

     The actual oxygen production rates by the four species exposed to atrazine
and its metabolites are presented in Table 3, Indicating significant differ-
ences (P<.05) between the means.  Hydroxyatrszine had no significant effect on
any of the plants at any of the three concentrations tested.  Deisopropylated
atrazine exhibited mixed effects in the four species at concentrations of 0.5
and 1.0 ppm and only significantly depressed photosynthesis 1n all species at
the 1.5 ppm concentration.  Deethylated atrazine had a significant effect at
all concentrations 1n all plants as did the parent compound.

Atrazine Metabolite Uptake

     The magnitude of the uptake of atrazine and Its metabolites 1n the tissue
of the four species (F1g. 3) can be related very closely to the relative polari-
ties of the molecules.   Relative polarity decreases 1n the order; hydroxyatra-
zine > deisopropylated atrazine > deethylated > atrazine.  The large llpold
component of the chlorophyllous tissue in these plants would enhance atrazine
uptake and retard the movement of the more polar compounds especially hydroxy-
atrazine into the plant.  This mechanism of selective herbicide penetration
was proposed earlier by Shimabukuro and Swanson (1969) for atrazine, but, no
data were presented.  The data presented here may support such a mechanism.

     However, whe.i the magnitude of photosynthetic Inhibition (% of control)
Is compared to the uptake of each compound (ug/gdw plant) (Fig. 4) an
indication of a more complex situation emerges.  For example, the uptake of
deethylated atrazine by Zannlchellia was nearly as great as the uptake of
atrazine (10 ug/gdw vs 11 pg/gdw, respectively) and yet the photosynthetic
inhibition 1n response to this uptake was dramatically different.  Atrazine
reduced photosynthesis by 85% while deethylated atrazine created a reduction
of only 40%.  A similar situation occurred In Potamogetpn and to a lesser
extent Ruppia.  Only Myrlophyllum had a close relationship between uptake and
photosyntnetlc Inhibition.


                                     VII-8

-------
                        O   «  O  «
(V.) NOI1I8IHNI  Dll3HiNAS010Hd
                 VII-9

-------
TABLE 3.  Apparent oxygen production (mg Op gdw^h"1)  of plants  exposed
          to three levels of atrazine and its  metabolites.
                                             Plant  Species
Concen-
tration

0.5 ppm




1.0 ppm




1.5 ppm




Treat-
mentt

Control
OH
Dei so
Deethyl
Atrazine
Control
OH
Dei so
Deethyl
Atrazine
Control
OH
Dei so
Deethyl
Atrazine
Zannichellia
palustris

59.18a*
52.95a>b
48.64b
36.85°
12.23d
63.93a
63.72a
56.14a
36.96b
8.73C
61.06a
57.15a«b
45.68b
25.02C
7.47d
Ruppia
mantima

44.45a u
42.35a«b
37.82b
41.04b
10.67C
40.16a
39.36a
30.99b
26.92b
6.21°
37.97a u
34.84a»b
30.90b
20.97C
3.53d
Myriophyllum
splcatum

51.76a
50.29a
54.55a
41.04b
10.72°
72.92a
73.98a
64.88a
49.46b
11.71C
54.59a'b
62.50a
51.15b
37.03°
7.23d
Potamogeton
perfoliatus

54.93a
52.23a
48.92?
36.47b
8.26°
58.15a
61.80a
49.91b
31.79°
6.08d
68.87a
69.88?
48.08b
26.26°
5.83d
*G1ven are mean values for 10 replicates followed by a letter designating
 statistically different groups for a given species (p<0.05)  using the  Student-
 Neuman-Keuls test.                                   ~"
tTreatments are control, hydroxyatrazine (OH),  deisopropylated atrazine (Deiso),
 deethylated atrazine (Deethyl), and atrazine parent compound.
                                     VII-10

-------
UJ  O

Z  UJ


N  <

<  -1
        UJ
<  O   j  UI
>  ac.   r*  z
X  0.   £  —
oo*"
*  «»   r~.  15
o  _   uj  o:
>.  UJ   UJ  I-
x  o   °  <

D  E3 S 0
MYRIOPHYLLUM
RUPP
POTAMOGETON
CHELL
solution concentration of
Atrazine metabolite uptake by the four species at
for each metabolite (x ± S.D., n = 4).
3
ZANN

igure
      o
      CM
                10
                   6/6r/)


                       VII-ll

-------
-
                 (%)NOI1I9IHNI Dli3HiNASOJLOHd
                                 VII-12

-------
     The observations reported above may indicate not only a solubility factor
1n the mechanism of atrazine toxicity, but also possibly steric factors as
discussed elsewhere in detail for £-tr1azines (Ebert and Dumford 1976).  The
current understanding of the mechanism of action of atrazine is that 1t binds
to a specific protein in the thylakoid membrane (Pfister et al., 1980)  which
requires that the compounds possess the appropriate configuration.   The removal
of groups from the ^-triazine ring or hydro^ylation of atrazine to form degraded
metabolites such as those tested here would create steric changes.   Pfister et
al. (1979) has further proposed a resistance mechanism where plants modify the
herbicide binding site such that the compound cannot bind and thus  produces no
effect.  The percent photosynthetic inhibition per ug deethylated atrazine
taken up by the plant was only 1/2 that for the same amount of atrazine absorbed.
Therefore, the compound was present in the cells but less effectual  in  reducing
photosynthesis compared to atrazine.  However the location of the compound is
most important for it still could be that the compound is unable to penetrate
the cMoroplasts due to polarity, and thus it remains chiefly in the cytosol.

     It would appear that the atrazine degradation metabolites tested here are
not as effectual in reducing photosynthesis in SAV species as atrazine  is  in a
system such as the Chesapeake toy.  The high concentration (0.5-1.5 ppm) of
the degradation products tested were necessary to ellicit detectable responses
and are far above the concentrations that would be expected 1n the  environment.
                                     VI1-13

-------
                                               REFERENCES
              1.  Armstrong, O.E., C. Chesters, and R.F.  Harris.  1967.  Atrazine hydrolysis
                 in soil.  Soil Sci. Soc. Am. Proc. 31:61-66.

              2.  Beer, S., Stewart, A.J., and R.G. Wetzel.  1982.  Measuring chlorophyll
                 a and ^C-labeled f/hotosynthate in aquatic angiosperms by the use of a
                 tissue solubilizer.  Plant Physiol. 69:54-57.

              3.  Correll, D.L. and T.L.  Wu.  1982.  Atrazine Toxicity to Submersed Vascular
                 Plants in Simulated Estuarine Microcosms.  Aquatic Botany 14:151-158.

              4.  Correll, D.L, J.W. Pierce, and T.L. Wu.  1978.   Herbicides and Submerged
                 Plants in Chesapeake Bay.   IN:  Amer. Soc. Civil  Eng.  (ed.) Coastal
                 Zone.  78 pp.  858-877.

              5.  Dao, T.H., T.L.  Lavy, and R.C. Sorensen.  1979.   Atrazine Degradation
                 and  Residue Distribution in Soil.  Soil Sci. Soc. Am. J.  43:1129-1134.

              6.  Ebert, E., and S.W. Dumford.  1976.  Effects of Triazine Herbicides on
                 the  Physiology of Plants.  Residue Rev. 65:1-103.

              7.  Forney, D.R., and D.E.  Davis.  1981.  Effects of  low concentrations of
                 herbicides on submerged aquatic vegetation.  Weed Sci.  29:677-685.

              8.  Goswami, K.P., and R.E. Green.  1971.  Microbial  degradation of the herbi-
                 cide atrazine and its 2-hydroxy analog in submerged soils.  Environ. Sci.
                 Tech. 5:426-429.

              9.  Gysin, H., and E. Knusli.  I960.  Chemistry and Herbicidal Properties of
                 Triazine Derivatives.   Adv. Pest Control Res. 3:289-358.

             10.  Jones, T.W., W.M. Kemp, J.C. Stevenson, and J.C.  Means.  1982.  Degrada-
                 tion of atrazine in estuarine water/sediment systems and soils.  J.
                 Environ.  Qual.  11:632-638.

             11.  Jones, T.W., W.M. Kemp, P.S. Estes, J.C. Stevenson.  1982.  Atrazine
                 uptake phytotoxicity, release, and short-term recovery for the submerged
                 aquatic plant Potamoqeton perfoliatus.  Report to U.S.  E.P.A., Annapolis,
                 Maryland, Grant  No. R805932-01-1.

             12.  Kaufman, D.D., and J. Blake.  1970.  Degradation  of atrazine by soil
                 fungi.  Soil Biol. Biochem. 2:73-80.
                                                 VII-14
1

-------
13.   Kemp, W.M., J.J. Cunningham,  J.C. Stevenson,  J.C. Means, W.R. Boynton.
     1982.  Response of submerged  vascular plants  to herbicide stress in estua-
     rine microcosms.  Report to U.S. E.P.A., Annapolis, Maryland, Grant No.
     R805932-01-1.

14.   Pfister, K., K.E. Steinback,  G.  Gardner, and  C.J. Arntzen.  198C.  Identi-
     fication of the photosystem II herbicide binding protein.  Plant Physiol.
     65(6)'.Supplement, Abstr. No.  48.

15.   Pfister, K., S.R. Radosevich, and C.J.  Arntzen.  1979.   Modification of
     Herbicide Binding to Photosystem II in  Two Biotypes of  Senedo vulgaris
     L. Plant Physio. 64:995-999.

16.   Shimabukuro, R.H.  1967.  Significance  of atrazine dealkylatlon 1n root
     and shoot of pea plants.  J.  Agri. Food Chem. 15:557-562.

17.   Shimabukuro, R.H., and H,R. Swanson.  1969.  Atrazine metabolism, selectiv-
     ity, and mode of action.  J.  Agr. Food  Chem.  17:199-205.

18.   Sirons, G.R., R. Frank, and T. Sawyer.   1973.  Residues of atrazine cyana-
     zine and their phytotoxic metabolites in a clay loam soil.  J. Agric. Food
     Chem. 21:1016-1020.

19.   Skipper, H.D., and V.V. Volk.  1972.  Biological and chemical degrada-
     tion of atrazine in soils.  Weed Sci. 20:344-347.

20.   Stevenson, J.C., and N.M. Confer.  1978.  Summary of available Infor-
     mation on Chesapeake Bay submerged vegetation.  Fish and Wildlife Service
     Pub. no. OBS-78/66, U.S. Dep. of the Interior.  U.S. Government Printing
     Office, Washington, D.C.

21.   Wolf, D.C., and J.P. Martin.   1975.  Microblal decomposition of ring-
     ll*C atrazine, cyanuric add,  and 2-chloro-4,6-diam1no-s-triazine.  J.
     Environ. Qual. 4:134-139.

22.   Wu, T.L.  1980.  Dissipation of the herbicides atrazine and alachlor in a
     Maryland corn field.  J. Environ. Qual. 9:459-465.
                                     VII-15

-------
                                                                      f I
                                  CHAPTER VIII
                         EFFECTS OF NUTRIENT ENRICWENT
                   IN EXPERIMENTAL ESTUARINE PONDS CONTAINING
                     SUBMERGED VASCULAR PLANT COMMUNITIES*
                               Robert R. Twllleyt
                                W. Michael  Kempt
                               Kenneth W. Stavert
                               Walter R. Boyntonf
                              J. Court Stevensont
                                  August 1982
Contribution No. 1431 Center for Environmental  and Estuarlne
 Studies, University of Maryland.

tHorn Point Environmental Laboratories, P.O.  Box 775,
 Cambridge, MD.   21613

^Chesapeake Biological Laboratory, P.O. Box 38,
 Solomons, MD.   20638
                                    VIII-1

-------
                               TABLE OF CONTENTS


                                                                          Page

LIST OF FIGURES	       VIIMv
INTRODUCTION	       VII1-1
MATERIALS AND METHODS	       VIII-2

     Ponds Design	       VIII-2
     Experimental Design	       VIII-2
     Bi oma s s	       V111- 4
          Submerged Vascular Plants	       VIII-4
          Epiphytic Material	       VIII-4
          PI anktoni c Materi a 1	       V111-5
     Productivity and Respiration	       VIII-5
          Vascular Plants and Epiphytes	       VIII-5
          Community Metabolism	       VIII-6
     Nutrient Dynamics	       VIII-6
          Denitrlfication	       VIII-6
          Nutrient Sampling	       VIII-6
          Chemical and Physical  Analyses	       VIII-7
     Light Measurements	       VIII-7

RESULTS	       VIII-8

     Nutrient Cycling	       VIII-8
          Water Column Concentrations	       VIII-8
          Plant Composition	       VIII-8
          Sediment Nutrients	       VIII-13
          Denitrification Rates	       VIII-13
     Light Attenuation	       VI11-17
          Epiphytic Solids	       VII1-17
          Water Column Attenuation	       VIII-17
     Plant Biomass	       VIII-17
          Submerged Vascular Plants	       VIII-17
          Epiphytic Biomass	       VII1-21
     Plant and Epiphyte Productivity	       VIII-24
          Plant Production and Respiration	       VIII-24
          Plant/Epiphyte Relations	       VI11-28
     Community Photosynthesis and Respiration	       VI11-30

DISCUSSION	       VIII-30

     Nutrient Uptake from the Water Column	       VIII-30
     Nitrogen Uptake and Storage by SAV	       VIII-34
     Sediment Nutrient Dynamics	       VIII-39

                                    VIII-11

-------
     Algal Response to Nutrient Enrichment	       VIII-39
     PAR Attenuation by Algal  Biomass	       VIII-42
     Epiphyte Effects on SAV	       VIII-45
     SAV Biomass Response to Nutrient Enrichment	       VIII-48
     Response of Community Metabolism to Fertilization	       VIII-50
     SAV Strategy for Growth and Production	       VIII-50

SUMMARY	       V111- 52
LITERATURE CITED	       VII1-54
APPENDICES	       VIII-61
                                    VIII-111

-------
t!
                                           LIST OF FIGURES
Figure  1.




Figure  2.




Figure  3.




Figure  4.




Figure  5.



Figure  6.




Figure  7.
              Figure  8.
                          Diagram of pond mesocosms at Horn Point Environmental
                          Laboratories.  Broken lines represent the flow of
                          estuarine water to and from the Choptank River
                          estuary ............................... .. ..... .. .....
                          Time series plot of inorganic nitrogen
                          and N(E) and phosphorus concentrations
                          in the pond water at four nutrient treatments
                          beginning on 16 July, 1981 .......... .........
                          Concentrations (uM) of total  nitrogen,  nitrate,
                          and ammonium in pond water at four nutrient treat-
                          ments on the third day following nutrient  applications
                          during the summer, 1981	,
                          Tissue nitrogen (N) concentrations (% dry wt.) of
                          Potamogeton perfoliatus subjected to four nutrient
                          treatments at 26 and 60 days following the initial
                          t rea tme nt	

                          Changes in the carbonrnitrogen ratios (atom weight)
                          in tissues of Potamogeton perfoliatus subjected
                          to four nutrient treatments during the summer, 1981.
                          Concentrations (pM) of nitrate plus nitrite
                          and ammonium in the pore waters of sediments with and
                          without vegetation in a low and high nutreint treated
                          pond on 15 July 1981	
                                                                                        Page
                                                              VIII-3
                                                              VIII-9
                                                              VIII-10
                                                              VIII-11
                                                              VIII-12
                                                              VIII-14
                          Concentrations (yM) of nitrous oxide (N20) in
                          the headspace at  various intervals  of a 24 h incuba-
                          tion of sediment  cores froma low and high nutrient
                          treatment.  Also  included are nitrate (NOo)
                          concentrations in water overlying the sediment  cores at
                          the beginning and end of the experiment	
Denitrification rates (ymol •nT2«h~1)
in the vegetated and bare sediments of a  low and
high nutrient treated pond................	.,
                                                              VIII-15
                                                                                        VII1-16
                                                  VHI-iv

-------
                                                          If-
Figure  9.


Figure 10.



Figure 11.



Figure 12.
Attenuation coefficients (m'1) of the water
column at four nutrient treatments	
Concentrations of chlorophyll _a (pg/*) and
total sestion (mg/je) in the pond waters at four
nutrient treatments	
Aboveground bioma^s (gdw/m2) of submersed
macrophytes in ponds subjected to four nutrient
treatments.  Fertilization began on 18 June, 1981.
                                  VIII-18
                                  VIII-19
                                  VI11-20
a) Changes in the total biomass of submersed macro-
phytes (kg dry wt.) in the ponds subjected to four
nutrient treatments,  b) Average daily changes in the
biomass of submersed macrophytes (gdw«m"2»d"1)
in ponds at four nutrient treatments	
Figure 1.3.
Figure 14.
Figure 15.
Epiphyte biomass (gdw/gdw(SAV)) on Ruppia maritima
and Potamogeton perfoliatus in ponds subjected to
four nutrient treatments.Fertilization began after
the June measurements..	
Epiphyte biomass (chl-a/gdw (SAV) on Ruppia maritima
and Potamogeton perfolTatus in ponds subjected to four
            nutrient treatments.
            June measurements...,
                      Fertilization began after the
                                                                          VII1-22
                                                                          VIII-23
                                                              VI11-25
Photosynthetic active radiation (pEin»m"2«
s'1) at the water's surface and at 0.5 m depth from
mi dooming to midafternoon along with measurements of
apparent production (Pa) and net respiration (R,,) for
Ruppia maritima and Potamogeton perfoliatus	
                                                                          VIII-26
Figure 16.
Net photosynthesis (mg C-gdw^h"1} along
the length of Potamogeton perfoliatus and Ruppia
maritima cleared of epiphytic material.   Included is
the amount of photosynthetic active radiation
(wEin'm-Z'S'1) from water's surface
to 0.5 m depth.....	
Figure 17.
Figure 18.
Apparent production and net respiration [mg

                     ?
nutrient treated pond
                        of Potamogeton perfoliatus  and
            associated epiphytes from a
                   (g dry wt.

low nutrient and a hiyh
                                                                          VIII-27
                                                                          VIII- 29
Net photosynthesis [mg C (g dry wt.
of the tip (0-15 cm) and basal (15-50 cm) sections
of Potamogeton perfoliatus and Ruppia maritima and
                           £
                          >rf
associated epiphytes	   V111-31
Figure 19.  Photosynthesis and respiration (g O.-m"3^"1)
            of the pond communities at four nutrient  levels
            during summer of 1981	   VI11-32
                                  VIII-v

-------
r
              Figure 20.
              Figure 21.
              Figure 22.
              Figure 23.
              Figure 24.
              Figure 25.
Changes in net production (g O'lrr^d'1) of
the pond communities at four nutrient treatments
during the summer of 1981.  Fertilization began
after the June measurements	
Uptake rate for NH^ and NUj from pond waters with
submersed macrophyte communities at various concen-
trations of nitrogen	
The relationship between tissue nitrogen content (%)
and biomass production of submersed macrophytes
The response of photosyntnetic active radiation
{attenuation coefficient, K), plankton chlorophyll £
(um/i) and total sestion (mg/i) of the
water in pond mesocosms subjected to four nutrient
treatments	

Correlations among the attenuation of photosynthetic
active radiation of the water column in ponds treated
with nutrients and A) total seston concentration
(mg/4), and B) total nitrogen concentration (um)	
                                                                                        VIII-33
                                                                                        VI1I-35
                                                                                        VII1-37
                                                                                        VI11-41
                                                                                        VI11-43
Correlations between A) attenuation coefficients
for photosynthetic active radiation and plankton
chlorophyll a, and B) total nitrogen concentrations
(um) and plaTikton chlorophyll £ 1n water column
of pr 4s treated with nutrients...	
                                                                                        VI11-44
              Figure 26.  Effects of epiphyte biomass [gdw/gdw(SAV)] on the
                          normal apparent production of Potamogeton
                          perfoliatus (% of production of plants cleared of
                          epiphytic material) based on experiments using 1(*C-
                          incorporation and dissolved oxygen evolution to measure
                          pi ant metabol1sm	
              Figure 27.
              Figure 28.
              Figure 29.
Effects of epiphyte biomass on Potamogeton perfoliatus
expressed as area (mg/cm2) and weight
[gdw/gdw(SAV)]	,
The response of submersed aquatic vegetation in
experimental ponds to various loading rates of
nitrogen and phosphorus	
The apparent production to nighttime respiration ratio
of ponds at four nutrient treatments	
VII1-46



VI11-47



VI11-49


VIII-51
                                                  VlII-vl

-------
                                   INTRODUCTION


      Over the past two decades there has been a  marked decline 1n abundance
 and distribution of submerged vascular plants in Chesapeake Bay (Ba>ley
 et al. 1978; Stevenson and Confer 1978; Orth and Moore 1981).  While densities
 of Zostera marina have been reduced in the higher salinity reaches of this
 estuary,  the most dramatic losses have occurred  in the upper and middle bay.
 Here, numerous plant species have been involved, such as Potamogeton perf ol -
 iatus, j>. pectinatus, Ruppia maritima, Zannichellia palustra,  Elodea
 canadensis and Valisnqria amencana. A number of factors have  been suggested
 as contributing to this decline,  including the increasing input of nutrient
 wastes (Boynton et al. 1981).  Nutrient enrichment is well  documented for
 several regions of Chesapeake Bay", as is the Increase in" phytoptankton chlor-
 opfcyH _a  -(Hfhursky and Boynton 1979; Heinle et al. 1981}. Prior to TRFgen-
 eral  demise of submerged aquatic  vegetation (SAV) in the Bay,  there was a
 proliferation of the exotic species Myriophyllum spicatum throughout the
 upper bay (Bayley et al . 1978). This is similar  to the pattern reported by
 Lind and  Cottam (1969), where Myriophyllum invasion of Lake Mendota occurred
 in response to eutrophi cation.

      Various investigators have postulated thac  observed losses of SAV in
 other freshwater systems were attributable to eutrophi cation,  Including Loch
 Leven, Scotland (Jupp and Spence  1976), Lake Erie, Ohio (Stuckey 1978), White-
 water Lake, Ontario (Dale and Miller 1978), and  Norfolk Broads, England
 (Phillips et al. 1978). Others have suggested that nutrient
 enrichment may be one of several  factors contributing to dimunition of sea-
 grasses in such coastal ecosystems, as the Dutch Waddenzee (Den Hartog and
 Polderman 1975), the French Medlterranian (Peres and Picard 1975), and Cockburn
 Sound, Australia (Cambridge 1979). The hypothesized mechanisms whereby nutrient
 additions could lead to reductions tn SAV involve the prolmptTbn of -a4-gaT .....
 growtfi, either phytoplanktonic (Jupp and Spence  1977) or ¥pTpfiy^Tc1pPln1ll1ps
 et al. 1978), and a resuTttrrg reduction 1n light available to  the vascular
 plants (Sand-Jensen 1977; Penhale 1977). Additional stress mechanisms jaybe
 ivolved  such as: inhibition of molecular
"Jensen 1977); enhanced deposition of sediments on SAV leaves (Menzie 1979);
 and Incresed growth of colonial animal  fouling on SAV leaves stimulated  by
 elevated availability of phytoplankton  (Staver et al.  1980).

      Much of the Information regarding  nutrient effects on algal -SAV relation-
 ships has been Inferred from indirect correlative evidence along natural  nutri-
 ent gradient (e.g.  Cattaneo and Kalff 1980;  Sand-Jensen and S^ndergaard
 1981). However,  direct experimental  approaches have been applied for fresh-
 water ecosystems utilizing pond or pool  microcosms (Mulligan and Baranowskl
 1969; Ryan et al. 1972; Moss 1976; Mulligan  et al. 1976; Phillips et al.
 1978). In fact,  early experiments with  fish  ponds demonstrated that  fertili-
 zation offered an effective means of controlling aquatic weeds (e.g., Smith

                                      VIII-1

-------
and Swingle 1941). In a recent paper Harlln and Thome-Miller (1981) reported
the results of semi control led experiments where nutrients were added to
seagrass beds 1n situ. While this study provided some provocative results,
rates of nutrient addlton could be determined only qualitatively. With this
exception, there have been no direct experiments Investigating nutrient enrich
ment Affects on SAV communities for marine or estuarine systems.

     In this paper we report the results of 2. study where we used eight repli-
cate experimental estuarine ponds stocked with mixed SAV species and flushed
with Chesapeake Bay waters. The purpose of this study was to Investigate the
effects of nutrient enrichment on: (1) abundance of phytoplankton and epiflora:
(2) community production and nutrient cycling; (3) Interactions between algae
and SAV (especially relative to light availability); and (4) SAV growth, abund
ance and nutrient status.
                          MATERIALS AND METHODS


                               Ponds Design

     The eight rectangular ponds were constructed at Horn Point Environmental
Laboratories during the spring and summer of 1979 (F1g. 1).  The dimensions  of
each pond are 27 m by 13 m at water surface, with sides sloped (1:1)  to a
nominal depth of about 1.2 m. Water from the Choptank River  1s pumped to a
centrally located reservoir, from which water flows Into each pond.   Estuarine
water entered each pond through a control valve and exited through a  stand-
pipe, so that flushing rate and water level could be controlled.  The ponds
have three piers each to facilitate sampling.  Two short (3m) piers  are lo-
cated at the Inflow and outflow ends of the ponds, and a longer pier  1s located
midway along one side, extending 9 m Into the pond.

The ponds were stocked with submersed aquatic vegetation during the summer  of
1979 from coves located along the Choptank River estuary. Cores (15 cm dia)
containing plant shoots, roots and sediments were transported from source
areas to the ponds, where they were transplanted in a grid with 1 m distance
between adjacent plugs. Chly two diagonally separated quarters of each pond
were seeded.  The dominant transplanted SAV species were Potamogeton  perfoliatus
and Ruppia maritime. By the summer of 1980, about 50-80% of  the sediment surface
was covered by SAV, and several other vascular plant species were observed  at
lower densities, including ZannichelHa palustrus, Myriophyllum spicatum. and
El odea canadensls.  This experiment was initiated in the early summer of 1981,
and represented the first manipulative treatment of these transplanted grass
beds.


                           Experimental Design

     Three different treatments of dissolved nutrients were  applied to duplicate
ponds, and two ponds served as controls. Commercial fertilizers 1n the form of
ammonium sulphate, potassium nitrate, and di -ammonium phosphate were  dissolved
in distilled water and applied to the ponds with a sprayer.   The nitrogen (N)
application rates were 1.68, 0.84, and 0.42 g/m2 for the high, medium, and

                                     VIII-2

-------
                                     To  Discharge
  From
Estuary
               ~-Hnf low  System

                   ^Discharge System

 Figure  1.  Diagram of pond mesocosms at Horn Point Environmental Labora-
           tories.  Broken lines represent the flow of estuarine water to
           and from theChoptank River estuary.
                            VIII-3

-------
low dose ponds, Respectively, and contained equal  amounts of nitrate (NOj)
and ammonium (NH.).  Concentrations of inorganic N immediately  after dosing
were 120, 60, and 30 vM in treated ponds, respectively.   Corresponding
phosphorus (P) loading rates were 0.37, 0.19,  and  0.09 g/mz which yielded
initial concentrations of 12, 6 and 3 uM in treated ponds.   The ponds were
treated on six occasions at 10 d intervals beginning 19  June 1981 and ending
on 4 August 1981.  The ponds were held in batch mode for 8-9 d  following  each
application, followed by 1-2 d of flushing with ambient  estuarlne water to
effect a complete turnover of water.


                                 Biomass

Submerged Vascular Plants

     Submersed aquatic vegetation was sampled  in each pond  twice before fer-
tilization began; twice during the treatment period and  once during  post  treat-
ment in November. A stratified random sampling procedure was used to select
six 0.1 m2 vegetated plots in each pond. Aboveground (leaves and stems) and
belowground (roots and rhizomes) biomass was harvested on 17 June whereas only
shoot? were sampled on the other occasions. Samples were rinsed with estuarlne
water, sorted by sped as, and dried in a forced draft oven  at 60°C to constant
weight.  Dry weights were determined to the nearest 0.01 g.  The dried biomass
was ground through a 40 mesh screen using a Wiley  Mill and  stored in air  tight
plastic bags. Aerial photography at 150 m was  used to determine the  percent
cover of SAV 1n each pond on each sampling date.  Kodak  Ektachrome ASA 400
daylight film was used to produce color slides from which the cover  of SAV  was
traced and area determined.

     The chlorophyll content of Pptamogeton was determined  from 3 mm diameter
disks sampled from leaves along the terminal 10 cm section  of the plant.
These samples were ground in a 10:1 dimethylsulfoxide (DMSO): acetone solution
(Wscox and  Israelstam 1979) and chlorophyll absorption  was measured on a Gary
219 spectrophotometer.


Epiphytic Material

     Epiphyte biomass was determined each time SAV biomass  was  measured except
for the first pretreatment and the post-treatment  occasions.  Individual
plants were sampled 1n triplicate and placed 1n 1  t plastic bottles.  Epi-
phytes were removed by agitating the bottles for several minutes. This water
sample was stirred 1n a beaker while three ten ml  aliquots  were removed and
filtered through prewelghed GFC glass fiber filt.ars. These three subsamples
were duplicated for analysis of chlorophyll £ and total  suspended solids.
Chlorophyll a filters were wrapped 1n aluminum foil and  frozen, and  the others
were dried for two days at 60°C. Background total  suspended iolld concentra-
tions were determined on pond water alone.  The plant from which epiphytes
were removed was dried 1n a plant press.  After determining dry weights,  these
samples were analyzed for leaf area with a LICOR Model 3100 area r*ter.

-------
            Planktonlc Material

                 Plankton pigments were measured by filtering 100-200 ml  of pond water
            through 0.45 y GFC filters.  Chlorophyll-a was extracted for  24 h in 10 ml  of
            a 1:1 solution of DMSO and acetone (Shoar~and Lium 1976) and  flourescence
            was measured using a Turner Model 111 flourometer.


                                   Productivity and Respiration

            Vascular Plants and Epiphytes

                 Primary production and respiration were measured using both dissolved
            oxygen (DO) and 14C techniques. Total seston was determined again by filtra-
            tion of 50-200 ml  of ambient water through GFC filters (0.45  p). Filters
            were washed, dried and weighed according to the methods of Strickland and
            Parsons (1972). SAV and/or epiphytes were incubated in 300ml  biological
            oxygen demand (BOO) bottles filled with filtered (1.0 p) estuarine water,
            and changes in 00 over time were used to estimate apparent productivity.
            Bottles were incubated for short time Intervals (1-3 h) in either the ponds or
            in a 0.25 m deep incubator under ambient light and temperature with continu-
            ously circulated river water. Light attenuation, salinity, and temperature of
            the water column were also measured during each experiment.

                 The productivity of the top 15 cm of plants with epiphytes of various
            densities were compared to the basal sections of SAV with epiphytes. The con-
            tribution of epiphytes to DO changes within these bottles was calculated by
            extrapolating measured rates of just epiphyte production to results with epi-
            phytes plus SAV. The epiphyte biomass on the plants (g/gdw(SAV)) was measured
            as discussed above and used to convert epiphyte production data (mg02 gdw
            (epiphyte)'1) to epiphyte production associated with SAV.  DO changes in
            light bottles represented apparent net production and no corrections were
            made for oxygen storage since lacuna! capacity in these plants is minor (West-
            lake 1978). DO changes in opaque bottle incubations were used to provide an
            estimate of dark respiration.
                 Net carbon incorporation was also measured using uptake of **C bicarbon-
            ate.  Whole plants were carefully removed from sediments and placed with epi-
            phytes intact Into clear plexiglass tubes (7.5 cm dia, 50 cm long) with the
            roots of the plants attached to one end of the core.  These cores were sealed,
            and floated vertically in the control pond such that the tip of the shoots
            were 2-3 cm below water's surface and the remainder of the plant extended to
j            0.5 m depth (the length of the core).  Carrierfree 14C-bicarbonate was
            injected along the length of the core via to a specific activity of 5 pCi/i.
            Plants were incubated for 1-2 h during midday.

                 Potampgeton perfoliatus was studied at 2 epiphyte levels and Ruppia mari-
            tima at only 1 epiphyte level, and controls for both plants were nearly free
            of epiphytes. Following incubation, the control plants were sectioned Into
            10 cm lengths, and the epiphyte colonized plants were sectioned into two lengths
            which were either less than or greator than 15 cm from the tip of the plant.
            The epiphytes were removed with a stream of water and a rubber policeman, and
            the entire contents filtered onto GFC glass fiber filters.  Plants and filters

                                                 VIII-5

-------
were dried to constant weight at 60°C and digested in nitric acid according to
Lewis et al. (1982). Radioassay was performed with a Packard Tri-Carb 460C
liquid scintillation counter standardized with a quench curve.  The absolute
amount of CO^ uptake was determined from specific activity  and  alkalinity
measurements, which represents net productivity.


Community Metabolism

     Diel changes of DO in the ponds were used to estimate  community produc-
tion and respiration following the approach  of Odum and Hoskin  (1958).  Measure-
ments were made prior to and during fertilization.  DO and  temperature  measure-
ments were made at the surface and 0.5 m at  3 stations (the end of each sampling
pier) in each pond at dawn and at dusk. This sampling scheme was performed for
2 consecutive days 7 times during the summer. Changes in DO were corrected for
oxygen transfer across the air-water interface using a modification of  the
method of Copeland and Duffer (1967). Changes in DO from dawn to dusk and from
dusk to dawn were taken to represent net production and nighttime respiration,
respectively.


                            Nutrient Dynamics

Denitrification

     Duplicate intact sediment cores (7.b cm diameter x 20  cm depth)  were
obtained in vegetated and unvegetated areas  of a low nutrient and high  nutri-
ent dosed pond.  Sediment denitrification was measured in each  core following
the acetylene reduction procedure of Sorensen (1978). The overlying water in
each core was removed to give a headspace of 4 cm resulting in  a water  depth
of about 15 cm.  Cores from the low and high nutrient dosed ponds, were en-
riched with dissolved nitrate at concentrations of 15 uM N  and  60 pM N
above ambient levels, respectively.  The cores were sealed  and  the headspace
purged with a gas mixture (80% Ar, 18% 02, 2% CO?) to remove nitrogen while
providing an in situ oxygen tension.  Acetylene (Cg^) was  injected into the
core to achieve a partial pressure of 0.2 atm (Patriquin and Denike 1978), and
the headspace of the core was sampled at 0,  3, 6, 17 and 24 h for N20 evolution.

     At 24 h the headspace was again purged  with the gas mixture, amended with
acetylene and incubated for another 24 h period. At 0, 24 and 48 h the  over-
lying water was sampled for dissolved inorganic nutrients (NH^, NOo
and DIP) and DO. Cores were incubated in Percival  incubators at 30*C and a 16
h light, 8 h dark photoperiod. Headspace samples were stored in yacutainers
and later analyzed by a Packard 111 gas chromatograph equipped  with a  63Ni
electron capture detector. Standards were run using pure nitrous oxide  (Air
Products, Inc.) to relate peak heights to NgO concent,ation.


Nutrient Sampling

     Water. Water in each pond was sampled at a depth of about  0.5 m at the
3 stations (end of each pier) in each pond throughout the summer.  Unfil-
tered subsamples were stored in plastic bottles and frozen  for  subsequent

                                     VIII-6

-------
analysis of total N and P. Triplicate 150ml aliquots were filtered through
preweighed GFC glass fiber filter paper (1 u) immediately following collec-
tion for analysis of both chlorophyll a and total suspended solids.  Filters
for chlorophyll £ were wrapped in alumThum foil and frozen, while the other
filters were dried for several days at 60°C and weighed. The filtered water
samples were analyzed for NH., NO^, NOo and DIP. Total nitrogen
and phosphorus, and pH were measured on unfiltered samples.

     Sediment. During the 17 June and 13 August biomass samplings, 6 sediment
samples were taken in each pond with a modified plastic core (2.5 cm dia, 15
cm long).  These sediment cores were sectioned at 2 cm intervals to a depth of
10 cm. Fresh weights were measured, the samples were dried to constant weight
at 60°C and weighed again to obtain water content.  Each section was ground
with a mortar and pestle, stored in plastic-pak bags and analyzed for total
carbon, nitrogen and phosphorus.  Four intact sediment cores (7.5 cm diameter
x 30 cm length) were taken in a low and high nutrient dosed pond on 15 .July
1982. Two cores in each pond were from vegetated and unvegetated areas. Each
core was sectioned into 2 cm subsamples to a depth of 10 cm in a glove-box
under N£ atmosphere.  Triplicate subsamples from each section were combined
and extracted with deionized water. Following centrifugation, the supernatant
was filtered through GFC filters and analyzed for inorganic nutrients '""
NO^ plus N0£, DIP).


Chemical and Physical Analyses

     Inorganic nutrients were assayed on a Technicon Auto Analyzer II using
the following techniques: ammonium determinations followed Solorzano (1969);
nitrate was reduced with a cadmium column and nitrite was reacted with sul-
fanilamide and N-1-dihydrochlomide njpthylene liamine; phosphorus results
represent filterable phosphorus reactive witr. Ammonium molybdate in ascorbic
acid.  Total phophorus and nitrogen were digested with persulphate using the
method of Valderrama (1981) followed by inorganic phosphate and nitrate assay
described above.  pH was measured with a Beckman Model 4500 digital pH meter
with a calomel reference electrode.  An Orbisphere Model 2603 oxygen meter and
probe (with stirrer) were used to measure DO in open water and in BOD bottles.
Total carbon and nitrogen of sediment, plant and filter paper were assayed
with a Perkin Elmer 240B elemental analyzer.


                            Light Measurements

     Attenuation of photosynthetically active radiation (PAR) in each pond was
determined periodically throughout the summer with a LICOR 185A light meter
equipped with a LICCri 1925B underwater quantum sensor. Daily integral light
energy and PAR were also obtained during all productivity experiments using a
Weather Measure Corp. Model No. R401 pyronometer and the LICOR meter.

     Light attenuation due to epiphytes was determined using a modification
of the method of Borum and Wiurn-Andersen (1980). Here, the reduction in PAR
was measured as It passed through a finger bowl containing different densities
of epiphytes removed from Potampgeton perfollatus leaves. Total suspended
solid concentration of the epiphyte medium was used to determine the weight

                                     VIII-7

-------
of epiphytes per cm^ in the bowls. This value was related to percent of
original light to calculate light attenuation coefficients.
                                 RESULTS


                             Nutrient Cycling
Water Column Concentrations
     During the week following each nutrient dosing,  inorganic nutrient  con-
centrations underwent a rapid and steady decline in all  ponds  (Fig.  2).
Initial concentrations varied according to treatment, but within  3 d  NHj
and DIP concentrations in all ponds were dramatically reduced.  The exception
was NOo in the high treatment ponds which remained above 25 uM even
after 7 d, Whereas NO^ in the other ponds was less than  1.0 yM after  3 d.

     Concentrations of NO^, NH^, and DIP, as well  as  total  P and
total N were measured throughout the summer on the third day following
treatment (Appendix, Table A-l to A-5).  The trend apparent in Fig.  2, where
NH< and DIP were reduced to similar levels among treatments within 3  d,
was exhibited for each dosage except the last. Following the sixth dosage,
both NH« and NOo remained relatively elevated in the  high dose ponds
(Fig. 3). Although the inorganic nutrients were similar  among  treatments
after 3 d, concentrations of total N and P reflected  treatment levels through-
out the summer (Fig. 3). Over 90% of the total nitrogen  levels were organic
N in the control, low and medium treatment ponds.   Inorganic N typically
comprised about 25% of the total for high treatment systems, except in the
6th wk when NHj and NOg were nearly 50% of total N (Fig. 3).

Plant Composition

     The N content in Potamogeton perfoliatus under control conditions changed
little from mid-June to August (Fig. 4, Table B-l). However, nitrogen levels
of plants in the low and medium nutrient dosed ponds  increased gradually
during the summer by about 0.3% N for each treatment. The most dramatic
change in % N was observed in the SAV in the high nutrient treatment, where
% N during the course of the experiment increased from 1.10-1.99% N in June
to 3.07% N in August.

     These changes in % N (g dry wt.) for Potamogeton are also reflected in
the carbon to nitrogen atom ratios in plants among the treatments (Fig.  5,
Table B-2).  C:N ratios of plants increased in the control  ponds  from 16 to
19, yet the ratios decreased in the other treatments. The magnitude of this
decrease in C:N ratios among the treatments followed  the nutrient application
rate, with the greatest decrease observed in SAV from the high nutrient
treatment. C:N ratios of Potamogeton in these ponds were only  10:1 in August.
These changes in C:N ratio and % N demonstrate the capacity of SAV to take
up N luxuriously in response to increased nutrient loading.
                                    VIII-8

-------
    10
                                        PHOSPHORUS
                        246

                 DAYS   AFTER  TREATMENT

Figure  2.  Time series  plot of inorganic nitrogen (NH^ and NO^) and phosphorus
           concentrations  (uM) in the pond water at four nutrient treatments
           beginning on 16 July, 1981.
                                VI11-9

-------


150
100
-^ 50

-
O
— 150
01 100
UJ 5°
o
o
0 1 50
z
O50

h-
1 50

00
e r\
DU



-


-



r
1





f


mi
EZ

.';":"•/ .V

•'•:••':'•:



••.•.'•.•'.'•a





m

1 T<
} N
I A















)ta
trc
mm
•/ •"•"•,•*


'•:':•••"•


KSNSN
WV1

:•'••/:"•;
• *• "•" *
• •" ."."•
^§§
^

Or
ite
on i i















gan
i m
*«•;•*.
'. ! J'.".' •

|$$


•_-ff-V/
'.'•'•'."'


."•••'•
::::.::
.'::•'•
^

ic r
(


L

N



1






sli t r
:or

•*\"v *".""
-°'
• ••*•"" " *
1EC

A'-:'.>".v-:
;•'•'•.':•'::
••• •;•.* i .•
HI C


"•"••*!•'/
V "• ." I • ;
^

oge
^JT


N

)l U



JH






• n
RO




M

l"."-:-'»
C'vCCf


•'.'•
•'•'f
*. •
^


L

















'•;•••
'** *\****

• • *J» • •

iv;-""--*
*.""*•**""
.*i |*-"*
.• • / % i*j
..vc^S
:.".*/.:-*
• * • * *
^si
^
^^
%

















             6/6     19/6    30/6     16/7   28/7    4/8
Figure 3.  Concentrations (uM) of total  nitrogen, nitrate, and ammonium in
          pond water at four nutrient treatments on the third day following
          nutrient applications during the summer, 1981.

                            VIII-10

-------
*
^1.5
•o
e
*.
Z 1.0
LJ
O
O

-------
1!
                           i-
                           <  20
                           or
                           UJ
                           o
                           o
                           CC
15
                               10
                           o
                           CD
                           or
                           <
                           o
                                Control
                                           20       40      60

                                TIME SINCE  INITIAL TREATMENT^
                          Figure 5.  Chanjes in the carbon:nitrogen ratios (atom weight)
                                   in tissues of Potamogeton perfoliatus subjected to
                                   four nutrient treatments during the summer, 1981.
                                               VIII-12

-------
ILL
                Sediment Nutrients

                     Nitrate concentrations in the interstitial  pore waters of sediments  were
                similar between the low and high nutrient dosed  ponds in July  (Fig.  6).   Neither
                were there any distinct trend in NOo of the pore waters  in  vegetated compared
                to unvegetated areas in either pond. In contrast, NH4 concentrations of
                the pore waters in the high nutrient dosed pond  were consistently  higher  than
                those in the low nutrient pond (Fig. 6).  Concentrations ranged from 150  uM
                to 400 pM in the former pond compared to a range of 40 pM to 210 pM  in
                the lower nutrient treatment.

                     Ammonium concentrations were about 180 pM in the vegetated and  unvege-
                tated areas of the high nutrient pond in the upper 6 cm.  From 6 to  10 cm
                sediment depth, NH* 1n the unvegetated areas increased sharply to  about
                350 - 400 yM, whereas NHt levels in the vegetated cores  remained at
                concentrations less than 200 uM. In the low nutrient o^osed  pond, pore-water
                NH. in the vegetated areas was usually only 25%  of NH^ concentrations
                in unvegetated areas to a depth of 10 cm.  Concentrations of the vegetated
                areas did not exceed 90 pM, whereas concentrations were  generally  180 pM
                in the unvegetated areas (Fig. 6).


                Denitrification Rates

                     Nitrous oxide (N20) production in the low NOo dosed core  (10  pM)
                varied considerably during the first 24 h incubation period (Fig.  7). N20
                evolution was low during the first 3 h of incubation reflecting the  lag for
                acetylene to diffuse into the sediments.  From 3 to 6 h  a sharp pulse of  N20
                was produced in all cores. Following this pulse, N20 production rates de-
                clined, and differences were observed among cores with and  without SAV. The
                change in N20 production during the time course  can be related to  the depletion
                of NK concentrations during the experiment. The initial NO^ concentration
                was 10 pM in the low nutrient dosage which decreased to  about  1.1  pM at
                17 h. Since denltrification rates are dependent  on NK concentrations, a
                reduction in N20 production would be expected as NO^ concentrations  in
                the overlying water decreased.

                     The time series of N^O production in the high nutrient dosed  cores also
                showed a lag in rates during the first 3 h (Fig. 7).  No0 production was
                generally linear from 3 to 17 h although some decrease in rates was  observed
                from 17 to 24 h. The change 1n NOjj concentration was from 72 to 4  pM,
                and therefore lower N20 production rates at 17 h could be expected because
                of the much lower NO^ levels 1n the water column.

                     Denltrification rates were calculated from  changes  1n  N20 from  3 to  6 h
                for the vegetated and unvegetatad cores of both  nutrient levels (F1g. 8). The
                rates were significantly greater 1n high dosed cores, with  values  ranging from
                19 to 22 pmol• m"2«h~l 1n low dosed cores and from 44 to 46 pmol«
                m"2.h"l under high nutrient dosage (Fig. 8). There were  no  significant
                differences between the vegetated and unvegetated cores  for either nutrient
                treatment.
                                                    VIII-13

-------
               LOW  NUTRIENT
                                 HIGH   NUTRIENT
0-2
E
u 2-4
I
^ 4-6
0.

_ *
yv 6-fl


8-IO
1 w
m o
g
i?



T 1 Ij
Y/////S 	
.'.fr1!
1 1









J
t
S-
/•y" ,f ss
TrJm
rf-
«1
y^/y/7^

b)

Vegetated
T .A^*


Bare
^.x- Sediment

      0-2

  E
  u   2-4
  h-   4-6
  Q.
  UJ
  O   6-8
     8-10
551-
•:-l:^:/>^h
                              10         0         5         10

                               NO" plus  NO"  ,/iM
                              //?-**•
                                         .••••; •".•:.• • ?.•••• •.:••.•-. 'iijj
                                         •.•..».*• . •••••. •.:•.•: •' 'tti !>U
                                         ':' »':•?: •':':';';'8. •.''• '•' .'•: •' •::':' :l'-\f.' J
200
                  400
                                                 200
400
Figure 6.  Concentrations (yM)  of nitrate plus nitrite and ammonium in the pore
          waters  of sediments  with and without vegetation in a low and high
          nutrient treated pond on 15 July 1981.
                               VIII-14

-------

-------
•^ 50
w*
'E
1 ^o
•»
z
2 30
t-
O
U. 20
DENITRI
o
DENITRIFICA

~ Vi
sgetated
Unvegetated
^ \
i
\

* •
• *•
•*• (
* • •




(6
TION;

/ >
//
^


6)
»
^
• *
• *
* *
* •
••
• *
.*/

           LOW-NUTRIENT  HIGH-NUTRIENT
Figure 8.  Denitrification ratas (gmol'm'^h"1)  in the vegetated
         and bare sediments of a low and high  nutrient treated
         pond.
                     VIII-16

-------
I
                                           Light Attenuation
               Epiphytic Solids
                    The attenuation of PAR by epiphytic solids on-tbe leaves of Potamogeton
               followed a negative exponential relation (Y = YO£"U*   , r2 =0.8/7:TfiTs
               estimate      is based on the dry weight of total  solids on the leaf surface.
               Presumably, the relationship between light attenuation and chlorophyll  content
               of epiphytes might indicate a greater effect per unit pigment due to high
               absorption in the spectral range from 400 to 700 nm.


               Mater Column Attenuation

                    Attenuation of PAR by the water column in experimental ponds was not
               significantly different among control, low, and meaium nutrient dosed systems,
               but was much greater under high nutrient treatment (Fig. 9).  Attenuation
               coefficients (K) in the control ponds were about 1 m"*, while values were
               slightly higher (1.0 to 1.5m-l) in low and medium nutrient treatments.
               Turbidity in high nutrient ponds was significantly greater overall,  with PAR
               attenuation coefficients averaging about 4.0 nr1 and reaching 8.0 m-1 during
               the final sampling.  This attenuation coefficient of 4 m-1 represented  a
               reduction of typical ambient midday light (1200 pE m-2«s-l) from
               730 pE m-2-s-l just below the water surface to about 160 uE m-l-s-1
               at 0.5 m deep.

                    High nutrient ponds had elevated concentrations of total suspended solids
               and chlorophyll £ compard to other treatments (Fig. 10). Whenever attenuation
               coefficients were greater than 3 m~l, chlorophyll a concentrations exceeded
               150 vg/t. Chlorophyll £ above 20 vg/t  was observed only under high
               dosage.  However, while total suspended solid (TSS) levels were about 30 mg/t
               for control and low nutrient treatments; they were significantly greater (ca.
               60 mg/t) in medium and to the high nutrient treatments (Fig. 10).


                                             Plant Biomass

               Submerged Vascular Plants

                    Mean values for plant blomass prior to treatment were slightly  higher 1n
               control ponds  (ca. 120 gdw/m') compared to 80-87 gdw/m^ among the other
               treatments (F1g. 11).  Biomass Increased 1n all treatments during June and
               July to  levels ranging from a low of 157 gdw/m2 1n medium treatment  ponds to
               a high of 237 gdw/m2 1n controls. M1d summer results also showed that Ruppia
               dominated the blomass of the control and medium treatments (71 and 63%,
               respectively) while Potamogeton dominated the low and high nutrient  treatments
               (68 and 98%, respectively).

                    By August the mean total blomass had decreased in the medium and high
               nutrient ponds to about 80 gdw/m^, similar to early summer values.  Blomiss
               1n the low (196 gdw/m2) and control (201 gdw/m2) nutrient ponds were signifi-
               cantly higher  (p<0.05) than values for the medium and high nutrient  treatment.

                                                   VIII-17

-------
     2
o
u_
u.
LU
O
<->   4
o

H  Z
<
3
Z
UJ
h-

<  4

           HIGH
                             7+
             I

                                                CONTROL
                                                LOW
                                                MEDIUM

                                                        1
            6/6      6/11^  6/22    7/21     7/24     7/30
Figure 9.  Attenuation coefficients
         treatments.
                                    of the water column at four nutrient
                               VIII-18

-------
      ,tn
                             15



                             10



                              5
en O oi

-------
\
320

CM IfiO
** 1 W V
X.
* 0

"5 320
o»
CO*
CO

O 0
CD
< 320
60

n
-



™

—

•



•

~


-
•




.

L ,
* •* * * *• •*




|
:"••'••":'
































/
• \
'. '•
• ':
* *•




r*
f


u4
* \
/



/
i

'/
"t * .
';
•*




"7
4


In
y

••-•


/
^
COI





LO





ME




HI


mt^fttm^i^fmmittt
N'
V
• /,
• •
*j

U

y
^


D
1
/
•* f
».* *

6
p i

/
2
r
i
t^
• .•
•
• .*•
* • •*

\

j
s
?:

L
,
(
• .*
- •.
.•;
H
'/
/
i
ROL











IM








rr-rr-

,*,
* •""
* *•
*S


XK
^T A
//
>8I



/i/
yT yl
£*<£


i
^/




I* •'•*•* i


PA! AmAo •! on
^ »r W1UIIIW y « IUII
_Ruppia






1^ 1 T\




                              4JUN   I7JUN   I3JUL  I3AUG   I3NOV

                                              rt
              Figure 11.  Aboveground biomass  (gdw/m£) of  submersed macrophytes  in ponds subjected
                        to four nutrient treatments.  Fertilization began on 18 June, 1981.
                                             VI11-20
  I

-------
fi
              The  biomass  loss  In  medium  ponds  was  marked  by  a  decline  In the presence of
              Rup£ia  from  63% of the  total  biomass  in  July (99.2  gdw/m2) to only 7% of the
              total biomass  in  August (6  gdw/m2).   Ruppia  was sparse  in the high nutrient
              ponds  in  July  (2% of the total) and absent in the high  nutrient ponds 1n August.
              By November  the effects of  the fall senescent period  on SAV biomass were evident
              in these  systems, although  all of the ponds  except  for  those receiving high
              nutrient  dose  had some  biomass remaining.

                  Changes in SAV  standing  crop in  the ponds  between  sampling dates provide
              a measure cf net  productivity for these  submersed grasses (Fig. 12b). There
              was  an  initial  pulbe of SAV production in the control pond followed by little
              growth  during  the remainder of the summer.   In  mid-June SAV net production
              ranged  from  0.64  gdw m-2»d-l  at the low  nutrient  treatment to 2.77
              gdwm~2.d-l  at the high nutrient  treatment which  follows the rank of
              nutrient  levels applied.  By mid-July  net SAV production continued to increase
              in the  low nutrient  treatment, but declined  slightly  in the medium and high
              nutrient  treatments  to  about  2.0  gdw-m~2«d~  .  Net  production of
              SAV  in  medium  and high  nutrient dosed ponds  continued to decrease in August to
              values  less  the -2 gdwm"2  «d"^,  while SAV net  production in the low
              treatment was  still  positive  at 1.2 gdw«nr2-d-l.

                   Since the stratified sampling strategy  for SAV biomass di^ not monitor
              unvegetated  areas of each pond, total  biomass per pond  using areal coverage of
              vegetation more c'early defines changes  in SAV  biomass  in response to nutrient
              enrichment.  Incremental changes in this  total biomass in  relation to initial
              pre-treatment  measurements  are given  in  Fig. 12a.  Both control and low
              nutrient  treatments  exhibited nearly  constant increases in SAV biomass during
              the  summer,  and increases in  SAV  biomass were also  observed for medium and
              high nutrient  treatments from early June to  mid-July.   However, from mid-July
              to mid-August  the absolute  biomass of SAV in both medium and high nutrient
              treatments declined.  August  biomass  levels  in  medium nutrient ponds remained
              at 4 kg dry  wt above early  June levels,  but  SAV biomass was nearly 4 kg dry wt
              below  the initial levels in high  nutrient ponds.  Therefore the biomass of SAV
              in the  high  nutrient pond showed  a significantly  greater negative response to
              nutrient  enrichment.


              Epiphytic Biomass

                   Dry  weight and  chl-a_ of  total epiphytic material was sampled on three
              occasions during  this experiment, once prior to treatment and twice during
              treatment.  Epiphytic biomass as  dry  weight, was  always highest for high and
              medium  treated systems  (Fig.  13), with mean  values  exceeding 3.7 gdw/gdw(plants).
              Low  dose  ponds exhibited a  steadily increasing  total  epiphytic material, where
              mean values  grew  from about 1.2 in June  to 2.4  gdw/gdw(plants) in August.
              Unfortunately, only  one pond  could serve reliably as  control for this experi-
              ment.   However, epiphytic biomass in  this control pond  remained at less than
              5% of  levels in all  other ponds.   This control  system contained Ruppia maritima
              as the  near  exclusive SAV species, while the two  high dose systems were virtu-
              ally devoid  of this  species.  However,  there  was no  distinct pattern in the
              accumulation of epiphytes of  J*. maritima versus Potamogeton perfoliatiis in the
              low  and medium dosed ponds  where  both plants were Important" to total biomass
              (Figs.  11 and  13). Epiphyte levels were  significantly greater for high nutrient

                                                 VIII-21

-------
f!
* 60
>s
h.
T3

o»
je 40

CO
(O
                    O

                    CO
                       20
                    C9
                    O
                            A)BIOMASS  INCREMENT
                                                              CONTROL
                                                     LOW  ^.-»
                                                   ... x      MEDIUM
                                                        *V*J	c
                                           HIGH
                                     B) NET  PRODUCTION
                              JUNE
                         JULY
AUGUST
                    Figure 12.  a) Changes  1n the total biomass of submersed macrophytes
                              (kg dry wt.) 1n the ponds subjected to four nutrient treat-
                              ments,  b)  Average daily changes in the biomass of submersed
                              macrophytes gdwnr2-d-l) in ponds at four nutrient treatments.
                                            VIII-22

-------
t:
4
> 2
<
CO
o*
X.
«>
•o 4
^
CO
o 2
o
H
CONTROL

LOW

Potomog«ton
Ruppia /

1 ^Ss
•".'•• 1 "•••: 'Vy
£% ^ ^^



(O
CO
< 4
2
O
CD 2
UJ
Q.
0. 4
UJ
2
MEDI,UM
J , As
•!•. ry>
• '.••/ /^^

! S ifi^
;:.{i^ v^'X
,i,

•;''•'•
•'."."•/
//
//
//















\




HIGH

fi

: I i

JUNE JULY
'•*•'.
• *.*
• * « *
•"* * "

\






i .
i
1









AU6
Figure 13. Epiphyte biomass (gdw/gdw(S/W)) on Ruppia
and Potamogeton perfoliatus in ponds subj
maritime
ected to
                                         four nutrient treatments.  Fertilization began after
                                         the June measurements.
                                                     VIII-23

-------
treatment than for low or control, whereas medium dosed epiphytes were signlfi-
cantly different only compared to controls.

     Epiphytic chlorophyll-^ biomass shows a much clearer response to nutrient
treatment, although data from a control  pond (#1) was not considered   (Fig. 14),
Chi-ai measurements were less than 0.3 mg/gdw(plants)  at pretreatment for both
specTes.  By August, chlorophy!1-a_ remained at this pretreatment  level  v,n
Ruppia in the control pond,  while concentrations on Ruppia and Potamogetpn
increased in relation to nutrient treatment up to 1.6 mg/gdw(plant) in the
high nutrient pond (Potamogaton) (Fig. 14).  The most dramatic Increase in
chlorophyll-a occurred on Potampgeton in the high nutrient applications of
nutreints.  The chlorophyl1 -£ biomass increased from  0.07 to 1.58 mg/gdw in
one month.  This Increase represented a  shift in the  chlorophyll-a:  total
solids ratio from 0.003 to 0.054, compared to a change of about 0.012 to 0.018
in the other two treatments  for Potamogeton.


                     Plant and Epiphyte  Productivity

Plant Production and Respiration

     Diurnal patterns of apparent production of P. perfoliatus (Fig.  14)  show
unvarying rates from 1100 -  1600 h on a  cloudlesT summer day (15  June 1981).
Production:respiration ratios (P/R) were also remarkably constant during
this period at about 1.8.  In contrast,  apparent production rates of £.
maritima exhibited a distinct temporal sequence, with rates at midday"~nearly 4
times greater than morning and afternoon rates, which were similar.  Respira-
tion, however, was constant  during the day for R. maritima, resulting In a
range of P/R ratios from 1.8 In the morning to.lT./.at midday. The peak apparent
production rate for Ruppia was 10.2 mg 02»gdw~1»hi compared to
8.1 mg 0«gdw  «h   for Potamogeton.
    mg O^^gdw  "h"1 for Potamogeton.

     Vertical distribution of production along plant stems  was  also different
    >en these two species (F1g. 16).   Net production varied line
length of Potamogeton ranging from 5  mg Ogdw^h'1 for the top
between these two species (Fig. 16).  Net production varied linearly along the
                 »ton ranging from 5 mg C»gdw~*»h~l for the top 5 cm
section to 1.2 mg Ogdw-l«h-I for the bottom section, more than 45 cm
from the tip (F1g. 16). However, net production was  nearly  constant  along  the
length of Ruppia at about 3.5 mg C^gdW^h'1.   The quantity of PAR
available from the water surface to 0.5 m 1s also shown  in  Fig. 16,  and  Ind-
icates that it was 1n excess of saturation values for photosynthesis (300-600
E1n nT2s'l) for both Potamogeton and Ruppia.

     The internodal length (i.e., distance between leaf  rodes  on the stem)
     ased along
plant to greater
decreased along Potamogeton from less than 1.0 cm in the terminal  10 cm of the
          reater than 2 cm at the base. Since the merlstem of Potamogetpn  1s
terminal, the leaves at the tip are usually younger than leaves  in the basal
section. The branching of Ruppj a 1 s more consistent along the length of the
plant and these branches, even those near the base of the plant,  are long
enough to reach the surface of the water. Therefore the net productivity dis-
similarities of these two submerged species seems to be a result  of the
morphological arrangement of the photosynthetic structures along  the stem of
the plants.


                                    VI 1 1-24

-------
2.0
< i.o
CO
•o

>•*
ol
J. 2.0
.c
o
6 1.0
CO*
CO
<[
jg
O
5 2-0
UJ
t 1 0
>- ••u
X
0.
Q.
!.•
CONTROL
f^+ffmf f^*+f*m lf*^4v^V

LOW
i— 4.«*_>_ -' '•' /* Jl
PtfTT^?^ ; ' '. '7>^/l • :. I '.' y^J

MEDIUM
Ruppio ,
* • • • *
i ^^^^^ t '•• • •.
^^^ ;;;^ ;^.^

     2.0
     1.0
Hl<
- Potamogeton ^^
trr-r\
3
/
/
/
H
£
/

.1 ,
I

             JUNE     JULY
AUG
Figure  14.  Epiphyte biomass (chl-a/gdw (SAV)) on Ruppia marltlma
           and Potamogeton perfolTatus in ponds subjected  to four
           nutrient treatments.   Fertilization began after the June
           measurements.
                          VIII-25

-------
     BOO
     400
 CC
 <
 CL
        8
 I

  •
 I

 •o
  CM
 O
       8
          1000
                            PAR
                                                SURFACE
                           POTAMOGETON
1200          1400
     TIME.h
1600
Figure 15.   Photosynthetic  active radiation  (yEin-m'2^"1)  at the water's
           surface and at  0.5 m depth from  midmorning to midafternoon
           along with measurements of apparent production  (Pa) and net
           respiration (Rn) for Ruppia maritima and Potamogeton
           perfoliatus.	
                           VIII-26

-------
                          'Hid 30
Figure 16.   Net photosynthesis  (ing C«gdw~l'h'l)  along  tha  length
            of Potamogeton perfoliatus  and  Ruppia morltima cleared
            of epiphytic material.Included  is  the amount of
            photosynhtetic active radiation (yEin'm"2^'1) from
            water's surface to  0.5 m depth.

                           VIII-27

-------
Plant/Epiphyte Relations

     Apparent production of Potamogeton taken from a low nutrient pond, was
    .«.. *h,» c „„ n ^^...-i.u-i f0r the tip sections, with rates above
                                 for plant tips cleaned of epiphytic
greater than 5 mg {X^gdw  'h"1 fc
8.1 mg Op'gdw  «h   on two dates
mater (Fig. 17). Pa of the basal
      (Fig. 17). Pa of the basal  sections of Potamogeton was always lower
than rates for the tip ranging from 4.8 to 6.0 mg 0^*  gdw  »h   (55% of
Pa for the 0-10 cm sections). The combined Pa of plant and epiphytes ranged
from 11.2 to 12.4.mg Oo-gdw  -h   for apical portions, and 5.8
to 7.2 mg ^'gdw'1' h~* for lower stem. In all cases,  plar.ts domin-
ated total apparent production in the low nutrient pond. Epiphyte levels on
Potamogeton in this pond were less than 2.0 gdw/gdw(SAV).

     Peak Pa for Potamogeton from high nutrient ponds  was only 4.0 mg 03*
gdw'1'!!'1 (Fig. 17). This rate was for a tip section of the plant
which had 1.40 gdw/gdw(SAV) of epiphytes, which was the lowest epiphyte biomass
value observed on the plants tested in this pond. At epiphyte biomass levels
above 3.0 gdw/gdw(SAV), SAV production was less than 1.0 mg 02«gdw"1-h"1,
which occurred for both the tip and basal  sections of Potamogeton in August.
Although the contribution of SAV to total production was minor, the elevated
epiphyte biomass resulted in total Pfl (epiphyte plus SAV) >.5 ;ng 02*gdw"1«h"1
for July and August with a peak value of 8.0 mg O^gdw  »h  .  The
contribution of epiphytes to this production ranged from 58 - 100%.

     SAV respiration ranged from 0.30 to 2.45 mg 02-gdw"1'h"1 with an
average of 1.4 mg Oo-gdw  «h   for Potamogeton from the low nutrient
treated pond resulting in P:R ratios from 4.0 to 30.  P:R ratios for epiphytes
from this pond ranged from 0 to 3.6; therefore the total production of the
SAV/epiphyte community was always greater than total respiration at an average
ratio of 4.4 (Table E-l).  Respiration of Potamogeton  from the high nutrient
treatment ranged from 0.40 to 6.03 mg Q^-gdw  »h   with an average of
1.9 g Og'gdw  »h  .  These higher respiration rates together with re-
duced net production values resulted in much lower P:R ratios ranging from 0
to 9.8. Only 2 of the 5 measurements of SAV metabolism from the high nutrient
treatment had P:R ratios greater than one. Although the P:R ratios of epi-
phytes from this treatment were similar to those ratios for epiphytes from the
low nutrient pond; the total P:R ratio of the SAV/epiphyte community under
high nutrient application was only 1.8 compared to 4.4 for the low nutrient
treatment.

     These data are based on DO changes of epiphytes 1n one chamber and epi-
phytes plus SAV in the other. By subtracting these res-ilts, the partitioning
of production and respira'cion of .just SAV and epiphytes were calculated for
each experiment.  The uptake of   C-C02 by epiphytes and SAV gives a more
direct measurement of the relative net productivities  of these plants under
varying levels of epiphyte biomass.

     Net production (as   C-CO? Incorporation) for the 0-15 cm section of
Potamogeton decreased to 73% of control values at an epiphyte density of
5.04 gdw/gdw(SAV). However, the total net production of the epiphyte and SAV

                                    VIII-28

-------
        12 -
         8 -
T   «•
CO
Mr
-

1
m
fffl
• •• ji



L(

DW
$'.•5:
v :':•":
•V'-'i

• • • •
• • ^


NUTRIENTS

.>•*•.•.{
\<-:-.:.r
1
^

I
1
l|
•::\j%1

%
• • • *




?f-«3
;:i;>:Ci
1
$
• • • •
It * ft



  *1
2      12
O
00
LU
 8 H



 4

(-f)
HIGH
' E3 Epiphytt P0
. E3 SAV P0
0 SAV Rn
• ESI EpiphyU Rn
"
' S,
5^ I
• • • '
B * * •
•
>^;-v
^
f::.-x\
W-
i
Y/<
* • * *



NUTRI

&$
• •£•
; )|
>^8
i
tt
7 1-
1
1
•
r
It .
EN
Si
:.:.:':"?;
!'•'.'.** 15
II
;'•';"•;«
:::H^
• IV •/!
• . .
• •
• •
• •
• •
•
• •
» •
T


S
',
•
>

tt.
• • *
• •
•
T * d



                TIP
                I8JUN
                   TIP  BASE
                     15 JUL
Tl?  BASE
 5 AUG
  Figure 17.  Apparent production and net respiration [mg 02 (9 dry wt.
            SAVJ-l-h-1] of Potamogeton perfoliatus and associated epi-
            phytes from a low nutrient and a high nutrient treated
            pond.
                          VIII-29

-------
community increased from 5.09 nig C-gdw-^h'1 (control) to 7.79 mg
C-gdw-Lh-1 at this high epiphyte biomass level  (Fig. 18).  This was
an increase of 153% of the net production of nonepiphytized plants.  At  the
same epiphyte biomass level, net production of the 15-55 cm section  of  Pota-
mogeton decreased to only 56% of the control.  As  observed for the tip oTTota-
mogeton. the basal section of the SAV/epiphyte community increased in totTT"
                                 -l-h'1 (control)  to 3.78 mg C-gdw-^h'1
net production from 2.32 mg C-gdw
(163%). At medium levels of epiphyte biomass, which ranged from 1-2 gdw/gdw(SAV),
the net production of neither the tip nor basal sections of the plant were
affected by the fouling.

     Epiphyte biomass was much lower on Ruppia compared to Potamogeton at  0,49
and 0.39 gdw/gdw(SAV) for tip and base, respectively.  However,  the  reduction
in photosynthesis on these plants was much higher at 67% of the control  for
the tip and only 23% of the control  for the basal  sections.  Also in  contrast
to Potamogeton, the total community net production of the lightly fouled Ruppia
was lower than the controls. Therefore epiphyte production was  not  enough  to
compensate for the reduced net production of Ruppia due to epiphytic  fouling.


                 Community Photosynthesis and Respiration

     Apparent production in control  and low nutrient pond communities exhibited
similar patterns with an increase in net production during the  latter week  of
July followed by a slight decrease in mid-August (Fig.  19JL The peak  production
rates in these two treatments were 8.71 and 8.23 g 02 • m  «d  ,
respectively. Through June and mid-July the medium and  high nutrient  dosed
ponds also increased in net production with peak values of 11.05 and  12.93 g
CU»m  »d  , considerably higher than for control and low treatments.
The peak increase in net production rates in these two  treatments was about
4.5 g O-'irf^'d'1 above pretreatment values, and the increase
occurred sooner in the high dose pond compared to medium nutrient treatment.
This rate of change in net production during the summer was more than double
the Increases observed in both the control and low nturient treatment (F1g. 20).

     Following these late July peaks, community Pa decreased In the control,
low, and medium nutrient treatments by similar magnitude during August.
However, only the medium nutrient pond had net production values greater than
pretreatment values. Net production 1n the high nutrient oonds  declined  dras-
tically during this period to levels more the 3.0 g 0?  nT^d"1  less
than pretreatment values. Thls.decline 1n net production represented  a total
decrease of more than 7 gO^'m" 'd"1 during a 10-day period, and
1t 1s coincident with the observed loss of SAV. Only under high nutrient dosaged
did P:R become less than 1.0. From the end of July to mid-August the  P:R
ratio of these ponds dropped from 1.5 to 0.6.


                                DISCUSSION


                  Nutrient Uptake from the Water Column

     Uptake rates of dissolved Inorganic nitrogen and phosphorus from the

                                    VIII-30

-------
              CONTROL
                           POTAMOGETON
MEDIUM
                                 HIGH
SURFACE


 E

 I
 H    0.25
 CL
 UJ
 O

      0.50

Plant
   Epiphytes

              2468     2468      2468
                PHOTOSYNTHESIS.mgC-gdwfSAV)-'.!.-1

                         RUPPIA
SURFACE


jf 0.251
a.
UJ
o
0.50
C<
'•:;• :•.*•

•"• •* • * *. *
'•':'•'••
* , . ". » •
3
i
.•


NTROL



Plant -^


''• '
•*#
•V
MEDIUM
• • •
'. .';

Epiphytts
i i i i
              2468      2468

               PHCTOSYNTHESIS.mg C-gdw (SAV)"'. h-1


 Figure 18.  Net photosynthesis [mg C (g dry wt. SAV)'1^'1] of the tip (0-15 cm)
          and basal  (15-50 cm) sections of Potamogeton perfoliatus and Ruppia
          maritima and associated epiphytes.
                            VIII-31

-------
 I
 •o
 *>•
  N
    o 8
  c
o:    4
   c
   •
        -4

         •,



       , 8


         4


         0
O
O
Figure 19.

                          CONTROL

                                            %
                                           y*,
                                           I
                   i
                           LOW
                                              I
                          MEDIUM
                   pb
                       HIGH
                  6    23  24

                 JUNE
                            ^
                   Ki
16T  17   23 T 30  31

        JULY
^
                                                            i
                                                    *   10   1!

                                                        AU6
         Photosynhtesis  and respiration  (g Oc«m" »d" ) of the pond communities
         at four nutrient levels during  summer of 1981.
                             VIII-32

-------
I1
                     O
                     o
                     O
                     O
                     tr
                     a.
!~-2
                     X
                     a
                                          CHANGE   IN

                                COMMUNITY   METABOLISM
                                      1   Medium
                                                                Control
                                                              \*  Low
                                                               i
                                                               i
                                 JUNE
                         JULY
AUGUST
                    Figure  20.  Changes in net production  (g 0-ni~3.d"l)  of the pond communities
                              at four nutrient treatments during the summer of 1981.
                              Fertilization began after  the June measurements.
                                              VI11-33

-------
r
                water  column  appear to have been dependent on concentration (Fig.  21).  Uptake
                (of  NOj and  NH^ were similar between  0 and 10 yM  N; above this concentr-
                ation  NH4 uptake  rates were much higher, indicating a preference for NH^.
                At  about 60 yM N  the uptake was only 18 yM/d for NOo compared to
                30  uM/d for NHj.  The curves in Fig.  21 suggest that NOj and NHj
                uptake by these SAV communities was  diffusion-limited without clear indication
                of  saturation up  to 90 uM.  Other investigators have, also, reported a prefer-
                ence of NO: over  NH+ for various SAV species (Nichols and Keeney 1976;
                Best 1980;  lizumi  and Hattori 1982)  as well as phytoplankton (McCarthy et al.
                1976).

                    These  NO^ and NH^ uptake curves are similar to patterns of nutrient
                uptake observed elsewhere for £. perfoliatus.  Laboratory uptake studies have
                shown  that  NH^ uptake by shoots of P. perToTiatus is 10 times greater than
                N(C uptake  (Stevenson et al. in prep)/Field studies in P. perfoliatus
                beijls along  the Choptank River estuary using 0.51 m deep chambers revealed that
                NH4 uptake  rates  were similar to those for NOj at+N concentrations up
                to  about 30 yM, but at concentrations > 75 yM, NH* uptake rates were
                generally twice those of NOj (Kaumeyer et al. 1981).  Thus, it appears that
                SAV communities have a higher absorption capacity for NO^ and NH^ compared
                to  plankton communities, and that the absorption rates for NH^ are much
                higher compared to NOj uptake.


                                   Nitrogen Uptake  and Storage by SAV

                    The relationship between tissue nutrient content of aquatic plants and
                nutrient enrichment observed in this study has two important ecological implica-
                tions.  Tissue nutrients in aquatic  plants have been proposed as an indicator
                of  the nutrient status of their ecosystem (Gerloff and Krombholz 1966; Fitz-
                gerald 1969;  Gossett and Morris 1971; Adams et al.  1971, 1973; Seddon 1972).
                Also,  aquatic plant communities may  be managed in a variety of schemes for the
                treatment of  nutrient wastes with subsequent economic benefits from the
                increased nutritional quality of the plants (Boyd 1970, 1971; Siedel 1971;
                McNabb 1976;  Wolverton et al.  1975).

                     In this  study, N content of Potamogeton leaves and stems increased markedly
                in  proportion to  nutrient enrichment.  The maximum t i ss> iie™F~cTjn~c"eTrt"r"attwt
                (3%) was observed in the high nutneTit treatment (120 yM N/wk).  In contrast,
                tissue N concentration of Potamogetpn growing in ambient estuarine waters was
                only  1.2% N.  Nitrogen content of plants at low and medium treatments were
                intermediate, both being about 2% N.

                    Boyd and Scarsbrook (1975) also observed significant increases in tissue
                N of Eichhornia crassipes for fertilized ponds; however, there was little
                distinction in tissue N (1.8%) among various nutrient enrichments ranging from
                7 to 35 yM  N  per  week. Under laboratory conditions, media concentrations of
                3000 yM N were required for the blades and floats of E. crassipes to reach N
                concentrations of 2.0% (Gosset and Norris 1971).  A cHange in N content for
                Vallisneria americana from  1-2% in the laboratory occurred in response to an
                increase from 500 to 2000 yM (as NOo) in the growth medium (Gerloff and
                Krombholz 1966).   Our study indicates that P. perfoliatus with tissue N >
                1.5% may be an indication of a nutrient enrTched environment, although precise

                                                    VI11-34

-------
                                      V)
                                      V
VIII-35

-------
relationships to the "trophic" status of those waters await further study.

     Luxurious accumulation of nutrients in submersed aquatic plant communi-
ties has been used successfully in ameliorating nutrient waste problems
(Sinclair and Forbes 1980).  Gerloft and Krombholz (1966) showed that J/.
americana accumulated nearly 75% of the N in growth media ranging from 15C to
3000 uM N, while at 6000 yM the percent recovery dropped to    56%.  During
their study, plant growth rates remained relatively constant while N content
of plants increased.  In our experiment, both SAV productvojT^ajjdJMjisue N
increased with fertilization (Fig. 22).  It appears fhait growth was stimulated
in associated' with tissue N >1.5%, which is similar to the critical concentra-
tion of 1.3% N proposed by Gerloff and Krombholz (1966).  Nutrient amended
plants accumulated about 1.45 g N/m' more than plants in unfertilized ponds
in July.  This excess accumulation was due to Increased tissue nitrogen and
biomass production for all nutrient treated ponds (Table 1).  The percentage
of applied nitrogen that was recovered in SAV biomass ranged from 113% in
the low nutrient ponds to 30% for the high nutrient treatment (Table 1).   By
13 August following six applications of nutrients to the ponds, the percent
recovery of the applied N by SAV tissue changed dramatically. Results for the
low nutrient treatment were similar to that observed on 13 July (127% recovery,
Table 1). However, N incorporation in the medium and high nutrient ponds  dropped
to 9 and 7%, respectively. Similar results were observed in pond fertilization
studies with JE. crassipes, where at 4 and 8 uM N treatments the plants
recovered larg"e amounts of N; but at the 34 pM N treatments recovery was
minimal (Boyd & Scarsbrook 1975).

     This change in the pattern of N accumulation by SAV can be explained by
examining the relationship between tissue N concentrations and SAV biomass
production on 13 August (F1g. 22).  Low nutrient treatments resulted 1n an
Increase in biomass production and an increase 1n tissue N concentrations to
1.8% in July.  However, even though tissue N concentrations were greater than
2*0% 1ft the high nutrient ponds, SAV production was negative by as much ps 3.0
gdwm-2. d-l (Fig. 22).  In short-term laboratory studies where               ^
plankton populations are controlled, SAV growth 1s observed for all nutrient
amendments.  In these field experiments, SAV biomass decreased under highly
eutrophic condiTions.  Therefore, eventhough the tissue N of SAV may increase
in response to nutrient enrichment; the "nutrient buffering effect" of SAV
is operative only under moderate eutrophic environments.                   i

     The luxuriant consumption of nutrients by aquatic plants may not only
ameliorate moderate eutrophic conditions; but also Increases the nutritional
quality of the macrophytes.  In this study, an increase in nutrient enrichment
resulted in a proportional decrease In tissue C:N ratios of SAV. Presumably
this Inorganic nitrogen was converted to ami no adds and proteins which would
increase the value of SAV as food (direct or indirect) for fish and waterfowl.
For macroalgae some of this excess tissue N may be stored as NO^ rather
than protein (Wheeler and North 1981). Although 1t has been shown that  season-
ally elevated tissue N is usually proteinaceous (Best 1977), further studies
are needed to determine the chemical nature of accumulated N in SAV in  response
to fertilization.
                                    VII I-36

-------
 \
y
                  M
                  'E
                                          13  JULY


O

O
O
cr
Q.
                  5   0
                  O  (-)

                  CD
                       I
                  -I   2
                  O.
1.0
                                  1.5
2.0
2.5      3.0
                                          13 AUGUST
                         1.0       1.5       2.0      2.5      3.0

                              TISSUE  NITROGEN (%)

                 Figure  22.  The  relationship between tissue nitrogen  content (%)
                            and  biomass production of submersed macrophytes
                            (gdwm-2«d-l).
                                        VIII-37

-------
Table 1. The amount of nitrogen applied  (g/m2)  compared  to  the  amount  of  nitrogen
     recovered in plant tissue (g/m2)  above controls for three  nutrient enrich-
     ment treatments.
NUTRIENT ENRICHMENT
LOW
13 July
Nutrients Applied 1.26
(g/m2)
Nutrients Recovered 1.42
in SAV (g/m2)
% Recovery 113
13 August
Nutrients Applied 2.52
Nutrients Recovered 3.19
in SAV (g/m2)
% Recovery 127
MEDIUM HIGH

2.52 5.04
1.40 1.49
56 30

5.04 10.08
0.47 0.72
9 7
                                    VIII-38

-------
 ,^                                     Sediment  Nutrient  Dynamics
1
     Denitrification rates measured in the low (~20  mol N m-2 h-1)
and high (~40 pmol N»m-2«h-l) nutrient treatments were within the
typical range of 10-100 ymol N«m-2.h-l previously reported for
denitrification in estuarine sediments (Sorsenson 1978; Koike and Hattori
1979;  Cren and Blackburn 1979; Seitzinger et al.  1980; Henrikson et al. 1981;
Nishio et al. 1982).  Denitrification rates in our study responded to NO,
enrichment; however while NOo levels were increased by a factor of 7 (10-70
the denitrification rates only doubled. This is consistent with the kinetic
relations reported by Billen (1978).  Koike and Hattori (1979) and others who
indicate that K$ is about 30-50 uM NO^.  Although denitrification
appears to have responded to eutrophication, the rates observed in this study
accounted for only 3-6% of the NO, applied weekly to the water column in
the low nutrient pond and 2-4% in the high nutrient pond.  Assimilatory uptake
was the dominant process causing loss of NOo from the water column in
these ponds following fertilization.

       Oxygen transport to estuarine sediments can be facilitated by the presence
of seagrasses (Oremland and Taylor 1977; lizumi  et al. 1980) which may promote
nitrification in otherwise anaerobic environments. This enhancement of N£
production should increase denitrification rates  if NOj is the limiting
factor. In this study we did not observe significant differences in denitrifi-
cation rates between vegetated and unvegetated sediments in either the high
or low nutrient treatments. Bremner and Blackburn (1979) reported that acetylene
inhibits nitrification as well as denitn*'  ;ion in agricultural soils, so
that our acetylene-blockage technique may i,.,* have been appropriate for detecting
such nitrification related effects.

     The significantly lower NH£ concentrations in the interstitial waters
of vegetated sediments in the low nutrient pond may reflect root uptake by
SAV.   The average NH4 difference in a m  of sediment, 10 cm deep, between
vegetated and unvegetated sites was 140 yM. Based on a bulk density of 1.8
g/cm3 and 20% water content for these sediments described above, this is
equivalent to 6300 umol N. Root biomass in this pond was 20 gdw/m2 and
root uptake was about 4 yg-at N«gdw-l« h-1 (Stevenson et al., 1n prep.;
resulting in total SAV uptake of 1900 ymol  N«m-2«d-l.  This is
equivalent to 30%/d of the difference between these two areas and represents
a turnover time of 1.4 and 4.7 d of the sediment  NH^ pools in the low and
high nutrient treatments, respectively.


                  Algal Response to  Nutrient Enrichment

     There have been numerous experimental  studies of phytoplankton response
to nutrient enrichment in microcosms of various sizes and shapes.  Generally,
these studies have observed rapid response of phytoplankton Pa to fertiliza-
tion (e.g., Edmondson and Edmonson 1947; Edmondson 1955; Abbott 1967, 1969;
Dunstan and Menzel 1971) and elevated but variable algal  populations and chlor-
ophyll  a concentrations (e.g., O'Brien and de Noyelles 1974; Gamble et  al.
1977).  Typically, the phytoplankton response 1s not linear with respect to
nutrient additions (O'Brien and de Noyelles 1977; Gamble et al. 1977; Nixon,
personal communication), and a variety of factors may be responsible for

                                    VII1-39

-------
r
mollifying or exacerbating this relation.  These include nutrient interactions
with sediment, grazing on plankton, competition with periphyton and macrophytes,
or limitation by carbon or other requisite substances.

     Several studies have involved fertilization of experimental  systems con-
taining submerged vascular plants, and these have similarly observed increases
of phytoplankton biomass at high doses.  However, a much reduced response was
reported at low and moderate nutrient doses, perhaps, as a result of competi-
tion with macrophytes (Mulligan and Baranowski  1969; Mulligan ?t  al.  1976;
Hall et al. 1970). In a recent paper Phillips et al. (1978) argued that phyto-
plankton response was a secondary effect of fertilization, following initial
periods of prolific epiphytic growth.  Apparently, this increased phytoplankton
growth in response to nutrient additions is not confined to microcosms  and
occurs also in nature, as strong correlations between the two have been ob-
served for lakes (Schindler 1978) and estuaries (Boynton et al. 1982).

     Phytoplankton biomass (chlorophyll a) exhibited a strong and very  signifi-
cant increase under high nutrient treatment in  our experimental ponds (Fig. 23).
However, low and medium treatments seemed to have little effect on plankton
biomass. While this comparatively reduced phytoplankton response seen at moder-
ate treatments may be a result of the inability of our sampling scheme  to
detect ephemeral blooms (with measurements 1-3/wk), it may also indicate an
inferior ability of phytoplantkon to compete with SAV and their epiphytes for
limited nutrients.  We did see increases in chlorophyll a of 1-2 orders of
magnitude occurring at high treatment over the  course of~l-2 days (Table C-2),
suggesting the highly variable nature of these  populations. For the most part,
when chlorophyll ^ levels were > 2 ug/t , plankton communities were
virtually monospecific, dominated by small coccoid chlorophytes.   Undoubtedly,
the low mixing rates in these ponds influenced  competitive selection conferring
advantage to motile and bouyant species.  Other investigators have also reported
widely fluctuating chlorophyll a levels in response to nutrient additions
(Edmondson 1955; O'Brien and deFToyelles 1974) and have attributed this  erractic
behavior to such factors as pH (carbon limitation) and grazing. Water column
attenuation of PAR followed a similar pattern relative to treatment, while
total seston exhibited a small but statistically insignificant increase with
treatment (Fig. 23).  Thus, it appears that living phytoplankton cells  generally
represented a small fraction of the total seston, but exerted a large effect
on PAR attenuation.

     Over the last decade an expanding literature has suggested that growth of
epiphytic algae in response to fertilization may be more important than phyto-
plankton relative to effects on submerged vascular plants (Moss 1976; Phillips
et al. 1978; Cattaneo and Kalff 1980; Harlin and Thome-Miller 1981; Sand-Jensen
and S0ndergaard 1981).  Interpretation of these  data is complicated by the
fact that micro and macro-algae may exhibit differential responses in time
(Phillips et al. 1978;  Harlin and Thorne-Miller 1981), and there are no previous
controlled experiments in brackish or marine environments.

     In our study epiphytes exhibited a different response to fertilization
than did phytoplankton, where all treatments (low, medium and high) experienced
elevated levels of epiphytic material compared  to controls.  By the final
sampling in August, epiphytic material averaged 0.1, 1.9, 2.8 and 2.9 g/gdw
(SAV) for control, low, medium and high treatments, respectively (Fig.  13).

                                    VI11-40

-------
2  E
o  ^r
 UJ
 U_
 UJ
 o
    50
 _  40
 o| 20
 _J

 S  I0
     0
x  40
 o»
 E
  -  30

O
H  20
UJ
    10


     0
                 PAR  ATTENUATION
                                                (7.2)
                                               (210)
                                                "
                   PLANKTON   CHLa
                 SESTON
         CONTROL    LOW     MEDIUM     HIGH

                NUTRIENT  TREATMENT

Figure 23.  The response of photosynthetic  active radiation (attenuation
           coefficient, K) , plankton chlorophyll a_ (urn/ft) and total
           sestion (mg/K,) of the water In  pond mesocosms subjected to
           four nutrient treatments.
                       VIII-41

-------
T!
The same rank occurred for epiphytic chlorophyll-a which averaged 0.08, 0.37,
0.75 and 1.58 mg/gdw(SAV) in August (Fig. 14).  TFe levels of epiphytic mate-
rial in nutrient treated ponds are similar to those observed by Staver et al.
(1981) for Chesapeake Bay and by Menzie (1979) for Hudson River upper estuary.
Control levels are comparable to those observed by Staver et al.   (1981)  at
Parson's Island in July. It appears that in the competition for light and
nutrients between phytoplanktonic and epiphytic algae, the latter have an
edge under all but the most extreme nutrient additions where phytoplankton
are very competitive.


                    PAR Attenuation by Algal  Biomass

     An effect of nutrient-stimulated algal  growth, which is of importance for
the production of submerged vascular plants, is the attenuance of PAR.  In our
experiment PAR downwelling attenuation coefficients (K) exhibited a  moderate
but statistically significant correlation (r2 = 0.32)  with total  seston
(Fig. 25A). The slope of this relation (0.0396) is about twice that  which we
determined experimentally (0.0174) for inorganic sediments (Kemp  et  al. 1981),
suggesting that a greater percentage of the material  suspended in the ponds
was pigmented (so as to absorb rather than merely reflect efficiently in  the
400-700 nm wavelength range). Another indication that  higher values  of K  were
associated with fertilized systems is seen in Fig. 24B, where a strong correla-
tion (r2 = 0.69) is obtained between total  nitrogen in the water  column and
PAR attenuance.

     Overall, the relation between K and planktonic chlorophyll ^ was vari-
able, particularly at low pigment levels where other  materials dominated  the
seston (Fig. 25A). However, if we consider only the few instances where algal
cells comprised a major fraction of sestion (>10%), a  very strong correlation
occurs (r2 = 0.95). Here (Fig. 25A) we estimated dry  weight of algal cells
by assuming a carbon:chlorophyll £ ratio of 100 (Hunter and Laws  1981) and a
dry weight:carbon ratio of 2.5 (Parsons et al. 1961).  The slope of this cor-
relation (0.0146) is remarkably consistent with low Input of Inorganic sedi-
ments (Riley 1956; Scott 1978).  These values, however, are about twice those
obtained experimentally for pure algal cultures of Dunallella sp. and
Pseudoisochrysis sj>. (0.0061 and 0.0081) at our laboratory (Kemp  et  al. 1981),
or those reported elsewhere (e.g., Parsons and Takahashl 1973; Bannister  1974).
This discrepancy is likely attributable to the additional  presence of consider-
able amounts of plankton detritus and Inorganic solids 1n pond water columns.

     Epiphytic material also reflects and absorbs PAR before reaching SAV leaf
surfaces. We obtained an excellent regression between PAR attenuance
through a plane and epiphytic biomass (r* = 0.87) with a slope of 0.36%
(mg/cm2)-l. Surprisingly, this 1s very similar to the relation obtained
for seston (with some simple unit conversions, the slope becomes  0.396% (mg/
cm2)"*), suggesting that the attenuance properties (per unit wt.) of seston
and epiphytes are comparable.

     Several previous studies have examined this relationship between epi-
phytic material and attenuance.  Phillips et al.  (1978) measured attenuations
at several wavelengths for glass slides colonized by  algal epiphytes. They
provide cell counts as a measure of biomass, and we converted these  to mg

                                    VII1-42

-------
r:
 . 2
h-  I


UJ

U
u.
U
o
o
h-
<  3
tu
I-
I-  2
       A)
           PAR  ATTENUATION  VS.  SESTON
                                                                         H
                                                       Y-0.0396 (X)+0.39
                                                        (r2-0.32)
                              10       20      30      40      50

                                   TOTAL SESTON,mg/l
      u B)
            PAR  ATTENUATION  VS.TOTAL N
             Y-O.I27(X)+0.66

                (r2-0.69)
             25
                                       50
75
100
125
                 TOTAL  NITROGEN,/JM
Figure 24.  Correlations among the attenuation of photosynhtetic active radia-
         tion of the water column in ponds treated with nutrients and A)
         total seston concentration (mg/fc) and B) total nitrogen concentration
         (um).
                         VIII-43

-------
 E
~L   4
          A)
               PAR  ATTENUATION

             VS. PLANKTON CHL a
 o
 LL
 U_
 UJ
 O
 O


 O
 H
 UJ
                              80
                                        !20
160
200
  . 150
UJ
e>
<
H-
O
    100
     50
          B)
            TOTAL NITROGEN  VS. PLANKTON  CHLa
        - /
                  40        SO        120      160

                 PHYTOPLANKTON  CHLa,/jg/l
                                                             200
Figure 25.  Correlations between A) attenuation coefficients for photosynthetic
          active radiation and plankton chlorophyll a_, and B) total nitrogen
          concentrations (um) and plankton chlorophyll a_ 1n water column of
          ponds treated with nutrients.
                                VIII-44

-------
dry wt. using chl a/cell data of Hickman (1971) from a nearby locale and dry
wt/chl a data of STaver et al.  (1981). The resulting relationship for 3 levels
of epipFyte colonization is almost identical with ours, where 10 mg/cm2 results
in about 84% absorption and 2 mg/cm2 causes 52% absorption at 615 nm. On the
other hand, the data of Borum and Wiurn-Andersen (1980) indicate a less pro-
nounced effect, so that 12 mg/cm2 results in only 56% reduction in PAR.
Presumably, the epiphytic material in the latter study contained a greater
percentage of nonabsorbing inorganic particulars.  Finally, the extensive
relationships reported by Sand-Jensen and S0ndergaard (1981) for 4 Danish
lakes appear to be very similar tc those observed in the present study. Un-
fortunately, each study has reported the results in different units, so that
these comparisons must be considered tentative.


                         Epiphyte Effects on SAV

     The development of epiphytic communities on SAV leaves can reduce vascular
plant production through several mechanisms other than PAR attenuation, in-
cluding the reduction of diffusive transport of N, P and C02. Furthermore,
the relationship between light reduction and loss of photosynthesis  is non-
linear except at PAR levels below I|< (intersection of initial slope  of
photosynthesis-irradiance curve with maximum photosynthesis value).  Thus, it
is important to examine directly the effects of epiphytic material on SAV
photosynthesis.

     We tested this relationship between epiphytes and SAV Pa twice, once^n
July usirg 02 production as a measure of Pa, and once in September using   C
incorporation. Significant relationships between-reduction in Pa and epiphyte
biomass were obtained (Fig. 26) in both cases (r  = 0.84 and 0.69 for Oo
and 14C, respectively). However, the slopes were markedly different, being
three times greater in the earlier study. This may be the result of  different
epiphytic communities during the two experiments, where the September epiphytes
were less detrimental per unit dry weight. These experiments were conducted
using natural light incubations, and the slightly higher ambient PAR in Sep-
tember could partially explain this difference in slopes. We can compare Figs.
9 and 27 by utilizing the relationship between leaf area and plant dry wt. for
_P. perfoliatus in these ponds (4.7 x 103 cm2/g, Fig. 27). If there was a
one-to-one relation between light and SAV Pa we would expect about a 90% loss
of Pa associated with an increase from 1.0 to 3.0 g epiphyte/g SAV.  In fact,
the September experiment ^..-icts a 27% reduction in Pa, while the July data
indicate a 67% loss of Pa. Thus, the relation to light was not linear for
either experiment.

     Other investigators have considered the effects of epiphytic coloniza-
tion on seagrass Pa, but only at a single level of epiphytes. Sand-Jensen
(1977) showed that from darkness to saturating PAR levels epiphyte colonized
Zostera marina exhibited lower Pa than clean plants. The % reduction was
greatest however (ca. 50%) through the initial linear portion of the curve. He
reported inhibition due to epiphytes even at PAR of about 1200 yEln  m~2s-1.
Yet, even with an 80% reduction of incident PAR, remaining light would be near
saturation. Sand-Jensen (1977) also showed a reduction in C02 uptake by _Z.
marina due to epiphytes except at the highest concentrations of inorganic
carbon, presumably capable of saturating plant demand. Penhale and Smith (1977)

                                    VII1-45

-------
    150
 E
 o
 a?
 2 100
 H
 O

 Q
 O
 (T
 0.
 Z
 Id
 ac
 <
 0.
 Q.
50
                 EPIPHYTE  EFFECTS  ON  SAV
                       0_- Production
                           (July)

                        Y-205.5-51.34 (X)

                            (r2«0.84
                                 ° Ti p,  C-lncorporotlon

                                 • BoM,l4C-lncorporatlon

                                 o Ti p, 02 Production
                                 • Bast, 02 Production
                                             o\
                                                      \
14
       C-lncorporation
           (Sipt.)
       Y-I2I.O-I4.57(X)

           (rz-0.69)
                 12345

             EPIPHYTE  BIOMASS.gdw/gdw  (SAV)


Figure 26.  Effects of epiphyte biomass  [gdw/gdw(SAV)l  on the normal apparent
          production of Potamogeton perfollatus (% of production  of plants
          cleared of epiphytic material) based on experiments using ^C-
          incorporation and dissolved  oxygen evolution to measure plant
          metabolism.
                             VI11-46

-------
 cn
 M  5
 E
 u
 V)
 CO
 < 3
 5
 O
 00 2
 Ld
 H-
 > I
 CL-
 E
 UJ
      EPIPHYTE   BIOMASS
PER UNIT  WEIGHT  VS. AREA
                 Y=2.60(X)-0.27
      0               I                2
      EPIPHYTE  BIOMASS,gdw/gdw(SAV)
Figure 27.  Effects of epiphyte biomass on Potamogeton perfoliatus
         expressed as area (mg/cm2) and weight [gdw/gdw(SAV)j.
                    VIII-47

-------
1
IS
               conducted  similar experiments, but found little epiphyte effect except above
               800  pEin m-2s-l, suggesting that the reduction of diffusive uptake of
               minerals (where demand was greatest at high PAR) may have been more important
               than  light attenuation.

                    In a  recent paper, Sand-Jensen and S0ndergaard (1981) have argued that
               light attenuation associated with epiphytic growth exerts a far greater impact
               of SAV than that due to phytoplankton in Danish lakes subjected to different
               degrees of fertilization. This is probably the case for these lakes where
               phytoplankton  pigments are < 10 pg/fc and watercolumn attenuation coeffi-
               cients are generally < 1.0/m.  In the shallow waters of upper Chesapeake Bay,
               however, attenuation coefficients of 1-4/m are not uncommon, and chl a typically
               ranges from 5-30 pg/4 (Boynton et al. 1981). Most of the attenuation TTere
               is probably due to suspended sediments, although plankton are important.

                    In our nutrient enrichment experiments, both plankton and epiphytes con-
               tributed substantially to PAR attenuation, particularly in the high nutrient
               dosed ponds.  For example, with a typical  summer PAR of 1200 uEin nr^s-l
               incident at the water surface and a water column K = 3.0/m, there would be 60,
               270  and 570 uEin m-2s-l available at the sediment surface, at 0.5 m, and
               at 0.25 m depth. Typical values of compensation light levels (Ic) are about
               50-75 pEin nr^s-l, so that net production is not possible at the sediment
               surface. Given a typical value of epiphytic biomass of 4.0 gdw/gdw (SAV), we
               would find 30  and 10 pEin m-2$-l available for leaves at 0.25 m and 0.5
               m, respectively, below the water surface. Thus, unless these vascular plants
               can  adapt  by pigment and enzymatic shifts to such rigorous light regimes, they
               cannot exist under the combined stress of turbidity plus epiphytic material.
                                  SAV Biomass Response to Nutrient  Enrichment

                     Nutrient loading rates >^ 60 uM N/wk and 6 uM P/wk during  the growing
                season caused a significant decline in SAV biomass  in our pond mesocosms.
                Negative biomass responses were not observed until  60 days following  the  initial
                fertilization and can be attributed to (among other things)  decreased light
                availability in response to increased phytoplankton and epiphyte  growth.   The
                results of this study agree with observations of other fertilization  experiments
                using ponds, and together these studies show the response of SAV  to a gradient
                of N and P loading rates (Fig. 28). All of these studies used  710 d fetilization
                frequencies for periods ranging from 6-23 wk.  Our  study is  the only  nutrient
                enrichment experiment reported here on estuarine plant communities.

                     Loading rates up to 30 pM N/wk and 3.0 pM P/wk above ambient nutrient
                conditions resulted in no apparent effect on SAV biomass (Fig. 28).   At 50 pM
                N/wk and 2.2 pM P/wk Moss (1976) observed lower biomass levels compared to
                low nutrient and control treatments.  However, at slightly higher N and P
                loading rates, Howard-Williams (1981) showed no odverse effects of nutrients
                to Potamogeton pectinatus. Above 60 pM N/wk and 6 pM P/wk, SAV consistently
                exhibited declines in biomass compared to control treatments (Fig. 28). At
                300 pM N/wk and 16 pM P/wk, SAV biomass was completely eliminated in  5 of
                6 experimental ponds (Mulligan et al. 1976). Investigators have attributed
                differences observed among these studies to species tolerance  to  nutrient
                enrichment (Mulligan and Baranowski 1969, Ryan et al. 1972,  Moss  1976). A

                                                    VIII-48

-------
                                            JI
   12.5
o
   10.0
o
-I  7.5

CO

cr
    5.0
a.
CO
o
                                                  X4
            NUTRIENT LOADING  EFFECTS

               ON  MACROPHYTE  BIOMASS

                                         -I
           o No Effect
           + Biomass Increased
           - Biomass Decreased
           x Biomass Eliminated
                              -2
                        -1
                -
I. THIS STUDY
2. Ryan et al.1972

3. Moss 1976
4. Mulligan et al.1976
5. Howard-Williams 1981
        45
        00
           3 05
                                           Y/
      v        30      60      90      120"   360

                NITROGEN  LOADING,/jM/wk


 Figure 28.  The response of submersed aquatic vegetation in experimental ponds to
         various loading rates of nitrogen and phosphorus.
                          VIII-49

-------
•n
 •  "            range of factors may influence these apparent taxonomic differences;  including
 *              nutrient absorption capacity,  adaptations  to shading,  and  high  initial  biomass
               conditions which may have ameliorated nutrient inputs  (Howard-Williams  1981).
               Thus, all of these studies show remarkable similarity  in the  response of SAV
               communities to increased nutrient  loadings.


                           Response of Community  Metabolism to Fertilization

                    Nutrient enrichment of pond ecosystems  is a standard  practice  for  increasing
               fish yields by stimulating primary production.   In this  study,  medium nutrient
               enrichments of 60 uM N/wk and  6 pM P/wk  increased net  community production
               over controls; however at doubled  fertilization rates  net  community production
               declined. Hephner (1962) observed  a similar  response of  fertilized  ponds  in
               which modest nutrient additions resulted in  a total production  rate of  462.0
               mg Om-2.h-l compared to 384.5 mg  C-m-2.h-l  at  higher
               fertilization. Although surface photosynthesis  were greater at  high nutrients,
               productivity at lower depths was inhibited from shading  by surface  plankton.
               In our study, high nutrient enrichmnet resulted in a decline  in SAV biomass
               and a subsequent reduction of  total community net production. This  decrease
               in SAV corresponded to increases in both phytoplankton and epiphytes.   However,
               since total community production decreased at high dosage, the  increased  phyto-
               plankton and epiphytic growth  was  insufficient  to compensate  for the  loss of
               vascular plants.

                    The role of SAV beds as habitat and indirect food sources  for  many sport
               and commercial fisheries has been  well established for Chesapeake Bay (Boynton
               1982).  The relative balance between production and respiration (P:R) for
               these SAV communities provides an  index  of their ability to fill this role.
               In our study, low nutrient levels  resulted in increased  P:R ratios  from 1.8-2.8;
               however, as nutrient levels increased, community P:R ratio decreased  to <1.0
               (Fig. 29).  Under these eutrophic  conditions, SAV communities no longer con-
               tribute net energy to estuarine food webs  .  The response of these experimental
               estuarine systems to elevated  nutrient levels followed the general  push-pull
               patterns discussed by Odum et  al.   (1979)  for ecosystems under  stress.


                                 SAV Strategy for Growth  and Production

                    Net photosynthesis (Pa) of Potamogeton  and Ruppia were distinctly  differ-
               ent in both vertical (along length of the  plant) and diurnal  patterns.   These
               diurnal trends suggest that _P. perfoliatus is more responsive to light  changes
               than is R. maritima.  Nonetheless, the total apparent  production of both  plants
               was abouT 25 mg (^•gdw~l»d">l.  Respiration  was relatively constant
               throughout the day for both plants, indicating that production  and  respiration
               may be coupled more closely in R.  maritima.

                    Maximum Pa along the length of Potamogeton was observed  in the tip of
   I            tne plant, whereas Pa in Ruppia was nearly constant along  its entire  length
               (Fig. 16).  A gradient or epiphyte biomass with lower  concentrations  at the
               tip has been observed in the field for mature Potamogeton  (Staver et  al.
               1981) and corroborated with data from the  ponds. These  gradients suggest that
               under ambient conditions, the  tip  of Potamogeton must  be paramount to the

                                                   VI11-50

-------
 0."
     3
 o
 CD
 UJ
             METABOLIC  RATIO

          FOR  POND  COMMUNIT IES
               30       60      90     120

          NITROGEN  LOADING, >uM/wk
Figure 29. The apparent production to nighttime respiration ratio
         of ponds at four nutrient treatments.
                      VIII-51

-------
energy economy of the entire plant, tui ticularly when demands for belowground
production and flowering are included. When this terminal  section of Potamogeton
is shaded either by epiphytic or suspended material, the balance between pro-
duction and consumption of energy becomes negative as observed experimentally
in this study (Fig. 17).  As the temporally and spatially  integrated energy
balance for these vascular plants approaches zero, a shift in reproductive
strategy may become evident.

     Potamogeton perfoliatus reproduces sexually by means  of seed production
during the summer, and asexually by formation of buds on the apex of rhizomes
at the growing season end. As observed for P. crispus, the advent of a new
growing season for this genus seems to be dominated by asexually produced
propagules rather than seeds (Rogers and Breen 1980).  The starch con'ent of
these structures enables plants to grow under adverse environments wh ;h may
exist early in the growing season (Rogers and Breen 1980). Fixed carton must
be translocated from sites of production to storage for this propagation scheme
to be successfull (HarrUon 1978).  Under high nutrient conditions, the shading
of the Potamogeton shoot tips may decrease the net energy  available for bud
formation In a survey of belowground vegetation in^.our^ experimental ponds
during November, buds were found in all  ponds except one of"tHose receiving
high nutrient treatment, which had no belowground vegetation.  During the
subsequent growing season, no vascular plants were observed in this pond.
Further studies are needed to investigate factors affecting net production in
these plants, particularly as related to respiration patterns, reproductive
structures and germination success.


                                 SUMMARY


     Eight experimental ponds containing submerged aquatic vegetation (SAV)
were subjected in duplicate to 4 levels (including control) of fertilization
from June to August 1981. Nutrients were added on a weekly ccycle with highest
loadings corresponding to those typical  for brackish environments in the upper
Chesapeake Bay receiving agricultural  runoff. Nutrients were removed to am-
bient levels within 1-3 d following treatment except in the high dose systems
after the fourth application, when NO: levels persisted well  above controls
for the entire period. Vascular plant tissue nutrients responded to treatment
with about 3% N at high dosage and 1.2% for controls. It appears that sediment
denitrification also followed treatment level; however, there was no evidence
for any influence of SAV on this process.  Seston and, in  particular, phyto-
planton chlorophyll ^ increased with fertilization, and pronounced blooms were
evident under high dosage. Of the total  seston, phytoplankton exerted the
greatest influence on attenuation of photoysnthetically active radiation
(PAR), so that there was insufficient light for SAV growth at the sediment
surface during blooms.  An extensive epiphytic community developed on plants
in all nutrienttreted ponds at densities similar to those  observed in nature
on senescent plants. Atter.uation of PAR by these epiphytes was considerable,
with >90% absorption and reflection of incident radiation  resulting from typical
epiphytic densities in treated systems. However, without any PAR attenuance in
the overlying water, such epiphytic growth would be Insufficient to reduce
light below photosynthetic compensation levels, and net SAV growth could occur.
Thus, PAR attenuation by both seston and epiphytes is necessary to effect

                                    VIII-52

-------
plant mortality. In fact, SAV biomass decreased significantly in high treatment
compared to controls within 60 d following initiation of the experiment. More-
over, the integrated primary production of these entire pond communities was
significantly reduced with the loss of SAV, eventhough phytoplankton and epi-
phytic growth were enhanced. This premature loss of SAV may only amount to a
shortening of the growing season rather than a population decimation unless
plant reproduction is also detrimentally affected.  Some preliminary evidence
suggests that the inability of these stressed plants to maintain a balance
between photosynthesis and respiration may prevent sufficient translocation
of energy to below-ground overwintering structures. It is hypothesized that
without this vegetative energy storage, spring budding may not be possible.
                                    VIII-53

-------
                             LITERATURE CITED
Abbott, W. 1969. Nutrient studies in hyperfertilized estuarine ecosystems.
     I. Phosphorus studies. In: S.H. Jenkins (ed.), Advances in water pollu-
     tion research. Pergamon~Tress, New York.

Abbott, W. 1967. Microcosm studies on estuarine waters. II. The effects of
     single doses of nitrate and phosphate. J.W.P.C.F. 39:113-122.

Adams,  Franklin, S., Herbert Cole, Jr., and Lowny B. Massie. 1973. Element
     constitution of selected aquatic vascular plants from Pennsylvania:
     submersed and floating leaved species and rooted emergent species.
     Environ. Pollut. 5:117-147.

Adams,  Franklin S., David R. MacKenzie, Herbert Cole, Jr., and Marilyn W.
     Price. 1971. The influence of nutrient pollution levels upon element
     constitution and morphology of El odea canadensis Lien, in Michx.
     Enviorn. Pollut. 1:285-298.

Bannister, T.T. 1974. A general theory of steady  state phytoplankton
     growth in a nutrient saturated mixed layer.  Limnol. Oceanogr. 19:13-30.

Bayley, S., V.D. Stotts, P.F. Springer, and J. Steenis. 1978. Changes in
     submerged aquatic macrophyte populations at  the head of Chesapeake
     Bay, 1958-1975. Estuaries 1:73-84.

Best, E.P.H. 1977. Seasonal changes in mineral and organic components of
     Ceratophyllum demersum and El odea canadensis. Aquatic Bot. 3:337-348.

Best, E.P.H. 1980. Effects of nitrogen on the growth and nitrogenous com-
     pounds of Ceratophyllum demersum. Aquatic Bot. 8:197-206.

Billen, 6. *978. A budget of nitrogen recycling 1n North Sea sediments off
     the Belgian coast. Estuarine Coastal Mar. Sci. 7:127-146.

Borum,  J. and S. Wiurn-Anderson. 1980. Biomass and production of epiphytes
     on eel grass (Zostera marina L.) in the 0resund, Denmark. Ophelia
     (Suppl.) 1:57^§4T

Boyd, C.E. 1971. The limnological role of aquatic macrophytes and their
     relationship to reservoir management. Res. Fish, and Limnol. 8:153-166.

Boyd, C.E. 1970. Vascular aquatic plants for mineral nutrient removal from
     polluted waters, Economic Botany, Vol. 24:95-103.
                                 VIII-54

-------
Boyd, C.E. and E. Scarsbrook. 1975. Influence of nutrient additions and
     initial  density of plants on production of water-hyacinth Eichhornia
     crassipes. Aquatic Botany 1:253-261.

Boynton, W.R. and K.L. Heck. 1981. Ecological role and value of submerged
     macrophyte communities: A scientific summary. Univ. of Md. Center for
     Environmental  and Estuarine Studies, Box 38, Solomons, Md.
     Ref. No, 81-194CBL.

Boynton, W.R., W.M. Kemp  and C.W. Keefe. 1982. A comparative analysis of
     nutrients and other  factors influencing estuarine phytoplankton
     production, pp. 69-90.  In: V.S.  Kennedy (ed.), Estuarine Compari-
     sons, Academic Press, N.77

Boynton, W.R. et al. 1981.  Influence of submerged macrophyte communities on
     turbidity and sedimentation in littoral zones of northern Chesapeake
     Bay, pp. 842-878. In: W.M. Kemp,  W.R. Boynton, J.C. Stevenson and
     J.C. Means (eds.),""Submerged aquatic vegetation in Chesapeake Bay.
     Univ. of Md., Center for Environmental and Estuarine Studies,
     Contribution No. 81-28-HPEL.

Bremner, J.M. and A.M. Blackmer. 1979. Effects of acetylene and soil water
     content on emission  of nitrons oxide from soils. Nature 280:380-381.

Cambridge, M.L. 1979. Cockburn ^ound study technical report on seagrass.
     Dept. Conservation and Environment, Western Australia. Rep. No. 7.

Cattaneo, A. and J. Kalff. 1980. The relative contribution of aquatic
     macrophytes and their epiphytes to the production of macrophyte beds.
     Limnol. Oceanogr. 25:280-289.

Cattaneo, A. and J. Kalff. 1979. Primary production of algae growing on
     natural  and artificial aquatic plants: A study of interactions between
     epiphytes and their substrate. Limnol. Oceanogr. 24:1031-1037.

Copeland, B.J. and W.R. Duffer. 1964.  Use of a clear plastic dome to
     measure gaseous diffusion rates in natural waters. Limnol. Oceanogr.
     9:494-499.

Dale, H.M. and G.E. Miller. 1978. Changes 1n the aquatic macrophyte flora of
     Whitewater Lake near Sudbury, Ontario from 1947 to 1977. Canad. Field-
     Naturalist 92:264-270.

Den Hartog, C. and  P.J.6. Poldcrman.  1975. Changes in seagrasses popula-
     tions of the Dutch Waddenzee. Aquatic Botany 1:141-147.

Dunstan, W.M. and D.W. Menzel. 1971. Continuous cultures of natural popula-
     tions of phytoplankton in dilute, treated sewage effluent. Limnol.
     Oceanogr. 16:623-632.

Edmondson, W.T. and Y.H.  Edmondson. 1947. Measurements of production in
     fertilized salt water. J. Mar. Res. 6:228-245.
                                 VIII-55

-------
!!'
Edmondson, W.T. 1955. Factors affecting productivity in fertilized salt
     water. Deep-Sea Res. 3:451-464.

Fitzgerald, George P. 1969. Field and laboratory evaluations of bioassays
     for nitrogen and phosphorus with algae and aquatic weeds.  Limnol. and
     Oceanogr.  206-212.

Gamble, J.C., O.M. Davies, and J.H. Steele. 1977. Loch Ewe bag experiment,
     1974. Bulletin of Marine Science 27:146-175.

Gerloff, G.C. and P.H. Krombholz. 1966. Tissue analysis as a measure of
     nutrient availability for the growth of angiosperms aquatic plants.
     Limnol. Oceanogr. 11:529-537.

Gossett, D.R. and W.E. Norris, Jr. 1971. Relationship between nutrient
     a/ailability and content of nitrogen and phosphorus in tissues of the
     aquatic macropnyte, Eichornia crassipes (Mart.) Solms. Hydrobiologia,
     38(1): 15-28.

Hall, D.J., W.E. Cooper, and E.E. Werner. 1970. An experimental approach
     to the production dynamics and structure of freshwater animal  com-
     munities. Limnol. Oceanogr. 15:839-928.

Harlin, M.M. and B. Thorne-Miller. 1981. Nutrient enrichment of seagrass
     beds in a Rhode Island coastal lagoon. Marine Biology 65:221-229.
Harrison, P.G. 1978. Patterns of uptake and translocatlon of lC by
     Zostera americana den Hartog in the laboratory. Aquatic Botany
     5:93-97.

Heinle, D.R., C.F. D'Elia, J.L. Taft, J.S. Wilson, M. Cole-Jones, A.B.
     Caplins, and L.E. Cronin.  1980.  Historical  review of water quality
     and climatic data from Chesapeake Bay with emphasis on effects of
     enrichment.  Univ. MO Center for Environ. Estuarlne Studies.  Ref.
     No. 80-15-CBL, Solomons, MD.

HenMksen, K., J.I. Hansen and T.H. Blackburn. 1981. Rates of nitrifica-
     tion, distribution of nitrifying bacteria, and nitrate fluxes 1n
     different types of sediment from Danish waters. Marine Biology
     61:299-304.

Hepher, B. 1962. Primary production in fishponds and Its application to
     fertilization experiments. Limnology and Oceanography 7:131-1.36.

Hlckman, M. 1971. The standing crop and primary productivity of the
     eplphyton attached to Equlsetum fluviatile L. in priddy pool.
     North Somerset. Br. Phycol. J. 6:51-59.

Hiscox, J.D. and G.F. Israelstam. 1979. A method for the extraction of
     chlorophyll from leaf tissue without maceration. Can. J. Bot.
     57:1332-1334.
                                 VIII-56

-------
ri
                Howard-Williams,  C.  1981.  Studies  on  the  ability of  a Potamogeton
                    pectinatus community  to  remove dissolved nitrogen and phosphorus
                    compounds  from  lake water. Journal of  Applied Ecology 18:619-637.

                Hunter,  B.L. and  E.A.  Laws. 1981.  ATP and chlorophyll £ as estimators
                    of  phytoplankton  carbon  biomass.  Limnol. Qceanogr. 26:944-956.

                lizumi,  H.  and  A.  Hattori.  1982.  Growth and organic production of eel grass
                    (Zostera marina L.) in temperate waters of the  Pacific Coast of
                    Japan.  IfHTRe  kinetics of  nitrogen  uptake.   Aquatic Botany 12:245-256.

                lizumi,  H.,  A.  Hattori, and C.P. McRoy. 1980. Nitrate and nitrite in
                    interestitial waters  of  eelgrass beds  in relation to the rhizosphere.
                    J.  Exp. Mar.  Biol. Ecol. 47:191-201.

                Jupp,  B.P.  and  D.H.N.  Spence.  1977.   Limitations of macrophytes in a
                    eutrophic  lake, Loch  Leven.   II.  Wave action,  sediments, and waterfowl
                    grazing.   Journal  of  Ecology  65:431-446.

                Jupp,  B.P.,  and D.H.N.  Spence.  1977.  Limitations on macrophytes in a
                    eutrophic  lake, Loch  Leven,  I.   Effects of phytoplankton.  Journal of
                    Ecology 65:175-186.

                Jupp,  B.P.,  and D.H.N.  Spence. 1977.  Limitations on  macrophytes in a
                    eutrophic  lake, Loch  Leven,  I. Effects of phytoplankton. Journal of
                    Ecology 65:175-186.

                Kaumeyer,  K. et al.  1981.  Nutrient dynamics cf submerged macrophyte com-
                    munities in  northern  Chesapeake  Bay, pp. 685-802.  In; W.M. Kemp,
                    W.R.  Boynton, J.C. Stevenson  and J.C.  Means (eds.), Submerged aquatic
                    vegetation in Chesapeake Bay. Univ.  of Md., Center for Environmental
                    and Estuarine Studies, Contribution  No. 81-28-HPEL.

                Kemp,  W.M.  et al.  1981. Experimental  observations of turbidity/light
                    relations  and their influence on submerged macrophytes in northern
                    Chesapeake Bay, pp. 2-29. In: W.M. Kemp, W.R. Boynton, J.C. Stevenson,
                    J.C.  Means (eds.), Submerged  aquatic vegetation in Chesapeake Bay.
                    Univ.  of Md., Center  for Environmental and Estuarine Studies,
                    Contribution  No.  81-28-HPEL.

                Koike,  I.  and A.  Hattori.  1979. Estimates of denitrification in sediments
                    of  the Bering Sea  shelf. Deep-Sea Research 26:409-415.

                Lewis, M.R., W.M.  Kemp, J.J.  Cunningham,  and J.C. Stevenson. 1982. A rapid
                    technique  for preparation of  aquatic macrophyte samples for measuring
                    l^C incorporation. Aquatic Botany 13:203-207.

                Lind,  C.T.  andd G. Cottam. 1969. The  submersed aquatics of University Bay:
                    A study in eutrophication. Amer. Midi. Nat. 81:353-369.

                McCarthy,  J.J., W.R. Taylor,  and J.L. Taft.  1977.   Nitrogenous nutrition
                    of  the  plankton in the Chesapeake Bay.  I.  Nutrient availability and
                    phytoplankton preferences.   Limnol.  Oceanogr. 22:996-1011.

                                                VII I-57

-------
McNabb, C.D. 1976. The potential of submersed vascular plants for reclama-
     tion of wastewater. In: J. Tourbier and R.V. Pearson (eds.), Biological
     Control of Water PolTTcion. The University Press, Philadelphia.

Menzie, C.A. 1979. Growth of th? aquatic pl»nt Myriophyllum spicatum in
     a littoral area of tne Hudson River estuary. Aquat. Bot. 6:365-375.

Mihursky, J.A. and W.R. Boynton. 1979. Review of the Patuxent Estuary data
     base. Chesapeake Biological Laboratory, Solomons, Maryland. University
     of Maryland, Center for Environmental  and Estuarine Studies. Ref. No.
     78-157-CBL.

Moss, B. 1976. The effects of fertilization and fish on community structure
     and biomass of aquatic macrophytes and epiphytic algal  populations:
     an ecosystem experiment. Journal of Ecology 64:313:342.

Mulligan, H.F. and A. Baranowski. 1969. Growth of phytoplankton and
     vascular aquatic plants at different nutrient levels. Verh. Internat.
     Verein. Limnol. 17:802-810.

Mulligan, H.F., A. Baranowski and R. Johnson. 1976.  Nitrogen and phosphorus
     fertilization of aquatic vascular plants and algae in replicated
     ponds. I.  Initial response to fertilization. Hydrobiologia 48:109-116.

Nichols, D.S. and D.R. Keeney. 1976. Nitrogen nutrition of Myriophyllum
     spicatum: uptake and translocation of "N by shoots and roots.
     Freshwater Biol. 6:145-154.

Nishio, T., I. Koike, and A. Hattori. 1982. Denitrificationn, nitrate
     reduction, and oxygen consumption in coastal and estuarine sediments.
     Applied and Environmental Microbiology 43:648-653.

O'Brien, W.J. and F. DeNoyelles, Jr. 1974.  Relationship between nutrient
     concentration, phytoplankton density,  and zooplankton density in
     nutrient enriched experimental ponds.  Hydrobiologla 44:105-125.

Odum, E.P., J.T. Finn, and E.H. Franz. 1979. Perturbation theory and the
     subsidy-stress gradient. Biosdence 29:349-352.

Odum, H.T. and C.M. Hoskin. 1958. Comparative studies on the metabolism
     of marine waters. Publ. Inst. Mar. Sci. Univ. Tex. 6:159-170.

Oremland, R.S. and B.F. Taylor. 1977. Diurnal fluctuations of Og, N£ and
     CH4 1n the rhizosphere of Thallassia testudinum. Limnology and
     Oceanography 22:566-570.

Oren, A. and T.H. Blackburn. 1979. Estimation of sediment devitrification
     rates at in situ nitrate concentrations. Applied and Environmental
     Mi croblology~37:174-176.

Orth, R.J. 197/. Effect of nutrient enrichment on growth of the eelgrass
     Zostera marina in the Chesapeake Bay, Virginia, USA. Marine Biology
     44:187-154T

                                 VIII-58

-------
Orth, R.J. and K.A. Moore. 1981. Submerged aquatic vegetation of the
     Chesapeake Bay: Past, present and future. Trans. 46th North American
     Wildlife and Nat. Res. Conference. Wildlife Management Institute,
     Washington, D.C.

Parsons, T. and M. Takahashi. 1973. Biological oceanographic processes.
     Pergamon Press, New York.

Parsons, T.R., K. Stephens, and J.D.H. Strickland. 1961. On the chemical
     composition of eleven species of marine phytoplankters. J. Fish. Res.
     Bd. Canada 18:1001-1016.

Penhale, P.  1977.  Macrophyte-epiphyte biomas and productivity in an
     eel grass (Zostera marina) community.  J. Ixp. Mar. Biol. Ecol.
     26:211-224.

Penhale, P.A. and W.O. Smith.  1977.  Excretion of dissolved organic carbon
     by eel grass (Zostera marina) and its epiphytes.  Limnol. Oceanogr.
     22:400-407.

Peres, J.M. and J. Picard.  1975.  Causes de la rarefaction et de la
     disparition des herbiers de Posidonia oceanica sur les cotes
     Francaises de la Mediterrane"e^  Aquatic Botany 1:133-139.

Phillips, 6.L., D. Eminson, and B. Moss.  1978.  A mechanism to account
     for macrophyte decline in progressively eutrophicated freshwaters.
     Aquatic Botany 4:103-126.

Rogers, K.H. and C.M. Breen.  1980.  Growth and reproduction of
     Potamogeton crispus In a south African lake.  Journal  of Ecology
     68:561-571.

Ryan, J.B., D.N. Rlemer, and S.J. Toth.  1972.  Effects of fertilization
     on aquatic plants, water, and bottom sediments.  Weed Sci.
     20:482-486.

Sand-Jensen, K.  1977.  Effect of epiphytes on eel grass photosynthesis.
     Aquatic Botany 3:55-63.

Sand-Jensen, K. and M. Sondergaard.  1981.  Phytoplankton epiphyte
     development and their shading effect on submerged macrophytes
     in lakes of different nutrient status.  Int. Revue ges. fydroblol
     66:529-552.

Schindler, D.W.  1978.  Factors regulating phytoplankto.i production and
     standing crop In the world's freshwaters.  Limnol. Oceanogr. 23:478-486.

Scott, B.D.  1C78.  Phytoplankton distribution and light attenuation in
     Port Hacking estuary.  Aust. J. Mar. Freshwater Res. 29:31-44.

Seddon, B.  1972.  Aquatic macrophytes as Hmnologlcal Indicators.   Freshwater
     Biol. 2:107-130.
                                 VIII-59

-------
                                                 •\
Seitzinger, S., S. Nixon,  M.E.Q.  Pilson,  and S.  Burke.  1980.  Denitrifica-
     tion and N20 production in near-shore marine sediments.  Geochimica
     et Cosmochimica Acta  44:1853-1860.

Shoaf, W.T. and B.W. Lium.  1976.   Improved extraction  of chlorophyll  a and
     ^ from algae using Dimethyl  Sulfoxide.  Limnol.  Oceanogr.  21:926-728.

Sinclair, L.R. and R.B. Forbes.  1980.   Nutrient  removal  from drainage
     waters with systems containing aquatic macrophytes.   American Society
     of Agricultural Engineers 23:1189-1194.

Smith, E.V. and H.S. Swingle.   1941.   The use of  fertilizer for controlling
     several  submerged aquatic plants  in  ponds.   Trans. Amer. Fish.  Soc.
     71:94-101.

Solorzano, L 1969. Determination  of ammonia in  natural waters  by  the
     phenolhypochlorite method. Limnol.  Oceanogr. 14:799-801.

Stfrensen, J.   1978.  Capacity for  denitrification and reduction to nitrate
     to ammonia in a coastal  marine sediment. Appl.  Environ. Microbiol.
     35:301-305.

Stevenson, J.C. and N.M. Confer.   1978.   Summary  of available information
     on Chesapeake toy submerged vegetation. Fish and Wildlife Service,
     Office of Biological  Services - 78/66.

Strickland, J.D.h  and T.R. Parsons. 1972. A practical  handbook of seawater
     analysis. Fish. Res.  Bd.  of Canada  Bull. 167 (2nd edition).

Stuckey, R.L. 1978. The decline of lake  plants.  Natural History 87:66-69.

Valderrama, J.C. 1981. The simultaneous  analysis  of total  nitrogen and
     total phosphorus in natural waters.  Marine  Chemistry 10:109-122

Westlake, D.F. 1978. Rapid exchange of oxygen between plant and water.
     Verh. Int. Ver. Theor. Angew. Limnol. 20:2362-2367.

Wheeler, P.A. and W.J. North.  1981. Nitrogen supply,  tissue composition
     and frond growth rates for Macrcrystis Pyr\^ra  off the  coast of
     southern California.  Marine Biology  64:59-69"!

Wolverton, B.C, R.M. Berlow, and R.C.  McDonald.  1975. Application  of
     vascular aqutic plants for pollution removal, energy and food
     production in a biological system.  NASA Technical  Memorandum
     TM-X-72726.
                                 VIII-60

-------
It"
4



.
o
E

3
00

': 01
4=
: 4J
' 60
C


•H
M
3
•o
: 0)
; *O




C
O
B.
r-l
eg
C
1 0)




i



















*






-
Is
Ol
X (0
01 ,H
O
^ 00 C
9 « 8
2-g
« «
~t
s«
^ Ol
§«
3
•H 4J
c e
i "
EJ ^4
CO 4J
3
*w C
o
Ol
CO 01
C M
0 J=
^J ^
0) 01
4J 4J
a
01 »-"
0 0
C «*4
o
CJ -H
oo
. 2
••H
1 144
< 0
I 01
• I-l
I -^
I 9-







§
^

vO


en


Q
m




U
H
^



O
Z

Jj C
O M
^ S
f** ^
W S
^
^3
c




w
0
CO
8



vo en ^ -4 ON -< CM in ~ en -« rx r- ON
O O O ^CM COO O CNr-CM OOO i-l -H
%^ Q\ ^5 PO ^^ ^^ CO
CM

r^ vD *"H vO «^ ON en in *^ r** f^ m *•* CM ON

CMCMO oo 'fr'-i en oo o o CM«* cMin
VO •* —I CM f» CM
—< CM




fM CM CM—. VD>9 i— 1 <-> \O •-> ^Hin CMCM
. . . g .. . ... .. ..
•^oo O z voo -4 rt^o i-io mo
en rn en ••* ^-i en

ON ^^ ^H «^ ^. ^H *^ ^j *^ co ^4 *^ en «^ *^
... * cd •• . ... .. ..
CMOO oz mo m enoo oo -nm
en CM <* ~* -H CM



Opp enin r»CM en vOvo—< ^ft encM
^•^Z oo r^o m «
^•-HO enr>« OCM CM -»•-<•» pp "*P
p-iOi-i o«-i eno CM VOCMO zz mz
CM in CM vo
"^ Q g "^ g ^ ^ "J ^ •* ^ O P *
_,ZZ O~ cJi-J C3 i-> >. >s x >^ x ><>>>> 33
(3CC CC Ci-H iH i-4iHrHlH iHrHi-l 0060
333 33 33 3 3333 333 33
^J Hj ^ ^3 h^ Hj Hj ^ (>-ij H^ H>^ )—j H^ h"J ^ ^(J <(J
vocsivo CNI\O O fn oo vor^^ro oo cs •-< -^ r**
^4^H CMCM CO ^i»H^HfM CMCMCO







OvOO cor** Oco O o*^cOf** O^*ro Of)
.— *






ij
c
i
4J
cd
01 T> JS
b 4J C T3 4-> ^3 jB
4J 00 O M I-l -U 4-1
V (-1 O T-I 3 IM X
>4 -H V -C O -H v4
P< (i, CO H fa fn M






































/MS
y;
3.

^H
.
O
V
01
J3
CD

U
01
4-1
Ol
•u
4J
O
C
n
p
z
*
                                      VIII-61

-------
tn
00
C
•H
^
3
"0
(0
"O
o
Q.
iH
to
4J
u
•iH
^
0
&«•
OJ .H
O
4J Id
_P* J *
4S **
00 C
•H O
01 0
0) T3
J3 C
4-> tO
C 09
•H 4J
C
£8
3. 4J
^ «J
m
0) h
4-) 4J
•H
H 4J
ij e
5 «
C2
§tj
iH C
a
0)
V 01
*j M
a £
M 4J
4J
•SJ8
4J

S
M
X
CM



£ oo
M
2 -tnco
0 ss i^s si ^^d


^gg «« VOCM o «^Q
"° -° s ^a85
ON 00 C"J CM vO ns. «^
. _ *^ >"' W» P** ^H c*> -*\
* • • ^^ *^1 5^ ^K
~°° a- s- R ,«i


^^" tn ^p ^ -H-.-
"°° nB sg si »«§

°§2 2§ "~ "i -^«
S° -1 -Hin85
CM VO CM


»n -*  IK
-< CM tn m
... .
ON 00 CO ON
CO vo CM r-»


CM ^ in ON
... .
ON O 00 CO
r^ m CM m


ON tn » r^
CM CM tn * ^


(
0 ^ co 0 ^ s


p
1
J
1
<
*
(
B
1
5 5
3 5 E
•* w 1


























^

D
•
mS
1»

j
j
J
j
I)
1
i
:


                                    VIII-62

-------
   O
   C.
   0)
   01
   o.
   X
   
-------
it!




(U
J-l
oo
C
t-c
•o
oo
0
a

«
4J
c
•H
(U 09
O« *H
X 0
01 U
u
ti c
•C O
oo -j
•H
ai -o
c
01 e
O 0)
h *H
*J M
•H 4J
C 3
C
t-H
QJ 0)
0 W

4J
<*•(
o at
CO 4J
O h
i-l O

4J 00
e ON
o *o
u

• i
** 1

t CO

0)
-H
•8

r-»

g
S
CM





gco
M
o



vO

5
Q
•-1
cn






•J m
§
g
8-





u
^


o
2
SB
^J 5«
J g
2 S
w H




M
!2

CO
2



-« cn
in r^
*H i— 4


r^ vo
•H VO
 "^ ^
vO CM vO
1-4 I— 1





O VO O



4J
e
§
4J
CO
01
h
Ol
(4
Cu
H
_,
•
m


m r~« m oo
r^ d -H -H
vo vo —i r^



cn -H
S 2

ON ON r-»
m o oo
CM m vo
vo cn
CO ON
-H %0


^1^ l^^ ^^ ^^
• • • •

cn CM CM CM



^D ^O
• •
f^ 4O
CM CM
cn cn CM r*.
ON in ON r^
CM CM CM CM



01 0) 0) >> X
C C C iH 1-4
33 33 3
"1 *"J 1 "^ ^3
CM NO o cn co
CM CM cn





co r^ o cn ^3








•o
u C -a
oo o ^
M O -H
•H 01 JS
fa W H

M ^
• • •
•-< r» m
r» i-i ON
"~ ^

ON 00 ^
^ O> -H
i— 1


CO cn ON
CO •* ON

ON in O
o *n vo
NO ^ cn
CM ON
o «»



O Ov «^
• • •
cn CM co
*^ cn cn



— i co m

cn i-i m
CM CM CO
CM NO O*
ON r^ o
CM CM •»



i-H f-H iH fH i-l r-l i-(
3333 333
•-J "-J >-j •-) <-> •-> <->
vO ^ Ov cn 00 ON ^H
ri rH i-i CM CM CM cn





o ^H cn r^ o *~* cn








J3
u j:
h 4J

O i-l


^H
•
oo
m
•"•

cn
oj
CM


vO
CO
CT>

r-»
^
°I
eo
CM


vO
1*^
sf



^.^
A
CM
CM
vo
vO
CO

4J 4J
09 00
II


-------



01
ff.
4J

00
C
•H
3
•O
CD
•O
C
0
a.
r-l
(8
^J

ol
E
•H
01
& 09
0)  cd
5 2
•M ^j
c
£§
Z3- C
v id
CO t-i
2"
O *J
j: c
B. 0)
go -H
o t-
fC U
CL 3
C
1-1
IB 0)
*» oi
o u

IH
0 0)
0 H
•H O
4J U-l
 1
4: •
d)
iH
•s

^
X
O
M
3B

CM






S 00
M

* "*




vO

o

CO






o ""*
ij^
H
§
0 ~







,o
^j
Q


O
z
i 1

O 3
PM S
C/)
^i "
*(
Q


U
«§
i


\o irt oo r-* cs o vo
C} CD ^D ^D ^3 *^ ^^
»-*



O\ ^H 00 ^^ 00 vO f*"' ^ H -HI-II-IPH rHrHi-i
3S3 33 33 3 3333 333
t—y *•) *^ f-j ^ ^-3 H^ Hj ^ ^ ^ Hj ^ ^ ^

>O CM VO CM O O CO 00 M? r*^ Q\ CO 00 ^ *>*4
^4^-1 CMCM CO ,M^Hi-iCM CMCMCO




O so O co r*^ o co o O ^^ co ^ O ^~* co
^^





4J
c
§
4J
01 TJ JS
V4 4J C *O 4J J2
4J CO O fc M 4J
01 M O -H 3 «-•
h -H 0) jj: O t-i

H
^
Ot




CM
.
CO




vO
.
•*
CO
CM



in
.
~*

in
•
CM




O
^


OS
^




4J 4-1
09 CO
3 3
60 60
3 5
^ ^

^ r—





O CO









4J
X

CO

VIII-65

-------
t
rH
oo
*""*
c.


3

C
"£
3
"°

•o
C
o
Q.
,_
IQ
*J
C
QJ
E
(.
0)
a.
X
0)

4J
cn

0)
c
V)
3

*-
t_
O)
a.
c
0

0)
en

10
0
a.

<*-
o
w
V

VI
JJ

c

^i i^
-g
en

is

+j
c
c
o
o
c
o>
o
t-
E
,
1—4
1
CO

O)

IO






















































*
10
o

c
0
o
•o
c
IO
«/>
4»)
C
1
IO
01
*-
**
c
V

c
co
L.
O













jZ
cn
£














3


*














2P
O














^_
o
*->
c
3























c
IO
£


-»



CM





C
IO
Ql
2?



oo



^.






c
ie
C




kO




ro





c
IO
£?


^





*"*








LU
^c
o







•-^ ^-» «— » ^*.^* ^~.
m c^ o^ """^ in ^^ to c^ f1*1* ^o ^^ ^~^
CM &\ CM n CM O CM ^ rH O CM CO
 ^^ •— «.
S § 1 c)f«? CM 2 5 CM 1
• • • # • » • »
CM CM ^HOCMO <-IO
*— * ~-* *~^


(_ ^ ^_^ ^_^ ^^
in o m •— i r^» ^ o ^r oo
oo in CM r^ CM ao <-* in CM i
CM »-i O "i O «-i O «-" O
^^ ^«^ *— ^ *«l 1*

^11 ^ ^^ ^1 ^ ^fH^,
CA r^ fN« 10 r^ CM CM r^ **^
^•CM to «— i r~ •— i "ft oo CM
t-t O «— I O •— i O •"* OO

^ n, ^" ^ _ir ^ ^"s
in oo "— i inin o**4 too
oo in *"4 r^ co c^ C? in CM i
CM »H O t-l O CMO rH O
«*_^ ^-* ^-^ ^^*



o o o *t c oo to
r^ <»• CM CM CM ^r i-i i i
i— 1 r-l O i— 1 O •— 1 O
^m^ *M^ ^^^f

o m
r^ i i o i i

i-l CM O

^^ ^"^ ^^
«<• CM CM CM ^1 r-l ^- rH tO
rH O rH O rH O rH rH O
**** %h^ «H^ ^-^





^
§(/) O
0) ai >> 3 z a) w
8c c i— en is c c
3333 O3 3
Of ^"3 *^ *^ ^C OH "^ '^
43 tO
2*^ r^»
§i—i
— i
CO '-


•
I/I
•o
0
Q.

%
O
^
a.
c-
n

t~
•o
ai
c
2
E
0

ie

IO
•o
J_
o


eT
"-*
•o
c
IO
c
IO
i

TJ
,2!
Q
O
a.

TJ
C
TJ
C
o
ex
jj
0
IO
0)
c
"^"

a
IQ
^•v
a.
to
i

t.
o

o
C-O
*~ *
•o
•a

V)
^
I
0»
IO

c
0)
>
•r-
o
*
VIII-66

-------
00
I-H
c


3

C
3

in
•o
c
o
Q.
^_
•D
c
CO
u
ft)
Q.
X
CO
en

"co

c
1—
vt
3
,)J
ID
*0
<*-
t_
CO
0.
,-
r
^
0)
O)
o
ID
0
Q.



V)
U
3
in
in
^

c


c
w
en
O

+J

C

C
o
i_

u
n_
o

0
*^»
+J
e?
,
CM
1
CD
CO
f\
IT}
H-



















































*
•
in
*o
c.
ij
c
o
u

•c
c
ID

in

C
i
1o
0)
t_
**
+•>
c
CO

°u
*->
c
CO
,_
o












x:
O1














E
.^
-o

3£














*
J3
















P_
o
{_
^
c

c3
























c
A3
"W

>^










c
, *— ^
vo m ^- ^- ^-i vo
^ ,_! ^H O O •-*

f-« oo o^ m esj co
10 «-H o o o cvj


•••— * .*-«^ j^^»
*-H ?— t r^ t" f— t in
co ^i ^t o o o
»-H v.^ f— 1 *— * rH v-^


>— *, —- *.
r*». vo *~^ m o>
• i • • • •
CO CO i-^ CO «— 4
*— 1 t— 1 
• 1 • • * •
vo co ^-t ro «— i
«— 4 r-t *— ' ^H -*-^-


^•1 ^ ^M^ ^H^
^? CO O^ CO ^* ^-
00 CVJ IT) O ID *-H
^^ ""Li-" V^ *T_'* ^H S^M*


^11^ ^^ ^1^
CM ID Tl- IO O> O
VO CM ID O ID CM


^ 	 ^ <|HMh ^^
O»^H T»- CT» «f O
O O VO O ID I-H
CM *M^ I-H SM^ f-H ^v'




^^.
«a- m
* 0
CO CM
I-H ^M*


<^^
!**• t^»
• •
VO CO
•-H ^.^


CO CO ID VO i— I O
in o o CM en CM
.-H^- CM «— •-I*—'








Q in o
5*0) CO >, 3 ZQJ CO
Sc c i— en 3 c c
3 3 3 3 O3 3
CO C9
uj ^- r^ co co 3!*^ r».
> •-* •-> I-H O i-H
O —1
CD ii i
•* VIII-67 «°

•
in
TJ
o
Q.

*J
0
"ex
(.


L.
"*-
•o
01
c
5
g

ID

•S
o


^~^
o
10
*
•o
c
<0
c
ID
i

•o
£
Q
0
Q.

•o
c
ID
•a
c
o
a.
F-
u
ID
CO

c
°*~
in
^J
c
ID
— —
ex

VO
1
CO

c_
o
it-
s'
IS)
•-^
T3
c
It)

in
c
ID
(U
c_
ID
c
CO
•^.
CD
*

-------
I?
f
c
•^
u
4->
c
Ol
Ol
t_
JC
» *
*
o
co
en
i— i
<*-
o
u

6
1
eu
C7>
c
T
•o
•o
c
o
ex
1o
c
E
°p
w
Q.
X
CU
c
1— 1
i
I/)

c
O>
_w
"4-
IV
o
u
c •
o o
_!__* Q
10 t-
3 4->
c c
O) O
4J U
10 T3
4-> ITS
cn 
Ij c
i
• 4-*
i— i ia
1 O>
0 (.
+J
4)

1




CD
i— <
•^
CM




co
=>
o
UJ




uo

8
ro




rt
.i*
0
ce.
0
CJ i-^


UJ
1—
0

CD
21
*— i
S z
0 ^
u. <
UJ
00 O£
>" f—
«c
o



»!l
CD
4^
 CM
1-4 1-1

CM VD
CM CM
O -H


CM ^~
r-. en
i-l O


OO ID
eno
O •-"

*— ^_
3 3
en co
i-l CM




CO ^^













JC
4J
C
s
u.

0
CM
r^


CO
VO
1— 1


VO
VO


o
o
f— 1

ID
en
o

CO
en
o


r-.
^•»
O


en
o

^—
3
^4
CO




CO














JZ
4->
c
                                                  VIII-68

-------

co
a\
f-H
t~
o

t_
f=
3
V)
O)
-C
4->

CD
C
•»•-
L.
3
•o

trt
T3
C
O
Q.

C
0)
E
^-
»
Ol
^^
X
Ol
4->
^—
0>
i—
CU

cu
JC
4->
C
•
^""* tf>

"o
•^ c_
014.)
c
— • o
,'J
10
i -o
•— c
i— ,
J= t/l
0.4J
o c
<- O)
0 E
JC 10
O o>
t_
o*1
4.)
in c
C Ol
0 •-
•«- t_
4J 4->

C CU
CU 0)
O (.
£Z f
a*1
HI
r-
• 4J
CM
1 t_
«-> 0
- t—
ff
o









UJ
(-
<
8
0

O r~ vo
• • •
•-i in o



•-* o o
• • •
^ O> t— «
f-H


CO CO VO f-.
• • • •
o ^- o o




O CM CM O
• • • •
«-• *-< O *-•




<*• •-< CM in
• • • •
O I-H O O


^ cr> in co
• • • •
t^ o o co


m r*. co in
• • • •
~* 0 0 O
•-• in t-^ oo
• • • •
"-1 O O O



ID
« 1 g>.
3 3 33
^> -3 -00
VO CM O CO
•-H CM CO






O PO O CO
•— l











4->
c
CU
E
4->
<0
cu -o
*- 4J C
<-> I/I O


oo o ^-i *r
• • • •
«— i CM ^r o
in ^-1
r-H CM

P~ If) CT> CM
• • • •
in CM ^ cr>
10


O CM I-H CM
• . . .
t-l i— I CD CM




CO •— I ^ CO
• • • •
O ^H 00 «-H




VO 00 O CM
• • • •
O r- 1 i— 1 .— 1
CM

CO i— 1 O> VO
• • • •
O CM U) CO
CM

CM m *r in
• • • •
O O O CO
m m cr> r-»
• • • •
0 00 -H




>> 222
*= "3*3 "3
"3 T3 ^
oo 10 en co
>— 1 i— 1 CM






o o co t>»
















JS
^ t!
•^ 3
J= 0
t— u.
VIII-69
00
•
CM



CM
•
03



OO
•
^~




VO
•
0




CM
•
in


cr>
•
CO


CM
•
O
00
•
«*•


4->
v>
o>
3
<
t-»







CO

















JC
4J
X
•r-
to









































•
L.
CU
Q.
CU

^
^3

"E
i

•c
c
<->

i.
«
a.
4.1
C

g
CU
c.
3
V)
to
i
c
CU
£^
o

>,
c
5
a


-------
*-
0
J^

C
1
C.
O)
Q.
X
Oi
^
JZ
cr»

OJ

O>
c

**^ •
O co
E —
— 0
CO +J
xi c
•«- O
i— O
o
«/> X)
c
^3 
0> C
a. at
3 +J
tn ID
Ol
^- L.

"o *j
+^ c
Ol
"^* "^"
o c.

l/> 3
C C
0
i- 0)
*J 01
ID t-
(. JC
c
Ol Ol
O -C
C 4->

o
UJ
^





L

^
3

no



in
^j
o
ec
•yp
0
O rH









UJ
(-
O



CD
21

S z
_j gr
0 t-
U. <
UJ
10 Q£

«c
a







UJ

^c


o



^ VO O
r-! o^ r~!
t-H t-H t-H

^» ^r CM
f^ 00 vo
^H t"H t—*


if f Ol
• • •
VO O^ vc*
r- < t— 1 «— 1


CM ^- CO
ID VO >*
t-H t-H t-H



VO «S- If)
If) f>» VO
t-H t— 1 t— 1

vo *r in
• • •
*r to r^
t-H t-H t—*
vo <• «*•
• • •
co ^r ^~
i— 1 1— 1 1— i


CM CO CO
o% cr^ r*^
l-t t-H t— t





ai ai ai
c c c

"O ^ -3
VO CM VO
t— 1 f-H






O VO O
1-1










*J
I

10
01

+j
Ol
L.
O.

CM O CO «»•
t-l 00 CM CM
CO t-l CM CO

O ^H VO «»
r^ oo r^ I-H
CO -H r-> CM


VO i-H CM t-H
• • • •
* *»• r>- O
CM t-H ,_| t-H


t— i * r- cr>
O vo «»• O
CM t— 1 rH ^~



rv. «»• o 0
CM r^ r~. CM
CM •-> t-l ^

rs. rv vo oo

00 CO ^H CM
i— I t— 1 t— I CM
VO VO C^ GO
• • • •
CO CO rH *f
CM fH f-l t-H


vo a> vo *
•«r r«» vo «M
CM t-l r-t CM





ai O) >> >> >>
C C f— r- t—
3 3 3 33
"3 ~s rs ^ ^3
CM o oo en co
CM CO t-l CM






co O o co r>-















XI JC
4-> C X) 4->
 0 l_ L.
«- 0 -^ 3
•^ ai .c o
u. co F- u.

C7<
t-H
VNJ

CO
VO



^.
9
1^
^


CM
CM
CO



VO
VO
CO

CM
t
^>
CO
CO
•
en
CM


VO
VO
CO



4J.
1/1
3
ot>
3

1^







CO
















^
^>
X

CO
VIII-70

-------


o
CU
+->
QJ
•T)
J3
3
in
•p*
W
c
o
Q.
IO
C
01
E
t_
01
O-
X
Ol
c


*3
c
IO
"a.
o

^J
IO
cr
ID
TJ
Cn
U
1
t_
o
CM
E
^j
X
•o
^

a*

J
4J
^
Ol


V)
10
0
5 •
VI

a 3
c
01

r\
IO
I-

























c
E

IO
a<


c
Ol
•r-
t













































C.
cn

r












E
3

5












3










0
^J
c
s



















• •
0.

ie
0
•*=
IO



r-^




CM


3^.


QJ
£




oo



,-T.







^n

*


PO


*
IO

£


^j




,_,



0)
3
IO
^*




tft
ai

"G
O)

CO



CM i-H CM
oo vo co
r-. to

oo o
en eo — '
Tf CO f-H
^— -

to rv oo
CM VO LO
o to
f-H •*-*

,_^
CM LO 00
10 VO f-*
CO i-H LO
*-*


^^
1
in * — o
«*•

CM LO CM
CO VO CM
CO I-H IO
^.^



TC o" to
en •— < r*.
CM CO ^t-
VO ^f 00
-5. s
^
CM CO CM
r~- f-H CM
TJ- Tt" CO

r~ co vo
r~ Tf r»
fH CM O


VO 00
to TJ- cn
i-H LO


o en CM
r-. en to
CM CM LO
^^


1 0* 1
CO
%•-•

c
o

Ol
cn
o  CX
3 O 3


Tf
LO <*• •— 1 VO
p^ to co o
f-H 00 IO CM

en o eo" o
ooo «r
CM to CO LO


co co r>. CM
«r eo to t^
-< o vo co

*—. *—*
TI- o oo to
oo r^ t*^ oo
•-H oo o r«.
f-H i-H


*-*• O ^^ CM
1 • 1 •
«~-- to > — r~
"* 2

^" *tao
oo to r^. o
•-H cn o *r
i-H f-H
^-> ^^


vo LO co cn
CM 0 *»• TJ-
r^ oo to *r
I-H (*» ^ 00
i-H O LO O"»
r^ 00 to  en
^t- o CM o
P** OO ^O TJ"

co cn Tj~ co
«*• O CM i-H
CM CM IO LO


co vo en to
CO LO CM O
r- to to


^ r~ o LO
00 CO CM fH
r» oo to O



O i a i
to co
^H^ ^^

§
4_>
01
cn
o
•• E
i— 0) W
IO C 4->
4-> 3 O
O T Q.
t—

f— 1
en oo co ^r to I-H
co I-H TJ* CM Tt* cn
vj* ^-^ CM Tj* VO
*^ i-l v — rH

co r^ vo r^ co CM
 CM LO CM
CO • — CM 00 CO

to LO CM to cn
to CO O CO OO CM
^r to r-« cn vo oc
«h_^ ^^ ^— ^



vo" i^ TT" oo r^ oo
oo TJ- I-H oo Tj" r**
T«- "*• VO 00 «* O
00 CM f-H VO 00 LO
CO i-H CO CM V£) VO
LO CM TT 1^. T» O
^ ^ ^
o CM cn f o CM
Tt- co CM CM ?: er>
TC to r^ o TT o

no cn Tt* cn vo vo
i-H LO CM VO TT VO
LO P~ CO CM O VO
^-- I-H f-H CM I-H

CO CO IO 00 CO
O r» co t*. Tf O
»— VO CO VO CO
CM i-H CM f-H

to" 10 co o o" co
CM «* f-H VO CM CO
f-. 00 CO OO VO CO
-—' i-H f-H -—^ 1— 1
^.^

o i a i a i
CO CO CO
«•— * +1.^ «•«*

0
4^
0}
Cn
io| o
ifw J f*~ t • g:
Q.I IO >, (T3
f-> 1 4^ p— 4^
3,O 3 O
as t— o a.

co
f>H
to I>»
TT f-H
to
	 ^
VO CO
OO CO
to
^*-

^
0 0
c^

»— »
cn CM
o en
to cn
*— *


er>
•
r^ eo
TJ- VO

10" co
-i" CD
r^ co
^«^ I-H



CO
r-! I-H
00 to
VO CM
CO ,-H
00 LO
^_^
en eo
0 0
en LO

oo* 10
LO O
^** r^
*» - i-H

r*.
o to
• — - to
CM

c\T TJ-
P^ LO
o oo
t-H
%^^

O 1
CO
^-^





IO
•f—
a.
o.
3
at



to COCM
*f CMTO
I^t£.
vo *-•>
CO VOO
to Ttr^
^— ^ Tjl/)
If-'

vocn
*-^ • •
o ooo
— - ovo
C0~—

*-~* *—.
to vocn
f^ IOVO
CM LOfH
f-H ' >« 1


cvT coo"
• • •
Tf* CDi-H
oo eDr^

vo coco
c^ cvjcn
LO f Ikj
f-H CMi-H
"~^ **^


10 r»-Tr
o encM
2. tfit^
LO LOVO
a COLO
^ ^
co enco
CM OOO
f-H VOOO

I-H CMLO
O r-o
CM COCft
f-H CM 	 	

Tp t^Tr"
o LOO

-H CMi-H

*-» r-.£T
• COTt-
en I-HVO
CM CM 	
f-H
«l_^
a i o"
CO CO
NlM^ ^-^






^_
IO

o




VIII-71

-------



























































•
X»
OJ
3
C

1 »
e
6

«
I-H
1
O
cu
r—
IO
h-
























4->
C
E
+•>
0)
c_
1—
c
•i-
t_
1































_^
en
£






Ifc
c



p-





CM








E
jf
•O
H



*
c
IO




00



^.








g
•I







*
IO
£


IO





PO








_
0
^
c
s






*
c

3

ie




t/>
Ol

"G
O)
.. ex
o> oo

CM


••H
00
t-H
t-H

CM
CM
00




^.
in
t-H


I-H
,^
"*"


f~-
CO



in
•
o
en




i



c
o
o>

4-> O
 E
3 IO
C71JJ
3 O
< a.
CO
I-H



•— I CM
r^ O
en t-H

^ 00
CO t-H
P""
*


*• i— <
VO 00
t-H r*»
i-H
*~*
co in
CO 00
*• CO
^-*

^i ^
CM r>>
oo f»-
in t-H
t-H t-H


r»- co
CO *
co in

^•^
i~* in
in CM
— vo
CM
^^
in
9
00 10

^— '



O 1
00
•t^,^




10
•y—
a.
a.
3
cc



CO CO*
^ - ^ ( 9
O 00 VO
in t-H
*— •» • •
O O VO
— in co

^_
O in

o vo m o i o i o i
• — o in • — *—* *~*

I-H CO O
• • •
vo in en
r* oo in
i

00 O CM"
CM o *r
CM «• CO

O I^I-H I-HCM cor^ *oo
**• ovo eyivo oo cnvo
. — co r». « — < — -^
t-H .

VO CM CM
r^ vo O i
^J Cft CM
^**
(O r-~ vo
oo o ey>
oo CM «cr
*— ' '~| *~->

^1 ^ n, ^ ^ ^1 ^ ^
N» r*^ in I-H I-H Cft CO O I-H
in i-Hr- CO* OI-H *in
in r- i-i CM I-H >^- CM t-i
i-H CM t-H >_• — ^
*"' ^^

r> co in
i— 1 I-H r-H
CO O *
t-H CM » —

*—** *~* ^-*— ^— «
IO CM 00 CM CO CM CO
in vovo 01 OCTI ocn
co vo co »-^ in t-H m I-H
^ CM^- *-* ^
^^ ^-«
r-> in oo
• • •
I-H vo in
r» co *
^-* t— i ^-^



o i a i o i o i o
00 00 00 00 00
*^ %H^ *— -• ^ f NH^

O
«- OJ
a> en
J3 O 10
^ E e i- —
10 cu 10 a.  > 4-> Q. 4J
O O O 3 O
r- Z o. on 1—
00
V-H





in
CU
o.



5
^_
4->

M
E
i-H

Ol
10
^
cu

IT)
E
t_
t-
M
e
£
>,
•o

o>
V)
CU
3
ie

CO
??
^ O
C O)
Oi_ 1
+J
•r- IO
4-> 0
IO f-
"> "cx
cu cu
•o «.
•o o
{_ X
IO 4->
TJ
IO O

(/) 

T)
C tt)
IO t—
Q.
CU
«- -o
IO C
IO
 10
3 £
* •*>
VIII-72

-------
Table D-3.   Specific leaf area (cm2/gdw)  of Potamogeton  perfoliatus  and  Ruppia
martima in some

Pond
1

2

3

7

8

Date
15 June
1 July
16 June
1 July
16 June
15 July
18 June
15 July
18 June
of the experimental ponds.
Section
Tip
Tip
Tip
Tip
Tip
Tip
Base
Whole
Tip
Tip
Base
Whole
Tip
Potamogeton
X
124.5
193.8
252.3
213.6
318.9
273.5
210.1
241.8
216.0
251.1
192.8
217.0
194.2
SD
35.4
52.4
54.2
59.4
49.0
43.1
45.1
53.9
54.0
115.1
102.6
108.0
80.0
%c
28
27
21
28
15
16
22
16
25
46
53
50
41
Ruppia
x 5D *L
12.4 5.3 43
71.0 38.3 54
49.3 9.1 19

40.4 6.1 15

58.9 14.0 24

48.8 13.7 28
                                    VIII-73

-------
1
• Table D-4. Chlorophyll a and b concentrations ( g/cm^) and ratios for
1 Potamogeton perfoliatus in a low nutrient dosed pond in relation to the
! dates nutrients were applied
Date o*" Sampling
Application
July 8
July 14
July 16
July 17
July 23
July 28
July 29 July 29
July 31
August 4 August 4
August 5
August 6
August 10
to the pond.
Chlor a
13.11
9.74
10.54
9.35
10.90
10.92
8.36
7.37
9.04
10.99

Chlor b^
2.01
2.25
2.35
2.83
1.53
2.36
1.87
2.39
2.68
1.73

a/b
6.53
4.33
4.48
3.31
7.15
4.62
4.47
3.08
3.38
6.34
VIII-74

-------
I/I
-o

ex
IO

c
i
O)
a.
X
W
c
(/)
4"*
c
TO

OL
3
U
IB
•o
o>
i
JQ
C
O

*T
C
IB
"ex

T3
$

O)
.-
•T-
OJ
I
U *
4-> V)
is C
.t o
CX-«-
1— 4-*
0.1-
O> "O
•o
"IB "
4-» 4->
O C
4-> 41
<•- C.
O 4->
3

l_
IB O
CO 0)
U
• a>
in -»-j
i .0
O 3
I/)





ai

IB
a





CM CM CO ^ CO CM
«!• C\J VO O CO 00
CM O CO ,-1 «»• O
*~* ^-^ ,-,
o CM m o oo o
^t" i— 1 CM CO ^x CO
CM O CO i-< CO O
*~* *-^ «-»
m r» mo CM •»»
in CM i— • vo co I-H
CM O CO O «S- O
^* — —

,—» -I—. *-»
1-4 CM ^" CO OO CO
i-4 CO 00 CM CM CO
•-I O ^ CM CO i-H
CM O IV. CO OO '-|
O CO CM *J" VO CO
•-I O ^- CO CO i-H
O^H" o Co" rv. o"
CM in oo CM co «t
1-4 O ^ O CM I-H

r- oo 1-4 o" co vo
O O CM CO CM CO
I-H ^H CM 1-4 ^ O

o if o CM co in
o m in vo o I-H
CM O T CD ^-i-4
^- *» ^*^ ^— ,
^t in CM oo * CM
i-H O CO CO «3- VO
O O -HO «* O

rv CM co oo ^ I-H
in o CM tx. 1-4 i— i
CM O CO O ^" i-l


1 1


r~. CM co oo r-» I-H
in o CM rv. i—i t-H
CM O CO O ^-i-H



i o i a i o
CO OO (/>

c 
O 0) >> 3
4J C •- en
ai 3 3 3
CM ^ *"^ ^C
O IB
E hv CO CO •!•—
to -4 i-i I-H ex
4-> a.
O 3







1



1



*~* ^~- ^*
ID Tj- O CM h«. CM
CM VO SO I-H i-H CM
CM CM ^-«-l in i-H
^^ ^-O" -nC
o r» r- in I-H CM
*t r-> «• i-i in I-H
O VO rv CM ^ CM
^- co ^-oo CM ^t-
O O ^- O in I-H

r>» in in r~- ^ «^
VO ^ 1— 1 ^f ^ fH
<-t i-H CM 1-4 CM I-H

*• ^^ co oo o oo
CM «» ^- VO O CO
CO O CO O CO i-4
^^ .— •» <-»
CM in r>. o oo m
o o oo CM co vo
o o o o -HO

y* co C2 co co ^*^
in r^ co «-4 o •-•
i-H —4 i-H CM CM CM

r ^ f ^ ^ >.
CM r~. CM CM o ^-
O O O O i-4 O
O O O O O O

CO CM CM CM VO i-H
O r-» in in CM in
CO O CO i-H CO O



1 O 1 O I O
oo co oo
i\
VI
0> >t 3
c *^~ c^
333
T -3 «

•v. ro co
-H 1-4 1-4




-------
M
CM
0 co
01 3
E 0
*^_^. e
Ol
0? 01
^-* 4J
>^
C JZ
o o.

4-> "a.
U 4)
3
T3 4J
O C
C. 1C

"a.
co
I/I L.
O 0
L. Vt-
o>
c
TJ O
1C +J
U
• 3

CO
ce i_
^-^ CL

c ,
CL j=

ffim
C CL
O Ol
4J t.
O t-
•S 41
•O JC
O 4->

CLT3
4-> 1C
c
Ol d
t_ 3
1C 4->
CL ic
CLt-
ic >—
O
V-  Ol
I/I O
Ol E
1C
VV- 4->
0 O
a.

W 0
6 *t—
g
AT
^
«—H
2^
O)
,_
«


































i








































^
•
o
4)
a
t_
(O
CL
01
I/I
•o

1C
u
£
+*
o
4-1







O
•^
4->
1C
at

Q£
• •
a.




._
1C
o

vt
1 O)
^ 4-1
CL >>
xl £
CL



^
^(
I/O











fm*
t

CL

*Q.
1. 1





D.





Q£




1C
a.









C
o

41
0)
C
1C
o
a.




a.





f-H
3J




 C

r— Cfi
OL ai



ai
41 1C
u o
t_
o



co ^-*i- aio oo P*.OO mcM
CM cnco coco -4 coca ^ -<



COIOVOOO CM>-IVOt-l
• ••• •••«
o CMCM cOi-H o min >— i CM




^ ' PN» CO C3 ^f ^^ ^0 ^H
• •• •• • ••
^r o in ft t^ en CM o oo
CO





^H o^ c^ ^^ ^Q ^^ o\ co f*4 ^^
o> ^fco r-m o co^t1 voin
i~^ CMC-* tnoi vo ooco <• cr
i— 1 t— 1 «-H •— 1


co co en CM oo en co co CM vo
*»• CMVO CMCO i-l CO '-> OOCM
• *• •• • •• ••
CM t-ii-i COCM CM i-ir^ in"*-



CO ^^ C3 ^O 00 C3 ^^ C? ^^ ^^
"^ ^0 r** if) ^H ^^ ^^ CO f^ CO
• *• •• • •• ••
in *— i LO cor^ co vo 10 oo LO
^-i ^H



^-^ O> U) «± 00 O* CO O CW O
o a* oo ir> io r^ ^- oo in I-H
^H CVJCSJ CO r-J ^H ^ uJ ^00
»H


fx. ^3 QT> i^. c> o^ ^H ^H r^fc «^
O cnr^ r-^ \o r** r^^n u>*o
• •* •• • •• ••
*-H oo oo f-t o<-< m oj


§^D i-. oo ca a» m a*
C? P-» C? t*** (O CT> ^*
• ••• ••••
o CMCM CM>-I o com oo in





^- oo ^f oo o vo* mi*>
oo in r- oj co co cnvo CMVO
• •• •* • •• ••
vo en»n CMP*. * covo o«-<
t— i


vo oo m co o g>fo &\f>
co coen «-r* *• oo CMVO
• • * •• • •• ••
«-i OO CM I-H O ~* «O O •-<


00 O O Ol O OP«.«-i
**• CM oo r~ ^H en oo vo
in o>* cnvo co CM o oo




01 Ol •• Ol Ol
•• CL CLI/I O. u> Q CL Q. co CL co
§P Poo P co oP Poo P- co
a.
h-
h- 4J Z 4.)
Z CO UJ CO
UJ Q) >, 3 —• 41 >) 3
•-H C ^— Ol OC C ^ Ol
OC33 3 h- 33 3
3 z
z vo m in co m m
i— i t— I 3C i— I i— I
8 2
~J ..1 T T 1C "^









•
^..^
Ol
I/I
1C
.0


X
ai
ex
1C

g
o

vt-
E
U

o
CO
1
m


A
E
01

to

t-


_.
01

**
c.
o

^-v
0.

I *
<>— '
CO C
ai o
^ -j- *
u c.
0 O
J3 CL
B —
 O
^
CT>  E
C Ol
ai ai
E to
F"
L_ AJ
01 C
CL ia
X f—
LU O.
* •*>

-------
Table E-2.   Epiphyte biomass and net productivity, and productivity of the host
     plant  Ruppia maritima for the upper (0-15 cm) and lower (+15 cm) sections
     of the plant. Total  production is expressed per gdw of the macrophyte.
 Plant      Epiphyte        Ruppia
Section      Biomass      Production
              (gdw)     (mg C gdw^h-
                                           Epiphyte Production
                                             (mg C
                                          Per Epiphyte  Per S
                                                         Total

                                                      Production
                                                   (mg C
 0-15
  15+
0.37

0.48

0.63



0.32

0.40

0.45
                              3.82

                              2.13

                              2.31

                              3.23

                              2.56

                              0.62

                              0.91

                              0.66
0.51

0.80

0.57



0.78

0.86

0.64
0.19

0.38

0.36




0.25

0.34

0.29
2.32

2.69

3.59



0.87

1.25

0.95
                                 VIII-77

-------
1
       Table  E-3.   Epiphyte biomass and productivity, and productivity of the host plant
           Potamogeton perfoliatus for the upper (0-15 cm) and lower (+15 cm) sections of
the
Plant
Section
(cm)
0-15






15+






plant. Total production is expressed
Epiphyte
Biomass
(g/gdw(SAV))
0.29
1.73
1.90
2.39
3.76
5.22
6.13
0.27
0.77
1.25
1.39
1.68
3.84
5.44
Potamogeton
Production
(mgC gdw1 h"1)
5.09
5.31
3.44
4.67
4.25
5.44
1.52
2.32
2.00
3.12
0.96
2.58
0.85
2.08
per gdw of the
macrophyte.
Epiphyte Production
Per Epiphyte
(mgC gdw~i h"1)
0.15
0.97
1.59
1.03
0.94
0.57
0.92
0.08
1.45
0.94
1.02
1.02
0.56
0.71
Per SAV
(mgC gdw1 h"1)
0.04
1.68
3.02
2.46
3.53
2.98
5.64
0.02
1.12
1.18
1.42
1.71
2.15
3.86

Total
Production
(mgC gdw1 h'1)
5.13
6.99
6.46
7.13
7.78
8.42
7.16
2.34
3.12 £
4.30
2.38
4.29
3.00
5.94
                                                 VIII-78

-------
                                            on
Treatment
Date
5 June
16 June
23 June
24 June
16 July
17 July
23 July
30 July
31 July
10 August
11 August
Control
1
7.16
5.89
4.69
(0.32)
4.53
(0.54)
3.09
(0.65)
7.83
(2.11)
8.89
(0.23)
9.47
(0.96)
8.35
(1.47)
7.36
(0.56)
6.52
5
6.34
6.31
9.19
(1.88)
4.65
(1.22)
3.84
(0.18)
8.03
(1.21)
7.43
(1.71)
7.95
(0.53)
7.17
(1.25)
4.71
(0.32)
6.92
Low
3
5.53
6.53
6.39
(0.84)
5.35
(0.34)
3.75
(0.38)
8.59
(0.66)
6.14
(0.50)
8.59
(1.01)
7.49
(0.39)
2.64
(0.33)
2. 6 ">
6
4.62
4.53
6.38
(0.71)
4.92
(0.36)
2.06
(0.67)
7.52
(1.40)
5.70
(0.96)
7.87
(1.37)
f.4s
5.13
(1.26)
5.24
Mid
4
5.19
4.33
5.77
(0.36)
6.55
(0.41)
3.18
(0.27)
10.26
(2.09)
6.23
(0.40)
10.21
(1.50)
8.83
(0.16)
9.29
(0.51)
6.80
8
5.50
6.48
6.28
(1.41)
5.35
(0.46)
2.22
(0.16)
11.31
6.64
(0.70)
11.88
(2.19)
8.75
(0.98)
9.77
(0.62)
5.63
Hi
2
-
5.94
5.35
(0.23)
7.05
(0.26)
4.33
(0.39)
10.06
(0.67)
9.18
(1.09)
8.57
(0.70)
3.17
1.09
(0.44)
1.46
gh
7
5.17
3.40
7.80
(0.70)
4.30
(0.15)
3.55
(0.95)
9.52
(1.85)
9.42
(0.39)
17.28
(4.02)
12.46
(1.96)
1.92
(0.17)
3.10
VIII-79

-------
Table F2.  Nighttime respiration (Rn) for total  communities in experimental  ponds based
     on dawn/dusk readings of dissolved oxygen (g02 m~3 d"1).
Treatment
Date
15 June
23 June
16 July
23 July
30 July
10 August
Control
1
3.2t
(0.77)
2.75
(0.36)
4.85
(0.43)
6.55
(0.22)
5.37
( )
2.48
(0.57)
5
3.05
(0.47)
3.31
(1.58)
4.07
(0.33)
5.17
(0.75)
4.33
(0.53)
-
Low
3
3.74
(0.67)
3.20
(0.46)
5.01
(0.07)
4.86
(0.63)
4.55
(0.38)
0.63
(0.22)
6
3.45
(0.44)
3.13
(0.26)
4.62
(0.92)
6.04
(0.31)
3.35
(0.27)
3.65
(0.69)
Mid
4
3.80
(0.36)
3.35
(0.13)
5.05
(0.31)
5.94
(0.25)
4.32
(0.78)
3.78
(0.89)
8
4.36
(0.11)
3.66
(0.80)
5.66
(0.30)
6.26
(0.18)
6.09
(1.47)
3.76
(0.18)
High
t
4.26
(1.14)
5.64
(0.33)
6.93
(0.15)
8.60
(0.69)
7.10
(0.13)
3.05
(0.20)
7
3.24
2.71
(0.61)
6.15
(0.54)
7.25
(0.24)
6.87
(2.00)
3.06
(0.11) i
                                            VIII-80

-------
f
           Table F-3.  Ratios of apparent productionrnighttime respiration (ParRn) for total communitit
                in experimental  ponds based on dawn/dusk readings of dissolved oxygen.
Treatment
Date
23 June
24 June
16 July
17 -July
23 July
30 J'dy
31 July
10 August
11 August
Control
1
1.71
1.65
0.64
1.61
1.36
1.76
1.55
2.97
2.63
5
2.78
1.40
0.94
1.97
1.44
1.84
1.66
-
-
Low
3
2.00
1.67
0.75
1.7i
1.26
1.89
1.65
4.19
4.22
6
2.04
1.57
0.45
1.63
0.94
2.35
1.93
1.41
1.44
Mid
4
1.72
1.96
0.63
2.03
1.05
2.36
2.04
2.46
1.80
8
1.72
1.46
0.39
2.00
1.06
1.95
1.44
2.60
1.50
High
2
0.95
1.25
0.62
1.45
1.07
1.21
0.45
0.36
0.48
7
2.88
1.59
0.58
1.55
1.30
2.52
1.81
0.63
1.01
                                                     VIII-81

-------
                                   CHAPTER IX
                       THE DECLINE OF SUBMERGED VASCULAR
                    PLANTS IN UPPER CHESAPEAKE BAY:   SUMMARY
                     OF RESULTS CONCERNING POSSIBLE  CAUSES1
                                W.  Michael  Kemp2
                               Walter R.  Boynton3
                               Robert R.  Twilley2
                               J. Court Stevenson
                                 Jay C. Means3
                                      1983
^Contribution No. 1431          University of Maryland
 Center for Environmental and Estuarine Studies (UMCEES),

2Horn Point Environmental Laboratories, UMCEES, Box 775,
 Cambridge, MD.   21613 USA

3Chesapeake Biological Laboratory, UMCEES, Box 38,
 Solomons, MD.   20638 USA
                                     IX-i

-------
                                ACKNOWLEDGEMENTS
     We would like to thank our many colleagues, students and research associ-
ates for their contribution to the conduct of this program: S. Bollinger, J.
Cunningham, D. Flemer, W. Goldsborough, T. Jones, M. Lewis, D. Marbury, L.
Murray, S. Nixon, R. Orth, P. Penhale, M. Shenton, K. Staver, R. Wetzel,  R.
Wijayaratne.  Special thanks are due to Al Hermann for his major contribution
to the numerical simulation modeling.  This reserach was supported by grints
from the US Environmental Protection Agency No. R805932010 and X00324801D and
the MD State Dept. Natural Resources, Tidewater Administration No. C18-8U-430
(82).
                                     IX-ii

-------
                                    ABSTRACT


     This paper provides a summary and synthesis  of research conducted to
investigate possible causes of the decline in abundance of submerged aquatic
vegetation (SAV) in upper Chesapeake Bay beginning in the late 1960's.  Three
factors potentially were emphasized in this study:  runoff of agricultural
herbicides; erosional  inputs of fine-grain sediments; nutrient enrichment and
associated algal growth.  Widespread use of herbicides in the estuarine water-
shed occurred contemporaneous with the SAV loss;  however, extensive sampling
of estuarine water and sediments during 1980-81 revealed that typical  bay
concentrations of herbicides (primarily atrazine) rarely exceeded  2 ppb.   On
two occasions relatively high values (20-45 ppb)  were observed for brief (2-4
h) periods in a small  cove following runoff events.  Short (2-6 h) and long
(4-6 wk) term experiments indicated that ephemeral phytotoxic effects  would be
expected in response to these highest herbicide concentrations followed by
rapid recovery.  However, normal concentrations (< 5 ppb) had little measurable
effect on plants.  Historical increases in turbidity have been documented for
some bay tributaries since the 1940's.  During our study light (PAR) attenua-
tion by suspended fine-grain sediments contributed more to total turbidity in
bay shallows (< 1.5 m) than did phytoplankton chlorophyll £.  Diel cycles of
PAR available in SAV beds indicated that plant photosynthesis was  light-limited
for much of the day, and PAR often fell below the compensation level (Ic)
needed for minimal plant growth.  Although some SAV species exhibited  consider-
able ability to adapt to reduced light by such mechanisms as increased pigmenta-
tion and stem elongation, increased turbidity has probably reduced overall
depth distributions of SAV markedly.  Effects of  the continual increase in
nutrient enrichment of the bay (documented since  1930) were tested by  experi-
mentally fertilizing pond mesocosms at levels common to the upper  estuary.
Moderate to high nutrient loadings resulted in significant increases in growth
of epiphytic and planktonic algae and c Teases in SAV production, as  well  as
premature seasonal senescence of fertil ..ad plant populations.  Direct measure-
ments demonstrated the inhibitory effect of epiphytic growth on SAV photosyn-
thesis, due largely to light attenuation.  The results of these various experi-
ments were synthesized into an ecosystem simulation model which demonstrated
the relative potential contributions of the 3 factors to SAV declines, where
nutrients > sediments > herbicides.  Other factors and mechanisms  are  also
discussed along with possible resource managements options.
                                      IX-1

-------
                                   INTRODUCTION


      It is widely recognized that  submerged vascular plants play an important
role  in the ecology of littoral regions of lakes, estuaries and oceans (1, 2,
3).   While a number of studies have noted the Ability of these plant communi-
ties  to attenuate variability of nutrient, sediment and production cycles (4,
5, 6), several such communities have themselves undergone extreme fluctuations
in distribution and abundance.  For example, in the mid 1930's a widespread
die-off of the seagrass, Zostera marina, was well documented throughout the
North Atlantic coastal regions (7JITFTe cause of this occurrence has never
been  unequivocally established, although recent suggestions have pointed to
subtle climatic shifts (8).  Other reports of regional declines in abundance
of submerged aquatic vegetation (SAV) have indicated the possible influence of
human activities (9, 10).

      Few of the  reported SAV declines have occurred in estuarine environ-
ments and most have involved 1-2 plant species.  However, in one of the world's
largest estuaries, Chesapeake Bay, a major loss of SAV has continued since
the mid 1960's to the present (11, 12, 13).  More than 10 species have experi-
enced significant decreases in abundancce, including Potamogeton perfoliatus,
_P.  pectinatus, Valisneria americana, Zannichellia palustris, Ruppia mantima
as well as the marine species Z. marina (12).In~the upper estuary this decline
in native species was proceeded" by an invasion of the exotic, Myriophyllum
spicatum (Fig. la) which eventually also died-back (11).  Studies of seed and
pol len distribution in sediment cores T'rom the upper bay have demonstrated
that  this dimunition in plant abundance is unprecedented (Fig. la) for at
least the last century (13).  In general, it appears that the recent decline
occurred first and with greatest intensity in the brackish waters of the estu-
ary,  with 1L. marina communities in the lower bay being affected less and some-
what  later (l"3JT

      Numerous mechanisms have been cited as possible causes of this occurrence
(12).  The concept that natural entrained population cycles or global  climatic
events might be responsible seems  unlikely in view of the range of biological
and physiological characteristics  for the numerous species invoked.  In addi-
tion, there is no parallel trend in plant abundance apparent in nearby coastal
regions (13).  Other factors including animal foraging and grazing and major
storm events are probably of occasional and local importance, but these are
part  of the normal milieu to which SAV are exposed and hence are insufficient
to explain this abnormal decline.  The absence of correlations between distri-
bution of SAV and industrial pollutants renders such anthropogenic wastes an
unlikely cause; however, more general changes in water quality associated with
diffuse sources (e.g. runoff) do represent a potential explanation(12).

      In an extensive review concerning SAV in Chesapeake Bay and elsewhere
(12), it was suggested that three  categories of environmental changes deserved
                                       IX-2

-------
a.
  o 300
Q Z
UJ < 200
      03
      <
     '00
      0
            .SAV-SUSQUEHANNA
               FLATS
                                      EXOTI C
                                           NATIVE
     *>0 200

     -* 100
      a
      *   0
          ATRAZINE  USE IN
          MARYLAND COASTAL
          COUNTIES
                JL
                        _L
e.
V00
_*600
*• 400
 * 200
     0

 '„ 180
 z 100
 O
 »-  20
         PATUXENT RIVER BASIN
         SEDIMENT  YIELD
                           INTENSE
                           CONSTRUCTION
             FERTILIZER SALES
             IN MARYLAND
                                                 AGNES
                                  TOTAL
                                  NUTRIENTS^
                                            (NtK-»P)
                                   NITROGEN^.,-—
f.
 2  22
•T  16
 •n
 z  10
 ^   4
        - WASHINGTON,D.C.
        _ SEWAGt DISCHARGE
         (NITROGEN)
            1930    1940
                        1950    I960
                          YEARS
                                        1970
1980
                         IX-3

-------
?!
                 Fig. 1.  Summary of long-term trends  (1930-1980)  in  selected  variables for
                          Chesapeake Bay and  its  tributaries:   a)  submerged aquatic vegetation
                          abundance in the upper  bay (1958-1976 (11),  prior to 1958 (14)); b)
                          use of atrazine in  coastal  plain  counties draining into the bay (12);
                          c) Susquehanna River flow (69); d) idealized sediment yield for Patux-
                          ent River basin (28, 29); 3)  fertilizer  sales 1n Maryland (35); f)
                          nitrogen in sewage  discharge  from Washington, D.C. into Potomac River
                          estuary (35, 69).
                                                        IX-3a

-------
               further attention:  increased  fine-grain  sediments  from  land erosion;  increased
               algal  growth from nutrient enrichment  of  estuarine waters; and aqueous concen-
               trations of herbicides  arising  from  agricultural runoff.

                    This  paper  summarizes the  results from  recent studies examining  possible
               causes of  the SAV decline in  upper Chesapeake  Bay.   Specifically, we  report
               the  salient findings  of investigations concerning: estuarine distributions,
               degradation, sorption,  and SAV  phytotoxicity of selected herbicides;  light
               availability and vertical attenuance by  suspended material, distribution of
               this suspended matter,  and photosynthetic responses  of  SAV to different
               light  regimes; effects  of nutrient enrichment  on algal  growth, and responses
               of SAV growth and production  to elevated  levels of planktonic and epiphytic
               algal  biomass.  We also synthesize these  results toward reconstructing a plaus-
               ible scena, io vis a vis this  loss of submerged vascular plants, and we
               provide comments on resource  management  options and  the future of SAV in upper
               Chesapeake Bay.


                                            APPROACH TO PROBLEM


                    In 1978 we  initiated a 3-year study  to  investigate various aspects of the
               ecology of SAV communities in Chesapeake  Bay.  While intensive research was
               conducted  at several  locations  along the  estuarine salinity gradient, our work
I               focussed on communities located in the low salinity  (5-15°/oo) region.  The
i               research was designed to address three general questions:  (1) factors poten-
               tially responsible for  the observed  decline  in SAV distribution and abundance;
               (2)  SAV community interactions  and characteristics;  and (3) resource  management
               options.

                    This  research program was  organized  in  a  hierarchical fashion to deal
               with the complexity of  the ecosystems  studied  in addressing these questions.
               Depicted in Fig. 2 1s the manner in  which (c)  specific  research approaches
               were Integrated  into  (b) broad  research objectives which were, in turn, related
               to (a) research  questions (15,  16).  General objectives of this reserach pro-
               gram involved:  the development of mechanistic relationships among ecological
               factors through  controlled experimentation;  the interpretation of these results
               in terms of actual  conditions in nature;  and the establishment of a framework
               to extend  specific results in to a general context.   The specific research
               approaches employed ranged from shortterm laboratory  experiments to seasonal
               mesurements in the estuary and  experimental  ponds, with ecological modeling at
               all  levels for analysis, interpretation and  generalization.  In the following
               sections we review the  results  of our  field, laboratory and modeling  studies
               to evaluate the  possible roles  of herbicides,  suspendable sediments and nutri-
               ents 1n the SAV  decline in the  bay.


                                   HERBICIDE INPUTS,  FATE AND PHYTOTOXICFY


                    For the last two decades the most widely  used herbicide in the Chesa-
               peake Bay  watershed (and particularly  in  the surrounding coastal plain) has
                                                     IX-4

-------
11
                       CO
                   X  HI
               "ox
               u_  cc  o
               -  <  <
               «  UJ  0
               W  CO  *
                       CO
                   X  UJ
                   o  >
               •2  (T  -
               <  <  H
               O  Id  O
               C  CO  UJ
               CD  UJ  "*
               CD



-J x z
J £°
< £ -
CC < H
UJ U CO
> tO QJ
O UJ ^
CC 0
•
<

"- z
0 z
2s
« Q
< >
0 <
£ft





CO
-1 Ul
< 0
o z
— UJ
0 13
0 0
_J UJ
0 <"
o z
UJ O
04 0





s £«
?So
iP
ro
                                                       IX-5

-------
Fig.  2.   Schematic flow chart illustrating the relationships  between  specific
         research approaches, broad research  objectives  and overall research
         questions.  Three thicknesses of arrows  suggest levels  of  influence
         for these items.
                                    IX-5a

-------
                been atrazine (12, 17).  Since its introduction to the region in the early
                1960's, its use, which is particularly important for corn crops, has grown
                steadily to the present (Fig. lb).  Atrazine was thus selected as the primary
                compound for intensive investigation.  Our research also considered the distri-
                bution and phytotoxicity of a second important compound, linuron, which is
                used for wetd control in soybean crops.


                                     Herbicide Distribution and Degradation

                     In general about 0.2-2.0% of atrazine applied to agricultural  fields is
                transported into adjacent waters in runoff, and resulting concentrations in
                the estuary depend on dilution, dispersion, adsorption and degradation (17,
                21, 27).   In the open waters of Chesapeake Bay observed concentrations of
                atrazine and linuron rarely have rarely exceeded 1 ppb (18, 19, 20).  In major
                tributaries such as the Choptank and Rappahanock Rivers concentrations of 5
                ppb may occur following a major spring runoff event.  Such events can generate
                transient (2-6 h) concentrations up to about 40 ppb in secondary tributaries
                and small cc-.cs (20, 21, 22).

                     Our data indicate that atrazine and linuron degrade rapidly in estuarine
                conditions with half-lives of 1-6 wk (17, 23).  In addition,  these  herbicides
                exhibit strong tendency for sorption to particles and colloids (20, 24), and
                degradation of particle-bound atrazine appears to be even more rapid (25).  As
                a result of dilution and degradation, concentrations of atrazine on suspended
                and deposited sediments in the estuary were seldom greater than 5 vg/kg
                suggesting little accumulation potential, (20), and moreover, once  bound to
                particles atrazine becomes virtually unavailable for plant uptake (25).


                                       Herbicide Uptake and Phytotoxicity

                     Using 14C-ring labelled atrazine Jones et al. (in 25) found that uptake
                by £. perfoliatus was rapid, reaching equilibrium within 1 h.  The  observations
                that atrazine concentrations were low in sediments (20), combined with the
                relatively slow translocation of the herbicide measured within plant vascular
                tissues (25) suggest the primary mode of uptake is foliar. A strong correla-
                tion was observed between atrazine uptake and short-term depression of photo-
                synthesis.  After a sequence of washes in atrazine-free water, a portion of
                the incorporated atrazine was released and photosynthesis returned  to control
                levels within 4-8 h (25).

                     In addition to short-term (2 h) experiments relating photosynthetic
                depression to atrazine exposure, microcosms containing estuarine water and
                sediments with £. perfoliatus or M. spicatum were used to test longer duration
                (4-6 wk) effects of atrazine and Tinuron on plant growth and  production.  At
                concentrations less than 100 ppb of either herbicide, both plants exhibited a
                significant trend of photosynthetic recovery within 1-2 wk after initial  ex-
                posure despite the continued presence of herbicides in microcosm waters (25).
                Using a linear dose-response model  for apparent photosynthesis (or  net biomass
                accumulation) versus the logarithm of initial herbicide concentration, signifi-
                cant (r2x).93) relationships were observed for all  combinations of  herbicides
5
                                                      IX-6

-------
and plants (Fig. 3).  Values of IJ.Q and *50 (concentrations resulting in
1% and 50% inhibition of photosynthesis) ranged from about 2-4 ppb and 70-120
ppb, respectively, for the 4 combinations of herbicides and plant species
tested (25).

     At 5-10 ppb (the upper limit of herbicide concentrations typically occur-
ring in most of the estuarine shallows) a 10-20% loss of photosynthesis would
be expected.  However, the ephemeral nature of such elevated herbicide levels
and the tendency toward rapid photosynthetic recovery in these plants  would
render such effects short in duration.  Overall, atrazine phytotoxicity has
been tested for 7 species in Chesapeake Bay in various experiments,  and some
differences in sensitivity were apparent (25,  26,  27).  For example, the exotic
species, M. spicatum, exhibited greater tolerance  than most native plants,  and
the seagrass. Z. marina, appeared also to be less  sensitive, while the fresh-
water member oT the Hydrocharitaceae, ±. americana, was among the least toler-
ant species tested.

     Some concern about the possible phytotoxicity of atrazine metabolites
formed during degradation was allayed in recent experiments, again using 14C-
labelled compounds (provided by Cipa-Geigy Corporation, Greensboro,  NC), where
150 values were found to be at least 10 times  greater than for the parent
compound (25).  Thus, while conditions certainly occur sporadically  in the
estuary where herbicide stress could induce some loss in SAV production, the
possibility that herbicides led to the SAV decline saems remote.


                 SUSPENDABLE SEDIMENTS AND SAV LIGKT RESPONSES
     Substantial changes in both sediment yield to and turbidity in Chesapeake
Bay and some tributaries have occurred over the last several  decades (Fig. Id).
In one major tributary (Patuxent River), there have been periods of increasing
and decreasing sediment yields, punctuated by occasional  major input events
during the last 50 years (14, 28, 29, 30).  Overall, it seems that  turbidity
in this tributary is higher now than 30 years ago (33) even though  inputs  of
fine-grain sediments have recently decreased (29), indicating there is not a
simple relationship between these factors.


                  Light Attenuation due to Suspended Sediments

     Laboratory and field experiments were conducted to assess the  relative
contributions of suspended sediments and chlorophyll-^ to diffuse,  down-welling
light attenuation.  Diffuse attenuation of phytosynthetically active radiation,
PAR (400-700 nm), due to different concentrations of suspended materials was
measured in stirred, opaque-sided, cylindrical  chambers (60 cm dia, 120 cm 'it)
by observing PAR at 10 cm vertical intervals.  Known amountd  of dried,  sieved
(62 v) sediment or algal culture suspensions were added to the chambers to
develop relations between attenuation coefficient (k) and suspended concentra-
tions.  The slope of k versus algal biomass (in terms of dry  wt.) was found to
be about twice that for k versus inorganic suspended sediments due  to the  high
absorbance of algal pigments at PAR wavelengths (31).  However, the relative
ranges of these two suspensoids typical of shallow estuarine  environments
                                      IX-7

-------
f.
                            • ^••••#»wa**  ••• •*•** «•••••• » •
                          .•.'.*•• •..•«.*•••*• *•/. . - . •••/•- •• •*••••••
                          • » ••*••••!» • ••*• l.» •• •••!••••"•• !••
                                                                  jQ
                                                                   Q.
                                                                   Q.
                                                                  o
                                                                  h-
                                                           - —   o:
                                                                  LU
                                                                  O
                                                                  o
                                                                  o

                                                                  LU
                                                                  Q

                                                                  O

                                                                  CD
                                                                  or
                                                                  LU
                 SIS3HlNAS010Hd   !N3dVddV
                        SNOI10nCI3d  !N30d3d
Nl
                                          IX-8

-------
Fig.  3.   Reduction in apparent  photosynthesis  (%) for two  species of  submerged
         aquatic  vegetation  resulting  from treatment with  the herbicide atrazine
         at 5 initial concentrations.   Three degrees of  shading  represent
         (from heavy  to  light)  maximum in situ atrazine  concentrations observed
         in:   open Chesapeake Bay waters; main bay  tributaries;  and a small
         cove within  one tributary (17, 25).
                                        IX-8a

-------
suggest that suspended solids are a far greater component to total turbidity.
Concentrations of suspended sediments (seston) often increase from 20-100
mg/i within 1-2 h due to moderate (5-10 m/s) summer winds (25, 31), while
chlorophyll-a concentrations typically range from 5-15 vg/t over a tidal
cycle (Fig. T) and are not markedly influenced by wind events.  These common
excurisons in seston and chlorophyll-^ would result in about 50% and 8% reduc-
tions in PAR at 50 cm depth, respectively.

     To assess jn-situ light conditions, measurements of seston, chlorophyll-
a, attenuation coefficient, water depth and PAR were made over tidal cycles in
Tittoral areas with and without SAV communities (25, 31).  Data from a typical
cycle (Fig. 4) indicate that both chlorophyl!-£ and suspended particulate
concentrations were lower within SAV communities than in non-vegetated littoral
areas, and the differences were greatest at low slack tide, probably due  to
settling of particulates in the quiescent waters of the bed.  It i* evident
here that PAR attenuation followed seston consistently at both sites, while
not so for chlorophyll-a.  Consistent with laboratory data, most (> 95%)  light
attenuation could be attributed to suspended solids other than chlorophyll-a.
Sediments may also be important in light attenuation when they settle from The
water column onto SAV leaves.  For example, Staver et al. (31) found 80-90%
of total epiphytic material (attached to leaves) to be inorganic sediments as
opposed to algae or other organic materials in a Chesapeake Bay SAV community.
However, there is a continuous cycle of deposition and resuspension of these
loosely attached sediments, and the integrated, long-term effect of these
processes on photosynthesis is not known.


                             Light Response of SAV

     In laboratory chambers the responses of P. perfoliatus and M. spicatum
photosynthesis to light intensity followed a "Basic rectangular hyperbolic
form.  Values of IK (intersection of initial slope and Pmax) ranged from
200-300 uEm^s'1 and were similar to values reported for other SAV
species (31).  A second important parameter of the photosynthesis-irradiance
(P-I) relationship is light compensation point (Ic), which is the PAR intens-
ity at which net production (in excess of respiration) approaches zero.  Values
of Ic at about 50 uEm^s"1 were reported for apical 10 cm growing
tips of £. perfoHatus (25).  However, Ic values for whole plants (including
non-photosynthetic root-rhizome tissue) would be substantially higher, probably
in the range of 75-150 yEjn~2s~1.  Given the ambient turbidity
levels generally observed in the bay (k > 2/m) in addition to light reduction
associated with epiphytic materials, it would appear that SAV are generally
exposed to light regimes approaching Ic.  Unless these plants can adjust  to
such light stress, net growth would not be possible.

           In fact, Goldsborough (32) found that at least one species (£.
perfoliatus) has considerable ability to adapt to reduced light conditions via
morphological changes (such as stem elongation and increased leaf area:weight
ratio) and biochemical changes (such as increased chlorophyll-£ per leaf  area)
which influence P-I relationships.  However, under extreme shading (< 75  yEm~2s"1)
the capacity of adaptive mechanisms was exceeded, resulting in reductions in
new shoots, number of flowering plants, and root storage nodules (which appear
                                      IX-9

-------
 O
 UJ
 03
(O

U_
O

UJ
O
o
UJ
o
UJ
m
V)

UJ
o

en
    r'
                                                                u_
                                                                o

                                                                UJ
       . I
                                          xnid
                          IX-10

-------
Fig.  4.   Time-course observations  for planktonic  chlorophyll £, total  seston,
         water depth, light  attenuation  coefficient and  light available at 10
         cm depth and at the bottom of a 1.0 m  (mean  depth) water  column for a
         bed of submerged plants (£. perfoliatus) and for an adjacent  unvege-
         tated area  on 22 September 1980.   Shaded regions represent typical
         ranges for  1^ (intersection of  initial slope and maximum  photosyn-
         thesis in a photosynthesis-irradiance  function) and Ic (compensation
         light level  for 10  cm  apical  growing tips of plants) (31, 32).
                                    IX-lOa

-------
to be overwintering organs, 1).


                   Overall  Effects of Suspendable  Sediments  on  SAV

     In the Patuxent Estuary, which formerly contained extensive SAV communi-
ties, chlorophyl l-a_ concentrations increased from  about 10 to 40pgA and
secchi  disc depths decreased from about 0.8 m to 0.35 m (or  k = 1.8 -
4.U/m,  assuming k = 1.4/secchi) for spring-summer  during the 1960's decade
(33).  Partitioning these attenuation coefficient  values between chlorophyl1-a
and other particulate and dissolved materials (34)  shows about  7% of total
attenuation could be attributed to the algal pigment  in the  early 1960's,
while about 14% could be so partitioned for the 1970 data.   While the relative
importance of phytoplankton attenuation increased  over this  period, it  appears
that most light attenuation was still associated with inorganic particulates
and other non-chlorophyllous organics.

      In any case, the observed increase in turbidity would  cause a significant
reduction in the depth distribution of SAV.  For example,  if we take summertime
irradiance at the water surface to be 1200 uEm"2s~1  (a typical  value
at noon on a sunny Aug-Sep day) and Ic to be 100 uEirf^"1, the
depth to which sufficient light penetrates to support SAV growth would  be
reduced from 1.4 m to 0.6 m for the aforementioned  scenario  in  the Patuxent
between 1960 and 1970.  Similar turbidities were observed (25)  throughout
littoral zones in upper Chesapeake Ba\  suggesting  that the  depth distribution
of SAV has been restricted due to redb  -. PAR availability.


                     NUTRIENT ENRICWENT AND ALGA1    V RELATIONS


      Watershed inputs of nitrogen (N) and phosphorus (P)  to Chesapeake Bay
have increased several-fold in the last two decades  (Fig.  1  e,f), and aqueous
concentrations of inorganic nutrients have likewise increased appreciably  in
many areas of the estuary (33, 35).  Correspondly,  phytoplankton populations
(as measured by chlorophyl1-a) have also expanded  during this period (33,  35).
Since it appears that submerged vascular plants can  obtain their N and  P re-
quirements from sediment pore waters (36, 37) which are rich in nutrients,  it
is expected that nutrient additions to overlying waters would increase  SAV
growth to a limited extent only.  In fact, there are a number of mechanisms
where fertilization may lead to decreases in SAV production  and abundance.
Nutrient enrichment tends to promote algal growth,  e'ther phytoplanktonic  (38)
or epiphytic (10) which in turn can inhibit SAV photosynthesis  via reductions
in light and molecular transport across SAV epidermis (39, 40).  In addition,
it has been postulated that increased phytoplankton  populations can stimulate
growth of animal fouling communities on SAV leaves  and other firm, stable
substrates (25, 41) leading to similar photosynthetic stress for the plants.
In this section we consider this C'.csclon terms of correlations from field
observations and results of nutrient enrichment studies in experimental  ponds.
                                     IX-11

-------
                     Algal  Responses to Nutrient Enrichment

     The effects of nutrient enrichment on algal and SAV growth were investi-
gated in 8 experimental ponds (area = 350m2; depth = 1.2 m) treated in
duplicate with 3 nutrient levels (plus controls) at 8-10 day intervals for a
10 week period during the summer of 1981.   Initial  nutrient concentrations
after dosing were 120, 60 and 30 yM inorganic nitrogen (50% NH^; 50%
NOj), with N:P (atomic) ratios of 10:1.  Nutrient loading rates at medium
dosage were equivalent to runoff from a typical  watershed (26% agriculture,
11% residential) draining into a small Chesapeake Bay tributary (42).

     Phytoplankton biomass (as chlorophyll-^) generally increased with treat-
ment levels (significantly greater at low and high  doses), particularly at the
highest nutrient amendments (Fig. 5).  A sequence of phytoplankton bloom events
were evident in high dose ponds, where chlorophyll-^ increased by 1-2 orders
of magnitude on several occasions within a 1-2 d period.   This response was
similar to those previously reported for nutrient enrichment studies (43, 44).
Epiphytic algal biomass was significantly higher than controls in all treated
ponds (Fig. 5).  Levels of epiphytic material in nutrient treated ponds were
similar to those observed in the estuary on scenescent plants while control
levels were comparable to those occurring early in  the SAV growing  eason (32,
45).  Again, this pattern of increased epiphytic algal growth is consistent
with results from nutrient enrichment experiments elsewhere (10, 46, 47).


                         Effects of Algal  Growth on SAV

     Correlations between PAR attenuation in the water and phytoplankton chloro-
phyll -a were obtained from measurements in the experimental ponds using only
those (Tata where the estimated weight of algal cells comprised a major faction
(> 10%) of total seston weight (25).  The slope of this relationship (0.0146
n^mg"1; r2 = 0.95) Is quite consistent with observations  from deep
coastal waters (48) and slightly greater than those derived from laboratory
cultures (31, 49).  We also obtained a strong relationship between PAR attenu-
ance through a plane  (estimated by scraping epiphytes from leaves of known
area into a petri dish and measuring PAR attenuation, 50) and epiphytic biomass
(r2 = 0.87).  The slope of this relation was 0.36% (mg/cm2)'1, which
again is comparable to previous reports (10, 50, 51).  Direct measurements of
SAV photosynthesis (using both 1J»C uptake and oxygen evolution methods) over
a range of epiphytic colonization levels (0-5 gdw epiphyte/gdw SAV) yielded
significant correlations with about 50% reduction in photosynthesis at 4
gdw/gdw (25).  After 8 wk of nutrient treatment, SAV biomass in medium and
high fertilization systems had decreased significantly compared to controls
(Fig. 5).  While the contribution of the water column to total attenuation
(from water surface to plant leaves) was generally  small  compared to that
associated with epiphytic material, without such light reduction due to tur-
bidity, epiphytic growth would have been insufficient to reduce light below
compensation levels (Ic).  Similar observations on the relative roles of
phytoplankton and epiflora have been reported for Danish lakes (51).
                                     IX-12

-------
I
_J
CL
o
tr
o
o
      10
    PHYTOPLANKTON
    V
I
     5.0
co <
UJ CO


— "°  25
x o»  *-D
Q. \
UJ
   -EPIPHYTIC BIOMASS
            '/.
                   A
    300
en

S
2
00
>
<
CO
150
   _ SAV  BIOMASS
              s
        Control   Low    Medium   High

              NUTRIENT DOSE

-------
!
             Fig.  5.   Summary of phytoplankton stocks (as  chlorophyll £),  weight  of  epiphytic
                      material,  and submerged aquatic vegetation  biomass  in  August  1981,
                      for experimental  ponds  treated with  4  levels  of  nutrient  enrichment
                      after 3 weeks.   Given are means   1  standard  error  (25).
                                                   IX-13a

-------
•I!
 •  -
 '.  '                                  Nutrient  Enrichment  and  SAV  Decline

   I                 While nutrient  fertilization  has  been occurring  1n  Chesapeake  Bay  over
   '            the last century or  more,  the upper bay  outbreak of M. spicatym  in  the  early
               1960's (just  proceeding the  general  SAV  decline) may~have marked the  onset  of
   ;            critical nutrient enrichment levels.   It appears that this  species  has  competi-
               tive advantage ov»r  other  SAV under conditions of  elevated  nutrient inputs
               (e.g,  52).  During the last  several years  numerous investigators have postu-
               lated that eutrophication  was responsible  for  observed losses of SAV  in fresh-
               water systems including:   Loch  Leven,  Scotland (38);  Lake Erie,  Chio  (53);
               Whitewater Lake, Ontario  (54);  and Norfolk Broads, England  (10).  Others have
               suggested that nutrient enrichment may be  one  of several factors contributing
               to dlmunition of seagrasses  in  su^h coastal ecosystems as the Dutch Waddenzee
               (9), the French Mediterranean (55) and Western Australia estuaries  (56).   In
               all  cases, these declines  have  occurred  progressively over  decade-long  periods,
               consistent with observations in Chesapeake Bay (13).

                    In our experimental pond studies, SAV biomass reductions in high dose
               ponds amounted to a  foreshortening of  the  normal growing sesaon; however, some
               effects on vegetative  reproduction were  also apparent.   For example,  shoot
               production from budding was  markedly reduced in experimental ponds  the  spring
               following a summer of  high nutrient treatment, and after a  second year  of
               identical treatment  vegetative  root nodules were also a  significantly reduced
               (25).  To interpret  these  results  in the context of long-term declines  in
               abundance, it may be necessary  to  relate premature seasonal scenescence to
               decreased reproductive success  in  a more quantitative fashion.


                                   SYNTHESIS  OF  FINDINGS AND IMPLICATIONS


                                 Ecosystem Modeling  as a Tool for Synthesis

                    The research discussed  in  this paper  has  identified and described  a wide
               range of detailed Interactions  between submerged vascular plants  and  their
               environment.   We used  a variety of numerical simulation models for  these SAV
               ecosystems to place  the complex interactions into  a unified ecological  frame-
               work.  Che of these  models was  designed  to address questions relating to man-
               agement of SAV resources in  the estuary, including both the role of these
               plants 1n secondary  production  and the environmental  factors influencing plant
               growth (57).   Generally, this model  consisted  of 14 simultaneous, non-linear
   {            differential  equations, with time  as the independent  variable.   Each  equation
   I            described the temporal behavior of a state variable,  with each term in  a given
   i            equation representing  a connection between that variable and another  internal
   |            or external variable.   The model was calibrated with  data collected from field
               and laboratory measurements  and was used to interpolate  between  discrete ob-
               servations in time and space and to both hindcast  and forecast consequences of
               changing external factors  (e.g. temperature, nutrient loading and flushing
               rate).

                    Digital  computer  simulations  were performed (using  finite difference
               numerical methods) to  consider  the effects of  herbicide, sediment and nutrient
                                                    IX-14

-------
  fl
.
inputs to the estuary individually and combined (Fig.  6).  Under simulated
conditions based on maximum observed herbicide concentrations in a shallow
embayment, annual mean SAV biomass (B) and total net production (NP)  decreased
by about 2% and 4% respectively (Fig. 6a), with greatest effects during May-
June runoff events.  A doubling of sedimer.. loading to the SAV bed and sur-
rounding waters resulted in J3% and 22% reductions in B and NP, and the effects
were most pronounced in the late summer and fall periods when heavy winds
stimualted sediment resuspension (Fig. 6b).  A two-fold nutrient errichment
resulted in marked reductions in B and NP (29% and 48%), with strongest depres-
sions occurring in June (Fig. 6c).  Apparently, relatively high nutrient and
light levels in spring promoted rapid growth of epiphytic algae which was
retarded subsequently in summer due to nutne-t depletion.  These simulation
results are similar to findings from pond fertilization experiments (25).
Combining these three stresses in a single simulation  resulted in an  average
of 35% and 56% losses of biomass and production, and the net effect of the
three factors thus appeared to be somewhat antagonistic (Fig. 6d).

     It is interesting to note that under combined stresses the growing season
(period of continuous biomass accumulation) was reduced from about 4.5 mo to
less than 2 mo.  In addition, plant biomass at the end of the simulated year
was considerably lower than initial levels under combined factors.  While this
might suggest an effect on reproductive success of this population, multi-year
simulations produced recurring patterns similar to that shown in Fig. 6.
However, this  is not surprising because reproductive biology, per se, was not
an explicit function in the model.


            Gradual Reductions in Abundance versus Population Demise

     Most of the species occurring in upper Chesapeake Bay undergo clear annual
cycles of above-ground structure, and new growth in spring may be generated
either from seeds or from underground storage organs.   The .nodel  simulations
indicated that under environmental stress, net growth  available for allocation
to sexual or vegetative reproduction is reduced.  Experimental observations
have demonstrated that extreme light stress, related either to turbidity or
epiphytic growth, results in significant decreases in  both flowering  and pro-
duction of underqround storage nodules (25, 32).  We hypothesize that the
observed gradual reductions in abundance of stressed plants in model  and field
experiments would eventually result in a local demise  cf these populations due
to reproductive failure.  This final outcome would require a repetitive se-
quence of annual stressed growth conditions, incrementally undermining repro-
ductive vitality.

     In general, SAV communities exhibit considerable  ability to modify their
local environment and thereby reduce stresses; however, this "buffering capac-
ity" can be exceeded and lead to premature seasonal die-back.  Two examples
involve nutrients ar I suspended particulates.  While SAV communities  are cap-
able of rapid uptake of nutrients from overlying water (4), nutrient  enrichment
beyond the assimilative capacity can promote epiphytic algal growth to the
detriment of vascular plar:t production (e,.g. 10, 25).   Secondly, the  well
established ability of seagrasses and other SAV to trap and bind suspended
particulate matter (e.g. 5, 58) is limited in silty estuarine environments
                                                    IX-15

-------
CNJ
 C
 o
 X)
 k.
 o
 o
V)
(ft
o

CD

I-
z
<
_J
o.

or
o
CO
36


27


18


 9



36


27


18


 9
         A) HERBICIDE

            RUNOFF
                       Basel i ne
                       Simulation


                         BasalIne
             b Herbicides

               Added
                     I   I   I
B) SEDI MENT

   LOADING
Baseline
            Sediments

            Added
36


27


18


 9



36 -


27 -


18 '


 9 -
C) NUTRIENT  j$j%jBa,.lin.

ENRICHME
D)COMBINED
                       Base!ine
         FMAMJ  JASONDJ

                MONTH  OF YEAR
                      IX-16

-------
Fig.  6.   Results of numerical  simulations of submerged vascular plant  biomass
         over an annual  cycle  under conditions  of  maximum herbicide  runoff,
         doubled sediment loading,  doubled nutrient  enrichment, and  a  combina-
         tion of these three stresses.   In all  cases  the  upper  line  (shown for
         comparative purposes) represents the baseline conditions  which  are
         compared to means  of  field observations in  the upper panel  (57).
                                    IX-16a

-------
(25), and associated turbid waters tend to reduce light availability and plant
production.  It appears that while moderate levels of these stresses result in
little if any reductions in plant production and distribution,  extreme levels
can overwhelm the ability of such plant communities to ameliorate these factors.


                         Reconstructing the SAV Decline

     Historical trends in data for SAV distribution and abundance, sediment
and nutrient loading and herbicide use (Fig. 1) coupled with experimental
results and model simulations reported here allow development of a plausible,
although not conclusive, scenario concerning the observed decline of SAV in
Chesapeake Bay.

     Accelerated soil erosion and sediment loading associated with extensive
land development in the late 1950's apparently led to elevated  turbidity in
several regions of the estuary (28, 29, 30, 35).  Nutrient loading also in-
creased from point sources associated with population centers and from regional
diffuse sources associated with intensifying agriculture.  In the Patuxent
River estuary, marked declines in SAV populations (59) coincided with increased
turbidity and nutrient concentrations.  In the Potomac River, invasion by M.
spicatum followed by proliferation of blue-green algal mats, correlated wiTh
expanding sewage discharges and incipient declines in native SAV (60).  The
establishment of M. spicatum in brackish portions of Chesapeake Bay proceeded
a general SAV decTine (11), which may have been a response to eutrophication
(52).   In the late 1960's, it appears that strict erosion control practices
reduced sediment runoff inputs and associated turbidities in some regions;
however, nutrient inputs continued to rise.  The general use of herbicides for
agricultural weed control in the watershed in the mid-1960's may have contrib-
uted slightly to deteriorate SAV growth conditions, and the greater tolerance
of ^. spicatum to herbicides is consistent with the observed brief prolifera-
tion of this species.  Possibly further compounding these stressed conditions
was the drought of the 1960's (Fig. 1) which led to increased salinities (61)
and probably salinity stress for various freshwater species (62).  In the
summer of 1972 a 200 year storm event resulted in rates of sediment deposition
up to 25 cm  (63) in some locations, virtually burying local plant communities.
While this storm occurred well after the initial evidence of SAV decline, its
effects are apparent in plant abundance data (Fig. la).  There is also evidence
for considerable damage to SAV populations from intense grazing and foraging
episodes by waterfowl and fish (64, 65).

     While it is impossible to definitely demonstrate what caused the SAV
decline, data from controlled experiments, as well as recent field observations
and historical records all suggest that nutrient enrichment and increased
turbidity probably played major roles. Herbicide runoff may represent an
ephemerally  and  locally important stress, but our data indicate its contribu-
tion to this general die-back has been minimal.  Major meteorological events
and direct grazing certainly have contributed to the  rigors of SAV existence
and may have accelerated the recent decline, but historical records indicate
these factors probably also played a secondary role.
                                      IX-17

-------
r
                                       Management of SAV Resources

                  The potential  for improving environmental  conditions for propagation and
             growth of submerged vascular plants in the estuary can be considered in terms
             of the relative importance and controllability  of  key factors.  Herbicide
             losses from croplands to adjacent watercourses  can be reduced by such measures
             as alternative weed control  techniques and "buffer stripping" at the margins
             of fields;  however, the socio-economic costs of such strategies  would be con-
             siderable (66).  Research results presented here indicate herbicides exert a
             minor influence on  SAV, so that the wisdom of such measures  may  be questionable.

                  Three  major sources of suspendable sediments  to the estuary are runoff,
             shoreline erosion and resuspension of bottom sediments (67).   Both shore ero-
             sion, which is important in the middle bay and  eastern shore tributaries (67,
             68),  and resuspension, which dominates in most  estuarine shallows (25),  are
             largely uncontrollable.  Existing soil conservation practices have been  reason-
             ably  effective in reducing sediment runoff (29, 30) and probably represents a
             realistic management option, given the apparent contribution of  suspendable
             sediments to the SAV decline.

                  Although nutrient inputs  to Chesapeake Bay from sewage  effluents represent
             less  than 20% of the total  for nitrogen (69), in some areas  such as the Potomac
             and Patuxent Rivers, they are  a major source which can be controlled,  albeit
             at considerable cost.  Diffuse sources, dominated  by agricultural  runoff,
             represent the largest nutrient input to the estuary, and efforts to develop
             effective controls  are crucial.  The substantial experimental  and correlative
             data  available strongly suggest that nutrient enrichment has resulted in losses
             of submerged vascular plants in a variety of aquatic systems, including Chesa-
             peake Bay.   This, coupled with other documented consequences of  excessive
             fertilization (e.g. noxious algal blooms and anoxic bottom waters) point to a
             considerable need to focus on  nutrient waste management.

                  The role of SAV communities In maintaining water quality and promoting
             secondary production has been  well established  in  general and in Chesapeake
             Bay in particular.   For example, it has been estimated that  about 50% of the
             current sewage derived nutrient inputs to the bay  could have been assimilated
             by healthy SAV communities which existed in 1960 (70).  Kahn (71) has  calcu-
             lated that  previous SAV populations supported a fishing resource valued in
             excess of a million USA dollars per year.  Nevertheless, estuaries are char-
             acteristically stressed environments, and such  natural stresses  combined with
             anthropogenic factors tend to  make estuarine littoral regions only marginally
             compatible  with SAV physiological requirements.  If a decision were made to
             attempt a reversal  of this SAV decline, it appears that several  steps would be
             required,  first, nutrient inputs should be reduced and efforts  should be made
             to continue and improve soil erosion control practices.  Second, a better
             understanding of propagation and reproduction mechanisms and capabilities of
             these plants is needed.  Finally, a substantial program (possibly involving
             direct transplanting) might be required to accelerate restoration of these SAV
             communities.
IX-18

-------
                                      REFERENCES


 1.   Hutchinson, G.E.   1975.  A treatise in limnology.  Vol. 3.  Limnological
     botany.  John Wiley, New York.

 2.   Thayer, G.W., D.A. Wolfe, and R.B. Williams.  1975.  The impact of man on
     seagrass systems.  Ameri. Scientist 63:288-96.

 3.   Den Hartog, C.  1977.   Structure, function and classification in seagrass
     communities, pp.  89-121 _In:  C.P. McRoy and C. Helfferich (eds.) Seagrass
     ecosystems:  A Scientific perspective.  Marcel Dekker, New York.

 4.   Howard-Williams,  C.  1981.  Studies on the ability of a Potamogeton
     pectinatus community to remove dissolved nitrogen and phosphorus compounds
     from lake water.   J. Appl. Ecol. 18:619-637.

 5.   Orth, R.J.  1977.  The importance of sediment stability in seagrass com-
     munities, pp. 281-300 _In:  B.C. Coull  (ed) Ecology of marine benthos.
     Univ. S. Carolina Press, Columbia.

 6.   Ott, J.A.  1980.   Growth and production in Posidonia oceanica.  Marine
     Ecol.  (P.S.Z.N.I.) 1:47-64.

 7.   Milne, L.J. and M.J. Milne.  1951.  The eelgrass catastrophe.  Sci.
     Amer. 184:52-55.

 8.   Rasmussen, E.  1977.  The wasting disease of eelgrass (Zostera marina)
     and its effects on environmental factors and fauna, pp. 1-52 In: C.P.
     McRoy and C. Helfferich (eds.) Seagrass ecosystems:  A scientTTic per-
     spective.  Marcel Dekker, New York.

 9.   Den Hartog, 9. and P.J.G. Polderman.  1975.  Changes in the seagrass
     populations of the Dutch Waddenzee.  Aquat. Bot. 1:141-147.

10.   Phillips, G.L., D. Eminson and G. Moss.  1978.  A mechanism to account
     for macrophyte decline in progressively eutrophicated freshwaters.  Aquat.
     Bot. 4:103-126.

11.   Bayley, S., V.D.  Stotts, P.P. Springer, J. Steenis.  1978.  Changes in
     submerged aquatic macrophyte populations at the head of the Chesapeake
     Bay, 1958-1975.  Estuaries 1:74-85.

12.   Stevenson, J.C. and N.M. Confer.  1978.  Summary of available information
     on Chesapeake Bay submerged vegetation.  U.S. Dept. Inter. FWS/OBS-78/66.
     NTIS, Springfield, VA.  333 pp.
                                     IX-19

-------
f  r
   f
                13.   Orth, R.J. and K.A. Moore.   1981.   Submerged aquatic  vegetation  of  the
                     Chesapeake Bay:  Past, present and  future.  Trans.  N. Amer. Wildl.  N«t.
                     Res. 46:271-283.

                14.   Brush, G.S.,  F.W. Davis and  C.A.  Stenger.   1981.  Sediment accumulation
                     and the history of submerged aquatic vegetation in  the Patuxent  and Ware
                     Rivers:  A stratigraphic study.   US EPA  Final Rep., Grant No. R80668001,
                     Annapolis, MO.

                15.   Boynton, W.R., W.M. Kemp and J.C. Stevenson.  1979.   A research  design
                     for understanding and managing complex environmental  resource systems,
                     pp. 169-177  In:  R.R. Johnson and J.F. McCormick (eds.) Strategies  for
                     protection ancT management of flood  plain wetlands and other riparion
                     ecosystems.   (GTR-WO-12).  U.S. Dept. Agric., Forest  Serv., Washington,
                     DC.

                16.   Kemp, W.M., M.L. Lewis, J.J. Cunningham, J.C. Stevenson, W.R. Boynton.
                     1980.  Microcosms, macrophytes, and hierarchies:  Environmental  research
                     in Chesapeake Bay, pp. 911-936   In:  J, Giesy (ed) Microcosm research in
                     ecology.  ERDA Conf. 781101.  NTT?, Springfield, VA.

                17.   Kemp, W.M., J.C. Means, T.W. Jones  and J.C. Stevenson.  1982.  Herbicides
                     in Chesapeake Bay and their  effects on submerged aquatic vegetation, pp.
                     503-567.   In:  U.S. Environ. Protect. Agen. Chesapeake Bay Program  tech-
                     nical studTes:  A synthesis.  U.S.  Govt. Printing Office, No. 509-660.
                     Washington, DC.

                18.   Newby, L.C.,  R.A. Kahrs, K.  Adams and M. Szolics.   1978.  Atrazine  resi-
                     dues in the Chesapeake Bay.  Proc.  N.E. Weed Sci. Soc. 32:339 (Abstract).

                19.   Zahnow, E.W.  and J.C. Riggleman.  1980.  Search for Unuron residues in
                     tributaries of the Chesapeake Bay.  J. Agric. Food  Chem. 28:974-978.

                20.   Means, J.C.,  R. Wijayaratne  and W.R. Boynton.  1983.  Fate and transport
                     of selected pesticides in estuarine environments: Implications for  ocean
                     pollution.  Can. J. Fish. Aquat.  Sci. (In press).

                21.   Wu, T.L.  1980,  Dissipation of the herbicides atrazine and alachlor in
                     a Maryland corn field.  J. Environ. Qual. 9:459-465,

                22.   Hershner, C., K. Ward and J. Illowsky.   1981.  The effects of atrazine on
                     Zostera marina in the Chesapeake  Bay, Virginia.  Rept. to US EPA. Virginia
                     Instit. Marine Sci., Gloucester Pt., VA.

                23.   Jones, T.W.,  W.M. Kemp, J.C. Stevenson and  J.C. Means.  1982.  Degradation
                     of atrazine in estuarine water/sediment systems and selected soils.  J.
                     Environ. Qual. 11:632-638.

                24.   Means, J.C. and R. Wijayaratne.   1982.  Role of natural colloids in
                     transport of  hydrophobic pollutants.  Science 215:968-970.
                                                     IX-20

-------
25.  Kemp, W.M., W.R. Boynton, R.R. Twilley, J.C. Stevenson and J.C. Means
     (eds.).  1982.  Submerged aquatic vegetation in upper Chesapeake Bay.
     Final Report Univ.  Md. Center Environ. Estuar. Stud. Ref. No. HPEL 82-193,
     Cambridge, MD.

26.  Forney, D.R. and  D. Davis.  1981.  Effects of low concentrations of
     herbicides on submersed aquatic plants. Weed Sci. 29:677-85.

27.  Correll, D.L. and T.L. Wu.  1982.  Atrazine toxicity to submersed vascular
     plants in simulated estuarine microcosms.  Aquatic Bot. 14:151-158.

28.  Wolman, G.W. and A.P. Shick.  1967.  Effects of construction and fluvial
     sediment form urban and suburban areas of Maryalnd.  Water Resources Res.
     3:451-464.

29.  Roberts, W.P. and J.W. Pierce.  1974.  Sediment yield in the Patuxent
     River (MD) undergoing urbanization, 1968-1969.  Sedimentary Geol.  12:179-
     197.

30.  Roberts, W.P. and J.W. Pierce.  1976.  Deposition in upper Patuxent Estu-
     ary, Maryland, 1968-1969.  Est. Coast. Mar. Sci. 4:267-280.

31.  Kemp, W.M., W.R. Boynton, J.C. Stevenson and J.C. Means (eds.).  1981.
     Submerged aquatic vegetation in Chesapeake Bay.  Univ. Md.  Center
     Environ.  Estuar. Stud., Ref. No. HPEL-81-28, Cambridge, MD.

32.  Goldsborough, W.J.   1983.  Light and shade adaptation in the submerged
     vascular plant, Potamogciton perfoliatus.  MS Thesis, Univ. MD, College
     Park.

33.  Mihursky, J.A. and W.R. Boynton.  1978.  Review of P--tuxent Estuary data
     base.  Univ. MD Center Environ. Estuar. Stud.  Ref. No. 78-157-CBL,
     Solomons, MD.

34.  Lorenzen, C.J.  1972.  Extinction of light in the ocean by phytoplankton.
     J. Cons. Int. Expor. Mer. 34:262-267.

35.  Heinle, D.R., C.F.  D'Elia, J.L. Taft, J.S. Wilson, M. Cole-Jones, A.B.
     Caplins and I.E. Cronin.  1980.  Historical review of water quality and
     climatic data from Chesapeake Bay with emphasis on effects of enrichment.
     Univ. MD Center Environ. Estuar. Stud. Ref. No. 80-15-CBL, Solomons, MD.

36.  Carignan, R. and J. Kalff.  1979.  Quantification of sediment phosphorus
     available to aquatic macrophytes. J. Fish. Res. Bd. Can. 36:1002-1005.

37.  lizumi, H. and A. Hattori.  1981.  Growth and organic production of eel-
     grass (Zostea marina) in temperate waters of the Pacific Coast of Japan.
     III.  The kinetics of nitrogen uptake.  Aquat. Bot.  12:245-256.

38.  Jupp, B.P. and D.H.N. Spence.  1977.  Limitations on macrophytes in a
     eutrophic lake, Loch Leven.  J. Ecol. 65:175-186.
                                     IX-21

-------
39.  Sand-Jensen, K.  1977.  Effect of ep.iphytes on eelgrass photosynthesis.
     Aquat. Bot. 3:55-63.

40.  Penhale, P.  1977.  Macrophyte-epiphyte biomass and productivity in an
     eelgrass (Zostera marina) community.  J. Exp. Mar. Biol. Ecol. 26:211-224.

41.  McNulty, J.K.  1961.  Ecological effects of sewage pollution  in Biscayne
     Bay, Florid i.  Bull. Mar. Sci. Gulf Carib. 11:394-447.

42.  Correll, D.L.  1981.  Nutrient mass balances for the watershed, head-
     waters, intertidal zone, and basin of the Rhode River Estuary.  Limnol.
     Oceanogr. 26:1142-1149.

43.  O'Brien, W.J. and F. deNoyelles.  1974.  Relationship between nutrient
     concentration, phytoplankton density, and zooplankton density in nutrient
     enriched experimental ponds.  Hydrobiologia 44:105-125.

44.  Edmondson, W.T. and Y.H. Edmondson.  1947.  Measurements of production in
     fertilized salt water.  J. Mar. Res. 6:228-245.

45.  Menzie, C.A.  1979.  Growth of the aquatic plant Myriophyllum spicatum in
     a littoral area of the Hudson River estuary.  Aquat. Bot. 6:365-375.

46.  Moss, B.  1976.  The effects of fertilization and fish on community struc-
     ture and biomass of aquatic macrophytes and epiphytic algal populations:
     An ecosystem experiment.  J. Ecol. 64:313-342.

47.  Mulligan, H.F., A. Baronowski and R. Johnson.  1976.  Nitrogen and phos-
     phorus fertilization of aquatic vascular plants and algae in  replicate
     ponds.   I.  Initial response to fertilization.  Hydrobiologia 48:109-116.

48.  Scott, B.C.  1978.  Phytoplankton distribution and light attenuation in
     Port Hacking estuary.  Aust. J. Mar. Freshwater Res. 29:31-44.

49.  Bannister, T.T.  1974.  A general theory of steady state phytoplankton
     growth in a nutrient saturated mixed layer.  Limnol. Oceanogr.  19:13-30.

50.  Borum, J. and S. Wiurn-Anderson.  1980.  Biomass and production of epiphytes
     on eelgrass (Zostera marina) in the jDresund, Denmark.  Ophelia (Suppl.)
     1:57-64.

51.  Sand-Jensen, K. and M. Stfndergaard.  1981.  Phytoplankton and epiphyte
     development and their shading effect on submerged macrophytes in lakes of
     different nutrient status.   Int. Rev. ges. Hydrobiol.  66:529-552.

52.  Lind, C.T. and G. Cottam.  1969.  The submerged aquatics of University
     Bay:  A study in eutrophication.  Amer, Midi. Nat. 81:353-369.

53.  Stuckey, R.L.  1978.  The decline of lake plants.  Natural History 87:66-
     69.
                                      IX-22

-------
•I  I
 1
 '.  f
                54.   Dale,  H.M.  and  G.E.  Miller.   1978.   Changes  in  the aquatic macrophyte
                     flora  of  Whitewater  Lake near Sudbury,  Oitario  from 1947 to 1977.   Canad.
                     Field-Naturalist  92:264-270.

                55.   Fere's, J.M.  and J. Picard.   1975.   Causes  de la rarefaction et de
                     la  disparition  des herbiers  de Posidonia oceam'ca  sur les cotes Francaises
                     de  la  Mediterranee.   Aquatic  Bot.  1:13" 3-139^

                56.   Cambridge,  M.L.   1979.  Cockburn Sound  study technical  report on seagrass.
                     Dept.  Conservation and  Environment,  Western  Australia.   Rep.  No. 7.

                57.   Kemp,  W.M.,  A.J.  Hermann and  W.R.  Boynton.   1981.   Resource dynamics and
                     ecology of  submerged aquatic  vegetation in Chesapeake Bay:   A modeling
                     approach  to demonstrate resource management  concepts.   Univ.  MD Center
                     Environ.  Estuar.  Stud.  Ref.  No.  HPEL-81-214,  Cambridge,  MD.

                58.   Ginsburg, R.N.  and H.A. Lowenstam.   1958.  The  influence of marine bottom
                     communities  on  the depositional environment  of  sediments.  J. Geol.  66:
                     310-318.

                59.   Anderson, R.R.   1966.   Plant  ecology  of the  upper  Patuxent River Estuary
                     with  special  reference  to  the effects of thermal pollution on macrophytes.
                     Ph.D.  Thesis, Univ.  MD, College Park, 99 p.

                60.   Jaworski, N.A., D.W. Lear  and 0. Villa. 1972.   Nutrient management  in
                     the Potomac estuary, pp. 246-273 In:  G.E.  Likens  (ed)  Nutrients and
                     eutrophication.  Am. Soc.  Limn.  Oceanog. Publ., Allen Press,  Lawrence,
                     KA.

                61.   Ritchie,  D.E. and J.B.  Genys.  1975.  Daily  temperature and salinity of
                     surface waters  of Patuxent River at Solomons, MD,  based on 39 years  of
                     records.  Ches. Sci. 16:127-133.

                62.   Haller, W.T., D.L. Sutton  and W.C.  Barlowe.   1974.  Effects of salinity
                     on  growth of several aquatic  macrophytes.   Ecology 55:891-894.

                63.   Schubel,  J.R. and D.J.  Hi rschberg.   1978.   Estuarine graveyards, climatic
                     change and  the  importance  of the estuarine environment,  pp. 285-303.  In;
                     M.L.  Wiley  (ed) rstuarine  interactions. Academic  Press, New York.

                64.   Kitfrboe,  T.   1980.   Distribution and production of submerged macrophytes
                     in  Tippen Grund,  and the  impact of waterfowl  grazing.  J. Appl. Ecol.
                     17:675-687.

                65.   Orth,  R.J.   1975. Destruction of  eelgrass,  Zostera marine, by the cownose
                     ray,  Rhinoptera bonasus,  in  the Chesapeake Bay"!CFfes.  Sci. 16:205-208.

                66.   Stewart,  B.A.  1975.  Control of water  pollution from croplands, Vol. 1.
                     USDA  Rept.  ARS-H-5-1, US  EPA Rept.  EPA-600/2-75-026a.  NTIS, Springfield,
                     VA, 111 p.
                                                     IX-23

-------
67.  Biggs, R.B.  1970.  Sources and distribution of suspended sediment in
     northern Chesapeake Bay.  Mar. Geol. 9:187-201.

68.  Yarbro, L.A., P.R. Carlson, R. Crump, J. Chanton, T.R. Fisher, N. Burger
     and W.M. Kemp.  1981.  Seston dynamics and a seston budget for the Chop-
     tank River Estuary in Maryland.  Univ. Md. Center Environ. Estuar. Stud.
     Ref. No. HPEL-81, Cambridge, MD.

69.  Smullen, J.T., J.u Taft and J. Macknis.  1982.  Nutrient and sediment
     loads to the tidal Chesapeake Bay system, pp. 147-262.  In:  US Environ.
     Protect. Agen. Chesapeake Bay Program technical studies:~A" synthesis.  US
     Govt. Printing Office No. 509-660.  Washington, DC.

70.  Boynton, W.R. and K.L. Heck.  1982.  Ecological role and value of sub-
     merged macrophyte communiteis, pp. 428-502.  In:  US Environ. Protect.
     Agen. Chesapeake Bay Program technical studies:  A syntheses.  US Govt.
     Printing No. 509-660.  Washington, DC.

71.  Kahn, J.R.  1981.  The economic losses associated with depletion of sub-
     merged aquatic vegetation in Chesapeake Bay.  Ph.D. Thesis.  Univ. MD,
     College Park, 120 p.
                                      IX-24

-------

-------