<>EPA
United States
Environmental Protection
Agency
600R06058
science for a changing world
USGS Fate and Effects of
Nitrogen and Phosphorus in
Shallow Vegetated
Aquatic Ecosystems
-------
EPA/600/R-06/058
June 2006
Fate and Effects of Nitrogen and
Phosphorus in Shallow Vegetated Aquatic
Ecosystems
James F. Fairchild
U.S. Geological Survey
Columbia Environmental Research Center
4200 New Haven Road
Columbia, Missouri 65201
Leigh Ann Vradenburg
Currently with: Willow Creek Reclamation Project
28 Tyndal Street
Monte Vista, Colorado 81144
Interagency Agreement No. DW14938599-01
Project Officer
Timothy J. Canfield
Ground Water and Ecosystems Restoration Division
National Risk Management Research Laboratory
Ada, Oklahoma 74820
National Risk Management Research Laboratory
Office of Research and Development
U.S. Environmental Protection Agency
Cincinnati, Ohio 45268
-------
Notice
The U.S. Environmental Protection Agency through its Office of Research and
Development funded, managed, and collaborated in the research described here
under contract IAG DW14938599-01 to the USGS, Columbia, MO. It has been
subjected to the Agency's peer and administrative review and has been approved for
publication as an EPA document. Mention of trade names or commercial products
does not constitute endorsement or recommendation for use.
All research projects making conclusions or recommendations based on en-
vironmental data and funded by the U.S. Environmental Protection Agency are
required to participate in the Agency Quality Assurance Program. This project did
not involve the collection or use of environmental data and, as such, did not require
a Quality Assurance Project Plan.
-------
Foreword
The U.S. Environmental Protection Agency is charged by Congress with protecting the Nation's land, air,
and water resources. Under a mandate of national environmental laws, the Agency strives to formulate
and implement actions leading to a compatible balance between human activities and the ability of natural
systems to support and nurture life. To meet this mandate, EPA's research program is providing data
and technical support for solving environmental problems today and building a science knowledge base
necessary to manage our ecological resources wisely, understand how pollutants affect our health, and
prevent or reduce environmental risks in the future.
The National Risk Management Research Laboratory is the Agency's center for investigation of technologi-
cal and management approaches for preventing and reducing risks from pollution that threatens human
health and the environment. The focus of the Laboratory's research program is on methods and their
cost-effectiveness for prevention and control of pollution to air, land, water, and subsurface resources;
protection of water quality in public water systems; remediation of contaminated sites, sediments and
ground water; prevention and control of indoor air pollution; and restoration of ecosystems. NRMRL
collaborates with both public and private sector partners to foster technologies that reduce the cost of
compliance and to anticipate emerging problems. NRMRL's research provides solutions to environmental
problems by: developing and promoting technologies that protect and improve the environment; advanc-
ing scientific and engineering information to support regulatory and policy decisions; and providing the
technical support and information transfer to ensure implementation of environmental regulations and
strategies at the national, state, and community levels.
This publication has been produced as part of the Laboratory's strategic long-term research plan. It is
published and made available by EPA's Office of Research and Development to assist the user community
and to link researchers with their clients.
The U.S. Geological Survey's Columbia Environmental Research Center is located in Columbia, MO.
The Center conducts interdisciplinary research on the existing and potential effects of various chemical,
physical, and biological stressors on fish and aquatic communities to provide scientific understanding
and technologies needed to support sound management and conservation of natural resources. The
Center includes five research branches: Ecology, Toxicology, Environmental Chemistry, Biochemistry and
Physiology, and Field Research. Activities are integrated among the Research Branches of the Center
via multi-disciplinary research teams. These research teams conduct research and provide technical
assistance for use by federal and state agencies to predict and evaluate the effects of contaminants
and other stressors on fish, invertebrates, aquatic plants, and other components of aquatic ecosystems.
Research results are widely recognized and applied by national and international research and manage-
ment organizations.
Stephen G. Schmelling, Director
Ground Water and Ecosystems' Restoration Division
National Risk Management Research Laboratory
-------
Contents
Foreword iii
Figures ix
Tables xii
Acknowledgments xiii
Executive Summary xv
Introduction 1
Methods 2
Study Site 2
Corral Construction and Design 2
Dosing 3
Water Chemistry 3
Phytoplankton 4
Periphyton 4
Zooplankton 4
Macrophytes 4
Sediment 5
System Metabolism 5
Statistical Analysis 5
Quality Assurance Summary for Nutrient Analyses 5
STUDY 1: Effects of N:P Ratio and Biweekly Nutrient Loading 7
Experimental Design 7
Results 7
Macrophytes 7
Macrophyte Biomass 7
Macrophyte Taxa 10
Macrophyte Nutrients 10
Macrophyte Nutrient Stock 12
Water Chemistry 13
Phosphorus 13
Nitrogen 16
Nitrogen:Phosphorus Ratio 22
pH 23
Alkalinity and Hardness 24
Conductivity 26
Turbidity 27
Phytoplankton 28
Periphyton 33
Zooplankton 34
Sediment 37
System Metabolism 39
Net Nutrient Balance 41
-------
STUDY 2: Effects of Dosing Prior to Macrophyte Development 43
Experimental Design 43
Results 43
Macrophytes 43
Macrophyte Taxa 43
Macrophyte Biomass 45
Macrophyte Nutrients 46
Macrophyte Nutrient Stock 48
Water Chemistry 50
Phosphorus 50
Nitrogen 52
Nitrogen:Phosphorus Ratio 55
pH 56
Alkalinity and Hardness 57
Conductivity 59
Turbidity 60
Phytoplankton 61
Periphyton 68
Zooplankton 69
Sediment 72
System Metabolism 73
Net Nutrient Balance 75
STUDY 3: Effects of Dosing in Relation to Macrophyte Stage 77
Experimental Design 77
Results 77
Macrophytes 77
Macrophyte Taxa 77
Macrophyte Biomass 79
Macrophyte Nutrients 80
Macrophyte Nutrient Stock 82
Water Chemistry 84
Phosphorus 84
Nitrogen 87
Nitrogen:Phosphorus Ratio 91
pH 92
Alkalinity and Hardness 93
Conductivity 95
Turbidity 96
Phytoplankton 97
Periphyton 101
Zooplankton 102
Sediment 103
System Metabolism 105
Net Nutrient Balance 106
VI
-------
Discussion 107
Nutrient Dissipation Rates 107
Nutrient Stocks 107
Phytoplankton Dynamics 107
Phytoplankton Biomass and Growth Rates 107
Phytoplankton Species Composition 108
Periphyton Dynamics 109
Periphyton Biomass and Growth Rates 109
Macrophyte Dynamics 110
Macrophyte Biomass and Growth Rates 110
Macrophyte Species Composition 111
Macrophytes as Nutrient Sinks 112
Macrophytes as Nutrient Sources 112
Zooplankton Dynamics 113
Sediments as Nutrient Sources and Sinks 113
Conclusions, Management Implications, and Research Needs 115
Literature Cited 117
Appendix 1 121
Figures 126
Tables 141
VII
-------
Figures
STUDY 1
1. Pond and corral diagram for Study 1 experiments experiments indicating corral
orientation and diameter 8
2. Changes in macrophyte biomass over time 9
3. Changes in macrophyte nitrogen content over time 10
4. Changes in macrophyte phosphorus content over time 11
5. Changes in macrophyte nitrogen stock over time 12
6. Changes in macrophyte phosphorus stock over time 13
7. Changes in total phosphorus by phosphorus and nitrogen dose levels over time 14
8. Changes in soluble reactive phosphorus by phosphorus dose levels overtime 15
9. Changes in soluble reactive phosphorus dissipation rates over dose periods 16
10. Changes in total nitrogen by nitrogen dose levels over time 17
11. Total and nitrate nitrogen over the season pooled by N-dose level 18
12. Changes in nitrate by nitrogen dose levels over time 19
13. Changes in nitrate dissipation rates over dose periods 19
14. Nitrate concentrations by N-dose over the 4-week extended monitoring
periods at the midseason and the end 20
15. Changes in ammonia by nitrogen dose levels overtime 21
16. Changes in the ratio of total nitrogen to total phosphorus by treatment overtime 22
17. Changes in pH by nitrogen dose levels overtime 23
18. Changes in alkalinity by nitrogen dose levels overtime 24
19. Changes in hardness by nitrogen dose levels overtime 25
20. Changes in conductivity by nitrogen dose levels over time 26
21. Changes in turbidity by nitrogen dose levels over time 27
22. Changes in phytoplankton chlorophyll by treatment over time 28
23. Changes in phytoplankton abundance of the four dominant divisions overtime 32
24. Changes in periphyton chlorophyll accrual rates overtime 33
25. Changes in total numbers of copepods over time 34
26. Changes in zooplankton species richness over time 35
27. Changes in sediment nitrogen pool overtime 37
28. Changes in sediment phosphorus pool overtime 38
29. Changes in community gross oxygen production by phosphorus dose levels
overtime 39
30. Changes in community gross respiration of oxygen by phosphorus dose levels
overtime 40
IX
-------
STUDY 2
31. Corral and pond diagram for Study 2 experiments indicating orientation of corrals 44
32. Changes in macrophyte biomass over time 45
33. Changes in macrophyte nitrogen content over time 46
34. Changes in macrophyte phosphorus content over time 47
35. Changes in macrophyte nitrogen stock over time 48
36. Changes in macrophyte phosphorus stock over time 49
37. Changes in total phosphorus concentrations over time 50
38. Changes in soluble reactive phosphorus concentrations over time 51
39. Changes in total nitrogen concentrations over time 52
40. Changes in nitrate concentrations overtime 53
41. Changes in nitrate dissipation rates over dose periods 53
42. Changes in ammonia over time 54
43. Changes in the ratio of total nitrogen to total phosphorus over time 55
44. Changes in pH overtime 56
45. Changes in alkalinity over time 57
46. Changes in hardness overtime 58
47. Changes in conductivity overtime 59
48. Changes in turbidity over time 60
49. Changes in phytoplankton chlorophyll concentrations overtime 61
50. Changes in particulate organic carbon concentrations overtime 62
51. Changes in the ratio of particulate organic carbon to phytoplankton chlorophyll
overtime 63
52. Changes in phytoplankton abundance of the dominant divisions over time 63
53. Changes in accrual rates of periphyton chlorophyll in 1- and 2-week exposures
overtime 68
54. Changes in zooplankton abundance overtime 69
55. Changes in the sediment nitrogen pool over time 72
56. Changes in the sediment phosphorus pool overtime 73
57. Changes in community oxygen and respiration production overtime 74
STUDY 3
58. Corral and pond diagram for Study 3 experiments indicating orientation of corrals 78
59. Changes in macrophyte biomass overtime 79
60. Changes in macrophyte nitrogen content over time 80
61, Changes in macrophyte phosphorus content over time 81
62. Changes in macrophyte nitrogen stock over time 82
63. Changes in macrophyte phosphorus stock over time 83
64. Changes in total phosphorus concentrations over time 84
-------
65. Changes in soluble reactive phosphorus concentrations overtime 85
66. Changes in soluble reactive phosphorus dissipation over dose periods 86
67. Changes in total nitrogen concentrations over time 87
68. Changes in nitrate concentrations over time 88
69. Changes in nitrate dissipation over dose periods 89
70. Changes in ammonia concentrations over time 90
71. Changes in the ratio of total nitrogen to total phosphorus overtime 91
72. Changes in pH over time 92
73. Changes in alkalinity over time 93
74. Changes in hardness overtime 94
75. Changes in conductivity over time 95
76. Changes in turbidity over time 96
77. Changes in phytoplankton chlorophyll concentrations over time 97
78. Changes in particulate organic carbon concentrations overtime 98
79. Changes in the ratio of particulate organic carbon to phytoplankton chlorophyll
over time 99
80. Changes in phytoplankton abundance of the dominant divisions 100
81. Changes in accrual rates of periphyton chlorophyll in 1- and 2-week exposures
overtime 101
82. Changes in zooplankton abundance over time 102
83. Changes in sediment nitrogen pool overtime 103
84. Changes in sediment phosphorus pool over time 104
85. Changes in community oxygen and respiration production overtime 105
XI
-------
Tables
1. Summary of Recovery Data for Nutrient Analyses Recoveries 6
2. Target Dose Ratios (N:P) for Treatments Receiving Both Nutrients in the Early Dosing
Period of Study 1 8
3. arget Dose Ratios (N:P) for Treatments Receiving Both Nutrients in the Late Dosing
Period of Study 1 8
4. List of Phytoplankton Species Collected during Study 1 29
5. List of Zooplankton Species Collected during Study 1 36
6. Summary Table of the Final Store of P and N in Water and Macrophytes in Study 1 41
7. List of Phytoplankton Species Collected during Studies 2 and 3 64
8. List of Zooplankton Species Collected during Studies 2 and 3 70
9. Summary Table of the Final Store of P and N in Water and Macrophytes in Study 2 75
10. Summary Table of the Final Store of P and N in Water and Macrophytes in Study 3...106
XII
-------
Acknowledgments
Rachael Bredfelt, Dan Carnegy, Brian Fuhr, Curt Gately, Mandi Hapner, Travis
Hill, David Hughes, Lynne Johnson, Julie Kisely, Patti Kohn, Ben Lakish, Steve
Olson, Jennifer Riddle, and Chris Witte provided technical assistance in data collec-
tion, entry, and analysis during the course of the study. Bill Mabee (Missouri Dept.
of Conservation, Columbia, MO) and John Beaver (BSA Environmental Services,
Beechwood, OH) identified the zooplankton. John Beaver (BSA Environmental,
Beechwood, OH) and Cheryl Gilpin (New Braunfels, TX) identified the phytoplankton.
Dr. William Richardson (USGS, LaCrosse, Wl), Nile Kemble (USGS, Columbia, MO),
and three anonymous reviewers from the U.S. EPA provided insightful reviews of
the report. This research was funded by the U.S. Environmental Protection Agency
under IAG DW14938599-01.
XIII
-------
Executive Summary
Nitrate concentrations have greatly increased in streams and rivers draining
agricultural regions of the Midwestern United States. Increasing nitrate transport
to the Gulf of Mexico has been implicated in the hypoxic conditions that threaten
the productivity of marine fisheries. Increases in nitrate concentrations have been
attributed to a combination of factors including agricultural expansion, increased
nitrogen application rates, increased tile drainage, and loss of riparian wetlands.
These landscape-level changes have resulted in a decreased natural capacity for
nitrogen uptake, removal, and cycling back to the atmosphere. Land managers
are increasingly interested in using wetland construction and rehabilitation as a
management practice to reduce loss of nitrate from the terrestrial systems. Yet,
relatively little is known about the limnological factors involved in nitrate removal
by wetland systems.
We conducted a series of studies from 1999-2000 to investigate the functional
capacity of shallow, macrophyte-dominated pond wetland systems for uptake,
assimilation, and retention of nitrogen (N) and phosphorus (P). We evaluated
four factors that were hypothesized to influence nutrient uptake and assimilation:
1) nitrate loading rates; 2) nitrogen to phosphorus (N-.P) ratios; 3) frequency of
dosing/application; and 4) timing of dose initiation.
Nutrient assimilation was rapid; more than 90% of added nutrients were re-
moved from the water column in all treatments. Neither variation in N:P ratios
(evaluated range: <13:1 to >114:1), frequency of application (weekly or bi-weekly),
nor timing of dose initiation relative to macrophyte development (0%, 15-25%,
or 75-90% maximum biomass) had significant effects on nutrient assimilation or
wetland community dynamics. Maximum loading of nitrate (60 g N/m2; 2.4 g P/m2)
applied as six weekly doses stimulated algal communities, but inhibited macrophyte
communities.
Predicted shifts from a stable state of macrophyte- to phytoplankton-domi-
nance did not occur due to nutrient additions. Macrophytes, phytoplankton, and
the sediment surface were all significant factors in the removal of nitrate from the
water column. Overall, these shallow, macrophyte-dominated systems provided
an efficient means of removing nutrients from the water column. Construction or
rehabilitation of shallow, vegetated wetlands may offer promise as land manage-
ment practices for nutrient removal in agricultural watersheds.
xv
-------
Introduction
Anthropogenic eutrophication of water bodies has been a major aquatic research and management focus since the
1950's (Jansson et al. 1994; Smith, 2003). Point sources of nutrients, such as effluents from municipal and industrial
facilities, have largely been identified and controlled via mechanical and engineering approaches to water pollution pre-
vention. However, non-point sources of nutrients to water bodies have continued to rise due to expanded agricultural
activities, increased application of fertilizers, fossil-fuel combustion, over-application of manure to crops, and runoff from
urban areas (Vitousek et al. 1997; Carpenter et al. 1998). These non-point sources of nutrients are continuing to rise
due to difficulties in source identification, lack of effective nutrient management strategies, and the lack of regulatory
focus (Carpenter et al. 1998).
Shallow aquatic systems such as wetlands and ponds can act as sinks for nutrients, thereby significantly decreasing
watershed export (Vitousek et al. 1997). Jansson et al. (1994) suggest that shallow ponds can provide the best means
of nitrogen retention through sedimentation, uptake by vegetation, and denitrification. Assimilative processes may be
facilitated in shallow environments because of the high surface areas of sediments and aquatic plants compared to
pelagic systems (Gasith and Hoyer 1998). Phosphorus may likewise be assimilated into vegetation or retained in sedi-
ments during periods of high nutrient loading. However, under senescent or anoxic conditions, sediments may act as
a nutrient source and result in release of nutrients to the water column (Scheffer 1998).
The establishment of macrophyte stands in shallow systems can increase nutrient retention and recycling (de Haan et
al. 1993). During the growing season, macrophytes act as a sink by accumulating nutrients in developing tissues (Engel
1990). Weisner et al. (1994) demonstrated that removal of nitrate from the water column was significantly higher in
vegetated than non-vegetated mesocosms due to uptake and denitrification. Macrophytes stimulate denitrification by
lowering the redox potential in microzones at the sediment surface and releasing dissolved organic carbon. Therefore,
shallow ponds utilized to reduce nutrients in surface waters may be most effective if macrophyte communities develop
and persist (Jansson et al. 1994).
One factor that may diminish the establishment and persistence of macrophyte stands is a dense community of phyto-
plankton that may develop with nutrient enrichment. Lake and reservoir investigations have generally found an inverse
relationship between macrophyte and phytoplankton communities where two alternative conditions may exist: 1) a mac-
rophyte-dominated system containing clear water and low phytoplankton biomass, or 2) high phytoplankton biomass,
with turbid water and poor macrophyte development. These "alternative stable states" of macrophyte or phytoplankton
dominance are relatively persistent and do not readily alternate unless conditions are disrupted by external or internal
forces (Scheffer 1990, 1998).
Nutrient ratio is a primary determinant of primary production in aquatic systems (Sakamoto 1966, Wetzel 1983). Optimum
ratios of nitrogen:phosphorus (N:P ratio) are approximately 13 (massimass basis) in aquatic systems; ratios that are
under 10 are generally considered nitrogen-limited, and ratios above 17 are generally considered phosphorus-limited
(Redfield et al. 1963; Sakamoto 1966). Thus, aquatic systems that receive nutrient inputs near the optimum level will
achieve a maximum level of primary productivity with efficient utilization of nutrients and minimal dissolved nutrient ac-
cumulation in the water column. Nutrient ratios that are limiting in one nutrient frequently exhibit elevated dissolved forms
of the nutrient in excess. Thus, nutrient uptake and retention is maximal when the N:P ratio is near the 13:1 optimum.
Nutrients other than nitrogen or phosphorus can be limiting (e.g., silica, carbon dioxide) in some systems, but most are
typically limited by phosphorus or nitrogen. Other factors can cause departures from expectations based on ratios of
dissolved nutrients. For example, internal sources of nutrients from sediments are an especially critical component in
shallow aquatic systems such as wetlands (Sand-Jensen and Borum 1991; Scheffer 1998). In addition, intense grazing
of phytoplankton by zooplankton (such as in the absence of fish predators) can increase turnover rates of phosphorus
and sustain productivity under conditions of low P supply and low algal biomass (Wetzel 1983).
Another primary factor influencing whether a system is macrophyte or phytoplankton- dominated is nutrient loading
(Scheffer 1998). When total phosphorus is below 20 ug P/L, algal turbidity and shading are minimal, thereby allowing
for the proliferation of the macrophyte community (Mjelde and Faafeng 1997). Conversely, at high N and/or P loadings,
algal biomass can rapidly increase beyond zooplankton grazing demands and thus dominate aquatic systems due to
shading of macrophytes. Dominance, however, is not absolute because other factors such as non-algal turbidity, water
depth, and season can alter predictions and outcomes (Scheffer 1998).
-------
Timing of nutrient loading, rather than the absolute amount of nutrient input, may also be influential in the determination
of macrophyte or phytoplankton dominance. Algal communities stimulated by nutrient enrichment early in the grow-
ing season can result in significant shading effects and thus hinder development of macrophytes (Phillips et al. 1978).
However, some macrophytes may out-compete phytoplankton by early season accumulation and storage of available
nutrients (Ozimek et al. 1990). It has been demonstrated that established macrophyte stands can maintain dominance
despite increases in loading (Balls et al. 1989). Such communities may respond with a change in composition to tall-
growing species that are better able to compete with epiphytes and phytoplankton shading (Moss 1990). Established
macrophyte stands can also reduce the amount of nitrogen in the water column, thereby inhibiting algal taxa that are
not able to fix atmospheric nitrogen. These nitrogen decreases may be the result of macrophyte uptake or the facilita-
tion of denitrification. Much less is known regarding the uptake and assimilation of phosphorus in shallow vegetated
systems (Scheffer 1998).
There are other biological factors that may influence the relative contribution of algal and macrophyte communities to
aquatic productivity and nutrient cycling. For example, shallow ponds and wetlands may not support fish communities
because of extreme temperature fluctuations and low dissolved oxygen (Bronmark and Hansson 1998). In the absence
of fish predation, large-bodied zooplankton frequently dominate and exert extreme grazing pressure on the phytoplankton
(Brooks and Dodson 1965) which can promote water clarity and increase growth and stability of macrophyte communities
(Moss 1995, Scheffer 1998). Zooplankton can exert variable grazing pressure on phytoplankton, though, and therefore
may not always be inversely related to phytoplankton biomass (Mitchell et al. 1988). Grazing may be ineffectual in
controlling a filamentous algal community, which is less palatable to grazers than smaller-celfed micro-algal species
(Mayer et al. 1997). It has also been observed that nitrogen-limited systems may have decreased grazing pressure by
zooplankton due to proliferation of large and generally unpalatable cyanobacteria (Jensen et al. 1994).
Much of the research regarding eutrophication of aquatic systems has been conducted using fertilization experiments.
Studies have demonstrated that macrophytes show a variable response to nutrient loading, and that the relative ca-
pacity of a system for nutrient retention may depend on the resulting dominant community. Mulligan et al. (1976) used
experiments at two fertilization levels in shallow ponds without fish to evaluate the fate of added nutrients and effects
on the macrophyte community. With the highest load, they found that dense communities of phytoplankton inhibited
or eliminated macrophyte development. Balls et al. (1989) conducted enrichment experiments in constructed ponds
to explore the mechanisms of macrophyte loss in local water bodies that had lost submerged plant communities. In
their experiments, macrophytes strongly buffered against all levels of nutrient enrichment and maintained dominance
whether or not fish were present; however, experimental treatments included several phosphorus levels but only one
nitrogen level. Stachowicz et al. (1994) fertilized a field pond over several years across various states of macrophyte
and phytoplankton dominance to evaluate nutrient retention. They found that phosphorus retention was high under
both phytoplankton and macrophyte dominance; however, macrophytes were far more effective at reducing the export
of nitrogen. Therefore, research has demonstrated the complexities between nutrient enrichment and community
interactions; yet few studies have comprehensively evaluated the full range of factors that may influence the uptake,
assimilation, and retention of nitrogen and phosphorus in experimental wetland systems. Such extensive studies are
logistically complex, but are necessary to isolate the relative effects of nutrient loading, ratios, frequency, and timing on
retention and community dynamics (Moss 1995; Havens et al. 1999).
The objective of this study was to systematically evaluate the assimilative capacity of shallow, vegetated experimental
wetlands for the uptake, removal, and retention of nitrogen and phosphorus. Three factors were evaluated: 1) the effect
of N:P ratios; 2) the effect of loading or dosing rates of N and P; and 3) the effect of timing of dosing of N and P. The
studies were conducted to explore the response of these experimental systems to nutrient manipulation under con-
trolled, experimental conditions. The results are provided to explore the functional utility of using constructed wetlands
as mitigation tools for removal of nutrients in runoff from agricultural watersheds.
Methods
Study Site
Studies were conducted at the experimental mesocosm facility located at the U.S. Geological Survey's Columbia En-
vironmental Research Center (CERC), Columbia, MO. This facility was constructed in 1968 to provide a controlled
experimental complex to evaluate the effects of environmental stressors on shallow aquatic systems. Individual impound-
ments are approximately 0.1 ha in area and range in depth from 0.1 to 1.5 m. Macrophyte communities are dominated
by Najas guadalupensis and Chara sp. Physical, chemical, and biological characteristics of the mesocosms have been
previously described (Fairchild et al.1992,1994; Fairchild and Sappington, 2002).
Corral Construction and Design
Circular corrals were used as the experimental treatment unit because they are highly replicable and reduce statisti-
cal variation typical of whole mesocosms. Multiple corrals were placed within each of 4 mesocosms; the mesocosm
served as the experimental block.
-------
Corrals were constructed of impermeable Scrimweave™ (StoCote Products, Chicago, IL) to create a circular enclosure
of approximately 4-m diameter. Corral walls were secured to a circular ring of 2.5 cm diameter black polyethylene water
pipe supported on steel fence posts (driven outside the corral) to maintain the upper edge of the corral approximately
20 cm above the water surface. The bottom edges of the corral wall were wrapped outward beneath a piece of circular
metal garden edging, which was then driven approximately 8 cm into the sediments before the ponds were filled with
water. The sides of the corrals were then weighted down with bricks while the ponds were filled with well water over a
2-d period. Once flooded, the mesocosms were allowed to mix and allow mobile biota to freely move within the system.
Prior to dosing, the corral edges were raised and secured to effectively isolate the contents (water, sediment, and biota)
within each individual corral.
Water exchange between the corral and the outside water was minimal as indicated by visual inspection (i.e., turbidity
during wading on the outside of the corral) and analytical chemical data. Ponds were occasionally refilled with water
during the season to maintain an average depth near 1 meter; water additions were conducted during non-critical pe-
riods of the dose/monitoring schedule to minimize artifacts of corral management. Depths ranged among the corrals
from 0.74-1.16 m and averaged 0.91 m. Levels of water in each individual corral remained similar to that outside of
the corral due to slow water diffusion through the sediments.
Dosing
The ranges of dose concentrations were selected based on the range of published spring and summer concentrations
of N and P from Midwestern streams subject to agricultural runoff (Hauck et al. 1997). Magnitude of dosing was cho-
sen based on literature reviews of studies (Johnston 1991; Mitsch et al. 1999) that indicate that natural wetlands can
assimilate a range of 0.03 - 28 g N /m2/yr and 0.07-3.48 g P/m2/yr depending on a range of factors including wetland
type, depth, vegetative structure, and hydrologic residence time.
Granular agricultural fertilizers (soda of nitrate and triple super phosphate) were used to dose the corrals. Amend-
ments were calculated according to corral volumes and the percent of available N and P in the fertilizer. Fertilizer was
pre-weighed, placed into a cotton bag, and agitated under the surface of the water inside the perimeter of the corral
for approximately 5 minutes. The water was then gently mixed with a paddle to ensure nutrient distribution. Once
granules were mostly dissolved, the bag was suspended in the water column to allow for release of residual nutrients
in the material. Laboratory experiments prior to the start of the study indicated that nutrients were rapidly released into
the water column. Rapid nutrient dissolution was also verified by measured nutrient concentrations in the corrals (see
results). Nutrient bags were specific for each corral and were used throughout the experiment. Control corrals were
similarly mixed to prevent experimental bias due to the physical disturbance of mixing.
Water Chemistry
Water samples were collected with a tube sampler (cylindrical sampler 7.62 cm diameter by 1 m length; vol.= 4560 cm3)
deployed in a rapid, vertical motion to collect a depth-integrated water sample. Three vertical samples were composited
in a clean 20-L polyethylene bucket. The composite was thoroughly mixed and then sub-sampled with a 1-L polyethyl-
ene bottle. The 1 -L samples were immediately chilled on ice and transported to the laboratory for analysis. Unfiltered
samples were kept on ice or refrigerated until processed or analyzed. All samples were analyzed within recommended
time limits according to EPA standards (U.S. ERA 1979).
Approximately 250 mL of each water sample was filtered for dissolved nutrients using a 0.45 urn membrane (47 mm
nitrocellulose filter; Whatman Inc., Clifton, NJ). All filtering equipment was thoroughly rinsed with deionized water be-
tween samples to minimize contamination. The filtrate (for dissolved constituents) was stored refrigerated (< 4°C) in
60-mL Nalgene HOPE bottles or 10-mL capped, borosilicate disposable tubes until analysis.
Soluble reactive phosphorus (SRP) was determined using a color reagent and an MR 1201 Spectrophotometer as
described by U.S. EPA Method 365.3 (U.S. EPA 1979). A Lachat 8000 Flow Injection Analyzer (Lachat Instruments,
Milwaukee, Wl) was used for analysis of nitrogen. The combined total of nitrate (NO3-N) and nitrite (NO2-N) was mea-
sured by a colorimeter after cadmium reduction to nitrite (Lachat Instruments 1997). Ammonia (NH3-N) was measured
colorimetrically after reactions with alkaline phenol, sodium hypochlorite, and sodium nitroprusside (Lachat Instruments
1997).
Unfiltered water was analyzed for total phosphorus and total nitrogen. Samples to be analyzed for total nitrogen and
total phosphorus were stored frozen (<0°C) in 60-mL HOPE bottles. Before analysis, the samples were thawed, mixed,
and pipetted into 10-mL glass tubes. Samples for total phosphorus were oxidized using potassium persulfate (Fisher
Scientific, Fairlawn, NJ) and then analyzed for orthophosphorus on an MR 1201 Spectrophotometer (U.S. EPA 1979). The
spectrophotometer was also used to determine total nitrogen as N after persulfate oxidation (Crumpton et al 1992).
The pH of each sample was determined using an Orion Model SA 290A pH meter in accordance with the manufac-
turer's recommendations. Alkalinity was determined by burette titration with 0.02 N sulfuric acid and expressed as mg
-------
CaCO3/L Hardness (mg CaCO3/L) was measured using a color indicator and burette titration with EDTA (APHA 1992).
Conductivity was determined using an YSI Model 33 S-C-T meter (YSI Corp., Yellow Springs, OH) and expressed as
uS/cm at room temperature (20-30°C). The HACH Model 2100A Turbidimeter (Hach Co., Loveland, CO) was used to
estimate the turbidity (NTU's).
Phytoplankton
Phytoplankton biomass was based on the chlorophyll a content of algae. A measured amount of sample (25-250 ml)
was filtered through a 47 mm glass fiber filter (Gelman type A/E; Fisher Scientific, Fairlawn, NJ). The filter was then
placed in a 15-ml vial of 90% buffered acetone and refrigerated overnight for extraction. This extract was subsequently
analyzed using a fluorometer (Turner Designs 10-AU-Fluorometer; La Jolla, CA) using EPA Method 445.0 (APHA 1992).
Particulate organic carbon samples were filtered onto a 47-mm Gelman type A/E filter and then combusted and ana-
lyzed using a Coulometrics Model 2010 Total Carbon Analyzer (UIC Corporation, Wheaton, IL). Phytoplankton were
sampled and preserved for taxonomic analysis on a monthly basis by preservation of 40 ml_ of the unfiltered water
sample using 1 mL Lugol's solution. Phytoplankton were counted and identified by John Beaver of BSA Environmental
Services, Beachwood, OH.
Periphyton
Periphyton biomass and accrual rates in the corrals were evaluated as chlorophyll concentrations extracted from growth
on artificial substrates (1 cm by 10 cm strips of Scrimweave™ suspended vertically just below the surface of the wa-
ter). In June of 1999, four strips were exposed for four weeks and then analyzed individually for chlorophyll. Future
exposures were shorter to reflect the rapid growth of periphyton that was observed. In July of 1999, four strips were
deployed. Two replicate strips were collected after 1-week and 2-week exposure intervals. In August and September,
six strips were deployed, and three strips were collected and analyzed as a composite after each of 1 -week and 2-week
exposures. Strips were carefully collected with forceps in the field and immediately put on ice in vials of 15 mL of 90%
buffered acetone. Periphyton chlorophyll was estimated using the same methods as for the phytoplankton; however,
values were expressed as accrual rates (ug Chl/cm2/wk).
Zooplankton
Monthly zooplankton samples were collected. On May 12, 1999, (Study 1) zooplankton samples were collected using
a 63 urn Wisconsin net and vertical tows to effectively sample a 10-L volume. Thereafter, zooplankton were sampled
using vertical migration samplers modified from the design of Whiteside et al. (1978). Samplers consisted of a funnel
and 2-L bottle assembly inverted and positioned in the water column just above the macrophyte layer. The funnel and
bottle used were clear so as to minimize avoidance due to darkened conditions. Samplers were deployed at dusk and
retrieved at dawn. These samplers passively trapped zooplankton during diurnal feeding movements. On retrieval, the
samplers were poured through a 63 urn Wisconsin net to isolate the zooplankton. Samples were stored in 90% ethanol.
Samples were analyzed by Bill Mabee, Missouri Dept. of Conservation, Columbia, MO. Sample numbers were then
calculated on an area basis by dividing zooplankton number by the surface area of the funnel surface.
Macrop/iytes
Macrophytes were qualitatively assessed each month based on visual assessment and ranking of the benthic plant and
filamentous algae communities; separate estimates were made within each of four quadrants of each corral. Assess-
ments included estimates of percent cover, height, species composition, and color. There were only two species of
macrophytes in the corrals (Chara sp. and Najas guadalupensis) which were easily distinguishable based on color and
morphometrics. Chara sp. is a macroalgae and has an upright and branched thallus, and is attached to the substrate
by rhizoids (Smith 1950; Kufel and Kufel 2002). Najas sp. is a submerged, branched macrophyte.
Macrophytes were quantitatively sampled each month from pre-set, buried standardized rings to minimize disturbance
and sampling bias. Each ring (5 cm height; 10 cm diameter) was cut from a cross-section of white PVC water pipe.
Replicate sampling rings were deployed in each of four corral quadrants to account for spatial variation within each cor-
ral. Prior to the study initiation, the rings were pushed into the sediments until flush with the top of the sediment layer.
This technique made the rings easy to locate, but minimized shading or enclosure effects.
Monthly composite macrophyte samples were collected (one ring from each of the four quadrants) from each corral
by divers wearing Neoprene wet suits. Wet suits allowed the divers to maintain neutral buoyancy and caused minimal
disturbance to surrounding sediments and macrophytes (Madsen 1993). Collection involved diving to locate a ring and
digging underneath it with a Plexiglas board. The board created a bottom for the encircled sediment and macrophyte
sample, and enabled it to be brought to the surface for careful processing. Macrophyte material originating from the
area enclosed by the ring was collected, including all above and belowground biomass. Composites of four rings per
corral were stored in plastic bags on ice during transport to the lab. In the lab, the macrophytes were washed on a
small mesh screen (<1 mm mesh) and any debris or attached sediment was carefully removed. Samples were then
-------
placed in pre-weighed aluminum foil packets, dried at 105°C, and weighed to get an estimate of dry weight biomass
(Madsen 1993). Biomass was expressed as dry weight (g/m2). A Wiley Mill was used to grind the dried samples, which
were then stored in airtight vials.
Dried and ground macrophyte samples were subsequently analyzed for total nitrogen and total phosphorus content.
Samples analyzed for N content were weighed (0.2 g) and then combusted in a LEGO FP-528 Analyzer. This appara-
tus transformed sample nitrogen to N2, which was then measured by thermal conductivity detection and expressed as
percent of dry weight.
Total phosphorus in macrophyte tissues was determined in pre-weighed samples (0.2 g) using perchloric acid diges-
tion (6% perchloric acid) based on the procedure of Sommers and Nelson (1972). This digestion process converts all
phosphorus to orthophosphate in a clear supernatant. Orthophosphate was then determined using the Lachat 8000
FIA (Lachat Instruments, Inc. 2000). The results, expressed as ug P/L, were converted to a percent basis (mass:mass)
normalized to the amount of plant material used in the digestion.
Sediment
Sediment samples were collected concurrently with macrophytes. One sediment plug of the top 2-5 cm of sediment
was taken from each macrophyte ring and deposited in a plastic bag. Care was taken not to include any plant material
with the sediment plug. The composite of sediment plugs for the four rings sampled in each corral was homogenized
by hand and dried in foil pans at 105°C. Dried samples were ground using a Wiley mill and stored in airtight vials.
Methods for analysis of nitrogen and phosphorus content of the sediments were the same as for macrophytes. At the
beginning of the season, four separate rings were placed in the ponds. The sediment enclosed by them was collected,
dried, and weighed to determine an average mass of sediment within a ring. Estimates of sediment nutrient pools (g N
or P/(m2*5 cm deep)) were calculated by multiplying the percent content of nutrient by the average mass of sediment
enclosed by the sampling ring, and then converting to a square meter of surface area.
System Metabolism
System metabolism was measured each week as a variation of the diurnal oxygen method outlined by Lind (1985). This
method was chosen over traditional light-dark bottle techniques because evaluations were desired for the total system
including macrophytes, phytoplankton, and sediments. Dissolved oxygen and temperature readings were taken with a YSI
Model 54 Oxygen Meter in every corral on a consecutive morning, evening, and morning sequence. The first readings
of each sequence coincided with water collection. Before field use, probes were calibrated in saturated air according
to manufacturers' specifications. Dissolved oxygen and temperature were measured by submerging the probe to mid-
depth of the water column to ensure homogeneity. All corrals were sampled in less than an hour to decrease temporal
variability. Oxygen readings (mg O2/l_) were designated M1 (morning 1), E1 (evening 1), and M2 (morning 2). Gross
production (GP) was calculated by: GP = (E1-M1)+(E1-M2). Gross respiration (GR) was calculated by: GR=2*(E1-M2)
as a modification of Lind (1985).
Statistical Analysis
Data were tested for normality of distribution by treatment using Proc Univariate in the Statistical Analysis System,
Release 6.12 (1996). The strong seasonal nature of the data (see Appendix 1) resulted in data that were not normally
distributed with homogeneous variance. Therefore, all datasets were subsequently transformed using the rank pro-
cedure prior to analysis (Conover and Iman 1981). Although some statistical power was lost by rank transformation,
this method provided the best means of analyzing all of the datasets uniformly (Snedecor and Cochran 1967, Green
1979). Transformed data were analyzed as a randomized complete block design using Analysis of Variance (ANOVA)
to determine influences due to N-dose, P-dose, time, and their interactions (using pond as the block and corrals as ex-
perimental units). When ANOVA indicated significant main effects, we statistically compared individual treatments using
the Student-T test. Significant differences between rank-transformed values were determined at the p<0.05 level.
Quality Assurance Summary for Nutrient Analyses
A summary of quality assurance results for nutrient analyses is presented in Table 1. Results indicated that recovery of
spiked standards ranged from 79-108% across the two years of study. Recoveries were within the range of acceptable
results for these analyses.
-------
Table 1. Summary of Recovery Data for Nutrient Analyses Recoveries. Numbers Represent Mean + 1 Standard
Deviation. Number in Parenthesis is Number of Independent Standards Analyzed Each Year
Nutrient
NH9
N02N03
SRP
TN
TP
Year
1999
90.6 + 15.2(63)
87.6 + 19.0(64)
94.2 + 4.6 (64)
85.9 + 21.1 (30)
94.3 + 2.0 (30)
2000
85.9+14.7(38)
95.4+15.7(38)
108 + 18.6(38)
79.1 +19.4(18)
101.6 + 3.7(18)
-------
STUDY 1: Effects of N:P Ratio and Biweekly Nutrient Loading
The nitrogen to phosphorus ratio (N:P ratio) is known to be a primary'determinant of the response of phytoplankton
to nutrient enrichment in deep, lentic aquatic ecosystems (Sakamoto 1966; Wetzel 1983). However, far less is known
regarding the response of phytoplankton and macrophyte communities in shallow, vegetated wetlands (Scheffer 1998).
Thus, the first experiment performed in 1999 evaluated the effects of N:P ratios on nutrient cycling, system metabolism,
and structural dynamics of the experimental mesocosms.
There were three objectives in Study 1:1) determine how nutrient loading and the N:P ratio influenced the concentra-
tion and relative distribution of nutrients in the water column, macrophytes, and sediments; 2) evaluate how nutrient
load and the N:P ratio influenced species composition, biomass, and/or abundance of macrophytes, phytoplankton,
and periphyton; and 3) characterize the assimilation and retention capabilities of shallow ponds for nutrient loads under
varying N:P ratios.
Experimental Design
Ponds were drained the April 15, 1999, and corrals were constructed over a 2-week interval. A total of 36 corrals
were constructed with nine corrals within each of 4 replicate ponds (experimental blocks). Ponds were reflooded with
well water on May 1 and allowed to biologically re-colonize from sediments. On May 18, the sides of the corrals were
raised above the water surface and secured to isolate each individual corral and its contents. Water sampling began
on May 20.Three levels of nitrogen (0, low, and high) and three levels of phosphorus (0, low, and high) were studied
in a balanced factorial design as described in Figure 1. Two dosing phases were evaluated: Early ( May 20-July 27,
1999) (Table 2) and Late (July 29-September 21, 1999) (Table 3). In the Early phase (Table 2), targeted nominal
concentrations of nitrogen ranged from 0 to 5 mg N/L; whereas, targeted nominal phosphorus concentrations ranged
from 0 to 88 ug/L. Thus, N:P ratios ranged from approximately 16:1 (approximate Early ambient conditions) to 114:1.
In the Late phase (Table 3), the amount of phosphorus was increased to standardize the N:P ratio at 13:1 (optimum N:
P ratio for productivity), 25:1 (lower end of phosphorus limitation), and 50:1 (high phosphorus limitation). Over the 12
weeks, the total loads of added nutrients ranged from 0 g NO3-N/m2 and 0 g P/m2 in the Control, to 30 g NO3-N/m2 and
0.864 P/m2 in the treatment receiving the highest concentrations of both nutrients. The frequency (6 doses) and timing
(every two weeks) of dosing were held constant and were initiated on June 2, 1999, when macrophytes had attained
approximately 25% surface cover as viewed from above.
Corrals were dosed six times during the study (Early 6/2/99, 6/16/99, 6/30/99; and Late 7/28/99, 8/11/99, 8/25/99). In
all additions, target concentrations for nitrate were 0, 2.5, and 5 mg NO3-N/L for ambient, low, and high treatments, re-
spectively. Phosphorus was applied in two phases constituting the Early (first three doses) and Late season (last three
doses). Target levels were 0, 44, and 88 ug P/L in the Early season, and 0, 100, and 200 ug P/L in the Late season.
The nitrogen or phosphorus portion of the dose was referred to as N or P, and for ambient target levels, the designa-
tion ON or OP was used. In some parameter analyses, P-dose did not significantly influence the results. For those
analyses, the three phosphorus treatments were combined within the nitrogen treatment for ease of presentation, i.e.,
(ON:OP), (ON:loP), and (ON:hiP) were averaged to create ON. Likewise, in analyses where P-dose was determined to
be the stronger influence, all nitrogen treatments (OP, loP, and hiP) were combined. Relative P-designations were used
for simplicity of presentation because dosing amounts were changed during the season. Analyses were conducted
across the entire season, and on subsets of the Early (May 20-July 27) and Late (July 29- September 21) seasons to
evaluate treatment effects.
Results
Macrophytes
Macrophyte Biomass
Macrophytes grew rapidly from May to July until a maximum biomass of approximately 800 g/m2 was reached; thereafter,
macrophyte biomass decreased during the remainder of the study in all treatments (Figure 2). There was a statistically
significant effect of Day on macrophyte biomass during the Early-, Late-, and Full-season analysis. However, there were
-------
4 meters _
diameter
©oo
ooo
ooo
Pondl
OOO
ooo
ooo
Pond 2
OOO
ooo
ooo
PondS
Pond 4
Figure 1. Pond and corral diagram for Study 1 experiments indicating corral orientation and diameter. Pond 4
shows an example of the random assignment of the nine treatments.
Table 2. Target Dose Ratios (N:P) for Treatments Receiving Both Nutrients in the Early Dosing Period of Study 1.
Treatments Consisting of at Least One Ambient Level are not Represented Because Initial Ambient Levels
for Both nutrients were at or below the Method Limit of Detection
^
0)
^_
Ambient
44
QD
OO
Ambient
•4ii.. .- ~-£' .
-i'.-; -If,".;' -? •-;*-.-•• •-
'•5-,' f ' : •',', :'.:•', ::••'•.""
' '"%lfc/. '•"•*; „,;„ ;"- - '
'•'•' ... ,. • "**-->--s „„ "'" A ,,
N (mg/L)
2.5
.'.."'• ,/r ';''~ ' J> ' • '?*' '•''*-••••
57:1
on-1
£.0. i
5
' ''• :l>!'f -:
114:1
KV-1
O/ . I
Table 3. Target Dose Ratios (N:P) for Treatments Receiving Both Nutrients in the Late Dosing Period of Study 1.
Treatments Consisting of at Least One Ambient Level are not Represented Because Initial Ambient Levels
for Both Nutrients were at or below the Method Limit of Detection
O)
3.
o.
Ambient
100
200
N (mg/L)
Ambient
',. ^*;^&',,, , '•*•'*$','.$$&;
2.5
:±;"' ^'^':diS :~^;^ •
25:1
13:1
5
Wf3
50:1
25:1
-------
no significant main effects of N or P dosing. In May, all treatments had initial dry weight biomass averaging 135 g/m2.
The midseason point (July 27) coincided with maximum macrophyte biomass, which reached 661 g/m2 (5 times initial
levels) in the ON treatment. However, there were no significant differences among N treatments at peak biomass levels
reached in July. Senescence during the Late season (August and September) resulted in a loss of one-third of the bio-
mass in ON, and final stands were significantly smaller than the maximums (p<0.05). Treatments 2.5N and 5N lost over
two-thirds of their maximum stand, and by September, had macrophyte biomass levels, 206 and 241 g/m2, respectively,
that were significantly lower (p<0.05) than ON (419 g/m2). Differences between 2.5N and 5N were not significant. Nutri-
ents appeared to slightly increase both macrophyte growth rates and rates of senescence; however, overall main effects
of nutrients were not significant. On the last sampling date, there were significant differences among treatments when
evaluated using a single LS Means test in which the 2.5N and 5N treatments contained significantly lower macrophyte
biomass than ON, implying that nutrient addition enhanced decomposition processes late in the study.
Variability in macrophyte biomass was caused to some degree by variation in depth across the experimental pond blocks.
Topography of the pond bottom varied from 0.91 to 1.16 m depth across locations in the 36 individual corrals. Macro-
phyte biomass levels "crashed" to zero in two of the deepest corrals in pond 4 in August. These corrals were treated
with intermediate levels of nutrients (2.5N:loP and 5N:loP) which would indicate that the crashes were not related to
necessarily high levels of enrichment (5N or hiP) or a particular TN:TP ratio. Rather, these two corrals were proximal
to each other in the deepest section of the pond. Therefore, we conclude that the macrophyte crashes were probably
due to a factor of location which was light-limited due to depth as opposed to algal-generated turbidity.
Macro.
Biomass
Full
Early
Late
N-dose
P-dose
N-dose *
P-dose
D
0.0001
0.0001
0.0001
N-dose *
Day
0.0447
o 0496
P-dose
Day
N-dose *
P-dose * Day
Macrophyte Biomass by N-Dose Level, for Study 1
September
0
N-dose 2.5
5
M
130
134
137
J
366
433
528
Month
j
661
772
802
A
585
470
365
S
419"
206b
241 b
Season
Avg.
422
395
412
Figure 2. Changes in macrophyte biomass over time. The upper table presents probabilities that dose, day, and
interactions of dose and day influenced macrophyte biomass in ANOVA of rank-transformed data. Dark-
ened values are not significant (p>0.05). The graph is a plot of macrophyte biomass over the entire
experimental season. Because N-dose*Day was a significant influence in ANOVA, values are pooled
by N-dose in the graph. Dark circles on X axis mark dose dates. The lower table lists LS Means (g dry
weight/m2) represented in the graph, along with statistical information based on the rank transformed data.
Within a column, values sharing a letter are not significantly different (p>0.05). In columns without letters,
values are not significantly different.
-------
Macrophyte Taxa
Qualitative, visual observations of the macrophyte and algal communities during Study 1 revealed seasonal succession,
although variability within and across treatments was high. Chara, an attached macroalgae, dominated most corrals
in early June. These populations declined by late June, and were not detectable by late July. The rooted macrophyte
community consisted of only one genus, A/a/as, throughout the season. These plants succeeded Chara in all corrals
except two that experienced biomass crashes (treatments 2.5N:loP and 5N:loP). In those cases, Chara populations
were replaced by filamentous algae, and Najas re-growth never occurred. Filamentous algal growth was observed in
all of the corrals on at least one date; however, filamentous biomass was qualitatively greater in treated corrals com-
pared to Control ponds. Filamentous algae, however, was not a major contributor of total macrophyte biomass in any
treatment.
Macrophyte Nutrients
Day was a significant main effect for both N (Figure 3) and P (Figure 4) content of macrophytes. Both N and P content
of macrophyte tissue significantly increased over the season in all treatments in 1999. Prior to dosing, macrophytes
averaged 1.89% N (Figure 3). Initial samples were composed of both Chara sp. and Najas guadalupensis. The mea-
sured N concentrations were intermediate between published literature values for similar species including Chara
vulgaris (2.43-3.19% N; Dykyjova and Kvet 1982) and Najas maritima (1.05-1.87% N; Royle and King 1991). There
was a significant main effect of N-dose on nitrogen content of macrophytes (p< 0.05) during the Late- and Full-season
analysis. The nitrogen content of ON macrophytes increased from 1.95% N to 3.02% N during the season. Following
dose initiation, macrophytes in N-dosed treatments had higher N content than ON, and differences among treatments
N-dose
P-dose
N-dose *
P-dose
Day
N-dose '
Day
P-dose'
Day
N-dose *
P-dose * Day
Macro.
%N
Full
00011
00001
00432
Early
•MMNiM**
0.0001
Late
0.0001
00493
Macrophyte % N, for Study 1
September
N °
dose
M
1.95
1.85
1.86
J
1.73"
1.99""
2.18b
Month
J
2.31a
2.28"
2.75"
A
2.95"
3.38ab
3.64"
S
3.02a
3.81b
3.95b
Season
Avg.
2.39a
2.66"
2.87C
Figure 3. Changes in macrophyte nitrogen content over time. The upper table presents probabilities that dose, day,
and interactions of dose and day influenced N content in ANOVA of rank-transformed data. Darkened
values are not significant (p>0.05). The graph is a plot of N content over the entire experimental season.
Because N-dose and N-dose*Day were significant influences in ANOVA, values are pooled by N-dose in
the graph. Dark circles on X axis mark dose dates. The lower table lists LS Means (% N of dry weight)
represented in the graph, along with statistical information based on the rank transformed data. Within a
column, values sharing a letter are not significantly different (p>0.05). In columns without letters, values
are not significantly different.
10
-------
N-dose
P-dose
N-dose *
P-dose
Day
N-dose *
Day
P-dose *
Day
N-dose *
P-dose * Day
Macro.
%P
Full
00001
0.0378
Early
0.0001
Late
0.0324
Macrophyte % P, for Study 1
September
0
P-dose io
hi
M
0.23
0.21
0.21
J
0.28
0.23
0.23
Month
J
0.44
0.41
0.41
A
0.38"
0.47"
0.46b
S
0.43"
0.47ab
0.52b
Season
Avg.
0.35
0.36
0.37
Figure 4. Changes in macrophyte phosphorus content over time. The upper table presents probabilities that dose,
day, and interactions of dose and day influenced P content in ANOVA of rank-transformed data. Darkened
values are not significant (p>0.05). The graph is a plot of P content over the entire experimental season.
Because P-dose*Day was a significant influence in ANOVA, values are pooled by P-dose in the graph.
Dark circles on X axis mark dose dates. The lower table lists LS Means (% P of dry weight) represented
in the graph, along with statistical information based on the rank transformed data. Within a column,
values sharing a letter are not significantly different (p>0.05). In columns without letters, values are not
significantly different.
increased during the season. The final nitrogen concentrations in 2.5N (3.81% N) and 5N (3.95% N) were nearly 30%
higher than final levels in the ON treatment (3.02% N; significant difference p<0.05). There was no significant effect of
P-dose on nitrogen content of macrophytes.
Macrophytes contained an average phosphorus content of 0.22% P at the beginning of the study (Figure 4). As with
the percent N, that value was an intermediate between published literature values for Chara vulgaris (0.36-0.46% P;
Dykyjova and Kvet 1982) and Najas guadalupensis (0.16% P; Boyd 1970). Day had a significant effect on P content
of macrophytes during the Early- and Full-season comparisons. P-dose had a significant effect on P content of mac-
rophytes during the Late season (p<0.05). The phosphorus content of OP macrophytes increased by 0.2% P from May
to July, but did not change substantially during senescence. Macrophytes in loP and hiP were similar in P content to
OP during the growing season, but continued to accumulate phosphorus in senescence. Phosphorus content of hiP
macrophytes was nearly 20% higher than OP in August and September (significant difference p<0.05). The P content
in loP was only 10% greater than that in the OP treatment, and was only significantly higher in August (p<0.05). N-dose
had no significant effects on P content of macrophytes.
11
-------
Macrophyte Nutrient Stock
The stock of N and P in macrophytes, expressed as g N/m2 or g P/m2, was calculated by multiplying the dry weight
biomass (g/m2) by the tissue content (g N/g dry wt or g P/g dry wt). Results are provided in Figures 5 and 6.
Nitrogen stocks in macrophytes followed seasonal trends of increase and decrease similar to biomass patterns. Day
was a significant main effect (p<0.0001) controlling N stock of macrophytes (Figure 5) during the Early dosing, Late
dosing, and Full-season analysis; N-dose had significant effects on N stocks of macrophytes during the Early season
but had no effect during the Late dosing season. Control N stocks increased 7-fold from May (initially 2.3 g N/m2) to
July (average 15.7 g N/m2 in July) (Figure 5). The 2.5N and 5N treatments developed maximum stocks of 17.5 and
21.8 g N/m2, respectively, in July, which were significantly higher than Control levels on the average. During the senes-
cent, Late treatment period nitrogen stocks decreased significantly (p<0.05) and declined to 13.0, 7.7, and 10.9 in the
Control, 2.5N, and 5N treatments, respectively. P-dose had no significant effects on nitrogen stocks of macrophytes.
Phosphorus stock in macrophytes increased 20-fold (from 0.15 to 3.28 g P/m2' in Control corrals during the May-July
Early period and was reflected in a significant main effect of Day. However, there were no significant main effects of
N-dose or P-dose in phosphorus stocks of macrophytes. Phosphorus pools were similar among treatments during all
phases of macrophyte growth. Phosphorus loss during biomass decline was weakly related to N-dose (p=0.11). Dur-
ing senescence, the store of P in ON decreased by 25% to 2.21 g P/m2, but this loss was not significant (Figure 6). In
N-dosed treatments, however, phosphorus stocks in September were <40% of July maximums (significant difference
p<0.05). Final values in 2.5N and 5N were 0.96 and 1.24 g P/m2, respectively, significantly lower than concurrent pools
in ON (p<0.05).
N-dose
P-dose
N-dose *
P-dose
Day
N-dose *
Day
P-dose '
Day
N-dose *
P-dose * Day
N
Stock
Full
0.0001
8.1680 "
Early
0.0265
0.0001
Late
00001
Macrophyte Nitrogen Stock, for Study 1
25 i
June
July
Month
August
September
N °
Hntr ^
5
M
2.3
2.3
2.4
J
6.1
8.4
10.8
Month
J
15.7
17.5
21.8
A
15.7
14.6
13.7
S
13.0
7.7
10.9
Season
Avg.
10.6
10.1
11.9
Figure 5. Changes in macrophyte nitrogen stock over time. The upper table presents probabilities that dose, day,
and interactions of dose and day influenced the N stock in ANOVA of rank-transformed data. Darkened
values are not significant (p>0.05). The graph is a plot of N stock of macrophytes over the entire experi-
mental season. There were no significant influences based on the full season. Because N-dose was
a significant influence in ANOVA of the early season, values are pooled by N-dose in the graph. Dark
circles on X axis mark dose dates. The lower table lists LS Means (g N/nf) represented in the graph.
12
-------
N-dose
P-dose
N-dose *
P-dose
Day
N-dose
Day
P-dose
Day
N-dose *
P-dose * Day
Full
W382
00001
P Stock Early
0.1972 ..
00001
Late
«,»*»
00001
Phosphorus Stock (g P/m2)
40 -i
I 5
Macrophyte Phosphorus Stock, for Study 1
- - A- - - ON l
/•'*"\t-
/'••' \^ * A
// • \x^ A i
si/' ^^* !
<*£•''* *
•^ ^ _ i
U.U 'WWW WWW '
May June July August September
Month
N °
rfnsr ^
5
M
0.27
0.27
0.27
J
0.85
1.11
1.33
Month
j
2.92
3.27
3.33
A
2.49
1.95
1.74
S
2.21
0.96
1.24
Season
A\g.
1.75
1.51
1.58
Figure 6. Changes in macrophyte phosphorus stock over time. The upper table presents probabilities that dose,
day, and interactions of dose and day influenced the P stock in ANOVA of rank-transformed data. Dark-
ened values are not significant (p>0.05). The graph is a plot of P stock of macrophytes over the entire
experimental season. Because N-dose*Day was weakly significant (p<0.1) in ANOVA based on the entire
season, values are pooled by N-dose in the graph. Dark circles on X axis mark dose dates. The lower
table lists LS Means (g P/m2) represented in the graph.
Water Chemistry
Phosphorus
Total phosphorus (TP) in water increased 8-fold in the Control and >14-fold in P or N-dosed treatments over the course
of the study. Day had a highly significant effect on TP (p<0.001), and both N- and P-dose had significant main effects
during the Early- and Full-season analysis; however, there was also significant P-dose*Day and N-dose*Day interac-
tions which complicate the interpretation of the effects of N-dose and P-dose alone. During the Early season, TP in OP
rose from 21 to 54 ug P/L (Figure 7A). In the amended treatments, TP was up to 20 ug P/L higher than OP from dose
initiation to July 7, but midseason values were similar. During macrophyte senescence, TP levels in OP quadrupled to
a final maximum of nearly 250 ug P/L, indicating internal loading from the sediments and/or macrophytes. LoP and hiP
maximums in September were >325 ug P/L, but were not significantly different from each other or OR
When evaluated by N-dose, total phosphorus in ON increased 11 -fold during the season from 20 to 220 ug P/L (Figure 7B).
During the Early season, TP in ON increased from 18 to 45 ug P/L, and N-dosed treatments were within 10% of ON
values. At midseason, TP in N-dosed treatments was 60 ug P/L, a third greater than in the ON Control (45 ug P/L;
significant difference p<0.05). During the Late dosing period, which corresponded with macrophyte senescence, TP
increased in all treatments, and differences between N-dosed and ON treatments increased significantly when compared
on single dates. TP in ON and 2.5N peaked at 222 and 305 ug P/L, respectively, on September 7, and then dropped
15% in both treatments by September 21. TP in 5N on September 7 and 21, was 283 and 366 ug P/L, respectively, but
due to variability among replicates, these levels were not significantly different from the other treatments.
13
-------
N-dose
P-dose
N-dose *
P-dose
Day
N-dose *
Day
P-dose *
Day
N-dose *
P-dose * Day
Full
0.0422
0.0003
0.0001
0.0014
00001
TP Early
0.0341
0.0001
0.0001
0.0056
0.0001
Late
0.0001
TP by P-Dose Level, for Study 1
5/20 6/3 6/17 7/1 7/15 7/29 8/12 8/26 9/9
p °
Lo io
dose hi
5/20
21'
21"
17"
6/1
16
16
16
6/3
17'
28b
37C
6/14
20'
22""
25"
6/24
22'
28b
30b
Dat
7/7
27*
36b
42°
e
7/27
54
56
55
8/9
99
110
107
8/24
108'
136b
143b
9/7
231
247
332
9/21
245
328
244
Season
Avg.
78*
93b
96b
B
TP by N-Dose Level, for Study 1
5/20
6/3
8/26
9/9
0
*- 2.5
dose 5
5/20
18'
22"
19""
6/1
17
17
15
6/3
29
25
27
6/14
25'
23*
19"
6/24
26'
31b
23'
Dat
7/7
35*"
39'
31"
B
7/27
45'
60b
60"
8/9
87'
118b
111'"
8/24
102*
151b
134"
9/7
222°
305b
283"
9/21
188
264
366
Season
Avg.
72*
96b
99'
Figure 7. Changes in total phosphorus by phosphorus and nitrogen dose levels over time. The upper table presents
probabilities that dose, day, and interactions of dose and day influenced TP in ANOVA of rank-transformed
data. Darkened values are not significant (p>0.05). The upper graph (A) is a plot of TP over the entire
experimental season. Because P-dose and P-dose*Day were significant influences in ANOVA, values are
pooled by P-dose in the graph. Dark circles on X axis mark dose dates. The table below Graph (A) lists
LS Means (mg P/L) represented in Graph (A), along with statistical information based on the rank trans-
formed data. Within a column, values sharing a letter are not significantly different (p>0.05). In columns
without letters, values are not significantly different. Because N-dose and N-dose"Day were significant in-
fluences in ANOVA, values are pooled by N-dose in Graph (B). The table below Graph (B) lists LS Means
represented in Graph (B) and follows the format described above.
14
-------
Soluble reactive phosphorus concentrations demonstrated cyclical response patterns due to significant main effects
of both Day and P-dose in Early-, Late-, and Full-season analysis (p<0.0001) (Figure 8). However, there were also
significant P-dose*Day interactions. Initial SRP concentrations (2 ug/L) were at or below the limit of detection. SRP
in the OP treatments fluctuated around the detection level during the first half of the experiment; SRP concentrations
gradually increased during the Late dosing period to an average of 27 ug/L by late September. Peaks in SRP in the
P-dosed treatments reflected the six amendments. During the Early season, SRP concentrations measured 24 hours
after dosing reflected <20% of the calculated additions (i.e., 80% loss/day) which is indicative of the rapid assimilative
capacity of the wetlands for dissolved phosphorus. SRP measurements taken one week after dosing were similar to the
Control values. Calculated dissipation rates, based on weekly declines, were <4 ug P/L/day during the Early study (first
3 doses) but were probably underestimates because additions were rapidly and completely dissipated within that time
period (Figure 9). After midseason, dissipation rates increased (Figure 9). During the Late dosing period there was a
net accumulation of SRP in the P-dosed corrals due to dose modifications and/or internal loading from the sediments
and/or macrophytes (Figure 8). In the week following doses three and four, dissipation rates in hiP were over twice
those in loP (p<0.05), but in neither treatment did SRP return to Control levels. On the final sample date (September 21,
1999), SRP levels in hiP averaged 86 ug/L and were significantly higher than Control (27 ug/L) and loP (57ug/L) treat-
ments (Figure 8). Final SRP concentrations did not differ between loP and hiP treatments.
N-dose
P-dose
N-dose * P-dose
Day
N-dose * Day
P-dose * Day
N-dose * P-dose * Day
Full
o.oooi
o.oooi
0.0001
SRP Early
0.0001
O.OOOI
0.0001
Late
0.0001
0.0001
00001
0.0001
0,9726
I
SRP by P-dose level, for Study 1
180 -,
160
20
0
5/20 5/27 6/3 6/10 6/17 6/24
7/8 7/15 7/22 7/29 8/5 8/12 8/19 8/26 9/2
Date
9/9 9/16
0
P-dose lo
hi
5/20
2
2
2
6/1
3
2
3
6/3
3"
llb
22C
6/8
4
4
4
6/14
3
3
4
6/17
3'
6b
10C
6/24
2
2
3
7/1
2'
5k
14C
7/7
5
6
6
7/14
4
8
5
1
7/27
7
10
g
)aie
7/29
13"
59"
108°
8/3
13'
35"
51"
8/9
11'
28b
40'
8/12
13'
61b
157'
8/17
12'
35"
89C
8/24
8'
29b
70°
8/26
12"
79b
171'
8/31
13"
64"
155"
9/7
17"
46"
109'
9/14
19"
48"
95"
9/21
27"
57"
86°
Season
Avg.
89"
27"
55'
Figure 8. Changes in soluble reactive phosphorus by phosphorus dose levels over time. The upper table presents
probabilities that dose, day, and interactions of dose and day influenced SRP in ANOVA of rank-trans-
formed data. Darkened values are not significant (p>0.05). The graph is a plot of SRP over the entire
experimental season. Because P-dose and P-dose*Day were significant influences in ANOVA, values are
pooled by P-dose in the graph. Dark circles on X axis mark dose dates. The lower table lists LS Means
(mg P/L) represented in the graph, along with statistical information based on the rank transformed data.
Within a column, values sharing a letter are not significantly different (p>0.05). In columns without letters,
values are not significantly different.
15
-------
SRP Dissipation Rates, for Study 1
a
o o
25
20
15
10
0 --
-5
• loP
• hiP
i=H
Dose
Figure 9. Changes in soluble reactive phosphorus dissipation rates over dose periods. Phosphorus dissipation (mg
P/Uday) was calculated from concentrations one day and one week after each dose.
Nitrogen
There were significant main effects of Day, N-dose, and N-dose*Day interactions on total nitrogen (TN) levels in the
study; however, P-dose had no effect on TN levels. TN significantly increased in all treatments during the study and
averaged a 6-fold increase over the 4-month study interval. In ON treatments, TN increased from 0.35 to 2.14 mg N/L
by the end of the study (Figure 10). During macrophyte growth, nitrate constituted 95% and 88% of the total nitrogen in
5N and 2.5N, respectively, due to inputs from the dosing regime (Figure 11). Nitrate values decreased to near Control
values at the end of the Early dosing period (approximately 1 mg/L; July 27) which illustrates the assimilative capacity
for nitrate when macrophytes were rapidly growing. During the Late dosing intervals, which corresponded to the ob-
served period of normal macrophyte senescence, TN in ON, 2.5M, and 5N increased 2, 3, and 4-fold, respectively. In
the Late dosing period, the relative proportion of TN as nitrate was <1% in 2.5N, and dropped from 42% to 9% in 5N,
indicating that a portion of the added nitrate was transformed and maintained in the water column in the organic form.
A net increase of nearly 2 mg N/L in ON treatment during the season indicated that these systems had large, internal
sources of nitrogen derived from sediment stores. In the final samples, TN concentrations in 5N (4.57 mg N/L) were
significantly larger than 2.5N (3.02 mg N/L) and the ON Control (p<0.05).
16
-------
N-dose
P-dose
N-dose *
P-dose
Day
N-dose *
Day
P-dose *
Day
N-dose *
P-dose * Day
Full
0.0001
0.0001
0.0001
TN Early
00001
0.0001
0.0001
0.4482
.9.1774
Late
ooooi
o.oooi
TN by Nitrogen Dose Level, for Study 1
5/20
9/9
0
N- 2.5
dose 5
5/20
0.35
0.39
0.36
6/1
0.41
041
0.40
6/3
0.40a
2.89b
5.33C
6/14
0.52"
079b
1 86C
6/24
0.61'
1.66b
4.10C
Dat<
7/7
0.74"
1.66b
4.54°
1
7/27
0.88"
1.051"
1.18b
8/9
1.01"
1.57b
2.34C
8/24
1.14'
1.95b
3.11C
9/7
1.71*
264b
5.59°
9/21
\ 2.14*
3.02b
457°
Season
Avg.
090*
1.64b
3.03'
Figure 10. Changes in total nitrogen by nitrogen dose levels over time. The upper table presents probabilities that
dose, day, and interactions of dose and day influenced TN in ANOVA of rank-transformed data. Darkened
values are not significant (p>0.05). The graph is a plot of TN over the entire experimental season. Be-
cause N-dose and N-dose*Day were significant influences in ANOVA, values are pooled by N-dose in the
graph. Dark circles on X axis mark dose dates. The lower table lists LS Means (mg N/L) represented in
the graph, along with statistical information based on the rank transformed data. Within a column, values
sharing a letter are not significantly different (p>0.05). In columns without letters, values are not signifi-
cantly different.
Nitrate in the ON treatments never exceeded 0.01 mg NO3-N/L, and was at or below the limit of detection during most of
the study (Figure 12). Nitrate values the day following the first dose showed that target levels of 2.5 and 5 mg NO3-N/L
were achieved by the additions (Figure 12). Statistical analyses confirmed that N-dose was a significant main effect on
nitrate (<0.0001), but that P-dose was not influential. Following dose initiation, nitrate in 5N remained significantly higher
than ON through the end of the season. Nitrate levels in 2.5N were similar to ON at the midseason point (July 27), and
in the last three post-treatment samples.
Nitrate dissipation rates were calculated each week after dosing (Figure 13). During the Early dosing period, uptake in
5N averaged over 0.4 mg NO3-N/L/day (8% applied N loss/day) and was significantly greater than rates in 2.5N which
averaged less than 0.3 mg NO3-N/L/day (12% applied P/day). Following the fourth amendment, when phosphorus
additions were increased (Table 3), the nitrate dissipation only slightly increased (0.42 mg NO3-N/L/day; 8% applied
P loss/day); however, nitrate uptake decreased thereafter. Nitrate dissipation of the last dose in 5N was less than
0.06 mg NO3-N/L/day (1% applied N loss/day). This rate was substantially lower than previous 5N rates and only 15%
of concurrent 2.5N rates.
17
-------
7
6
5
4
3
2
1
0 X-
5/20
TN and NQs-N; Dose=0, for Study 1
-ONNO3 ---A--- ONTN
A-A-
-A—A-
-A—AA—A AA—A-
7/31
6/3
6/17
7/1
7/15
7/29
8/12
8/26
9/9
Date
TN and NQ,-N: Dose=2.S, for Study 1
6/3 6/17 7/1 7/15 7/29 8/12 8/26 9/9
TN and NQ-N by N-dose, for Study 1
[--* 5NN03 ------- 5NTN I
5/20 6/3 6/17 7/1 7/15 7/29 8/12 8/26 9/9
Date
Figure 11. Total and nitrate nitrogen over the season pooled by N-dose level. Nitrate values for ON are not visible
due to scale. Dark circles on X axis mark dose dates.
18
-------
N-dose
P-dose
N-dose * P-dose
Day
N-dose * Day
P-dose * Day
N-dose * P-dose * Day
0.0001
9,8444
O.HSS8
0.0001
0.0001
N
Early
0.0001
0.0001
0.0001
Late
00001
0,8762
0.0001
0.0001
NO3-N by N-dose level, for Study 1
9/9
9/16
N °
i 2'5
dose 5
5/20
0005
0006
0.006
6/1
0005
0005
0005
6/3
0007'
2535b
5080'
6/8
0005'
1.0311
3 120'
6/14
0.005"
0.2 15b
1 528s
6/17
0.005'
31051
7453'
6/24
0.006'
0934'
3735'
7/1
0.007'
2360b
6333'
7/7
0.007'
0.984'
4374'
7/14
0006"
0058b
1.910C
L
7/27
0.005'
0.005"
0.093°
(ate
7/29
0006'
2493b
4.660'
8/3
0006*
0434°
2509C
8/9
0007"
0021'
0949"
8/12
0007'
2.196b
5.187'
8/17
0008"
0665°
4018'
8/24
0.006"
0020°
1 517'
8/26
0009"
2770°
5.229'
8/31
0.007'
0937°
4.957'
9/7
0.006"
0083"
2160"
9/14
0.005"
0.034"
0921°
9/21
0005"
0017"
0.379°
Season
Avg.
0026"
0951°
3.006°
Figure 12. Changes in nitrate by nitrogen dose levels over time. The upper table presents probabilities that dose,
day, and interactions of dose and day influenced NO3-N in ANOVA of rank-transformed data. Darkened
values are not significant (p>0.05). The graph is a plot of NO3-N over the entire experimental season. Be-
cause N-dose and N-dose*Day were significant influences in ANOVA, values are pooled by N-dose in the
graph. Dark circles on X axis mark dose dates. The lower table lists LS Means (mg NO3-N/L) represent-
ed in the graph, along with statistical information based on the rank transformed data. Within a column,
values sharing a letter are not significantly different (p>0.05). In columns without letters, values are not
significantly different.
NOs-N Dissipation Rates, for Study 1
Dose
Figure 13. Changes in nitrate dissipation rates over dose periods. Nitrate dissipation (mg NO3-N/L/day) was calcu-
lated from concentrations one day and one week after each dose. Analyses showed that dissipation rates
were 30- 70% greater in 5N than 2.5N in the first half of the season. Mean dissipation rates of the 6th
dose in 5N were only 15% of concurrent 2.5N rates. Values plotted are LS Means pooled for N-dose, and
bars represent one standard deviation above and below the mean. Stars indicate significant difference
(p<0.05) between 2.5N and 5N treatments.
19
-------
Nitrate dissipation during the midseason inter-treatment period (July 1-27) and at the end of the season (Aug. 26-Sept. 1)
resulted in complete nitrate uptake in 2.5N, and nearly complete uptake in 5N (Figure 14). During both periods, nitrate
concentrations in 2.5N fell below 0.05 mg NO3-N/L within three weeks following the amendment, and were similar to
ambient levels in ON (p>0.05). Nitrate in 5N decreased below 0.4 mg NO3-N/L within four weeks, but remained sig-
nificantly higher (p<0.05) than ON and 2.5N. The sustained uptake of nitrate indicated that the assimilative capacity of
these shallow, macrophyte-dominated systems was still high, but that the time required for nitrate assimilation >5 mg
NO3-N/L was increasing.
o
NO3-N Extended Monitoring, for Study 1
5/20 6/3
6/17
7/15 7/29 8/12 8/26
Date
9/9
Figure 14. Nitrate concentrations by N-dose over the 4-week extended monitoring periods at the midseason and the
end. Dark circles on X axis mark dose dates. Nitrate in 2.5N and 5N was significantly different (p<0.05)
on all but the last day of each period.
20
-------
Ammonia levels in ON treatment fluctuated between the limit of detection (0.005 mg NH3-N/L) and 0.025 mg NH3-N/L
during the study (Figure 15). Both N-dose (p<0.0001) and Day (p<0.0001) had significant main effect on ammonia
concentrations. P-dose had no significant main effect on ammonia. Ammonia generally peaked the week following
additions, resulting in concentrations in 5N that were 2-11 times those in concurrently measured ON corrals. Ammonia
in the 2.5N corrals was intermediate between the ON and 5N levels, and likewise showed periodic increases the week
after dosing. Simultaneous but smaller peaks (<0.02 mg NH3-N/L) also occurred in the ON corrals, indicating a possible
effect of the stirring procedure on sediment release of ammonia. Ammonia in ON averaged 0.009 mg NH3-N/L during
the study. Seasonal averages in the 5N (0.037 mg NH3-N/L) and 2.5N (0.024 mg NH3-N/L) treatments were significantly
greater (four and three times, respectively) than those in the ON treatment.
N-dose
P-dose
N-dose * P-dose
Day
N-dose * Day
P-dose * Day
N-dose * P-dose * Day
NH3-
N
Full
0.0001
0.0001
0.0001
Early
0.0001
0.0001
0.0021
0.0309
Late
0.0001
0.0001
0.0014
NH3-N by N-dose level, for Study 1
0125 7
5/20 5/27
9/9
9/16
N °
L» 2-s
dose 5
5/20
0024
0023
0022
6/1
0005
0.005
0.005
6/3
0005
0.005
0005
6/8
0.005
0005
0.006
6/14
0010*
0014''
0018°
6/17
0.006*
0.014°
0014"
6/24
0.018
0.029
0050
7/1
0009'
0018b
0.025°
7/7
0.014*
0031°
0045"
7/14
0008*
0.008'
00195
Da
7/27
0.005*
0.007*
0.022°
te
7/29
0.008"
0.023*
0.02?
8/3
JHW6"
0019b
0038°
8/9
0006'
Lp013b
0.067e
8/12
0.005"
0.013°
0.042C
8/17
0014*
0023b
0.045Q
8/24
0.010"
0026b
0058'
8/26
0.010"
0022b
0045Q
8/31
0.010"
0053b
0109°
9/7
0.013"
0.037k
0087C
9/14
0005"
0.013°
0037C
9/21
0009"
0017"1
0024b
Season
Avg.
0009*
0.024b
0.037'
Figure 15. Changes in ammonia by nitrogen dose levels over time. The upper table presents probabilities that dose,
day, and interactions of dose and day influenced NH3-N in ANOVA of rank-transformed data. Darkened
values are not significant (p>0.05). The graph is a plot of NH3-N over the entire experimental season. Be-
cause N-dose and N-dose*Day were significant influences in ANOVA, values are pooled by N-dose in the
graph. Dark circles on X axis mark dose dates. The lower table lists LS Means (mg NH3-N/L) represented
in the graph, along with statistical information based on the rank transformed data. Within a column,
values sharing a letter are not significantly different (p>0.05). In columns without letters, values are not
significantly different.
21
-------
Nitrogen:Phosphorus Ratio
Sakamoto (1966) observed an optimum range otTN:TP ratio of approximately 13 (range 10:1-17:1; mass:mass basis)
for algal productivity. Sakamoto (1966) proposed that above a TN.TP ratio of 17:1 algal populations were limited by
phosphorus; aTN:TP ratio less than 10:1 was likely nitrogen-limited. All treatments hadTN:TP ratios >17 during most
of the Early dosing period, indicating they were P-limited. N:P ratios decreased below 10:1 in some treatments (ON:loP,
ON:hiP, 2.5N:loP, 2.5N:hiP) during the Late dosing period which indicates that some nitrogen limitation may have occurred
(Figure 16). The Control (ON:OP) was P-limited through August (TN:TP>17) but approached 15:1 (near optimum) in the
final two samples. Analyses indicated thatTN:TP was significantly influenced by N-dose (p<0.0001), P-dose (p<0.02),
and N-dose*P-dose interactions (p<0.02). In samples taken the day following the first dose, TN:TP ratios in the P-dosed
treatments were 2-3 times larger than predicted based on the loading ratios, whereas ratios in all OP treatments were
within 12% of calculations. Therefore, added nitrogen remained in the water column, while phosphorus additions were
rapidly lost through sedimentation or uptake by periphyton and macrophytes. At midseason, TN:TP ratios in all treat-
ments ranged from 17:1 and 25:1. Following the P-dose increase during the Late dosing period (beginning July 29),
N:P ratios decreased (range 4-30). Internal P-loading (evidenced in SRP and TP data) and P-dose modifications low-
ered ratios to between 10 and 17 in all treatments at some point in the Late season.
TN:TP by Treatment, for Study 1
------- 1 (ON;OP)
I--••-•• 3(ON;loP) --
- ] - - -A- - - 2 (ONjbiP) - -
—• — Sakamoto
- 7 (2.5N;OP) — •— 4 (5N;OP)
- 9 (2.5N,loP) — •— 6 (5N,loP)
- 8 (2.5N;hiP) - - A— 5 (5N;hiP)
5/20 6/3 6/17 7/1 7/15 7/29 8/12 8/26 9/9
TN:TP by Treatment, for Study 1
- 7 (2.5N;OP)
- 9 (2.5N;loP)
- 8 (2.5N,hiP)
4 f5N;OP) U^_L_
6 (5N;loP) *
5 (5N;hiP)
5/20 6/3 6/17 7/1 7/15 7/29 8/12 8/26
Date
9/9
Figure 16. Changes in the ratio of total nitrogen to total by treatment over time. A) TN:TP ratios in all treatments pre-
sented with a heavy dashed line representing the range for optimal algal growth proposed by Sakamoto
(1966). B) The lower graph is the same as (A), but with the scale enlarged to delineate ratios in the late
season. Dark circles on X axis mark dose dates.
22
-------
pH
N-dose, Day, and the N-dose*Day interaction had significant effects on pH; however, P-dose had no significant ef-
fect. Levels of pH in ON increased from 8.4 to 9.7 during the Early season, and then decreased to 9.2 by September
(Figure 17). The pH levels significantly increased (p<0.001) in the N-dosed treatments compared to the ON treatment
within two weeks of the first dose and remained above 9.5 the remainder of the season. This increase in pH occurred
concurrently with the observed increase in primary productivity which is expected as available dissolved carbon dioxide
decreases due to increase photosynthetic uptake of carbon (Wetzel 1983). Thus pH was a good surrogate indicator
of the positive effects of N-dose and Day on primary productivity. However, there were significant interactions among
N-dose and Day during the Early, Late, and Full-study components which indicate that neither main effect, alone, was
solely responsible for observed increases in pH.
N-dose P-dose N-dose * Day
P-dose
Full ooooi Wftifp- llifc!iSl?||fllt o.oooi
pH Early 0.0001 0.0001
Late 0 0001 o.OOOl
N-dose * P-dose * N-dose * P-dose
Day Day * Day
0.0001 WfS4 ^ X ^fUHS! _; A
0.0033 / t|Mt - 8tJ!Wi*k*"
0.0001 -^ &Z&S% ^ ss&ZWl
10.5 -,
10 0 -
D.
ft < -
8.5 !
pH by N-dose Level, for Study 1
i
>fc^ A
J^^^'^^\^^
/'•*''"'
2_mz
jr
- - - - m-.. \
"A.....^
- - A- - • ON 1
— -* — 2.5N j
r
A A A A A A
5/20 6/3 6/17 7/1 7/15 7/29 8/12 8/26 9/9
Date
N °
rincr
5
5/20
84
8.4
8.4
6/1
8.9
8.7
88
6/14
91'
9.5"
97b
7/1
93'
9.6"
98"
Da
7/14
9.7"
10.1b
10.1b
te
7/27
9.5'
9.7*
99b
8/9
95'
9.9b
98b
8/24
9.5*
10.0b
10.0b
9/7
9.3'
98b
99C
9/21
9.2'
93b
9.8C
Season
Avg.
9 ]•
9.1b
92'
Figure 17. Changes in pH by nitrogen dose levels over time. The upper table presents probabilities that dose, day,
and interactions of dose and day influenced pH in ANOVA of rank-transformed data. Darkened values are
not significant (p>0.05). The graph is a plot of pH over the entire experimental season. Because N-dose
and N-dose'Day were significant influences in ANOVA, values are pooled by N-dose (data were averaged
by H-ion concentration, then converted to pH: calculated pH= - log (H-ion). Dark circles on X axis mark
dose dates. The lower table lists pH represented in the graph, along with statistical information based on
the rank transformed data. Within a column, values sharing a letter are not significantly different (p>0.05).
In columns without letters, values are not significantly different.
23
-------
Alkalinity and Hardness
Alkalinities in ON ranged from 97-150 mg CaCO3/L during the study; the lowest values of alkalinity occurred in mid-July
(Figure 18). N-dose, Day, and the N-dose*Day interaction had significant effects on alkalinity, which rose continuously
after dose initiation to over 175 and 250 mg CaCO3/L in 2.5N and 5N, respectively. Alkalinities in the 5N treatment were
significantly higher (p<0.05) than ON and 2.5N from July through the end of the season. Values in the 2.5N were only
significantly higher than ON on July 14 and the last three dates of the season.
Full
Alk. Earl)
Late
N-dose
0.0001
r 0.0033
0.0001
300
18"
a n
* U
£
P-dose
3W34S
0.05). The graph is a plot of alkalinity over the entire experimental season.
Because N-dose and N-dose*Day were significant influences in ANOVA, values are pooled by N-dose
in the graph. Dark circles on X axis mark dose dates. The lower table lists LS Means (mg CaCO/L)
represented in the graph, along with statistical information based on the rank transformed data. Within a
column, values sharing a letter are not significantly different (p>0.05). In columns without letters, values
are not significantly different.
24
-------
Hardness values in ON decreased from May (184 mg CaCO3/L) to a seasonal minimum in mid-July (92 mg CaCO/L),
and then increased to 129 mg CaCO3/L by the end of September (Figure 19). Hardness values were negatively influ-
enced by N-dosing and Day (p<0.0001). The apparent effect of Day corresponds to a decrease in cations in the water
column over time following flooding of the experimental wetlands with CERC well water. Loss of hardness levels over
time is frequently observed in these systems due to precipitation losses as dissolved carbon dioxide decreases and pH
increases due to primary productivity. In treatments receiving either N-dose, hardness values were significantly lower
(p<0.05) than in the ON treatment which corresponds to overall positive effects of N-dosing on primary productivity and
carbon dioxide removal. Hardness did not differ significantly in the 2.5N and 5N treatments, however.
N-dose
P-dose
N-dose * P-
dose
Day
N-dose * Day
P-dose *
Day
N-dose * P-dose *
Day
Full
0.0001
00001
0.0001
Hard. Early
00001
0.0001
0.0001
Late
0.0001
Hardness by N-dose Level, for Study 1
5/20
6/17
7/1
7/15 7/29
Date
8/12
8/26
9/9
N °
Hrwr 2.5
5
5/20
184
182
183
6/1
135
128
133
6/14
110*
104ab
102b
7/1
103a
89b
87b
I
7/14
92"
84"
74C
)ate
7/27
96"
73"
68"
8/9
107"
80b
73"
8/24
104a
66"
66b
9/7
107a
79"
76b
9/21
129"
92b
86b
Season
Avg.
117"
98b
95b
Figure 19. Changes in hardness by nitrogen dose levels over time. The upper table presents probabilities that dose,
day, and interactions of dose and day influenced hardness in ANOVA of rank-transformed data. Darkened
values are not significant (p>0.05). The graph is a plot of hardness over the entire experimental season.
Because N-dose and N-dose*Day were significant influences in ANOVA, values are pooled by N-dose
in the graph. Dark circles on X axis mark dose dates. The lower table lists LS Means (mg CaCO^L)
represented in the graph, along with statistical information based on the rank transformed data. Within a
column, values sharing a letter are not significantly different (p>0.05). In columns without letters, values
are not significantly different.
25
-------
Conductivity
Both N-dose, Day, and N-dose*Day interaction had significant effects on conductivity of the corrals. Conductivity in ON
corrals declined from initial levels of 450 uS/cm to seasonal minimums (297 uS/cm) at mid-season (Figure 20). During
the Late season, ON conductivity steadily rose to 383 uS/cm. Conductivity was contributed by two major factors: 1) well
water used to fill the corrals prior to study initiation, and 2) N-dosing using sodium nitrate. Conductivity decreased early
in the study due to the gradual loss of ions due to precipitation reactions in the water column. N-dosing, initiated in early
June, significantly increased conductivity (p<0.0001) in the 2.5N and 5N treatments due to the influence of sodium (Na)
in the fertilizer. Conductivity in 5N was significantly higher (p<0.05) than both ON and 2.5N from June 14 through the
end of the season, attaining over 600 uS/cm by September. The 2.5N treatments had significantly higher conductivity
than the ON treatments from July 1 through September. Thus, conductivity was an artifact of the experimental treatment
as opposed to a response variable related to eutrophication.
N-dose
P-dose
N-dose *
P-dose
Day
N-dose *
Day
P-dose *
Day
N-dose * P-dose
*Day
Full
0.0001
0.0001
0.0001
••. ,-0.05). The graph is a plot of conductivity over the entire experi-
mental season. Because N-dose and N-dose*Day were significant influences in ANOVA, values are
pooled by N-dose in the graph. Dark circles on X axis mark dose dates. The lower table lists LS Means
(uS/cm) represented in the graph, along with statistical information based on the rank transformed data.
Within a column, values sharing a letter are not significantly different (p>0.05). In columns without letters,
values are not significantly different.
26
-------
Turbidity
The experimental enclosures were minimally affected by wave action, in-flow, and disturbance. Therefore, turbidity
values primarily reflected living and detrital material as opposed to suspended sediment. In the Early dosing period,
turbidity values were similar among treatments and only ranged from 2-3 NTU's (Figure 21). Turbidity increased during
the Late dosing period due to the main treatment effects of N-dose, Day, and the N-dose*Day interaction. P-dose had
no significant effect on turbidity; however, the P-dose*Day and the N-dose*P-dose*Day interactions were significant
due to the Late dosing regime. Overall, turbidity increased in all treatments (4 to 6-fold). Although this increase was
significantly related to N-dose, the increases in the ON treatment indicated that turbidity increases, in part, were due
to release of nutrients and organic matter as macrophytes senesced as discussed below in relation to phytoplankton
dynamics.
Full
Turb. Early
Late
N-dose
0.0292
0.0033
P-dose
N-dose *
P-dose
Day
0.0001
0.0001
0.0001
N-dose *
Day
0.0001
0.0005
P-dose *
Day
0.0295
N-dose * P-dose
*Day
0.0200
Turbidity by N-dose Level, for Study 1
14 -T—
12
S
H
5/20 6/3 6/17 7/1 7/15 7/29 8/12 8/26 9/9
N °
UOaC
5/20
2
3
2
6/1
2
2
2
6/14
3
3
3
7/1
2
2
2
Da
7/14
2
3
3
te
7/27
2
3
2
8/9
3a
4b
3b
8/24
3a
7b
5b
9/7
6"
6'
8b
9/21
81
gab
12b
Season
Avg.
f
4b
4b
Figure 21. Changes in turbidity by nitrogen dose levels over time. The upper table presents probabilities that dose,
day, and interactions of dose and day influenced turbidity in ANOVA of rank-transformed data. Darkened
values are not significant (p>0.05). The graph is a plot of turbidity over the entire experimental season.
Because N-dose and N-dose*Day were significant influences in ANOVA, values are pooled by N-dose in
the graph. Dark circles on X axis mark dose dates. The lower table lists LS Means (NTU) represented in
the graph, along with statistical information based on the rank transformed data. Within a column, values
sharing a letter are not significantly different (p>0.05). In columns without letters, values are not signifi-
cantly different.
27
-------
Phytoplankton
Chlorophyll a was used as an indicator of phytoplankton biomass. Day (p<0.0001), and to a lesser extent phosphorus
(p=0.0399), had significant effects on chlorophyll during the Early dosing period. Chlorophyll remained <25 ug/L in all
treatments from May through July (Figure 22) during the period of initial macrophyte growth. Neither N-dose or P-dose
had significant effects on chlorophyll during the Late dosing period, in part due to the high inherent variability among
replicates; standard deviations frequently exceeded 100 ug/L (n=4 replicates). In the Control (ON:OP), chlorophyll was
<12 ug/L until the last two days of the season when values increased to above 50 ug/L. In the Control, chlorophyll
was significantly correlated with turbidity (r2= 0.82; p<0.0001), TM (r2= 0.70; p<0.0001), and TP (r2= 0.66; p<0.0001).
Relations between chlorophyll and turbidity, TN, and TP were generally weaker in the dosed treatments than in the
Control. These results indicate that there was tight coupling between macrophyte growth and nutrient uptake which
limited phytoplankton growth during the Early dosing period. As rnacrophytes began to senesce, macrophyte:nutrient
relationships were less tightly coupled as reflected in increased chlorophyll, turbidity, and dissolved nutrients.
Monthly planktonic algal identifications yielded 99 species during Study 1, representing the divisions Chlorophyta (greens),
Cyanophyta (cyanobacteria or blue-greens), Cryptophyta (cryptomonads), Bacillarophyta (diatoms), Euglenophyta
(euglenoids), Xanthophyta (yellow-greens), Pyrrophyta (dinoflagellates), and Chrysophyta (golden-browns) (Table 4).
Total species richness was similar among treatments and averaged approximately eight species during summer, with
slightly fewer species in May and September.
N-dose
P-dose
N-dose *
P-dose
Day
N-dose *
Day
P-dose *
Day
N-dose * P-
dose * Day
Full
0.0001
0,051*
Chi Early
0.0399
0.0001
Late
0.0001
0.0236
'-4-,
":• 0,2754-
350 1
300
Phytoplankton Chlorophyll by Treatment, for Study 1
I
250
200
a
150
U 100
50
1 (ON;OP) --m-~7 (2.5N;OP)
3(ON;loP) - -*- - 9 (2.5N;loP)
2(ON;hiP) - -^ - 8 (2.5N;hiP)
4 (5N;OP)
6 (5N;loP)
5 (5N;hiP)
5/20
6/3
6/17
7/1
7/15 7/29
Date
8/12
8/26
9/9
Figure 22. Changes in phytoplankton chlorophyll by treatment over time. The table presents probabilities that dose,
day, and interactions of dose and day influenced chlorophyll in ANOVA of rank-transformed data. Dark-
ened values are not significant (p>0.05). The graph is a plot of chlorophyll over the entire experimental
season. Dark circles on X axis mark dose dates.
28
-------
T3
^
GO
O3
c
^5
Q
I
"o
—
0
0
O3
CD
'o
CD
Q-
00
C
c
CO
Q-
o
J2-
CL
"5
03
-1
m
^
(D
J5
s
'o
0)
a
CO
Genus
>,
«
iL
v
T3
CD
(0
*»
o
E§
- «
.C
0.5
E
o
TJ
O>
£
i^
CD
CO .^
«" a-
ci o 2
O) sr .£
nabaena
nabaena
nabaena
< < <
0)
CO
CD
O
CO
0
0
03
O
•z.
stocales
0
~Z-
CD
CO
CD
0
f—
Q.
O
CO
O
CO
0
O
f—
Q.
O
C
CO
O
CO
CD
C
o
^
CO
O 0)
C •- 0
ll § 1
-55 co cLd-cLS; cicicicLQ-E
Q.>0303O3i±!o303030303^=
C CO
11 rt „ « |8 § -
^^E-S"300 '^9-cDoo
COCOi-cO'C'C -^CO^-fiO
^C9coox°":£oo
CDCD.NcO-^S'C'COogo
cococ-ojocO-c-gccoo
-Q^DC03."=^;'XfiiCOcOoo
coco-CcDOo^xjr^ict:
CCQ.03W")CO^Q.CL^:-C
<< ^Z
O QC O
03
CD
CO
o
o
o
0
o
2
.c
0
CO
03
CO O
*-• c
03 3 03
.3 ° S
CO 3 03 Q. CL
> Q. CO O3 03
s|»«
0 CD 03 0 OT
0 0 H ^§ 0>
o F ?T c c
§ 1 1 1 §
.C 03 .2 o O
O 2 S < O
fll
03 CO
03 O
0 CO
c: c
CO 0
C 0
"o o
< o
03
CD
CO
c:
CO
o
<
CD
CO
03
0
QL
o
'l
m
'o
CO
m
CO
Q_
o
'l_
_CO
'o
CO
m
03
c
CO
Q-
co
^5
C CO
03 "3
J5 cL cL 'F
Q. 03 03 b
§111
0200
ro 55
03 Q)
11
CO >•
m o
CO U3
CD 03
"cO CO
IE
o E
CO >^
m o
03
omphonema
03
CO
CD
o
05
CD
1
o
(D
COOO
...
Q.CD Q.Q.
O3Q.03O3
CL H3 Q. Q.
O3 O3 O3 C/3
HI -|
03
CO
CD
O
CO
o c
DC LU
03
_03
CO
O
LU
m S g
Is! 1
CO 5 -C
-Z. Q. CC
03 O3 CO
® ® =6
_CO _CO _p
3 3 CO
O O Q.
'> '> O
z £ a:
03
CO
03
O
>>
Q-
g
.§
'o
CO
m
co
Q.
.g
^
'o
CO
m
o
_cO
Q.
CO
1
U3
29
-------
CD
.c
"c
o
O
1 —
T3
3
03
O)
c
^
Q
•D
Q)
"o
_0
"o
o
CO
0
"o
CD
Q.
CO
C
o
c
co
Q_
o
c~
0.
"o
15
^"
fli
_£•
13
(0
0
E
CD
LL.
0)
•a
6
w
co
co
O
||
^>i
0-5
E
o
TJ
D>
C
r^
CO
c
— .5°.
O CO 'p
10 .0 CO Q)
w -a c CD
JD c .55 c
CO rrt
^— ^ CO
CO 0 CO CO ~— ~ CO °~~
~~™ ° °~~ °~~ i • ^~ tO
CD
CO
8 „ 1 »
11 fi
1. 2 ° ™
— — *TT *** TO
O "o *^
m 2E 55 it
CO
CD
1 1 II
"a_ 2 'co -c
-9- -^ co co
3 Ji: CO ^^
"^ "co S1
m 5 f LL
i
0)
CO
CD
0
a.
_o
^
'a>
2
LL
CO £ E <" CO
co '£ « S E co -2 m « |
!!.!«. 1.. 1.1. !!!!.!
co t. Q_CD£ CL^-JQ Q-Q.^ a.Q-'p Q.CL^ ^a) Q-°
CO
2 co w
t ^3
EC CZ
^, CO _CO CO CO CO ff. ,yj
.5? .CO CO "^ 'C "o ^ W w 3 E "C 'C -C "C 'C CD CD
'~T "P^» Q) Q) Q) fl) Ql ._ ("^ ^^* ^^ ^^ ~^ ^^ O O O O O ^~ C
CT C C r^ r^ v ^~* O O fy Q) Q} ^ -^ — < — ^ — ^ rt) nj
CD
CO
0
0 0
CO 0 CO
0 CO P
! 1 1
^ CO 0
"0 >^ C
^0 I
0
C/J CO
0 0
1 §
~fn **~
_Q ^
.CO g
0
CO
0
O
^:
o
^
6
-2
>,
a.
^
O
arcuatus
bijuga
brasiliensis
CO CO CO
Hi 13 ^
& E E
CO CO CO
0 CD 0
Q) Q) Q)
02 Q) Q)
desmaceae
Q)
c:
Q)
O
C/3
CD
03
o
o
o
o
0
i_
o
Q
CD
CO
0
O
:>,
Q.
Q
c>
6
CO
^
Q.
_g
6
0
s
'c
CO
il
quadricauda
sp.
sp.
CO
^ fll f-
C ^_* ;-
CO ^ •
0 J5 c
c o o5
joniaceae
uj
o
0
O
CO
"CO
'c
o
O)
o
"O
O
Q.
CO
to
o
o
m
rrt
f—
Q_
C/5
CD
liopsidaceae
0
C
"co
0.
1
O
Q.
co
^
.0
30
-------
T5
Q)
C
"c
o
O^
-*— •
CO
c
Q
"0
.2
"o
JCD
"o
O
CO
CD
'o
CD
Q.
CO
c
2
c
CO
CL
O
jl>
a.
bi
o
w
1
— I
•
^T
0>
JQ
W
V
'o
Q)
Q.
CO
W
3
C
0)
0
I
(0
u.
^
(D
0
(A
M
BJ
O
l-i
— .2
r** ^
0.0
E
o
•o
0)
_c
2
* s
IS £E js
>- O CO CO o CO
CD _Q -Jg CD s= Q
oLolS ci -2 > cl -5? ~co CL r:
COCO-H;cOC')OCOCDOCO:=
co co en
CO CO CO
•— o o o
CO CZ C C
^> o o o c c
2 .x .x "5, ">,">, .c _c 'g •:: -c
IlilllSSloo
CD
co
CD
CJ
co
CO
CD C ^
CO g CD ffl
CD C CO JU
00 CD "
CO T3 O .°°
-C >, CO -D
•- E o 1
K -S > w
-2 -C "5 CD
Z) O > Q
CO
CD
$ CO «
"CO ^ ^
-c co E
'•5 S c
^ 0 5
D 5 N
CO
•^ 55
. -t CC
Q. O Q,Q.Q_Q. -P'F
c/)-Q(/Dc/)c/)coCC
J2
r^ r; J5 C 2
'~ — ' ^ rrt O
lllllll_
c3c5illdl^
0
CO
Cl) O
CO CO
5 I
1 E
§ 2
C Q.
cn >s
*>^ x-
N 0
CO
CD
CO
~s
O
E
Q.
0
CD
CO
CD
0
^— *
Q.
-g_
^^
^_
O
CO
o
T^
nannoplant
A
o
-~.
^>
5
—\
J
c
cL
CO
1
co
ynuraceae
co
CO
CD
CO
T3
CO
c
0
_c
0
o
CD
CO
CD
0
f~
Q.
O
CO
i—
6
^
o.
o
CO
^
r~
O
sp.
longicauda
g CO
CD
LJJ Q.
uglenaceae
UJ
CD
co
-&>
3
UJ
CD
CO
CD
O
•>,
CL
O
C
_CD
Oi
LU
ro
o
o
>,
Q.
O
c
CD
O)
UJ
CO J3
CL O CO
CO CO C
co co
^ ^
o o E
cz cz ^
o o 'E
Q) CD T3
£ £ I
-
CO
CD
C1
lenodiniopsidai
a
CO
CD
.CO
1
°-
CD
CO
CD
O
^Z
CL
^
Q
crt
*->
^
Q.
0
^*
Q.
i_
^_
>,
CL
ci
CO
CO
'd
CD
:>
aucheriaceae
>
CO
CD
.CO
*L_
CD
o
>
CD
CO
CD
O
SI
Q.
O
£
C
X
CO
J,
CL
O
t:
CO
X
31
-------
Algal enumeration at the division level revealed seasonal patterns in total and relative abundances that were similar
among treatments (Figure 23). In general, total algal densities were highest in mid-August (mean = 7.1*106cells/L).
Densities declined by September to approximately half of August levels. In the Early season, cryptomonads averaged
8.8*105/L and were the numerically dominant taxa in May (92% of populations) and June (38% of populations) dur-
ing the early dosing interval. However, during the Late summer season, cryptomonad populations contributed <5%
to the total phytoplankton community. Blue-green populations peaked in summer and averaged 5*105 cells/L during
the June-August period, but were never the numerically dominant taxa. Green algae were poorly represented in May
samples (<1*104 cells/L) but represented 30% of the algae (5* 105 cells/L) in June. Chlorophytes were the dominant taxa
in July, August, and September, comprising 48, 91, and 90% of the community, respectively. Chlorophyte populations
peaked in August (average >6*106 cells/L). Diatoms represented <5% of the total phytoplankton community throughout
the 4-month study. Euglenophyta, Pyrrophyta, Xanthophyta, and Chrysophyta were infrequently encountered; their oc-
currence was not consistent among corrals within a given treatment and did not show a seasonal relationship.
Cyanobacteria observed during the study included Anabaena, Aphanizomenon, Aphanothece, Calothrix, Gloeotrichia,
Oscillatoria, and Pseudanabaena sp. Statistical analysis indicated that there was not a substantial difference in the
total abundance of these genera among treatments. Two corrals from the Control (ON:OP) and one corral from 5N:OP
experienced blooms of Anabaena and Aphanizomenon in July and August. TN:TP ratios in those corrals prior to the
blooms (>16:1) indicated that nitrogen was not strongly limiting. Likewise, TN in bloom corrals was similar to replicate
corrals prior to and subsequent to the blooms. Although cyanobacteria are capable of "fixing" atmospheric nitrogen
under nitrogen-limited conditions, there were no trends to indicate that nitrogen fixation was a nominal factor in nitrogen
dynamics or primary productivity in the loP or hiP treatments during the study.
Phytoplankton Division Abundance, for Study 1
0 4-
May
Chlorophyta B Cyanophyta 8 Cryptophyta d Bacillariophyta
Figure 23. Changes in phytoplankton abundance of the four dominant divisions over time. Columns represent aver-
ages of all corrals in each month to show general successional trends over the season.
32
-------
Periphyton
Periphyton biomass was measured as the accrual rate of chlorophyll a on Scrimweave™ strips. Accrual rates of pe-
riphyton biomass (expressed as ug Chl/cm2/wk) were measured in July, August, and September which corresponded
to the Early, Late, and post-dosing periods. During each interval, strips were incubated and retrieved for both 1-week
and 2-week intervals to determine penphyton accrual rates related to nutrient dosing. In this analysis, the data were
statistically analyzed for main effects using the entire combined dataset.
Day (p<0.0001), N-dose (p<0.0001), and P-dose (p=0.0336) had significant effects on periphyton accrual rates in
the 1-week periphyton growth interval; the N-dose*Day interaction (p=0.0043) was also significant (Figure 24). Day
(p<0.0001) and N-dose (p<0.0001) had significant effects on the 2-week periphyton response; however, P-dose had
no effect (Figure 24).
Control periphyton accrual rates averaged 0.05 ug Chl/cm2/wk for both 1-week and 2-week growth intervals during the
July, August, and September sampling intervals. Neither N-dose nor P-dose was a significant factor in the July data,
however, which corresponded to the Early dosing interval. In contrast, nutrient dosing significantly increased the 1 -week
periphyton accrual rates during August which occurred during the Late dosing period; however, there were no signifi-
cant nutrient dosing effects on the 2-week data. The weaker associations between periphyton accrual and dosing in
the 2-week exposures, in which accrual rates began to decrease, could have been due to nutrient limitation, biomass
Peri.
l-wk
Peri.
2-wk
Full
Full
N-dose
o.oooi
ooooi
P-dose
00336
N-dose »
P-dose
Day
00001
00001
N-dose'
Day
00043
P-dose"
Day
N-dose * P-dose *
Day
Periphyton Chlorophyll Accrual, for Study 1
August
Month
September
N °
,L 2-5
Month
July
1
0.03
003
004
2
0.03
0.04
004
Aug
1
004'
020b
024"
2
006
0.10
016
Sept
J
0.04'
0.08b
024'
2
0.05
008
0.22
Season
Avg.
l
0.04'
o.io"
0.17"
2
004'
0.07b
0.1 4C
Figure 24.
Changes in periphyton chlorophyll accrual rates over time. The upper table presents probabilities that
dose, day, and interactions of dose and day influenced periphyton accrual in ANOVA of rank-transformed
data. Data from one and two-week exposures were analyzed separately. Darkened values are not sig-
nificant (p>0.05). The graph is a plot of accrual rates in both datasets. Solid and striped bars represent
(1) and (2) week exposures, respectively. Because N-dose was a significant influence in ANOVA of both
datasets, values are pooled by N-dose in the graph. The lower table lists LS Means (mg Chl/crrf/wk) for
both datasets represented in the graph, along with statistical information based on the rank transformed
data. Within a column, values sharing a letter are not significantly different (p>0.05). In columns without
letters, values are not significantly different. Significant differences between monthly values within the 2-
week data set are not shown due to the lack of significant influence on the model.
33
-------
loss, or shifts in the species composition of the periphyton community. N-dose at the 5N dosing level had significant
effects on both the 1-week and 2-week accrual rates in September during the post-dosing interval. Both N-dose and
P-dose had significant effects on 1-week and 2-week periphyton accrual rates when tested across the seasonal aver-
age, resulting in an average of 0.4, 0.1, and 0.17 ug Chl/cm2/wk for the Control, 2.5N, and 5N treatments, respectively
(1-week data); and 0.4, 0.07, and 0.14 ug Chl/cm2/wk for the Control, 2.5N, and 5N treatments, respectively (2-week
data). Collectively, the data indicated that both N-dose and P-dose influenced periphyton accrual during the Late and
post-dosing periods but had minimal effects early in the study. Thus, periphyton productivity most likely played a much
greater role in overall system productivity late in the study compared to the early component of the study when macro-
phytes probably were more dominant factors.
Zooplankton
N-dose had significant effects (p=0.0295) on total numbers of copepods; however, neither N-dose or P-dose had signifi-
cant effects (p>0.05) on total numbers of zooplankton, total numbers of cladocerans, or total numbers of rotifers in Study
1 (Figure 25). However, Day was a significant factor for total numbers of zooplankton (p= 0.0212) and total numbers
of cladocerans (p=0.0013). Highest actual numbers of zooplankton occurred during the early, pre-treatment period
when total numbers reached approximately 6.8 * 105 zooplankton/m2; over 95% of total zooptankton was represented
by cladocerans during this pre-treatment period. Total numbers of zooplankton appeared to decline across treatments
Total # zooplankton
Total # cladocerans
Total # copepods
Total # rotifers
N-dose
is.$m
. ftjjKjfo,
^H
Hi
Control
N-dose
2.5
5
Average number of copepods (# x!05/m2)
5/12/1999
0.3
0.3
0.3
7/16/1999
0.42*
0.57'
0.44'
8/19/1999
0.26*
0.56b
0.52*
Figure 25. Changes in total numbers of copepods over time. The upper table presents probabilities that dose, day,
and interactions of dose and day influenced total numbers of zooplankton by major group in ANOVA of
rank-transformed data. Darkened values are not significant (p>0.05). Because N-dose was a significant
influence in ANOVA for total number of copepods, values are pooled by N-dose in the graph. The lower
table lists LS Means (total number of copepods) along with statistical information based on the rank trans-
formed data. Within a column, values sharing a letter are not significantly different (p>0.05). In columns
without letters, values are not significantly different.
34
-------
in the July and August sampling periods to less than 3.2 * 105 zooplankton/m2; however, these samples were taken
with the passive activity traps as opposed to the pre-treatment period which was net samples corrected to the same sq.
meter basis of comparison. In addition, proportions of copepods and rotifers increased compared to cladocerans. The
observed increase in total numbers of copepods was statistically related to N-dose (p=0.0295) but not P-dose or Day.
Total number zooplankton species was significantly related to Day (p=0.0005) (Figure 26). Total number cladoceran
species was significantly related to P-dose (p=0.0041), Day (p=0.0003), and the P-dose*day interaction (p=0.0281).
There were a total of 31 species of zooplankton species identified in Study 1; the species list is presented in Table 5.
Thirteen species were cladocerans; 5 species were copepods; 13 species were rotifers; and one species was an
ostracod. Prior to corral construction (May 12, 1999), the zooplankton community was dominated by two species of
cladocera: Ceriodaphnia reticulata and Daphnia pulex in a 4:1 ratio. By July, however, the zooplankton community
had shifted to a dominance of seven species: Hexartha mira (Rotifera), Cypridopsis sp. (Ostracoda), Microcylcops
rubellus (Copepoda), Platyias patulus (Rotifera), Chydorus brevilabrus (Cladocera), Alona monocantha (Cladocera),
N-dose
P-dose
N-dose *
P-dose
Day
N-dose *
Day
P-dose *
Day
N-dose * P-
dose * Day
Total # sp.
0.0005
0.0069
# Cladoceran sp.
0.0041
0.0003
00281
# Copepod sp.
# Rotifer sp.
•jam.
0,3292
0.2822
ft.0871.
00038
OJ&3
Zooplankton Species Richness by P-dose Level, for Study 1
1 S
.2 12-
u
J5 '»
* 8 -
«
S 6-
o -
D Total
• Cladoceran
H Copepod
B Rotifer
T
L
1L.
Pre-Trt
5/12/1999
-
•
_i
n
J
31
•
•
_i
y
n
Sa
3i
B
\ffl
31
Control Lo Hi Control Lo Hi
7/16/1999 8/19/1999
Treatment Date and Phosphorus Dose
Control
P-dose
Lo
Hi
Average number of cladoceran species
5/12/1999
2.3
2.3
2.3
7/16/1999
6.9"
8.2b
8.1b
8/19/1999
6.5"
6.4"
7.3b
Figure 26. Changes in zooplankton species richness over time. The upper table presents probabilities that dose,
day, and interactions of dose and day influenced zooplankton species richness by major group in ANOVA
of rank-transformed data. Darkened values are not significant (p>0.05). Because P-dose was a significant
influence in ANOVA for number of cladoceran species, values are pooled by P-dose in the graph. The
lower table lists LS Means (number cladoceran species) along with statistical information based on the
rank transformed data. Within a column, values sharing a letter are not significantly different (p>0.05). In
columns without letters, values are not significantly different.
35
-------
1
^^
^
CO
en
3
Q
T3
"o
"5
O
CO
CD
"o
CD
Q.
CO
c
O
2
c
CO
CL
o
o
N
'to
— '
if>
o>
TIT
(0
H
Species
Genus
E
(C
u.
,_
0)
^j
o
3
w
Q)
"S
O
If)
co
(0
u
.a
3
(0
to
(0
(0
o
E
3
^^
t-
Q.
CO
guttata
monocan
CO CO
c c
O 0
hydoridae
O
CO
CD
0
0
T3
CO
O
CO
0
CO
+-J
CO
o
CL
b
CO
T3
O
CL
J3
•^
f—
CL
CO
T)
O
CL
O
f~
O
c
CO
CD
CO
T3
0
CL
2
.C
COcOm"rr0l^C::39
sss^SS'-g-s-isg
co^DO^i:cO"Cii;-CLCO>^3
CO CO ™
CO -c 3 E
CO 'c CD -m 0
§1 ffl ^fco^^i
gll^ll«lt§I
0 ^ § g ST 0 "fe & g E 1
,
o
LU
iaptomidae
yclopidae
Q O
CO
CO tJ
~ o
0 CL
1 f
O 0
CO
T3
0
CL
CD
CL
O
O
co
o
CL
0
—
"x
CO
2
CO
T5
"co
CO
r*>
Macrocycloj:
rubellus
CO
Microcyclop
d
CO
Cypridopsis
m
ypridopsidat
0
CO
CL
0
0
o
o
CL
to
o
o
CO
1_
(/)
O
longiseta
co
"c:
u;
iliniidae
LL.
Q)
CO
Q)
0
CO
jo
-3
o
CO
0
III
03
C
Q
c:
0
o
c
o
2
2
CD
;—
O
en
'E
Hexarthra
exarthridae
X
co
32 CD -§
•§ d "§ ^
'co co o S
CO
!•§«.«.
C 03 CD 01
cO f — *~* "^^
— — -~— co co
co :3 CD CD
< LU X X.
CD
splanchnida
rachionidae
< CO
CO
E
"o
CL
CO
.« "co
E "H
1 § 8 g « co
s 1 -o i co ^ ^
5 Q. CT l> I 0- 2
CO cO
CO JO CO (U gj ^£
r- CO CO ;=- ;=- CO i_
— -^r '^r •= -^ O CO
= ^ ^ co co c £,
>^ JO _CO ^ CD 5 "0
S CL Q 1 ._J 2 CL
CD
CO
CD S
CO Q
P CO
C -C
s g
CD >*
_J CO
36
-------
and C. reticulata (Cladocera). Slight shifts in dominance of the top 7 species were observed in August: Cypridopsis
sp. (Ostracoda), P. patutus (Rotifera), Dunhevedia crassa (Cladodera), Hexartha mira (Rotifera), Chydorus brevilabrus
(Cladocera), M. rubellus (Copepoda), and C. reticulata (Cladocera). Total numbers of cladocerans were significantly
correlated with chlorophyll a (r=0.258; p=0.0285); however, no single species of cladoceran was correlated with chl
a. Ceriodaphnia reticulata was negatively correlated with macrophyte biomass; however, no other cladoceran species
demonstrated statistical associations with macrophytes.
Sediment
Nitrogen content of sediment in the Control nearly doubled between May (0.21% N) and June (0.37% N), and then
remained near that level for the remainder of the season. ANOVA did not indicate treatment influences, but concentra-
tions in all treatments in May were significantly lower (p<0.05) than the other months (Figure 27). Sediments contained
around 84 g N/(m2*5 cm deep) in May, and between 150 and 160 g N/(m2*5 cm deep) during the remainder of the
season. The apparent sediment pool of nitrogen increased seasonally in all treatments due to a combination of macro-
phyte mobilization/deposition and possibly physical remixing of sediments due to macrophyte and sediment sampling
activity. Thus, it is apparent that even in the highest dosing treatment where a total of 30 g of nitrogen was applied
(5N treatment), it was difficult to measure temporal increases in loading and transfer of dosed nitrogen to sediments
due to the inherent error involved in the procedures used. In fact, early attempts using deposition trays demonstrated
significant, yet highly variable accumulations of sediment and detritus in trays due to various factors including biomass
sloughing, disturbance by physical activity, and disturbance by gaseous evolution by sediment decomposition processes.
Denitrification, not directly measured in this study, may also have been a factor in our inability to detect nitrogen ac-
cumulation in sediments over time.
N-dose
P-dose
N-dose *
P-dose
Day
N-dose *
Day
P-dose *
Day
N-dose *
P-dose * Day
Sed
N
Full
0.0001
Early
0,4488
0.0001
Late
200
180
"o _ 160
* S
s*141
| g 120 +
g £ 100 - -
a "a 80 -
60
40
20
0
.1
•8
CO
Sediment Nitrogen Pool, for Study 1
M
10.34
%N
J
Month
0.36
%N
Figure 27. Changes in sediment nitrogen pool over time. The table presents probabilities that dose, day, and interac-
tions of dose and day influenced N content in sediments in ANOVA of rank-transformed data. Darkened
values are not significant (p>0.05). The graph is a plot of the estimated pool of N in the sediments for an
area of 1 rrf and a depth of 5 cm. Because treatment influences were not significant, monthly values are
presented as averages of all treatments. The average %N in sediments is shown within the column for
each month.
37
-------
Phosphorus content of the Control sediments fluctuated between 0.05 and 0.06% P during the experiment (Figure 28).
ANOVA indicated that treatment responses were not influenced by N- or P-dosing, and that sediment phosphorus was
significantly lower (p<0.05) in September than earlier months although differences were slight. Sediments contained
between 23 and 25 g P/(m2*5 cm deep) during the season. Dosing in the highest treatments (hiP) delivered 0.86 g/m2;
therefore, the background levels of phosphorus in sediments inhibited our ability to measure additional phosphorus ac-
cumulation. However, due to the fact that phosphorus is conserved (i.e., not cycled to the atmosphere), it is assumed
that phosphorus not accounted for in macrophytes or the water column was transferred to sediment.
N-dose
P-dose
N-dose *
P-dose
Day
N-dose *
Day
P-dose *
Day
N-dose *
P-dose * Day
Sed
Ful1
OJ018
0.0001
Early
0,1828
0.0011
Late
$3755,
0.0001
Sediment Phosphorus Pool, for Study 1
Sediment Phosphorus Poo
(g P/m2 x 5cm depth)
i-- — NJ KJ
3 <-* O Lrt O <-"
f
i 0.05 j
< %P
M
T T T
8?
lo.oe'
•- %P ••
• i
iooe;
%P f
1
\ 0.06 ^
i %P j
I
0.05;
i %p ;
Jim Jul A S
Month
Figure 28. Changes in sediment phosphorus pool over time. The table presents probabilities that dose, day, and in-
teractions of dose and day influenced P content in sediments in ANOVA of rank-transformed data. Dark-
ened values are not significant (p>0.05). The graph is a plot of the estimated pool of P in the sediments
for an area of 1 nf and a depth of 5 cm. Because treatment influences were not significant, monthly
values are presented as averages of all treatments. The average %P in sediments is shown within the
column for each month.
38
-------
System Metabolism
Dissolved oxygen was measured on a consecutive morning, evening, and morning sequence and was used to evaluate
community metabolism of the corrals through estimations of gross community primary production and respiration. Day
(p<0.0001), P-dose (p=0.0430), and the P-dose*Day (0.0249) interaction were significant factors related to community
gross primary production during the 4-month study (i.e., Full dataset) (Figure 29). Prior to dosing, all treatments had
similar levels of gross primary productivity and averaged 9 mg/L. As dosing began, gross primary productivity increased
in all treatments concurrently with macrophyte development. N-dose was a significant, positive main effect which in-
creased productivity compared to Controls during the Early dosing period; P-dose had no significant effect during the
Early dosing period. P-dose had a significant, negative effect on gross primary production. Production estimates in
loP and hiP were 20% lower than OP when averaged over the course of the study; however, the loP and hiP treatments
did not differ between each other.
N-dose P-dose
Gross
Productic
Full a*fesli*tff 0 0430
Early 0 0470
11 Late !gp$'$ggf|f::; 00126
N-dose*P-dose*Day Day
0.7831 - 0.0001
<•- ' . 0.8292, s k 0 0001
OjWW-J " OOOOI
N-dose*Day
vf""f8j]$r-!'>
;*^iiMWi?, -
IliftSm'-.x'
Community Production by P-dose
n (mg O2/L)
; g t
S
K
10
0 -
5
^ ^''
P-dose*Day
0.0249
'.'i* ^iiftSSEl^SSJiS.,
::S*r'*tt>». r-^.u'
N-dose*P-dose*Day
i**»8'r ^-»'.WI^^»»:f%-%'
^' iv! . . • u '~j4;I^^^^^P's- '*i i£j
s-g^&f^'tif^&^^&A;: IS'
level, for Study 1
A--'
.A.
^f^
^\ /f^t^^+S'***^
^^$^*^^f
fc-*"^^^*
• . A
V W ' ' ' V ' V '
20 5/27 6/3 6/10 6/17 6/24 7/1 7/8 7/15 7/22 7/29 8/5 8/12 8/19
Date
5/20 6/3 6/8 6/17 6/
0 9.1 11.8 13.1 12.8 14
. lo 8.2 11.3 11.3 13.2 15
hi 8.9 11.1 11.6 15.7 18
Dat
24 7/1 7/7 7/14 7/27
.la 11.3 21.1 17.2 20.8"
.1" 10.6 18.3 16.0 15.9b
.3b 12.0 21.0 17.7 16.0"
e
8/3
23.4a
19.4b
20.1ab
8/9
25.0a
19.7b
18.8b
8/17
25.9a
21.6ab
20.7b
8/26
•
•«.
,->--
^. ^--^Jfr"^.
- - A- - • OP
— -» — loP
8/24
29.5a
23.0b
24. lb
9/2 9/9
8/31
25.5
22.8
21.8
A
k '
^
9/7
28.8
24.3
23.9
9/16
9/14
26.3a
20.2b
22.7*
Season
9/21 Avg.
26.9a 20.2a
23.2ab 17.3b
21.0b I8.0ab
Figure 29. Changes in community gross oxygen production by phosphorus dose levels over time. The upper table
presents probabilities that dose, day, and interactions of dose and day influenced production in ANOVA
of rank-transformed data. Darkened values are not significant (p>0,05). The graph is a plot of produc-
tion over the entire experimental season. Because P-dose and P-dose*Day were significant influences
in ANOVA, values are pooled by P-dose in the graph. Dark circles on X axis mark dose dates. The lower
table lists LS Means (mg O/L) represented in the graph, along with statistical information based on the
rank transformed data. Within a column, values sharing a letter are not significantly different (p>0,05). In
columns without letters, values are not significantly different.
39
-------
Trends in community gross respiration were similar to gross primary production in direction and magnitude (Figure 30).
Day was a significant factor influencing community respiration; however, N-dose had no effect, and P-dose was signifi-
cant for only the Late dosing interval and had negative effects compared to the Control OP treatment. The negative
effect of P on both gross community primary productivity and community respiration reflects, and perhaps exacerbates
to some degree, the observed macrophyte senescence. Macrophyte stands declined during the Late dosing period in
all treatments including the Controls. HiP treatments accumulated more dissolved phosphorus than other treatments
(Figure 8), which indicates to some degree that the nutrient assimilatory capacity of the corrals declined along with
macrophyte biomass. Phytoplankton productivity increased late in the study in response to increased nutrient availability
(Figure 22); however, it was not sufficient to maintain levels of productivity observed in Control treatments. Macrophytes
dominate the productivity of these systems, and therefore as macrophyte productivity declined so did the overall esti-
mates of community metabolism.
N-dose
P-dose
N-dose*P-dose*Day
Day
N-dose*Day
P-dose*Day
N-dose*P-dose*Day
Gross
Respiration
Full
Early
Late
0.0337
00001
0.0001
0.0001
Community Respiration by P-dose level, for Study 1
5/20 5/27 6/3 6/10 6/17 6/24 7/1 7/8 7/15 7/22 7/29 8/5 8/12 8/19 8/26 9/2 9/9 9/16
Date
0
l*-dose i0
hi
5/20
7.1
6.5
6.9
6/3
9.6
9.3
8.6
6/8
13.6
12.1
12.6
6/17
10.5
10.3
13.2
6/24
12.2
13.0
16.0
7/1
11.7
10.8
12.0
7/7
20.6
18.2
20.8
7/14
17.1
15.4
16.7
Date
7/27
21.7
17.1
16.9
8/3
23.3
20.2
21.2
8/9
24.0
18.6
17.4
8/17
27.2
23.5
22.1
8/24
29.5
23.0
24.4
8/31
25.9
22.2
21.7
9/7
30.7
26.5
26.0
9/14
26.4
20.4
22.8
9/21
25.2
21.5
19.2
Season
Avg.
19.8
17.0
17.6
Figure 30. Changes in community gross respiration of oxygen by phosphorus dose levels over time. The upper table
presents probabilities that dose, day, and interactions of dose and day influenced respiration in ANOVA
of rank-transformed data. Darkened values are not significant (p>0.05). The graph is a plot of respiration
over the entire experimental season. Because P-dose and P-dose*Day were weakly significant (p<0.1)
in ANOVA, values are pooled by P-dose in the graph. Dark circles on X axis mark dose dates. The lower
table lists LS Means (mg O/L) represented in the graph. Statistical comparisons are not presented.
40
-------
Net Nutrient Balance
During this study, a maximum of 30 g N/m2and 0.86 g P/m2was applied. A final mass balance of nutrients was ca-
culated among various nutrient pools to determine the net efficiency of uptake and assimilation of nutrients in these
experimental systems (Table 6). At the end of the study, macrophytes contained a total of 13.0, 7.7, and 10.9 g N/m2
in the Control, 2.5N, and 5N treatments, respectively; water contained an additional 2.14, 3.02, and 4.57 g N/m2 in the
Control, 2.5N, and 5N treatments, respectively. Combined (macrophytes + water), these two nutrient pools contained
15.14, 10.72, and 15.47 g N/m2 stocks at the end of the study. Thus, a total of 71% (2.5 N treatment) and 52% (5N
treatment) of total nitrogen added during the study were found in these two major biological pools at the end of the
study. This implies, under simple mass balance conditions (i.e., no loss to the atmosphere), that up to 29% (2.5N
treatment) and 48% (5N treatment) of total nitrogen added were absorbed or lost to the sediments as detritus. Actual
attempts to measure the amount of nitrogen in sediments were unsuccessful due to a combination of the large mass
of pre-existing nitrogen in sediments and the error contributed by our sampling procedures. However, we know that
these numbers are conservative, since the Control treatment alone exhibited a total sequestration of 15.14 g N/m2 in
the absence of external nitrogen addition. Thus, it is evident that macrophytes, algae, and sediments combined as an
efficient biological, chemical, and physical sink for nitrogen under the study as designed.
Phosphorus, likewise, was efficiently assimilated and retained in the study. At the end of the study, macrophytes contained
a total stock of 2.21, 0.96, and 1.24 g P/m2 in the Control, loP, and hiP treatments, respectively; water contained an ad-
ditional stock of 0.24, 0.33, and 0.24 P/m2 in the Control, loP, and hiP treatments, respectively. Thus, macrophytes alone
contained more phosphorus, including the Control treatment, than the total externally added to even the hiP treatment.
Thus, the 2.5N and 5N treatments at the end of the study had negative sediment transfer coefficients, which mean that
even under these conservative assumptions, the macrophytes and water contained more phosphorus than could be
explained by external addition and that phosphorus assimilation from water (i.e., added dose) was extremely efficient.
Table 6. Summary Table of the Final Store of Phosphorus and Nitrogen in Water and Macrophytes in Study 1
Nitrogen
Control
2.5
5.0
Load
(g N/m2)
0
15
30
Macrophytes
(9 N/m2)
13.0
7.7(51%)
10.9(36%)
Water
(g N/m2)
2.14
3.02 (20%)
4.57(15%)
Macrophytes +
Water Total
(g N/m2)
15.14
10.72(71%)
15.47(52%)
Presumed Sediment
Transfer
(g P/m2)
NA
4.28 (29%)
14.53(48%)
Phosphorus
Control
Lo
Hi
Load
(g P/m2)
0
0.43
0.86
Macrophytes
(g P/m2)
2.21
0.96 (223%)
1.24(144%)
Water
(g P/m2)
0.24
0.33 (77%)
0.24 (28%)
Macrophytes +
Water Total
(g P/m2)
2.45
1 .29 (300%)
1.48(172%)
Presumed Sediment-
Transfer
(g P/m2)
NA
-0.85 (-198%)
-0.62 (-72%)
41
-------
STUDY 2: Effects of Dosing Prior to Macrophyte Development
Study 2 was conducted in year 2000 to evaluate the effect of nutrient loading on nutrient assimilation, cycling, and
community responses to enrichment in shallow, vegetated aquatic systems. There were four objectives in Study 2: 1)
determine how nutrient loads influenced the concentration and relative distribution of nutrients in the water column,
macrophytes, and sediments; 2) evaluate how nutrient loads influenced species composition, biomass, and/or abundance
of macrophytes, phytoplankton, periphyton, and zooplankton; 3) determine if nutrient enrichment prior to macrophyte
growth induced a phytoplankton-dominated state that persisted throughout the season; and 4) characterize the assimi-
lation and retention capabilities of shallow ponds receiving set weekly nutrient additions, starting prior to macrophyte
development.
The hypothesis for Study 2 was that weekly additions of N and P starting prior to macrophyte development would have a
negative impact on the macrophytes due to shading by stimulated periphyton and phytoplankton communities in both Lo
and Hi treatments. This hypothesis was based on findings in Study 1 that indicated that 25% macrophyte coverage had
provided a stable state that could not be shifted by nutrient addition. We predicted that phytoplankton would establish
dominance early in the season; zooplankton grazing would not maintain algal biomass at a low level, because nutrient
stimulation would allow for an algal growth rate that was higher than the grazing rate. Ultimately, phytoplankton would
persist and reduce macrophyte development by shading (Scheffer 1990,1998). Alternatively, in the absence of fish
predators, large-bodied zooplankton communities would graze expanding algal populations and maintain water clarity
and macrophyte dominance (Brooks and Dodson 1965).
Experimental Design
In Study 2, the frequency of dosing was increased to six weekly additions rather than six bi-weekly additions studied in
Study 1 of 1999. We standardized N:P ratios at 25:1 based on the evidence that N:P ratio was not a significant factor
in nutrient dynamics or production in Study 1. Finally, nutrient enrichment in Study 2 was initiated prior to observable
macrophyte growth. Mesocosms were drained April 11, 2000, for corral construction. A total of 12 corrals were con-
structed in each of 4 mesocosms (blocks) (Figure 31). After a 26-d draw-down period for corral construction, the ponds
were refilled over a 2-day period (May 7-8) and allowed to mix for 2 days prior to raising of sides of corrals on May 9.
Water sampling was begun on May 10 and terminated September 12. Dosing began on May 11, 2000, and continued
weekly for 6 weeks. A different set of ponds were used in Study 2 to prevent bias due to the previous year's study.
There were four replicate corrals for each of the three experimental treatments (n=12 total corrals). There were three
treatments in Study 2:1) a Control, in which no nutrients were added; 2) "Lo", in which the load was 30 g NO3-N/m2 and
1.2 g P/m2 (dosed as 5 mg/L N and 200 ug/L P each of 6 weeks; and 3) "Hi", in which the load was 60 g NO3-N/m2 and
2.4 g P/m2 (dosed as 10 mg/L N and 400 ug/L P each of 6 weeks). The loads in these dosed treatments were two-fold
greater than the maximum in 1999 (total load 30 g NO3-N/m3and 0.864 g P/m3).
Results
Macrophytes
Macrophyte Taxa
The aquatic macrophyte community consisted of >99% Najas during the experiment. The macroalgae Chara sp., was
sparsely present in May but was not noted in subsequent months due to the dominance of Najas guadalupensis. The
relative absence of Chara sp. in the ponds in Study 2 was in contrast to the results observed in Study 1. The lack of
Chara sp. may have resulted due to the fact that ponds were drawn down longer (26-d draw-down) in year 2000 com-
pared to 1999 (14 days) which may have altered normal seasonal succession of the macrophyte community.
43
-------
O
o
O O
o
Pondl
Pond 2
Pond 3
Pond 4
Figure 31. Corral and pond diagram for Study 2 experiments indicating orientation of corrals. Pond 1 shows an ex-
ample of the random assignment of the treatments.
44
-------
Macrophyte Biomass
There was no measurable growth of macrophytes above the sediment surface at the initiation of the study; macrophyte
surface coverage was <1 %. Macrophytes grew rapidly in the Control from May to early August and reached a maximum
biomass of 213 g/m2 (Figure 32). Thereafter, macrophytes lost biomass (32%) between August and September. Both
Day and Dose had significant effects on macrophyte biomass. Nutrient enrichment negatively influenced macrophyte
growth, resulting in stands in the dosed treatments that were significantly smaller than the Control when averaged
across the season (p<0.0013). However, ANOVA indicated that enrichment did not significantly influence biomass in
any given month. Biomass in Lo peaked in August (167 g/m2) at 75% of Control stands, and subsequently decreased
to 105 g/m2 in September. In Lo, timing of growth and senescence periods was similar to the Control. In Hi, however,
the growth period ended in July with a maximum biomass of 59 g/m2. During senescence, biomass in Hi declined by
53%, ending the season at only 31 g/m2. These results were contrary to those in Study 1 in which neither macrophyte
dosing or nutrient ratio had significant effects on macrophyte biomass. Macrophyte biomass in Study 2 was only 25%
of that observed in Study 1. However, different ponds were used in each study.
Macrophyte
Biomass
Dose
0.0013
Day
0.0001
Dose*Day
;;.^';;;"ftfla^t -.^
T'?*' •;-•;.:„ • . .-::S*k-/*T
Macrophyte Biomass, for Study 2
250 T
June
July
Month
August
September
Control
Dose Lo
Hi
June
21
14
4
Mo
July
137
36
59
nth
August
213
167
38
September
144
105
31
Season
Avg.
130"
80b
33°
Figure 32. Changes in macrophyte biomass over time. The upper table presents probabilities that dose, day, and
interactions of dose and day influenced macrophyte biomass in ANOVA of rank-transformed data. Dark-
ened values are not significant (p>0.05). The graph is a plot of macrophyte biomass over the experimen-
tal season pooled by treatment. The final dose occurred one week subsequent to June samples. The
lower table lists LS Means (g dry weight/m2) represented in the graph, along with statistical information
based on the rank transformed data. Within a column, values sharing a letter are not significantly different
(p>0.05). In columns without letters, values are not significantly different.
45
-------
Macrophyte Nutrients
Day (p= 0.0001), Dose (p= 0.0061), and the Dose*Day interaction (P= 0.0129) had significant effects on nitrogen content
of macrophytes in Study 2 (Figure 33). Nitrogen content of macrophytes in the Control significantly increased (p<0.05)
from 1.78% N in June to 3.06% N in September. Nitrogen uptake in the amended treatments was enhanced during
the dose period. Nitrogen content in Lo and Hi peaked at >4% N in early July at levels 2-fold higher than Controls.
Thereafter, N levels in macrophytes decreased in both the Lo and Hi treatments. Macrophytes in the Lo treatment
contained significantly higher nitrogen (p< 0.05) compared to Controls in June, July, and September. Macrophytes in
the Hi treatment were significantly higher (p< 0.05) than Controls in July and August; percentage nitrogen in macro-
phytes in Hi was significantly higher (p< 0.05) than those in Lo in August, only. Nitrogen concentrations were similar in
macrophytes in Study 2 (Figure 33) compared to Study 1 (Figure 3)
Macrophyte
%N
Dose
0.0061
Day
0.0001
Dose*Day
0.0129
Is
•S w>
o '3
5 t
s s?
«> -8
Kfl ^
£•8
Zt
A $
A
"1 ^
1 -
9 ^
t c
i
Oc
Macrophyte % Nitrogen, for Study 2
x-'IS-^
^' ^-^^
^r ^ ^»- — ^
w/ -^-' ...
w .-•
A A"'
A
June July August
Month
-t
^A
- - A- - • Control
-•*- Lo
— • — Hi
September
Control
Dose LO
Hi
June
1.78a
2.91b
2.40ab
Mo
July
2.13a
4.13b
4.35"
nth
August
2.11a
2.63a
3.32b
September
3.06a
3.67b
3.57ab
Season
Avg.
2.27a
3.34b
3.41b
Figure 33. Changes in macrophyte nitrogen content over time. The upper table presents probabilities that dose, day,
and interactions of dose and day influenced N content in ANOVA of rank-transformed data. The graph
is a plot of N content over the experimental season pooled by treatment. The final dose occurred one
week subsequent to June samples. The lower table lists LS Means (% N of dry weight) represented in
the graph, along with statistical information based on the rank transformed data. Within a column, values
sharing a letter are not significantly different (p>0.05). In columns without letters, values are not signifi-
cantly different.
46
-------
Dose (p=0.0061) and Day (p= 0.0001) had significant effects on phosphorus content of macrophytes (Figure 34). Phos-
phorus content of macrophytes in the Control averaged 0.26% P in June, and increased to a maximum of 0.54% P in
September. Seasonal averages of phosphorus in macrophytes were significantly higher in the Lo and Hi treatments,
with peak concentrations in early July at 0.68% P and 0.95% P, respectively. Phosphorus concentrations of macrophytes
in Study 2 (Figure 34) exceeded those in Study 1 (Figure 4) in all treatments including the Control.
Macrophyte
%N
Dose
0.0061
Day
0.0001
Dose*Day
0.0129
1 %
£ M
o -3
U f
8 ¥
b£ ^
fc*
2£
-1
A ^
1 *\
a _
? ^
H
1 C
1
0 S
n
Macrophyte % Nitrogen, for Study 2
xJS-^
^ ^-^^
^f "-. "• — ^
*/ ^- ..-
A A- '
A
June July August
Month
., *j
*r
-A
- - Ar - • Control
--»- Lo
— • — Hi
September
i
,
|
Control
Dose LO
Hi
June
1.78a
2.91b
2.40ab
Mo
July
2.13a
4.13b
4.35b
nth
August
2.11a
2.63a
3.32b
September
3.06"
3.67b
3.57ab
Season
Avg.
2.27a
3.34b
3.41"
Figure 34. Changes in macrophyte phosphorus content over time. The upper table presents probabilities that dose,
day, and interactions of dose and day influenced P content in ANOVA of rank-transformed data. Darkened
values are not significant (p>0.05). The graph is a plot of P content over the experimental season pooled
by treatment. The final dose occurred one week subsequent to June samples. The lower table lists LS
Means (% P of dry weight) represented in the graph, along with statistical information based on the rank
transformed data. Within a column, values sharing a letter are not significantly different (p>0.05). In col-
umns without letters, values are not significantly different.
47
-------
Macrophyte Nutrient Stock
Both Day (p= 0.0001) and Dose (p= 0.0157) had significant effects on stocks of nitrogen in macrophytes. The N stock
in macrophytes in the Control treatment increased 10-fold, from 0.4 g N/m2 to a maximum of 4.3 g N/m2 over the course
of the study (Figure 35). The N stock in Controls did not decrease during macrophyte senescence because the actual
percentage of nitrogen continued to increase late in the study (Figure 35). Nitrogen stocks of macrophytes were similar
in the Lo and Control treatments. The N stocks of macrophytes in the Hi treatment were similar to the Control in June
and July. However, due to premature senescence, N stocks in Hi in August (1.4 g N/m2) and September (1.1 g N/m2)
were 30% lower than in the Control and Lo treatments.
| N Stock
Dose
0.0157
Day
0.0001
Dose*Day
"_
.S
j?
Sw1
X
u
•+rf
t»
a
o>
en
o
•~
Z.
C
A S
A -
1 S
-3
9 ^
9
1 <
i
Oc
Macrophyte Nitrogen Stock, for Study 2
A ^
, *+• — • — __ »
* S — — -,_ .
.'•''' ~*
- f
y«s\,/ --»-Lo
7 S' ^^«-_
•y ^ " ^ " — -m
ff/ "^
&
June July August September
Month
Control
Dose LO
Hi
June
0.4
0.4
0.1
Mo
July
2.7
1.5
2.8
nth
August
4.3
4.3
1.4
September
4.3
3.9
1.1
Season
Avg.
2.9a
2.5a
1.4b
Figure 35. Changes in macrophyte nitrogen stock over time. The upper table presents probabilities that dose, day,
and interactions of dose and day influenced N stock in ANOVA of rank-transformed data. Darkened
values are not significant (p>0.05). The graph is a plot of N stock over the experimental season pooled
by treatment. The final dose occurred one week subsequent to June samples. The lower table lists LS
Means (g N/m2) represented in the graph, along with statistical information based on the rank transformed
data. Within a column, values sharing a letter are not significantly different (p>0.05). In columns without
letters, values are not significantly different.
48
-------
Both Day (p= 0.0001) and Dose (p= 0.0274) had significant effects on P stocks of macrophytes (Figure 36). The P stock
in Control macrophytes increased 12-fold, from 0.08 g P/m2to a maximum of 0.98 g P/m2in August. During senescence,
P stock in the Control decreased to 0.80 g P/m2, but was not significantly lower than the August maximum. The P stock
in Lo was 25% lower than the Control from July through September, but the treatments were not significantly different
overall (p>0.05). The P stock in Hi peaked in July at levels comparable to the Control (0.75 g P/m2 g P/m2) but declined
thereafter to 0.28 and 0.21 g N/m2 in August and September, respectively, due to macrophyte senescence. Overall,
the P stock in Hi was significantly smaller (p<0.05) than the Control and Lo treatments.
| P Stock
Dose
0.0274
Day
0.0001
Dose*Day
'•***.<'
Macrophyte Phosphorus Stock, for Study 2
June
July August September
Month
Control
Dose LO
Hi
June
0.08
0.06
0.03
Mo
July
0.65
0.50
0.75
nth
August
0.98
0.78
0.28
September
0.80
0.65
0.21
Season
Avg.
0.58a
0.45a
0.30b
Figure 36. Changes in macrophyte phosphorus stock over time. The upper table presents probabilities that dose,
day, and interactions of dose and day influenced P stock in ANOVA of rank-transformed data. Darkened
values are not significant (p>0.05). The graph is a plot of P stock over the experimental season pooled
by treatment. The final dose occurred one week subsequent to June samples. The lower table lists LS
Means (g P/m2) represented in the graph, along with statistical information based on the rank transformed
data. Within a column, values sharing a letter are not significantly different (p>0.05). In columns without
letters, values are not significantly different.
49
-------
Water Chemistry
Phosphorus
Day (p=0.0001), Dose (p=0.0048), and the Dose*Day interaction were significant factors controlling total phosphorus
concentrations (Figure 37). Total phosphorus concentrations increased in the Controls from 14 ug P/L in early May to
a maximum of 87 ug P/L in early September (Figure 37). Total phosphorus concentrations averaged 110 ug/L in the Hi
treatment and was significantly greater (p<0.05) than the Control on all but two dates. Concentrations of TP in the
Lo treatment did not differ from the Control treatment on a pooled, study basis but was frequently greater than Control
levels on individual dates.
| TP
Dose
0.0048
Day
0.0001
Dose*Day
0.0001
TP Experiment 1, for Study 2
0-
0 *
5/10
5/24
6/7
6/21
7/5 7/19
Date
8/2
8/16
8/30
Control
Dose LO
HI
5/10
14
13
14
5/23
19'
30'
64"
5/30
24'
31'
66"
6/6
31'
36"
69"
6/13
22'
47"
148C
6/20
25'
56b
193'
6/27
18'
40"
106°
7/4
18'
54"
83°
Da
7/11
21'
54"
114'
y
7/]8
30'
5ib
84"
7/25
70
112
84
8/1
72'
117*
128"
8/8
59"
104"b
122°
8/15
66'
79'
123"
8/22
77"
123'
162"
8/29
69'
70'
123"
9/5
87'
148*
177"
9/12
67'
80*
117"
Season
Avg.
44'
69'
110"
Figure 37. Changes in total phosphorus concentrations over time. The upper table presents probabilities that dose,
day, and interactions of dose and day influenced TP in ANOVA of rank-transformed data. The graph is
a plot of TP over the experimental season pooled by treatment. Dark circles indicate dose dates. The
lower table lists LS Means (mg/L) represented in the graph, along with statistical information based on the
rank transformed data. Within a column, values sharing a letter are not significantly different (p>0.05). In
columns without letters, values are not significantly different.
50
-------
Soluble reactive phosphorus in the Controls averaged 7 ug P/L (range 2 ug to 16 ug P/L) during the season (Figure 38).
Dose (p=0.0001), Day (p=0.0001), and the Dose*Day interaction (p= 0.0001) were all significant factors in SRP dynam-
ics. Due to rapid loss of SRP observed in Study 1, we sampled SRP in Studies 2 and 3 within approximately 1 hr of
application. Peaks in SRP in Lo and Hi indicated an average of 60% and 70% dose recovery, respectively. On May
25, the recovery of only 11% of the third amendment in Hi indicated some unexplained problem in dosing. Dissipation
in the dosed treatments was rapid. The Lo treatment assimilated 100% of each dose within a week (approximately
29 ug P/L/day; 14% P loss/day); whereas, the SRP in the Hi treatment had not yet dissipated to Control levels prior to
the subsequent dose. Calculated dissipation rates in the Hi treatment averaged 40 ug P/L/day (25% applied P loss/day;
excluding anomalous dose three) during the May to early June dosing period.
| SRP
Dose
0.0001
Day
0.0001
Dose*Day
0.0001
SRP, for Study 2
• --A- - • Control
--«--Lo
1
5/10
5/24
6/7
6/21
7/5 7/19
Date
8/2
8/16
8/30
Day
Control
Dose Lo
Hi
Control
Dose Lo
Hi
5/10
5.55
5.58
5.75
7/4
2.05
3.00
25.43
5/11
5.93
12840
320.30
7/5
3.80
323
2050
5/16
12.58
11.33
33.80
7/11
3.43
3.58
20.68
5/18
13.25
89.30
229.08
7/12
6.78
5.28
23.35
5/23
455
3.73
8.93
7/18
9.40
565
26.58
5/25
415
9300
46.25
7/19
16.33
6.23
30 13
5/30
4.35
9.73
28.10
7/25
8.85
978
22.95
5/31
535
16015
342.08
7/26
7.23
9.60
25.05
6/1
6.45
102.48
255.13
8/1
995
16.90
50.13
6/6
6.08
6.90
41.18
8/2
8.00
13.90
47.03
6/7
5.68
159.88
39853
8/8
11.98
16.63
37.13
6/13
480
10.78
9618
8/9
953
12.48
29.95
6/14
7.15
127.10
375.65
8/15
7.75
1035
21.13
6/20
3.60
9.00
161.28
8/22
8.83
1465
3380
6/21
3.55
5.78
144.10
8/29
818
1025
24.53
6/27
4.08
5.60
60.85
9/5
14.35
14.48
42.70
6/28
5.70
5.70
57.60
9/12
9.95
6.00
22.23
Season
Avg.
7.33'
32.25b
91.41'
Figure 38. Changes in soluble reactive phosphorus concentrations over time. The table presents probabilities that
dose, day, and interactions of dose and day influenced SRP in ANOVA of rank-transformed data. The
graph is a plot of SRP over the experimental season pooled by treatment. Dark circles indicate dose
dates. The lower table lists LS Means (ug/L) represented in the graph, along with statistical information
based on the rank transformed data. Within a column, values sharing a letter are not significantly different
(p>0.05). In columns without letters, values are not significantly different.
51
-------
Nitrogen
Dose (p<0.001), Day (p<0.0001), and the Dose*Day interaction (p<0.0001) were all significant factors related to total
nitrogen dynamics (Figure 39). Total nitrogen in the Control ranged from lows of 0.43 mg N/L in mid-May to highs of
approximately 1 mg N/L in August. From the start of the season to July 18, TN in the Control fluctuated between 0.4 and
0.8 mg N/L. Between July 18 and 25, TN increased by 33%, and subsequent concentrations (range: 0.85-1.03 mg N/L)
were significantly larger (p<0.05) than values before July 18. The Lo and Hi treatments exhibited TN levels that were
significantly larger (p<0.05) than the Control on most dates (p<0.05). During the dose period, most of the measured
TN consisted of nitrate (90-100% nitrate) in the Lo and Hi treatments; whereas, nitrate was at background levels in
the Control (Figure 40). Following dose six, TN in the Lo (9.83) and Hi (33.08 mg N/L) treatments reached seasonal
maximums. TN declined throughout the remainder of the season and averaged approximately 1 mg N/L at the end of
the study. Final TN levels in the Lo (0.83 mg N/L) and Hi (1.10 mg N/L) treatments were significantly greater (p<0.05)
than the Control (0.63 mg N/L).
| TN
Dose
0.0001
Day
0.0001
Dose*Day
0.0001
TN, for Study 2
I
5/10
5/24
6/7
6/21
7/5
7/19
8/2
8/16
8/30
Date
Control
Dose LO
HI
5/10
0.78
0.33
0 38
5/23
043"
5.80b
14.88°
5/30
0.65'
6.98b
18.43'
6/6
0.63'
7.68b
24.10°
6/13
0.53'
8.68b
2833°
6/20
0.58'
983b
33.08'
6/27
050"
6.60b
26.08'
7/4
0.53"
463b
22.30'
D
7/11
060"
288b
15.38'
ay
7/13
065"
1.88b
13.78°
7/25
0.88"
1.73b
9.53°
8/1
0.93"
1.55b
6.30°
8/8
0.85'
123b
3.08°
8/15
0.95"
1 08"
1.73"
8/22
1.03"
1 15"
1.65"
8/29
0.95'
1.03"
1.45"
9/5
098"
143b
1 63"
9/12
0.63'
0.83b
1.10C
Season
Avg.
0.70'
3.63b
12.40'
Figure 39. Changes in total nitrogen concentrations over time. The upper table presents probabilities that dose, day,
and interactions of dose and day influenced TN in ANOVA of rank-transformed data. The graph is a plot
of TN over the experimental season pooled by treatment. Dark circles indicate dose dates. The lower
table lists LS Means (mg/L) represented in the graph, along with statistical information based on the rank
transformed data. Within a column, values sharing a letter are not significantly different (p>0.05). In col-
umns without letters, values are not significantly different.
Dose (p<0.0001), Day (p<0.0001), and the Dose*Day interaction (p<0.0001) were all significant factors in nitrate dy-
namics. Nitrate in Control fluctuated near the limit of detection (0.005 mg NO3-N/L) during most of the season. Dur-
ing the dose period, nitrate peaks in Lo and Hi indicated that dosing achieved 87% and 96% of target concentrations,
respectively (Figure 40). Rates of nitrate dissipation in the week following each amendment did not differ between Lo
and Hi, but did significantly increase (p<0.05) from an average of 0.43 mg NO3-N/L/day (4% N loss/day) after the first
two doses, to 0.75 mg NO3-N/L/day (7.5% N loss/day) after the final four doses as macrophyte biomass increased
(Figure 41). Nitrate amendments were not completely dissipated within a week, and therefore, dissipation rates were
accurate estimates of uptake in these systems in May and early June. Following the final amendment, nitrate in the Lo
and Hi treatments peaked at 12.63 and 38.57 mg NO3-N/L, respectively, and subsequently decreased throughout the
remainder of the season. Due to slower dissipation rates with each successive week following the final dose, nitrate in
Lo did not fall below the limit of detection until August 15 (Figure 40). Nitrate in the Hi treatment steadily decreased at a
rate of 0.59 mg NO3-N/l_/day (6% N loss/day) from June 21 (29.67 mg NO3-N/L) to August 8 (1.71 mg NO3-N/L) (Figure
40). After August 9, nitrate dissipation in the Hi treatment continued at <0.2 mg NO3-N/L/day until concentrations fell
below the limit of detection on August 29.
52
-------
[ NO3-N
Dose
00001
Day
0.0001
Dose*Day
0.0001
Control
Dose Lo
Hi
Control
Dose Lo
Hi
5/10
0207
0)80
0226
7/4
0015
3630
21 345
5/11
0156
4586
10141
7/5
0010
3 147
20251
5/16
0073
2878
7681
7/11
0017
1 785
17 197
5/18
0168
8081
18676
7/12
0009
1 539
16.299
5/23
0030
6 134
16156
7/18
0015
0811
12475
5/25
0018
9451
23740
7/19
0013
0706
11509
5/30
0009
6454
18771
7/25
0011
0323
8273
5/31
0011
10489
27882
7/26
0010
0250
7588
Day
6/1
0009
10149
27510
8/1
0007
0150
5079
6/6
0009
6849
23793
8/2
0008
0122
4580
6/7
0009
12072
35209
8/8
0005
0013
1 708
6/13
0008
8685
30191
8/9
0.005
0017
1364
6/14
0014
12634
38566
8/15
0006
0006
0329
6/20
0010
9155
32561
8/22
0005
0005
0013
6/21
0010
8078
29666
8/29
0005
0006
0005
6/27
0011
5657
25343
9/5
0005
0005
0005
6/28
0014
5.294
24572
9/12
0005
0005
0005
Season
Avg.
0027'
4098b
15256'
Figure 40. Changes in nitrate concentrations over time. The upper table presents probabilities that dose, day, and
interactions of dose and day influenced nitrate in ANOVA of rank-transformed data. The graph is a plot
of nitrate over the experimental season pooled by treatment. Dark circles indicate dose dates. The lower
table lists LS Means (mg/L) represented in the graph, along with statistical information based on the rank
transformed data. Within a column, values sharing a letter are not significantly different (p>0.05). In col-
umns without letters, values are not significantly different.
N
Dissipation
Dose
Day
ooooi
Dose*Day
ft.-
» ^ 1 2 -
*% i
II
O.O OX
f z °8
5 s n ft n
0 9
0
NO3-N Dissipation, for Study 2
+LO \
•HI T
('
T .1
I „ !
I 1 i S I t l
J i1 it
* A i
1
123456
Dose
Figure 41. Changes in nitrate dissipation rates over dose periods. Nitrate dissipation rates (mg NO3-N/L/day) in the
week following each dose shown as LS Means for Lo and Hi treatments. Bars indicate the one standard
deviation above and below the mean. Dissipation rates were not significantly different between Lo and Hi
(p>0.05).
53
-------
Ammonia in the Control was near the limit of detection (0.005 mg NH3-N/L) throughout most of the season (Figure 42).
During the dose period, however, ammonia in the Control rose to 0.045 mg NH3-N/L, possibly due to the disturbance of
stirring. Dose (p=0.0003), Day (p<0.0001), and the Dose*Day interaction (p<0.0001) were significant factors in dynam-
ics of ammonia. Ammonia in the amended treatments peaked on dates of nutrient application at 4 times Control levels.
Seasonal maximums in the Lo treatment (0.152 mg NH3-N/L) and Hi treatment (0.281 mg NH3-N/L) occurred following
the fourth dosing (May 31). For four weeks following the dose period, ammonia in the Lo treatment was significantly
greater (p<0.05) than Control levels. After July 18, ammonia in Lo was not significantly different (p>0.05) from the Control.
After the dose period, ammonia in the Hi treatment varied between 0.1 and 0.25 mg NH3-N/L until August 8. Thereaf-
ter, ammonia in the Hi treatment decreased >85%, and in September, concentrations were comparable to the Control.
Overall, study averages of ammonia in the Lo treatment (0.038 mg NH3-N/L) and Hi treatment (0.127 mg NH3- N/L) were
significantly different from each other and exceeded the Control average (0.012 mg NH3-N/L).
NH3-N
Dose
0.0003
Day
0.0001
Dose*Day
0.0001
NH3-N, for Study 2
0.005
5/10
5/24
6/21
7/5
7/19
8/2
8/16
8/30
Date
Control
Dose Lo
Hi
Control
Dose Lo
Hi
5/10
0008
0.011
0012
7/4
0.005a
0.022b
0.238C
5/11
0006a
0074b
0178b
7/5
OOOSa
0020b
0195c
5/16
0016
0021
0032
7/11
OOOSa
0.026b
0.226c
5/18
0.01 7a
0083b
0122b
7/12
OOOSa
0020b
0201c
5/23
0.0 17a
0 020ab
003 Ob
7/18
OOOSa
0014a
0135b
5/25
0028a
0.1 44b
0.244b
7/19
0007a
0.014a
0128b
5/30
0045
0105
0.141
7/25
0006a
0012a
0093b
5/31
0.033a
0.1 52b
0.281b
7/26
O.OOSa
0.013a
0096b
Day
6/1
0044a
O.lOlab
0214b
8/1
O.OOSa
OOOSa
0184b
6/6
0030
0038
0071
8/2
OOOSa
0006a
0142b
6/7
0021a
0094b
0217b
8/8
0012a
OOlla
0082b
6/13
0.006a
0033b
0137b
8/9
O.OOSa
0009a
0054b
6/14
OOOSa
0069b
0256b
8/15
0006a
0007a
0.044b
6/20
O.OOSa
0028b
0.1 OSc
8/22
OOOSa
0009b
0027b
6/21
O.OOSa
0.027b
OlOOc
8/29
0006
0007
0012
6/27
OOOSa
OOSSb
0162b
9/5
0010
0.009
0029
6/28
O.OOSa
0.027ab
0.136b
9/12
0009
0008
0.008
Season
Avg.
0012'
0038b
0.127'
Figure 42. Changes in ammonia over time. The table presents probabilities that dose, day, and interactions of dose
and day influenced ammonia in ANOVA of rank-transformed data. The graph is a plot of ammonia over
the experimental season pooled by treatment. Dark circles indicate dose dates. The lower table lists LS
Means (mg/L) represented in the graph, along with statistical information based on the rank transformed
data. Within a column, values sharing a letter are not significantly different (p>0.05). In columns without
letters, values are not significantly different.
54
-------
NitrogerrPhosphorus Ratio
Dose (p<0.0001), Day (p<0.0001), and the Dose*Day interaction (p<0.0001) were significant factors influencing the TN:
TP ratio. The TN:TP ratio in the Control fluctuated between 20 and 30 from the beginning of the experiment through
July 18, indicating P-limitation (Figure 43). The TP increase on July 25 dropped TN:TP to 15:1, and ratios were between
10 and 20 during the remainder of the season. The calculated TN:TP of the amendments was 25:1, but after the first
two doses, ratios in Lo and Hi had doubled to more than 50:1, indicating that added phosphorus was rapidly lost from
the water column whereas nitrate accumulated (Figures 38 and 40). Following the third amendment, TN:TP ratios in
the dosed treatments exceeded 200, nearly ten times the Control. During July, TN:TP ratios in Lo and Hi decreased
because of the TN decline, and in August reached levels between 10 and 20, comparable to the Control.
Dose Day
TN:TP 0.0001 00001
450 -i
400
TCA
•3 fin
H 250
1 CA _j
100 -
1
TN:TP, for Study 2
S \
-/ \ S* *\
L^-'*~~+ V '^l* ""^
r ''^ v ^
1 v ^
Dose* Day
0.0001
--A--- Control
--»- Lo
— •— Hi
h<^"-A A A i A A A - - A - - - - A- . . ."J.- .-. * -.^Ir~~~t»=— . •
^. - -. - i
0 • • • • • 9 1 1 1 ~ = 1 " "i ™ •
5/10 5/24 6/7 6/21 7/5 7/19 8/2 8/16 8/30
Date
Control
Dose LO
HI
5/10
29
27
29
5/18
23"
58"
53"
5/23
22'
201b
238"
5/30
27'
232"
310"
6/6
21"
220b
411'
6/13
25*
191b
203"
6/20
26'
23 lb
182"
6/27
29"
203b
274"
7/4
30"
112b
293C
Day
7/11
30'
77'
223"
7/18
23"
40'
180"
7/25
15'
26'
112"
8/1
14'
17"
51"
8/8
15"
15"
24"
8/15
15
17
15
8/22
14
14
11
8/29
14
16
12
9/5
12
12
10
9/12
10*
14"
11"
Season
Avg.
21"
91b
139'
Figure 43. Changes in the ratio of total nitrogen to total phosphorus over time. The upper table presents probabilities
that dose, day, and interactions of dose and day influenced TN:TP in ANOVA of rank-transformed data.
The graph is a plot of TN:TP over the experimental season pooled by treatment. Dark circles on X axis
circles indicate dose dates. The lower table lists LS Means represented in the graph, along with statistical
information based on the rank transformed data. Within a column, values sharing a letter are not signifi-
cantly different (p>0.05). In columns without letters, values are not significantly different.
55
-------
pH
Day (p<0.0001) was a significant main effect on pH response; however, dose had no effect. Levels of pH in the Control
averaged 8.3 in May and early June, but then rose steadily to nearly 10 by mid-July (Figure 44). With the exception
of September 5, when pH was 8.8, pH in the Control fluctuated between 9 and 10 during July to September. Nutrient
enrichment did not significantly influence pH (p>0.05) in analyses based on the entire season. However, values in the
Lo and Hi treatments averaged >0.5 pH units higher than the Control during the May-June dosing period.
pH
Dose
?,;,• • ;,-;• -i«4K',., ii,(MI4ftl»'"««>p;Y;; •; ,• ••
Day
0.0001
Dose*Day
'" ^itis? ':;H5:':M'"
100 -
•3, *•»
8.5,
7 5 -
5/
pH Experiment 1, for Study 2
— j | * 0 j ,
sj£-~*' •*' S > ' * * ^\jV-r-^
^*^~*-^*r ^'' - -A- -Control \$S
**-*.--A>..A..A-" -«-Lo
— • — Hi j
»m m m m m
10 5/24 6/7 6/21 7/5 Date 7/19 8/2 8/16 8/30
Control
Dose LO
HI
5/10
84
84
83
5/16
83
85
8.7
5/23
83
87
90
5/30
8 1
88
87
6/6
83
93
91
6/13
86
96
93
6/20
89
99
97
6/27
92
1 0.0
99
I
7/4
96
100
100
)ay
7/11
100
99
99
7/18
99
98
98
7/25
9.8
9.7
99
8/1
97
91
9.7
8/8
94
99
95
8/15
97
99
9.8
8/22
97
95
97
8/29
97
94
96
9'5
88
8.5
90
9/12
9.2
92
92
Season
Avg.
87
90
91
Figure 44. Changes in pH over time. The upper table presents probabilities that dose, day, and interactions of dose
and day influenced pH in ANOVA of rank-transformed data. Darkened values are not significant (p>0.05).
The graph is a plot ofpH over the experimental season pooled by treatment (data were averaged by H-
ion concentration, then converted to pH: calculated pH= - log (H-ion)). Dark circles indicate dose dates.
The lower table lists LS Means represented in the graph.
56
-------
Alkalinity and Hardness
Alkalinity significantly responded to Day (p<0.0001), Dose (p=0.0019), and the Dose*Day interaction (p<0.0001) (Fig-
ure 45). Alkalinity in the Control decreased from a maximum of 233 mg CaCO3/L on May 10 to 80 mg CaCO3/L in
early July, and then fluctuated between 60 and 110 mg CaCO3/L through September. Initial alkalinity in 2000 (233 mg
CaCO3/L) was higher than initial values observed in 1999 (150 mg CaCO3/L) because water samples in 2000 were
collected within two days of filling the ponds, whereas over two weeks elapsed between filling and sampling in 1999.
Alkalinities ranged from 100-170 mg CaCO3/L in the Lo and Hi treatments during the dosing period and were 25% lower
than Control levels. Following the dosing period, alkalinities in the Lo treatment ranged from 94 to 105 mg CaCO3/L and
were similar to Control values. Alkalinities in the Hi treatment remained between 129 and 173 mg CaCO3/L through
September and were significantly higher than the other treatments (p<0.05).
| Alkalinity
Dose
0.0019
Day
0.0001
Dose*Day
0.0001
Alkalinity, for Study 2
50
- A - * '
..:.;.: .•X-A-. I -*
5/10
5/24
6/7
6/21
7/5 7/19
Date
8/2 8/16 8/30
Control
Dose LO
a
5/10
233
237
233
5/16
201
169
161
5/23
193'
135b
133"
5/30
179'
139'"
123"
6/6
171'
116b
127"
6/13
136'
107k
129"
6/20
111"
114'
146"
6/27
90'
103'
135"
7/4
80*
101'
129"
Day
7/11
77'
101b
129C
7/18
105'
100'
147"
7/25
93'
115b
167C
8/1
78'
102b
168C
8/8
69"
94"
155C
8/15
84*
97'
165"
8/22
80'
104b
172'
8/29
78'
99k
162'
9/5
90*
110b
173e
9/12
90'
103*
168"
Season
Avg.
117*
118°
154'
Figure 45. Changes in alkalinity over time. The upper table presents probabilities that dose, day, and interactions
of dose and day influenced alkalinity in ANOVA of rank-transformed data. The graph is a plot of alkalin-
ity over the experimental season pooled by treatment. Dark circles indicate dose dates. The lower table
lists LS Means (mg CaCOJL) represented in the graph, along with statistical information based on the
rank transformed data. Within a column, values sharing a letter are not significantly different (p>0.05). In
columns without letters, values are not significantly different.
57
-------
Hardness was similarly affected by Day (p=0.0001) but not by Dose (p=0.1353). Trends in hardness paralleled those
of alkalinity early in the study when water quality was highly influenced by the groundwater source at corral filling.
Maximum hardness (273 mg CaCO3/L) was observed at the beginning of the season (Figure 46). Hardness values
decreased approximately 50% by June 21 in the Control and averaged 148 mg CaCO3/L. Hardness values in the Lo
and Hi treatments were approximately 25% lower compared to the Control during the dosing period, but were similar
to Control values late in the study. Hardness averaged 140, 105, and 112 mg/L mg CaCO^L in the Control, Lo, and Hi
treatments, respectively, over the course of the entire study.
Dose
Day
Dose*Day
[Hardness
0.0001
Hardness, for Study 2
5/10
5/24 6/7 6/21 7/5 7/19 8/2 8/16 8/30
Date
Control
Dose LO
Hi
5/10
273
273
270
5/16
243
208
197
5/23
234
161
165
5/30
231
149
149
6/6
220
131
143
6/13
176
115
130
6/20
148
111
129
6/27
120
94
112
7/4
108
78
94
Day
7/11
101
74
86
7/18
101
68
78
7/25
100
69
79
8/1
95
64
74
8/8
79
62
67
8/15
83
64
70
8/22
87
70
76
8/29
78
63
68
9/5
91
73
75
9/12
88
70
73
Season
Avg.
140
105
112
Figure 46. Changes in hardness over time. The upper table presents probabilities that dose, day, and interactions of
dose and day influenced hardness in ANOVA of rank-transformed data. Darkened values are not signifi-
cant (p>0.05). The graph is a plot of hardness over the experimental season pooled by treatment. Dark
circles indicate dose dates. The lower table lists LS Means (mg CaCOJL) represented in the graph.
58
-------
Conductivity
Conductivity averaged 382, 429, and 601 |jS/cm in the Control, Lo, and Hi treatments, respectively; Day (p=0.0001),
Dose (p=0.0007), and the Dose*Day (p=0.0001) had significant effects (Figure 47). Conductivity in the Control de-
creased 60% during the season, from an initial maximum of 625 uS/cm, to 268 uS/cm in September. Conductivity in
the Lo treatment was similar to the Control during the dose period, and around 25% higher than the Control during
the remainder of the season. Conductivity in the Hi treatment increased substantially after dose two, and fluctuated
between 600 and 800 uS/cm during most of June and July. In early August, conductivity in Hi decreased to 500 uS/cm,
but remained significantly larger than the other treatments through the end of the season (p<0.05).
| Conductivity
Dose
0.0007
Day
00001
Dose*Day
0.0001
"E
_o
i
u
•o
s
u
7AA
-inn
200
5/
Conductivity, for Study 2
--*•-• Control
m-— — •" "^"^-B.. xA - •» — Lo I
L ^*^^ *\^ • B/' \ 1 • Hl i
^~9z*^¥^-"±~--t'- --» ^* ^
"*" *•--. "*--»- — _^_ ^^+^. m — ~ • — • •
" " - -A - - - -A - - - -* - - .~T "t. - - ' A - ^ s^_ _ -» A I
"A — A — I- . . . J: . . - f- - -"^
;
10 5/24 6/7 6/21 7/5 7/19 8/2 8/16 8/30
Date
Control
Dose LO
HI
5/10
625
622
622
5/16
573
538
571
5/23
5201*
466'
545"
5/30
556"
514'
639"
6/6
548"
523'
712"
6/13
464'
506"
739'
6/20
402"
526b
792€
6/27
346"
476"
714'
7/4
329"
446b
675'
Day
7/11
357"
417*
570"
7/18
305"
380b
624'
7/25
305"
385b
607C
8/1
383"
471 b
749'
8/8
256"
314b
4921
8/15
263"
330b
503'
8/22
259"
313b
482C
8/29
241"
291'
454r
9/5
266'
311b
469C
9/12
268"
315"
465'
Season
Avg.
382"
429"
601'
Figure 47. Changes in conductivity over time. The upper table presents probabilities that dose, day, and interac-
tions of dose and day influenced conductivity in ANOVA of rank-transformed data. The graph is a plot
of conductivity over the experimental season pooled by treatment. Dark circles on X axis indicate dose
dates. The lower table lists LS Means (mS/cm) represented in the graph, along with statistical information
based on the rank transformed data. Within a column, values sharing a letter are not significantly different
(p>0.05). In columns without letters, values are not significantly different.
59
-------
Turbidity
Turbidity ranged from 1.4 to 12.0 NTU's during the study. Day and the Dose*Day interaction had significant effects; Dose
had no effect. Turbidity in the Control ranged from 1 to 4 NTU's through mid-July, and then increased to a maximum of
8.8 NTU's on September 5 (Figure 48). Overall, turbidity in Lo and Hi was similar to the Control, but significant differ-
ences (p<0.05) were observed on four dates during the season based on Day-specific T-tests. After two doses, turbidity
in the Hi treatment (4.8 NTU's) was significantly greater than the Control (2.8 NTU's) (p<0.05). Also, for three weeks in
late June and early July, amended treatments exhibited turbidities of 4-8 NTU's, 2-3 times Control levels.
| Turbidity
Dose
tyiSI-t,* 'Jf >,:•>*:,>,'!.% j^aff:" ,~'.~- ',
^l...... W^. ^k^M^Y&Wffi^-M" ' ^^W'^'
Day
00001
Dose*Day
0.0199
Turbidity, for Study 2
5/10
5/24
6/21
7/5
7/19
8/2 8/16 8/30
Date
Control
Dose LO
HI
5/10
3.9
3.8
3.5
5/16
33
39
58
5/23
28"
3.1-"
48l
5/30
25
2 1
33
6/6
2.1
34
3.3
6/13
4 1
55
60
6/20
2.0
1.8
1.9
6/27
1.4'
5.4k
42b
7/4
1 9'
51*
7.9"
Day
7/11
14'
4.7b
5.1"
7/18
21
30
34
7/25
49
68
41
8/1
50
59
7.4
8/8
3.3
5.4
2.8
8/15
76
120
12.4
8/22
7.0
6.7
5.3
8/29
8.3
72
5.3
9/5
8.8
103
101
9/12
4.6
6.0
4.7
Season
Avg.
40
54
5.3
Figure 48. Changes in turbidity over time. This combination of tables and a graph presents the data and statistical
information pertaining to turbidity (NTU). The upper table presents probabilities that dose, day, and inter-
actions of dose and day influenced turbidity in ANOVA of rank-transformed data. Darkened values are
not significant (p>0.05). The graph is a plot of turbidity over the experimental season pooled by treatment.
Dark circles indicate dose dates. The lower table lists LS Means (NTU) represented in the graph, along
with statistical information based on the rank transformed data. Within a column, values sharing a letter
are not significantly different (p>0.05). In columns without letters, values are not significantly different.
60
-------
Phytoplankton
Phytoplankton biomass, measured as chlorophyll a, responded significantly to the effects of Day and the Dose*Day
interaction (Figure 49). Chlorophyll a averaged 13, 25, and 26 ug/L Chi in the Control, Lo, and Hi treatments, respec-
tively; however, levels varied seasonally within treatments, and therefore, there was no significant main effect of Dose.
In the Control, phytoplankton biomass as Chi was <4 ug/L from May through mid-July. In late July, as macrophytes
matured, Chi increased to 32 ug/L by late July, and then varied between 14 and 46 ug/L through September. Signifi-
cant deviations from the Control were noted in the Lo and Hi treatments between dose initiation and mid-July (p<0.05),
but not during the remainder of the study. In the Hi treatment, chlorophyll peaked on May 15 (35 ug/L) and again on
July 11 (86 ug/L). Chlorophyll in the Lo treatment was not significantly different from Hi, but exhibited peaks of 33 ug/L
on June 20, and 50 ug/L on July 11.
Dose
Day
Dose*Day
PkytoplaBkton
Chlorophyll
0.0086
Chlorophyll, for Study 2
5/10
5/24
8/30
Control
Dose LO
HI
5/10
5
6
6
5/16
2"
10"
35"
5/23
2'
6'"
11"
5/30
3
3
3
6/6
r
7b
14"
6/13
2"
23"
6a
6/20
3'
33"
8'"
6/27
r
2*
11"
7/4
2'
20"
30"
Day
7/11
3'
50"
86"
7/18
4
15
19
7/25
32
55
20
8/1
29
24
35
8/8
15
28
24
8/15
25
29
22
8/22
38
27
60
8/29
46
75
63
9/5
17
27
23
9/12
14
45
19
Season
Avg.
13
25
26
Figure 49. Changes in phytoplankton chlorophyll concentrations over time. The upper table presents probabilities
that dose, day, and interactions of dose and day influenced Chi in ANOVA of rank-transformed data.
Darkened values are not significant (p>0.05). The graph is a plot of Chi over the experimental season
pooled by treatment. Dark circles on X axis indicate dose dates. The lower table lists LS Means (mg/L)
represented in the graph, along with statistical information based on the rank transformed data. Within a
column, values sharing a letter are not significantly different (p>0.05). In columns without letters, values
are not significantly different.
Particulate organic carbon (POC) significantly increased over time (p<0.0001) but did not respond to nutrient dosing
(Figure 50). The Control, Lo, and Hi treatments averaged 5.14, 5.71, and 5.29 mg C/L, respectively, for the entire study.
Control POC levels fluctuated around 2 mg C/L from May through early July; POC's gradually increased in the Control
treatment to a maximum of 18.4 mg C/L on August 22, and then decreased to 7.6 mg C/L by the end of the season.
Day (p<0.0001), Dose (p=0.0099), and the Dose*Day interaction (p=0.0028) had significant effects on the POC:Chl ratio
of water (Figure 51). The POC:Chl ratio averaged 1044, 530, and 485 in the Control, Lo, and Hi treatments, respectively,
when averaged over the entire study. Control POC:Chl ratios were significantly greater than those in Lo and Hi treatments
during the early part of the study due to the observed increase in chlorophyll from phytoplankton stimulation (Figure 49).
POC:Chl ratios in the Control ranged from 1000:1 to 2500:1 from late May to mid-July, but decreased to less than 500:1
during the remainder of the season as chlorophyll concentrations increased in water (Figure 51). Though carbon and
chlorophyll increased in August and September, the smaller ratio was due to the greater proportional contribution of
61
-------
POC
Dose
Day
0.0001
Dose*Day
Participate Organic Carbon, for Study 2
20 -,
i
§
u
o
5/10
5/24
6/21
7/5 7/19
Date
8/2
8/16
8/30
Control
Dose LO
HI
5/10
261
2.80
2.36
5/16
205
2.46
5.49
5/23
2.71
2.44
309
5/30
2.10
261
240
6/6
2.25
356
3.63
6/13
2.04
3.75
2 16
6/20
1.58
256
1 33
6/27
140
1 14
146
7/4
1.84
4.41
4.12
Day
7/11
3.32
941
8.29
7/18
3.52
4.59
10.29
7/25
5.84
940
3.95
8/1
450
243
2.89
8/8
5.28
7.75
451
8/15
7.26
7.49
5.47
8/22
18.38
954
10.90
8/29
13.45
15.21
9.12
9/5
10.02
6.87
1142
9/12
7.59
10.01
767
Season
Mg.
5.14
5.71
5.29
Figure 50. Changes in paniculate organic carbon concentrations over time. The upper table presents probabilities
that dose, day, and interactions of dose and day influenced POC in ANOVA of rank-transformed data.
Darkened values are not significant (p>0.05). The graph is a plot of POC over the experimental season
pooled by treatment. Dark circles on X axis indicate dose dates. The lower table lists LS Means (mg
POC/L) represented in the graph.
chlorophyll. POC:Chl ratios in the dosed treatments averaged <1000:1 and were significantly lower than the Control
on most dates between May and mid-July. Because POC values were similar at that time, lower POC:Chl ratios in the
dosed treatments reflect a greater proportion of living algal biomass in the suspended carbon pool as compared to the
Control. POC:Chl ratios in the dosed treatments averaged <500:1 in August and September, similar to the Control.
Dose had no significant effects (p>0.05) on the major taxa of phytoplankton; however, Day was a significant fac-
tor (p=0.0001) as phytoplankton increased seasonally (Figure 52). Initial numbers of phytoplankton were less than
0.4*106cells/L in May samples, and numbers were evenly distributed across major divisions. Phytoplankton numbers
increased in June, however, to a community dominated by chlorophytes. Phytoplankton numbers increased further in
July, after dosing had ended, to an average of 11.8*106 cells/L; approximately 80% of the community was chlorophytes,
whereas cyanophytes comprised 20% of the community. Total numbers of algae declined between July and August to
approximately 6.0*106 cells/L but shifted in proportions to equal numbers of Chlorophytes and Cyanophytes. By Sep-
tember, however, algal cell numbers had increased to a seasonal high of 25.7*106 cells/L and a community dominated
by Cyanophytes. On all dates Euglenophytes, Bacillariophytes, and Cryptophytes were rare. Thus, the phytoplankton
community was more sensitive to seasonal changes in light and temperature than to nutrient dosing.
A total of 137 species of algae/cyanobacteria were observed in Study 2; the list of observed phytoplankton species is
presented in Table 7. We observed 62 species of Chlorophytes; 37 species of Bacillarophytes; 15 species of Cyanophytes;
13 species of Euglenophytes; 5 species of Chysophytes; 4 species of Cryptophytes; and 1 species of Pyrrophycota.
Early in the season, the Chlorophytes were dominated by Gleocystis, Scenedesmus, and Oedegonium sp.; whereas,
the Cyanophytes were dominated by the filamentous Oscillatoria sp. By the end of Study 2, the algal community was
dominated by the filamentous cyanophyte Oscillatoria sp. and the chlorophytes Pleodorina, Oocystis, Characium, and
Oedeogonium sp.; lesser amounts of the Euglenophytes (Trachlemonas and Euglena sp.) and the Bacillariophytes
(Navicula, Nitzschia, Fragilaria, and Gomphonema sp.) were observed.
62
-------
| POC.-Chl
Dose
0.0099
Day
0.0001
Dose*Day
0.0028
POC:ChI, for Study 2
5/10
5/24 6/7
6/21
7/5 7/19
Date
8/2
8/16
8/30
Control
Dose LO
HI
5/10
670
562
488
5/16
2413'
301b
223"
5/23
1854"
496b
276"
5/30
1135
1418
979
6/6
2008'
875*
397"
6/13
1730'
1404"
689'"
6/20
1400'
222*
245"
6/27
1427'
479b
206C
7/4
1419'
242b
684"
Day
7/11
1400"
695b
409"
7/18
1010"
434b
1572'
7/25
285
259
226
8/1
235
181
397
8/8
353"b
632'
208"
8/15
281
334
260
8/22
434
380
310
8/29
369
342
276
9/5
892
419
732
9/12
523
404
634
Season
Avg.
1044"
530b
485"
Figure 51. Changes in the ratio of paniculate organic carbon to phytoplankton chlorophyll over time. The upper table
presents probabilities that dose, day, and interactions of dose and day influenced POC:Chl in ANOVA of
rank-transformed data. The graph is a plot of POC:Chl over the experimental season pooled by treatment.
Dark circles indicate dose dates. The lower table lists LS Means represented in the graph, along with
statistical information based on the rank transformed data. Within a column, values sharing a letter are
not significantly different (p>0.05). In columns without letters, values are not significantly different.
Total
Chlorophyta
Euglenophyta
Bacillanophyta
Cryptophyta
Cyanophyta
5/11/00
0.3479
02280
00010
00725
00357
00107
6/7/00
10865
09270
00003
01225
00000
00367
7/5/00
11 8264
83968
00143
04033
00120
30000
8/2/00
60000
26262
00003
00711
00079
21320
9/5/00
25.7251
79720
08118
04051
00738
164624
Phytoplankton Division Abundance, for Study 2
OA
Number (Individuals /L)
oc
§
I.
in -
5
n
KM
&
«
H _ S 1L a
5/1 1 1QQ 6/7/00 7/5/00 8/2/00
Sample Date
1
L
9/5/00
•
E
z
a Total • Chlorophyta Q Euglenophyta n Bacillariophyta a Cryptophyta e Cyanophyta
Figure 52. Changes in phytoplankton abundance of the dominant divisions over time. Columns represent averages
of all corrals in each month to show general successional trends over the season.
63
-------
CD
C
CD
CM
CO
0)
T3
CO
O)
•^
Q
T3
CB
O
0)
"o
0
CO
0)
'o
CD
Q.
CO
c
_o
2
c
CD
Q.
O
.C
Q_
"b
•»— »
CO
(>-!
JB
.a
(0
'o
Q)
a
CO
Genus
E
re
ai
^
O
U)
in
re
O
c
0
Q
3
^*
a.
E
o
•o
u>
c
'2
-2
§
|~:
dd ddddddddddd u a
(0(0 (0(0(0(0(0(0(0(0(0(0(0 O) CO
s § 1 H
S^jslSS'-sSeo 5 -g 5 5
°oora£o'$c P'5.o.y.ycc
go§-u£g^"3o^^^-i=-fc'£
SooooSoSoop^Soo^S
8 %
1 « 8 re
§ S | §
O m O ra
§ ill
2 t» o =
^ o w .2
O Z O DC
(0
J>
S (0
U CD
O re
0 0
8 2
>:. w
£ O
O Z
Q)
re
a>
u
_^"
a.
o
re
O
re
o
u
.c
Q.
O
C
re
O
re
0)
o
s
d d
(0 (0
^2 03
a>
idiacea
:eae
Achnanth
Bacillarial
CD
re
0)
u
re
:•§ s
•C "75
*; •—
re IS
-G '5
u re
< m
0)
re
o>
u
^*
Q.
O
'w
re
'o
re
m
re
•»-•
JZ
a.
o
_re
J3Z
U
CO
m
0)
5
c
re
0.
a
o
3 'R
l
'55
.«
CD
•§ -2 •§ -5 -2
2 S S 2
CCCC
CDCDCDCD
o o o o
.w
re §
o> o
^ S
21 §
o E E
c >. o
< O CJ
w
>.
O
o>
re
o>
o
re
£
'u>
re
re
8
re
'C
_ro
'u>
re
Q)
re ^
0) U
Q re
JS '=
rr re
3
_o
re
w
re
O)
w
re
O)
re
u
a
o
_ra
'o
re
m
$
a
o
_re
'o
re
m
S
jo
a.
JO
3
_O
Q
64
-------
(0
0)
'o
CD
I*
p
§
'
I?
Q)
£ cS
ESS
- - •
-
O) O)
3 3
E E
O) O)
-S
'qj
3 3
E E
in
3
. <°
O
s
CD
c
o
O
I
1 o
O (B
!= Q.
CO OC
g> c
.± 0>
1- 4=
3 eo
co O
CO
0)
CD
0)
M
o o>
fi
5 5
OO
d>
CO
0)
u
co
u
u
o
u
o
O
ro
!
£
8
8
O
0>
CO
0>
u
§.
S
Q
5.
8
O
Q) CO
to
^)
(0
'•5
^
(0
Q.
O
(0
2 'to
»%
'^ CO
3 £
CO I-
w w
_0> _Q)
•
II
0 0
10 0)
-2 I
•c o
O. o
o "
TJ O
J5 Z
O O
CD
o
^
"o
o
tfl
CD
'o
0)
Q.
CO
_cO
Q.
o
">>
^z
CL
M—
O
"5
jo
10
c
_o
'55
'>
IP
&
o>
s
u
>
£
Q.
O
_o
o
2
>>
o.
2
5
~E
o
Q.
E
j3
o
CO
I-
Q.
2
o
o
0)
'!
65
-------
CD
C.
O
CO
f—
CO
C\J
CO
CD
! .
CO
en
^c
Q
~Q
0)
o
—
"o
Q
CO
CD
'o
Q.
CO
C
o
^
c
CO
CL
O
JZ
CL
"o
to
LJ
0
^3
i^
co
'o
0)
a
CO
CO
c
to
0
1
0)
J^
O
co
co
CO
O
c
o
'55
,^
Q
|
^
a.
o
•o
D)
C
^
armatus
bernardii
'esmus
'esmus
o o
CD CD
c c
CD CD
O O
CO CO
"CD
o>
CD
co
c
bijuga
dimorphus
longus var.
'esmus
'esmus
'esmus
a a a
CD CD CD
c c c
CD CD CD
o o o
co co co
en
quadricaut
'esmus
o
CD
c
CD
C_5
to
a. a.
co co
CO S
O CO
11
li
^^ fV"y
Microsporaceae
Oedogoniaceae
co co
_CU Q)
2 «
o 'E
Q. 0
v) O)
o P
.ii o>
S O
sp.
schroeteri
sp.
if
"c "
o S x
o> CD •£:
-S co £:
^ -C r?
cu ci -2
O c/)3
o>
Palmellopsidacea
Ulotrichaceae
CO
o> w
CO 0)
^ ™
o ***
(A ^
E "S
fli ™™
i— ^
tenerrima
sp.
sp.
3
'domonas
'^'^ &
fc t~ CD
-9 co 5
3 G 0
0)
CO
0)
u
Chlamydomonadc
co
_o>
CO
u
1
o
:>
1
ionium
w)
5
G
^ ^^ 1
y -2 CD Q. "co Q.
-3 CD Q. co u co
1
8 co g
S 2 r- -S P
2 -i 1 o o x
c k ^ -2 "o R
5 •§ '5 ^ § §
CO ^3 O CO -5 "o
0)
m
Haematococcacei
Volvocaceae
sp.
incurvum
£ £
sz c
CD CD
CO CO
o o
0 O
Desmidiaceae
co
_g>
to
(0
E
c
^^
N
CD
CO
0)
U
£
0.
0
o
£
0
(0
*^
^
Q.
O
J
O
CD
5
c
co
Q.
I
-S
CO
*-»
:g | 1 6 1 |co 1 §
-=:-3m P3CD -srC Co
Q)^2s fccop 3 . 3 COCOX;
3^c -E52« . &•£ .S1^^
IK ^ CD ao2 o Q..C d.a-STo'S
-S-C^COUOOCOOOlOCDO-S
.CO
1 1 1 i i i 1 1 1 -s -s t co
cj3c!3cD| ll^ilHI^
ooog gotaco^ooo-^
OOGO OOcococo§§5co"
a>
(0
0>
u
la
0)
0)
66
-------
CD
13
EZ
'c
0
0
CD
"0
CO
Studies 2
D)
g
13
Q
"O
0)
0
0)
—
o
0
00
CD
'o
CD
a.
CO
o
•y
c
CO
Q.
O
Q.
"o
To
F»l
fl)
—
co
(A
'5
0)
a
(0
Genus
CO
li.
o5
O
(A
(0
co
0
o
(A
_>
5
E
_3
>>
Jd
CL
E
o
•o
D)
_C
^
a. Q.
(A (A
Zygnema
Chromulina
hromulinaceae
o
(A
0)
hromalina
O
CD
co
0)
u
>
.c
Q
0
(A
>
0
CO
4-i
^,
*^
a
o
(A
^
^3
o
a a.
(A (A
Chrysococcus
Mallomonas
a>
hrysococcacea
vnuraceae
0 03
(A
(U
CO
•o
chromona
O
t
Q. a
(A (A
Synura
Chrysamoeba
V
CO
hizochrysidace
DC
(A
Q)
CO
•u
hizochrysi
cc
-2
Q. a. Q. 5
in tz
unknown
Chroomonas
Rhodomonas
Rhodomonas
V
co
ryptomonadace
O
(A
—
co
TJ
co
O
E
0
Q.
0
n\
CO
(1)
o
f-
Q.
Q.
^_
O
to
-S
_ to to
-S 3 co
S 3 "cS "5.
CO "5 O O
CO t3 ^ CJ J~
• 3 3 c o -2
OL u co O C; i;
(A CO O O CO O
q) Q) Q) 03 ^ ?
O) Oi Ol O) CO
u
CO
_0)
D)
UJ
CO
(D
0
^i
j£
Q.
O
C
Q)
D)
UJ
nj
O
U
>*
Q.
O
0)
O)
3
UJ
CO
CO "O ,
•§ co a
^ O) ^5
^^ c Cr
5 -2 '£
o o o
CD co co
£ £ £
uglenaceae
UJ
uglenales
UJ
Q>
CO
0)
u
^1
J£
Q.
O
C
0)
CO
3
UJ
ro
o
o
>»
Q.
O
0>
D)
3
UJ
(D
TO
C
E
Q.
(A
CO
Trachelomona
co
c
o
§
§
Trachlemonas
cc
'S.
'-c
Trachlemonas
CO
W •—
ll
Trachlemonas
Peridinium
eridiniaceae
CL
eridiniales
CL
o>
CO
0)
u
Q.
O
Q
2
.c
Q.
o
u
£
O
^>
^^
Q.
67
-------
Periphyton
Periphyton accrual rates were significantly affected by Day (p<0.0001) and Dose (p<0.003) for both the one- and two-
week accrual intervals (Figure 53). In addition, the Dose*Day interaction was significant (p=0.0174) for the 1-week ac-
crual data. One-week periphyton accrual rates averaged 0.08, 0.44, and 0.88 ug Chl/cm2/wk in the Control, Lo, and Hi
treatments, respectively. Two-week periphyton accrual rates averaged 0.08, 0.39, and 0.78 ug Chl/cm2/wk, respectively,
in the Control, Lo, and Hi treatments. Control values peaked in June and September, averaging 0.14 ug Chl/crrWwk,
and were 3-fold rates in May, July, and August (average 0.04 ug Chl/cm2/wk). Highest levels of periphyton accrual in
the Lo (1.23 |jg Chl/cm2/wk) and Hi (2.75 ug Chl/cm2/wk) treatments occurred in May following the initiation of dosing
when nutrients and light were un-limited. Accrual levels in this study were 10-fold higher than in Study 1 due to the
earlier timing of dosing, higher levels of dosing, and decreased competition with macrophytes due to the study design.
Biomass remained significantly higher in the Lo and Hi treatments in June and July but at lower levels than in May as
nutrients and light began to limit periphyton growth. In general, the 1 -week and 2-week accrual rates were similar within
both the Control and Lo treatments. However, the 1-week and 2-week accrual rates varied in the Hi treatment most
likely due to variation in levels of self-shading of periphyton and macrophytes.
Periphyton 1-
week
Periphyton 2-
week
Dose
0.0011
0.0026
Day
0.0001
0.0001
Dose*Day
0.0174
,nW • ®'"fJ*«f'. •- '
Periphyton Accrual in 1- and 2-Week Exposures, for Study 2
May
September
Control
Dose LO
Hi
Month
May
I-wk
0.07*
1 23b
2.75"
2-wk
0.04
1.22
1.29
June
1-wk
0.15"
075b
0.69"
2-wk
013
0.53
1.35
July
l-wk
0.02"
006b
0.20"
2-wk
0.03
0.04
0.37
August
1-wk
004"
0.07"
0.59"
2-wk
0.04
0.06
0.27
September
1-wk
0.13
0.06
0.16
2-wk
0.16
0.12
0.64
Season
Avg.
1-wk
008'
0.44b
0.88C
2-wk
008"
0.39b
0.78"
Figure 53. Changes in accrual rates of periphyton chlorophyll in 1- and 2-week exposures over time. The 1- and
2-week datasets were analyzed separately. The upper table presents probabilities that dose, day, and
interactions of dose and day influenced accrual rates in ANOVA of rank-transformed data. Darkened val-
ues are not significant (p>0.05). The graph is a plot of accrual rates over the experimental season pooled
by treatment. Solid and striped bars represent 1-and 2-week exposures, respectively. The lower table
lists LS Means (mg Chl/cnf/wk) represented in the graph, along with statistical information based on the
rank transformed data. Within a column, values sharing a letter are not significantly different (p>0.05). In
columns without letters, values are not significantly different.
68
-------
Zooplankton
Day (p=0.0001) was a significant main effect influencing total numbers of zooplankton as well as numbers of cladocer-
ans, copepods, and rotifers (Figure 54). Dose had no effect on zooplankton numbers. The Dose*Day interaction was
significant only for total zooplankton numbers.
Identification of monthly zooplankton samples yielded 35 genera, including 16 rotifers, 12 cladocerans, and 7 copepods.
A list of observed species is presented in Table 8. Numerically, macrozooplankton (cladocerans and copepods) and
microzooplankton (rotifers and nauplii) represented 40% and 60%, respectively, of the total organisms collected each
month. This relative abundance did not significantly change (p>0.05) during the season or with treatment.
In the Control, abundance of zooplankton increased 20-fold from May to September (Figure 54). Total zooplankton
abundance was at a seasonal minimum in May (0.1*105 zooplankton/m2) and was similar among treatments. In June,
total zooplankton increased 6-fold in the Hi treatment (1.0*105 zooplankton /m2) and was significantly larger (p<0.05)
than in the Control (0.3*105 zooplankton /m2) and Lo (0.3*105 zooplankton /m2) treatments. In July total zooplankton
abundance increased in both the Lo (2.3*105/m2) and Hi (1.8*105/m2' treatments compared to the Control. Total zoo-
plankton numbers were similar across treatment in August and September. Regressions of Chi to total zooplankton
abundance indicated that relationships were significant for the Control (r2=0.41; p<0.005) but not for the Lo (r2= 0.16;
p<0.1) or Hi (r2= 0.08; p<0.5) treatments.
Total Zooplankton
Cladocerans
Copepods
Rotifers
Dose
0.6669
Day
0.0001
0.0013
0.0001
0.0001
Dose*Day
00167
Zooplankton Abundance, for Study 2
H Cladocerans • Copepods S Rotifers
C Lo Hi C Lo Hi C Lo Hi C Lo Hi C Lo Hi
May June July August September
Treatment: C Control, Lo, Hi
Control
Dose LO
Hi
May
13337
11420
16108
June
30719'
29409'
98059"
Month
July
65669'
232999*
181778b
August
223446
238095
339782
September
271863"b
335594'
154545b
Season
Avg.
121007
169503
158055
Figure 54. Changes in zooplankton abundance over time. The upper table presents probabilities that dose, day,
and interactions of dose and day influenced abundance of the total zooplankton community, cladocerans,
copepods, and rotifers in ANOVA of rank-transformed data. Darkened values are not significant (p>0.05).
The graph is a plot of abundance over the experimental season (Control (C), Lo, Hi). Abundances in the
respective categories (Cladocerans- stripes; Copepods- solids; Rotifers- checks) are stacked to indicate
totals. The lower table lists LS Means (# of indiv./rn2) for abundance of the total zooplankton community,
along with statistical information based on the rank transformed data. Within a column, values sharing a
letter are not significantly different (p>0.05). In columns without letters, values are not significantly differ-
ent.
69
-------
CO
c:
CO
CM
CO
CD
TD
CO
D)
C
Q
-£}
-2
CD
"o
o
co
CD
'o
CD
Q.
CO
O
V
C
CO
Q.
0
0
N
"6
00
1 1
co
JD
.Q
U)
.2
'o
Ct)
Q.
CO
(/)
3
C
0)
O
E
CO
(D
O
12
C/>
0)
6
(0
(0
CO
u
J3
3
(0
(0
co
O
3
>
Q.
.co ctj
•£ 2
d. € to
Q) O " — • C_ Q) O 3r ^^
"5cL-2cL''m °-5-S 0-5.0 m C^ **• c O- §" ^ O °-^ o"cc
D)(/)C3)co-ScT3JiTDOTCrv::-iSfc:C!.wc:
0) "o ro
CO » Q> TJ
~ £ 0) TJ ^
^ 2 5 1 f
Q. o ~ JS 9-
co co 2 ro .co
Q 2 a) O Q
CO
TJ
'5
C
CO
"5
O
CO
TJ
0
a.
a.
0
O
CO
TJ
o
a.
o
^z
'S
CO
2
c
1 ^ 1 «. .a 1
£ m -^2 -9 "S ^*
"S. ' — " ^ "^ C
•^ t/J ^ CtJ Q) O
^ CXJ fc Q^ ^ O
§ ^ ^ «
1 s s 8-
§ a s u
Ii-i-tii
||| |||
Q 5 CO CO ^C 0
o
CO
TJ
'a.
o
u
>.
O
CO
TJ
'6
Q.
2
O
>
0
•5S
J3
^ --2
ll
11
It
Q UJ
0)
CO
TJ
'5.
O
o
>>
O
CO
TJ
'o
a.
2
u
0
co
TJ
0
Q.
H
0
0
CO
TJ
O
Q.
O
!ZS
'jj
CO
5
CO
TJ
0
Q.
|
r
"*
CO
.c
CO
ci
opocyclops
K
70
-------
Q)
Ji
O
CO
c
CO
CM
CO
CD
,.
CO
O)
c:
3
Q
T3
-2
"o
.9?
O
0
co
O
'o
Q.
CO
C
o
X?
c:
CO
Q.
O
O
N
H
O
"co
1
co
CD
.Q
0)
'o
0>
Q.
CO
Genus
.-^
E
CO
u.
0>
^
o
CO
<5
TJ
6
co
CO
o
^
3
CO
(0
CO
CO
o
E
3
>,
Q.
05
g
.O
j|~
^
§
Conochil
Q)
CO
•JO
^M
!E
u
o
c
o
o
CO
TJ
'_5
Q*
CD
co
^
CO CO O
•0 0) C
0 "0 0
£ 1 I*
w -o o
o m s
CO
T3
o
Q. CO
2 oj
n ^
t; o
< cc
Is .to
CC 2w-§'Cco| SS'c
qj ^^COcCojSlS •&? -S C^^^CRcn
3>co .^^2"ca-S"oca:mca S ^ .^oS'^^?
^ — Q. CC Ct3 ^ ^ -2 O ^ ^ ^ CL Cl CL ^ CL cL ^ ^ ^ r~ -^ •£
5 coooacr^ oc.oc.co o,^ , co , cx^ a S
^C/)C/)C/)W) "Q\ mCT3CX3fQCXJ
£i^§S^|iB^^^§^^llo£(£^^^iSf£
0) (DO)
CD 5 <» ^ 2J 5
CO ~ n ' .5 2
•S c T> 0 -g 2 o
o "^ ^ c (0 JS ^rj Ji*
(C-Ci2O "O Cm M
== fe ™ Z 1 £ -g ?
— CO — Q (0 O *•* -^
._ X u. (Q t) *^ C O
— Q) (/) ^ fl) O>*'r"
U. I < CD _l ZCOH
1
'o
E
71
-------
Rotifers accounted for approximately 75% of the zooplankton numbers during the season. Rotifer populations signifi-
cantly increased (p<0.05) during the season, but the effect of nutrient enrichment on rotifer numbers was not significant
(p>0.05). In May and June, rotifers averaged <0.5*105/m2 and were dominated by Bdelloidspp., Euchlanisspp., Hexar-
tha mira, Lecane spp., and Monostyla spp. Rotifers increased from an average of 1.0*105/m2 in July, to 1.8*105/m2 in
September, when Brachionus spp., Euchlanis spp., H. mira, and Monostyla bulla dominated.
Cladocerans and copepods were at seasonal minimums in May (0.1*105/m2), peaked over 0.8*105/m2 in August, and
averaged 0.6*105/m2 in September. Cladocerans represented 97% of the macrozooplankton in May and were domi-
nated by Sididae spp. In June and July, Cladocerans accounted for 75% of the macrozooplankton and were dominated
by Simocephalus serrulatus and Ceriodaphnia spp. Copepod populations in June and July were approximately 43%
calanoids and 57% cyclopoids. In August, at the peak of macrozooplankton abundance, Cladocerans accounted for
67% of macrozooplankton numbers and were predominantly Alona spp., Ceriodaphnia spp., Chydorus sphaericus, and
Simocephalus serrulatus. Copepods were primarily cyclopoids (>95%) in August. Cladoceran populations declined at
the end of the season, and final macrozooplankton communities were close to 1:1 Cladocerans to copepods. In Sep-
tember, dominant genera in both Cladocerans and copepods were the same as in August.
Sediment
Nitrogen and phosphorus in sediments were significantly influenced by Day (p=0.0043), but were not affected by nutrient
amendments. Nitrogen content in sediments was greatest in June and September, averaging 0.35% N (Figure 55) which
was similar to nitrogen concentrations observed in Study 1 (Figure 27). In July and August, N content averaged 0.29%
N. Sediments contained approximately 157 g N/(m2*5 cm deep) in June and September, and around 137 g N/(m2*5 cm
deep) during July and August. Total nitrogen pools in sediment were similar to those observed in Study 1 (Figure 27).
Sediment
Nitrogen
Dose
- <***• (, _ -<."?/,
j*^;-03*tt> ^
-x ;, „ >**• ,
Day
0.0043
Dose*Day
Sediment Nitrogen Pool, for Study 2
June
July August
Month
September
Figure 55. Changes in the sediment nitrogen pool over time. The table presents probabilities that dose, day, and in-
teractions of dose and day influenced N content in sediments in ANOVA of rank-transformed data. Dark-
ened values are not significant (p>0.05). The graph is a plot of the estimated pool of N in the sediments
for an area of 1 rrf and a depth of 5 cm. Because treatment influences were not significant, monthly
values are presented as averages of all treatments. The average %N in sediments is shown within the
column for each month.
72
-------
Sediment phosphorus pools were not statistically related to Dose. Day (p=0.0004) was a significant main effect as sedi-
ment concentrations declined slightly over time (Figure 56). Sediment phosphorus concentrations were at a maximum
in June at 0.07% P and decreased slightly in subsequent months to approximately 0.06% P. Sediment phosphorus
pools ranged from 26 to 32 g P/(m2*5 cm deep) during the season. Sediment phosphorus data was similar to that
observed in Study 1 (Figure 28).
System Metabolism
Day had a significant main effect on observed values of gross primary production and community respiration (Figure 57).
Dose had no significant effect on either parameter when combined over the entire study; however, lack of effect is biased
by the large number of observations in the dataset, and differences among treatments were significant on several dates
during the dosing period (p<0.05).
u ,.
[Phosphorus
j«ifN
Dose
V< ' .x ~
$$p
'"SSf;
.
|ir«- ••:
Day
0.0004
Dose*Day
^>*«sS? ".:• ,. - '-"*
... , • £%'?. ;1
Sediment Phosphorus Pool, for Study 2
June
July August
Month
September
Figure 56. Changes in the sediment phosphorus pool over time. The table presents probabilities that dose, day, and
interactions of dose and day influenced P content in sediments in ANOVA of rank-transformed data. Dark-
ened values are not significant (p>0.05). The graph is a plot of the estimated pool of P in the sediments
for an area of 1 nf and a depth of 5 cm. Because treatment influences were not significant, monthly
values are presented as averages of all treatments. The average %P in sediments is shown within the
column for each month.
73
-------
Estimates of both diel gross primary production and community respiration increased 3-fold over the season, from around
4 mg O2/L in May and early June, to around 15 mg O2/L in August and September (Figure 57). Oxygen production in
the Lo treatment was significantly larger than the Control on May 18, May 31, June 1, and June 7. Oxygen production
in the Hi treatment was significantly larger (p<0.05) than the Control on May 16, May 18, May 31, June 1, and June 21.
Near the end of the season, oxygen production in the Hi treatment was significantly smaller than the Control on August 1,
August 2, August 8, and August 15. Respiration values were similar to production in magnitude, seasonality, and
responses to enrichment. Oxygen respiration in Lo was significantly larger than the Control on May 16, May 18, May
31, and June 1. Community respiration in the Hi treatment was significantly greater than the Control on May 16, May 18,
May 31, June 1, and June 21. Respiration in the Hi treatment was significantly less than the Control (p<0.05) throughout
much of August. Collectively, the data reflect the significant increases in primary productivity of phytoplankton (Figure 49)
and periphyton (Figure 53) due to dosing of nutrients early in the season prior to macrophyte development.
Production
Respiration
Dose
Day
0.0001
0.0001
Dose*Day
0.0001
0.0001
Community Production, for Study 2
5/10 5/17 5/24 5/31 6/7 6/14 6'21 6/28 7/5 7/12 7/19
Date
7/26
8/2 8/9 8/16 8/23 8/30 9/6
B
a 12 "
2 !
A- - - Control
• Lo
m _«^
Community Respiration, for Study 2
^ _^
/ A ^** \ ,
A. ..
/ \
""--- .---/' *-.\^'—
'A '•*" T*-^*r--*' •*- ~ ' jr^*-\
_n //*^ _ .':-1^--#rrr-\.-- *• ^^-- --__ /^^ *
/£*">- r^ '/•• . •"^TN,--*' s NJ f — x^
/• \ ^
>' A---ir.~
jr vs^^ f • ... - ^* ^^xl/' ~ ^ P^ j/'
7 A- - - *^s /•' ^V / "v>«x"
M A - - ~^J* •»
5/10 5/17 5/24
1
Control
Dose Lo
Hi
3.08
2.80
368
50
Control
Dose Lo
Hi
880
9.40
11.25
2
3.65
625
6.78
57
955
843
7.00
5/31 6/7 6/14 6/21 6/28 7/5 7/12 7/19 7/26 8/2 8/9 8/16 8/23 8/30 9/6
Dale
7
3.90
7.40
1000
63
11.
13
50
15
7.65
9
3.75
8.55
9.60
64
1085
13.03
13.93
14
478
4.50
710
70
1008
14.45
7.25
21
2.45
1.50
4 10
71
11 03
16.65
12.55
22
285
7.63
8.33
77
10.83
1623
12.30
Day
23
4.25
9.63
10 10
78
9.98
1325
788
28
6.95
7.28
9.08
84
12.98
12.95
4.83
29
4.43
8.45
743
85
1188
1320
6.90
35
465
4.43
4.55
91
15.88
12.88
8.75
36
2.88
5.90
548
98
17.10
1238
8.88
42
6.28
11.23
8.08
112
13.75
8.73
9.53
43
9.45
14.58
15.50
119
17.00
2003
1278
49
933
1050
10.78
126
11.33
10.65
15.55
Season
Avg.
8.51
10.20
8.92
Figure 57. Changes in community oxygen and respiration production over time. The table presents probabilities that
dose, day, and interactions of dose and day influenced production and respiration in ANOVA of rank-trans-
formed data. Darkened values are not significant (p>0.05). The graphs are plots of production (A) and
respiration (B) over the experimental season pooled by treatment. Dark circles indicate dose dates. The
lower table lists LS Means (mg/L) represented in the graph.
74
-------
Net Nutrient Balance
During Study 2, a maximum of 60 g N/m2and 2.4 g P/m2was applied. A final mass balance of nutrients was calculated
among various nutrient pools to determine the net efficiency of uptake and assimilation of nutrients in these experi-
mental systems (Table 9). At the end of the study, macrophytes contained a total of 4.29, 3.92, and 1.05 g N/m2 in the
Control, Lo, and Hi treatments, respectively; water contained an additional 0.63, 0.83, and 1.10 g N/m2 in the Control,
Lo, and Hi treatments, respectively. Combined (macrophytes + water), these two nutrient pools contained 4.92, 4.75,
and 2.15 g N/m2 stocks at the end of the study. Thus, a total of 16% (Lo) and 4% (Hi treatment) of total nitrogen added
during the study were found in these two major nitrogen pools at the end of the study. This implies, under simple mass
balance conditions (i.e., no loss to the atmosphere) that up to 84% (Lo treatment) and 96% (Hi treatment) of total nitro-
gen added were absorbed or lost to the sediments as detritus. Attempts to measure sediment nutrient dynamics did not
reveal the amount of nitrogen transferred due to a combination of the large mass of pre-existing nitrogen in sediments
and the error contributed by our sampling procedures. However, we know that these numbers are conservative, since
the Control treatment alone exhibited a total sequestration of 4.92 g N/m2 (combined macrophytes and water) at the end
of the study in the absence of external nitrogen addition. The large percentage of nitrogen that was not accounted for
at the end of the study (84%, Lo treatment; 96% Hi treatment) indicates that these shallow, vegetated aquatic systems
served as efficient biological, chemical, and physical sink for nitrogen.
Phosphorus, likewise, was efficiently assimilated and retained in the study. At the end of the study, macrophytes con-
tained a total stock of 0.80, 0.65, and 0.21 g P/m2 in the Control, Lo, and Hi treatments, respectively; water contained
an additional stock of 0.07, 0.08, and 0.12 g/P/m2 in the Control, Lo, and Hi treatments, respectively. Using conserva-
tive mass balance estimates, subtracting the amount of phosphorus in macrophytes and water from that applied in
dosing indicates that sediments had a net accrual of 0.47 (Lo treatment) and 2.07 (Hi treatment) g P/m2. Macrophytes
in the Control treatment contained 4-fold more phosphorus than the Hi treatment at the end of the study even though
no phosphorus was applied. This difference in macrophyte storage is largely due to the significant higher biomass
observed in the Control compared to the Hi treatment (Figure 32). Thus, even though the Hi treatment significantly
reduced macrophyte biomass compared to Controls, the system efficiently retained phosphorus.
Table 9. Summary Table of the Final Store of Phosphorus and Nitrogen in Water and Macrophytes in Study 2.
Load Represents the Total Amount of Fertilizer P or N Added in Each Treatment. Stores in Macrophytes
are Considered as Grams Per Cubic Meter Because Water Depth was 1 m
Nitrogen
Control
Lo
Hi
Load
(g N/m2)
0
30
60
Macrophytes Water ^^ +
(g N/m2) (g N/m2) Wfte' 2
VM ' \s / ^ N/m2)
4.29
3.92(13%)
1.05(2%)
0.63
0.83 (3%)
1.10(2%)
4.92
4.75(16%)
2.15(4%)
Presumed Sediment
Transfer
(g N/m2)
NA
25.25 (84%)
57.85 (96%)
sill* ft
Phosphorus
Control
Lo
Hi
Load
(g P/m2)
0
1.2
2.4
Macrophytes
(g P/m2)
0.80
0.65 (54%)
0.21 (9%).
Water
(g P/m2)
0.07
0.08(7%)
0.12(5%)
Macrophytes +
Water Total
' (g P/m2)
0.87
0.73(61%)
0.33(14%)
Presumed Sediment
Transfer
(g P/m2)
NA
0.47 (39%)
2.07 (86%)
75
-------
STUDY 3: Effects of Dosing in Relation to Macrophyte Stage
Study 3 was conducted concurrently with Study 2 in 2000 to evaluate the effect of the timing of nutrient additions in
relation to stage of macrophyte development. The results of Study 2 demonstrated that early, intense dosing of nutri-
ents significantly reduced macrophyte biomass compared to the Control treatment, but that shallow, vegetated aquatic
systems were still efficient in removing both nitrogen and phosphorus. There were three objectives in Study 3: 1) to
determine how the stage of macrophyte development influenced the concentration and relative distribution of nutrients in
the water column, macrophytes, and sediments; 2) to determine if the timing of nutrient addition in relationship to mac-
rophyte stage influenced species composition, biomass, and/or abundance of macrophytes, phytoplankton, periphyton,
and zooplankton; and 3) to characterize the assimilation and retention capabilities of shallow ponds receiving doses at
different stages of macrophyte development.
The hypothesis for Study 3 was that timing of additions would be influential in determination of community dominance;
formation of alternative stable states (i.e., phytoplankton or macrophyte dominance) would be created based on the stage
of macrophyte development at the initiation of nutrient dosing. At one extreme, phytoplankton and periphyton would be
stimulated by enrichment in the Early treatment. With little initial competition from the macrophytes, algal communities
would establish and maintain dominance throughout the season due to a growth rate that was higher than the grazing
rate of zooplankton consumers (Scheffer 1998), and by imposing light limitation on the macrophytes (Phillips et al. 1978).
In addition, overall nutrient uptake and assimilation would be reduced in the Early treatment. Mid and Late treatments,
however, would be macrophyte dominated and resist a shift to phytoplankton dominance due to removal of nutrients by
epiphytes/macrophytes (Scheffer 1990, 1998) and shifting to tall growth forms or species (Moss 1990). Large-bodied
zooplankton would promote water clarity through grazing, which would further stabilize macrophyte dominance (Brooks
and Dodson 1965, Scheffer 1998). Systems dominated by macrophytes at the tinning of nutrient addition would be more
efficient in nutrient uptake compared to a phytoplankton dominated system.
Experimental Design
Mesocosms were drained April 11, 2000 for corral construction. A total of 16 corrals were constructed in each of 4 me-
socosms (blocks). Ponds were refilled over a 2-day period (May 7-8) and allowed to mix for 2 days prior to raising of
sides of corrals on May 9. Water sampling began on May 10 and terminated September 12. Dosing began on May 11,
2000 and continued weekly for 6 weeks. A different set of ponds was used in Study 3 than in Study 1 to prevent bias
due to the previous study. Study 3 was conducted over the period of May 10 to September 12 of 2000. In Study 3,
nutrient additions were added during one of three stages of macrophyte growth: Early (0% cover; initiated May 11), Mid
(15-25% cover; June 12), or Late (75-90% cover; July 5) (Figure 58). The three dosed treatments received the same
nutrient load (30 g NO3-N/m2; 1.2 g P/m2) applied as six successive weekly additions of 5 mg NO3-N/L and 200 ug P/L
(25:1 N:P ratio).
Results
Macrophyfes
Macrophyte Taxa
The aquatic macrophyte community consisted of >99% A/ayas during the experiment. Chara, an attached macroalgae,
was sparsely present in May, and was not noted in subsequent months in any treatment. The lack of Chara sp. may
have resulted due to the fact that ponds were drawn down longer (26-day draw-down) in year 2000 compared to 1999
(16 days) which may have altered normal seasonal succession of the macrophyte community.
77
-------
o o
o o
o o
o o
o o
o o
Pondl
Pond 2
PondS
Pond 4
Figure 58. Corral and pond diagram for Study 3 experiments indicating orientation of corrals. Pond 1 shows an ex-
ample of the random assignment of the treatments.
78
-------
Macrophyte Biomass
Study 3 was initiated on May 10, 2000. There was no measurable growth of macrophytes above the sediment surface
at the beginning of the study (i.e., Early dosing). Macrophyte stage (Stage) had no significant effect on observed mac-
rophyte biomass (Figure 59). However, Day (p<0.0001) had a significant effect. Macrophyte biomass averaged 130, 80,
143, and 139 g/m2 in the Control, Early, Mid, and Late treatments, respectively, when averaged over the four monthly
sample dates (Figure 59). Maximum macrophyte biomass occurred in all treatments in early August; thereafter, macro-
phytes began to senesce. Macrophyte biomass was highly variable within treatments; at the peak of the growing season
biomass in replicate corrals differed by as much as one order of magnitude. In the Control, the maximum biomass was
213 g/m2 in August. Control biomass decreased by 32% to a final biomass of 144 g/m2 in September. Nutrient enrich-
ment had a weak influence on macrophyte development in the Early treatment (p<0.1), but had no influence in either
the Mid or Late treatments. Biomass in the Early treatment was only 36 g/m2 in early June, or approximately 25% of
macrophyte biomass in the other treatments. During July, biomass in the Early treatment increased 4-fold and peaked
at 167 g/m2. Peak macrophyte biomass (August sampling) in the Early treatment was 75% of the maximum biomass
in the Control (213 g/m2), Mid (222 g/m2), and Late (227 g/m2) treatments, but differences were not significant. Percent
loss of macrophytes due to senescence was similar in the Control and dosed treatments.
| Biomass
Stage
o$mi
Day
0.0001
Stage*Day
\ ,*' " «22B5, _v7!
250
3
m ~™
fr,-v 150
§ 1
£> >3 inn
Q.
2
o 0.05). The graph is a plot of macrophyte bio-
mass over the experimental season pooled by treatment. Corresponding arrows mark dose periods. The
lower table lists LS Means (g dry weight/nf) represented in the graph.
79
-------
Macrophyte Nutrients
Stage (p=0.0033), Day (p=0.0001), and the Stage*Day interaction (p=0.0421) were significant factors affecting
concentrations of nitrogen in aquatic macrophytes. In the Control, N in macrophyte tissues increased during the season
(Figure 60). Nitrogen content in the Control averaged around 2% N from June to August, then increased to 3.06% N in
September. In amended treatments, nitrogen uptake was enhanced during and immediately following the dose period. In
the Early treatment, N content in June (2.91% N) and July (4.13% N) was significantly higher than the Control (p<0.05),
but subsequent values were similar. Nitrogen content of macrophytes in the Mid treatment significantly increased in
July (3.61% N) after dose initiation, but was similar to the Control in August and September. Nitrogen content in the
Late treatment significantly increased in August (3.31% N) but was similar to the Control in September. Total season
averages indicated that the Late (2.92% N), Mid (3.21% N), and Elarly (3.34% N) treatments were significantly greater
than the Control average (2.27% N) which indicates the positive response to nitrogen dosing in all treatments.
Macrophyte
%N
Stage
0.0033
Day
0.0001
Stage*Day
0.0421
^— V
*•»
§ SP
Mg
£ £>
>• *"
A -O
v 0.05). In columns without letters, values are not signifi-
cantly different.
80
-------
Stage (p=0.0128) and Day (p<0.0001) were significant main effects on phosphorus concentration of macrophytes;
however, there was no Stage*Day interaction (Figure 61). Season total averages were significantly higher in the Early
(0.58% P), Mid (0.64% P), and Late (0.58% P) treatments compared to the Control (0.43% P). Phosphorus content
of macrophytes in individual treatments within the season did not reflect the effect of dosing due to high variability with
treatments.
Macrophyte
%P
Stage
0.0128
Day
0.0001
Stage*Day
'^4Wiyfc40» ' y *.;
^ ^
»> ^
3 s-
£ M
o '3
•a*
1 &
f ~°
a- <«
« °
££
£g
U +4
« s
« u
On
OS -,
0 7
n £
0 S -
0 4
0 ^
0? -1
0 1 -
0
Macrophyte % Phosphorus, for
/* — •-----_
^^.^ -j-j- - • -^
vC"' """ "^" """ "
^/S-' .- A A
f-'--''
A' - - A- - -Control —
— • — Mid -
June July August
Month
Study 3
~5"trt
..--A
-» - Early
•• — Late
September
1
i
\
|
i
i
t
i
i
i
Control
ItAVC Eai"ly
kM4SiC
Mid
Late
June
0.24
0.47
0.36
0.31
Mo
July
0.48
0.67
0.80
0.67
nth
August
0.47
0.55
0.76
0.70
September
0.54
0.62
0.64
0.65
Season
Avg.
0.43"
0.58b
0.64b
0.58b
Figure 61. Changes in macrophyte phosphorus content over time. Treatments are differentiated by stage, which
refers to the stage of macrophyte development at the start of the dose period. The upper table presents
probabilities that stage, day, and interactions of stage and day influenced P content in ANOVA of rank-
transformed data. Darkened values are not significant (p>0.05). The graph is a plot of P content over
the experimental season pooled by treatment. Corresponding lines and arrows mark dose periods. The
lower table lists LS Means (% P of dry weight) represented in the graph, along with statistical information
based on the rank transformed data. Within a column, values sharing a letter are not significantly different
(p>0.05). In columns without letters, values are not significantly different.
81
-------
Macrophyte Nutrient Stock
Stage had no significant effect on nitrogen stock of macrophytes; however, Day was a significant main effect (p=0.0001)
as nitrogen stocks increased seasonally in all treatments (Figure 62). The N stock in the Control increased 10-fold during
the growing season, from 0.4 to 4.3 g N/m2, but did not substantially decrease during senescence. The maximum stocks
in the Mid (8.6 g N/m2) and Late (7.6 g N/m2) treatments in August were nearly 2-fold those in the Early (4.3 g N/m2)
and Control (4.3 g N/m2) treatments; however, by the end of the study, total macrophyte stocks were similar among all
treatments (range 2.5-4.9 g N/m2).
| N Stock
Stage
I,-.,,,-;,--,-., kVjQiyj
;sS*ws. "* xif -,:;», iSSik
Day
0.0001
Stage*Day
:-- ;J5^" 0.6595 '•"
a
OJ
£
+•>
'$• 0.05). The graph is a plot of N stock over the experimental
season pooled by treatment. Corresponding lines and arrows mark dose periods. The lower table lists LS
Means (g N/m2) represented in the graph.
82
-------
Both Stage (p=0.0308) and Day (p=0.0001) significantly affected phosphorus macrophyte stocks (Figure 63). The
phosphorus stock in the Control increased 12-fold during macrophyte growth, from 0.08 to 0.98 g P/m2 (August peak)
and then decreased to 0.80 g P/m2 during senescence. The Mid (0.95 g P/m2) and Late (0.94 g P/m2) treatments were
significantly greater than Controls (0.58 g P/m2) at the end of the study period.
Stage
Day
Stage*Day
P Stock
0.0308
0.0001
2 -j
1.5
Macrophyte Phosphorus Stock, for Study 3
- - A- - - Control
- •* - Early
~ -Mid
June
July
August
September
Month
Control
pt___ Early
kM0.05). The graph is a plot of P stock over the experimental
season pooled by treatment. Corresponding lines and arrows mark dose periods. The lower table lists LS
Means (g P/m2) represented in the graph, along with statistical information based on the rank transformed
data. Within a column, values sharing a letter are not significantly different (p>0.05). In columns without
letters, values are not significantly different.
83
-------
Water Chemistry
Phosphorus
Stage (p=0.0287), Day (p<0.0001), and the Stage*Day interaction (p=0.0001) each were significant factors related toTP
concentrations in water (Figure 64). Total phosphorus concentrations averaged over the entire study were 102, 89, 69,
and 44 ug/LTP jn fne Late, Mid, Early, and Control treatments, respectively; all treatments were significantly greater than
the Control (p<0.05). Total phosphorus increased from early May (14 ug P/L) to a maximum in September (87 ug P/L).
From May through mid-July, TP in the Control ranged from 14 to 30 ug P/L on a weekly basis. Total phosphorus in the
Control more than doubled in late July from 30 to 70 ug P/L. After July 25, values in the Control continued to increase.
Nutrient additions increased TP during and following dose periods; dosing effects were more pronounced in the Mid
and Late treatments compared to the Control and Early treatments.
| TP
Stage
0.0287
Day
0.0001
Stage*Day
0.0001
TP, for Study 3
p-
5/10
5/24
6/7
6/21
7/19
8/2
8/16
8/30
Date
Control
ot,or Early
StagC M,d
Late
5/10
14
13
13
18
5/23
19'
30"
18"
17*
5/30
24ab
31'
20b
22*
6/6
31
36
29
28
6/13
22'
47"
21"
23*
6/20
25'
56b
32'
21'
6/27
18"
40b
49"
20'
7/4
18*
54"
55"
27"
D
7/11
21'
54b
90C
68*
ay
7/18
30*
55"
146C
83"
7/25
70*
112"
187b
163b
8/i
72'
117*"
131b
152b
8/8
59'
104*
117b
153b
8/15
66"
79*
115'b
175b
8/22
77"
123"
183b
241b
8/29
69*
70"
]22bc
178'
9/5
87"
148*
167b
288b
9/12
67*
80"
105*
154b
Season
Avg.
44'
69"
89b
102b
Figure 64. Changes in total phosphorus concentrations over time. Treatments are differentiated by stage, which
refers to the stage of macrophyte development at the start of the dose period. The upper table presents
probabilities that stage, day, and interactions of stage and day influenced TP in ANOVA of rank-trans-
formed data. The graph is a plot of TP over the experimental season pooled by treatment. Correspond-
ing lines and arrows mark dose periods. The lower table lists LS Means (mg/L) represented in the graph,
along with statistical information based on the rank transformed data. Within a column, values sharing a
letter are not significantly different (p>0.05). In columns without letters, values are not significantly differ-
ent.
84
-------
Stage (p=0.0001), Day (p=0.0001) and the Stage*Day interaction (p=0.0001) were significant factors in SRP concentrations
(Figure 65). Soluble reactive phosphorus (SRP) in the Control and Stage treatments ranged between 2 and 16 ug/L until
the initiation of dosing within Stages. Soluble reactive phosphorus dissipated rapidly after addition, indicating that SRP
was rapidly assimilated or lost from the water column. Dissipation rates averaged 20 ug/L/day in the Early treatment,
but were probably underestimates because additions were completely dissipated in 5 days; calculations based on loss of
nominal concentrations applied indicated an approximately 25% P loss/day (Figure 66). Following the final amendment,
SRP in the Early treatment was <20 ug/L and similar to the Control. In the Mid and Late treatments, dissipation rates
were only 17 and 13 ug/L/day (8 and 6% P loss/day), respectively, and SRP accumulated in the water column during
the dosing period within each Stage. At the end of the season, SRP in the Mid and Late treatments decreased to 33
and 65 ug/L, respectively, but remained significantly higher than in the Control and Early treatments (p<0.05).
SRP
Stage
0.0001
Day
0.0001
Stage*Day
0.0001
SRP, for Study 3
1
c«
0 \^ -^—i
5/10 5/24
6/21
T/F7/19 "8/2"^ 8/16 8/30
Date
Cover Control
0%
25%
75%
Cover Control
0%
25%
75%
5/10
6
6
5
6
7/4
2
3
23
6
5/11
6
128
6
7
7/5
4
3
137
121
5/16
13
11
11
11
7/11
3
4
47
17
5/18
13
89
11
13
7/12
7
5
144
69
5/23
5
4
4
4
7/18
9
6
104
27
5/25
4
93
4
5
7/19
16
6
233
156
5/30
4
10
4
4
7/25
9
10
106
36
5/31
5
160
6
5
7/26
7
10
89
94
Day
6/1
6
102
7
7
8/1
10
17
70
57
6/6
6
7
6
5
8/2
8
14
58
149
6/7
6
160
6
5
8/8
12
17
70
82
6/13
5
11
6
4
8/9
10
12
66
187
6/14
7
127
154
6
8/15
8
10
62
86
6/20
4
9
13
3
8/22
9
15
72
88
6/21
4
6
143
3
8/29
8
10
48
76
6/27
4
6
26
5
9/5
14
14
70
137
6/28
6
6
125
8
9/12
10
6
33
65
Season
Avg.
T
32"
58C
46"
Figure 65. Changes in soluble reactive phosphorus concentrations over time. Treatments are differentiated by stage,
which refers to the stage of macrophyte development at the start of the dose period. The table presents
probabilities that stage, day, and interactions of stage and day influenced SRP in ANOVA of rank-trans-
formed data. The graph is a plot of SRP over the experimental season pooled by treatment. Correspond-
ing arrows mark dose periods. The lower table lists LS Means (ug/L) represented in the graph, along with
statistical information based on the rank transformed data. Within a column, values sharing a letter are
not significantly different (p>0.05). In columns without letters, values are not significantly different.
85
-------
SRP
Dissipation
Stage
0.0236
Dose
0.0015
Stage* Dose
0.0030
SRP Dissipation, for Study 3
&•
i
c
35
30
15
5
0
!:={'
t
m
+ Early
• Mid
• Late
Dose
Early
Stage Mid
Late
1
23"
24ab
17"
2
17a
20a
7b
Do
3
17
17
20
ise
4
19"
15ab
6b
5
25a
7b
llb
6
20
21
17
Season
Avg.
20a
17ab
13b
Figure 66. Changes in soluble reactive phosphorus dissipation over dose periods. Treatments are differentiated by
stage, which refers to the stage of macrophyte development at the start of the dose period. The upper
table presents probabilities that stage, dose, and interactions of stage and dose influenced SRP dis-
sipation in ANOVA of rank-transformed data. The graph is a plot of SRP dissipation rates for each dose
pooled by treatment. Bars indicate one standard deviation above and below the mean. The lower table
lists LS Means represented in the graph, along with statistical information based on the rank transformed
data. Within a column, values sharing a letter are not significantly different (p>0.05). In columns without
letters, values are not significantly different.
86
-------
Nitrogen
Stage (p<0.0001), Day (p<0.0001), and the Stage*Day interaction (p<0.0001) were significant factors related to TN in
the water column (Figure 67). Total nitrogen averaged 0.70, 3.63, 3.89, and 3.67 mg N/L in the Control, Early, Mid, and
Late treatments, respectively. Total nitrogen in the Control ranged from 0.38 to 0.98 mg N/L over the course of study.
Total nitrogen exhibited a pattern of increase in dosed treatments that corresponded to the dosing period and peaked
near 10 mg N/L the week following the final dose (Figure 67). During a dose period, TN in an amended treatment was
composed of 90% added nitrate, indicating that persistence of the dose was driving the TN pattern. Total nitrogen val-
ues in the Early, Mid, and Late treatments dropped to around 1 mg N/L within four weeks of the sixth weekly nutrient
addition. Total nitrogen in the Early and Mid treatments was comparable to the Control near the end of the season.
TN
Stage
0.0001
Day
0.0001
Stage* Day
00001
TN, for Study 3
5/10 5/24 6/7 6/21
Control
cfaojl Early
Mage
Mid
Late
5/10
038
033
035
038
5/23
0.43'
5.80b
040'
040'
5/30
0.65"
6.98"
055'
0.60'
6/6
0.63'
768"
060'
1.28'
6/13
053a
8.68b
058a
0.55a
6/20
058'
9.83"
335'
0.55'
6/27
0.50'
660"
565*
0.50'
7/4
0.53'
4.63"
7.75'
0.60'
I
7/11
0.60'
2.88b
8 80'
4 13b
)ay
7/18
065'
1.88"
9.531
608d
7/25
088'
1.73'
I0.78C
8.13s
8/1
093'
1.55b
775'
905'
8/8
085'
123'
445'
888'
8/15
095'
1.08'
330b
1025'
8/22
1.03'
1.15"
233'
640b
8/29
095'
103'
165'
448b
9/5
0.98'
1.43"1
US'"
280b
9/12
0.63'
0.83b
0.78"
1 03b
Season
Avg.
0.70'
3.63"
3.891"
367'
Figure 67. Changes in total nitrogen concentrations over time. Treatments are differentiated by stage, which refers to
the stage of macrophyte development at the start of the dose period. The upper table presents probabili-
ties that stage, day, and interactions of stage and day influenced TN in ANOVA of rank-transformed data.
The graph is a plot of TN over the experimental season pooled by treatment. Corresponding lines and
arrows mark dose periods. The lower table lists LS Means (mg/L) represented in the graph, along with
statistical information based on the rank transformed data. Within a column, values sharing a letter are
not significantly different (p>0.05). In columns without letters, values are not significantly different.
87
-------
Stage (p<0.0001), Day (p<0.0001), and the Stage*Day interaction (p<0.0001) were also significant factors related to
nitrate in the water column (Figure 68). Nitrate in the Control fluctuated near the limit of detection (0.005 mg NO3-N/L)
during the study. In contrast, nitrate concentrations in the Stage treatments closely reflected temporal patterns of dos-
ing. Calculations of nitrate dissipation rates indicated that post-dose sampling accounted for 87, 80, and 71% of nitrate
added in the Early, Mid, and Late treatments, respectively (Figure 69). Nitrate dissipation rate was inversely related to
Stage: Early (0.52 mg NO3-N/L/day; 10% N loss/day); Mid (0.42 rng NO3-N/L/day; 8% N loss/day); and Late (0.34 mg
NO3-N/l_/day; 7% N loss/day). Day was a significant main effect (p<0.0001) in addition to the Stage*Day interaction
(p=0.0041) (Figure 69). Nitrate dissipation continued at similar rates in amended treatments after the termination of
dosing until concentrations fell below the limit of detection (August 22, Early treatment; September 12, Mid and Late
treatments).
NO3-N
Stage
0.0001
Day
0.0001
Stage*Day
0 0001
14 T
NO3-N, for Study 3
5/10
5/24
8/16
8/30
Cover Control
0%
25%
75%
Cover Control
0%
25%
75%
5/10
0207
0.180
0.212
0.202
7/4
0.015
3630
7043
0037
5/11
0156
4.586
0.159
0.151
7/5
0010
3147
11000
3.955
5/16
0073
2.878
0034
0.037
7/11
0017
1 785
7887
3245
5/18
0.168
8.082
0.152
0.187
7/12
0.009
1 539
10.959
5.861
5/23
0.030
6134
0.012
0.015
7/18
0.015
0811
8343
5275
5/25
0018
9451
0.014
0.017
7/19
0.013
0706
11 745
9242
5/30
0009
6.454
0.011
0.007
7/25
0.011
0.323
9252
6617
5/31
0011
10.489
0023
0014
7/26
0.010
0250
8543
9.346
Day
6/1
0.009
10149
0.029
0022
8/1
0007
0150
6.730
8.238
6/6
0.009
6.849
0.023
0737
8/2
0008
0.122
6.364
12654
6/7
0009
12072
0.024
0.096
8/8
0005
0013
3.560
8.004
6/13
0.008
8.685
0.025
0.017
8/9
0005
0017
3.200
11.779
6/14
0.014
12.634
5.030
0.036
8/15
0006
0.006
2.359
9.254
6/20
0010
9155
2739
0.013
8/22
0.005
0.005
1.226
son
6/21
0.010
8078
7694
0.041
8/29
0.005
0.006
0.716
3.120
6/27
0.011
5.657
5.038
0.012
9/5
0.005
0005
0.295
1498
6/28
0.014
5.294
8.807
0.032
9/12
0005
0005
0.005
0.462
Season
Avg.
0.027°
4.098b
3801b
3.097°
Figure 68. Changes in nitrate concentrations over time. Treatments are differentiated by stage, which refers to the
stage of macrophyte development at the start of the dose period. The table presents probabilities that
stage, day, and interactions of stage and day influenced nitrate in ANOVA of rank-transformed data. The
graph is a plot of nitrate over the experimental season pooled by treatment. Corresponding lines and
arrows mark dose periods. The lower table lists LS Means (mg/L) represented in the graph, along with
statistical information based on the rank transformed data. Within a column, values sharing a letter are
not significantly different (p>0.05). In columns without letters, values are not significantly different.
88
-------
NO3-N
Dissipation
Stage
I , «$*TW ,
Dose
0.0001
Stage*Dose
0.0041
__ ,-— v
» x A 7
O a U, /
'•0 >Q
* > n A
a, ,j 0.6
8 ^ n.
a a °-5
f* *7 (\ A -
" M
Oc n T -
Z S 0.3
* Early
-•—Mid
—•—Late
J
ft
1
1
nvj3-i
i
J
> uissipa
I
t
'
I 3
moil, lor
f
T f
1
i
i
.
(
4
Dose
aiuay j
< >
L
i
_
< '
•~ <
.
»
1
6
1
Early
Stage Mid
Late
1
0.34ab
0.38*
0.12"
2
0.39"
0.44a
0.1 Ob
Dose
3
0.60'
0.29b
0.44ab
4
0.66"
0.52"
0.18"
5
0.56ab
0.44"
0.78b
6
0.58
0.42
0.42
Season
Avg.
0.52
0.42
0.34
Figure 69. Changes in nitrate dissipation over dose periods. Treatments are differentiated by stage, which refers
to the stage of macrophyte development at the start of the dose period. The upper table presents prob-
abilities that stage, dose, and interactions of stage and dose influenced nitrate dissipation in ANOVA of
rank-transformed data. Darkened values are not significant (p>0.05). The graph is a plot of nitrate dis-
sipation rates for each dose pooled by treatment. Bars indicate one standard deviation above and below
the mean. The lower table lists LS Means represented in the graph, along with statistical information
based on the rank transformed data. Within a column, values sharing a letter are not significantly different
(p>0.05). In columns without letters, values are not significantly different.
89
-------
Stage (p=0.0046), Day (p<0.001) and the Stage*Day interaction (p=0.0001) had significant effects on ammonia (Fig-
ure 70). Ammonia in the Control was near the limit of detection (0.005 mg NH3-N/L) during most of the season. During
the Early dose period, however, ammonia in the Control, Mid, and Late treatments rose to 0.045 mg NH3-N/L, possibly
due to disturbance of the bare sediment surface during stirring. A similar response was not seen in the Control dur-
ing other periods because disturbance of the stirring effect may have been dampened by macrophytes. In amended
treatments, ammonia peaked on the day of additions. Ammonia peaks during the dose period decreased in the order
of dosing initiation (Early, 0.07 to 0.15 mg NH3-N/L; Mid, 0.03 to 0.11 mg NH3-N/L; and Late, 0.01 to 0.04 mg NH3-N/L);
indicating that ammonia responses were less as the growing season progressed. Seasonal maximums in the Early
(0.152 mg NH3-N/L) and Mid (0.105 mg NH3-N/L) treatments occurred during their respective dose periods; however,
maximum ammonia concentration in the Late (0.085 mg NH3-N/L) treatment occurred latently in September. Following
the dose period, ammonia concentrations in the Early and Mid treatments gradually decreased over time and were
comparable to the Control after July 12 and August 29, respectively. Ammonia in the Late treatment was similar to the
Control with the exception of a sudden observed increase in early September. The proportion of ammonia compared
to nitrate was minimal (<1%) in all treatment/day combinations.
NH3-N
Stage
0.0046
Day
0.0001
Stage*Day
0.0001
Cover Control
0%
25%
75%
Cover Control
0%
25%
75%
5/10
0.008
0-011
0.012
0.017
7/4
OOOSa
0022b
0014b
OOOSa
5/11
0.006a
0074b
O.OOSa
0015a
7/5
OOOSa
0.020b
0.045C
OOlSbc
5/16
0016
0.021
0015
0017
7/1]
OOOSa
0026bc
0048c
0007ab
5/18
0017a
0083b
OOlSa
0019a
7/12
OOOSa
0020bc
0074c
0013ab
5/23
0017
0020
0016
0024
7/18
OOOSa
0014ab
0.063b
0017ab
5/25
0028a
0144b
0020a
0046ab
7/19
0007a
0014ac
OlOSb
0038bc
5/30
0045
0105
0048
0035
7/25
0006a
0012a
0053b
0012ab
5/31
0033a
0.1 52b
0043a
0025a
7/26
O.OOSa
001 Sac
0090b
0026bc
Day
6/1
0044ab
OlOlb
0043ab
0028a
8/1
OOOSa
OOOSa
OlOOb
0013b
6/6
0 030ab
0 038b
0045a
0012a
8/2
O.OOSa
0006a
0091b
0 029b
6/7
0.021a
0094b
0045ac
OOOSc
8/8
0012a
OOllac
OOSlbc
0039b
6/13
0006a
0033b
0 038ab
OOOSa
8/9
OOOSa
0009a
0034b
0035b
6/14
OOOSa
0069b
OOSOb
OOOSa
8/15
0006a
0007a
0026b
0028b
6/20
OOOSa
0028b
0031ab
OOOSa
8/22
OOOSa
0 009ac
0012b
0022bc
6/21
O.OOSa
0027b
0082b
OOOSa
8/29
0006ab
0007a
0013bc
0.046C
6/27
OOOSa
OOSSb
0.025C
0009ac
9/5
OOlOab
0.009a
0025bc
0 OSSc
6/28
OOOSa
0 027ab
0032b
OOOSa
9/12
0 009ab
OOOSa
OOlOb
0012ab
Season
Avg.
0012"
0038*
0043b
0021*
Figure 70. Changes in ammonia concentrations over time. Treatments are differentiated by stage, which refers to
the stage of macrophyte development at the start of the dose period. The table presents probabilities that
stage, day, and interactions of stage and day influenced ammonia in ANOVA of rank-transformed data.
The graph is a plot of ammonia over the experimental season pooled by treatment. Corresponding lines
and arrows on X axis mark dose periods. The lower table lists LS Means (mg/L) represented in the graph,
along with statistical information based on the rank transformed data. Within a column, values sharing a
letter are not significantly different (p>0.05). In columns without letters, values are not significantly differ-
ent.
90
-------
NitrogeirPhosphorus Ratio
The TN:TP ratio was significantly related to Stage (p=0.0001), Day (p<0.0001), and the Stage*Day interaction (p<0.0001)
(Figure 71). Nutrients were added during each Stage at a targeted N:P ratio of 25:1; in the absence of internal load-
ing, this ratio was expected to be phosphorus limited. Similarly, the N:P ratio in the Control fluctuated between 20 and
30 from May through July 18, indicating the potential for P limitation; thereafter, the TN:TP decreased to between 10
and 20 during the remainder of the season. Although the calculated TN:TP of the amendments was 25:1, the ratio
in the Early treatment doubled to more than 50:1 after the first dosing because the added P was rapidly lost from the
water column in proportion to nitrate. Following the third amendment, TN:TP ratios in the Early treatment exceeded
200 (10-fold greater than the Control) and remained around that level through June. During July, TN:TP ratio in the Early
treatment dropped due to the decline in TN, and thereafter, levels were similar to the Control. In the Mid treatment,
the TN:TP ratio increased from 29 to 122 following the first amendment, and peaked at 164 following the third dose.
The ratio in the Mid treatment decreased through July and was comparable to the Control (<20) by mid-August. The
TN:TP ratio peaked at 89:1 following the second dose and gradually decreased to around 9:1 by September. Seasonal
averages of TN:TP ratios were 21, 91, 52, and 38 in the Control, Early, Mid, and Late treatments, respectively, which
revealed the overall effect of Stage on the TN:TP ratio of water.
TN:TP
250 -|
H 15°
a
50
1
n
Stage
00002
Day
0.0001
Stage*Day
0.0001
TN:TP, for Study 3
X*"V x'^N
r
! V
/ ~ 7-^
/
-'*'^ J
5/10 5/17 5/24 5/31 6/7 6/14 6/21 6/28
^
^
>^Sf_-!_ .^
A ^">-- ^*~~~*-
I 1 * *
— Mid — - •— - Late
V-..
~^B *£i*J=6^=t
1
1^-..^. .«...-. — _.* '
7/5 7/12 7/19 7/26 8/2 8/9 8/16 8/23 8/30 9/6
Date
Control
ofaop Ear'y
stage
Mid
Late
5/10
29
27
29
24
5/18
23'
58"
22"
25'
5/23
22"
201b
23'
23'
5/30
27'
232b
29"
28*
6/6
21'
220b
23'
46"
6/13
25'
191b
29'
25'
6/20
26"
231"
I22b
27'
6/27
29"
203b
127"
28'
B
7/4
30'
112"
164"
27'
ay
7/11
301
77-"
110b
72"
7/18
23'
40'
74"
89b
7/25
15'
26'
63"
56"
8/1
14"
17"
60"
60"
8/8
15'
15"
37"
59"
8/15
15"
17"
28"
64'
8/22
14'
14'
14'
25"
8/29
14'
16"
13'
23"
9/5
12
12
9
10
9/12
10
14
10
9
Season
Avg.
21'
91"
52"
38"
Figure 71. Changes in the ratio of total nitrogen to total phosphorus over time. Treatments are differentiated by
stage, which refers to the stage of macrophyte development at the start of the dose period. The upper
table presents probabilities that stage, day, and interactions of stage and day influenced 77V:TP in ANOVA
of rank-transformed data. The graph is a plot of TN:TP over the experimental season pooled by treat-
ment. Corresponding lines and arrows mark dose periods. The lower table lists LS Means represented in
the graph, along with statistical information based on the rank transformed data. Within a column, values
sharing a letter are not significantly different (p>0.05). In columns without letters, values are not signifi-
cantly different.
91
-------
PH
Both Stage (p=0.0160) and Day (p<0.0001) had significant effects on pH of the water column (Figure72). Seasonal
averages of pH in the Early treatment (pH=9.0) were significantly greater (p<0.05) than in the Control (pH=8.7) treat-
ment; however, the seasonal average pH in the Mid and Late treatments (pH=8.8) were similar to the Control. Initial
pH values in all treatments ranged from 8.3-8.4 and continually increased over time. Dosing in the Early treatment
led to increases in pH in the Early treatment of approximately 0.5 units. In early July, the pH levels in all treatments
converged to approximately 10 and remained similar among treatments for the remainder of the study. Thus, nutrient
amendments influenced pH only when added prior to macrophyte growth (Early treatment). The observed increase in
pH reflects the stimulation of periphyton communities (Figure 81) as discussed below.
Stage
Day
Stage*Day
0.0160
0.0001
11.0!
10.5
K
D.
10.0
9.5
7.5
7.0
pH, for Study 3
1
*
I
5/10
5/24
:•= — **•
6/7
6/21
7/5
7/19
8/2
8/16
8/30
Date
Control
Ctann E^
stage
Mid
Late
5/10
8.4
8.4
8.4
8.3
5/16
83
85
8.2
8.2
5/23
8.3
8.7
83
8.3
5/30
8.1
88
8.2
82
6/6
8.3
9.3
8.4
85
6/13
8.6
96
86
88
6/20
89
9.9
8.9
92
6/27
92
99
9.4
9.6
]
7/4
96
10.0
100
9.8
)ay
7/11
100
100
102
10.1
7/18
99
99
101
10.1
7/25
9.8
98
10.0
100
8/1
97
97
98
98
8/8
94
91
98
98
8/15
9.7
9.9
10.0
99
8/22
97
95
98
97
8 '29
9.7
9.4
9.7
97
9/5
8.8
85
89
8.7
9/12
9.2
9.2
9.5
9.5
Season
Avg.
87'
9.0b
88'
88'
Figure 72. Changes in pH over time. Treatments are differentiated by stage, which refers to the stage of macrophyte
development at the start of the dose period. The upper table presents probabilities that stage, day, and
interactions of stage and day influenced pH in ANOVA of rank-transformed data. Darkened values are not
significant (p>0.05). The graph is a plot of pH over the experimental season pooled by treatment (data
were averaged by H-ion concentration, then converted to pH: calculated pH= - log (H-ion)). Correspond-
ing lines and arrows mark dose periods. The lower table lists LS Means represented in the graph, along
with statistical information based on the rank transformed data. Within a column, values sharing a letter
are not significantly different (p>0.05). In columns without letters, values are not significantly different.
92
-------
Alkalinity and Hardness
Stage (p=0.0111), Day (p<0.0001), and the Stage*Day interaction (p<0.0001) had significant effects on alkalinity (Fig-
ure 73). Seasonal averages of alkalinity indicated that the Mid (133 mg CaCO3/L) and Late (130 mg CaCO3/L) treat-
ments were significantly greater (p=0.05) than the Control and Early treatments (117-118 mg CaCO3/L). Alkalinity
in the Control decreased from a maximum of 233 mg CaCO3/L soon after corral filling on May 10 to 80 mg CaCO3/L
in early July prior to stabilization in a range between 60 and 110 mg CaCO3/L through the remainder of the season.
Early dosing substantially decreased alkalinities by 25% compared to Control levels in May and early June due to loss
of carbonate to primary productivity. However, alkalinity in the Early treatment was similar to the Control by late June.
Alkalinities in the Mid and Later treatments increased above Control levels as dosing was initiated at each respective
Stage. Thus, Early treatment decreased alkalinities soon after dosing; whereas, dosing had the opposite effect in the
Mid and Late treatments which indicates a differential system response in periphyton-dominated communities (Early
treatment) compared to macrophyte-dominated systems (Mid and Late treatments).
Alkalinity
Stage
0.0111
Day
0.0001
Stage*Day
0.0001
250
Alkalinity, for Study 3
50
5/10
5/24
6/7
6/21
-d E. -^
7/? 7/19 8/2
8/16
8/30
Date
Control
•Ctucro Early
Mage
Mid
Late
5/10
233
237
228
236
5/16
201
169
203
200
5/23
193
135
194
191
5/30
179
139
173
170
6/6
171'
116b
154'
144'b
6/13
136'
107b
]25,b
110'b
6/20
111"
114'
114"
95"
6/27
90'b
103'
107'
85"
7/4
80'
101b
95'"
79*
Day
7/11
77'
101b
101b
se"
7/18
105
100
106
100
7/25
93'
115"
130b
123b
8/1
78'
102b
118b
112b
8/8
69'
94b
107b
102b
8/15
84"
97'"
114^
114C
8/22
80'
104b
119b
126b
8/29
78'
99b
lllb
123"
9/5
90*
110b
118b
138b
9/12
90"
103"
no"
138e
Season
Avg.
117'
118"
133"
130b
Figure 73. Changes in alkalinity over time. Treatments are differentiated by stage, which refers to the stage of mac-
rophyte development at the start of the dose period. The upper table presents probabilities that stage,
day, and interactions of stage and day influenced alkalinity in ANOVA of rank-transformed data. The
graph is a plot of alkalinity over the experimental season pooled by treatment. Corresponding lines and
arrows mark dose periods. The lower table lists LS Means (mg CaCO/L) represented in the graph, along
with statistical information based on the rank transformed data. Within a column, values sharing a letter
are not significantly different (p>0.05). In columns without letters, values are not significantly different
93
-------
Stage (p=0.0048), Day (p=0.0001), and the Stage*Day interaction (p=0.0001) were all significant factors affecting hard-
ness (Figure 74). Hardnesses, averaged over the entire study, were 140,105,135, and 139 mg CaCO3/L in the Control,
Early, Mid, and Late treatments, respectively; the Early treatment was significantly higher (p<0.05) than the Control on a
total-study basis. Hardness averaged 273 mg CaCO3/L at the beginning of the study and reflected conditions within the
well water source. Hardness values declined from May to September in a linear trend over time. Hardness decreased
relative to Control values due to the Early treatment dosing which resulted in an approximate 30% decrease in hard-
ness values. However, neither the Mid nor Late treatments influenced hardness values. The decrease in hardness
in the Early treatment reflected the precipitation losses of major divalent cations as carbonate was consumed due to
increased productivity of periphyton in the absence of macrophytes. Hardness continued to decline in all treatments
as macrophytes developed. Macrophyte biomass, and hence overall system primary productivity, was relatively similar
in the Mid, Late, and Control treatments; hence, hardness levels exhibited similar trends over time as carbonate was
consumed and cations were precipitated.
Hardness
Stage
0.0048
Day
0.0001
Stage*Day
0.0001
. 5
Sfl **>
at Q
*2 *
w -
^
300-1
i
OCA _
o -
5/
N>T^^*H^|
X
10 5/24
Hardness, for Study 3
-A---Control --
- - Early
-Mid --•-- Late
6/7
6/21
— •••.
7/5 7/19
Date
8/2 8/16 8/30
Control
Cf-,™,, Ear|y
Mid
Late
5/10
273
273
272
271
5/16
243
208
243
239
5/23
234'
161*
233'
2281"
5/30
231'
149*
218'
213*
6/6
220"
131*
198"
187'*
6/13
176"
115*
160"*
149*
6/20
148"
III*
142"
131""
6/27
120'
94"
125'
116'
7/4
108'
78"
106'
106-
Day
7/11
101'
74*
104"
107"
7/18
loi"
68*
101"
105'
7/25
100"
69*
98"
109'
8/1
95"
64*
92"
ior
8/8
79ab
62'
78"
86b
8/15
83"
64'
79'
115"
8/22
87""
70'
71'
95*
8/29
78
63
72
86
9/5
91"b
73'
82"
97b
9/12
88
70
81
92
Season
Avg.
140'
105*
135'
139"
Figure 74. Changes in hardness over time. Treatments are differentiated by stage, which refers to the stage of mac-
rophyte development at the start of the dose period. The upper table presents probabilities that stage,
day, and interactions of stage and day influenced hardness in ANOVA of rank-transformed data. The
graph is a plot of hardness over the experimental season pooled by treatment. Corresponding lines and
arrows mark dose periods. The lower table lists LS Means (mg CaCO/L) represented in the graph, along
with statistical information based on the rank transformed data. Within a column, values sharing a letter
are not significantly different (p>0.05). In columns without letters, values are not significantly different.
94
-------
Conductivity
Stage (p=0.0046), Day (p=0.0001), and the Stage*Day interaction (p=0.0001) were highly significant factors controlling
changes in conductivity (Figure 75). All three dosing stages (Early, 429 uS/cm; Mid, 451 uS/cm; and Late, 430 uS/cm)
contained significantly higher (p<0.05) conductivity values compared to the Control (382 uS/cm). Conductivity in the
Control decreased 60% during the season, from an initial maximum of 625 uS/cm, to 268 uS/cm in September. The
primary decrease in conductivity (50%) occurred by early July, and values fluctuated around 260 uS/cm the final six
weeks of the experiment. Although conductivity was significantly influenced by amendments (p<0.05), the dosed treat-
ments were not significantly different from each other.
Conductivity
Stage
0.0046
Day
0.0001
Stage*Day
0.0001
u
X
^i
'>
•o
V
3
•O
B
O
O
f.f\f\ i
•3 Art
0
5/
Conductivity, for Study 3
,, . j
"^>>*^^^4 *^ ^\ !
^" *^i'- g m^^*^-.+.~^+-~~~~?'s'*^\ --•• i
•A-----A-----A- .-.^.---* -i
i
;
10 5/24 6/7 6/21 7/5 7/19 8/2 8/16 8/30
Date
Control
Staero Early
stage.
Mid
Late
5/10
625
622
609
620
5/16
573
538
572
568
5/23
520
466
516
510
5/30
556
514
538
530
6/6
548
523
505
493
6/13
4641"
506'
432"
404e
6/20
402"
526"
429'
364'
6/27
346'
476b
426'
330*
7/4
329"
446b
432"
318'
Day
7/11
357'
417'
511e
367'b
7/18
305'
380"
444C
389b
7/25
305'
385"
478C
422b
8/1
383*
471"
582b
557b
8/8
256'
314b
375e
404[
8/15
263'
330b
376b
436C
8/22
259"
313"
353"
395'
8/29
241"
29 l'b
325be
348'
9/5
266'
311*
327b
364b
9/12
268'
315-
337b
354"
Season
Avg.
382'
429"
451"
430b
Figure 75. Changes in conductivity over time. Treatments are differentiated by stage, which refers to the stage of
macrophyte development at the start of the dose period. The upper table presents probabilities that
stage, day, and interactions of stage and day influenced conductivity in ANOVA of rank-transformed data.
The graph is a plot of conductivity over the experimental season pooled by treatment. Corresponding
lines and arrows mark dose periods. The lower table lists LS Means (mS/cm) represented in the graph,
along with statistical information based on the rank transformed data. Within a column, values sharing a
letter are not significantly different (p>0.05). In columns without letters, values are not significantly differ-
ent.
95
-------
Turbidity
Turbidity changes were associated with Day (p<0.0001) but not with Stage (Figure 76). Overall, study averages of the
Control (4.0 NTU's), Early (5.4 NTU's), Mid (3.2 NTU'.s), and Late (3.9 NTU's) treatments were quite similar with a
range of less than 2.2 NTU's. Turbidity in the Control ranged from 1 to 4 NTU's through mid-July, and then increased
to a maximum of 8.8 NTU's on September 5 (Figure 76). Although the Stage of dosing had no significant main effect
on turbidity, average values were generally highest in the Early Stage due to the significant effect of Early dosing on
phytoplankton biomass discussed below. Such effects were not observed in the Mid and Late treatments due to the
dominance of macrophytes on system productivity and nutrient dynamics.
Stage
Day
Stage*Day
Turbidity
0.0001
Turbidity, for Study 3
5/10
5/24
7/5 7/19
Date
8/16
8/30
Control
0.05). The graph is a plot of turbidity over the experimental season pooled
by treatment. Corresponding lines and arrows mark dose periods. The lower table lists LS Means (NTU)
represented in the graph.
96
-------
Phytoplankton
Stage (p=0.0279), Day (p=0.0001), and the Stage*Day interaction (p=0.0461) were all significant effects controlling
changes in phytoplankton biomass as measured as Chi a (Figure 77). Total study averages indicated that Chi a con-
centrations in the Early treatment (25 ug/L) were significantly greater (p<0.05) than the Control (13 ug/L), Mid (9 ug/L),
and Late (19 ug/L) treatments. Chi a averaged 5 ug/L across all treatments at study initiation in early May. Chi a ranged
from 1-5 ug/L from May through mid-July; increased to 31 ug/L by late July; and then varied between 14 and 46 ug/L
through September. Early nutrient enrichment resulted in significant increases (p<0.05) in Chi a from May through
mid-July; during this period macrophyte biomass was low, and both phytoplankton and periphyton increased due to
the nutrient subsidy. Phytoplankton in the Mid and Late treatments frequently departed from Control values during the
study but varied and were not significantly different from the Controls.
Phytoplankton Chi
80-i
" SO
1" 40
D.
£ ™
3
U on
10
I
Stage
0.0279
Day
0.0001
Stage*Day
0.0461
Phytoplankton Chlorophyll, for Study 3
- -A- -Control — «— Early
— • — Mid --•-- Late
*
\
/
^
- --
\
it
* / \
1 \ ir \
1 \ f' *'\.
' ...A-> .• ^r*"
/ s^\ / /" V/^^ V*
n
/ \ •
/ \ _^!_
/. - A, ^ / *
/^'•x \ ^ />
-%-} — ^^/
/ *•-*
£-*^* T — *- ^t "• - r^ - * *V-W^-*»-*''x "Sl1 — " — *
5/10 5/24 6/7 6/21
'^-.. ..J: '.. >
7/5 7/19 8/2
Date
8/16 8/30
Control
Ctanc Early
stage
& Mid
Late
5/10
5
6
5
5
5/16
2"
10b
2'
23'
5/23
2'
6b
2'
1"
5/30
3
3
2
2
6/6
1'
7b
1'
3"
6/13
2'
23"
3'
4"
6/20
3"
33"
15"
15"
6/27
1
2
4
1
E
7/4
2'
20b
2'
3'
'ay
7/11
3'
50"
5'
20*
7/18
4*
15'
4"
22'
7/25
32
55
13
41
8/1
29-
24*
6"
28'
8/8
15
28
7
15
8/15
25'
29*
7"
21*
8/22
38
27
24
40
8/29
46
75
24
29
9/5
17
27
21
24
9/12
14
45
34
64
Season
Avg.
13'
25"
9'
19""
Figure 77. Changes in phytoplankton chlorophyll concentrations over time. This combination of tables and a graph
present the data and statistical information pertaining to phytoplankton chlorophyll (mg/L). Treatments are
differentiated by stage, which refers to the stage of macrophyte development at the start of the dose pe-
riod. The upper table presents probabilities that stage, day, and interactions of stage and day influenced
Chi in ANOVA of rank-transformed data. The graph is a plot of Chi over the experimental season pooled
by treatment. Corresponding lines and arrows mark dose periods. The lower table lists LS Means (mg/L)
represented in the graph, along with statistical information based on the rank transformed data. Within a
column, values sharing a letter are not significantly different (p>0.05). In columns without letters, values
are not significantly different.
97
-------
Particulate organic carbon (POC) was not influenced by Stage, but did significantly increase over time due to the main
effect of Day (p<0.0001) (Figure 78). Levels of POC averaged 5.14, 5.71, 3.91, and 4.29 mg/L in the Control, Early,
Mid, and Later treatments, respectively. Initial levels of POC ranged from 2.61 to 2.84 mg/L among treatments and
remained less than 5 mg/L through July 4. Thereafter, POC levels increased in all treatments, typical of the late season
senescent period dominated by internal nutrient release.
POC
Stage
ft. -i-*.'*^»-:-'-».5^*s: ^i2?r ;'
:• ^ ,,,„•,'" • • ""*?t*-',; «*;.:. f ;;;
Day
0.0001
Stage*Day
' ' '" '•'•• m. < '' " '-; ,> ^0^ ".'$>
• ; < ^fev^: &835&^7" • ' -^^--i
u
o
OH
Particulate Organic Carbon, for Study 3
5/10
5/24
7/5 7/19
Date
8/16
8/30
Control
ct.,r,fl Early
Mage
Mid
Late
5/10
2.61
2.80
2.84
2.72
5/16
2.05
2.46
2.55
417
5/23
2.71
2.44
2.49
1 97
5/30
2.10
2.61
226
243
6/6
225
3.56
252
340
6/13
204
375
206
2.01
6/20
1.58
2.56
1 68
246
6/27
140
1.14
1.75
1.73
7/4
1 84
4.41
2.05
231
Da>
7/11
3.32
9.41
2.57
453
r
7/18
352
459
342
427
7/25
584
9.40
350
5.69
8/1
4.50
2.43
3.09
416
8/8
5.28
7.75
365
491
8/15
7.26
7.49
4.55
481
8/22
18.38
9.54
654
7.51
8/29
13.45
15.21
5.07
8.59
9/5
10.02
6.87
11.10
7.57
9/12
7.59
10.01
10.70
626
Season
Avg.
5.14
5.71
3.91
4.29
Figure 78. Changes in paniculate organic carbon concentrations over time. Treatments are differentiated by stage,
which refers to the stage of macrophyte development at the start of the dose period. The upper table
presents probabilities that stage, day, and interactions of stage and day influenced POC in ANOVA of
rank-transformed data. Darkened values are not significant (p>0.05). The graph is a plot of POC over the
experimental season pooled by treatment. Corresponding lines and arrows mark dose periods. The lower
table lists LS Means (mg C/L) represented in the graph.
98
-------
POC:Chl ratios varied significantly in relation to Stage (p=0.0028), Day (p=0.0001), and the Stage*Day interaction
(p=0.0323) (Figure 79). Total season averages of the POC:Chl ratio were significantly lower in the Early (530) and
Late (813) treatments compared to the Control (1044); POC:Chl ratio was lowest in the Mid treatment (992) which was
significantly greater (p< 0.05) than in the Early treatment. On May 10, the POC:Chl ratios ranged from 562-670 among
treatments and increased in all treatments (upper limit 2500) as productivity began to increase. The POC levels signifi-
cantly decreased (p<0.05) in the Early treatment as dosing was initiated due to the stimulatory effect on phytoplankton.
Similar decreases in POC:Chl ratios were observed in the Mid and Late treatments as dosing was initiated in relation
to macrophyte stage; however, the dominance of macrophytes dampened the level of response to dosing.
POC:Chl
Stage
0.0028
Day
0.0001
Stage*Day
0.0323
<5
0
o
BH
JOUU -
0 -
51
POCrChl, for Study 3
^ , ^ T---1..
/\ — • — Mid --»--Late
/.--/W<*>>... , ^
:;/ K. ;->i -.^•^r--.1\ ^
r / ¥ v/f vM^.^ Vj^S
-v- V" -^' *-%Cj^'-' '*-*^*"'*:"*
10 5/24 6/7 6/21 7/5 7/19 8/2 8/16 8/30
Date
Control
s:t-,0n Ear'y
Mid
Late
5/10
670
562
668
589
5/16
2413'
301b
1009'
1599'
5/23
1854'
496b
1879'
1924*
5/30
1135
1418
1610
1128
6/6
2008
875
1833
1724
6/13
1730'
1404b
1240"
1026*
6/20
1400'
222b
224b
562"
6/27
1427'
479b
796*
1549'
D
7/4
1419"
242"
1264'
1324'
ay
7/11
1400'
695b
1431'b
492b
7/18
1010"
434b
1683'
748b
7/25
285
259
310
187
8/1
235
181
482
144
8/8
353
632
742
580
8/15
281
334
1238
285
8/22
434
380
569
398
8/29
369
342
488
662
9/5
892
419
798
355
9/12
523'
404*
593'
175b
Season
Avg.
1044'
530b
992«
813'
Figure 79. Changes in the ratio of particulate organic carbon to phytoplankton chlorophyll over time. Treatments are
differentiated by stage, which refers to the stage of macrophyte development at the start of the dose pe-
riod. The upper table presents probabilities that stage, day, and interactions of stage and day influenced
POC:Chl in ANOVA of rank-transformed data. The graph is a plot of POC:Chl over the experimental
season pooled by treatment. Corresponding lines and arrows mark dose periods. The lower table lists
LS Means represented in the graph, along with statistical information based on the rank transformed data.
Within a column, values sharing a letter are not significantly different (p>0.05). In columns without letters,
values are not significantly different.
99
-------
Stage of treatment had no significant effects (p>0.05) on the distribution of the five major classes of phytoplankton;
however, Day was a significant (p=0.0001) factor. Low numbers of phytoplankton were observed in May sampling
which occurred soon after corral-filling (Figure 80). By June, the phytoplankton community was dominated by the chlo-
rophytes with the cyanophytes being low in number. Total number of algae significantly increased in July to a total of
9.9*106 cells/L; approximately 71 % of the phytoplankton were Chlorophytes, and 25% were Cyanophytes. Phytoplankton
numbers decreased in August but remained similar in distributions among major groups. Peak seasonal numbers of
phytoplankton were observed in September, when total numbers of phytoplankton reached 25.4*106 cells/L, and the
community shifted to a community slightly dominated by Cyanophytes (50% of the community) compared to Chloro-
phytes (44% of the community); numbers of Euglenophytes and Baccillariophytes also increased but comprised less
than 6% of the total community.
A total of 137 species of algae/cyanobacteria were observed in Study 2; the list of observed phytoplankton species is
presented in Table 7. We observed 62 species of Chlorophytes; 37 species of Bacillarophytes; 15 species of Cyanophytes;
13 species of Euglenophytes; 5 species of Chysophytes; 4 species of Cryptophytes; and 1 species of Pyrrophycota.
Early in the season, the Chlorophytes were dominated by Gleocystis, Scenedesmus, and Oedegonium sp.; whereas
the Cyanophytes were dominated by the filamentous Oscillatoria sp. By the end of Study 2, the algal community was
dominated by the filamentous cyanophyte Oscillatoria sp. and the Chlorophytes Pleodorina, Oocystis, Characium, and
Oedeogonium sp.; lesser amounts of the Euglenophytes (Trachlemonas and Euglena sp.) and the Bacillariophytes
(Navicula, Nitzschia, Fragilaria, and Gomphonema sp.) were observed.
Total
Chlorophyta
Euglenophyta
Bacillariophyta
Cryptophyta
Cyanophyta
5/11/00
03870
0.2541
0.0009
0.0733
00468
0.0121
6/7/00
08640
0.7711
0.0008
00874
0.0000
00043
7/5/00
9.9290
70589
0.0049
0.2259
0.0938
2.5453
8/2/00
6.0210
3.9591
0.0005
0.0854
0.0050
1.9706
9/5/00
25.4040
11.3575
0.7509
0.5586
0.0553
12.6819
J 04
c«
'a
•o I?20"
51,6
?i
8 i? -
S
I „
A
H Total
Phytoplankton Division Abundance, for Study 3
§
Ij IL
. — — rc^H OH H vH tS
5/11/00 6/7/00 7/5/00 8/2/00
Sample Date
• Chlorophyta D Euglenophyta II Bacillariophyta S Cryptophyta
SI
1
li I
^ m
tJ
9/5/00
H Cyanophyta
Figure 80. Changes in phytoplankton abundance of the dominant divisions over time. Columns represent averages
of all corrals in each month to show general successional trends over the season.
100
-------
Periphyton
Both 1-week and 2-week periphyton accrual rates responded to Stage (p=0.0173), Day (p=0.0020), and Stage*Day
interaction (p=0.0042) (Figure 81). Periphyton accrual averaged 0.08 ug Chl/cm2/wk for both 1-week and 2-week ac-
crual intervals in the Control when averaged across all dates. Individual weekly accrual rates in the Control ranged
from a minimum of 0.02 ug Chl/cm2/wk in July to a maximum of 0.17 ug Chl/cm2/wk in September. One- and two-week
accrual rates in the Control were similar. Periphyton accrual was substantially enhanced by nutrient enrichment, but
the response varied with the timing of the dose period. Dosing in the Early macrophyte stage resulted in a 19-fold
increase (1.23 ug Chl/cm2/wk) in periphyton accrual rates. However, during the post-dosing period from July through
September, rates in Early treatments averaged <0.1 ug Chl/cm2/wk and were comparable to the Control. Mid treatment
rates peaked at 1.65 ug Chl/cm2/wk in 2-week exposures in June; 2-week rates were twice those based on 1-week
exposures. These differences may be because 2-week exposures received two amendments, whereas 1-week expo-
sures had received only one amendment. Rates in Mid treatments averaged <0.2 ug Chl/cm2/wk from July through
September. Accrual rates in Late treatments were minimally enhanced during the dose period in July, averaging 0.13
and 0.26 ug Chl/cm2/wk in 1- and 2-week exposures, respectively, due to the substantial competition by increased
macrophyte stands late in the study.
Periphyton 1-week
Periphyton 2-week
Stage
0.0173
0.0119
Day
0.0020
0.0001
Stage*Day
0.0042
0.0002
Periphyton Accrual in 1- and 2-week Exposures, for Study 3
• Control (1)
B Early (1)
HMid(l)
QLate(l)
H Control (2)
U Early (2)
B Mid (2)
B Late (2)
September
Control
StaoP Early
Mid
Late
Month
M
l-wk
0.07'
123"
006'
004'
ay
2-wk
0.051
123*
004'
006'
Ju
l-wk
O15'b
0.75'
080lb
012'
ne
2-wk
013lb
054'
1 65'
008"
Ju
l-wk
0.02'
006b
003*
013b
iy
2-wk
0.03"
0.05''
0161"
026'
Au|
1-
wk
004
0.07
003
007
;ust
2-
wk
004
006
0 14
010
Septei
l-wk
013"
0.06'
003*
022"
nber
2-
wk
017
012
0.11
0.50
Season
l-wk
0.08'
043"
018'
0121*
2-wk
008'
0.40"
042"
018"
Figure 81. Changes in accrual rates of periphyton chlorophyll in 1- and 2-week exposures over time. The 1- and 2-
week datasets were analyzed separately. Treatments are differentiated by stage, which refers to the stage
of macrophyte development at the start of the dose period. The upper table presents probabilities that
stage, day, and interactions of stage and day influenced accrual rates in ANOVA of rank-transformed data.
The graph is a plot of accrual rates over the experimental season pooled by treatment. The lower table
lists LS Means (mg Chl/cnf/wk) represented in the graph, along with statistical information based on the
rank transformed data. Within a column, values sharing a letter are not significantly different (p>0.05). In
columns without letters, values are not significantly different.
101
-------
Zooplankton
Dosing in relation to macrophyte development (i.e., Stage) had no statistically significant effect on the zooplankton com-
munity. However, Day (p<0.0001) and the Stage*Day interaction (p=0.0108) had significant effects (Figure 82). Total
numbers of zooplankton, in addition to major group classifications (cladocerans, copepods, and rotifers) increased in
numbers over time in all treatments. A total of 36 genera, including 16 rotifers, 13 cladocerans, and 7 copepods, were
identified in this study. A list of zooplankton taxa observed in Study 3 are presented in Table 8.
Lowest numbers of zooplankton were observed in all treatments in May soon after corral filling (Figure 82). Observed
significant differences among treatments at study initiation, and prior to treatments, indicate the relative variability of the
zooplankton data. For example, the Late treatment contained significantly higher numbers of zooplankton compared to
the Early and Mid treatments even though dosing had not begun. Similarly, there were no significant differences in total
zooplankton in June in any treatment. Dosing effects became apparent, however, in July as total numbers of zooplankton
significantly tripled in the Mid (2.3*105 zooplankton/m2) and Late (1.6*105 zooplankton/m2) treatments compared to the
Control (0.6*105 zooplankton/m2). Total zooplankton numbers declined in the Early treatment in August but remained
significantly higher than the Control in the Mid and Late treatments. In September, total numbers of zooplankton were
elevated in all treatments compared to the Control; however, only the Late treatment was significantly greater than the
Control.
Total Zooplankton
Cladocerans
Copepods
Rotifers
Stage
.^-^m&M®-.-^
,,„> jyga%s.«. .,v}«*
, .„ ,'.0.05). The graph is a plot of abundance over the experimental season by treatment
(Control (C); Early (E); Mid (M); Late (L)). Abundances in the respective categories (Cladocerans-diago-
nals; Copepods-solids; Rotifers-stripes) are stacked to indicate totals. The lower table lists LS Means (# of
indivVn?) for abundance of the total zooplankton community, along with statistical information based on the
rank transformed data. Within a column, values sharing a letter are not significantly different (p>0.05). In
columns without letters, values are not significantly different.
102
-------
Rotifers accounted for approximately 95% of the microzooplankton numbers during the season. In the Control, rotifer
abundance increased significantly from May to September (p<0.05). Overall, nutrient enrichment substantially increased
rotifer communities in Mid and Late treatments (p<0.1), but not in the Early treatments. In May and June, rotifer abundance
was <0.2*105/m2 in all treatments, and communities were dominated by Bdelloid spp., Euchlanis spp., Hexartha mira,
Lecane spp., and Monostyla spp. In the Control, rotifers increased from 0.2*105/m2 in July to 2.0*105/m2 in September,
and were dominated by Brachionus spp., Euchlanis spp., H. mira, and Monostyla bulla. Rotifer abundance in the dosed
treatments was up to 3-times that in the Control, but dominant genera were similar among treatments.
Macrozooplankton populations in the Control were at a minimum in May (0.1*105/m2) and peaked in August at
0.8*105/m2. Nutrient enrichment did not significantly influence cladoceran or copepod numbers, or the cumulative mac-
rozooplankton community (p>0.05). In May, cladocerans represented 96% of the macrozooplankton in all treatments,
and were dominated by Sididae spp. In June and July, cladocerans accounted for 67% and 59% of the macrozoo-
plankton, respectively, and were dominated by Simocephalus serrulatus and Ceriodaphnia spp. Copepod populations
in June and July in all treatments averaged 50% calanoids and 50% cyclopoids. In August, at the peak of macrozoo-
plankton abundance, cladocerans accounted for 77% of the macrozooplankton numbers in all treatments. In August,
cladocerans were predominantly Alona spp., Ceriodaphnia spp., Chydorus sphaericus, and Simocephalus serrulatus,
and copepods were primarily cyclopoids (>94%). Cladoceran abundance declined at the end of the season, but still
accounted for 68% of final macrozooplankton numbers. Dominant genera of cladocerans and copepods in September
were the same as in August.
Sediment
N and P stocks in sediments changed significantly during the season (p<0.01), but were not significantly affected by
nutrient amendments at various macrophyte stages. The N content of sediments was highest in June and September,
averaging 0.35% N (Figure 83). Sediment content in July and August averaged 0.28% N. Sediments contained around
149 g N/(m2*5 cm deep) in June and September, and around 127 g N/(m2*5 cm deep) during July and August. Sediment
P was at a maximum in June at 0.07% P (Figure 84). In subsequent months, P content fluctuated around 0.06% P.
Sediment phosphorus pools ranged from 25 to 29 g P/(m2*5 cm deep) during the season.
Sediment
Nitrogen
Stage
WS8$
"„¥ ^ x\ x ^
Day
0.0033
Stage*Day
'>"'#•'.- •*y%8%%i,r'**v*!*i"1^*?81
'-,.$£ ' • .'; !: V||. ^9'|||C!l:^l,l;,,v;J ';,",
Sediment Nitrogen Pool, for Study 3
June
August
September
Month
Figure 83. Changes in sediment nitrogen pool over time. Treatments are differentiated by stage, which refers to the
stage of macrophyte development at the start of the dose period. The table presents probabilities that
stage, day, and interactions of stage and day influenced N content in sediments in ANOVA of rank-trans-
formed data. Darkened values are not significant (p>0.05). The graph is a plot of the estimated pool of N
in the sediments for an area of 1 rrf and a depth of 5 cm. Because treatment influences were not signifi-
cant, monthly values are presented as averages of all treatments. The average %N in sediments is shown
within the column for each month.
103
-------
Sediment
Phosphorus
Stage
Day
0.0051
Stage*Day
Sediment Phosphorus Pool, for Study 3
July August September
Month
Figure 84. Changes in sediment phosphorus pool over time. Treatments are differentiated by stage, which refers to
the stage of macrophyte development at the start of the dose period. The table presents probabilities that
stage, day, and interactions of stage and day influenced P content in sediments in ANOVA of rank-trans-
formed data. Darkened values are not significant (p>0.05). The graph is a plot of the estimated pool of P
in the sediments for an area of 1 rrf and a depth of 5 cm. Because treatment influences were not signifi-
cant, monthly values are presented as averages of all treatments. The average %P in sediments is shown
within the column for each month.
104
-------
System Metabolism
Estimates of gross community primary production (GCPP) and community respiration (CR) increased 3-fold over the
season among all treatments; the effect of Day was significant (p<0.0001) (Figure 85). In addition, there was significant
effect of the Stage*Day interaction (p=0.0374) on GCPP. Stage had no significant effect on either parameter.
Gross community primary production in the Control increased from around 4 mg O2/L in May and early June, to around
15 mg O2/L in August and September. Levels of GCPP significantly increased (p<0.05) in the Early treatment in early
May and ranged from 5.9-9.6 mg O2/L, or 2-fold Control levels. Thereafter, GCPP was similar in the Early treatment
and Controls through the end of the season with the exception of August 29 when the Early (8.7 mg O2/L) treatment was
lower than the Control (13.8 mg O2/L). Production in the Mid treatment was similar to the Control on all but two dates
during the season. On June 20, the Mid treatment (10.2 mg O2/L) was significantly higher (p<0.05) than the Control
(6.3 mg O2/L), and on August 15, Mid (11.6 mg O2/L) was lower than the Control (17.1 mg O2/L). Oxygen production
in the Late treatment was similar to the Control throughout most of the season.
Community respiration was similar in magnitude and seasonality to GCPP. There was no stimulation of CR due to the
Early treatment. Therefore, CR was less sensitive to dosing than GCPP.
Production
Respiration
Stage
'''-" 0.4523
1 "'/f'*, f . X*. 0,XNN
,' " -x * fcttlt%>" ^
Day
0.0001
0.0001
Stage*Day
0.0374
,.;/.^^v . ,,
Community Production, for Study 3
^ ?^ .'flr'*"'-*
;=^w a^v^-
5/10 5/17 5/24 5/31 6/7 6/14 6/21 6/28 7/5 7/12 7/19 7/26 8/2 8/9 8/16 8/23 8/30 9/6
Date
B
Community Respiration, for Study 3
5/10 5/17 5/24 5/31 6/7 6/14 6/21 6/28 7/5 7/12 7/19 7/26 8/2 8/9 8/16 8/23 8/30 9/6
Date
Figure 85. Changes in community oxygen and respiration production over time. Treatments are differentiated by
stage, which refers to the stage of macrophyte development at the start of the dose period. The table
presents probabilities that stage, day, and interactions of stage and day influenced production and res-
piration in ANOVA of rank-transformed data. Darkened values are not significant (p>0.05). The graphs
are plots of production (A) and respiration (B) over the experimental season pooled by treatment. Corre-
sponding lines and arrows mark dose periods.
105
-------
Net Nutrient Balance
In Study 3, we applied equivalent loads of nutrients at different stages of macrophyte growth: Early (0% cover; initiated
May 11), Mid (15-25% cover; June 12), or Late (75-90% cover; July 5) (Figure 59). The three dosed treatments received
the same nutrient load (30 g NO3-N/m2; 1.2 g P/m2) applied as six successive weekly additions of 5 mg NO3-N/L and
200 ug P/L (25:1 N:P ratio). We calculated a final mass balance of nutrients among various nutrient pools to determine
the net efficiency of uptake and assimilation of nutrients in these experimental systems (Table 10).
At the end of the study, macrophytes contained a total of 4.29, 3.92, 4.81, and 4.60 g N/ m2 in the Control, Early, Mid,
and Late treatments, respectively (Table 10). Water contained an additional 0.63, 0.83, 0.78, and 1.03 g N/m2 in the
Control, Early, Mid, and Late treatments, respectively. Combined (macrophytes + water), these two nutrient pools con-
tained 4.92, 4.75, 5.59, and 5.63 g N/m2 stocks at the end of the study. Thus, a total of 16% (Early treatment), 19%
(Mid treatment), and 19% (Hi treatment) of total nitrogen added during the study were found in these two major nitrogen
pools at the end of the study. This implies, under simple mass balance conditions (i.e., no loss to the atmosphere) that
up to 84% (Lo treatment) and 81% (Mid and Late treatments) of total nitrogen added was sorbed to sediments or lost
to the sediments as detritus. Attempts to measure actual sediment transfers were unsuccessful due to a combination
of the large mass of pre-existing nitrogen in sediments and the error contributed by our sampling procedures. How-
ever, we know that these numbers are conservative, since the Control treatment alone revealed a total sequestration
of 4.92 g N/m2 (combined macrophytes and water) at the end of the study in the absence of external nitrogen addition.
The large percentage of nitrogen that was not accounted for at the end of the study (84%, Early treatment; 81 % Mid
and Late treatments) indicates that these shallow, vegetated aquatic systems served as efficient biological, chemical,
and physical sinks for nitrogen.
Phosphorus, likewise, was efficiently assimilated and retained in the study (Table 10). At the end of the study mac-
rophytes contained a total stock of 0.80, 0.65, 0.86, and 0.84 g P/m2 in the Control, Early, Mid, and Late treatments,
respectively. Water contained an additional stock of 0.07, 0.08, 0.11, and 0.15 g P/m2 in the Control, Early, Mid, and
Late treatments, respectively. Using conservative mass balance estimates, subtracting the amount of phosphorus in
macrophytes and water from that applied in dosing indicates that sediments had a net accrual ot 0.47 (Early treatment),
0.23 (Mid treatment), and 0.21 (Hi treatment) g P/m2. Macrophytes in the Control, Mid, and Late treatments contained
equivalent amounts of phosphorus at the end of the study even though no phosphorus was applied to Controls. In
contrast, the Early treatment contained 25% less phosphorus than the other treatments. This difference in macrophyte
storage is largely due to the significantly lower biomass (Figure 59) and P stock (Figure 63) in macrophytes observed
in the Early treatment compared to the Control, Mid, and Late treatments. Thus, Early treatment, prior to macrophyte
development, somewhat inhibited the growth of macrophytes, though the overall phosphorus uptake and storage were-
similar and efficient across treatments.
Table 10. Summary Table of the Final Store of Phosphorus and Nitrogen in Water and Macrophytes in Study 3.
Load Represents the Total Amount of Fertilizer P or N Added in Each Treatment. Stores in Macrophytes
are Considered as Grams Per Cubic Meter Because Water Depth was 1 m
Nitrogen
Control
Early
Mid
Late
Load
(g N/m2)
0
30
30
30
"W J55, « +
4.29
3.92(13%)
4.81 (16%)
4.60(15%)
0.63
0.83 (3%)
0.78 (3%)
1 .03 (3%)
4.92
4.75(16%)
5.59(19%)
5.63(19%)
?m: ^;.i.. ...'.. ii <***• ,.-',; „.'',*, '.'*'•'* ';w ••••?•+<'••'•"' •- ~ •„ • ' "- • • - * ,-:. •; \,,;,^; ;%t
Phosphorus
Control
Early
Mid
Late
Load
(g P/m2)
0
1.2
1.2
1.2
Macrophytes
(g P/m2)
0.80
0.65 (54%)
0.86 (72%)
0.84 (70%)
Water
(g P/m2)
0.07
0.08 (7%)
0.11 (9%)
0.15(13%)
Macrophytes +
Water Total
(g P/m2)
0.87
0.73(61%)
0.97(81%)
0.99 (83%)
Presumed Sediment
Transfer
(g P/m2)
NA
25.25 (84%)
24.41 (81%)
24.37(81%)
j ;«a«*4- ««"" _,, ,,v~^ ,.**••
Presumed Sediment
Transfer
(g P/m2)
NA
0.47 (39%)
0.23(19%)
0.21 (18%)
1Mean, (n); [STD.COV]
106
-------
Discussion
Three studies were conducted over a 2-year period to determine the nutrient assimilation capacity of shallow vegetated
aquatic systems. We evaluated the effects of magnitude of dosing; N:P ratio; frequency of dosing, and timing of dosing.
We hypothesized that these various nutrient loading regimes would stimulate periphyton and phytoplankton growth,
and subsequently, macrophyte development would be inhibited due to light limitation. An alternative stable state, domi-
nated by phytoplankton, would occur that was less efficient than the original macrophyte community in assimilating
and retaining nutrients from the water column. Ultimately, these results were not obtained. Under all scenarios, the
macrophyte-dominated system persisted, and nutrient uptake and assimilation were maintained at over 90% efficiency.
In some cases, macrophyte biomass was reduced below control levels; however, these experimental aquatic systems
retained their ability to sequester excess nutrients, as described below.
Nutrient Dissipation Rates
Both nitrogen and phosphorus rapidly dissipated in these studies regardless of nutrient loading rate, N:P ratio, or tim-
ing of nutrient application in relation to macrophyte stage. Target concentrations of nitrogen were usually quite close to
nominal dose applied. However, phosphorus losses were rapid, and concentrations were frequently less than 25% of
nominal (i.e., Study 1). Rapid loss rates of phosphorus cannot be explained by biological uptake alone; rather, much
of the phosphorus was probably lost to precipitation with iron and calcium carbonate similar to that observed in high
alkalinity marl ponds (Otsuki and Wetzel 1972, Wetzel 1983). Phosphorus precipitation was greatest early in dosing
intervals when hardness and alkalinity were still high due to the influence of the well water source. Therefore, nutrient
dissipation rates on a mass basis varied and underestimated actual losses in the case of phosphorus due to rapid up-
take and precipitation. Percentage losses, calculated as the percentage nominal concentration applied to the amount
remaining immediately prior to the next dose, indicate that phosphorus dissipation rates (14-80% P/day) exceeded
nitrogen dissipation rates (4-12% N/day) but that both were rapid. Nutrient uptake was greatest during the period rang-
ing from May-July. Thereafter, however, nutrient uptake decreased due in part to macrophyte senescence, decomposi-
tion, and possible release of nutrients from sediments. Highest nutrient additions, added in the Hi treatment in Study
2 (60 g/m2 N; 2.4 g/m2 P) resulted in an overall loss of 95% of dosed N and 86% of dosed P by the end of the study.
Similar results were observed by Balls et al. (1989) when various rates of nitrogen (up to 29 g N/m2) and phosphorus
(up to 2.3 g P/m2) were added to vegetated ponds over the period May-October: half of added nutrients were gone in
2 days, and nutrients were near background levels within a 2-week period.
Nutrient Stocks
Nutrient stocks were evaluated in major ecosystem compartments (macrophytes, water, and sediments) to determine
relative pools of nutrients over time. Sediments represented the single highest nutrient stock (and remained stable over
time in spite of dosing) and contained an average of 150 g/m2 N and 23 g/m2 P. Macrophytes contained the second
highest stock of nutrients. Maximum macrophyte nutrient stocks ranged from 2.8-22 g/m2 N and from 0.7-3.3 g/m2 P at
peak macrophyte biomass; macrophyte nutrient stocks were usually lower at the end of each study (1.05-13.00 g/m2 N
and from 0.65-2.21 g/m2 N) due to loss at senescence. Thus, macrophytes at peak biomass contained less than 15%
nitrogen and phosphorus even at maximum stock levels. Water contained significantly lower nutrient stocks compared
to macrophytes or sediments, and only became significant pools of nutrients at the end of the study as macrophytes
senesced.
Phytoplankton Dynamics
Phytoplankton Biomass and Growth Rates
Phytoplankton contributes the greatest amount of primary productivity to many deep aquatic ecosystems including
oceans, large lakes, and reservoirs (Wetzel 1983). However, the contribution of phytoplankton productivity to shallow
aquatic systems is variable and depends on a combination of physical (e.g., depth, turnover time, inorganic turbidity),
chemical (e.g., nutrient status), and biological (e.g., presence or absence of macrophytes, zooplankton grazing pressure).
Phytoplankton, under unlimited conditions of light and nutrients, have greater potential for production per unit carbon
compared to epiphytes or macrophytes for several reasons: 1) phytoplankton have a greater surface area: volume ratio
107
-------
compared to epiphytes or macrophytes; 2) cell walls are thinner in phytoplankton; 3) diffusional gradients are less for
phytoplankton due to the combination of thin cell walls and frequent mixing in the water column; and 4) phytoplankton
are less light limited due to water mixing, and in some cases motility, which maintains cells in the photic zone (Sand-
Jensen and Borum 1991).
We hypothesized that nutrient additions, under some cases, would stimulate phytoplankton to the point where shad-
ing would result in light limitation of macrophytes and epiphytes; system dominance would then shift from macrophyte
dominance to an alternative stable state (Scheffer 1990) dominated by phytoplankton (Scheffer 1998). This shift is
similar to that documented in a eutrophic, shallow lake in Austria by Mayer et al. (1997). These predictions, however,
did not occur in our studies. In Study 1, we varied N:P ratio and loading rates throughout the summer. Early in the
season, phytoplankton biomass responded positively to phosphorus dosing, and Chi a significantly increased up to
35 ug/L through late July; however, water clarity was retained. Maximum phytoplankton biomass rapidly increased in
all treatments in September in Study 1 (up to 200-320 ug/L Chi a) approximately 2 weeks after the last dose. However,
only Day and the N-dose*Day were significant factors late in the study; main effects of N dose and P dose were not
significant. Chi a increased in all treatments (unrelated to treatment effects) as nutrients were released from senesc-
ing macrophytes in combination with possible sediment release of nutrients. In a field study, Landers (1982) reported
that phytoplankton chlorophyll increased from 10 to around 90 ug/L during senescence in macrophyte enclosures, but
chlorophyll remained around 10 ug/L in denuded enclosures. Based on those findings, he concluded nutrient release
by senescing macrophytes caused the proliferation of phytoplankton.
In Study 2, both Lo and Hi dosing early in the study resulted in phytoplankton increases up to 33 ug/L Chi a; however,
Chi a did not linearly increase with dosing due to internal damping mechanisms. Maximum phytoplankton biomass in
Study 2, which intensified loading of nutrients, was observed in the Hi dose in early July (90 ug/L Chi a) just prior to the
end of dosing; however, chlorophyll declined soon thereafter and was not maintained. In Study 3, we varied nutrient
dosing by applying nutrients (Early, Mid, and Late) under varying stages of macrophyte development. Chlorophyll a
was significantly related to timing of the dose, but varied in 2-week cycles. In contrast, nitrates gradually increased with
dosing, whereas SRP cycled in 7-day spikes due to rapid dissipation. Therefore, a combination of differential nutrient
controls and increasing grazing pressure by zooplankton acted to maintain Chi a levels below what would be expected in
the absence of internal control mechanisms. Maximum phytoplankton biomass in Study 3, which varied nutrient loading
in relation to macrophyte stage, peaked in July and late August (55 and 75 ug/L Chi a, respectively) in the treatment
receiving the Early dosing. However, phytoplankton biomass in Ihis treatment was not consistently different from the
Control, which indicates that internal mechanisms again masked any lasting effect of nutrient timing or dose.
In summary, phytoplankton responses in all three studies were similar in that we saw positive phytoplankton responses
due to nutrient dosing early in the study. However, the degree of phytoplankton response was less than expected. Balls
(1989) observed similar responses in mesocosm experiments when phytoplankton response to nutrient enrichment
was much lower than predicted in bioassays. They concluded that macrophytes and epiphytes may have assimilated
nutrients thereby reducing nutrient availability to phytoplankton; they also suggested that zooplankton grazing became
important in the regulation of algal biomass at some point in late spring. Therefore, phytoplankton was regulated due
to a combination of top-down (zooplankton grazing) and bottom-up (nutrient competition by epiphytes) internal control
forces (Bronmark and Hansson 1998; Sondergaard and Moss 1998). In our studies, peak phytoplankton biomass ranged
from 75-310 ug/L Chi a across all studies, which is extremely high for aquatic systems such as lakes and reservoirs and
would lead to a trophic classification of hypereutrophy and loss of macrophytes (Carlson 1977). However, numerous
studies have indicated that traditional trophic state indices based on phytoplankton do not apply to shallow, vegetated
aquatic systems because of interacting internal mechanisms that allow macrophytes and clear-water conditions to persist
even though nutrient loading is high (Gasith and Hoyer 1998; Bachman et al. 2002).
Phytoplankton Species Composition
None of the experimental treatments in our three studies resulted in significant shifts in dominant in phytoplankton com-
munity composition. Some of our treatments were designed to result in nitrogen limitation, which should favor develop-
ment of cyanophytes due to their ability to fix atmospheric oxygen (Wetzel 1983). In spite of nitrogen-limited conditions
in some treatments, we observed a dominance of chlorophytes rather than blue-greens. Research has demonstrated
that some macrophytes such as Ceratophyllum demersum (Kogan and Chinnova 1972; Koerner and Nicklisch 2002;
Gross et al. 2003), Myriophyllum spicatum (Koerner and Nicklisch 2002), Najas marina (Gross et al. 2003), and Chara
sp. (van Donk and van de Bund 2002) can produce allelopathic substances that inhibit growth of both green algae and
cyanobacteria. However, allelopathic effects are rather species specific as opposed to acting on general taxonomic
groups of algae and cyanobacteria (Koerner and Nicklisch 2002; Gross et al. 2003). It is unclear in our study if macro-
phytes produced allelopathic substance that may have inhibited some species of the phytoplankton.
Others have suggested that in shallow lakes the sediments may release nutrients that promote chlorophytes over blue-
greens. Based on their survey of 178 lakes, Jensen et al. (1994) concluded that chlorophytes, with a higher growth rate
108
-------
than blue-greens, may have a competitive advantage in shallow systems where sediments release pulses of nutrients.
In Lake Sobygard, a system in which sediment release has been the major source of P for phytoplankton, chlorophytes
dominate in the summer and autumn (Sondergaard et al. 1990). In our study, sediment release of P was consistent
among treatments, and may explain the continued dominance of chlorophytes into the late season.
In studies on small, shallow, macrophyte-dominated ponds, Mulligan et al. (1976) observed that phytoplankton in
control treatments were dominated by chlorophytes and cyanophytes. Following heavy fertilization (total load: 75 mg
N/L; 7.5 mg P/L), they observed Chi increases up to 300 ug/L and dominance by different phytoplankton species that
were not observed in the Control; however, the community was still dominated by chlorophytes and cyanophytes. We
observed similar results in our studies. Therefore, relative shifts of phytoplankton under nutrient-enriched conditions
may differ than that frequently observed in deeper, limnetic systems.
Periphyton Dynamics
Periphyton B/omass and Growth Rates
Periphyton is defined as the sessile assemblage of diatoms and algae that forms on underwater surfaces including mac-
rophytes, sediments, and other surfaces such as corral sides. The role and dynamics of penphyton have received far
less focus in studies compared to macrophytes and phytoplankton because of the difficulty in sampling and subsequent
bias associated with incubations for productivity estimates. Therefore, researchers frequently use colonization rates of
artificial substrates for estimates of periphyton productivity. In our studies, we incubated Scrimweave™ strips below the
water surface to measure accumulation of Chi a at 1 and 2-week intervals. This technique does not account directly
for periphyton on plants, sediments, or corral sides which are subject to differential influence of light and invertebrate
grazing. It does, however, standardize the effects of surface area and light in order to partition the relative influence of
experimental dosing and nutrient availability.
Periphyton productivity, or accrual rate, was low in July of Study 1 and ranged from 0.03-0.04 ug Chl/cm2/wk; accrual
rates were similar across treatments. Maximum biomass accrual (0.24 ug Chl/cm2/wk) was observed in the August and
September samples as nutrient dosing rates were increased. Biomass levels were comparable to those observed on
artificial substrates in other studies (Cattaneo and Kalff 1980; Brock et al. 1995; James et al. 2000). Both N and P dosing
had significant effects on periphyton growth rates; however, the N dose had a greater relative effect than P dose. In his
review of nutrient cycling in shallow lakes, Lijklema (1994) concluded that spring algal biomass is determined by initial
nutrient concentrations and external loading, whereas later in the season, internal loading may play a greater role as a
nutrient source. This theory is supported by the periphyton dynamics in our study. In June, periphyton accrual rates were
higher in P-dosed than OP treatments, indicating P-limitation. Later in the season, August and September, periphyton
responded to N-dosing; P availability was not dependent on phosphorus amendments, as indicated by equal or higher
levels of periphyton growth in OP than P-dosed treatments. Water column SRP and TP increased during the late season
in all treatments indicating that internal P loading, whether from macrophytes or sediments, was occurring.
Periphyton dynamics differed in Studies 2, however, when nutrient doses were applied prior to macrophyte development.
Highest periphyton biomass accrual (2.75 ug Chl/cm2/wk) was observed in May in the Hi dose treatment. Thereafter,
biomass accrual decreased as nutrient competition became more intense in June. A residual dosing effect was observed
in July, August, and September, but biomass was much lower (<0.6 ug Chl/cm2/wk) after the dosing period (end June 10)
when nutrients became limiting. Periphyton biomass accrual in Study 3 was also greatest in May and June (maximum
1.65 ug Chl/cm2/wk) during the Early and Mid dosing intervals; however, dosing had much less effect during the Late
dosing interval. These observations imply that nutrient availability was high early in the study when both macrophyte
and phytoplankton biomass was low, but that late in the study, as phytoplankton biomass increased, periphyton was
limited by both nutrients and light.
We did not measure actual nutrient pools in periphyton. However, approximate nutrient stocks in periphyton can be es-
timated using conversion rates of chlorophyll and nutrient relationships from the literature. Stelzer and Lamberti (2002)
found that chlorophyll, nitrogen, and phosphorus comprised approximately 0.03%, 0.7%, and 0.06%, respectively, of
periphyton on a dry-wt basis. Therefore, chlorophyll conversion factors of 23 and 2 could be used to estimate nutrient
stocks of periphyton in our study. Using these conversion factors, peak levels of periphyton observed (2.75 ugChl/cm2;
Study 2), respectively, and the total area of Scrimweave™ material in each corral (12.54 m2), we estimate that the en-
tire nutrient stock associated with periphyton on the Scrimweave™ corral sides would be approximately 9.0 g N and
0.69 g P per corral. Values adjusted to an area basis (12.56 m2 surface area of each corral) would be 0.71 g N/m2 and
0.06 g P/m2. Thus, on a standing crop basis, periphyton associated with the corral sides in Study 2 contained approxi-
mately equivalent amounts of nutrients as water; 18% compared to macrophytes; but less than 1% as compared to
sediments. Percentages in Studies 1 and 3 are much less than this calculation due to proportionately less periphyton
biomass in these studies.
109
-------
It is difficult to estimate the total surface area of macrophyte stands due to the high level of surface area associated
with individual leaves or whorls. However, Bachman et al. (2002) estimated that periphyton biomass associated with
macrophytes accounted for approximately 1.8% dry weight of submerged macrophytes in a survey of 319 shallow,
vegetated lakes in Florida. In addition, the percentage of nitrogen (2-4% dry wt) and phosphorus (0.2-0.5% dry wt) of
macrophyte tissue in our study exceeds that estimated for periphyton based on the findings of Stelzer and Lamberti
(2002). Thus, it is apparent that periphyton represented a relatively minor component of nutrient pools associated with
macrophytes or the corral sides in this study.
However, biomass is not a good estimator of productivity for periphyton. Periphyton can exhibit high rates of productivity
under conditions of high nutrient availability even though standing crop is held low due to grazing (Cattaneo and Kalff
1980, Sand-Jensen and Borum 1991). Nutrient uptake by periphyton can be rapid from the water column. Cell walls
of periphyton are thicker than phytoplankton, but thinner than macrophytes, which allow intermediate rates of nutrient
uptake; furthermore, the diffusional boundary of periphyton can vary depending on nutrient concentrations and the
degree of mixing of the water column (Sand-Jensen and Borum 1991). In this study, however, mixing was undoubtedly
low due to the confinement of the water column by the corral sides which minimized wind-mixing; this may have de-
creased productivity of periphyton due to decreased nutrient exchange. Highest levels of periphyton growth rates were
observed early in Study 2 when dosing occurred and invertebrate grazing was low due to the lack of colonization time;
in addition, macrophyte shading was minimal due to low biomass. In contrast, in Study 1 and portions of Study 3, it is
likely that macrophyte shading and nutrient limitation led to decreased periphyton productivity. Thus, we believe that
periphyton productivity was probably not a major sink for nitrogen or phosphorus compared to other compartments. In
some systems, however, constant nutrient renewal and invertebrate grazing can mask high rates of periphyton produc-
tivity as long as light is not limiting by turbidity or macrophyte shading. The relative role of periphyton in nutrient uptake
and assimilation therefore is a priority for future studies of nutrient dynamics in shallow water bodies.
Macrophyte Dynamics
Macrophyte Biomass and Growth Rates
Macrophyte growth rates were robust in these studies; however, there were differences across years in maximum bio-
mass. In Study 1 (1999) macrophyte biomass in Controls peaked at maximum dry weight biomass of 800 g/m2 in July.
Macrophyte biomass in Studies 2 and 3 (2000) peaked in August at 213 g/m2, which is 75% less than the previous year.
Macrophyte biomass can vary substantially within these experimental systems. For example, maximum biomass aver-
aged 122 g/m2 (Fairchild and Sappington 2002), 170 g/m2 (Fairchild et al. 1994), and 330 g/m2 (Fairchild et al. 1992)
among mesocosms in three different studies conducted at CERC. The peak macrophyte biomass observed in Study
1 (800 g/m2) is higher than previously observed; however, in these corral studies, sampling was restricted to waters
< 1-m deep. Lower biomass observed in year 2000 likely occurred due to the extended drawdown period for corral
construction in 2000 (26 days) compared to 1999 (16 days).
Various nutrient regimes applied during these studies had variable effects on macrophyte growth. Maximum macrophyte
biomass was not affected by nitrogen load, phosphorus load, or the N:P ratio when loads of up to 30 g N/m2 and 0.86 g
P/m2 were applied on a bi-weekly basis to corrals containing approximately 25% surface coverage of macrophytes (Study
1). Similarly, there was no effect of nutrient dosing (30 g N/m2 and 1.2 g P/m2) when dose frequency was increased to
weekly dosing (Study 3) when nutrients were applied either mid-season (i.e., Mid; 15-25% macrophyte coverage, June
12) or late-season (i.e., Late, 75-90% cover; July 5). Balls et al. (1989) also found macrophytes to persist in 0.8-m
deep ponds dosed over 5 months (thirteen doses totaling loads of 29 g N/m2 and 2.3 g P/m2). In contrast, we found
that nutrient dosing prior to macrophyte development (early in growing season) caused significant decreases in macro-
phyte growth and biomass when applied at weekly low (30 g N/m2 and 1.2 g P/m2) and high dose (60 g N/m2 and 2.4 g
P/m2) levels. However, early dosing did not totally eliminate macrophyte stands or shift the systems to a phytoplankton-
dominated alternative stable state (Scheffer 1990, 1998). These results indicate that shallow, macrophyte-dominated
systems can persist at dosing of up to 60 g N/m2 and 2.4 g P/m2:; however, this may be the upper limit of early loading
that will maintain a macrophyte-dominated system structure. Results also indicated that A/a/as relied primarily on the
sediments as its primary nutrient source since added nutrients did not stimulate macrophyte growth. Laboratory studies
have also shown that sediments are the primary source of nutrients for A/a/as sp. (Moeller et al. 1988).
There are several mechanisms that can limit the existence and persistence of macrophytes in shallow aquatic systems.
Inorganic turbidity, for example, can decrease light penetration and therefore eliminate submerged macrophytes (Wetzel
1983). Inorganic turbidity can result from erosional runoff, wind activity/wave action, and bioturbation due to fish and
invertebrates (Engel 1990; Gasith and Hoyer 1998; Horppila and Nurminen 2003). None of these factors were present
in our study due to the experimental design and tight control of experimental conditions. Depth is also a significant fac-
tor limiting growth of submerged macrophytes (Wetzel 1983; Haekanson and Boulion 2002). Macrophytes can persist
in clear, oligotrophic systems at depths of up to 10 meters due to a combination of light attenuation and increasing
hydrostatic pressure (Wetzel 1983). However, actual, realized maximum depth of macrophyte distributions in most
mesotrophic-eutrophic littoral aquatic systems is much less.
110
-------
We observed the loss of macrophytes in Study 1 in two dosed corrals in the deepest part of the mesocosm (approxi-
mately 1.1 m depth); in these corrals, phytoplankton biomass and resulting turbidity were noteably higher, suggesting
that loss of macrophytes occurred due to light attenuation by phytoplankton. Phytoplankton turbidity due to nutrient
enrichment is a primary limiting factor for macrophytes in eutrohpic systems (Moss 1976). Mulligan et al. (1976)
demonstrated that high loading of nutrients (75 g N/m3; 7.5 g P/m3) prolonged phytoplankton blooms (>3 months with
Chi >100 ug/L), which in turn eliminated some plants (Chara spp., Myriophyllum spicatum), and temporarily inhibited
others (Elodea canadensis) by shading. The highest loading rates applied in our studies (60 g N/m2 and 2.4 g P/m2)
prior to macrophyte development did not result in phytoplankton blooms; chlorophyll levels remained less than 40 ug/L
during the dose period, and turbidity was less than 10 NTU's. Zooplankton numbers were low in the first month of the
study, and therefore zooplankton grazing did not appear intense. It is probable that epiphytes on macrophytes, corral
sides, and perhaps sediments were primary factors in nutrient dissipation that minimized phytoplankton dominance.
Subsequently, light penetration remained sufficient for macrophyte survival, growth, and persistence. Seasonal tim-
ing of nutrient additions can also be a significant factor in the ability of phytoplankton to out-compete macrophytes. As
macrophytes grow, the canopy moves higher in the water column and light is less of a limiting factor (Scheffer 1998).
Therefore, factors that reduce light, such as phytoplankton or inorganic turbidity, will have the greatest impacts in the
early stages of macrophyte development.
Periphyton can also limit the growth rates and ultimate standing crop of macrophytes by excessive growth on leaves and
stems resulting in shading and light limitation (Phillips et al. 1978; Cattaneo and Kalff 1980; Sand-Jensen and Borum
1991). In laboratory enrichment experiments, Phillips et al. (1978) demonstrated that periphyton restricted the growth
of Najas marina under enriched conditions; in separate field experiments, they linked poor development and prema-
ture decline of macrophytes (including Najas marina) in the field to light limitation imposed by dense epiphytic growth.
Sand-Jensen and Borum (1991) also suggest that periphyton may also inhibit macrophyte growth by slowing diffusion of
carbon dioxide and oxygen. In our experiments, periphyton accrual rates in the early, high weekly dosing regimes were
40-times and 5-times rates measured in the Controls in May and June, respectively. Therefore, persistent development
of periphyton may have reduced light reaching macrophyte surfaces, thereby inhibiting macrophyte growth.
Self-shading of macrophytes is also a significant factor in many aquatic ecosystems. As macrophytes grow in height,
the amount of light reaching the lower leaves becomes limiting, and may result in "shedding" of old leaves that cannot
receive enough light to be photosynthetically active. This is a form of self pruning, and can present a significant mecha-
nism of nutrient deposits to sediments in the form of detrital material (van Donk et al. 1993). Such "leaf shedding" was
not measured directly in this study. However, the significance of the effect was often noted late in the studies during
deployment of zooplankton traps. It became difficult to place the traps into the corrals without disturbing accumulated
organic debris resulting from senescent macrophytes. In addition, overnight release of gases from sediment respira-
tory processes entrained significant amounts of organic material from sediments and plants which were carried up into
the funnel traps.
Macrophyte Species Composition
Macrophyte stands in the CERC experimental mesocosms are dominated by two species of macrophytes: Chara sp.,
a macroalgae; and Najas guadalupensis, a submerged vascular angiosperm (Fairchild et al. 1992, 1994; Fairchild and
Sappington 2002). Najas normally comprises approximately 70-90% of the biomass in these systems, whereas Chara
ranges from 10-30% biomass (Fairchild et al. 1994; Fairchild and Sappington 2002). Chara populations were observed
early in Study 1 which was conducted in year 1999. In Study 1, Chara was observed early in the study but declined
in abundance during June until a monospecific stand of N. guadalupensis developed. Chara sp. was not observed in
Studies 2 and 3 that occurred in year 2000, macrophyte stands consisted of a monospecific stand of N. guadalupensis.
The lack of Chara sp. early in 2000 was most likely due to the extended draw-down period of 2000 (26 days) compared
to 1999 (16 days).
Seasonal succession from Chara to stands of Najas, Potamogeton, and other angiosperms has frequently been ob-
served in aquatic systems (e.g., Crawford 1977; Wood 1950). Seasonal declines in Chara abundance can occur due
to excessive nutrients, lack of nutrients, and light limitation. Forsberg (1964) observed poor charophyte development in
the laboratory at phosphorus levels around 20 ug P/L; similar field observations were made. Forsberg (1964) proposed
that phosphorus concentrations at or above 20 ug P/L had an inhibitory effect on Chara, although the mechanisms of
inhibition were not known. Such an inhibitory effect was probably not a factor in this study, because Control populations
of Chara declined in Study 1 when TP was <20 ug P/L.
Nutrient limitation can also be a factor in Chara declines. Najas and Chara are capable of absorbing nutrients from the
surrounding water, but their ability to utilize nutrients from the sediments depends on the efficiency of roots in Najas and
rhizoids (root-like filaments) in Chara. Chara has demonstrated the ability to assimilate nutrients from water through its
rhizoid and other structures (Box 1986, 1987; Kufel and Kufel 2002). Najas has been shown to take up N and P in fertil-
ized sediments, and in other lab experiments Najas relied on the sediment for nearly all of its phosphorus requirement
111
-------
(Moeller et al. 1988). A limited supply of nutrients in the water column, especially phosphorus, necessitates utilization of
sediment stores for growth. Najas may have had an advantage over Chara under those circumstances, because roots
provided a more efficient means of nutrient uptake than rhizoids (Moeller et al. 1988; Kufel and Kufel 2002).
Other authors have indicated that light limitation was the main factor causing Chara replacement due to turbidity or
shading by macrophytes (Blindow 1992; Crawford 1977). Turbidity observed in May and June of 1999 (<4 NTU's) was
not at a level that would likely cause Chara to decline. However, physical shading may have been a factor. Fairchild
et al. (1994) previously provided evidence that shading by Najas was a factor in competition with Chara, when applica-
tion of 50 ug/L atrazine reduced Najas populations, released Chara sp. from light limitation, and allowed Chara sp. to
dominate the aquatic macrophyte community. Chara was only observed in Study 1 and nutrient addition itself did not
negatively alter Chara incidence or macrophyte growth. Collectively, these results indicate that shading by Najas and
epiphytes was the primary factor causing decline of Chara.
Macrophytes as Nutrient Sinks
Although nutrient dosing did not increase the growth rates of macrophytes, it did significantly increase the apparent
nutrient concentrations of macrophytes. Macrophytes in the Control treatment averaged 1.90%N and 0.22%P at the
beginning of Study 1 and increased to 3.02%N and 0.43%P by the end of the study. Nutrient dosing significantly
increased concentrations of both N and P in macrophytes in Study 1 at up to 3.95%N and 0.52%P in the high dose
treatments. In Studies 2 and 3, we observed similar concentrations and trends of N and P in the Control treatment. We
also observed significant increases in nutrient content of macrophytes due to the effect of Dose (Study 2) and Stage
(Study 3) of macrophytes related to timing of dose; nutrient concentrations in macrophytes increased in proportion to
dosing of both N and P and in some cases increased over 200% greater (4.35%N and 0.95%P; July, Study 2) than
nutrient concentrations compared to the Control (2.13%N and 0.48%P; July, Study 2). Although nutrient concentra-
tions of macrophytes increased under all treatments, macrophyte nutrient stocks varied and strongly reflected trends in
macrophyte biomass. For example, in Study 2, when Early dosing significantly decreased macrophyte biomass, the
nutrient stock declined accordingly.
Our measurements of N and P concentrations in Najas from Control macrophytes are similar to those in the literature.
Royle and King (1991) examined nutrient concentrations in macrophytes in Lake Liddell, New South Wales; nutrient
concentrations in Najas marina, Vallisneria spiralis, Potomogeton perfoliatus, and Potomogeton pectinatuswere similar
when based on means by species (1.51-1.91% N; 0.15-0.19% P). Boyd (1970) also found similar concentrations of P
(0.15% P) in Najas guadalupensis; concentrations ranged from 0.12% to 0.27% P among five other submerged species.
Therefore, Najas appears to be intermediate in its ability to store nutrients.
Macrophytes as Nutrient Sources
Macrophytes do not provide significant dissolved nutrient sources to the water column during periods of active growth.
However, macrophytes serve as significant sources to sediments via nutrient translocation from roots and through
shedding and sloughing of leaves and associated periphyton (Barko et al 1991). In addition, senescing macrophytes
release significant amounts of dissolved phosphorus and nitrogen to the water column that are utilized by periphyton
and phytoplankton (Landers 1982; Engel 1990).
Release of nutrients during macrophyte senescence varied in our studies. In Study 1, we applied varying levels of
nitrogen and phosphorus over a 16-week period. Macrophyte growth was robust, reaching a maximum of biomass
levels in July of 661-802 g/m2 dry weight among treatments. Subsequent macrophyte senescence resulted in a loss
of up to 50% of N and P stocks. Much of this loss was not accounted for in the water column; therefore, it is likely that
the majority of these nutrients were transferred to sediments as observed by van Donk et al (1993) and Stachowicz et
al. (1994). However, these transfers were not observed in actual measurements of sediment stocks due to the degree
of error of the chosen sampling method as compounded by the large pool of N in sediments (160 g/m2). Significant
amounts of nitrogen and phosphorus were also transferred to the water column as indicated by dramatically increasing
amounts of phytoplankton chlorophyll a and concurrent 3-fold increases in TN and TP across all treatments.
Similar, but less dramatic results were obtained in Study 3 in the Mid and Late dosing. In contrast, macrophyte senes-
cence and nutrient releases were less profound in Studies 2 and 3 even though total nutrient loading was increased.
In Study 2, when nutrients were applied before macrophytes emerged, total macrophyte biomass reached a maximum
of 213 g/m2 dry weight in the Control in August. Macrophytes began to senesce over the next 30 days, but N and P
stocks in macrophytes were maintained, and phytoplankton biomass, as indicated by chl a,TN, andTP did not substan-
tially increase. The decreased effects of nutrient release in Studies 2 and 3 may be due to the proportionately smaller
macrophyte biomass (75% less; maximum 213-227 g/m2 dry weight) compared to Study 1 (maximum 661-802 g/m2
dry weight). Therefore, fewer nutrients were available for release, and degree of senescence at the end of the study
was less due to decreased macrophyte self-shading and perhaps a truncated sampling duration. It is evident, however,
that macrophytes can serve as both sources and sinks for nutrients in shallow vegetated aquatic systems, but likely
112
-------
depend on factors such as biomass, species, and other physical and chemical factors (Barko et al. 1991; van Donk et
al. 1993; Stachowicz et al. 1994).
Zooplankton Dynamics
Zooplankton are known to play a major role in water quality and clarity of deep-water limnetic systems (Brooks and
Dodson, 1965; Irvine et al. 1989). Zooplankton, especially the cladocerans, are efficient filter feeders that serve to graze
down zooplankton populations in systems where fish are absent. When fish are present, they frequently selectively feed
on larger bodied zooplankton such as the cladocerans; under high levels of fish predation, the zooplankton community
often shifts to small-bodied rotifers and copepod nauplii which are less efficient in grazing phytoplankton. Subsequently,
in systems containing fish predators, cladoceran zooplankton numbers decrease; algae tends to increase; and water
clarity deceases.
In Study 1, the zooplankton community was dominated by the cladoceran Ceriodaphnia reticulata early in the study
prior to corral construction and treatment. In the absence of fish, C. reticulata was able to effectively graze down phy-
toplankton and maintain a high level of water clarity. Subsequently, macrophyte growth and biomass development were
high. Cladocerans continued to dominate the zooplankton community throughout the study in spite of increased nutrient
loading. Late in the study, algae began to increase due to nutrient release from macrophytes, and total cladocerans
increased concordantly; therefore, water clarity was maintained (turbidity < 7 NTU's) until the final two weeks of the study.
These findings support those of Hansson (1992), which suggested that in systems with two main trophic levels (algae
and zooplankton), phytoplankton biomass would show a minor increase with nutrient enrichment, but algal populations
would be largely regulated by grazers. Dominance of cladocerans in Study 1 contrasted results in a previous study at
CERC when fish were present and rotifers dominated the zooplankton community (Boyle et al. 1996).
In contrast, in Studies 2 and 3, zooplankton numbers were low early in the study in May and June in spite of early nutri-
ent dosing. Low numbers were likely due to the early flooding of the corrals prior to macrophyte emergence. Although
cladocerans, primarily C. reticulata, were dominant early in Studies 2 and 3, they were replaced in July and August by
high numbers of rotifers including Brachionis spp., Eudhlanis spp., H. mira, and Monostyola bulla. Larger cladocerans
and copepods generally exert greater grazing pressure on phytoplankton than smaller zooplankton species due to high
intake rates and a large range of particle sizes that they are able to ingest (Thorpe and Covich 1991). However, Jeppesen
et al. (1990) found that phytoplankton declines in shallow Lake Sobygard were associated with proliferation of rotifers
and/or cladocerans. In our Studies 2 and 3, rotifer abundance increased midsummer and persisted as the numerically
dominant division (60% of the cumulative zooplankton abundance), regardless of date or treatment. Rotifers may have
provided a steady grazing pressure on the algae and may have influenced the size, structure, and abundance of the
phytoplankton by regulating small algal species (Thorpe and Covich 1991). Nutrient dosing regime apparently had
little effect on zooplankton communities in our studies. Rather, differences in zooplankton community dynamics among
years (1999, Study 1 versus 2000, Studies 2 and 3) were more likely due to differences in the operational aspects of
corral construction and flooding chronology.
Although abundance in this study can be expressed volumetrically (number/m3) based on the 1 meter depth, it must be
noted that we sampled the zooplankton using behavioral traps set overnight as opposed to instantaneous tow samples
frequently used in limnological studies. Therefore, our zooplankton counts were based on numbers vertically migrating
overnight on an area basis (number/m2). Therefore, highly active zooplankton taxa are more likely to appear in samples
compared to those less prone to vertical night-time migration.
The average abundance of rotifers in our study (range 0.7-1.0*106/m2 across 3 studies) was comparable to the rotifer
abundance reported by Irvine et al. (1989) in other macrophyte-dominated mesocosms devoid of fish (average: 1.0*105/m2;
range: 0.3 *105/m2-1.7*105/m3). The seasonal average of cladoceran abundance (range 0.3-1.0*105/m2) was similar to
the average of 0.3*105/m2 reported by Irvine et al (1989). Maximum cladoceran abundance in the Control (1.1*106/m2;
Study 1) was also comparable to abundance (range: 0.4*106/m2-0.8*105/m2) measured using vertical migration sam-
plers in Lake Itasca in August by Williams (1983). Irvine et al. (1989) reported copepod abundance around 2.0*105/m2,
which was an order of magnitude greater than averages in our study (range 0.2-0.3*105/m2). Vertical migration in
copepods is well noted in Hutchinson (1967). Therefore, it is not known why preferential samples of vertical migrants
accounted for fewer copepods, but comparable numbers of rotifers and cladocerans, than tube samples used by Irvine
et al. (1989). Site and species variability may have been more influential in estimates of zooplankton abundance than
sampling techniques.
Sediments as Nutrient Sources and Sinks
Sediments are known to be the primary pool of nutrients in shallow aquatic systems (Johnston 1991; Barko et al. 1991).
The stock of nutrients in the sediments varies with consideration of depth, but in general is at least one or two orders
of magnitude greater than the macrophytes or water (Johnston 1991). Sediments in our study contained approximately
150 g N/m2 and 23 g P/m2 as measured in the nutrient pool in the upper 5 cm of sediment. Comparatively, macrophytes
113
-------
had less than 22 g N/m2 and 3.3 g P/m2 under maximum conditions dosage and biomass conditions. Water contained
less than 1 g N/m2 and 0.1 g P/m2 under Control conditions and up to 10 g N/m2 and 0.4 g P/m2 under highest dos-
age conditions; higher levels generally were due to high levels of dissolved nutrients due to dosing as opposed to true
steady-state conditions. Thus, sediments dominated as the primary source and sink of nutrients in these studies. We
did not observe an increase in total nutrient pools in sediments, however, due to the size of the nutrient pool and the
inherent error in our measurement technique.
Phosphorus amendments were effectively conserved in our experimental corrals because these systems had no outflow.
Sediments served as both a source and a sink for phosphorus in this study. For example, Control macrophytes accrued
up to 3.3 g P/m2 from sediments even though no external phosphorus was added. In contrast, phosphorus additions
to the corrals rapidly dissipated in as little as 7 days; the major portion of these additions were transferred to sediments
either directly (sorption or precipitation) or indirectly due to detrital transfer. Johnston (1991) reviewed the literature
regarding uptake and retention of P by natural wetlands and indicated that values averaged 0.34 g P/m2/yr (range
0.07-3.48 g P/m2/yr). Richardson and Qian (1999) evaluated a North American wetland database and determined that
the assimilative capacity of most wetlands for phosphorus is around 1 g P/m2/yr. Our studies support this estimate be-
cause at dosing levels of 0.86 g P/m2 (Study 1) and 1.2 g P/m2 (Study 2) phosphorus was effectively assimilated without
significant effects on community or nutrient dynamics. In our highest dose (2.4 g P/m2; Study 2), macrophyte biomass
decreased, which indicates that there is an upper limit to phosphorus assimilation by submerged macrophytes.
Nitrogen removal and transfer to sediment was also efficient. Effective doses of up to 60 g N/m2 were assimilated in our
studies. Mitsch et al. (1999) reviewed the literature regarding nitrogen assimilation in natural wetlands and indicated that
assimilation of up to 28 g N/m2/yr can occur; even higher assimilation can occur in engineered wetlands. In contrast with
phosphorus, nitrate is not necessarily conserved because it may be lost to the atmosphere as a gas via denitrification
processes (Seitzinger 1988). Johnston (1991) indicated that denitrification can be a major factor in loss of nitrogen in
wetland soils; denitrication losses can range from 0.002 - 0.34 g N/m2/yr (mean 0.19 g N/m2/yr) at unammended sites
and from 16-134 g N/m2/yr (mean 60 g N/m2/yr) in sites amended with inorganic nitrogen. Denitrification involves the
microbial transformation of nitrate to N2 under saturated anaerobic conditions (Scheffer 1998). Macrophytes can stimu-
late denitrification by providing a source of organic carbon to sediments; as organic matter decomposes, near-surface
layers of sediments become anoxic and stimulate denitrification processes (Weisner et al. 1994). If denitrification is
related to loading and water TN, as suggested by Jensen et al. (1992), nitrate additions may enhance loss rates. We
did not measure actual denitrification rates in our studies. Greater denitrification in relation to nitrate amendments in
our study may explain why N-stores in N-dosed treatments were not as high, compared to the Control, as their loading
would have predicted.
114
-------
Conclusions, Management Implications, and Research Needs
We determined that corral experiments, used to simulate shallow vegetated aquatic systems, were highly efficient in
removal of nitrogen (applied as nitrate) and phosphorus (applied as phosphate) at dose levels of up to 60 g N/m2and
2.4 g P/m2; nutrient uptake, assimilation, and retention were efficient regardless of magnitude of dose, timing of dose in
relation to macrophyte development, or frequency of dose. In one treatment (high early dose), we observed significant
reductions in macrophyte biomass; however, stands persisted throughout the study, and nutrient removal was efficient.
Total nutrient removal was over 90% as indicated by dissolved nutrients remaining at the end of the study.
Sediments served as the largest storage pool of nutrients, followed by macrophytes, phytoplankton, epiphytes, and
water (dissolved forms). Sediment was such a dominant factor in nutrient dynamics because it provided the primary
source of nutrient for macrophytes; this was dramatically illustrated by macrophyte production in Control treatments in
the absence of macrophyte dosing. Much of the applied nutrients were returned to the sediments by macrophyte se-
nescence, shedding of leaves, grazing of epiphytes, and zooplankton grazing. Precipitation of phosphorus was also a
likely factor based on observed decreases in hardness over time which was greater in dosed treatments. In this study,
we attempted to determine the rates of sediment deposition of nutrients using precipitation trays. However, these at-
tempts were unsuccessful due to the disturbance created in macrophyte sampling. Other ancillary observations support
the contention that grazing of epiphytes and shedding of macrophyte leaves were significant factors, including high
accumulations of fine particulate organic matter in zooplankton traps and observed particles settled on macrophyte
surfaces. Future studies should focus more closely on direct quantification of detrital nutrient transfer to sediments for
epiphytes and macrophytes. In addition, direct measurements of denitrification rates are needed.
Although the results of these studies demonstrated that shallow, vegetated aquatic systems are highly efficient in nutrient
uptake, assimilation, and removal, there may be some limitations in the direct application of these data. Our wetlands
were operated as "closed systems" which eliminated any losses of nutrients due to hydrologic discharge. In addition,
our systems were isolated from wind and wave action, common in natural systems, which can increase turbidity and
decrease availability to macrophytes. Wetlands in the actual environment experience variable inputs and outputs of
water and nutrients based on the season, frequency, and magnitude of rainfall events, size of the watershed, and wetland
dimensions. Therefore, dissolved nutrients and suspended particles can be lost from the system as "leakage" in overflow
or runoff. In addition, we focused primarily on the active growing season of the wetland cycle; the over-winter period
was not studied. Under aerobic conditions, phosphate forms insoluble complexes with iron which facilitate phosphorus
retention in sediments; however, under anaerobic conditions, phosphorus is released to interstitial waters and the wa-
ter column. During the over-winter period, biological activity is low and so dissolved phosphorus and nitrogen can be
leached from wetland systems during colder temperatures. In addition, newly constructed wetlands may lack the fine
organic sediments that promote macrophyte development and critical sediment processes. Each of these factors could
alter the efficiency of wetlands for nutrient removal. Therefore, additional studies on the nutrient removal efficiencies
of both constructed and natural wetlands are needed to determine the ultimate role of wetlands as management tools
for nutrient reduction streams and other receiving bodies.
115
-------
Literature Cited
American Public Health Association. 1992. Standard methods for the examination of water and wastewater. Amer. Pub.
Health Assoc. Inc., New York.
Bachman, R.W., C.A. Horsburgh, M.V. Hoyer, L.K. Mataraza, and D.E. Canfield Jr. 2002. Relations between trophic state
indicators and plant biomass in Florida lakes. Hydrobiologia 470: 219-234.
Balls, H., B. Moss, and K. Irvine. 1989. The loss of submerged plants with eutrophication: experimental design, water
chemistry, aquatic plant and phytoplankton biomass in experiments carried out in ponds in the Norfolk Broadland.
Fresh. Biol. 22:71-87.
Barko, J.W., D. Gunnison, and S.R. Carpenter. 1991. Sediment interactions with submersed macrophyte growth and
community dynamics. Aquat. Bot. 41:41-65.
Blindow, I. 1992. Decline of charophytes during eutrophication: comparison with angiosperms. Freshwat. Biol. 28:9-
14.
Box, R.J. 1986. Quantitative short-term uptake of inorganic phosphate by the Chara hispida rhizoid. Plant, Cell and
Environ. 9:501-506.
Box, R.J. 1987. The uptake of nitrate and ammonium nitrogen in Chara hispida L: the contribution of the rhizoid. Plant,
Cell, and Environ. 10:169-176.
Boyle, T.P., J.F. Fairchild, E. F. Robinson-Wilson, P.S. Haverland, and J.A. Lebo. 1996. Ecological restructuring in experi-
mental aquatic mesocosms due to the application of diflubenzuron. Environ. Tox. Chem. 15:1806-1814.
Boyd, C.E. 1970. Chemical analyses of some vascular aquatic plants. Arch. Hydrobiol. 67(1):78-85.
Brock, T.C.M., R.M.M. Roijackers, R. Rollon, F. Bransen, and L. Van der Heyden. 1995. Effects of nutrient loading and
insecticide application on the ecology of Elodea-dominated freshwater microcosms II: responses of macrophytes,
periphyton, and macroinvertebrate grazers. Arch. Hydrobiol. 134:53-74.
Bronmark, C. and L. Hansson. 1998. The biology of lakes and ponds. Oxford University Press, Oxford.
Brooks, J.L. and S.I. Dodson. 1965. Predation, body size and composition of the plankton. Science 150:28-35.
Carlson, R.E. 1977. A trophic state index for lakes. Limnol. Ocean. 22:361-369.
Carpenter, S.R., N.F. Caraco, D.L. Correll, R.W. Howarth, A.N. Sharpley, and V.H. Smith. 1998. Nonpoint pollution of
surface waters with phosphorus and nitrogen. Ecol. Appl. 8(3):559-568.
Cattaneo, A., and J. Kalff. 1980. The relative contribution of aquatic macrophytes and their epiphytes to the production
of macrophyte beds. Limnol. Oceanogr. 25(2):280-289.
Conover, W.I., and R.L. Iman. 1981. Rank transformations as a bridge between parametric and nonparametric statistics.
Amer. Statistician 35:124-128.
Crawford, S.A. 1977. Chemical, physical, and biological changes associated with Chara succession in farm ponds.
Hydrobiol. 55(3):209-217.
Crumpton, W.G., T.M. Isenhart, and P.O. Mitchell. 1992. Nitrate and organic N analyses with second derivative spectros-
copy. Limnol. Oceanogr. 37:907-913.
de Haan, H., L. Van Liere, S.R Klapwijk, and E. Van Donk. 1993. The structure and function of fen lakes in relation to
water table management in The Netherlands. Hydrobiol. 265:155-177.
Dykyjova, D. and J. Kvet, 1982. Mineral nutrient economy in wetlands of theTrebon Basin Biosphere Reserve, Czecho-
slovakia, in B. Gopal, E.R.Turner, R.G Wetzel, D.F. Whigham, eds. Proceedings of the Interational Conference of
Wetlands: Ecology and Management. International Scientific Publications, Jaipur, India.
Engel, S. 1990. Ecosystem responses to growth and control of submerged macrophytes: a literature review. Technical
Bulletin No. 170. Dept. Natural Resources., Madison, Wl.
117
-------
Fail-child, J.F., T.W. La Point, J. Zajicek, M.K. Nelson, F.J. Dwyer. and RA. Lovely. 1992. Population, community, and
ecosystem-level responses of mesocosms to pulsed doses of a pyrethroid insecticide. Environ. Toxicol Chem
11:115-130.
Fairchild, J.F., T.W. La Point, and T. Schwartz. 1994. Effects of an herbicide and insecticide mixture in experimental
aquatic mesocosms. Arch. Env. Tox. Chem. 27:527-533.
Fairchild, J.F. and L.S. Sappington. 2002. Fate and effects of the triazinone herbicide metribuzin in outdoor aquatic
mesocosms. Archives. Environ. Toxicol. Chem. 43(21 ):198-2002.
Forsberg, C. 1964. Phosphorus, a maximum factor in the growth of Characeae. Nature 201:517-518.
Gasith, A. and M.V. Hoyer. 1998. Structuring role of macrophytes in lakes: changing influence along lake size and depth
gradients, in Jeppesen, E., M. Sondergaard, M. Sondergaarcl, and K. Christoffersen, eds. The structuring role of
submerged macrophytes in lakes. Springer, New York.
Green, R.H. 1979. Sampling Design and Statistical Methods for Environmental Biologists. Wiley and Sons, New York
NY.
Gross, E.M., D. Erhard, and E. Ivanyi. 2003. Allelopathic activity of Ceratophyllum demersum L. and Najas marina ssp
intermedia (Wolfgang) Casper. Hydrobiologia 509: 583-589.
Haekanson, L., and V.V. Boulion. 2002. Empirical and dynamical models to predict the cover, biomass, and production
of macrophytes in lakes. Ecol. Model. 151:213-243.
Hansson, L. 1992. The role of food chain decomposition and nutrient availability in shaping algal biomass development.
Ecology 73: 241 -247.
Hauck, H.S., L.G. Huber, and C.D. Nagal. 1997. Water Resources Data Missouri Water Year 1996. U.S. Geological Survey
Water Data Report MO-96-1. Rolla, MO. 292 p.
Havens, K.E., T.L. East, A.J. Rodusky, and B. Sharfstein. 1999. Littoral periphyton responses to nitrogen and phosphorus:
an experimental study in a subtropical lake. Aquat. Bot. 63:267-290.
Horppila, J., and L. Nurminen. 2003. Effects of submerged macrophytes on sediment resuspension and internal phos-
phorus loading in Lake Hiidenvesi (southern Finland). Water Res. 37:4468-4474.
Hutchinson, G.E. 1967. A treatise on limnology. Volume II: Introduction to lake biology and the limnoplankton. Wiley,
New York.
Irvine, K., B. Moss, and H. Balls. 1989. The loss of submerged plants with eutrophication, II. Relationships between fish
and zooplankton in a set of experimental ponds, and conclusions. Freshwat. Biol. 22:89-107.
James, M.R., I. Hawes, and M. Weatherhead. 2000. Removal of settled sediments and periphyton from macrophytes by
grazing invertebrates in the littoral zone of a large, oligotrophic lake. Fresh. Biol. 44:311-326.
Jansson, M., R. Andersson, H. Berggren, and L. Leonardson. 1994. Wetlands and lakes as nitrogen traps. Ambio
23(6):320-325.
Jensen, J.P., E. Jeppesen, P. Kristensen, RB. Christensen, and M. Sondergaard. 1992. Nitrogen loss and denitrification
as studied in relation to reductions in nitrogen loading in a shallow, hypertrophic lake (Lake Sobygard, Denmark).
Int. Revue ges. Hydrobiol. 77(1):29-42.
Jensen, J.P., E. Jeppesen, K. Olrik, and P. Kristensen. 1994. Impact of nutrients and physical factors on the shift from
cyanobacterial to chlorophyte dominance in shallow Danish lakes. Can. J. Fish. Aq. Sci. 51:1692-1699.
Jeppesen, E., M. Sondergaard, O. Sortkjaer, E. Mortensen, and P. Kristensen. 1990. Interactions between phytoplank-
ton, zooplankton and fish in a shallow, hypertrophic lake: a study of phytoplankton collapses in Lake Sobygard,
Denmark. Hydrobiol. 191:149-164.
Johnston, C.A. 1991. Sediment and nutrient retention by freshwater wetlands: effects on surface water quality. Crit.
Rev. in Env. Control. 21(5,6):491-565.
Koerner, S., and A. Nicklisch. 2002. Allelopathic growth inhibition of selected phytoplankton species by submerged
macrophytes. J. Phycol. 38:862-871.
Kogan, S.J., and G.A. Chinnova. 1972. Relations between Ceratophylum demersum and some blue-green algae. Hy-
drobiol. J. [Gidrobiol. Zh.] 8:14-19.
Kufel, L., and I. Kufel. 2002. Chara beds acting as nutrient sinks in shallow lakes--a review. Aquat. Bot. 72:249-260.
Lachat Instruments, Inc. 1997. Determination of Ammonia by Flow Injection Analysis Colorimetry. QuikChem Method
10-107-06-1-A. Milwaukee, Wl.
Lachat Instruments, Inc. 1997. Determination of Nitrate/Nitrite in Surface and Wastewaters by Flow Injection Analysis.
QuikChem Method 10-107-04-1-B. Milwaukee, Wl.
118
-------
Lachat Instruments, Inc. 2000. Orthophosphate in Waters. QuikChem Method 10-115-01-1-A. Milwaukee, Wl.
Landers, D.H. 1982. Effects of naturally senescing aquatic macrophytes on nutrient chemistry and chlorophyll a of sur-
rounding waters. Limnol. Oceanogr. 27(3):428-439.
Lijklema, L. 1994. Nutrient dynamics in shallow lakes: effects of changes in loading and role of sediment-water interac-
tions. Hydrobiol. 275/276:335-348.
Lind, O.T. 1985. Handbook of common methods in limnology. Kendall/Hunt, Dubuque, IA.
Madsen, J. D. 1993. Biomass techniques for monitoring and assessing control of aquatic vegetation. Lake and Reserv.
Manage. 7(2):141 -154.
Mayer, J., M. T. Dokulil, M. Salbrechter, M. Berger, T. Posch, G. Pfister, A.K.T. Kirschner, B. Velimirov, A. Steitz, and T.
Ulbricht. 1997. Seasonal successions and trophic relations between phytoplankton, zooplankton, ciliate and bacteria
in a hypertrophic shallow lake in Vienna, Austria. Hydrobiol. 342/343:165-174.
Mitchell, S.F., D.P. Hamilton, W.S. MacGibbon, P.K.B. Nayar, and R.N. Reynolds. 1988. Interrelations between phyto-
plankton, submerged macrophytes, black swans (Cygnus atratus) and zooplankton in a shallow New Zealand lake.
Int. Revue ges. Hydrobiol. 73(2):145-170.
Mitsch, W.J., J.W. Day, Jr., J.W. Gilliam, P.M. Groffman, D.L. Hey, G.W. Randall, and N.Wang. 1999. Reducing nutrient
loads, especially nitrate-nitrogen, to surface water, ground water, and the Gulf of Mexico. Topic 5 Report for the
Integrated Assessment on Hypoxia in the Gulf of Mexico, NOAA Coastal Ocean Program Decision Analysis Series
No. 19.111 p.
Mjelde, M. and B.A. Faafeng. 1997. Ceratophyllum demersum hampers phytoplankton development in some small Norwe-
gian lakes over a wide range of phosphorus concentrations and geographical latitude. Freshw. Biol. 37:355-365.
Moeller, R.E., J.M. Burkholder, and R.G. Wetzel. 1988. Significance of sedimentary phosphorus to a rooted submersed
macrophyte (Najas flexilis (Willd.) Rostk. and Schmidt) and its algal epiphytes. Aquat. Bot. 32:261-281.
Moss, B. 1976. The effects of fertilization and fish on community structure and biomass of aquatic macrophytes and
epiphytic algal populations: an ecosystem experiment. J. Ecol. 64:313-342.
Moss, B. 1990. Engineering and biological approaches to the restoration from eutrophication of shallow lakes in which
aquatic plant communities are important components. Hydrobiol. 200/201 -.367-377.
Moss, B. 1995. The microwaterscape: a 4-dimensional view of interactions among water chemistry, phytoplankton,
penphyton, macrophytes, animals, and ourselves. Water Sci. Technol. 32:105-116.
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. Hydrobiol. 48(2):109-116.
Otsuki, A., and R.G. Wetzel. 1972. Coprecipitation of phosphate with carbonates in a marl lake. Limnol. Oceanogr.
17:763-767.
Ozimek, T, R.D. Gulati, and E. van Donk. 1990. Can macrophytes be useful in biomanipulation of lakes? The Lake
Zwemlust example. Hydrobiol. 200/201:399-407.
Phillips, G.L., D. Eminson, and B. Moss. 1978. A mechanism to account for macrophyte decline in progressively eutro-
phicated freshwaters. Aquat. Bot. 4:103-126.
Redfield, A.C., B.H. Ketchum, and F.A. Richards. 1963. The influence of organisms on the composition of seawater, in
Hill, M.N. (ed.), The Sea, Vol. 2, Wiley Interscience, New York, NY, p. 26-77.
Richardson, C.J. and S.S. Qian. 1999. Long-term phosphorus assimilative capacity in freshwater wetlands: A new para-
digm for sustaining ecosystem structure and function. Environ. Sci. and Technol. 33(10):1545-1551.
Royle, R.N. and R.J. King. 1991. Aquatic macrophytes in Lake Lindell, New South Wales: biomass, nitrogen and phos-
phorus status, and changing distribution from 1981 to 1987. Aquat. Bot. 41(4):281-298.
Sakamoto, M. 1966. Primary production by phytoplankton community in some Japanese lakes and its dependence on
lake depth. Arch. Hydrobiol. 62:1-28.
Sand-Jensen, K. and J. Borum. 1984. Epiphyte shading and its effect on photosynthesis and diel metabolism of Lobelia
dortmanna during the spring bloom in a Danish lake. Aquat. Bot. 20:109-120.
Sand-Jensen, K. and J. Borum. 1991. Interactions among phytoplankton, periphyton, and macrophytes in temperate
freshwaters and estuaries. Aquat. Bot. 41:137-175.
SAS Institute, Inc. 1996. Statistical Analysis System, Release 6.12., SAS Institute, Inc., Gary, N.C.
Scheffer, M. 1990. Multiplicity of stable states in freshwater systems. Hydrobiol. 200/201:475-486.
Scheffer, M. 1998. Ecology of shallow lakes. Chapman & Hall, London. 357 p.
119
-------
Seitzinger, S.P. 1988. Denitrification in freshwater and marine coastal ecosystems: ecological and geochemical signifi-
cance. Limnol. Oceanogr. 33:702-724.
Smith, G.M. 1950. The freshwater algae of the United States. McGraw-Hill, New York.
Smith, V. 2003. Eutrophication of freshwater and coastal marine ecosystems: a global problem. Environ. Sci. Res. Int.
10:126-139.
Snedecor, G.W. and W.G. Cochran. 1967. Statistical Methods. University of Iowa Press, Ames, IA.
Sommers, I.E., and D.W. Nelson. 1972. Determination of total phosphorus in soils: a rapid perchloric acid digestion
procedure. Soil Sci. Soc. Amer. Proc. 36:902-904.
Sondergaard, M., E. Jeppesen, P. Kristensen, and O. Sortkjaer. 1990. Interactions between sediment and water in a shallow
and hypertrophic lake: a study on phytoplankton collapses in Lake Sobygard, Denmark. Hydrobiol. 191:139-148.
Sondergaard, M. and B. Moss. 1998. Impact of submerged macrophytes on phytoplankton in shallow freshwater lakes,
in Jeppesen, E., M. Sondergaard, M. Sondergaard, and K.Christoffersen, eds. The structuring role of submerged
macrophytes in lakes. Springer, New York.
Stachowicz, K., M. Czernoch, and E. Dubiel. 1994. Field pond as a sink for nutrients migrating from agrocenoses to
freshwaters. Aquat. Sci. 56(4):363-375.
Stelzer, R.S., and G.A. Lamberti. 2002. Ecological stoichiometry in running waters: periphyton chemical composition
and snail growth. Ecology 83:1039-1051.
Thorpe, J.H. and A.R Covich. 1991. Ecology and classification of North American freshwater invertebrates. Academic
Press, San Diego.
U.S. Environmental Protection Agency (USEPA). 1979. Methods for chemical analysis of water and wastes. EPA-600/4-
79-020. U.S. Environmental Protection Agency, Washington, D.C. 430 p.
van Donk, E., R.D. Gulati, A. ledema, and J.T. Meulemans. 1993. Macrophyte-related shifts in the nitrogen and phos-
phorus contents of the different trophic levels in a biomanipulated shallow lake. Hydrobiol. 251:19-26.
van Donk, E., and W.J. van de Bund. 2002. Impact of submerged macrophytes including charophytes on phyto-and
zooplankton communities: allelopathy verus other mechanisms. Aquat. Bot. 72:261-274.
Vitousek, P.M., J.D. Aber, R.W. Howarth, G.E. Likens, P.A. Matson, D.W. Schindler, W.H. Schlesinger, and D.G.Tilman.
1997. Human alteration of the global nitrogen cycle: sources and consequences. Ecol. Appl. 7(3):737-750.
Weisner, S.E.B., RG. Eriksson, W. Graneli, and L. Leonardson. 1994. Influence of macrophytes on nitrate removal in
wetlands. Ambio 23(6):363-366.
Wetzel, R.G. 1983. Limnology (2nd ed.). Saunders College Publishing, Philadelphia, PA.
Whiteside, M.C., J.B.Williams, and C.R White. 1978. Seasonal abundance and pattern of Chydorid Cladocera in mud
and vegetative habitats. Ecol. 59(6):1177-1188.
Williams, J.B. 1983. A study of summer mortality factors for natural populations of Chydoridae (Cladocera). Hydrobiol.
107:131-139.
Wood, R.D. 1950. Stability and zonation of Characeae. Ecol. 31 (4):642-647.
120
-------
Appendix 1: Experimental Error Associated with Physical, Chemical, and
Biological Variables in Shallow, Vegetated, Outdoor Experimental Corrals
Abstract
in years 1999 and 2000 the U.S. Geological Survey was funded by the U.S. Environmental Protection Agency to conduct
a series of experiments to evaluate the fate and effects of nutrients in simulated shallow, vegetated wetlands. These
studies were conducted in outdoor experimental corral systems located at the Columbia Environmental Research Center
in Columbia, MO. Results of these studies were published in a final report by Fairchild and Vradenburg (2004). Data
from these studies were not normally distributed and exhibited heterogeneity of variance. Therefore, the data were
statistically analyzed using non-parametric statistics. Although data means by treatment were presented, the actual
experimental error rates, defined as the coefficient of variation (COV), were not reported. This study re-evaluates the
experimental error rates of the control treatments from the 1999 and 2000 studies to allow a relative comparison of the
variability of numerous physical, chemical, and biological variables across treatments. Results indicated that physi-
cal parameters, including pH and temperature exhibited low experimental error rates (COV range <10%). Chemical
variables ranged widely in experimental error rates. For example, alkalinity, hardness, and conductivity ranged from
2 - 25% COV. Total nutrients in water (total nitrogen and total phosphorus) ranged from 0 - 55% COV. Dissolved nu-
trients, including ammonia, nitrate-nitrite, and soluble-reactive phosphorus, ranged from 5 - 200% COV because they
frequently were low and near detection limits. Nutrient content (total nitrogen and total phosphorus) of macrophytes
and sediments exhibited relatively low experimental error rates and ranged from 5 - 55% due to the inherently large,
static pools of nutrients. Highest rates of experimental error were associated with biological variables such as mac-
rophyte biomass (28 - 96% COV), periphyton biomass (20 - 110% COV), chlorophyll a (i.e., algal biomass; 20 - 125%
COV), phytoplankton community structure (50 - 450% COV), and zooplankton community structure (30 - 200% COV).
Relative rates of experimental error of physical, chemical, and biological variables varied with season. For example,
error rates of physical parameters were lowest early in the season since these were driven by characteristics of the
well water source. In contrast, experimental error rates of biological variables were highest early in the season when
numbers were low, and species were rapidly colonizing. Highest average rates of experimental error were also associ-
ated with variables influenced by biological processes (e.g., nutrient supply, zooplankton grazing, etc.) and seasonal
species succession controlled by nutrient supply and temperature constraints. An understanding of the ranges and
sources of experimental error can be valuable in planning experiments where statistical power to differentiate among
treatments is desired. However, increased statistical power often requires large increases in statistical replication which
may have tradeoffs in terms of the number of treatment effects that may be studied. Ultimately, design of such stud-
ies must be driven by the goals and objectives of a study as constrained by the experimental error rates that occur in
complex ecological test systems.
Introduction
The U.S. Environmental Protection Agency (USEPA) is the primary federal agency whose mission is to conduct research
to support environmental regulation in the United States. Research sponsored by the USEPA is conducted across a
vast array of experimental test systems ranging from standardized in-vivo assays, conducted at the cellular or enzyme
level, to landscape level analysis reflecting broad spatial and temporal scales. Each of these test systems has its own
intent and merit.
In vitro cellular assays are intended to be highly replicable (e.g., across laboratories) and repeatable (i.e., over time
within a laboratory). Such assays allow rapid through-put of large numbers of samples. For example, an in-vivo cel-
lular assay may allow one to determine the response to a given chemical stressor. However, in many cases, these tests
may only reflect a relative response of a cell or tissue to one factor, with no potential homeostasis or inherent capacity
to recover from an insult.
In contrast, large scale biological experiments, conducted at the landscape level, allow studies of broad temporal and
spatial scales to examine environmental changes. For example, long-term ecological experiments can be conducted
to determine the effects of climate change on forest productivity. Such studies, while having perhaps the ultimate envi-
ronmental relevance, may take years to decades to determine a trajectory or response. Such experiments are difficult
to repeat or replicate due to inherent differences across landscapes and stochastic error.
Other experimental approaches fall in the middle of the continuum of environmental realism and statistical confidence.
For example, mesocosm experiments allow a researcher to study the responses of a simulated ecosystem or community
to an environmental stressor or manipulation. Examples of mesocosms include experimental ponds, streams, field plots,
or other physical models of natural ecosystems. Mesocosms can be cost-effective approaches for ecological studies
that allow a researcher to replicate treatments and derive statistical inference regarding system responses.
121
-------
The USEPA, in 1999 and 2000, funded the U.S. Geological Survey under IAG DW14938559-01 to conduct mesocosm
experiments to determine the assimilative capacity of simulated wetlands for nutrients. The results of these studies were
reported to the USEPA in a final report entitled "Fate and Effects of Nitrogen and Phosphorus in Shallow, Vegetated
Aquatic Ecosystems" (Fairchild and Vradenburg, 2004). These studies examined the response of aquatic mesocosms
to several factors: 1) nitrate loading rates; 2) nitrogen to phosphorus (N:P) ratios; 3) frequency of dosing/application;
and 4) timing of dose initiation.
Analysis of the data in these studies indicated that in most cases the data were not normally distributed and exhibited
non-homogeneous variance. In spite of several statistical attempts to normalize the data, the data did not fit the as-
sumptions required for Analysis of Variance (ANOVA). Therefore, the data were analyzed using ANOVA of ranked data
as suggested by numerous statistical references (Snedecor and Cochran, 1967; Green, 1979). ANOVA of ranked data
allows a researcher to evaluate the differences among experimental treatments using a robust, unbiased approach.
However, there are tradeoffs in some cases in terms of estimates of statistical power or the true degree of differences
among treatments because the actual data are analyzed in terms of relative rank differences only.
Even though the use of ranked data in ANOVA is widely accepted and commonly done, there is still some utility in
evaluating the experimental error of variables in a dataset. Calculations of experimental error can be useful in designing
future experiments. For example, a researcher may wish to determine the number of replicates needed, for example,
to detect a 50% difference among variables (Snedecor and Cochran, 1967; Ellersieck and LaPoint, 1995). One might
also be interested in evaluating the relative experimental error among different response variables in order to choose
those that respond to experimental manipulations yet vary little in time and space due to inherent and external forces.
Objective
Herein, we analyze the results of Fairchild and Vradenburg (2004) to determine the experimental error of various physical,
chemical, and biological variables. This analysis is done to allow comparison of this mesocosm data to other datasets
derived from other experimental test systems used in research sponsored by the USEPA. Actual comparisons to other
data are not conducted in this analysis. Rather, these data are provided to researchers and quality assurance personnel
of the USEPA for use in experimental design of future experiments, and as an objective dataset for use in evaluating
the cost effectiveness, utility, and value of various test systems currently used in ecological research. As previously
described, the choice of the test system is driven by multiple factors which range from ecological relevance to other
objectives such as high statistical precision and accuracy. There are tradeoffs associated with each choice.
Methods
Data for this analysis were derived from Fairchild and Vradenburg (2004). Experimental error was defined as the coef-
ficient of variation (COV), which is calculated as follows:
COV = STD/Mean x 100, where
STD - standard deviation, and
Mean = arithmetic average of the data
The COV is the proportion of variation of a variable reflected as the average value of the observations. Ideally, the
lower the COV the fewer replicates are needed to determine a desired difference in an experiment. For this analysis,
we calculated the COV using the grand mean, in addition to the mean within a given date for each variable. Note that
a different set of ponds were used in year 2000 compared to year 1999 to minimize effects due to the previous year's
experimental manipulations. Differences across years can be due to inherent pond differences in addition to annual
differences in the timing of flooding, temperature, etc. Therefore, data were calculated separately for each year of the
study. Only the control data are used for this analysis, since they represent the true variability of a metric in the absence
of the experimental manipulation. All statistics were calculated using the Statistical Analysis System (SAS, 2000).
Results and Discussion
Raw data summaries (grand mean, mean, standard deviation, and COV by week) for 1999 are presented in Tables
A1-A6. Raw data summaries (grand mean, mean, standard deviation, and COV by week) for 2000 are presented in
Tables A7-A11. The data (years 1999 and 2000) are combined and plotted for visualization of trends in the COV over
time for each variable (Figures A-1-A-43).
Water quality
The COV for algal biomass, measured as chlorophyll a content of water, ranged from 24 -128% across the two years of
study. Chlorophyll varied much more in year 2000 than in 1999, and appeared to cycle in intensity over 30-day intervals.
Chlorophyll data fluctuated seasonally in 1999, but to a lesser extent than in year 2000 (Figure A-1). The COVs for
periphyton biomass, represented as chlorophyll a content associated with the surface of Scrimweavetm strips, ranged
from 21 -110% over the two years. Relative differences in the COV for periphyton varied within a year, but the ranges
were similar across years 1999 and 2000.
122
-------
The COVs for particulate nutrients (TN and TP) are presented in Figures A-3 and A-4. The COV for TN ranged from
0 -41% (Figure A-3). Higher temporal variation occurred in year 2000 (COV range 0 - 41%) than in year 1999 (COV range
13 - 27%). The COV in TP values ranged from 14-52 % in Year 1999 and from 11 - 54% in year 2000 (Figure A-4).
The COVs for dissolved nutrients, including ammonia (NH3), nitrate-nitrite (NO2NO3), and soluble reactive phospho-
rus (SRP) are presented in Figures A-5-A-7. The COVs for NH3 ranged from 5 - 200% within a year but were similar
in ranges across years (Figure A-5). The COVs for NO2NO3 also ranged from 5 - 200% in a manner similar to NH3
(Figure A-6). The COVs for SRP ranged from 3 - 63% and were considerably lower than those for forms of dissolved
nitrogen (Figure A-7). The high COVs for dissolved nutrients were in part due to the extremely low concentrations in
the control corrals which were near the limits of detection in most cases (Tables A-1 and A-7). In contrast to the control
corrals, dosed corrals typically (data not shown) exhibited COV values of dissolved nutrients in the range of 10 - 20%
on dosing days due to the high levels of nutrients added to those corrals. The COVs in treated corrals increased with
time after nutrient addition, however, as dissolved nutrients were being rapidly being assimilated by plants.
The COVs of physical/chemical parameters of pH, alkalinity, hardness, turbidity, conductivity, and temperature ranged
from 0 - 30% and were much more uniform across time and years compared to the chlorophyll/nutrient data. This de-
creased variation is largely due to the highly buffered water quality conditions derived from the well water source used
to fill the experimental corrals. Routinely, the COVs for these variables were lowest early in the study soon after corral
flooding and prior to divergence due to biological, chemical, and physical influences. The COVs for pH were similar
across years and varied little across the control corrals (COV 0 - 5%) (Figure A-8). Highest COVs occurred in late June
in both years when primary productivity was generally highest; during this period, carbon dioxide is highly assimilated
leading to an increase in both magnitude and variation of pH. The COV values for alkalinity, hardness, and conductivity
were uniform (within and across years) and generally ranged from 2 - 20% (Figures A-9, A-10, and A-11). Although
absolute values of these parameters changed over time (Fairchild and Vradenburg, 2004) due to internal processes,
the actual level of variation among control replicates did not change much over time. These variables are much less
variable since they are driven by physical/chemical limnological conditions (e.g., precipitation, dissolution) as opposed
to biological interactions (e.g., photosynthesis, grazing, etc.). The COVs for turbidity, in contrast, fluctuated both within
and across years and ranged from 0 - 89%. Although turbidity values themselves were less than 10 NTU's, the COVs
can vary due to many factors including algal turbidity, physical disturbance due to sampling, and in some cases bubbling
of gaseous releases from sediments. Therefore, variation in turbidity is caused by numerous biological and chemical
factors.
The COVs for temperature, measured at dusk (Figure A-13) and dawn (Figure A-14) varied less than any other variable
in these studies and ranged from 1 - 7% across experimental replicates. Although absolute measures of temperature
range considerably in the diurnal and annual cycle, the COV does not because of the large latent thermal mass of the
experimental systems. Since the systems are constructed to uniform standards (depth, circumference, and volume),
temperature variations are damped across replicates.
Macrophytes
The COV of macrophyte weights (macrophyte biomass) was two-fold higher in year 2000 (62 - 97%) compared to year
1999 (28-58%) (Figure A-15). Differences across years may have varied due to differences in the draw-down/flooding
regime as well as inherent pond differences since different ponds were used each year. For both years, the COV of
macrophyte biomass decreased over the season as macrophyte biomass increased and stand patchiness decreased.
The COVs for nutrient content of aquatic macrophytes ranged from 6 - 30 % forTN (Figure A-16) and from 6 - 53% for
TP (Figure A-17). The COVs for nutrients in macrophytes decreased seasonally for both constituents. It is not clear
why such a dramatic trend in nutrient content occurred over time; however, it is likely due to changes in growth status
with higher variation occurring early in the season prior to stand maturity.
Sediments
The COV of TN in sediments ranged from 12 - 32% across years (Figure A-18) and did not vary appreciably across
years. There was also no appreciable seasonal trend in the pattern of COV forTN in sediments. The COV forTP in
sediments ranged from 8 - 18% (Figure A-19); there was likewise no seasonal trend in the COV forTP in sediments.
The relative lack of variation in nutrient content of sediments was due to two factors: 1) the large, relatively static pool
of nutrient stocks and 2) the use of composite sampling which reduced error.
Zooplankton
The COV for total zooplankton densities ranged from 21 - 94% (Figure A-20). Highest variation occurred in year 2000
at the beginning of the study. Such high variation is likely due to the rapid rate of flooding that occurred prior to mac-
rophyte development. Variation in zooplankton numbers decreased over time as the systems matured in year 2000.
In contrast, in year 1999 the highest variation in total zooplankton density was observed at the end of the year, most
probably due to macrophyte senescence and altered physical habitat conditions. The COV for total zooplankton species
123
-------
richness ranged from 4 - 23% across and within years (Figure A-21); this variation was much lower than that observed
for total zooplankton numbers. The COV for Simpson's dominance ranged from 9 - 77% over the two years of study
(Figure A-22).
The COV for density of cladocerans ranged from 12 - 28%, and was similar between the two years (Figure A-23). Av-
erage COV for cladoceran species richness ranged from 12 - 30% between the two years, and fluctuated little during
the year (Figure A-24). Similarly, Simpson's dominance of the cladoceran community ranged from 18 -50% across and
within years (Figure A-25). Average variation among the cladoceran community parameters varied less than those
calculated based on the combined, total zooplankton community which is ecologically significant since the cladocerans
generally contribute the greatest to overall zooplankton grazing pressure in aquatic systems (Brooks and Dodson, 1965;
Wetzel, 1983).
The COV for copepod numbers ranged from 6 - 278% across years (Figure A-26). The COV for copepod species rich-
ness ranged from 23 - 141% over two years (Figure A-27). The COV for Simpson's dominance of the copepod com-
munity ranged from 9 - 130% (Figure A-28). Highest COV for copepod parameters, as previously noted, occurred on
the first sampling date in year 2000 due to the early flooding effects on zooplankton variability.
The COVs for rotifer numbers were intermediate between those of cladocerans and copepods and ranged from 50 -175%
within a given year but were relatively similar across years (Figure A-29). The COVs for rotifer species richness var-
ied considerably across years and ranged from 0 - 100% in year 1999 but only ranged from 16 - 40% in year 2000
(Figure A-30). The COVs for Simpson's dominance of the rotifer community were similar to that for species richness
both among and across years, with higher variation noted in year 1999 compared to year 2000 (Figure A-31).
Phytoplankton
The COVs for parameters associated with the phytoplankton community were some of the highest observed for any
biological, physical, or chemical variable observed (Figures 32 • 43). For example, the COV for total phytoplankton
numbers ranged from 195 - 460% in year 1999 but ranged from 80 - 100% in year 2000 (Figure A-32). The COVs for
phytoplankton species richness, however, were lower than for total numbers and ranged from 10 - 88% (Figure A-33).
Similar levels of variation were noted for total numbers and species richness of major groups of phytoplankton (e.g.,
Bacillariophyta, Chlorophyta, Cryptophyta, Cyanophyta, and Euglenophyta) (Figures 34-43). The high levels of varia-
tion in the phytoplankton community parameters across years are commonly observed due to many factors including
changes in temperature, nutrient supply, and intensity of zooplankton grazing pressure (Wetzel, 1983) or simply differ-
ences due to inherent pond differences across years.
Conclusions
Coefficients of variation (COVs) ranged from 1 - 450% in outdoor aquatic experimental corrals used by studies in
Fairchild and Vradenburg (2004). Lowest COVs were associated with physical parameters such as temperature and pH
(range 1 - 8%). Some chemical parameters, such as alkalinity, hardness, and conductivity exhibited low COVs (range
1 - 26%) because they were not strongly influenced by biological interactions. Other parameters, such as nutrient
concentrations, varied widely in COVs. Total nutrients (e.g., TN and TP) in water, macrophytes, and sediment COVs
ranged from 8 - 55%. Dissolved nutrients varied much more across replicates (range 3 - 200% COV) because they
were frequently low, near detection limits and were intimately tied to biological processes such as plant uptake and
decomposition. Highest levels of variation occurred in some biological endpoints such as zooplankton, phytoplankton,
and macrophyte community structure. The COV for chlorophyll a, used to estimate phytoplankton biomass, ranged from
22 -130% due to differences in nutrient supply and zooplankton grazing pressure. The COVs for macrophyte biomass
(28 - 96%) and periphyton biomass (20 - 110%) reflected spatial and temporal variation in response to shading and
other factors. Although common groups of zooplankton (e.g., cladoceran species richness; COV range 12 - 30%) and
phytoplankton (e.g., chlorophyte species richness; COV range 10 - 120%) exhibited moderate levels of variation, rare
community groups were much more variable, with COVs commonly exceeding 200%. An understanding of these levels
of variance can be used to design experiments based on anticipated statistical (e.g., precision) and ecological (e.g.,
relevance) objectives.
References
Brooks, J.L. and S.I. Dodson. 1965. Predation, body size and composition of the plankton. Science 150:28-35.
Ellersieck, M.R. and T.W. LaPoint. 1995. Statistical analysis. In Rand, G., ed., Fundamentals of aquatic toxicology: effects,
environmental fate, and risk assessment. Taylor and Francis Publishers, Washington, D.C. p. 307-344.
Fairchild, J.F. and L. Vradenburg. 2004. Fate and effects of nitrogen and phosphorus in shallow, vegetated aquatic eco-
systems. Final Report to the U.S. Environmental Protection Agency, IAG DW14938559-01 Sept. 30, 2004.187 p.
Green, R.H. 1979. Sampling design and statistical methods for environmental biologists. Wiley and Sons, New York,
N.Y.
124
-------
Snedecor, G.W. and W.G Cochran. 1967. Statistical methods. University of Iowa Press, Ames, IA.
Statistical Analysis System Institute (SAS). 2000. SAS/STAT Guide for personal computers, Version 6, 4th ed., Volumes
1 & 2. SAS Institute Inc., Gary, N.C. 1685 p.
Wetzel, R.G. 1983. Limnology (2nd ed.). Saunders College Publishing, Philadelphia, PA.
125
-------
140
Figure A-1.
Coefficient of variation (%) of chlorophyll a in water (ug/L) over time for two-yr study.
120
-
in in
Date
Figure A-2.
Coefficient of chlorophyll a in periphyton (pg/crn2) over time for two-yr study.
25 if)
Date
Figure A-3. Coefficient of variation (%) of total nitrogen (mg/L) over time for two-yr study.
126
-------
Figure A-4.
Figure A-5.
Date
Coefficient of variation (%) of total phosphorus (ug/L) over time for two-yr study.
250 -,
1999
-2000
in in 10
i- r-~-cocMOTcocoocDcoor--
"S^ddPr-cCISoSo^gCSi^:™™
cS cS <5
r- r~- r-
Date
CO 06 CO 3> O>
Coefficient of variation (%) of ammonia (ug/L) over time for two-yr study.
250
Figure A-6.
re r^ rC oo oo oo
Date
Coefficient of variation (%) of nitrate / nitrite (ug/L) over time for two-yr study.
O>
127
-------
Figure A-7.
Figure A-8.
OOlOCVJCncOCMOJCOCOOCOCO
Date
Coefficient of variation (%) of soluble reactive phosphorus (ug/L) over time for two-yr study.
6 -,--
-4-2000
CXI CO
In uo in 25
Coefficient of variation (%) of pH over time for two-yr study.
30 !
25
O)O5O>
Date
Figure A-9. Coefficient of variation (%) of alkalinity (mg/L) over time for two-yr study.
128
-------
Figure A-11.
— . • ~. — ,
roo>o>
Date
Figure A-10. Coefficient of variation (%) of hardness (mg/L) over time for two-yr study.
1999
.2000
.1999
.2000
Oh^^-»-h-^t-r-OOlOCMC7>CDCNO>tDeOO--
T-T-CNCO;ST-CMCN:>X-->-IN;R;S«-CMCO;S-<-CMCN
•— •— ~— -~C CD —^ ^-. -~; n- —-, ~—. —, OO CO —.. —Z ~-^ OJ —^ — *~^
iniomio CDCDCD N-r^h- oooooo ojoicn
Date
Coefficient of variation (%) of conductivity (mS/cm) over time for two-yr study.
100
90 --
80
70
_ 60
g^
~ 50
40
8
o
55 25 io 25
5
-------
Figure A-13.
Figure A-14.
c3 co CD F^P^r-^ oo co co oo
Date
Coefficient of variation (%) of dusk temperature (°C) over time for two-yr study.
4
3.5
LOIOU5
CDCDUD
Date
Coefficient of variation (%) of dawn temperature (°C) over time for two-yr study.
100
90
Figure A-15. Coefficient of variation (%) of macrophytes (g dry wt.) over time for 1999 study.
130
-------
F\gure A-16.
Date
Coefficient of variation (%) of total nitrogen in macrophytes (% of dry wt.) over time for two-yr study.
60
50
1999
.2000
Date
Figure A-17. Coefficient of variation (%) of total phosphorus in macrophytes (% of dry wt.) over time for two-yr
study.
35
£! co- 5
OOinCNO>COOOOf-OOOt^-'»T-|v-'*J-
Jo-
CD CO
Ci £! So ^ ^
r^ r^ i^
00
o5 oo
*»
o>
Date
Figure A-18. Coefficient of variation (%) of total nitrogen in sediment (g N/rrf * 5 cm depth) over time for two-yr
study.
131
-------
Figure A-19.
Figure A-20.
Date
Coefficient of variation (%) of total phosphorus in sediment (g P/m2 * 5 cm depth) over time for two-yr
study.
1999
.2000
Date
Coefficient of variation (%) of total number zooplankton (# individuals/rrf) over time for two-yr study.
25
.1999 —*—2000
20
g15
8 10
CNOCQCNOCOCOON-'*
Date
Figure A-21. Coefficient of variation (%) of species richness total zooplankton over time for two-yr study.
132
-------
Figure A-23.
in in in
CD CD to
r-- c- i--.
Date
CO CO OO
Figure A-22. Coefficient of variation (%) of Simpson's dominance of total zooplankton over time for two-yr study.
120
2000
Date
Coefficient of variation (%) of total numbers of cladoceran zooplankton (#/rrf) over time for two-yr
study.
CM 05
Figure A-24.
~-^ ~~~* ~^». I"** ~~-~. ~~~i "~^ ^W --^. ^^. ~**~
COCDCO r-t^-r- oococo
Date
Coefficient of variation (%) of species richness of cladoceran zooplankton over time for two-yr study.
133
-------
1999
Figure A-25.
.2000
Coefficient of variation (%) of Simpson's dominance of cladoceran zooplankton over time for two-yr
study.
350 -,
Ht-1999 -*-2000
300,
in 10 m
Date
Figure A-26. Coefficient of variation (%) of total numbers copepod zooplankton (#/rrf) over time for two-yr study.
a>
Date
Figure A-27. Coefficient of variation (%) of species richness of copepod zooplankton over time for two-yr study.
134
-------
Figure A-28.
en o>
CNO>COCNO5COCOON-^r-<-CQ
^Z ^^ ^i! CD CD i^ £^j E^ h~- ^1 ^, G.
LO LO IO CD CO CD h— I**— Is*1*
Date
Coefficient of variation (%) of Simpson's dominance of copepod zooplankton over time for two-yr
study.
1999
.2000
iB 55 55
CD tB
r~ t^ r~
Date
CO 00 OO
Figure A-29. Coefficient of variation (%) of # individuals of rotifer zooplankton (#/rrf) over lime for two-yr study.
Date
Figure A-30. Coefficient of variation (%) of species richness of rotifer zooplankton over time for two-yr study.
135
-------
figure A-32.
Figure A-33.
Date
Figure A-31. Coefficient of variation (%) of Simpson's dominance of rotifer zooplankton over time for two-yr study.
Date
Coefficient of variation (%) of total numbers phytoplankton (# cells/L) over time for two-yr study.
100 -,
90
80
70
1999
.2000
^~- -^ ^^ — ^
ooraoooo
^^ i^ -~^
cncnoj
Date
Coefficient of variation (%) of species richness of phytoplankton (# species/ L) over time for two-yr
study.
136
-------
250 T
Figure A-34.
Figure A-35.
in 25 U>
co CD
CO CO CO
h- h- h-
Date
CM CO
co co co co
h- ^ T- co
05 C Cf S!
OT O) O>
Coefficient of variation (%) of total numbers bacillariophyta (# cells /L) over time for two-yr study.
140
,1999 _«_ 2000
120 4—
If)
CDCOCO
CO O
CO ^
oo co co
5
co
5i
o>
Date
Coefficient of variation (%) of species richness bacillariophyta (# species/L) over time for two-yr
study.
200 r—
.1999 _«_2000
180 4 -T~
Figure A-36. Coefficient of variation (%) of total numbers chlorophyta (# cells/L) over time for two-yr study.
137
-------
Figure A-37.
Figure A-38.
1999
.2000
Date
Coefficient of variation (%) of species richness of chlorophyta (# species/L) over time for two-yr study.
250 -, ~
T-oomT-ooLncMajcoroor^-c^oi^-^-^— t^-Tj-^-oo
JI^C!SS^;CJC!r=^C!C!23i;^:SiS235i;C!C!
m in LO CD CD CD r^-r^s- oooooooo cnojoi
Date
Coefficient of variation (%) of total numbers of cryptophyta (# cells/L) over time for two-yr study.
2*50 - — ~~ -• ~ . . .. ...... ~ , ~ ~ .-.. - .......
,1999 —•— 2000
CO 00 00
CO
O> O> C7>
Date
Figure A-39. Coefficient of variation (%) of species richness cryptophyta (# species/L) over time for two-yr study.
138
-------
250 ^
Figure A-40.
Date
Coefficient of variation (%) of total numbers cyanophyta (# cells/L) over time for two-yr study.
250
T- co in
T- J- CVJ
25 in 25
co in
CO CO
" £
CD CO CD
OJCOCOOh-COON-*
CM •
1^ ^t T- OO
T-OJCN£o-T-^C\lrOg5T--CNCNI
h- r^- i^-
Date
00 OO CO CO
O) O)
Figure A-41. Coefficient of variation (%) of species richness cyanophyta (# species/L) over time for 1999 study.
Hi
200
g150
" 100
50
• 1999
.2000
—i 1—i—i 1—\—i—\ \—i—*—
ooLOt-ooioo-JcncDcoor^
to LO co co co f^- r*^ f^
Date
coor^^T-r^.^- T- oo
»r;^;G!£255^Cl^
CO CO 00 CO CD O> O)
Figure A-42. Coefficient of variation (%) of total numbers euglenophyta (# cells/L) over time for two-yr study.
139
-------
250 -,
200
£150
100
50
1999
.2000
CO O
CD CD CO
^ 5! £2
CO CO CO
T- CO
CM CSI
Date
Figure A-43. Coefficient of variation (%) of species richness of euglenophyta (# species/L) over time for two-yr
study.
140
-------
Table A-1.
Experimental Error Statistics for Nutrient Data from 1999 Study
Variable
Chlorophyll in water
(Mg/L)
Chlorophyll in
periphyton(ug/cm2)
TN
(mg/L)
TP
(Mg/L)
NH,
(Mg/L)
SRP
(Mg/L)
NO2/NO3
(Mg/U
Grand Mean
8.57 (44)
[14.91,174]
0.04 (22)
[0.03,83]
0.87 (44)
[0.63,72]
48.7 (44)
[52.93,109]
0.01 (88)
[0.01,125]
5.98 (88)
[4.66,78]
0.06 (88)
[0.42,672]
5/20/1999
0.75 (4)
[0.39,51]
.2
0.33 (4)
[0.04,13]
17.68(4)
[2.68,15]
0.03 (4)
[0.01,25]
1.38(4)
[0.38,27]
0(4)
[0,0]
6/1/1999
0.66 (4)
[0.41 ,63]
.2
0.44 (4)
[0.09,20]
17.13(4)
[3.90,23]
0(4)
[0,0]
2.13 (4)
[0.45,21]
0(4)
[0,0]
Mean By Date1
6/3/1999
.2
_2
0.43 (4)
[0.06,14]
16.23(4)
[2.01,12]
0(4)
[0,0]
2.03 (4)
[0.45,22]
0(4) [0,1 18]
6/8/1999
0.74 (4)
[0.38,51]
.2
_2
.2
0(4)
[0,200]
4.65 (4)
[0.79,17]
0(4)
[0,0]
6/14/1999
5.30 (4)
[1 .28,24]
.2
0.46 (4)
[0.12,27]
20.98 (4)
[5.28,25]
0.01 (4)
[0.01,1327]
3.1 (4)
[0.55,18]
0(4)
[0,0]
Variable
Chlorophyll in water
(M9/L)
Chlorophyll in
periphyton (ug/cm2)
TN
(mg/L)
TP
(ug/L)
NH3
(M9/L)
SRP
(Mg/L)
NO2/NO3
(Mg/L)
6/17/1999
_2
_2
.2
_2
0(4)
[0.01,200]
2.3 (4)
[0.55,24]
0(4)
[0,200]
6/24/1999
_2
_2
_2
_2
0.01 (4)
[0.01,76]
1 .35 (4)
[0.66,49]
0(4)
[0,61]
Mean
7/1/1999
2.07 (4)
[0.93,45]
.2
0.46 (4)
[4.20,27]
18.72(4)
[4.20,22]
0.01 (4)
[0.01,148]
0.98 (4)
[0.59,60]
0.01 (4)
[0,90]
By Date1
7/7/1999
_2
_2
_2
_2
0.01 (4)
[0.01,80]
4.03 (4)
[1 .54,38]
0 (4) [0,128]
7/14/1999
4.09 (4)
[4.68,114]
.2
0.67 (4)
[0.15,23]
26.03 (4)
[3.60,14]
0.01 (4)
[0.01 ,85]
3.85 (4)
[1 .53,40]
0(4)
[0,88]
7/27/1999
4.46 (4)
[4.07,91]
0.01 (3)
[0,24]
0.78 (4)
[0.12,16]
34.85 (4)
[8.84,25]
0(4)
[0,200]
4.33 (4)
[1 .56,36]
0(4)
[0,115]
1 Mean, (n); [STD.COV]
2 Denotes no sample taken.
141
-------
Table A-1.
Experimental Error Statistics for Nutrient Data from 1999 Study (Continued)
Variable
Chlorophyll in water
(M9/L)
Chlorophyll in periphyton
(ug/cm2)
TN
(mg/L)
TP
(ug/L)
NH3
(ug/L)
SRP
(ug/L)
NO2/NO3
(ug/L)
Variable
Chlorophyll in water
(ug/L)
Chlorophyll in periphyton
(ug/cm2)
TN
(mg/L)
TP
(ug/L)
NH3
(ug/L)
SRP
(ug/L)
NO,/NO3
(ug/L)
7/29/1999
_2
0.03 (3)
[0.01 ,45]
_2
_2
0.01 (4)
[0,76]
9.08 (4)
[3.42,38]
0(4)
[0,68]
8/26/1999
.2
.2
_2
_2
0.01 (4)
[0.01,69]
7.55 (4)
[3.54,49]
0.90 (4)
[1.78,198]
8/3/1999
_2
_2
_2
.2
0(4)
[0,15]
11.7(4)
[3.72,32]
0(4)
[0,122]
8/31/1999
_2
_2
_2
.2
0.01 (4)
[0.01 ,70]
6.53 (4)
[3.40,52]
0.41 (4)
[0.81 ,200]
Mean
8/9/1999
5.47 (4)
[3.53,65]
_2
0.84 (4)
[0.12,14]
49.33 (4)
[14.82,30]
0(4)
[0,0]
8.05 (4)
[5.08,63]
0.01 (4)
[0,88]
Mean
9/7/1999
40.88 (4)
[25.04,61]
-'
2.05 (4)
[0.26,13]
140(4)
[57.39,41]
0.02 (4)
[0.02,90]
5.6 (4)
[0.71,13]
0.01 (4)
[0.01,125]
By Date1
8/12/1999
.2
.2
.2
_2
0 (4) [0,72]
9.7 (4)
[4.55,47]
0.01 (4)
[0,8]
By Date1
9/14/1999
_2
0.03 (4)
[0.02,69]
_2
_2
0(4)
[0,75]
10(4)
[3.19,32]
0(4)
[0,155]
8/17/1999
.2
0.03 (4)
[0.01 ,55]
_2
.2
0.02 (4)
[0,21]
10.3(4)
[3.57,35]
0(4)
[0,36]
9/21/1999
25.64 (4)
[17.64,69]
0.05 (4)
[0.06,102]
2.12 (4)
[0.32,15]
144 (4)
[74.55,52]
0(4)
[0,110]
15.98(4)
[6.99,44]
0(4)
[0,200]
8/24/1999
4.17(4)
[2.84,68]
0.07 (4)
[0.02,21]
1 .05 (4)
[0.14,13]
50.65 (4)
[21 .60,43]
0.01 (4)
[0.01,162]
7.08 (4)
[2.89,41]
0(4)
[0,77]
'Mean, (n); [STD,COV]
2Denotes no sample taken.
142
-------
Table A-2.
Experimental Error Statistics for Water Quality Data from 1999 Study
Mean By Date1
Variable
PH
Alkalinity
(mg/L)
Hardness
(mg/L)
Turbidity
(NTU)
Conductivity
(mS/cm)
Variable
pH
Alkalinity
(mg/L)
Hardness
(mg/L)
Turbidity
(NTU)
Conductivity
(mS/cm)
Grand Mean
9.35 (40)
[0.48,5]
119.78(40)
[23.08,19]
114.03(39)
[33.05,29]
2.97 (40)
[2.31,78]
337.9 (40)
[52.87,16]
7/27/1999
9.76 (4)
[0.23,2]
104(4)
[5.89,6]
93(4)
[10.65,11]
1.73(4)
[0.33,19]
296 (4)
[12.57,4]
5/20/1999
8.43 (4)
[0.12,1]
155(4)
[22.72,15]
186.25(4)
[22.95,12]
2.00 (4)
[0,0]
449.25 (4)
[43.92,10]
8/9/1999
9.70 (4)
[0.20,2]
128.25(4)
[32.99,26]
103.5(4)
[12.48,12]
1 .83 (4)
[0.33,18]
310(4)
[9.42,3]
6/1/1999
8.87 (4)
[0.41,4]
131 (4)
[24.79,19]
154.5(4)
[25.89,17]
1.68(4)
[0.36,21]
379.75 (4)
[54.41,14]
Mean By Date1
8/24/1999
9.57 (4)
[0.05,1]
112.5(4)
[8.54,8]
93(4)
[8.87,10]
1 .93 (4)
[0.57,30]
314.25 (4)
[15.33,5]
6/14/1999
9.15 (4)
[0.45,5]
108 (4)
[3.65,3]
112.5(4)
[4.12,4]
3.25 (4)
[0.50,15]
322.75 (4)
[8.77,3]
9/7/1999
9.38 (4)
[0.17,2]
116.5 (4)
[15.61,13]
94(4)
[15.58,17]
7.88 (4)
[2.29,29]
340 (4)
[18.51,5]
7/1/1999
9.64 (4)
[0.40,4]
101 (4)
[5.03,5]
100 (4)
[2.83,3]
1.65 (4)
[0.48,29]
297 (4)
[4.32,1]
9/21/1999
9.21 (4)
[0-12,1]
140 (4)
[16.73,12]
103.3(3)
[15.28,15]
6.00 (4)
[2.31 ,38]
376 (4)
[18.35,5]
7/14/1999
9.77 (4)
[0.18,2]
101.5(4)
[7.55,7]
97.5 (4)
[11.12,11]
1.73(4)
[0.33,19]
294 (4)
[12.25,4]
1Mean, (n); [STD,COV]
143
-------
Table A-3.
Experimental Error Statistics for Diurnal Temperature Data from 1999 Study
Mean By Date1
Variable
Dusk
Dawn
Grand Mean
27.49 (68)
[3.84,14]
23.78 (68)
[3.68,15]
5/20/1999
21.4(4)
[0.41,2]
25.8 (4)
[0.29,1]
6/3/1999
21 .95 (4)
[0.33,2]
24.2 (4)
[0.24,1]
6/8/1999
27.95 (4)
[0.30,1]
31 .08 (4)
[0.15,0]
6/17/1999
20.13(4)
[0.25,1]
23.68 (4)
[0.54,2]
6/24/1999
25.3 (4)
[0.36,1]
29.68 (4)
[0.32,1]
7/1/1999
24.85 (4)
[0.40,2]
26(4)
[0.16,1]
Mean By Date1
Variable
Dusk
Dawn
Variable
Dusk
Dawn
7/7/1999
27.9 (4)
[0.14,1]
31 .93 (4)
[0.15,0]
8/31/1999
21 .88 (4)
[0.66,3]
26.9 (4)
[0.66,2]
7/14/1999
25.03 (4)
[0.57,2]
30.55 (4)
[0.24,1]
Mean By
9/7/1999
23.1 (4)
[0.70,3]
28.73 (4)
[0.66,2.]
7/27/1999
30.75 (4)
[0.33,1]
34.25 (4)
[0.24,1]
Date1
9/14/1999
18.28(4)
[0.57,3]
22.08 (4)
[0.32,1]
8/3/1999
25.94 (4 )
[0.38,1]
26.93 (4)
[0.29,1]
9/21/1999
15.5(4)
[0.57,4]
18.98(4)
[0.65,3]
8/9/1999
25.25 (4)
[0.30,1]
31 .05 (4)
[0.42,1]
8/17/1999
25.73 (4)
[0.53,2]
29.45 (4)
[0.51 ,2]
8/24/1999
23.35 (4)
[0.51 ,2]
26.1 (4)
[0.52,2]
1Mean, (n); [STD,COV]
144
-------
Table A-4. Experimental Error Statistics for Macrophyte (dry wt., TN, and TP) and Sediment (TN.TP) Data from
1999 Study
Variable
Macrophyte
(g dry wt.)
Macrophyte TN
(% of dry wt.)
Macrophyte TP
(% of dry wt.)
Sediment TN
(g N/m2*5 cm depth)
Sediment TP
(g N/m2*5 cm depth)
Variable
Macrophyte
(g dry wt.)
Macrophyte TN
(% of dry wt.)
Macrophyte TP
(% of dry wt.)
Sediment TN
(g N/m2*5 cm depth)
Sediment TP
(g N/m2*5 cm depth)
Grand Mean
14.6(24)
[8.53,58]
2.48 (20)
[0.60,24]
13.9(16)
[4.64,33]
0.33 (20) [0.10,
31]
0.06 (20)
[0.01,13]
Mean By
8/19/1999
16.35(4)
[5.24,32]
2.85 (4)
[0.30,10]
17.41 (4)
[0.93,5]
0.42 (4)
[0.11,27]
0.06 (4)
[0.01,13]
5/25/1999
2.07 (4)
[1.19,58]
2.18(4)
[0.50,23]
9.86 (4)
[2.12,21]
0.21 (4)
[0.06,28]
0.05 (4)
[0.01,11]
Date1
9/24/1999
14.38(4)
[3.70,26]
3.22 (4)
[0.52,16]
,2
0.33 (4)
[0.04,12]
0.05 (4)
[0.01,18]
Mean By Date1
6/24/1999 7/20/1999
9.48 (4) 22.67 (4)
[3.11,33] [6.48,29
1 .84 (4) 2.32 (4)
[0.11,6] [0.25,11]
9.66(4) 18.68(4)
[1.75,18] [2.53,14]
0.37 (4) 0.35 (4)
[0.07,20] [0.11,32]
0.06 (4) 0.06 (4)
[0.01,13] [0.01,9]
7/27/1999
22.67 (4)
[6.48,29]
_2
_2
.2
_2
1Mean, (n); [STD.COV]
2Denotes no sample taken.
145
-------
Table A-5.
Experimental Error Statistics for Zooplankton Data from 1999 Study
Mean By Date1
Category
Zooplankton
Zooplankton
Zooplankton
Cladoceran
Cladoceran
Cladoceran
Copepods
Copepods
Copepods
Rotifers
Rotifers
Rotifers
Variable
totnum
nsptot
simpdom
totnum
nsptot
simpdom
totnum
nsptot
simpdom
totnum
nsptot
simpdom
Grand Mean 5/12/1999
172575(11) 2
[84650,49]
14.45(11) 2
[2.25,16]
0.21(11)
[0.07,33]
86189(11) 2
[83034,96]
6.64(11) 2
[1 .43,22]
0.34(11) 2
[0.12,36]
24040(11) j
[12950,54]
1.64(11)
[0.67,41]
0.86(11) 2
[0.18,21]
37192(11) 2
[33315,90]
1.73(11) 2
[1.01,58]
0.85(11) 2
[0.24,28]
7/16/1999
128137(4)
[44710,35]
15.25 (4)
[1.26,8]
0.23 (4)
[0.10,45]
46733 (4)
[23687,51]
6.25 (4)
[0.96,15]
0.32 (4)
[0.12,39]
22096 (4)
[11991,54]
2(4)
[0.82,41]
0.90 (4)
[0.08,9]
50369 (4)
[25876,51]
2.5 (4)
[1.29,52]
0.70 (4)
[0.29,42]
8/19/1999
200703 (4)
[42998,21]
12.25 (4)
[0.50,4]
0.20 (4)
[0.02,8]
71462 (4)
[20683,29]
5.75 (4)
[0.96,17]
0.44 (4)
[0.08,18]
34607 (4)
[11348,33]
1.25(4)
[0.50,40]
0.88 (4)
[0.24,27]
49234 (4)
[38116,77]
1 (4)
[0,0]
1 (4)
[0,0]
9/22/1999
194322(3)
[154322,79]
16.33 (3)
[2.52,15]
0.19(3)
[0.07,38]
158437 (3)
[147012,93]
8.33 (3)
[1.15,14]
0.24 (3)
[0.08,34]
12544(3)
[729.07,6]
1.67(3)
[0.58,35]
0.80 (3)
[0.26,33]
3569 (3)
[5738,161]
1 .67 (3)
[0.58,35]
0.84 (3)
[0.27,32]
1Mean, (n); [STD.COV]
2Different sampling method taken.
146
-------
Table A-6. Experimental Error Statistics for Total Numbers of Phytoplankton (#/m2) Data from 1999 Study
Category
Phytoplankton
Phytoplankton
Bacillariophyta
Bacillariophyta
Chlorophyta
Chlorophyta
Cryptophyta
Cryptophyta
Cyanophyta
Cyanophyta
Euglenophyta
Euglenophyta
Grand
V3i I3DI6 MB
Mean
. .n 219650(191)
totnum [576313>262]
«*• ««
<°<- (M8,
n^otot
r"^ ^^ >1 C QOl
. . 1166782 (24)
totnum [311 0909,267]
nsptot \2.77(j°\
.. 390337 (24)
totnum [827294,212]
-<°< K>
totnum 244625 (24)
LWU ILJI 1 1 r » 00700 -1 QQ1
nsptot 1.4 (20) [1.10,78]
119 (24)
totnum [583,490]
0.20 (20)
nSPt0t [0.41 ,205]
5/12/1999
8073 (23)
[36995,458]
6(1)
0(4)
[0,0]
4(1)
0(4)
[0,0]
0(1)
0(4)
[0,0]
1 (D
0(4)
[0,0]
0(1)
0(4)
[0,0]
1 (D
Mean By
5/20/1999
77495 (29)
[292374,377]
4(3)
[2.65,66]
12885 (4)
[13392,104]
2(3)
[1.73,87]
66.67 (3)
[115,173]
0.67 (3)
[0.58,87]
1460085(4)
[1603986,110]
1 (3)
[0,0]
92124(4)
[113131,123]
0.33 (3)
[0.58,173]
714 (4)
[1428,200]
0(3)
[0,0]
Date1
6/21/1999
193927(34)
7.75 (4)
[5.85,76]
914(4)
[1348,148]
1 .25 (4)
[0.96,77]
348010(4)
3.75 (4)
[4.11,110]
375039 (4)
1 (4)
[0,0]
661615(4)
1.5(4)
[1.29,86]
0(4)
[0,0]
0.25 (4)
[0.50,200]
7/14/1999
330473 (36)
[635370,192]
7.5 (4)
[1 .73,23]
1164(4)
[1629,140]
0.75 (4)
[0.50,67]
2617842 (4)
4(4)
[2.45,61]
492572 (4)
[599606,122]
0.75 (4)
[0.50,67]
350367 (4)
[216299,62]
1.75(4)
[0.50,29]
0(4)
[0,0]
0.25 (4)
[0.50,200]
1Mean, (n); [STD,COV]
2Denotes no sample taken.
147
-------
Table A-6. Experimental Error statistics for Total Numbers of Phytoplankton (#/m2) Data from 1999 Study (Con-
tinued)
Category
Phytoplankton
Phytoplankton
Bacillariophyta
Bacillariophyta
Chlorophyta
Chlorophyta
Cryptophyta
Cryptophyta
Cyanophyta
Cyanophyta
Euglenophyta
Euglenophyta
Variable
totnum
nsptot
totnum
nsptot
totnum
nsptot
totnum
nsptot
totnum
nsptot
totnum
nsptot
Mean
8/17/1999
394908 (32)
[836675,212]
9.25 (4)
[0.96,10]
17531 (4)
[12631,72]
0.75 (4)
[0.96,128]
3774233 (4)
[6261811,166]
5.75 (4)
0.50,9]
14325(4)
[28651 ,200]
0.5 (4)
[0.58,115]
286507 (4)
[385525,135]
2.25 (4)
[0.96,43]
0(4)
[0,0]
0(4)
[0,0]
By Date1
9/23/1999
233126(36)
[692818,297]
11.5(4)
[1.91,17]
93118(4)
[76066,82]
3.5 (4)
[0.58,16]
260560 (4)
[238524,92]
5.5 (4)
[1.29,23]
0(4)
[0,0]
1 (4)
[0,0]
77137 (4)
[154273,200]
1 .25 (4)
[1.26,101]
0(4)
[0,0]
0.25 (4)
[0.50,200]
1Mean, (n); [STD.COV]
2Denotes no sample taken.
148
-------
Table A-7.
Experimental Error Statistics for Nutrient Data from 2000 Study
Variable
Chlorophyll in water
(M9/L)
Chlorophyll in
periphyton
(ug/cm2)
TN
(mg/L)
TP
(ug/L)
NH,
(M9/L)
SRP
(Mg/L)
N02/N03
(ug/L)
Variable
Chlorophyll in water
(ug/L)
Chlorophyll in
periphyton
(ug/cm2)
TN
(mg/L)
TP
(ug/L)
NH.
(ug/L)
SRP
(ug/L)
NO2/NO3
(ug/L)
Grand Mean
6.09 (76)
[8.94,147]
1.01 (40)
[0.01,114]
0.68 (76)
[0.26,38]
42.22 (76)
[29.11,69]
0.01 (136)
[0.02,155]
7.32 (136)
[4.30,59]
0.03(136)
[0.07,288]
5/25/2000
.2
1.01 (4)
1 .02 [0,43]
_2
.2
0.03 (4)
[0.03,99]
4.15 (4)
[0.79,19]
0.02 (4)
[0.02,96]
5/10/2000
2.16(4)
[0.97,45]
_2
0.38 (4)
[0.10,26]
13.68(4)
[2.68,20]
0.01 (4)
[0.01,126]
5.55 (4)
[0.40,7]
0.21 (4)
[0.10,50]
5/30/2000
1.35(4)
[0.74,55]
.2
0.65 (4)
[0.06,9]
24.42 (4)
[3.56,15]
0.05 (4)
[0.03,70]
4.35 (4)
[0.79,18]
0.01 (4)
[0.01,94]
Mean
5/11/2000
_2
,2
_2
,2
0(4)
[0,128]
5.93 (4)
[0.91,15]
0.16(4)
[0.11,69]
Mean By Date1
5/31/2000
_2
_2
.2
.2
0.03 (4)
[0.03,94]
5.35 (4)
[0.70,13]
0.01 (4)
[0.01,136]
By Date1
5/16/2000
0.77 (4)
[0.91,118]
.2
_2
_2
0.02 (4)
[0,29]
12.58(4)
[0.84,7]
0.07 (4)
[0.12,165]
6/1/2000
_2
,2
.2
_2
0.04 (4)
[0.04,85]
6.45 (4)
[0.60,9]
0.01 (4)
[0.01,87]
5/18/2000
.2
0.01 (4)
[0.01,105]
0.3 (4)
[0.08,27]
14.18 (4)
[7.67,54]
0.02 (4)
[0.01,37]
13.25(4)
[5.23,39]
0.17 (4)
[0.30,179]
6/6/2000
0.68 (4)
[0.31,46]
.2
0.63 (4)
[0.13,20]
30.5 (4)
[3.68,12]
0.03 (4)
[0.03,103]
6.08 (4)
[0.72,12]
0.01 (4)
[0.01,121]
5/23/2000
0.79 (4)
[0.34,43]
_2
0.43 (4)
[0.05,12]
19.43 (4)
[2.08,11]
0.01 (4)
[0.02,137]
4.55 (4)
[0.56,12]
0.03 (4)
[0.03,119]
6/7/2000
.2
_2
_2
.2
0.02 (4)
[0.02,103]
5.68 (4)
[0.76,13]
0.01 (4)
[0.01,109]
1Mean, (n); [STD.COV]
2Denotes no sample taken.
149
-------
Table A-7.
Experimental Error Statistics for Nutrient Data from 2000 Study (Continued)
Variable
Chlorophyll in water
(ug/L)
Chlorophyll in periphyton
(ug/cm2)
TN
(mg/L)
TP
(M9/L)
NH
(ug/L)
SRP
(ug/L)
N02/N03
(M9/L)
6/13/2000
1 .04 (4)
[0.98,94]
.2
0.53 (4)
[0.05,10]
21.58 (4)
[2.53,12]
0.01 (4)
[0,33]
4.8 (4)
[1.21,25]
0(4)
[0.01,170]
6/14/2000
_2
0.01 (4)
[0.01,65]
.2
.2
0(4)
[0,159]
7.15(4)
[1 .43,20]
0.01 (4)
[0.01,62]
Mean
6/20/2000
1.59(4)
[2.02,127]
-?
0.58 (4)
[0.05,9]
25.48 (4)
[1302,51]
0(4)
[0,200]
3.6 (4)
[1.33,37]
0.01 (4)
[0.01,134]
By Date1
6/21/2000
_2
0.02 (4)
[0.02,73]
.2
.2
0(4)
[0,115]
3.55 (4)
[0.50,14]
0.01 (4)
[0,42]
6/27/2000
0.47 (4)
[0.14,30]
_2
0.5 (4)
[0,0]
17.75 (4)
[2.76,16]
0.01 (4) [0,22]
4.08 (4)
[0.99,24]
0.01 (4)
[0.01,122]
6/28/2000
_2
_2
_2
_2
0.01 (4)
[0,32]
5.7 (4)
[0.92,16]
0.01 (4)
[0,29]
Variable
Chlorophyll in water
(ug/L)
Chlorophyll in periphyton
(ug/cm2)
TN
(mg/L)
TP
(ug/L)
NH3
(ug/L)
SRP
(ug/L)
N02/N03
(ug/L)
7/4/2000
0.85 (4)
[0.56,66]
_2
0.53 (4)
[0.05,10]
17.85(4)
[3.49,20]
0(4)
[0,120]
1 .83 (4)
[0.29,16]
0.02 (4)
[0.01,71]
7/5/2000
_2
_2
_2
.2
0(4)
[0,0]
3.8 (4)
[1.52,40]
0.01 (4)
[0.01,70]
Mean
7/11/2000
1.24(4)
[0.70,56]
_2
0.6 (4)
[0,0]
21 .05 (4)
[4.27,20]
0(4)
[0,52]
3.43 (4)
[0.97,28]
0.02 (4)
[0.02,118]
By Date1
7/12/2000
_2
0(4)
[0,47]
_2
.2
0(4)
[0,141]
6.78 (4)
[1.88,28]
0.01 (4)
[0.01 ,80]
7/18/2000
1 .82 (4)
[0.55,30]
_2
0.65 (4)
[0.10,15]
30.13(4)
[6.35,21]
0.01 (4)
[0,6]
9.4 (4)
[4.88,52]
0.01 (4)
[0.01,76]
7/19/2000
.2
0(4)
[0,81]
.2
_2
0(4)
[0.01,128]
16.33 (4)
[9.08,56]
0.01 (4)
[0,31]
Wean, (n); [STD,COV]
2Denotes no sample taken.
150
-------
Table A-7.
Experimental Error Statistics for Nutrient Data from 2000 Study (Continued)
Variable
Chlorophyll in water
(Mg/L)
Chlorophyll in periphyton
(ug/cm2)
TN
(mg/L)
TP
(Mg/L)
NH3
(Mg/L)
SRP
(Mg/L)
NO2/NO3
(Mg/L)
7/25/2000
15.04(4)
[13.32,89]
.2
0.88 (4)
[0.25,29]
69.58 (4)
[33.74,49]
0 (4) [0,141]
8.85 (4)
[2.98,34]
0.01 (40
[0,17]
7/26/2000
_2
_2
_2
.2
0.01 (4)
[0,37]
7.23 (4)
[2.42,34]
0.01 (4)
[0,22]
Mean By
8/1/2000
13.89(4)
[17.80,128]
.2
0.93 (4)
[0.38,41]
71.78(4)
[36.06,50]
0(4)
[0,115]
9.95 (4)
[2.91,29]
0.01 (4)
[0,40]
Date1
8/2/2000
.2
.2
.2
.2
0(4)
[0,62]
8(4)
[1.69,21]
0.01 (4)
[0,35]
8/8/2000
7.01 (4)
[5.41,77]
_2
0.85 (4)
[0.17,20]
58.88 (4)
[8.81,15]
0.01 (4)
[0.01,104]
11.98(4)
[5.15,43]
0(4)
[0,200]
8/9/2000
_2
0(4)
[0,54]
.2
_2
0.01 (4)
[0,49]
9.53 (4)
[1.84,19]
0(4)
[0,0]
Variable
Chlorophyll in water
(Mg/L)
Chlorophyll in periphyton
(ug/cm2)
TN
(mg/L)
TP
(M9/L)
NH3
(Mg/L)
SRP
(Mg/L)
N02/N03
(M9/L)
8/15/2000
1 1 .90 (4)
[5.56,47]
1.01 (4)
[0,36]
0.95 (4)
[0.33,35]
66.25 (4)
[26.22,40]
0.01 (4)
[0,48]
7.75 (4)
[1.77,23]
0(4)
[0,200]
8/22/2000
18.26(4)
[9.07,50]
_2
1.03(4)
[0.17,17]
76.68 (4)
[18.32,24]
0(4)
[0,29]
8.83 (4)
[1 .78,20]
0(4)
[0,0]
Mean By
8/29/2000
21.89(4)
[13.41,61]
.2
0.95 (4)
[0.06,6]
68.65 (4)
[10.74,16]
0.01 (4)
[0,24]
8.18(4)
[0.29,4]
0(4)
[0,200]
Date1
9/5/2000
8.11 (4)
[5.58,69]
.2
0.98 (4)
[0.10,10]
87.38 (4)
[25.09,29]
0.01 (4)
[0.01,51]
14.35 (4)
[7.75,54]
0(4)
[0,0]
9/12/2000
6.77 (4)
[3.26,48]
0.01 (4)
[0.01,110]
0.63 (4)
[0.10,15]
66.93 (4)
[20.88,31]
0.01 (4)
[0.01,83]
9.95 (4)
[6.05,61]
0(4)
[0,0]
9/21/2000
_2
0.03 (4)
[0.02,74]
.2
.2
.2
.2
.2
1Mean, (n); [STD,COV]
2Denotes no sample taken.
151
-------
Table A-8.
Experimental Error Statistics for Water Quality Data from 2000 Study
Variable
PH
Alkalinity
(mg/L)
Hardness
(mg/L)
Turbidity
(NTU)
Conductivity
(us/L)
Grand Mean
9.17(76)[0.67,7]
116.65(75)
[52.43,45]
139.66 (76)
[66.63,48]
4.04 (76)
[2.93,73]
382.39 (76)
[126.66,33]
5/10/2000
8.37 (4)
[0.04,0]
233 (4)
[6.48,3]
273 (4)
[6.83,3]
3.88 (4)
[0.28,7]
625.25 (4)
[4.79,1]
5/16/2000
8.26 (4)
[0.02,0]
200.75 (4)
[3.40,2]
243 (4)
[7.75,3]
3.28 (4)
[0.94,29]
572.75 (4)
[12.92,2]
Mean By Date1
5/23/2000
8.28 (4)
[0.08,1]
192.5(4)
[7.55,4]
234 (4)
[11.66,5]
2.78 (4)
[0.21,7]
519.5 (4)
[18.70,4]
5/30/2000
8.14(4)
[0.03,0]
178.75 (4)
[10.75,6]
231 (4)
[18.07,8]
2.48 (4)
[0.60,24]
556.25 (4)
[31.67,6]
6/6/2000
8.35 (4)
[0.17,2]
170.5 (4)
[26.80,16]
220.25 (4)
[32.54,15]
2.1 (4)
[0.43,21]
548.25 (4)
[53.92,10]
Variable
PH
Alkalinity
(mg/L)
Hardness
(mg/L)
Turbidity
(NTU)
Conductivity
(mS/cm)
6/13/2000
8.64 (4) [0.34,4]
135.5 (4)
[26.95,20]
175.5 (4)
[30.51,17]
4.1 (4) [1.1 2,27]
464 (4)
[56.91,12]
6/20/2000
9.05 (4)
[0.46,5]
110.5(4)
[20.74,19]
147.5 (4)
[22.05,15]
2.03 (4)
[1 .79,88]
402 (4)
[39.88,10]
Mean By
6/27/2000
9.41 (4)
[0.50,5]
89.5 (4)
[12.58,14]
119.75 (4)
[12.23,10]
1.43 (4)
[0.19,13]
345.75 (4)
[24.10,7]
Date1
7/4/2000
9.77 (4)
[0.41,4]
79.75 (4)
[7.41,9]
108.25 (4)
[9.32,9]
1 .85 (4)
[0.70,38]
328.75 (4)
[10.94,3]
7/11/2000
9.97 (4)
[0.09,1]
76.5 (4)
[5.51 ,7]
101 (4)
[8.25,8]
1 .35 (4)
[0.17,13]
357 (4)
[67.46,19]
7/18/2000
9.90 (4)
[0.06,1]
79.67 (3)
[4.73,6]
100.75(4)
[10.18,10]
2.05 (4)
[0.13,6]
305 (4)
[15.36,5]
1Mean, (n); [STD.COV]
152
-------
Table A-8.
Experimental Error Statistics for Water Quality Data from 2000 Study (Continued)
Variable
PH
Alkalinity
(mg/L)
Hardness
(mg/L)
Turbidity
(NTU)
Conductivity
(mS/cm)
7/25/2000
9.76 (4) [0.05,0]
92.5 (4)
[10.25,11]
99.5 (4)
[9.57,10]
4.93 (4)
[3.14,64]
305 (4)
[14.45,5]
8/1/2000
9.66 (4)
[0.08,1]
77.5 (4)
[2.52,3]
94.5 (4)
[6.40,7]
5.03 (4)
[3.34,66]
382.75 (4)
[33.48,9]
Mean By
8/8/2000
9.37 (4)
[0.13,1]
68.5 (4)
[3.00,4]
79(4)
[10.65,13]
3.28 (4)
[1 .41 ,43]
256.25 (4)
[13.72,5]
Date1
8/15/2000
9.66 (4)
[0.08,1]
84(4)
[20.20,24]
82.5 (4)
[7.37,9]
7.58 (4)
[3.84,51]
263.25 (4)
[11.70,4]
8/22/2000
9.71 (4)
[0.08,1]
80(4)
[6.93,9]
87(4)
[6.63,8]
6.98 (4)
[3.06,44]
259 (4) .
[10.95,4]
8/29/2000
9.74 (4)
[0.20,2]
78(4)
[14.14,18]
78.25 (4)
[7.50,10]
8.33 (4)
[3.92,47]
241 (4)
[12.49,5]
'Mean, (n); [STD,COV]
153
-------
Table A-9.
Experimental Error Statistics for Diurnal Temperature Data from 2000 Study
Variable
Dusk
Dawn
Variable
Dusk
Dawn
Variable
Dusk
Dawn
Variable
Dusk
Dawn
Grand Mean
27.82 (132)
[3.17,11]
24.71 (136)
[2.98,12]
5/30/2000
24.08 (4)
[0.10,0]
27.33 (4)
[0.39,1]
6/20/2000
24.43 (4)
[0.15,1]
25.48 (4)
[0.19,1]
7/12/2000
29.6 (4)
[0.36,1]
32.7 (4)
[0.42,1]
5/11/2000
16.55 (4)
[0.44,3]
20.63 (4)
[0.25,1]
5/31/2000
24.63 (4)
[0.72,3]
29.9 (4)
[0.20,1]
6/21/2000
23.33 (4)
[0.10,0]
25.7 (4)
[0.28,1]
7/18/2000
27.1 (4)
[0.20,1]
28.75 (4)
[0.13,0]
Mean By
5/12/2000 5/15/2000
19.08(4) 19.75(4)
[0.15,1] [0.17,1]
23.78(4) 21.63(4)
[1.69,7] [0.13,1]
Mean By Date1
6/1/2000 6/6/2000
26.74(4) 19.78940
[0.19,1] [0.22,1]
30.05 (4) 23.88 (4)
[0.13,0] [0.15,1]
Mean By Date1
6/27/2000 6/28/2000
24.51 (4) 24.1 (4)
[0.18,1] [0.14,1]
25.98 (4) 28 (4)
[0.10,0] [0.57,2]
Mean By Date1
7/19/2000 7/25/2000
26.58 (4) 24.68 (4)
[0.25,1] [0.47,2]
27.53 (4) 29.1 (4)
[0.10,0] [0.28,1]
Date1
5/18/2000
21.89(4)
[0.30,1]
24.7 (4)
[0.24,1]
6/7/2000
21.18 (4)
[0.19,1]
25.19 (4)
[0.25,1]
7/4/2000
27.9 (4)
[0.27,1]
_ 2
7/26/2000
25.13(4)
[0.41 ,2]
29.83 (4)
[0.61,2]
5/23/2000
22.7 (4)
[0.35,2]
25.13(4)
[0.25,1]
6/13/2000
25.75 (4)
[0.31,1]
29.13(4)
[0.15,1]
7/5/2000
28.13(4)
[0.22,1]
32.73 (4)
[0.95,3]
8/1/2000
25.3 (4)
[0.40,2]
29.88 (4)
[0.26,1]
5/25/2000
23.88 (4)
[0.15,1]
26.73 (4)
[0-22,1]
6/14/2000
26.03 (4)
[0.10,0]
26.93 (4)
[0.29,1]
7/11/2000
30.38 (4)
[0.33,1]
33.55 (4)
[0.33,1]
8/2/2000
26.55 (4)
[0.47,2]
28.6 (4)
[0.24,1]
Mean By Date1
Variable
Dusk
Dawn
8/8/2000
24.85 (4)
[0.24,1]
29.38 (4)
[0.24,1]
8/9/2000
26.65 (4)
[0.47,2]
32.28 (4)
[0.67,2]
8/15/2000 8/22/2000
27.08(4) 25.13(4)
[0.53,2] [0.15,1]
30.68 (4) 28.68 (4)
[1.81,6] [1.48,5]
8/29/2000
28.48 (4)
[0.33,1]
32.05 (4)
[0.70,2]
9/5/2000
23.63 (4)
[0.45,2]
25.93 (4)
[0.38,1]
9/12/2000
24.7 (4)
[0.23,1]
26.35 (4)
[0.52,2]
1Mean, (n); [STD.COV]
2Denotes no sample taken.
154
-------
Table A-10. Experimental Error Statistics for Macrophyte and Sediment Data from 2000 Study
Mean By Date1
Variable Grand Mean 6/9/2000 7/3/2000 7/31/2000 9/4/2000
Macrophyte 1.05(16) 0.17(4) 1.11(4) 1.73(4) 1-17(4)
(gdrywt.) [0.94,89] [0.16,97] [0.71,64] [1.28,74] [0.73,62]
Macrophyte TN 2.27(16) 1.78(4) 2.13(4) 2.11(4) 3.06(4)
(%ofdrywt.) [0.60,26] [0.53,30] [0.44,21] [0.23,11] [0.26,8]
Macrophyte TP 0.45(15) 0.26(3) 0.48(4) 0.47(4) 0.54(4)
(%ofdrywt.) [0.13,30] [0.14,52] [0.11,23] [0.11,22] [0.03,6]
Sediment TN 0.33(16) 0.39(4) 0.27(4) 0.32(4) 0.34(4)
(g N/m2*5 cm depth) [0.08,25] [0.07,18] [0.08,29] [0.07,23] [0.09,27]
Sediment TP 0.06(16) 0.07(4) 0.06(4) 0.06(4) 0.06(4)
(g N/m2*5 cm depth) [0.01,14] [0.01,8] [0.01,14] [0.01,17] [0.01,9]
1Mean, (n); [STD,COV]
155
-------
Table A-11. Experimental Error Statistics for Zooplankton Data from 2000 Study
Mean By Date1
Category
Zooplankton
Zooplankton
Zooplankton
Cladoceran
Cladoceran
Cladoceran
Copepods
Copepods
Copepods
Rotifers
Rotifers
Rotifers
Variable
totnum
nsptot
simpdom
totnum
nsptot
simpdom
totnum
nsptot
simpdom
totnum
nsptot
simpdom
Grand Mean
61996(20)
[47761,77]
13(20)
[2.60,20]
0.36 (20)
[0.22,60]
32522 (20)
[26900,83]
4.2 (20)
[0.95,23]
0.57 (20)
[0.19,33]
17997(20)
[21714,121]
2.35 (20)
[1.31,56]
0.65 (20)
[0.30,47]
3894 (20)
[2.12,38]
5.5 (20)
[2.12,38]
0.44 (20)
[0.14,33]
5/12/2000
9699 (4)
[9140,94]
10.5 (4)
[2.38,23]
0.48 (4)
[0.37,77]
9080 (4)
[9485,104]
4.25 (4)
[0.5,12]
0.58 (4)
[0.29,50]
188.29(4)
[334.9,178]
1 (4)
[1.41,141]
0.37 (4)
[0.48,130]
290.98 (4)
[426.0,146]
4.5 (4)
[0.58,13]
0.43 (4)
[0.21 ,49]
6/9/2000
32395 (4)
[8879,27]
13.25 (4)
[1.50,11]
0.33 (4)
[0.23,70]
16838(4)
[12094,72]
5(4)
[1.41,28]
0.47 (4)
[0.15,33]
11941 (4)
[6997,59]
3(4)
[1.41,47]
0.69 (4)
[0.24,35]
174.25(4)
[134.6,77]
4.25 (4)
[1 .71 ,40]
0.38 (4)
[0.12,31]
7/7/2000
51647(4)
[23553.46]
12(4)
[1.15,10]
0.40 (4)
[0.12,31]
36472 (4)
[31391,86]
4(4)
[0.82,20]
0.68 (4)
[0.18,27]
11075 (4)
2.5 (4)
[0.58,23]
0.68 (4)
[0.26,37]
1230(4)
[918.2,75]
4.5 (4)
[0.58,13]
0.46 (4)
[0.10,21]
8/4/2000
119845(4)
[39122,33]
12.25(4)
[0.96,8]
0.35 (4)
[0.21,60]
65376 (4)
[27971,43]
3.75 (4)
[0.96,26]
0.60 (4)
[0.20,33]
30844 (4)
[30551,99]
2.25 (4)
[0.96,43]
0.67 (4)
[0.22,33]
6909 (4)
[6091,88]
5.25 (4)
[0.96,18]
0.55 (4)
[0.18,32]
9/7/2000
96395 (4)
[34670,36]
17(4)
[0.82,5]
0.24 (4)
[0.06,26]
34845 (4)
[7145,21]
4(4)
[0.82,20]
0.53 (4)
[0.11,21]
35934 (4)
[26162,73]
3(4)
[1-41,47]
0.82 (4)
[0.15,19]
10864(4)
[9501 ,87]
9(4)
[1.63,18]
0.36 (4)
[0.06,17]
1Mean, (n); [STD.COV]
2Denotes no sample taken.
156
-------
TableA-12.
Experimental Error Statistics for Total Numbers of Phytoplankton (#/m2) Data from 2000 Study
Category
Phytoplankton
Phytoplankton
Bacillariophyta
Bacillariophyta
Chlorophyta
Chlorophyta
Cryptophyta
Cryptophyta
Cyanophyta
Cyanophyta
Euglenophyta
Euglenophyta
Variable Grand Mean
5700600 (20)
totnum [6698397,118]
«~ S
. . 221150(20)
totnum [422420,191]
nsptot 4.8 (20) [3.78,79]
2766000 (20)
roinum [3392408,123]
nsPt°t I A 77 661
. 23400 (20)
totnum [44882,192]
1 25 (20)
nsptot M T7 1 101
. . 2635650 (20)
totnum [5394883,205]
nsptot 2.1 (20) [2.27, 108]
, . 57263(19)
roinum [220880,386]
-P- M
5/11/2000
335000 (4)
[323417,97]
12.25 (4)
[8,66]
45000 (4)
[24083,54]
2.75 (4)
[1,18]
214000(4)
[230853,108]
5.5 (4)
[4,79]
62000 (4)
[81191,131]
1.75(4)
[1,55]
13000(4)
[12490,96]
2(4)
[2,108]
1000(4)
[2000,200]
0.25 (4)
[1,200]
Mean By Date1
6/7/2000
1754250 (4)
[1987828,114]
10(4)
[9,87]
47750 (4)
[45828,96]
5(4)
[4,71]
1693000(4)
[1933972,114]
4.25 (4) [5, 124]
0 940 [0,0]
0.25 (4) [1 ,200]
9500 940
[19000,200]
0.25 (4) [1,200]
1000(4)
[2000,200]
0.25 (4) [1 ,200]
7/5/2000
5227500 (4)
[5262102,101]
15.5 (4) [10,64]
195500(4)
[284288,145]
4.50 (4)
[4,99]
3594000 (4)
[4946223,138]
6(4)
[4,65]
22500 (4)
[32388,144]
2(4)
[1,58]
1413000(4)
[2509204,178]
2.75 (4)
[2,62]
2500 (4)
[5000,200]
0.25 (4) [1 ,200]
1Mean, (n); [STD,COV]
157
-------
Table A-12. Experimental Error Statistics for Total Numbers of Phytoplankton (#/m2) Data from 2000 Study (Con-
tinued)
Category
Phytoplankton
Phytoplankton
Bacillariophyta
Bacillariophyta
Chlorophyta
Chlorophyta
Cryptophyta
Cryptophyta
Cyanophyta
Cyanophyta
Euglenophyta
Euglenophyta
Variable
totnum
nsptot
totnum
nsptot
totnum
nsptot
totnum
nsptot
totnum
nsptot
totnum
nsptot
Mean By
8/2/2000
9569500 (4)
[7340385,77]
18.75(4)
[13,70]
118250(4)
[112760,95]
5(4)
[4,71]
3095750 (4)
[2391557,77]
9(4)
[4,47]
20000 (4)
[40000,200]
1.5(4)
[2,159]
6334500 (4)
[8421168,133]
3(4)
[4,119]
1000(4)
[2000,200]
0.25 (4)
[1 ,200]
Date1
9/5/2000
11619750 (4)
[8508006,73]
23(4)
[13,58]
699250 (4)
[795404,114]
6.75 (4)
[6,87]
5233250 (4)
[4417499,84]
1 1 .25 (4)
[5,41]
12500(4)
[25000,200]
0.75 (4)
[1,128]
5408250 (4)
[7633214,141]
2.5 (4)
[2,95]
355333 (3)
[529881,149]
2.33 (3)
[1.53,65]
1Mean,(n);[STD,COV]
158
-------
?/EPA
United States
Environmental Protection
Agency
National Risk Management
Research Laborator>
Cincinnati, OH 45268
Official Business
Penalty for Private Lse
S3 00
EPA'600 R-06-05X
June 2006
Please make alt necessary changes on the below label
aetach or copy and return to the address in the upper
lelt-hand corner
I' /ou do not wish to receive these reports CHECK HERED
detach or copy this cover and return to the address in the
upper left-hand corner
PRESORTED STANDARD
POSTAGE & FEES PAID
EPA
PERMIT No. G-35
Recycled/Recyclable
Printed with vegetable-based ink on
paper that contains a minimum of
50% post-consumer fiber content
processed chlorine free
------- |