Citation
Mineral and nutrient cycles and their effect on the proton balance of a softwater, acidic lake

Material Information

Title:
Mineral and nutrient cycles and their effect on the proton balance of a softwater, acidic lake
Creator:
Baker, Lawrence Alan, 1951-
Publication Date:
Language:
English
Physical Description:
vii, 159 leaves : ill. ; 28 cm.

Subjects

Subjects / Keywords:
Acid precipitation (Meteorology) ( fast )
Aquatic ecology ( fast )
Dissertations, Academic -- Environmental Engineering Sciences -- UF
Environmental Engineering Sciences thesis Ph. D
Lake ecology ( fast )
City of Gainesville ( local )
pH ( jstor )
Lakes ( jstor )
Sediments ( jstor )
Genre:
bibliography ( marcgt )
theses ( marcgt )
non-fiction ( marcgt )

Notes

Thesis:
Thesis (Ph. D.)--University of Florida, 1984.
Bibliography:
Includes bibliographical references (leaves 144-152).
Additional Physical Form:
Also available online.
General Note:
Typescript.
General Note:
Vita.
Statement of Responsibility:
by Lawrence Alan Baker.

Record Information

Source Institution:
University of Florida
Holding Location:
University of Florida
Rights Management:
The University of Florida George A. Smathers Libraries respect the intellectual property rights of others and do not claim any copyright interest in this item. This item may be protected by copyright but is made available here under a claim of fair use (17 U.S.C. §107) for non-profit research and educational purposes. Users of this work have responsibility for determining copyright status prior to reusing, publishing or reproducing this item for purposes other than what is allowed by fair use or other copyright exemptions. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder. The Smathers Libraries would like to learn more about this item and invite individuals or organizations to contact the RDS coordinator (ufdissertations@uflib.ufl.edu) with any additional information they can provide.
Resource Identifier:
030454519 ( ALEPH )
11588189 ( OCLC )

Downloads

This item has the following downloads:

mineralnutrientc00bake.pdf

EAVDGX6AU_YXXEUY.xml

mineralnutrientc00bake_0149.txt

mineralnutrientc00bake_0158.txt

mineralnutrientc00bake_0037.txt

mineralnutrientc00bake_0112.txt

mineralnutrientc00bake_0013.txt

mineralnutrientc00bake_0079.txt

mineralnutrientc00bake_0041.txt

mineralnutrientc00bake_0075.txt

mineralnutrientc00bake_0082.txt

mineralnutrientc00bake_0001.txt

mineralnutrientc00bake_0105.txt

mineralnutrientc00bake_0132.txt

mineralnutrientc00bake_0062.txt

mineralnutrientc00bake_0029.txt

mineralnutrientc00bake_0025.txt

mineralnutrientc00bake_0086.txt

mineralnutrientc00bake_0095.txt

mineralnutrientc00bake_0100.txt

mineralnutrientc00bake_0038.txt

mineralnutrientc00bake_0092.txt

mineralnutrientc00bake_0017.txt

mineralnutrientc00bake_0063.txt

mineralnutrientc00bake_0107.txt

mineralnutrientc00bake_0019.txt

mineralnutrientc00bake_0116.txt

mineralnutrientc00bake_0139.txt

mineralnutrientc00bake_0166.txt

mineralnutrientc00bake_0032.txt

mineralnutrientc00bake_0129.txt

mineralnutrientc00bake_0074.txt

mineralnutrientc00bake_0007.txt

mineralnutrientc00bake_0057.txt

mineralnutrientc00bake_0042.txt

mineralnutrientc00bake_0135.txt

mineralnutrientc00bake_0024.txt

mineralnutrientc00bake_0109.txt

mineralnutrientc00bake_0064.txt

mineralnutrientc00bake_0044.txt

mineralnutrientc00bake_0089.txt

mineralnutrientc00bake_0118.txt

mineralnutrientc00bake_0108.txt

mineralnutrientc00bake_0121.txt

mineralnutrientc00bake_0087.txt

mineralnutrientc00bake_0162.txt

mineralnutrientc00bake_0125.txt

mineralnutrientc00bake_0049.txt

mineralnutrientc00bake_0071.txt

mineralnutrientc00bake_0076.txt

mineralnutrientc00bake_0153.txt

mineralnutrientc00bake_0004.txt

mineralnutrientc00bake_0106.txt

mineralnutrientc00bake_0090.txt

mineralnutrientc00bake_0040.txt

mineralnutrientc00bake_0111.txt

mineralnutrientc00bake_0097.txt

mineralnutrientc00bake_0066.txt

mineralnutrientc00bake_0030.txt

mineralnutrientc00bake_0084.txt

mineralnutrientc00bake_0142.txt

mineralnutrientc00bake_0122.txt

mineralnutrientc00bake_0088.txt

mineralnutrientc00bake_0021.txt

mineralnutrientc00bake_0117.txt

mineralnutrientc00bake_0124.txt

mineralnutrientc00bake_0150.txt

mineralnutrientc00bake_0012.txt

mineralnutrientc00bake_0034.txt

mineralnutrientc00bake_0055.txt

mineralnutrientc00bake_0005.txt

mineralnutrientc00bake_0045.txt

mineralnutrientc00bake_0147.txt

mineralnutrientc00bake_0119.txt

mineralnutrientc00bake_0102.txt

mineralnutrientc00bake_pdf.txt

mineralnutrientc00bake_0035.txt

mineralnutrientc00bake_0101.txt

mineralnutrientc00bake_0098.txt

mineralnutrientc00bake_0000.txt

mineralnutrientc00bake_0072.txt

mineralnutrientc00bake_0145.txt

mineralnutrientc00bake_0091.txt

mineralnutrientc00bake_0155.txt

EAVDGX6AU_YXXEUY_xml.txt

mineralnutrientc00bake_0039.txt

mineralnutrientc00bake_0152.txt

mineralnutrientc00bake_0068.txt

mineralnutrientc00bake_0052.txt

mineralnutrientc00bake_0026.txt

mineralnutrientc00bake_0136.txt

mineralnutrientc00bake_0048.txt

mineralnutrientc00bake_0137.txt

mineralnutrientc00bake_0003.txt

mineralnutrientc00bake_0113.txt

mineralnutrientc00bake_0080.txt

mineralnutrientc00bake_0006.txt

mineralnutrientc00bake_0047.txt

mineralnutrientc00bake_0067.txt

mineralnutrientc00bake_0031.txt

mineralnutrientc00bake_0046.txt

mineralnutrientc00bake_0011.txt

mineralnutrientc00bake_0053.txt

mineralnutrientc00bake_0083.txt

mineralnutrientc00bake_0160.txt

mineralnutrientc00bake_0027.txt

mineralnutrientc00bake_0144.txt

mineralnutrientc00bake_0099.txt

mineralnutrientc00bake_0069.txt

mineralnutrientc00bake_0070.txt

mineralnutrientc00bake_0094.txt

mineralnutrientc00bake_0114.txt

mineralnutrientc00bake_0073.txt

mineralnutrientc00bake_0103.txt

mineralnutrientc00bake_0078.txt

mineralnutrientc00bake_0085.txt

mineralnutrientc00bake_0138.txt

mineralnutrientc00bake_0154.txt

mineralnutrientc00bake_0008.txt

mineralnutrientc00bake_0023.txt

mineralnutrientc00bake_0015.txt

mineralnutrientc00bake_0059.txt

mineralnutrientc00bake_0140.txt

mineralnutrientc00bake_0156.txt

mineralnutrientc00bake_0036.txt

mineralnutrientc00bake_0028.txt

mineralnutrientc00bake_0014.txt

mineralnutrientc00bake_0134.txt

mineralnutrientc00bake_0104.txt

mineralnutrientc00bake_0131.txt

mineralnutrientc00bake_0159.txt

mineralnutrientc00bake_0115.txt

mineralnutrientc00bake_0167.txt

mineralnutrientc00bake_0130.txt

mineralnutrientc00bake_0110.txt

mineralnutrientc00bake_0128.txt

mineralnutrientc00bake_0143.txt

mineralnutrientc00bake_0060.txt

mineralnutrientc00bake_0096.txt

mineralnutrientc00bake_0077.txt

mineralnutrientc00bake_0157.txt

mineralnutrientc00bake_0018.txt

mineralnutrientc00bake_0126.txt

mineralnutrientc00bake_0010.txt

mineralnutrientc00bake_0093.txt

mineralnutrientc00bake_0081.txt

mineralnutrientc00bake_0133.txt

mineralnutrientc00bake_0022.txt

mineralnutrientc00bake_0065.txt

mineralnutrientc00bake_0120.txt

mineralnutrientc00bake_0127.txt

mineralnutrientc00bake_0165.txt

mineralnutrientc00bake_0056.txt

mineralnutrientc00bake_0043.txt

mineralnutrientc00bake_0163.txt

mineralnutrientc00bake_0123.txt

mineralnutrientc00bake_0058.txt

mineralnutrientc00bake_0016.txt

mineralnutrientc00bake_0051.txt

mineralnutrientc00bake_0002.txt

mineralnutrientc00bake_0009.txt

mineralnutrientc00bake_0146.txt

mineralnutrientc00bake_0050.txt

mineralnutrientc00bake_0054.txt

mineralnutrientc00bake_0020.txt

mineralnutrientc00bake_0141.txt

mineralnutrientc00bake_0161.txt

mineralnutrientc00bake_0164.txt

mineralnutrientc00bake_0061.txt

mineralnutrientc00bake_0033.txt

mineralnutrientc00bake_0148.txt

mineralnutrientc00bake_0151.txt


Full Text














MINERAL AND NUTRIENT CYCLES AND THEIR EFFECT ON THE PROTON BALANCE OF A SOFTWATER, ACIDIC LAKE



By

LAWRENCE ALAN BAKER




























A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN
PARTIAL FULFILLMENT OF THE REQUIREMENTS
FOR THE DEGREE OF DOCTOR OF PHILOSOPHY


UNIVERSITY OF FLORIDA 1984
















ACKNOWLEDGEMENTS

I would like to express my appreciation to those who helped make this dissertation possible. Dr. P. L. Brezonik, my original major professor, provided expert assistance and guidance throughout this research project. Several parts of ths research were done in collaboration with other students, particularly Walter Ogburn and Eric Edgerton, and their efforts are appreciated. I would also like to thank James Heaney, for serving as chairman of my supervisory committee

following the move by Dr. Brezonik and myself to the University of Minnesota, and the other committee members -- G. R. Best, G. Bitton, and D. Graetz -- for their continued interest and advice during this project.

Finally, I would like to thank my friends Jack and Debbie Tuschall, Carl Miles, and Sue Zoltewicz for technical advice and comradeship at the University of Florida and my fiance, Nancy Rodenborg, for her patient support during the writing process.

















ii















TABLE OF CONTENTS

ACKNOWLEDGEMENTS ....................................... ii

ABSTRACT ........................................ vi

CHAPTER I -- INTRODUCTION ................................... 1
Background ............................................. 1
Objectives ........................................ 2
Site Description ....................................... 3

CHAPTER 2 -- METHODS ....................................... 7
Sample Collection ..................................... 7
Analytical Methods ..................................... 8
Cations ........................................... 8
Nutrients ........................................ 8
Major Anions ...................................... 9
pH ........................................ 9
Isotope Ratios .................................... 9
Quality Assurance ................................. 11

CHAPTER 3 -- SEDIMENT BUFFERING IN McCLOUD LAKE .............. 12
Introduction ........................................ 12
Mechanisms for Sediment Buffering ...................... 15
Methods ........................................ 20
Sediment-Water Batch Experiments .................. 20
Seepage Column Experiment ......................... 22
Microcosm Experiment .............................. 22
Littoral Mesocosms ................................. 24
Pore Water Profiles ............................... 25
Results and Discussion ................................. 25
Sediment-Water Batch Experiments .................. 25
Magnitude of neutralization .................. 25
Cation exchange .............................. 27
Aluminum solubility .......................... 31
Sulfate adsorption ........................... 35
Seepage Column Experiment ......................... 37
Sediment-Water Microcosms ......................... 40
Littoral Mesocosms ................................. 41
Pore Water Profiles ............................... 49
Conclusions ........................................ 51

CHAPTER 4 -- McCLOUD LAKE MASS BALANCE ....................... 53
Introduction ........................................ 53
Background ........................................ 54
Methods ....................................... 56



iii









Wet Precipitation .................................. 56
Dry Deposition ..................................... 56
Seepage ......................................... 57
Groundwater Wells .................................. 59
Lake Storage ....................................... 59
Evaporation ........................................ 59
Results and Discussion .................................. 60
Water Budget ..................................... 60
Chemistry of McCloud Lake, Past and Present ........ 64 Dry Deposition ..................................... 71
Wet Precipitation .................................. 81
Groundwater Chemistry .............................. 83
Fate of Major Ions in McCloud Lake ................. 86
Chloride ..................................... 88
Sodium ...................................... 90
Sulfate ..................................... 90
Calcium, magnesium, and potassium ............. 92
Nitrogen species ............................. 93
Proton balance ................................ 94

CHAPTER 5 -- DECOMPOSITION AND NUTRIENT CYCLING IN McCLOUD LAKE ............................................. 96
Background ........................................... 96
Effects of Low pH on Decomposition ................. 97
Effects of Low pH on Nitrogen Cycling .............. 99
Ammonification ................................ 99
Nitrification ................................ 100
Denitrification .............................. 100
Nitrogen fixation ............................ 101
Methods ................................................ 101
Littoral Mesocosms ................................ 101
Microcosm Experiments ............................. 104
Nitrogen Mass Balance .............................. 106
Results .............................................. 106
Enclosure Experiments .............................. 106
Ammonium Adsorption ............................... 117
Microcosm Experiments ............................. 119
Water-only microcosms ......................... 119
Sediment-water microcosms ..................... 121
McCloud Lake Nitrogen Budget ....................... 129
Precipitation ................................ 129
McCloud Lake nitrogen ........................ 132
Mass balance ................................ 132
Conclusions ............................................ 140

CHAPTER 6 -- CONCLUSIONS .................................. 142
Sediment Neutralization ................................ 142
McCloud Lake Mass Balance ............................. 142
Nitrogen Cycling and Decomposition ...................... 143






iv









BIBLIOGRAPHY ..................................... *. ** *** 144

APPENDIX -- McCLOUD LAKE HYDROLOGIC AND CHEMICAL DATA ........ 153 BIOGRAPHICAL SKETCH ......................................... 159




















































V















Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy

MINERAL AND NUTRIENT CYCLES AND THEIR EFFECT ON THE PROTON BALANCE OF A SOFTWATER, ACIDIC LAKE

By

Lawrence Alan Baker

April 1984


Chairman: James P. Heaney
Cochairman: Patrick L. Brezonik Major Department: Environmental Engineering Sciences Mineral and nutrient cycles in a softwater, acidic lake were studied

using laboratory experiments, littoral mesocosms, and wholelake mass balances. The primary study site was McCloud Lake, a small (5 ha), acidic (pH 4.5), seepage lake in the Trail Ridge Region of northcentral Florida.

Titrations of sediment slurries from McCloud Lake and other softwater lakes showed that profundal seediments have an acid neutralizing capacity (ANC) up to 10 meq/100 g dry weight. Cation exchange of calcium and magnesium accounted for over 50% of the ANC in the sediments examined, while solubilization of aluminum accounted for up to 20% of the ANC. Exchanges of sodium and potassium were unimportant, as was

sulfate adsorption. An experiment to simulate subsurface seepage to McCloud Lake showed that littoral sediments could neutralize synthetic groundwater (pH 3.4 to 6.8) by sulfate reduction and cation exchange.



vi









Chemical analysis of in situ seepage and sediment pore waters confirmed these results.

Water budget calculations showed that McCloud Lake received 90% of its water from precipitation and 10% from subsurface flow. Sulfate entering the lake was lost by an internal sink (sulfate reduction) that accounted for 37 to 73% of the total input. Assimilation and denitrification consumed over 95% of the nitrate entering the lake. Internal sinks for nitrate and sulfate nearly balanced the proton input to the lake. Although the mass balance was not sensitive enough to show the internal source of calcium and magnesium expected from lab experiments, an analysis of historical trends showed that calcium levels have nearly doubled while magnesium levels have increased by 34% during the past 14 years in which the proton concentration has doubled.

Littoral mesocosm and laboratory microcosm experiments showed that nitrification can occur at the sediment-water interface when the pH of the overlying water is as low as 3.5. Sediment respiration did not appear to be affected by pH in the littoral mesocosms. An analysis of historical trends in McCloud Lake nutrient levels revealed no significant changes in major nutrient species during the past 14 years. These results cast doubt on the hypothesis that nutrient regeneration and decomposition are inhibited in acidified lakes.













vii















CHAPTER 1
INTRODUCTION

Background

Numerous studies have shown that precipitation acidity has increased in large areas of Scandinavia and North America (Dovland et al. 1976; Cogbill and Likens 1974; Brezonik et al. 1983b). It is widely believed that increased precipitation acidity has resulted in

the acidification of thousands of poorly buffered lakes and streams (Grahn et al. 1974; Gjessing et al. 1976; Norton et al. 1980; Schofield 1976; Impact Assessment Group 1983).

Acidification has major and often serious effects on lake and

stream ecosystems. In addition to the direct effect of H+ on osmoregulation (Leivestad et al. 1976), decreased pH affects biota through several indirect mechanisms. Decreased pH enhances the solubility of toxic metals, and increased concentrations of aluminum, cadmium, zinc,

and lead have been observed in lakes and streams that have become acidified (Almer et al. 1978; Norton et al. 1980). Humic acids

precipitate at low pH, accounting in part for the increase in clarity observed in acidified lakes, and the precipitation of aluminum phosphate complexes may decrease lake phosphorus levels (Almer et al. 1978).

Acidification produces distinct changes in biological communities. The most notable effect has been the decline of fish populations in many rivers and lakes in the Scandinavian countries and in


1






2


the Adirondack Mountains of New York (Leivestad et al. 1976; Almer et al. 1978; Schofield 1976). Major changes occur in the structure of phytoplankton and zooplankton communities, and declining pH clearly results in decreased species diversity of plankton communities (Hutchinson et al. 1978; Leivestad et al. 1976; Almer et al. 1978), although it is not clear whether primary productivity decreases (see Hendrey et al. 1976). Grahn et al. (1974) postulated that acidification retards decomposition and nutrient regeneration and initiates a "self-oligotrophication" process, although subsequent studies have produced conflicting results (Schindler et al. 1980; Dillon et al. 1979; Hultberg and Andersson 1982).

Objectives

One of the major problems in acid precipitation research is to elucidate the relationship between precipitation acidity and lake pH. To develop models for this purpose requires an understanding of the processes that produce and consume H+ ions in lakes and their watersheds. Current lake acidification models are based on watershed weathering reactions (Hendriksen 1980; Wright 1982) and ignore in-lake neutralization processes. Thus, the objectives of the first phase of this study (Chapter 3) were to identify H +-neutralizing mechanisms that occur in the sediments of softwater lakes and to evaluate the potential of these reactions in neutralizing the overlying water. Chapter 4 addresses the question of in-lake neutralization using a mass balance of major ions for McCloud Lake, Florida, to quantify source and sink terms.

The second phase of this research (Chapter 5) is a study of decomposition and nitrogen cycling in McCloud Lake. The hypothesis






3



that decomposition and nutrient regeneration are inhibited in acidified lakes has not been thoroughly evaluated, and several recent studies indicate that nutrient levels do not change substantially upon

acidification or neutralization of whole lakes (Dillon et al. 1979; Schindler et al. 1980; Hultberg and Andersson 1982). Chapter 5 includes in situ mesocosm and laboratory microcosm studies to evaluate nutrient regeneration, a nitrogen mass balance for McCloud Lake, and an evaluation of historical trends in nutrient levels for this lake.

Site Description

McCloud Lake is a small (5 ha) softwater lake in the Trail Ridge Region of north-central Florida, approximately 40 kilometers east of Gainesville. The region is characterized by karst topography and includes numerous sand hills interspersed by small lake basins that resulted from solution of the underlying limestone.

The 95 ha watershed is uninhabited and is located on the Katherine Ordway Ecological Preserve maintained by the University of

Florida. Surficial soils in the watershed are nonspodic marine deposits of fine sand, gravel, and sandy clays of the Citronelle Formation. These surficial deposits are underlain by 24-30 m of phosphatic sands, sandy clays, and clays of the Hawthorne Formation. These clays act as a confining layer (aquiclude) that separates the deep Floridan aquifer from the unconfined, perched water table of the Citronel le Formation (Brezonik et al. 1983a).

The lake is a simple doline (Figure 1-1) with no defined inlets or outlets. Most of the water to the lake comes from direct precipitation; the remainder enters as subsurface seepage. Water levels


















o O
0






12








0 0














0 50 100 Seepage meters meters
o Groundwater wells Figure 1-1. McCloud Lake bathymetric map.






5

Table 1-1. Summary of morphometric information for McCloud Lake. Elevation 36.5-38.0 above NVGDa Surface area (at 35.6 m) 4.81 ha Volume (at 35.6 m) 11.9 x 104 m3 Mean depth 2.5 m Maximum depth 4.6 m Watershed area 95 ha

aNational Vertical Geocletic Datum (approximately mean sea level)






6


fluctuate in response to changes in relative rates of precipitation and evaporation. During 1980-81, the water level was very low due to a prolonged drought and the maximum depth was 4-5 m. During 1966-67, the maximum depth was 6.4 m and the lake had a surface area of 9 ha (Brezonik et al. 1969), nearly double the 1980-81 area of 5 ha.

McCloud Lake has very low conductivity (< 50 uS/cm at 250 C) and no alkalinity. The lake has become increasingly acidic during the past 14 years (see Chapter 4), and now has a mean pH of 4.6. The algal standing crop is small (mean chlorophyll a was 6 ug/L in 1981), and the phytoplankton community is dominated by six genera: Oocystis, Dinobryon, Mougeotia, Chlamydomonas, Glenodinium, and Euglena (Crisman et al. 1983). A narrow band of emergent macrophytes surrounds the lake, and the sparse submergent macrophyte community is composed primarily of two species: Eleocharis sp. and Websteria confervoides (Crisman et al. 1983).














CHAPTER 2
METHODS

A major problem in evaluating historical data relative to the effects of acid precipitation is the lack of detail provided by earlier investigators on the analytical methods they used. The purpose of

this chapter is to provide details of sample collection, preservation, and laboratory analyses used in this study. Experimental procedures are described in the appropriate chapters.

Sample Collection

Throughout this study, samples were collected in linear polyethylene (LPE) bottles that were washed with soap and water, then thoroughly rinsed with distilled water. Samples collected for analyses of pH, sulfate, chloride, and conductivity were stored in bottles that were new at the beginning of the study and were never acidwashed. At the University of Florida, these were rinsed with distilled, deionized water (DDW) following the soap-and-water wash. During later experiments (at the University of Minnesota), bottles for these analyses were purchased new and simply rinsed between use and

filled with DDW during storage. Samples for nutrient analyses were stored in LPE bottles that had been soaked overnight in 1:1 sulfuric acid (at the University of Florida) or 10 % HCl (at the University of Minnesota), then rinsed with DDW. Nutrient samples were preserved by adding 40 mg/L HgC12 to the sample bottles (USEPA 1974) or by freezing. Samples for metals analyses were stored in LPE bottles that had been soaked in 10 % HNO3 and rinsed with DDW; these were acidified


7






8



with HNO3 to reach a pH < 2. Glassware used in sample analyses were washed the same way as collection bottles.

Samples were collected from McCloud Lake using an acrylic Kemmerer sampler and were filtered using in-line Gelman Type A-E glass fiber filters that had been rinsed with 250 mL lake water. In-line glass fibers were also used to collect samples for 15N analysis from the microcosm experiments; filters in this experiment were rinsed with 100 mL HCI + 250 mL DDW. Samples in the sediment titration studies and in the pore water study were filtered using glass fiber filters mounted in a syringe-type holder; the entire apparatus was rinsed with

3 fillings of DDW between samples.

Analytical Methods

Cations

Calcium, magnesium, potassium, and sodium were determined by

flame atomic absorption (AA)(Varian Model 175 or Perkin Elmer Model 5000) following methods recommended by the manufacturers. Lanthanum

oxide was added to prevent ionization interferences in calcium and magnesium analyses (APHA 1981). Aluminum was analyzed by flameless AA on the Perkin Elmer instrument using settings recommended by the manufacturer.

Nutrients

Nitrogen species (ammonium, nitrite, and nitrate) were generally determined by AutoAnalyzer techniques (Table 2-1). Kjeldahl nitrogen was determined by a scaled-down micro-Kjeldahl technique (APHA 1981) in which 50 mL samples were digested in a block digester. Ammonium in






9


the digestate was determined by AutoAnalyzer. Nitrate in the pore water study was determined by ion chromatography (described below). Major Anions

Sulfate and chloride were determined by AutoAnalyzer methods (Table 2-1) for 1) McCloud Lake water, 2) precipitation and seepage meter samples, and 3) samples from the sediment-water microcosms. Ion chromatography (Dionex Model 10) was used to determine sulfate and chloride concentrations in the seepage column experiments and in the pore water study. The ion chromatograph was operated at a flow rate of 40% of maximum (nr 1.0 mL/min) with a 10 cm pre-column, a 250 cm separator column, and a 250 cm supressor column (Dow Chemical Corporation).



Analyses of pH were conducted using an Orion Model 811 pH meter with a Corning Model 910200 pH electrode at the University of Minnesota) and an Orion Model 801 pH meter with the same probe at the University of Florida. All pH analyses were conducted within eight hours of sample collection. Fischer Certified pH Standards were used to calibrate the pH meter at 4.00 and 7.00 prior to each use. Measurements were made under quiescent conditions, as recommended by Galloway et al. (1979).

Isotope Ratios

Samples for isotope ratio analyses (15N/14N) of ammonium from the microcosm experiment (Chapter 5) were prepared by alkaline distillation (APHA 1981) of 400 mL aliquots into 0.01 N H2SO4. The first 50 mL of distillate was evaporated to < 2 mL by gentle boiling on a hot plate, cooled, and poured into an AutoAnalyzer tube. To assure






10


Table 2-1. AutoAnalyzer methods.


Detection
limit,
Parameter Method and referencea mg/L Modifications


Sulfate Methylthymol blue 0.2 For precipitation samples,
IM 226-72W BaC12 and MTB reduced to improve sensitivity.

Chloride Ferricyanide 0.2 Sample cups rinsed with
DDW.

Nitrate Cd reduction/ 0.001 Cadmium wire used instead and diazotization. of granules at U of F. nitrite IM 158-71W/B

Ammonia Phenate 0.003 EDTA substituted for
IM 154-71W/B citrate-tartrate reagent (APHA 1981). Fresh DDW used for standards. In later experiments, standard treated as standard addition to compute NH3 in DDW.

TKN Kjeldahl digestion Digestion modified by using
APHA 1981 50 mL aliquots and block digester. NH3 determined by automated phenate method.

aIM = Technicon, Inc. Industrial Methods Series.







11


adequate N2 production for isotope ratio analysis, 1.00 mg N as NH4Cl

was added to each sample prior to distillation. Isotope ratio analyses were conducted by Bill Portier at the University of Florida Soil Science Department using a Micromass 602D double beam mass spectrometer.

Quality Assurance

The correctness of chemical analyses was checked periodically using external quality control standards (EPA "mineral" and "nutrient" standards). A low ionic strength pH standard, prepared using H2S04, was used to check pH calibrations, as recommended by Galloway et al. (1979). Sulfate analysis was used to compute the exact pH of the low ionic

strength stock.

Analytical errors were also checked by ion balance calculations. If all significant ions are measured for a water sample, the sum of anions should equal the sum of cations. The difference between these terms can be assumed to represent the analytical errors. Errors in ion balances were calculated from the equation

E = (A M) x 100 (2-1)
(A- + M")
where E = error, as %,
A- = sum of anions, meq/L, and
M+ = sum of cations, meq/L.
















CHAPTER 3
SEDIMENT BUFFERING IN McCLOUD LAKE


Introduction

The development of models to predict the pH response of lakes to inputs of acid precipitation is a major objective of acid precipitation research. Models currently in use or being developed are either empirical (e.g. Almer et al. 1978) or focus on processes occurring in the watersheds of lakes. Several models treat watershed acidification as a large scale titration in which protons entering the

system via precipitation are replaced by cations, principally calcium and magnesium (Henriksen 1980; Wright 1982; Thompson 1982). Other models have been formulated to model short-term variations in streamflow chemistry (Chen et al. 1979; Christopherson and Wright 1981).

The role of pH buffering mechanisms that occur at the sedimentwater interface has received little attention in models of lake acidification although Schnoor et al. (1983) included an in-lake neutralization term in their "trickle-down" model. The ability of sediments to neutralize acidity may be particularly important for lakes that have a small ratio of watershed area to lake surface area and receive most of their inflow directly from precipitation. These "Type I" lakes (Schnoor et al. 1983), which are particularly susceptible to acidification, are common in northern Wisconsin and north-central

Florida. For example, the 13 soft water lakes in the Trail Ridge Region of north-central Florida surveyed by Brezonik et al. (1983b)


12






13



have a mean watershed area:lake surface area of 7.0 (range = 1.4 to 20) and a mean pH of 5.2 (Table 3-1). However, the role of precipitation in the water budgets of these lakes is even more important than indicated by the watershed:lake area ratios since much of the precipitation falling on the sandy soils of this region passes directly to the regional water table.

The objective of this chapter is to examine the potential for sediment neutralization of acid in softwater lakes. The addition of acid or base to sediment-water slurries was used to determine the potential magnitude of inorganic neutralization mechanisms under completely mixed conditions. Other experiments were designed to simulate natural sediment-water interactions. Chemical changes that occur during subsurface seepage were evaluated by passing synthetic groundwater through columns of littoral sediment. Acidification experiments involving laboratory sediment water microcosms and littoral mesocosms were used to examine the capacity of sediments to neutralize inputs of acidity to the overlying lake water. Finally, analyses of sediment pore waters were used to evaluate the extent of sediment neutralization processes in situ.

The principal site for this phase of the investigation was McCloud Lake. McCloud Lake is an ideal site for the study of in-lake neutralization mechanisms since the lake receives nearly all of its water directly from precipitation (Chapter 4). Furthermore, the lake has become more acidic during the past 14 years. Water chemistry data collected during 1968-69 (Brezonik and Shannon 1971) and 1978-79 (Brezonik et al. 1983b) show that the pH of McCloud Lake has dropped







Table 3-1. Characteristics of lakes in sediment neutralization experiments.

Lake Location Surface Mean depth, pH Alkalinity Conductivitg area, ha m ueq/L uS/cm @ 25 C Anderson-Cuea Putnam Co., Fla. 8 2.0 4.94 10 38 McCloud Putnam Co., Fla. 5 2.0 4.56 0 45 Adaho Alachua Co., Fla. 216 3.6 6.0 54 59 Johnson Putnam Co., Fla. 140 5.24 10 23 Geneva Clay Co., Fla. 650 4.1 6.10 130 65 Clarab Vilas Co., Wis. 34 4.5 6.03 49 37 McGrath Vilas Co., Wis. 21 3.0 5.26 -6 18 Sand Vilas Co., Wis. 15 3.2 5.01 -6 27 aData on Florida lakes from Brezonik et al. (1983). bData on Wisconsin lakes from K. Webster, Wisconsin Dept. Natural Resources.






15

from 5.0 to 4.6. The littoral sediments are generally sandy, interspersed with pockets of peat-like material while the profundal sediments are a soft, highly organic ooze that extend to a depth of approximately 6 meters.

Sediments from four other lakes in the Trail Ridge group and from three lakes in northern Wisconsin were included in the batch neutralization experiments (Table 3-1). Like McCloud Lake, these lakes lack well-defined inlets or outlets and have water with little buffering capacity (alkalinity <200 meq/L) and pH values < 7.0.

Mechanisms for Sediment Buffering

Cation exchange and mineral dissolution are undoubtedly responsible for much of the acid neutralization of watersheds (Henriksen 1980; Thompson 1982). In the well-weathered soils of Florida, kaolinite and gibbsite are the major clay minerals. Both minerals neutralize acidity during congruent dissolution (Stumm and Morgan 1981); kaolinite also has a limited cation exchange capacity (CEC) of less than 10 meq/100 g resulting from isomorphic substitution and from edge charges. In the sensitive watersheds of northern Wisconsin, the dissolution of feldspars and other clay minerals may neutralize acidity. Organic matter, which has a CEC up to 300 meq/100 g, may contribute substantially to the total cation exchange capacity of sandy soils (Yuan et al. 1967) and must be a major component of the total exchange capacity of lake sediments, which often have an organic matter content over 80 %. Humic acids in lakes are known to precipitate upon acidification, a process that consumes protons and may account for the increased clarity observed in acidified lakes (Almer






16



et al. 1978). Conversely, precipitated humic acids may buffer inputs of base, such as lime, that may be added to neutralize acidity.

There is some evidence that these reactions may result in the

depletion of cations in surficial sediments of recently acidified lakes. Norton et al. (1981) found lower zinc concentrations near the

surface of sediment cores of 20 lakes in northeastern U.S. and Norway and suggested that the zinc mobilization was caused by cation exchange or solubilization of ZnS. Kahl et al. (1982) reported evidence of accelerated leaching of Ca2+, Mg2+, Mn2+, and Zn+2 in surficial

sediments of three ponds in Maine and proposed that the trend may be the result of recent acidification. Neither of these studies (Norton

et al. 1981 or Kahl et al. 1982) related the depletion of cations to the hydrogen ion buffering capacity of the sediments. Cation exchange was found to be of minor importance in the neutralization of Lake 223 in the Experimental Lakes Area (ELA) of western Ontario during a whole-lake acidification project (Cook et al. 1982). However, Lake

223 was acidified to only pH 5.1 and it is not known whether cation exchange may have played a more significant role in buffering at lower pH levels.

Although the role of sulfate reduction in pH-buffering of sediments is well-known (Nriagu and Hem 1978), this process has received little attention as a neutralization mechanism in acid-sensitive lakes. The one exception is the study by Cook et al. (1982) which found sulfate reduction to be a major neutralizing mechanism in an

experimentally acidified lake. Over a five-year period approximately half of the H2SO4 added to ELA 223 was neutralized by sulfate reduction in the littoral sediments and the anaerobic hypolimnion. This






17



work clearly suggests the potential for sulfate reduction as a neutralizing mechanism in lakes receiving acid precipitation. Although sulfate reduction in culture is inhibited at pH levels below 5 (Zinder and Brock 1978), this process undoubtedly occurs in neutral microzones in the environment. Several reports indicate that sulfate reduction occurs at significant rates in acidic peat bogs. Hemond (1980) reported that 77% of the S042- entering Thoreau's Bog, Mass., was retained and suggested that dissimilatory reduction may have accounted

for approximately half of this loss. Collins et al. (1978) isolated and enumerated sulfate-reducing bacteria, including Desulfovibrio desulfuricans, from a bog that had nominal pH levels of 3.0-4.6 but gave no indication of the activity of the organisms in the bog. Rippon et al. (1980) concluded, from S042- and H2S profiles and 355042experiments, that sulfate reduction was a significant process in a

Danish bog (pH 4.8-5.2). These reports clearly suggest that sulfate reduction may be a significant process in acidic environments, including moderately acidic lakes.

Sulfate adsorption can occur by two mechansims. Reversible, or non-specific adsorption occurs when S042- acts as a counterion on a

positively charged surface (Hsu 1977). In some watersheds, reversible adsoption results in seasonal stabilization of stream sulfate levels.

During summer, when precipitation is relatively acidic, soil surfaces become positively charged and S042- is adsorbed. During snowmelt, more neutral water containing less sulfate passes through the soil; surfaces are neutralized and sulfate is desorbed. This process has been in watersheds of the Integrated Lake-Watershed Study (ILWAS) in






18




the Adirondack Mountains (Chen et al. 1979) and in the Birkeness watershed in Norway (Christopherson and Wright 1981).

Irreversible sulfate adsorption (also called specific adsorption) involves a substitution of S04-2 into the inner sphere of a metal hydroxide surface. The mechanism proposed by Rajan (1978) results in the displacement of two hydroxide ions for each sulfate ion adsorbed:

Al Al\
S OH 0 0

+ 2H+ + SO4 2- 2 2H20 + S

\ OH ,0 0
Al Al

(3-1)

Irreversible sulfate adsorption results in a net (permanent) accumulation of sulfate and protons in a watershed. Johnson et al. (1980) reported sulfate accumulation rates up to 9.3 kg/ha-yr for a North Carolina watershed, although many watersheds show a balance between sulfate input and output. Sulfate adsorption has been correlated with iron and aluminum content, clay content and the presence of sesquioxides, and it is inversely correlated with organic content

(Johnson et al. 1980). Specific adsorption occurs only under acidic conditions and is not a significant process above pH 6-7.

Reactions of the nitrogen cycle also consume or generate protons in sediment environments (Figure 3-1). Since approximately one-third of the acidity in Florida precipitation is HNO3 (Brezonik et al. 1983) these reactions are potentially significant in controlling the pH of softwater lakes. In forest and bog ecosystems, most of the nitrate in






19











PLANT
NH
R-C-R NH NH NO3 4 3 R organic nitrogen H 2H+


SOIL


NH2 H+ 2H+
R-C--R a NH+ NH NO3"
3 4 3 R organic nitrogen


Source: Impact Assessment Group, 1983




Figure 3-1. Effect of nitrogen cycle on proton balance.






20


precipitation is retained (Kerekes et al. 1982; Hemond 1980; Wright and Johannessen 1980). Retention of nitrate by assimilation into organic matter or by denitrification has the same effect on pH since both reactions consume one H+ per N03- ion consumed. Although the assimilation of NH4+ results in the production of H+, nitrate deposition exceeds ammonium deposition in Florida and the net effect of terrestial retention of inorganic nitrogen species is the consumption of protons. The effect of nitrogen cycling on the pH of softwater lakes has received little attention. In addition to nitrate and ammonium assimilation, mineralization of organic nitrogen and other reactions of the nitrogen cycle, such as nitrification, could potentially alter lake pH. Although some of these reactions-- notably nitrification and denitrification-- are inhibited at pH levels below 5 in laboratory cultures (NAS 1978), ample evidence indicates that they may occur in acidic environments (see Chapter 5).

Methods
Sediment-Water Batch Experiments

Initial experiments to determine the neutralizing capacity of sediments were conducted by adding 25 g of wet sediment, 100 mL of distilled water, and 1.0 mL of chloroform to 10 Erlenmeyer flasks. Each flask received a single dose of acid (0.1 N H2S04) or base (0.1 N NaOH) and was placed on a shaker table for one week. Upon termination of the experiment, water from each flask was filtered and analyzed for pH, Ca2+, Mg2+, Na+, K+, Si(OH)4, and total aluminum. In later experiments, conducted with sediments from Wisconsin lakes, sequential doses of acid or base were added daily, after first determining that the results were similar to those of the single dose method. Several






21



other modifications included the use of LPE bottles rather than glass Erlenmeyer flasks and the use of 20 g of sediment. After each experiment, the sediment in each flask was dried to 1030 C to determine dry

weight, and then ashed at 5000 C to determine the volatile solids content.

The neutralizing capacity of the sediments was calculated on a dry weight basis:


NCi = 100 x [ViNa,b 10 pHi-1OPHo (Vf)]/W (3-2) where Vi = cumulative volume of acid at iith addition, mL,
Nb = normality of acid or base,
po = original pH (no acid or base),
pHi = final pH,
Vf = volume of water in flask(including water content of wet
wet sediment), mL,
W = dry weight of sediment, g, and
NCi = neutralizing capacity of sediment at the ith acid or
base addition, meq/100 g.

The exchange of metal cations also was computed on a dry weight basis in order to calculate the contribution of each metal to the total

buffering capacity.

The potential for sulfate adsorption by McCloud Lake sediments was evaluated in a separate batch experiment in which a single dose of 0.1 N H2SO4 (0.1 to 1.0 mL) was added to each of ten LPE bottles containing 20 g sediment and 50 mL DDW. Since chloroform could potentially alter sulfate adsorption and/or sulfate analysis, the

experiment was conducted with and without the addition of 1.0 mL chloroform. The entire experiment was also conducted with a series of

bottle blanks that received no sediment. The bottles were placed on a shaker table for a week; upon termination of the experiment water from

each bottle was filtered and analyzed for sulfate and pH. Sulfate






22



adsorption was evaluated by comparing sulfate recovered with sulfate added, after accounting for sulfate levels in controls (no acid added).

Seepage Column Experiment

An upflow column study was conducted to simulate the flow of groundwater through McCloud Lake littoral sediment. The synthetic groundwater used in this experiment (Table 3-2) represented soil water

from a depth of 190 cm at a site adjacent to Lake McCloud that had been repeatedly irrigated with simulated pH 3.0 rain (J. Byers, University of Florida, pers. comm.). This soil water was circumneutral (pH 6.6) and contained appreciable alkalinity (estimated by ion balance).

The synthetic groundwater was passed through three 20 cm intact columns of McCloud Lake littoral sediment at a flow rate of "40 mL/day, which corresponded to the highest seepage rates observed during 1981. Three additional columns received the same groundwater

acidified to pH 3.5 with H2SO4 (designated "low pH" treatment). A third set of three columns received groundwater acidified to pH 6.2 (through week six) and then acidified to pH 4.0 to observe the effects of rapid acidification. Following a one-month period of pH stabilization, the eluate from each column was collected weekly and analyzed for sulfate, chloride, and nitrate (ion chromatography), alkalinity, major cations, and pH.

Microcosm Experiment

The influence of sediments on the composition of overlying lakewater was simulated in 10 L microcosms containing 1.0 L of







Table 3-2. Synthetic groundwater used in seepage experiments.

COMPOSITION OF INFLOW WATERa
Constituent Study site b Low pH Intermediate pH inflow High pH
groundwater inflow Before week 7 After week 7 inflow
pH 6.86 3.47 6.22 4.00 6.56 Ca2+ 3.42 3.68 3.44 3.44 3.65 Mg2+ 1.09 1.25 1.20 1.31 1.17 K+ 1.92 2.21 2.21 2.17 2.22 Na+ 4.09 4.31 4.31 4.12 4.26 HCO3 13.7 0.00 NA 0.00 13.3 SO42- 3.6 35.0 13.3 22.6 3.8 C1 5.5 5.9 5.5 6.01 5.74 NO3 0.33 0.33 0.40 0.33 0.42 avalues in mg/L (except pH).
bGroundwater from acidified forest plots.






24


McCloud Lake littoral sediment plus 9.0 L of synthetic Lake McCloud water (based on 1981 data). Following a one week period of pH stabilization, duplicate microcosms were acidified to pH 5.0, 4.5, 4.0, and 3.5 using 1.0 N H2SO4; an additional set of duplicates was used as controls. Repeated additions of acid were required to maintain the

desired pH levels. Acid inputs were based on pH values determined immediately before each addition; the vol ume of H2SO4 added on each

date was sufficient to reach the desired pH level in the absence of neutralization. Samples collected after 40 days (prior to the addition of labelled algae and the beginning of the decomposition experimentsee Chapter 5) were filtered and analyzed for alkalinity, sulfate and chloride, major cations, and pH. Components of buffering were inferred by comparing the chemical composition of the microcosms at 40 days with the initial composition.

Littoral Mesocosms

A mesocosm experiment conducted in the littoral region of Lake McCloud to evaluate the effects of pH alterations on biological processes (see Chapter 5) also served as a preliminary experiment on insitu pH buffering. During March, 1981, three 4.0 m diameter enclosures (designed after Landers 1979) were placed in the lake at a depth of rJ1 meter and anchored firmly to the bottom with wooden stakes to

enclose an area of 12 m2. One enclosure was acidified to pH 3.6 with 1.0 L aliquots of 0.72 N H2SO04 while a second received 1.0 L aliquots of 0.1 N NaOH to raise its pH to 5.6. A third bag received no acid or base and served as a control. Acid or base was added weekly with the intention of reaching the desired pH level within one month. Continued additions of acid or base were required throughout the 26 week






25


study to reach or maintain the desired pH levels. During this experiment, samples were analyzed weekly for pH, major cations, sulfate, chloride, and major nutrients.

Pore Water Profiles

Littoral and profundal cores were obtained from McCloud Lake using a Livingston-type corer during Febuary, 1983. Three cores (3.8 cm interior diameter) from each site were dissected into 2.b cm seyments in a glove box that had been purged with 02-free nitrogen gas. Individual segments were placed into centrifuge tubes, treated with

0.1 mL 2 N zinc acetate and 0.06 mL NaOH to precipitate ZnS (APHA 1981), and centrifuged for 20 minutes at 15,000 rpm. The supernatant

was filtered and analyzed for sulfate, chloride, and nitrate by ion chromatography.

Results and Discussion

Sediment-Water Batch Studies

Magnitude of neutralization

Results of the sediment batch studies are shown as buffering capacity curves for the McCloud Lake littoral and profundal sediments

(Figure 3-2). Each point in Figure 3-2 represents the neutralizing capacity for a given pH, as calculated from equation 3-2. The profundal sediment had far more neutralizing capacity than the littoral sediment, as seen in the much steeper slope for the profundal sediment. To facilitate comparisons of neutralizing capacities among sediments, neutralizing capacities were computed over defined pH ranges. Thus, the NC4.5-5.0 is the neutralizing capacity, expressed

as meq/100 g, over the pH range 4.5 to 5.0 and the NC5.0-5.5 is the






26








15





10




5


.+














1 0 1
-I
I-

5 -O Profundal
0 ~ Littoral




3.0 4.0 5.0 6.0 7.0 8.0

pH Figure 3-2. Titration curves for McCloud Lake sediments.






27



neutralizing capacity between pH 5.0 and pH 5.5. As seen in Table 3-3, the profundal sediments typically have NC4.5-5.0 values of 8-10 meq/100 g while the littoral sediments, which are considerably more sandy, have much lower buffering capacities (NC4.5-5.0 =0.6-1.2 meq/100 g). Although there are not enough data points for sediments having intermediate levels of volatile solids to permit formal regression analysis, it is clear that the volatile solids content, or some factor associated with volatile solids content (such as clay fraction) is strongly related to the buffering capacity of sediments. Cation exchange

Changes in cation concentrations during titration of the McCloud Lake sediment (Figure 3-3) suggest that ion exchange was the major buffering mechanism. Although mineral dissolution and precipitation could account for these results, minerals containing these cations are probably not present in McCloud Lake. Furthermore, the H+ neutralization occurred rapidly and was essentially complete within a few hours. This behavior is typical of cation exchange, whereas mineral dissolution, especially of clay minerals, is generally much slower.

When NaOH was added, protons on sediment exchange sites were replaced by cations from solution (primarily Ca2+ and Mg2+), and their solution concentrations decreased. When H2SO4 was added, base cations on sediments surfaces were replaced by protons, and cation concentrations in solution increased. Calcium and magnesium exchange accounted for over 50% of the total buffering capacity of McCloud Lake sediments throughout most of the pH range (Figure 3-4). Exchange of Na+ was significant at pH levels over 6 when NaOH was the source of OH- in the experiment. Thus, some of the Na+ added displaced protons on the






28

Table 3-3. Buffering capacity of some Florida and Wisconsin lake
sediments.


Acid neutralizing
capacity
meq/100 g Volatile
pH pH solids Sand Dry weight Sediment 4.5-5.0 5.0-5.5 % dry wt % dry wt % wet wt


Littoral

Anderson-Cue, 1.15 1.0 18.6 30.2 45.8
Fla.

Adaho, Fla. 0.6 0.5 9.0 41.7 47.6 Johnson, Fla. 0.6 2.1 11.3 38.0 46.2 Geneva, Fla. 1.1 12.4 27.5 37.0 McCloud, Fla. 1.0 0.5 10.4 31.2


Profundal

McCloud, Fla. 8.3 2.2 73.3 0 8.6 Clara, Wis. 8.7 3.3 62.6 0 5.8 McGrath, Wis. 10.3 7.0 37.1 0 3.2 Sand, Wis. 9.0 7.0 67.3 0 5.4






29



1000











_J
Cr 1100
x
z
0


z

z
o
0
U

-IJ S10


e Co2+ o Mg2+

a dis. Al
SA6 No

2 I I I I I
3.9 4.4 4.9 5.4 5.9 6.4 pH Figure 3-3. Cation concentrations versus pH in batch neutralization experiment with McCloud Lake profundal
sediment.






30












100
Unknown Unknown Z 0 Unknown

LL 80 L2+ M 70 Mg Mgz
0

Zo
O I- C S 50 /

Z 3o
40



+No
20 No
Z
0 10 M3: free -AI W fre Al /// CL 0
3.8 4.2 4.6 5.0 5.4 5.8 6.2 pH

Figure 3-4. Contribution of cations to total neutralizing
capacity of McCloud Lake profundal sediment.





31



sediment surfaces, which contributed to the neutralization of the added OH-. At pH 6.0, Na+ exchange accounted for nearly 50% of the

total neutralization capacity. Concentrations of dissolved aluminum increased significantly upon acidification, but aluminum solubilization contributed less than 10% to the total neutralizing capacity

(even assuming a charge of +3) throughout the pH range 4.0-7.0.

Finally, a portion of the buffering could not be attributed to cation exchange or mineral dissolution. This component increased to 30% of the total neutralizing capacity at pH 6.6 and probably resulted from the dissolution of humic acids. This hypothesis is supported by the observation that the sediment-water solutions became

visibly colored at pH levels over 6.0.

Experiments with sediments from other lakes in the Trail Ridge Region and from northern Wisconsin lakes show that exchange of Ca2+ and Mg2+ is the major mechanism of pH buffering in all of these lakes. Calcium exchange accounted for 41-72% of the total buffering in these lakes, while Mg2+ exchange accounted for 20-40% of the buffering in

the Florida lakes but only 8-9 % of the total buffering in the three Wisconsin lakes (Table 3-4).

Aluminum solubility

Although dissolution of aluminum accounted for less than 20% the total neutralizing capacity of the Florida lakes (again assuming a charge of +3/mole), aluminum solubility is of interest because aluminum toxicity is a major problem associated with lake acidification.

Previous studies (Johnson et al. 1981; Driscoll et al. 1982) have shown that A13+ solubility is controlled by gibbsite, although several






32

Table 3-4. Components of buffering capacity for selected Florida and
Wisconsin lakes titrated to pH 4.


% of total neutralizing capacity

Final
Final 2+ 2+ + 3+ Lake pH Ca Mg Na+ K Al ECa+Mg Johnson 4.46 35.7 39.3 1.2 2.5 75.0 Geneva 4.83 63.3 24.7 2.3 11.1 88.0 Anderson-Cue 4.07 46.5 22.0 1.2 18.4 68.5 Adaho 3.93 46.2 40.0 <.1 2.6 86.2 McCloud:
center 3.96 41.2 20.9 0.5 15.5 62.1 littoral 3.90 71.8 20.7 .25 .25 92.5 Clara 3.61 44.0 8.9 1.1 1.2 52.9 Sand:
center 3.72 49.6 8.3 0.5 1.3 57.9 McGrath 3.44 43.3 7.6 0.2 1.0 50.9





33



other aluminum-bearing minerals, including kaolinite, amorphous aluminum hydroxide, and feldspars potentially could control aluminum solubility. Nordstrom (1982) have recently shown that several sulfatebearing minerals, notably jurbanite (A1(S04)(OH)) and alunite (K(S04)(Al)3(OH)6), may control aluminum solubility in highly acidic water such as mine drainage.

Total aluminum concentrations in the batch experiments with Florida sediments increased with increasing acidity below pH 5 but

were relatively constant (10-5 to 10-6 M) above pH 5 (Figure 3-5). Also shown in Figure 3-5 are total aluminum concentrations (sum of hydrolysis species) predicted from the solubility of gibbsite (log K =

-33.2 from May et al. 1979) and from the solubility of kaolinite (log

K=-38.7, Stumm and Morgan 1981). Since silica levels remained nearly constant thoughout the pH range, the mean concentration of 44 uM was

used to compute kaolinite solubility. These data clearly show that total aluminum in the sediment titrations was generally above the levels predicted from the dissolution of these two minerals. The most likely reason for the discrepancy between measured total aluminum and predicted solubility is that the measured total aluminum includes

organic complexes and fluoride complexes that were not accounted for in these solubilitiy calculations. Driscoll et al. (1982) have shown that organic Al complexes accounted for an average of 44% of the total monomeric aluminum and that Al-F complexes accounted for 29% of the total monomeric aluminum in a group of Adirondack lakes. Fluoride

complexes are not likely to be important in McCloud Lake because the lakewater contains very little fluoride (<0.003 meq/L). However, water in the experimental flasks often became visibly colored at pH






34




3.0


3 _A Johnson V Geneva
0 Anderson-Cue & Adaho
4.0 0 McCloud central







5.0 A\







6.0 V A




4.0 4.5 5.0 5.5 6.0 6.5 7.0
pH
Figure 3-5. Aluminum concentration versus pH in sediment neutralization experiments for five softwater Florida lakes.






35



levels above 5, indicating the solubilization of humic acids. These humic acids undoubtedly formed Al complexes that kept total aluminum levels in solution far above levels predicted on the basis of inorganic complexes. Without further information on the extent of organic complexes, the control of Al solubility cannot be determined precisely.

Solubility calculations for the Trail Ridge lakes, using data from Brezonik et al. (1983b), were conducted with the assumption that total aluminum represented free A13+. These calculations show that several minerals, including kaolinite, gibbsite, and alunite, may control aluminum solubility. The possibility of alunite precipitation is particularly interesting, since this process would be a sink for

sulfate as well as a H+-neutralization mechanism. However, alunite precipitation would be accompanied by decreased concentrations of K+ and S042- in the batch studies and these ions were conservative for the McCloud Lake sediments.

Sulfate adsorption

The sulfate adsorption experiment showed that sulfate recovery was at least 100% of sulfate added for both littoral and profundal sediments (Figure 3-6); the addition of chloroform to the bottles had virtually no effect on sulfate recovery in either blanks or treatment bottles. Although it is not clear why sulfate recovery was usually slightly greater than 100%, this experiment shows that sulfate adsorption or other abiotic sulfate-reducing processes, such as the precipitation of alunite or other sulfate-containing mineral (See discussion

of aluminum solubility), were unimportant as neutralizing mechanisms within the pH range of this experiment (4.0-5.5).






36

















2.5



2.0
/
N/

1.5 / = LTTRAL
Sc/ Y= LITTORAL + QCL3 / / Bl= PROFUDAL 0= PROFUNDAL + cICL3
1.0 /
8 /


- 0.5 /
/

0

0.5 1.0 1.5 2.0 2.5


[S-] ADIEED, TEQ/L



Figure 3-6. Sulfate recovery in batch acidification of McCloud Lake sediments.






37


Seepage Column Experiment

Neutralization of H+ occurred in the upflow column experiment to the extent that the eluate from all three sets of columns had pH

values between 5.0 and 7.0 following the stabilization period (Figure 3-7). Changes in the composition of major ions in the column eluates (Figure 3-8) occurred in all three sets of columns. In the high pH columns (inflow pH = 6.6), concentrations of Ca2+, Mg2+, HC03 S042and N03- decreased, but pH remained nearly constant. During weeks 5-8, the reduction in divalent cations (Ca2+ and Mg2+) of 0.20 meq/L was similar in magnitude to the reduction in alkalinity (0.18 meq/L). These results can be accounted for by the following mechanism:


nHCO3- + Mn+ + nH+-X ----> nH2CO3 + M-X (3-3)


The inference to be made is that alkalinity and divalent cations leached from the soil by acidic precipitation are removed from seepage water as it passes into the lake through the littoral sediments.

In the low pH columns (inflow pH = 3.5), nearly all (99%) of the influent protons were neutralized by the sediment. Calcium, which was initially retained by the columns (weeks 5-7), later was leached from the columns (Figure 3-8). Although magnesium was leached from the column throughout the study, the extent of leaching increased by week 14. Calcium retention nearly balanced magnesium leaching during weeks 5-8; cation exchange was therefore unimportant in pH-buffering. Of much greater significance was the reduction in sulfate levels. The average loss of sulfate (0.410 meq/L) during weeks 5-8 accounted for 86% of the sum of proton loss plus alkalinity gain. Although sulfate adsorption potentially could account for the observed results, sulfate















7.0




6.0


SInflow pH:O S* -3.4 u 5.0
0--- 06.2/4.0

---- 6.7
4.0
4.0 I I I I I I I I I I I I I I 1 2 3 4 5 6 7 8 9 10 11 12 13 14 WEEK

Figure 3-7. pH of eluate passing through 20 cm cores of littoral McCloud Lake sediment.








+0.5
INFLOW pH: 6.7

H+ Ca2+ Mg2+ HCO3- S04 NO3
0.0
5 6 7 8 14 LLJ.
Week
-0.5
-j +0.5 INFLOW pH: 6.2 (through week 6); then 4.0 (weeks 7-14) H+ 2+ 2+ HCO 2- NO H Ca Mg HCO3 SO4 NO
_4 3 D 0.0 I-L- ..

0 5678 14 SWeek oz -0.5z +0.5 INFLOW pH: 3.5 + 2+ 2+ 2H Ca Mg HCO3 S4 NO3
0 .0 -- -



-0.5
-05 6 77 14
Week



Figure 3-8. Effect of littoral sediments on groundwater eluate in
seepage experiment.





40




adsorption was found unimportant in the sulfate adsorption study (see Figure 3-6). Biological sulfate reduction is a more likely mechanism; sulfate entering the columns was either reduced to H2S and lost to the atmosphere or precipitated in the columns as FeS.

Both sulfate reduction and cation exchange were important H +buffering mechanisms in the intermediate pH columns. During weeks 56, a loss of sulfate (0.14 meq/L) was exactly balanced by a loss of cations and the pH remained constant. Upon further acidification of the inflow to pH 4.0 during week 6, the loss of sulfate increased to 0.31 meq/L and was accompanied by a comparable increase in the loss of protons. Thus, even with rapid acidification the pH of the eluate remained above 5.0. By week 14 however, sulfate reduction was diminished and the pH of the eluate decreased slightly.

Concentrations of Na+, K+, or Cl- were unchanged in all three sets of the columns. These results support the conclusion from the batch studies that monovalent ions do not contribute to ion exchange buffering in the sediments.

It is interesting to note that N03- concentrations were decreased by 90 % in all columns. This reduction could be the result of either nitrate assimilation or dissimilatory nitrate reduction. Although both reactions consume protons (see Figure 3-1), the level of N03- in the column inflows was so low (< 40 ueq/L) that the effect of nitrate loss mechanisms on eluate pH was minor.

Sediment-Water Microcosms

The McCloud Lake sediment-water microcosms exhibited a substantial capacity to neutralize acid added to the overlying water.





41



Maintenance of the acidified microcosms at the desired pH levels required continued inputs of H2SO4 during the 20-week experiment. As seen in Figure 3-9, predicted pH values, calculated from the cumulative acid additions, were far lower than measured pH levels. By the end of the experiment over 90% of the acid added to each microcosm had disappeared (Figure 3-9). An analysis of major ions during week 7 (prior to the addition of 15N-labelled algae to begin the decomposition experiment described in Chapter 5) was used to determine the components of pH buffering. Alkalinity was not measured due to time constraints but was inferred from the ion balance of other major constituents. As seen in Table 3-5, cation exchange was responsible for over 60% of the total buffering in all of the microcosms. Calcium exchange accounted for 45-60% of the buffering, while magnesium exchange comprised 15-25% of the total buffering. Sodium and potassium concentrations changed little and were unimportant in buffering. Sulfate reduction accounted for 25-40% of the total buffering in the

acidified microcosms, although sulfate concentrations actually increased slightly ("15%) in control microcosms. This small increase may have resulted from sulfate oxidation but also may have been the result of experimental error. In the acidified microcosms, cation exchange plus sulfate reduction accounted for 81% to 104% of the observed buffering. As with the groundwater seepage experiment, the importance of sulfate reduction as a buffering mechanism increased with decreasing pH and increasing sulfate concentration. Littoral Mesocosms

The littoral mesocosms also exhibited a substantial capacity for pH buffering (Figure 3-10). During the first 14 weeks of treatment,






42



6.0
Measured


5.0
predicted NOMINAL pH: 5.0
-I-0
4.0 I --- I6.0



5.0 Predicted NOMINAL pH: 4.5



4.0 I

6.0
Measured


5.0 NOMINAL pH: 4.0 Predicted
4 0 I I -1 *-*-- I1


6.0



5.0 Measured


4.0 .4Predicted NOMINAL pH: 3.5
4.0


3.0 I I
0 5 10 15 20 WEEK

Figure 3-9. pH of acidified sediment-water microcosms.







Table 3-5. Buffering components in sediment-water microcosms during first 40 days of acidification.

2- Ca2+ Mg2+ Na+ K+ SO4 Mg Na K Loss of H % % % Sum of buffering Micro- + gain of contri- contri- cotri- contri- contri- components, % cosm HCOS,meq/L meq/L bution meq/L bution meq/L bution meq/L bution meq/L bution total buffering

3.5 .9192 -0.141 15.3 0.452 49.2 0.156 16.9 -0.013 1.4 0.005 0.5 80.5

4.0 .4360 -0.126 29.0 0.248 56.8 0.094 21.6 -0.020 4.5 0.003 0.7 103.6

4.5 .2143 -0.048 22.6 0.129 60.1 0.050 23.4 -0.024 -11.2 0.002 1.0 95.9

5.0 .1165 -0.050 43.3 0.052 44.8 0.022 19.1 -0.021 -18.2 0.004 3.3 92.3

5.5 .0959 0.021 -21.6 0.052 54.4 0.024 25.0 -0.016 -16.2 0.006 5.7 47.3











ADDITIONS )F ASE 1 cn =0.1 MOLES OH




ADDITIONS OF ACID. 1 mrn = 0.7 MO ES H

5.5
A Acidified mesocosm e Control mesocosm oNeutralized mesocosm
5.0 Littoral mesocosm



4.5



4.0



3.5
2 4 6 8 10 12 14 16 18 20 22 WEEK

Figure 3-10. Additions of acid and base and pH of McCloud Lake littoral mesocosms.






45




0.46 meq H+/L was added to the acidified mesocosm, and 80% was neutralized. While the areal buffering rate of the acidified mesocosm

(1.2 meq/m2-yr) during the first 14 weeks was comparable to the areal buffering rate in the pH 3.5 laboratory microcosms (1.3 meq/m2-yr), the nature of the buffering was different. Cation exchange, which accounted for 65% of the buffering in the pH 3.5 microcosms, was apparently unimportant in buffering of the mesocosms. Although calcium levels fluctuated considerably during the mesocosm study, there was no sustained increase in calcium or magnesium with acidification (Figure 3-11). The neutralization of added protons was paralleled by a loss of added sulfate, suggesting that sulfate reduction probably was responsible for the neutralization (Figure 3-12).

In the neutralized mesocosm, repeated additions of NaOH failed to

cause a sustained increase in pH (Figure 3-10). As in the acidified mesocosm, calcium levels fluctuated considerably, but the expected decrease of calcium and magnesium concentrations did not occur (Figure 3-11). Sodium, however, did disappear, and its rate of disappearance closely paralleled the neutralization of added OH- (Figure 3-13). The loss of Na+ is disturbing because Na+ was conservative in the batch studies below pH 5. The sodium loss may possibly have occurred as the result of water exchange between the mesocosm and the lake, although it is unlikely that the rapid disappearance of OH- observed during the first month of base additions occurred as the result of water exchange. By week 14, the loss of sodium was equivalent to 71%

of the loss of OH-.






46








.J
CY

z0.20
o pH 3.5 I
S0.10

(
1-0.00
-J


0.20
o pH 4.6

0.10

0



- 0.20
z pH 5 C-- Ca Ca 2+ 0oMg 0.10 I



LU
0.00
1 2 3 4 5 6 7 8 9 19 11 12 13 14 WEEK Figure 3-11. Calcium and magnesium concentrations in McCloud Lake littoral mesocosms.










100





O s







50 o I U
L
1 2 8 12 1
U



so 2

+4








0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 WEEK
Figure 3-12. Loss of sulfate and protons in the acidified mesocosm.









100 - - -.- -90 /


80

I
70





70
I
I
60
co
I
50 I
I I
40


_ 30
I -- Na+
I
20 *------ OH
I I
10
I


1 2 3 4 5 6 7 8 9 10 11 12 13 14 WEEK Figure 3-13. Loss of sodium and hydroxide ions in neutralized mesocosm.






49



Precautions taken to prevent mixing of lakewater with the water in the experimental enclosures included anchoring the bottom with

stakes and providing a floating collar to minimize wave action, but some exchange undoubtedly occurred. Furthermore, inputs of groundwater and precipitation were not determined and presumably had some effect on the composition of water in the enclosures, even during the relatively short experimental period. In future experiments, it would be useful to use an inert tracer (such as bromide) in control mesocosms in order to ascertain the extent of water exchange between the mesocosms and the open lake.

Pore Water Profiles

The pore water profiles show that sulfate reduction occurs not only in laboratory microcosms but in Lake McCloud itself. As seen in Figure 3-14, sulfate pore water concentrations for the three profundal cores decreased from 150 ueq/L in the overlying water to < 15 ueq/L below 10 cm. The flux of sulfate to the sediment in this part of the lake was calculated from Fick's law:

F = Dc(Ci Ci+1)/z (3-4) where Dc = diffusion coefficient, cm2/sec,
Ci= concentration at depth i, mg/L, and
Ci+1 = concentration at depth i+1.

According to Lerman (1978), the effective diffusion coefficient, Dc, is approximately equal to DoG2, where Do is the bulk diffusion coefficent and 0 is the porosity. Li and Gregory (1974) reported that Do for SO42- = 8.9 x 10-6 cm2/sec at 18 o C. If the porosity is estimated from the water content (0.9), Dc is 7.2 x 10-6 cm2/sec. From d = 0 to d = 10 cm, the concentration gradient is 136 ueq/L, so the flux rate of S042- to the profundal sediment is approximately 225






50






0
0 A
5 A. PROFUNDAL o = Core 1
ao Core 2
10 o = Core 3
x = Lake water
Solid line represents mean

LUJ
C 0

20 1 I I
0 50 100 150 200
SULFATE CONCENTRATION, UEQ/L


0 0 A


5 B. LITTORAL 0 =Core 1
0 D a Core 2 10 =Core 3
o


15 S0

20 0
0 100 200 300 400
SULFATE CONCENTRATION, UEQ/L


Figure 3-14. Sulfate profiles in pore water of McCloud Lake
sediments, February, 1982.






51



meq/m2-yr. This rate of sulfate reduction is 54% of the wet-only deposition of 420 meq/m2-yr for the annual period May, 1981-April, 1982.

Individual pore water profiles in the littoral sediment pore water were very irregular (Figure 3-14), with concentrations ranging from 56 to 320 ueq/L. These data suggest that distinct pockets of sulfate reduction and oxidation may occur in the sandy, littoral sediments. The irregularity of sulfate reduction was confirmed by the irregular pattern of mottling seen on silver-coated rods inserted in sediment 24 hours prior to collection of the littoral cores. While some reduction and reoxidation may occur in individual core segments, the ratio of S042-/C1- in the pore water profile is nearly constant with depth (1.21) and is almost identical to the S042-/C1- ratio in the lake water (1.24). Chloride concentrations in the pore water were comparable to lakewater concentations but higher than concentrations in the more dilute in-seepage (see Chapter 4), suggesting that water was moving out of the lake when these cores were taken. The movement of water may have obliterated sulfate profiles that formed during the sumer, resulting in the irregular profiles observed. Thus, while this pore water study confirms the occurrence of sulfate reduction in the profundal sediments, there appears to be little evidence of sulfate reduction occurring in the littoral zone.

Conclusions

These experiments demonstrate that reactions at the sedimentwater interface are potentially important in neutralizing inputs of H+ to McCloud Lake. The major mechanisms of H+ neutralization observed






52


were cation exchange and sulfate reduction. Cation exchange, in which

Ca2+ and Mg2+ were displaced by H+ on sediment surfaces, provided a neutralizing capacity of up to 10 meq/100 in the profundal sediments.

Exchange of Na+ and K+ was unimportant, as was sulfate adsorption. Although aluminum was solubilized in the batch experiments, aluminum

solubility was relatively unimportant as a neutralizing mechanism. Evidence of sulfate reduction was obtained in the seepage column experiments, in both the mesocosm and microcosm experiments, and in the pore water profiles. This mechanism is interesting because it means that sulfate is not a conservative ion in softwater lakes, as is

commonly believed, and because sulfate reduction is a proton-consuming process that may contribute to the neutralization of acid precipitation.















CHAPTER 4
McCLOUD LAKE MASS BALANCE

Introduction

In the previous chapter, results of lab experiments showed that McCloud Lake sediments are capable of neutralizing acid by several mechanisms. Cation exchange, particularly the replacement of divalent cations (Ca2+ and Mg2+) by protons resulted in the neutralization of up to 10 meq/100 g dry sediment. Evidence of sulfate reduction was obtained by groundwater seepage and sediment-water microcosm

experiments and from in situ pore water profiles of sulfate.

The objective of this chapter is to evaluate the magnitude of inlake neutralization processes in McCloud Lake by constructing mass balances for major ions. Based on the results from the previous chapter, two hypotheses were generated:

1) cation exchange at the sediment-water interface results in the

enrichment of the overlying water with base cations, particularly Ca2+ and My2+, and

2) sulfate reduction in the sediments reduces the quantity of sulfate in the overlying water and neutralizes an equivalent amount of acidity.

A third hypothesis is that assimilation of NO3- by phytoplankton and macrophytes, which consumes protons (Brewer and Goldman 1976), is a significant sink for acidity.


53






54


Mass balances of major ions entering and leaving McCloud Lake during the period September, 1981-August, 1982 were used to evaluate these hypotheses. Fluxes measured or estimated in the mass balances included wet precipitation, aerosol and gaseous dry deposition, and groundwater seepage to and from the lake. Further evidence of neutralizing mechanisms was obtained by evaluating historical water chemistry data for McCloud Lake collected during 1968-69 (Brezonik and Shannon 1971), 1978-79 (Brezonik et al. 1983b), and 1981-82 (present study).

Background

Mass balance models have been widely used to evaluate the role of watersheds in neutralizing acid precipitation (Johnson et al. 1981; Galloway et al. 1980; Thomson 1982). Cation exchange and mineral dissolution have been recognized as the most important neutralizing mechanisms in watershed soils, and several regional scale sensitivity models are based exclusively on mineral weathering reactions (e.g., Henriksen 1980). Assimilation of nitrate by vegetation in watersheds may also contribute to the neutralizing capacity of forested watersheds. As noted in Chapter 3, the assimilation of nitrate produces hydroxide ions that consume protons. Several studies have shown that most of the nitrate entering forested watersheds is retained with an equivalent number of protons (see Impact Assessment Group 1983). Finally, sulfate adsorption may be an important neutralizing process in some watersheds, particularly those with sesquioxide-rich subsoils (Johnson et al. 1980). In addition to permanent (ligand exchange) sulfate adsorption, reversible adsorption is believed to account for






55


short-term stabilization of S042- concentrations in watersheds (Chen et al. 1979; Christopherson and Wright 1981).

Few mass balance models have been compiled for sensitive, softwater lakes and the role of in-lake processes on the H+ balance of lakes is not well understood. However, several recent reports suggest that anion-consuming processes may be important in regulating lake pH. The potential role of sulfate reduction in neutralizing inputs of acid

precipitation to softwater lakes was elucidated by Schindler et al. (1980) and Cook and Schindler (1983). Following the artificial acidification of Lake 223, these authors found that the reduction of sulfate in the anaerobic hypolimnion and littoral sediments accounted for most of the in-lake neutralization in the first two years of acidification (Schindler et al. 1980) and consumed 50% of the total sulfate input during the five years following acidification (Cook and Schindler 1983).

Kilham (1982) reported that acid precipitation may actually increase the alkalinity of non-sensitive lakes. His hypothesis is that the protons in precipitation are neutralized by cation exchange and mineral weathering. Since the consumption of acidic anions (N03 and S042-) also consumes protons (or produces OH-, i.e. alkalinity), the addition of nitric and sulfuric acids to non-sensitive watersheds can result in an increase in alkalinity production. Kilham (1982) demonstrated that nearly all of the NO3- and 65% of the SO42- entering the watershed of eutrophic Weber Lake, Michigan, was retained by the lakewatershed system and concluded that nitrate assimilation and sulfate reduction were sufficient to neutralize acid precipitation. Acid precipitation also appeared to have increased the mineralization rate






56



in the watershed and may have caused a doubling of alkalinity observed

over the past 30 years. Kelley et al. (1982) also postulated that hypolimnetic OH- production resulting from anion-consuming processes may be sufficient to neutralize inputs of acid precipitation to

eutrophic lakes.

Finally, in what appears to be one of the only mass balance models reported for an undisturbed, softwater lake, Wright and Johannessen (1980) found that 30% of the H+ entering Langtern Lake, Norway

(6 keq/km2-yr) was neutralized together with 20% (9 keq/km2-yr) of the SO42-. These authors concluded that sulfate reduction may have accounted for the observed H+ and S04 2- depletion.

Methods

Wet Precipitation

Wet-only precipitation was collected using an Aerochem-Metrics Model 101 sampler placed on a raft in the center of McCloud Lake. This sampler is designed so that the wet sample bucket is open only during precipitation events and is tightly sealed with a lid between events to prevent evaporation and contamination by birds, insects, and other debris. Samples were collected monthly and analyzed for pH, major ions, and nutrients.

Dry Deposition

Ambient air concentrations of aerosols (S042-, NH4+) and gases (SO2, NO2, and HN03) were measured during 20 24-hour collection periods during August and September, 1981. Dry deposition rates of these constituents were computed using the equation (Chamberlain and Chadwick 1953)





57



F = Vd x C, (4-1) where F= areal deposition rate,
Vd = deposition velocity, and
Cz = concentration at a reference height, usually one meter.

Dry deposition rates of other aerosol constituents (Ca2+, Mg2+ K+, Na+, and Cl-) were estimated from two northern Florida data bases. Edgerton (unpublished data) recently conducted a study of dry deposition throughout northern Florida that included one remote site similar to McCloud Lake. The study of Brezonik et al. (1983b) included wet and dry deposition measurements for four inland Florida sites and bulk deposition measurements for 25 sites, including four near McCloud Lake (Gainesville, Jasper, Hastings, and Waldo). Two methods were used to estimate dry deposition of aerosol constituents at McCloud Lake from these data. First, the ratios of wet to total deposition for the four wet/dry collectors were used to estimate dry deposition at McCloud Lake from measured wet deposition. The second method was to subtract McCloud Lake wet deposition from bulk deposition measured at the four local bulk collectors.

Seepage

Seepage flow to and from McCloud Lake was measured using 22 seepage meters (Fellows and Brezonik 1981) placed along six transects perpendicular to the shoreline (Figure 4-1). Flows were measured once a month. When the direction of flow was towards the lake, samples of the inflow groundwater were collected for analyses of major ions and nitrogenous species. Flux rates of water and ions were calculated by dividing the lake into five concentric rings having boundaries that paralleled the shoreline and were equidistant from adjacent seepage






58


2




/0e\

// / \ \

/ / ii

i / / 1 I // 1 / ** 3 6 */ I I /
/ i i II

\I I lI / I I







4
fl II / I I I I '
N






/Fi 4 M L I / N

I IJ I
f t t lI I I
II//I II -/ I
I ~Il
I..., I I
/ 1 1.,,I






I II
--. / / I /..-.- _/-.
II, '









Boundaries between regions I--Figure 4-I. McCloud Lake seepage meter transects.





59


meters along each transect. The outer three rings were further divided into 6 subdivisions whose boundaries were equidistant from adjacent transects. Thus, a total of 20 regions were formed, each represented by one, or, for the two inner regions, two seepage meters (Figure 41). The flux of constituent i was computed from the equation
n
Fi = JCi,j x Qj (4-2)
j=1
where Cij concentration of constituent i in seepage region j,
Qj flow in region j. (Sign convention: "+" = inflow;
"-" = outflow).

Groundwater Wel 1 s

Two to four wells (2 diameter PVC pipes) were placed in the ground along each transect (see Figure 1-1) in order to permit sampling of the local groundwater. During the second half of 1982, samples from these wells were collected for analyses of major ions and nutrients.

Lake Storage

A bathymetric map of McCloud Lake was used to construct stagevolume and stage-area relationships. (See Figure 1-1). Stage measurements were made 2-6 times per month and the mean stage reading was used to compute lake volume and area for the month. Water samples collected monthly at 1-meter intervals near the center of the lake were analyzed for major ions and nutrient species. Since the lake never stratified, storage calculations were based on mean concentrations of samples collected at 3-4 depths. Evaporation

Pan evaporation measured at Lisbon, Gainesville, and Lake City (Climatological Abstracts 1981 and 1982) was used to compute a distance-weighted mean representative of evaporation at McCloud Lake:






60



n n
Edwm = -(D/di x Ei)/ (D/di) (4-3)
i=1 i=l
where E = distance-weighted mean of pan evaporation, cm/month,
dw = distance to station i,
D = total distance, i.e., di, and
Ei = pan evaporation at station i, cm/month.

A pan coefficient of 0.70 was used to estimate lake evaporation from the Class A pan evaporation data (Linsley et al. 1975).

Results and Discussion

Water Budget

Components of the McCloud Lake water budget are shown in Figure 4-2 (see Table A-i). To check the accuracy of the water balance, monthly precipitation (P), evaporation (E), and net seepage (G) were used to model the change in storage (AS) and the new storage (S) for each month, beginning with the measured storage in August, 1981:

ASi = Pi Ei + G (4-4)

Si = Si-1 + ASi (4-5) where i = month.

Although the computed S often differed from the measured S for individual months (Figure 4-2), computed S and AES agree well with measured values for the entire year. At the end of the year, computed storage was only 2% less than the measured storage, and the computed annual change in storage (+ 17.45 x 103 m3) was 15% less than the measured change in storage of 20.47 x 103 m3. The difference between predicted and measured AS was only 3.6% of the total water input (precipitation + in-seepage).

Precipitation accounted for 90 % of the total water input, while in-seepage accounted for the remaining 10%. It should be noted that





61


the contribution of seepage to the water budget was far less than suggested by the watershed:lake surface area ratio of 20:1. Because

of the sandy soils, gentle slopes, and a lack of defined stream channels, precipitation falling on most of the watershed passes directly to a regional water table rather than to a lake or outflow stream.

Although precipitation during the model year was 6% above the 40 year mean, the previous year was very dry and had an annual precipitation of only 57% of the 40 year mean. Thus, seepage during the first seven months of the model year was outward, resulting in recharge of a depleted water table. During the last five months, continued precipitation and a rising water table resulted in seepage

flows into the lake (Figure 4-2).

Seepage occurred mainly in the sandy littoral area. As shown in Figure 4-3, seepage rates during both outflow periods (e.g., October, 1981) and inflow periods (e.g., April, 1982) diminished with distance

from shoreline. This was particularly evident during the inflow period. During April, 1982, for example, the mean flow rate decreased from 1.1 cm/d at 3 meters from shore to nearly 0 cm/d at 25 m (Figure 4-3). However, since much of the total lake area was in the central

low-seepage zone, the central region of the lake was significant in terms of total seepage flow. Thus, 72% of the total outflow but only 4% of the total inflow occurred in the 71% of the lake within the interior of the 17.5 m contour.

The water residence time, based on total inflow (precipitation + seepage), was 1.7 years. Most of the water entering the lake evaporated (63% of the total inflow) and only 16% left the lake via outseepage (the remaining inflow resulted in a change in storage). The






62



12
PRECIPITATION
10

8

Co 6
~ 4

~ 2
0

10
0
EVAPORATION
8



6
S 06
C
4

2







S-4
SEEPCHANGE IN STORAGE
a 2

-4
-J
-65 __10
5 CHANGE IN STORAGE
5




-10 _- Measured
--j Predicted o -15
Sept. Nov. Jan. Mar. May July
Oct. Dec. Feb. Apr. June Aug.
1981 1982

Figure 4-2. Water balance for McCloud Lake, September,
1981, to August, 1982.






63






OCTOBER, 1981 (FLOW FROM LAKE)

0


Solid line shows mean
o 0



-1.0 I I I I I
0 10 20 30 40 50 DISTANCE FROM SHORE, M

2.0


APRIL, 1982 (FLOW TO LAKE)
1.5



1.0
Solid line shows mean


0.5








-0.5
0 10 20 30 40 50 DISTANCE FROM SHORE, M Figure 4-3. Seepage flow versus distance from shore.





64



residence time based on outflow was 9.6 years. Since volatilization of substances other than water entering the lake is minimal or nonexistent, the outflow-based water residence time should be the residence time for conservative ions. Chemistry of McCloud Lake, Past and Present

Data collected during 1968-1969 (Brezonik et al. 1969), 1978-79 (Brezonik et al. 1983b), and in the current study (Tables A-2 and A3) allow an evaluation of historical water chemistry trends for McCloud Lake. Before discussing these data, several problems concerning the evaluation of historical water chemistry trends must be considered. The most serious problem with comparing data collected over a span of 14 years is that certain analytical methods have changed significantly. While methods for analyzing cations were similar in all three studies (calcium, magnesium, sodium, and potassium

were analyzed by flame atomic adsorption), methods for collecting pH samples and for analyzing sulfate and chloride have changed. In the 1968-69 study, pH samples were col 1 elected in glass D.O. bottles "to prevent CO2 transfer with the atmosphere" (Brezonik et al. 1969, pg. 83). Although Kramer and Tessier (1982) have pointed out that soft

glass bottles may contribute 20 100 ueq/L alkalinity, the same authors concluded that the use of glass containers would not be be a problem in studies where analyses were conducted immediately following

sample collection. In the study of Brezonik et al. (1969), prompt (same day) analysis was done, as inferred by the stated concern for carbon dioxide transfer. Thus, comparisons of the 1968-69 pH data with the data from the two more recent studies (in which pH samples were





65


collected in distilled water-soaked linear polyethylene (LPE) bottles and analyzed within eight hours) probably are valid.

The most significant methodological change involves sulfate analyses. Sulfate was analyzed by the turbidimetric method in the 1968-69

study (APHA 1971) which, in addition to being an awkward technique, has a detection limit of only 1 mg/L. The reliability of the reported mean sulfate concentration of 2.2 mg/L is therefore questionable. Sulfate analyses in 1978-79 and in the current study were done by the automated methylthymol blue technique, modified to give a

detection limit of 0.1 mg/L. Methods of chloride analysis also changed during the 14 year observation period. Chloride in the 196869 study was determined by mercuric nitrate titration, using the "low

level" modification (APHA 1971), while in the recent studies chloride was determined by the automated ferricyanide procedure (APHA 1981), also modified to give a sensitivity of 0.1 mg/L. Both techniques, however, are generally free of interferences and considered reliable.

Although concentrations of chloride in McCloud Lake are so low that only 1 mL of titrant would be required to reach the endpoint with the mercuric nitrate titration, small bore burets were used to assure that

small volumes could be delivered accurately.

The excess of measured anions in the 1968-69 study (10% of the sum of ions) indicates that analytical errors were a problem (Table 41). Ion balances were more exact in the 1978-79 study (mean error = 5%) and in the current study ( mean error = 3.5 %).

A second problem in evaluating historical trends in acidification is that McCloud Lake undergoes substantial changes in volume as a result of long-term variations in precipitation. Between September,






66



1981, and August, 1982, for example, the volume increased by 20% because of increased precipitation following several drought years. These volume fluctuations cause changes in the concentrations of dissolved substances by dilution or evaporative concentration that may obfuscate long-term changes in water quality. One way to eliminate changes caused by dilution/concentration effects is to normalize the concentrations of dissolved ions to the concentration of a conservative ion such as chloride. Both concentrations and ion/chloride ratios are shown in Table 4-1.

As shown in Table 4-1, McCloud Lake has become more acidic during the past 14 years. Hydrogen ion concentrations have nearly doubled from 14 ueq/L (pH 4.9) in 1968-69 to 32 ueq/L (pH 4.5) in 1982, while sulfate concentrations appear to have increased from 104 ueq/L in 1968-69 to 142 ueq/L in 1978-79 and 173 ueq/L in 1981-82 (Table 4-1).

Reported values of sulfate in 1968-69 may be erroneously high due to analytical problems associated with the turbidimetric method (discussed above), and sulfate levels may have increased more than these data indicate. The excess of measured anions in the 1968-69 data suggests that reported values of one or more anions were too high or that reported cation values were too low.

Increased acidification has apparently resulted in enhanced leaching of base cations. Calcium concentrations in McCloud Lake have doubled from 30 ueq/L in 1968-69 to 66 ueq/L in 1981-82, while magnesium levels have increased by 34%, from 47 ueq/L in 1968-69 to 63 ueq/L in 1981-81. These data must be interpreted with caution because of the poor ion balance associated with the 1968-69 data. The





67


Table 4-1. Chemical composition of McCloud Lake, 1968 to present.


1968-69a 1978-79b 1981-82


Conc. Conc. Conc.
peq/L Ci/Clc peq/L Ci/C1c peq/L Ci/Cl H+ 14 0.08 19 0.13 32 0.19 Ca2+ 30 0.18 47 0.32 66 0.39 Mg2+ 47 0.28 51 0.35 63 0.37 K 6 0.04 15 0.10 6 0.04 Na+ 122 0.73 121 0.83 153 0.90
2
SO- 104 0.62 142 0.98 173 1.02 C1 167 1.00 145 1.00 170 1.00 SM+ 219 253 320 ZA 271 287 343 % Error 10.6 6.3 3.5

aBrezonik and Shannon (1971).

Brezonik et al. (1983b).

CRatio of ion i to chloride.





68


calcium data are particularly suspect, since ratios of Ca2+/Mg2+ have changed significantly since 1968-69. Although the Ca2+/Mg2+ ratios in

1978-79 and 1981-82 were close to 1.0, the ratio in 1968-69 was only 0.6. No other investigators have reported such a shift in cation ratios associated with acidification, and it is not clear to what extent the observed trend reflects analytical errors. The early

calcium data are particularly suspect since calcium analysis is susceptible to interferences and because an increase in the early

calcium values would produce Ca2+/Mg2+ ratios closer to 1.0.

Chloride concentrations were relatively stable during this period. Although the mean chloride concentration in 1978-79 (145 ueq/L) was slightly lower than the 1968-69 mean (167 ueq/L), the current mean chloride concentration (170 ueq/L) is nearly identical with the 196869 mean. These data indicate that dilution or concentration resulting

from variations in relative rates of precipitation and evaporation cannot account for the observed differences in H+, S042-, Ca2+, or Mg2+. Sodium and potassium levels have changed little since the first study; this is consistent with the observation that these ions were nearly conservative in the sediment titration experiments (Chapter 3).

Concentrations of major cations and anions in the lake during 1981-82 are shown in Figures 4-4 and 4-5. Constituents which are

derived primarily from precipitation, including sodium, chloride, and sulfate, fluctuated considerably during the study as a reflection of variations in lake volume. Sodium concentrations, for example, increased from 125 ueq/L during October, 1980, when the lake volume was 157 x 103 m3 to 229 ueq/L by December, 1981, when the volume had decreased to 120 x 103 m3. As the lake volume increased, sodium












250


200

-C
150- H O)--- Ca24
100 --a M


50



0 N D J F M A M J J A S 0 N D J F M A M J J A
1980 1981 1982 DATE

Figure 4-4. Cations in McCloud Lake, 1980-1982.













250


200


150



- 2
---4 SO2

o 50 A C I


0 I t I I I I I I l I I I I I I I I '
0 N D J F M A M J J A S 0 N D J F M A M J J A
1980 1981 1982
DATE

Figure 4-5. Anions in McCloud Lake, 1980-1982.






71



levels declined. By the end of the study, the lake volume had increased to 148 x 103 m3 while the sodium levels declined to 139 ueq/L.

Concentrations of other base cations were relatively stable. Although calcium concentrations decreased from 112 ueq/L tor-'60 ueq/L during the first three months for some unknown reason, levels were relatively stable throughout the remainder of the study. Magnesium levels were even more stable; concentrations in the two-year study consistently were within +10% of the mean value. Potassium levels were consistently less than 10 ueq/L and the observed variablity at these low levels (4.1 to 8.5 ueq/L) may reflect analytical errors.

Hydrogen ion concentrations show a moderate response to lake volume. Levels increased from 25 ueq/L in October, 1980, to 44 ueq/l in January, 1981, then decreased to 25 ueq/L by August, 1982. Dry Deposition

Measured concentrations of gases (SO02, NO2, and HNO3) and aerosol constituents (NH4+, NO3-, and S042-), shown in Table 4-2, were used to compute fluxes of these constituents using equation 4-1. The accuracy of this approach depends on the selection of appropriate deposition velocities (vd). Deposition velocities depend on the nature of the gas or aerosol, the deposition surface, and meteorological conditions. Published values of vd for a given substance thus may vary over several orders of magnitude. However, when only one type of surface is considered (in this case, water) the range of deposition velocities decreases. For example, while published values of vd for SO2 including all surfaces and conditions span four orders of magnitude (see





72

Table 4-2. Atmospheric concentrations and fluxes of nitrogen
and sulfur species at McCloud Lake site, AugustSeptember, 1982.



2- + a SO4 NH~ HNO3 S02 NO2

Mean concentration, Mean concentration, 2.7 0.4 0.5 3.0 6.0 ug/m (n=20)

Deposition Low 0.2 0.2 0.5 0.5 0.5 velocity (vd), High 0.6 0.6 1.5 1.5 1.5 cm/sec
Flux, Low 40 14 12 150 210 eq/ha-yr High 120 43 36 450 620

aDetermined at Gainesville.






73


Sehmel 1980), deposition velocities for water surfaces under most conditions range from 0.4 to 2.2 cm/sec. Garland (1978) concluded that the mean vd for SO2 is approximately 0.8 cm/sec; several researchers have used values of 0.5 to 1.0 cm/sec (Joranger et al. 1980, Edgerton 1981). For this study, the SO2 flux was estimated

using a range of vd value of 0.5-1.5 cm/sec. Relatively few deposition velocities have been reported for NO2, NH3, and HNO3 (Soderlund

1981). However, since these compounds are highly soluble in water, their deposition to wet surfaces is probably limited by surface resistance (Fowler 1980), and their deposition velocities are probably similar to those for S02. Thus, in this study the deposition velocities of these nitrogenous species was estimated as 0.5-1.5 cm/sec.

The value of vd for sulfate aerosols is less well known. Garland (1978) reported that most of the sulfate aerosol has a diameter of 0.1 to 1.0 um and has a mean deposition velocity of 0.1 cm/sec. However, based on a comparison between measured sulfate deposition in dry

buckets (Aerochem-Metrics collector) and ambient air concentrations, Edgerton (unpublished data) has concluded that the mean deposition

velocity for sulfate aerosol in Florida is 0.4 cm/sec. Accordingly, values of 0.2-0.6 cm/sec were used to estimate the deposition of sulfate aerosol in this study.

The dry deposition of other aerosol constituents (Ca2+, Mg2+, K2+ Na+, C1-, and NO3-) and of NH4 and S042- was estimated using two northern Florida data bases. Edgerton (unpublished data) has recently

collected dry deposition data from four sites in northern Florida using an Aerochem-Metrics wet/dry collector (Table 4-3). These data show considerable variability in deposition rates among sites,






74

Table 4-3. Dry deposition at five northern Florida sites, 1982.


Annual dry deposition, eq/ha-yr

Site Ca2+ Mg2+ K Na2+ C1 Gainesville 115 70 12 76 60 St. Augustinea 74 35 12 90 50 Cross City 649 30 12 73 66 Archibald 95 36 12 89 105 Panhandle 33 17 12 37 40

aCorrected for sea salt influence.

Source: Edgerton, unpublished data.






75



reflecting local anthropogenic influences. However, the Panhandle site closely resembled the McCloud Lake site in its remoteness from urban areas and its sandy soils and pine-forested surroundings. Deposition rates at this collector were the lowest observed and represent reasonable minimum deposition rates for McCloud Lake. Deposition rates of Ca2+, Mg2+, and K+ at this site were used as lower limits of deposition of these constituents McCloud Lake.

The study of Brezonik et al.(1983b) included wet/dry deposition

measurements at three inland Florida sites (Gainesville, Belle Glade, and Apopka) as well as bulk precipitation data for 25 sites, including four in the vicinity of McCloud Lake (Gainesville, Jasper, Waldo, and Hastings). Two methods were used to estimate dry deposition at the McCloud site from these data. First, the ratios of wet to

total deposition at the three inland wet/ dry collectors (Table 4-4) were used to estimate dry deposition from measured wet deposition at McCloud Lake. This method should produce accurate estimates of dry deposition for substances that have uniform wet/total deposition ratios, such as chloride (wet/total ratios = 0.55-0.59), sodium (wet/ total ratios = 0.55-0.60), potassium (wet/total ratios = 0.55-0.60), nitrate (wet/total ratios = 0.68-0.74), and magnesium (wet/total ratios = 0.40-0.45). For substances such as calcium (wet/total ratios = 0.29- 0.54), sulfate (wet/total ratios = 0.71-0.88), and ammonium (wet/total ratios = 0.73-0.88) this method is less reliable. The second method of estimating dry deposition was to subtract the measured wet depositon at McCloud Lake from the bulk deposition measured








Table 4-4. Relative loadings in wet versus total (wet + dry) precipitation. Values correspond to
% deposition from wet-only precipitation.


+ + + +2 +2 + -ess Location H Na+ K Ca+2 Mg +2 NH -N NO-N S


Gainesville 100 59 59 31 40 88 68 74 61
(semi-urban)

Apopka 100 57 55 52 45 73 74 77 61
(agricultural)

Belle Glade 100 59 60 40 45 85 72 75 60
(agricultural)





77



at the four regional collectors (Jasper, Gainesville, Waldo, and Hastings) during the 1978-79 study (Table 4-5).

Dry deposition rates estimated from these two data bases are shown in Table 4-6. For each method, some of the measured deposition rates were excluded because they reflect extreme conditions that are not representative of conditions at McCloud Lake. Calcium deposition, in particular, appears to be strongly influenced by local anthropogenic influences, and data from three sites were excluded. Dry deposition of calcium at Cross City (Table 4-3) reflected the proximity of a school playground, whereas bulk deposition of calcium at Gainesville was influenced a parking lot and city streets and bulk deposition of calcium at Jasper reflected phosphate mining activity. Dry deposition of sodium and chloride deposition at St. Augustine and bulk deposition of these ions at Hastings and Jasper reflected a strong sea-salt influence that was not representative of McCloud Lake, and these data were not used. Finally, ammonium deposition at Jasper seemed excessively high (283 eq/ha-yr). Although the reason for the high ammonium deposition is not known, this value was rejected as an outlier.

Dry deposition of alkalinity in the study of Brezonik et al. (1983b) was inferred from the difference between the sum of measured cations and the sum of measured anions (Table 4-5). From Table 4-5 it appears that the inferred alkalinity was approximately 0.6 times the calcium deposition. To estimate alkalinity deposition in this study, this ratio was used in conjunction with estimated calcium deposition.

Dry deposition of NO2, SO2 HNO3, and alkalinity were used to compute an "effective dry deposition" for protons. This computation








Table 4-5. Bulk deposition at four northern Florida sites, 1978-1979.


Bulk deposition, eq/ha-yr

2+ 2+ K+ + +NO NH Site Ca2+ Mg K Na+ CI- NO NH4 Gainesville 356 91 33 191 264 508 247 137 Jasper 529 104 80 219 241 670 227 283 Hastings 185 119 88 395 510 450 158 183 Waldo 197 92 61 367 412 617 208 185


Source: Brezonik et al. (1983b).






79


assumes that the gases dissolve in water and undergo the fol lowing reactions:

SO2 + H20 + 1/2 02 ---- > So42- + 2H+ (4-6) 2NO2 + 1/2 02 + H20 ----> 2NO3" + 2H+ (4-7) HNO3 ----> H+ + NO3- (4-8) The effective dry deposition of H+ (DH+) was thus

DH+ = DS02 + DN02 + DHN03 Dalk (4-9)
where DS02.... Dalk = deposition rates of subscripted substances, eq/ha-yr.

The last two columns in Table 4-6 show dry deposition rates used in the McCloud Lake model. These values bracket the measured or estimated dry deposition rates for north Florida. Sodium and chloride were further constrained to occur in a ratio of 0.88:1 (the ratio of these ions in seawater--see Stumm and Morgan 1981), although this constraint required very little modification of observed deposition rates (Table 4-6). As shown in Table 4-6, the estimated deposition of Na+ and Cl- varied by a factor of three, while estimates of other aerosol constituents varied by factors of three to ten. Estimates of gaseous deposition varied by a factor of three, reflecting the use of vd values from 0.5-1.5 cm/sec. Deposition rates for NH4+ and S042estimated from aerosol concentrations (Table 4-2) fell within the range of estimates made from bulk and dry collector data. Dry deposition constituents were assigned to three groups: 1) sea-salt components (Na+ and Cl-), 2) anthropogenic aerosols (S042-, NO3-, Ca2+, Mg2+, K and alkalinity), and 3) gases (SO02, HN03, and NO2). To compile the McCloud Lake mass balances, constituents within each group were assigned the upper or lower estimates of dry deposition







Table 4-6. Aerosol deposition at McCloud Lake (all values as eq/ha-yr).


Estimated from Estimated from Dry deposition wet/total bulk values used in deposition ratio deposition McCloud model

Measured bulk
Measured Dry deposition McCloud deposition at McCloud
wet-only for north wet/total dry depo- north Florida dry
Constituent deposition Florida sitesa ratios sition sites deposition Low High

Ca2+ 50 33-115 0.31-0.52 46-111 185-197b 135-147 35 200 Mg2+ 42 17- 70 0.40-0.45 51- 63 90-120 48- 72 20 70 co K+ 11 n12 0.55-0.60 7- 9 33- 88 22- 77 10 75 Na+ 154 37- 89 0.57-0.59 107-116 191-219c 37- 65 35 105 Cl- 180 40-105 0.60-0.61 115-120 241-264c 61-84 40 120 SO2- 418 80 0.74-0.88 57-147 450-670 32-252 30 250 NO 144 49- 87 0.68-0.44 51- 68 158-247 14-103 10 90 NH4 73 0.73-0.88 10- 27 127-185d 64-112 10 100

aEdgerton, unpublished data.
bGainesville and Jasper included.

CHastings, Jasper, and St. Augustine excluded.
djasper excluded.






81


simultaneously. Thus, eight scenarios were constructed representing combinations of high and low sea salts, high and low anthropogenic aerosols, and high and low gases. All combinations of aerosol deposition (i.e., high sea-salt + low anthropogenic aerosols, low sea-salt + low anthropogenic aerosols, high sea-salt + high anthropogenic aerosols, and high sea-salt + low anthropogenic aerosols) had ion balances (equation 2-1) within 5%.

Since deposition of gases contributed to the effective H+ deposition and the deposition of alkalinity reduced the effective H+ deposition (equation 4-9), the higher estimate of H+ deposition was computed using the high estimate of gas deposition and the low estimate of aerosol deposition. Conversely, the lower estimate of H+ deposition was obtained using the lower estimate of gas deposition and the higher estimate of alkalinity deposition (Table 4-6). Wet Precipitation

Wet-only precipitation at McCloud Lake, like precipitation throughout northern Florida (Brezonik et al. 1983b), was acidic (volume-weighted mean pH: 4.5). The source of H+ in wet precipitation can be inferred from the anion composition. The occurrence of chloride and sodium in a ratio of 0.85:1 in wet precipitation clearly suggests a sea-salt origin for these ions rather than an HC1 source for chloride. The N03- concentration in wet precipitation, 9.8 ueq/L, could account for no more than 34% of the 29 ueq H+/L. Sulfuric acid was therefore the primary source of protons in wet precipitation: the S042- concentration of 29 ueq/L was adequate to account for 100% of the protons in wet precipitation.







Table 4-7. Precipitation chemistry at McCloud Lake, September 1981-August 1982.


Estimated dry
deposition,
Wet/only Wet eq/ha-yr
concentration, deposition, Total Constituent peq/L eq/ha-yr Low High eq/ha-yr % wet Ci/Cl

H+ 28.5 420 250 1090 670-1510 28-63 2.2-6.9 Ca2+ 3.4 50 35 200 85-250 20-59 0.3-1.1 Mg2+ 2.9 40 20 70 60-110 36-67 0.2-0.5 K+ 0.75 10 10 75 20- 85 12-50 0.1-0.4 Na+ 10.5 150 35 105 185-255 59-81 0.86 C1 12.3 180 40 120 220-300 60-82 1.0 Sulfur (total)a 420 180 700 600-1120 37-70 2.0-5.1 SO2 (gas) - 150 450 - SO- 28.5 420 30 250 450-670 37-93 1.5-3.0 NO3 (total)a 140 230 750 370-890 16-38 1.2-4.0 NO2 (gas) 210 620 - HNO3 (gas) - 10 40 - NO3 9.8 140 10 90 160-230 61-93 0.6-1.1 NH4 5.0 75 10 100 85-175 75-88 0.3-0.8 Alkalinity 0 0 20 120 20-120 0 0.1-0.5

a Includes wet + aerosol + gas species.





83



Wet precipitation accounted for 60-82% of the chloride, 59-81% of the sodium, 62-93% of the nitrate, 75-88% of the ammonium, and 37-93% of the sulfate deposition to McCloud Lake. Gaseous deposition was an additional major source of sulfate and nitrate. Sulfur dioxide (SO02) deposition, estimated to be 150-450 eq/ha-yr (expression on equivalent basis assumes conversion to SO42- at lake surface), was 25-40% of the total sulfur deposition. The adsorption of NO2 at the lake surface contributed 210-620 eq NO3-/ha-yr, or 57-70% of the total NOx (NO3 + NO2 + HN03). Interestingly, although S042- was the dominant acidic anion in wet-only precipitation, the dry deposition loading of NO, was equal or greater than the dry deposition of S042- + SO2 (Table 4-7).

Dry deposition of H+ (equation 4-10) was a major source of protons to McCloud Lake. Although the range of estimated values was large (250-1090 eq/ha-yr), these estimates show that deposition of gaseous species (primarily SO2 and NO2) may be an important source of protons to McCloud Lake, and may account for 54-71% of the total proton deposition.

Groundwater Chemistry

Table 4-8 shows water quality data for the 12 wells sampled during May-October, 1982 (arithmetic means) and for the 1982 inseepage (flow-weighted means). These data show that precipitation passing through the soils surrounding McCloud Lake was altered by several mechanisms, even though these soils are extremely sandy and permeable. First, evaporation concentrated the precipitation, resulting in increased concentrations of conservative ions such as chloride and sodium. Based on observed chloride values, well water






84


Table 4-8. Chemistry of groundwater at McCloud Lake.

Seepage Meters Well Water Conc. Conc.
Constituent peq/L Ci/Cl E.F. 1eq/L CiCl E.F.

H+ 1.9 0.05 0.007-0.03 1.9 0.02 0.003-0.01 Ca2+ 92 2.36 1.5 -7.9 204 2.27 1.4 -7.6 Mg2+ 51 1.30 2.6 -6.5 90 1.00 2.0 -5.0 Na+ 34 0.87 1.0 91 1.01 1.0 K+ 20 0.50 1.3 -5.0 37 0.41 1.0 -4.0
2
SO2- 36 0.92 0.2 -0.3 49 0.54 0.1 -0.2 C1 39 1.00 90 1.00 NH4 3 0.09 0.1 -0.2 - N03 4 0.10 .02 .07 4.9 .05 0.01 -0.03 HCO3 a 123 3.14 280 3.11

alInferred from ion balance.






85


was concentrated by a factor of -14 while in-seepage was concentrated by a factor of v3 (Table 4-8). The fact that in-seepage was less concentrated than well water suggests that seepage flows to the lake originated as precipitation that fell very near the lake shore and passed rapidly through the soil and into the lake. Groundwater further

from the lake apparently remains in the soil longer and becomes more concentrated by further evaporation.

Both seepage and wel 1 water had pH values between 5.5 and 6.0. Neutralization of precipitation acidity probably occurs as the result cation exchange that produces both alkalinity and dissolved Ca2+ and

Mg2+. This enrichment can be seen by comparing ion/Cl- ratios in the seepage and well water with the ion/C1- ratios in the total (wet + dry) precipitation using an enrichment factor (EF):

EF = (Ci/Cl-) (4-10)
(Ci/C-)p
where (Ci/Cl-) = ratio of ion i to Cl- in compartment j;
( /Cl) precipitation ion/Cl- ratio.
An enrichment factor of < 1 indicates a depletion of the ion

resulting from chemical precipitation or assimilation, while an enrichment factor > 1 indicates enrichment from mineral dissolution or mineralization. An EF of 1.0 indicates that the ion is conservative.

As seen in Table 4-8, calcium was enriched by a factor of 1.58.0 in both the well water and in-seepage, while magnesium was enriched by a factor of 2.6-6.5 in the seepage and 2.0-5.0 in the wells. Alkalinity, calculated by inference from the ion balance, roughly balances Ca2+ Mg2+ in both seepage and well water. Potassium may also have been enriched in the seepage and well waters, although when the lower estimates of K+ dry deposition were used to compute the EF,





86



potassium appeared to be conservative in both compartments

(Table 4-8).

Sulfate was depleted in both seepage water (EF = 0.2 0.3) and in well water (EF = 0.1-0.2). Sulfate adsorption or biological assimilation was probably responsible for the sulfate depletion in the well

water although sulfate reduction could deplete sulfate if the well water became anaerobic. Sulfate reduction undoubtedly caused the sulfate depletion in the seepage water, since this process occurred in the laboratory seepage experiment (Chapter 3).

Both nitrate and ammonium were depleted in the subsurface

waters. Much of this depletion undoubtedly occurs as the result of assimilation of these nutrients by terrestial vegetation and soil microorganisms. These results are consistent with reports that inorganic nitrogen is retained in forest soils (Impact Assessment Group 1983). Denitrification in littoral in-seepage also may account

for some of the nitrate depletion; laboratory seepage column experiments (Chapter 3) also showed a dramatic reduction of nitrate in the eluate.

Fate of Major Ions in McCloud Lake

Data on wet and dry precipitation, seepage flows, and changes in storage were used to evaluate the fate of ions entering McCloud Lake. Several approaches were used to evaluate in-lake sinks and sources of ions. Annual mass balances were compiled for major ions and nitrogen species using the equation

dM/dt = P + Gin Gout + S (4-11) where dM/dt = change in storage, eq,
P = precipitation (wet + dry) inputs, eq/yr,






87


Gin,Gout = groundwater inputs and outputs, respectively,
m /yr, and
S = sink (negative value) or source (positive value), eq/yr.

This approach has the advantage of not requiring steady state conditions. However, evaluation of the sink/ source term is sensitive to errors that result determining small differences between large numbers. This problem is discussed in greater detail with respect to the chloride mass balance below.

Sinks and sources of major ions also were evaluated by comparing

the ion/chloride ratios in the inflow (wet + dry precipitation + inseepage) to ion/chlorides ratios in the lake. Since chloride is conservative in aquatic systems, the ion/chloride ratio of a substance entering the lake can change only occur as the result of sinks or sources of that substance within the lake. The magnitude of the sink for substance i (Si) can be evaluated as:

Si = Li/Lr1 [Cill/[Cl]I x 100 (4-12) i/LcI

where Li,l 1 = annual input of substance i and chloride, eq/yr,
[Ci]L, [C1]L = concentration of substance i and of chloride in lake, meq/L, and
Si = sink/source term as per cent of input.
Although this approach assumes that the lake is at steady state, it does not depend on measured changes in storage and is less sensitive to errors in input load. Input loads and mean storage were also used to compute the residence time (T) for various substances entering McCloud Lake, where:

T = mean storage, e (4-13)
annual input, eq/yr

The residence time of conservative substances such as chloride and sodium is the same as the water residence time based on outflow.





88


Shorter residence times indicate that a substance has an internal sink, while longer residence times indicate an internal source. Chloride

Most of the chloride entering McCloud Lake came from wet + dry precipitation (Table 4-9); groundwater seepage accounted for < 25% of the total chloride input. Although the mass balance equation for the 1981-82 model year indicated a source of 4960-5370 eq/yr, chloride was undoubtedly conservative and the calculated source reflects a problem with the mass balance calculation. Since the mass balance equation (equation 4-11) computes the sink/source term as the difference between inputs, outputs, and the change in storage, it is subject to errors that accrue from estimating small differences between large numbers. The change in storage term is the most likely source of error. During the model year, storage of chloride increased by 4470 eq (20% of the mean storage), compared with an annual precipitation input of only 1120-1520 eq. However the change in storage term is subject to errors in both chloride measurement and estimation of water volume. The error term can be calculated from the equation:

EdM/dt = EC1*V1 + C1*Ev1 + Ec2*C2 + V2*EC2 (4-14) + EC1*EV1 +EC2*EV2

where EdM/dt = error in change in storage for a given constituent,
V1,V2 = initial and final volume,
C1,C2 = initial and final concentration,
EV1,EV2 = error in initial and final estimates of volume, and
EC1,Ec2 = error in initial and final estimates of
concentration.

If each measured term had an error of only 5% of the true value, and if initial and final storage were approximately equal, the potential error in dM/dt could be as high as 25% of the mean storage. Although








Table 4-9. Fate of ions entering McCloud Lake.


Seepage

Mean % loss Residence Precipitation, In Out storage AS, Sink,sourcea from ion time (T) Constituent eq/yr eq/yr eq/yr eq eq/yr eq/yr ratios b yr C1 1120-1520 340 2360 22,440 +4470 5370-+ 4860 12.0-15.4 Na+ 1000-1360 300 2620 23,340 -3540 -1360+-1220 11 13.5-17.2 SO2- 2990-5580 320 2790 24,360 + 220 230+-*2820 39-73 4.1- 7.4 Ca +
460-1290 810 750 7,050 320 840+-1560 53-77 3.4- 5.6 Mg2+ 570- 320 450 890 8,460 -1950 -1820+-2070 7-56 8.4-11.1 K+ 100- 430 170 80 740 + 230 + 30+- 300 78-92 1.2- 2.7 H+ 3300-7560 -480 20 4,100 90 -2440+-1815 89-96 0.7- 1.4 NH+ 420- 880 30 80 600 990 -1350+-1815 89-96 0.7- 1.3 NO- 1910-4570 30 80 670 320 -2140+-4810 97-99 0.1- 0.4

a,,+ = source; = sink. Estimated from annual mass balance (equation 4-11).
See equation 4-12






90


this example represents a worse-case scenario (all errors in the same direction), it illustrates the type of problem encountered in using a mass balance model for a substance that has a large storage relative to inputs during the modeling period. For chloride, with a residence time of 12.0-15.4 years in McCloud Lake, a much longer modeling period would be needed to accurately estimate the sink/source term. Sodium

Precipitation was also the major source of sodium to McCloud Lake, accounting for 77% of the total input. The mass balance equation indicates a sink of 1360 to 2220 eq/yr, but this is < 10% of the mean storage and is probably insignificant for reasons discussed above. Calculations based on Na+/Cl- ratios indicated a small sink (11% of annual inputs), suggesting that sodium is nearly conservative. The residence time for sodium in McCloud Lake, 13.5 to 17.2 years, is very near that of chloride.

Sulfate

Wet + dry precipitation accounted for over 90% of the total inorganic sulfur (S042- + SO02) to McCloud Lake (Table 4-9). Both ion ratios and the annual mass balance indicate that the lake was a sink for sulfate. Although the mass balance yields a sink of 230 to 2820 eq/yr, the magnitude of this sink is small relative to the lake storage of 24,360 eq and therefore subject to the type of errors discussed above. The presence of a sink is confirmed, however, by ion ratio calculations which indicate that 39-73% of the inorganic sulfur entering the lake is lost internally and by the calculated residence time of only 4.1 to 7.4 years.






91



The most likely mechanism for an internal sulfate sink is biological sulfate reduction at the sediment-water interface. This hypothesis is strongly supported by the sulfate pore water profiles (Figure 3-14), which show a distinct gradient of sulfate in the profundal sediments, and by the seepage column experiments, which showed a loss of sulfate in the groundwater eluate (see Chapter 3). These data alone do not conclusively prove that biological sulfate reduction is responsible for the observed sulfate sink since end products (FeS or H2S) have not been identified. On the other hand, sulfate adsorption did not occur to an appreciable extent in McCloud Lake sediments (Chapter 3), and it is unlikely that other mechanisms such as biological assimilation or chemical precipitation of sulfate-bearing minerals (Nordstrom 1982) could account for the observed sink. Furthermore, the organic-rich sediment (over 90% volatile solids in the profundal sediments) and warm temperatures (annual mean = 200 C) are conducive to sulfate reduction in McCloud Lake. Although there is some evidence that sulfate reduction may be inhibited below pH 5 (Zinder and Brock 1978), the pH of McCloud Lake sediments is generally above 5, perhaps as a result of the pH-buffering effect of the sulfate reducers. Sulfate reduction has been observed in other acidic environments including peat bogs (Rippon et al. 1980) and experimentally acidified Lake 223 (Kelley et al 1982). The evidence for sulfate reduction in McCloud Lake suggests that sulfate reduction may be an important sink for sulfate in acidic seepaye lakes.






92


Calcium, magnesium, and potassium

Mass balance calculations and comparisons of ion/C1- ratios of

inputs and lakewater indicate that McCloud Lake is a sink for all three cations. Ion ratio calculations, for example, show that 53-77%

of the calcium, 7-45% of the magnesium, and 78-92% of the potassium entering the lake is lost through an internal sink. Potential sinks,

such as biogenic accumulation or precipitation of inorganic solids, are unlikely to be important in McCloud Lake. Potassium generally is assumed to be conservative. Potassium concentrations are rarely affected by phytoplankton blooms (Wetzel 1975), K+ was essentially nonreactive in the neutralization experiments described in Chapter 3, and there are few potassium-containing minerals that would precipitate in a softwater lake. Chemical precipitation of calcium and magnesium is also unlikely in this lake. It therefore appears that inaccuracies in the estimated inputs of these cations are responsible for the apparent sinks. Of the input components, wet deposition is undoubtedly the most accurate. Dry deposition for these elements was estimated only to within a range of approximately one order of magnitude, although the values used undoubtedly bracket the actual dry deposition rates.

The term most likely subject to errors is in-seepage. As shown in Table 4-9, in-seepage was a major source of calcium (39-64% of the total loading), magnesium (44-58% of the total loading), and potassium (29-62% of the total). Furthermore, a large (50%) error in calculated in-seepage would be nearly undetectable in the water balance, since in-seepage only accounted for 10% of the total inflow to the lake. For example, a 50% reduction in in-seepage would change the computed endof-year water storage by only 3% but would decrease the total calcium






93



loading by as much as 32% and increase the residence time to 4.4-8.2 years. It is apparent that more intensive sampling of seepage would be required to obtain accurate estimates of fluxes of these cations in McClould Lake. It should be noted that seepage was a minor source of sulfate, protons, chloride, sodium, nitrate, and ammonium, and relatively large errors in seepage loadings would have minimal effect of budgets of these substances.

Nitrogen species

Wet + dry precipitation accounted for nearly all (over 95%) of the inorganic nitrogen input to McCloud Lake. The total N03- load

(including NO2 and HNO3 adsorption at the lake surface) was 1940 to 4610 eq/yr, about five times that of NH4+ (450-910 eq/yr). Both species of nitrogen were removed from the water column by in-lake mechanisms. The mass balance indicated an annual sink of 1350-1820 eq for NH4 (several times the mean storage of 603 eq) and NH4+/CIratios show that 86-96% of the NH4+ entering McCloud Lake was removed. Nitrate had an annual sink of 2140-4810 eq during the model year, compared with a mean storage of 670 eq. Ion ratio calculations indicated that 97-99% of the input nitrate disappeared via an internal sink. The residence times for these substances is appoximately one year. Biological assimilation by phytoplankton, bacteria, and macrophytes and the accumulation of organic nitrogen in the sediment are probably the most important mechanisms for removing inorganic nitrogen from the water column of McCloud Lake. Additional loss of NH4+ may occur as the result of adsorption to sediment surfaces and additional




Full Text

PAGE 1

MINERAL AND NUTRIENT CYCLES AND THEIR EFFECT ON THE PROTON BALANCE OF A SOFTWATER, ACIDIC LAKE By LAWRENCE ALAN BAKER A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 1984

PAGE 2

ACKNOWLEDGEMENTS I would like to express my appreciation to those who helped make this dissertation possible. Dr. P. L. Brezonik, my original major professor, provided expert assistance and guidance throughout this research project. Several parts of ths research were done in collaboration with other students, particularly Walter Ogburn and Eric Edgerton, and their efforts are appreciated. I would also like to thank James Heaney, for serving as chairman of my supervisory committee following the move by Dr. Brezonik and myself to the University of Minnesota, and the other committee members -G. R. Best, G. Bitton, and D. Graetz ~ for their continued interest and advice during this project. Finally, I would like to thank my friends Jack and Debbie Tuschall, Carl Miles, and Sue Zoltewicz for technical advice and comradeship at the University of Florida and my fiance, Nancy Rodenborg, for her patient support during the writing process.

PAGE 3

TABLE OF CONTENTS ACKNOWLEDGEMENTS ii ABSTRACT vi CHAPTER I — INTRODUCTION 1 Background 1 Objectives 2 Site Description 3 CHAPTER 2 -METHODS 7 Sample Collection 7 Analytical Methods 8 Cations 8 Nutrients 8 Major Anions 9 pH 9 Isotope Ratios 9 Quality Assurance 11 CHAPTER 3 — SEDIMENT BUFFERING IN McCLOUD LAKE 12 Introduction 12 Mechanisms for Sediment Buffering 15 Methods 20 Sediment-Water Batch Experiments 20 Seepage Column Experiment 22 Microcosm Experiment 22 Littoral Mesocosms 24 Pore Water Profiles 25 Results and Discussion 25 Sediment-Water Batch Experiments 25 Magnitude of neutralization 25 Cation exchange 27 Aluminum solubility 31 Sulfate adsorption 35 Seepage Column Experiment 37 Sediment-Water Microcosms 40 Littoral Mesocosms 41 Pore Water Profiles 49 Conclusions 51 CHAPTER 4 — McCLOUD LAKE MASS BALANCE 53 Introduction 53 Background 54 Methods 56 iii

PAGE 4

Wet Precipitation 56 Dry Deposition 56 Seepage 57 Groundwater Wells 59 Lake Storage 59 Evaporation 59 Results and Discussion 50 Water Budget 50 Chemistry of McCloud Lake, Past and Present 64 Dry Deposition 71 Wet Precipitation 81 Groundwater Chemistry 83 Fate of Major Ions in McCloud Lake 86 Chloride 88 Sodium 90 Sulfate 90 Calcium, magnesium, and potassium 92 Nitrogen species 93 Proton balance 94 CHAPTER 5 -DECOMPOSITION AND NUTRIENT CYCLING IN McCLOUD LAKE 96 Background 96 Effects of Low pH on Decomposition 97 Effects of Low pH on Nitrogen Cycling 99 Ammonifi cation 99 Nitrification 100 Denitrification 100 Nitrogen fixation 101 Methods 101 Littoral Mesocosms 101 Microcosm Experiments 104 Nitrogen Mass Balance 106 Results 106 Enclosure Experiments 106 Ammonium Adsorption 117 Microcosm Experiments 119 Water-only microcosms 119 Sediment-water microcosms 121 McCloud Lake Nitrogen Budget 129 Precipitation 129 McCloud Lake nitrogen 132 Mass balance 132 Conclusions 140 CHAPTER 6 — CONCLUSIONS 142 Sediment Neutralization 142 McCloud Lake Mass Balance 142 Nitrogen Cycling and Decomposition 143 iv

PAGE 5

BIBLIOGRAPHY 144 APPENDIX -McCLOUD LAKE HYDROLOGIC AND CHEMICAL DATA 153 BIOGRAPHICAL SKETCH 159 V

PAGE 6

Abstract of Dissertation Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy MINERAL AND NUTRIENT CYCLES AND THEIR EFFECT ON THE PROTON BALANCE OF A SOFTWATER, ACIDIC LAKE By Lawrence Alan Baker April 1984 Chairman: James P. Heaney Cochairman: Patrick L. Brezonik Major Department: Environmental Engineering Sciences Mineral and nutrient cycles in a softwater, acidic lake were studied using 1 aboratory experiments, littoral mesocosms, and wholelake mass balances. The primary study site was McCloud Lake, a small (5 ha), acidic (pH 4.5), seepage lake in the Trail Ridge Region of northcentral Florida. Titrations of sediment slurries from McCloud Lake and other softwater lakes showed that profundal seediments have an acid neutralizing capacity (ANC) up to 10 meq/lOO g dry weight. Cation exchange of calcium and magnesium accounted for over 50% of the ANC in the sediments examined, while solubilization of aluminum accounted for up to 20% of the ANC. Exchanges of sodium and potassium were unimportant, as was sulfate adsorption. An experiment to simulate subsurface seepage to McCloud Lake showed that littoral sediments could neutralize synthetic groundwater (pH 3.4 to 6.8) by sulfate reduction and cation exchange. vi

PAGE 7

Chemical analysis of in situ seepage and sediment pore waters confirmed these results. Water budget calculations showed that McCloud Lake received 90% of its water from precipitation and 10% from subsurface flow. Sulfate entering the lake was lost by an internal sink (sulfate reduction) that accounted for 37 to 73% of the total input. Assimilation and denitrification consumed over 95% of the nitrate entering the lake. Internal sinks for nitrate and sulfate nearly balanced the proton input to the lake. Although the mass balance was not sensitive enough to show the internal source of calcium and magnesium expected from lab experiments, an analysis of historical trends showed that calcium levels have nearly doubled while magnesium levels have increased by 34% during the past 14 years in which the proton concentration has doubled. Littoral mesocosm and laboratory microcosm experiments showed that nitrification can occur at the sediment-water interface when the pH of the overlying water is as low as 3.5. Sediment respiration did not appear to be affected by pH in the littoral mesocosms. An analysis of historical trends in McCloud Lake nutrient levels revealed no significant changes in major nutrient species during the past 14 years. These results cast doubt on the hypothesis that nutrient regeneration and decomposition are inhibited in acidified lakes.

PAGE 8

CHAPTER 1 INTRODUCTION Background Numerous studies have shown that precipitation acidity has increased in large areas of Scandinavia and North America (Dovland et al. 1976; Cogbill and Likens 1974; Brezonik et al. 1983b). It is widely believed that increased precipitation acidity has resulted in the acidification of thousands of poorly buffered lakes and streams (Grahn et al. 1974; Gjessing et al. 1976; Norton et al. 1980; Schofield 1976; Impact Assessment Group 1983). Acidification has major and often serious effects on lake and stream ecosystems. In addition to the direct effect of H'*' on osmoregulation (Leivestad et al. 1976), decreased pH affects biota through several indirect mechanisms. Decreased pH enhances the solubility of toxic metals, and increased concentrations of aluminum, cadmium, zinc, and lead have been observed in lakes and streams that have become acidified (Aimer et al. 1978; Norton et al. 1980). Humic acids precipitate at low pH, accounting in part for the increase in clarity observed in acidified lakes, and the precipitation of aluminum phosphate complexes may decrease lake phosphorus levels (Aimer et al. 1978). Acidification produces distinct changes in biological communities. The most not able effect has been the dec line of fish populations in many rivers and lakes in the Scandinavian countries and in 1

PAGE 9

2 the Adirondack Mountains of New York (Leivestad et al. 1976; Aimer et al. 1978; Schofield 1976). Major changes occur in the structure of phytopi ankton and zooplankton communities, and declining pH clearly results in decreased species diversity of plankton communities (Hutchinson et al. 1978; Leivestad et al. 1976; Aimer et al. 1978), although it is not clear whether primary productivity decreases (see Hendrey et al. 1976). Grahn et al. (1974) postulated that acidification retards decomposition and nutrient regeneration and initiates a "sel f-ol i gotrophi cati on" process, although subsequent studies have produced conflicting results (Schindler et al. 1980; Dillon et al. 1979; Hultberg and Andersson 1982). Objectives One of the major problems in acid precipitation research is to elucidate the relationship between precipitation acidity and lake pH. To develop models for this purpose requires an understanding of the processes that produce and consume H"*" ions in lakes and their watersheds. Current lake acidification models are based on watershed weathering reactions (Hendriksen 1980; Wright 1982) and ignore in-lake neutralization processes. Thus, the objectives of the first phase of this study (Chapter 3) were to identify H''"-neutral i zi ng mechanisms that occur in the sediments of softwater lakes and to evaluate the potential of these reactions in neutralizing the overlying water. Chapter 4 addresses the question of in-lake neutralization using a mass balance of major ions for McCloud Lake, Florida, to quantify source and sink terms. The second phase of this research (Chapter 5) is a study of decomposition and nitrogen cycling in McCloud Lake. The hypothesis

PAGE 10

3 that decomposition and nutrient regeneration are inhibited in acidified lakes has not been thoroughly evaluated, and several recent studies indicate that nutrient levels do not change substantially upon acidification or neutralization of whole lakes (Dillon et al. 1979; Schindler et al. 1980; Hultberg and Andersson 1982). Chapter 5 includes in situ mesocosm and laboratory microcosm studies to evaluate nutrient regeneration, a nitrogen mass balance for McCloud Lake, and an evaluation of historical trends in nutrient levels for this lake. Site Description McCloud Lake is a small (5 ha) softwater lake in the Trail Ridge Region of north-central Florida, approximately 40 kilometers east of Gainesville. The region is characterized by karst topography and includes numerous sand hills interspersed by small lake basins that resulted from solution of the underlying limestone. The 95 ha watershed is uninhabited and is located on the Katherine Ordway Ecological Preserve maintained by the University of Florida. Surficial soils in the watershed are nonspodic marine deposits of fine sand, gravel, and sandy clays of the Citronelle Formation. These surficial deposits are underlain by 24-30 m of phosphatic sands, sandy clays, and clays of the Hawthorne Formation. These clays act as a confining layer (aquiclude) that separates the deep Floridan aquifer from the unconfined, perched water table of the Citronel le Formation (Brezonik et al. 1983a). The lake is a simple doline (Figure 1-1) with no defined inlets or outlets. Most of the water to the lake comes from direct precipitation; the remainder enters as subsurface seepage. Water levels

PAGE 11

o Seepage meters meters Groundwater wells Figure 1-1. McCloud Lake bathymetric map.

PAGE 12

Table 1-1. Summary of morphometric information for McCloud Lake. Elevation 36.5-38.0 above NVGD^ Surface area (at 35.6 m) 4.81 ha 4 3 Volume (at 35.6 m) 11.9 X 10 m Mean depth 2.5 m Maximum depth 4.6 m Watershed area 95 ha ^National Vertical Geodetic Datum (approximately mean sea level)

PAGE 13

6 fluctuate in response to changes in relative rates of precipitation and evaporation. During 1980-81, the water level was very low due to a prolonged drought and the maximum depth was 4-5 m. During 1966-67, the maximum depth was 6.4 m and the lake had a surface area of 9 ha (Brezonik et al. 1969), nearly double the 1980-81 area of 5 ha. McCloud Lake has very low conductivity (< 50 uS/cm at 25 C) and no alkalinity. The lake has become increasingly acidic during the past 14 years (see Chapter 4), and now has a mean pH of 4.6. The algal standing crop is small (mean chlorophyll a^ was 6 ug/L in 1981), and the phytopl ankton community is dominated by six genera: Oocysti s Pi nofa ry on Mougeotia Chi amydomonas Gl enodinium and Eugl ena (Crisman et al. 1983). A narrow band of emergent macrophytes surrounds the lake, and the sparse submergent macrophyte community is composed primarily of two species: El eocharis sp. and Websteria confervoides (Crisman et al. 1983).

PAGE 14

CHAPTER 2 METHODS A major problem in evaluating historical data relative to the effects of acid precipitation is the lack of detail provided by earlier investigators on the analytical methods they used. The purpose of this chapter is to provide details of sample collection, preservation, and laboratory analyses used in this study. Experimental procedures are described in the appropriate chapters. Sample Collection Throughout this study, samples were collected in linear polyethylene (LPE) bottles that were washed with soap and water, then thoroughly rinsed with distilled water. Samples collected for analyses of pH, sulfate, chloride, and conductivity were stored in bottles that were new at the beginning of the study and were never acidwashed. At the University of Florida, these were rinsed with distil led, deionized water (DDW) following the soap-and-water wash. During later experiments (at the University of Minnesota), bottles for these analyses were purchased new and simply rinsed between use and filled with DDW during storage. Samples for nutrient analyses were stored in LPE bottles that had been soaked overnight in 1:1 sulfuric acid (at the University of Florida) or 10 % HCl (at the University of Minnesota), then rinsed with DDW. Nutrient samples were preserved by adding 40 mg/L HgCl2 to the sample bottles (USEPA 1974) or by freezing. Samples for metals analyses were stored in LPE bottles that had been soaked in 10 % HNO3 and rinsed with DDW; these were acidified 7

PAGE 15

8 with HNO3 to reach a pH < 2. Glassware used in sample analyses were washed the same way as collection bottles. Samples were collected from McCloud Lake using an acrylic Kemmerer sampler and were filtered using in-line Gel man Type A-E glass fiber filters that had been rinsed with 250 mL lake water. In-line glass fibers were also used to collect samples for -^^N analysis from the microcosm experiments; filters in this experiment were rinsed with 100 mL HCl + 250 mL DDW. Samples in the sediment titration studies and in the pore water study were filtered using glass fiber filters mounted in a syringe-type holder; the entire apparatus was rinsed with 3 fillings of DDW between samples. Analytical Methods Cations Calcium, magnesium, potassium, and sodium were determined by flame atomic absorption (AA)(Varian Model 175 or Perkin Elmer Model 5000) following methods recommended by the manufacturers. Lanthanum oxide was added to prevent ionization interferences in calcium and magnesium analyses (APHA 1981). Aluminum was analyzed by flameless AA on the Perkin Elmer instrument using settings recommended by the manufacturer. Nutrients Nitrogen species (ammonium, nitrite, and nitrate) were generally determined by AutoAnalyzer techniques (Table 2-1). Kjeldahl nitrogen was determined by a scaled-down micro-Kjel dahl technique (APHA 1981) in which 50 mL samples were digested in a block digester. Ammonium in

PAGE 16

9 the digestate was determined by AutoAnal yzer. Nitrate in the pore water study was determined by ion chromatography (described below). Major Anions Sulfate and chloride were determined by AutoAnalyzer methods (Table 2-1) for 1) McCloud Lake water, 2) precipitation and seepage meter samples, and 3) samples from the sediment-water microcosms. Ion chromatography (Dionex Model 10) was used to determine sulfate and chloride concentrations in the seepage column experiments and in the pore water study. The ion chromatograph was operated at a flow rate of 40% of maximum {r^ 1.0 mL/min) with a 10 cm pre-column, a 250 cm separator column, and a 250 cm supressor column (Dow Chemical Corporation) M Analyses of pH were conducted using an Orion Model 811 pH meter with a Corning Model 910200 pH electrode at the University of Minnesota) and an Orion Model 801 pH meter with the same probe at the University of Florida. Al 1 pH analyses were conducted within eight hours of sample collection. Fischer Certified pH Standards were used to calibrate the pH meter at 4.00 and 7.00 prior to each use. Measurements were made under quiescent conditions, as recommended by Galloway et al. (1979). Isotope Ratios Samples for isotope ratio analyses (^^N/^'*N) of ammonium from the microcosm experiment (Chapter 5) were prepared by alkaline distillation (APHA 1981) of 400 mL aliquots into 0.01 N H2SO4. The first 50 mL of distillate was evaporated to < 2 mL by gentle boiling on a hot plate, cooled, and poured into an AutoAnalyzer tube. To assure

PAGE 17

10 Table 2-\. AutoAnalyzer methods. r u 1 ui 1 1 Mpthnd and rpfprpnrp^ Detection 1 imit, mn / I MnHifiratinn^ Sul fate Mpthvl thv/mnl hi up 1 IV 1 uiiyiiivji L^iuc IM 226-72W 0 Z Fnr nrpr i ni t t i on <^amnlp^ BaCl2 and MTB reduced to improve sensitivity. Phi nri dp FprrirVf^mdp 1 C 1 1 1 ^ Jr Oil 1 U C 0.2 "^arrinl p run*; rin<;pd with DDW. Nitrate and nitrite Cd reduction/ diazotization. IM 158-71W/B 0.001 Cadmium wire used instead of granules at U of F. Ammonia Phenate IM 154-71W/B 0.003 EDTA substituted for r i t r3 tpta rt r3 tp rpaopnt (APHA 1981). Fresh DDW used for standards. In later experiments, standard treated as standard addition to compute NH^ in DDW. TKN Kjeldahl digestion APHA 1981 Digestion modified by using 50 mL aliquots and block digester. NH3 determined by automated phenate method. ^IM = Technicon, Inc. Industrial Methods Series.

PAGE 18

11 adequate N2 production for isotope ratio analysis, 1.00 mg N as NH4CI was added to each sample prior to distillation. Isotope ratio analyses were conducted by Bill Portier at the University of Florida Soil Science Department using a Micromass 602D double beam mass spectrometer. Qual ity Assurance The correctness of chemical analyses was checked periodically using external quality control standards (EPA "mineral" and "nutrient" standards). A low ionic strength pH standard, prepared using H2SO4, was used to check pH calibrations, as recommended by Galloway et al. (1979). Sulfate analysis was used to compute the exact pH of the low ionic strength stock. Analytical errors were also checked by ion balance calculations. If al 1 significant ions are measured for a water sample, the sum of anions should equal the sum of cations. The difference between these terms can be assumed to represent the analytical errors. Errors in ion balances were calculated from the equation E = (A~ m;^) X 100 (2-1) where E = error, as %, A^ = sum of anions, meq/L, and M = sum of cations, meq/L.

PAGE 19

CHAPTER 3 SEDIMENT BUFFERING IN McCLOUD LAKE Introduction The development of models to predict the pH response of lakes to inputs of acid precipitation is a major objective of acid precipitation research. Models currently in use or being developed are either empirical (e.g. Aimer et al. 1978) or focus on processes occurring in the watersheds of lakes. Several models treat watershed acidification as a large scale titration in which protons entering the system via precipitation are replaced by cations, principally calcium and magnesium (Henriksen 1980; Wright 1982; Thompson 1982). Other models have been formulated to model short-term variations in streamflow chemistry (Chen et al. 1979; Christopherson and Wright 1981). The role of pH buffering mechanisms that occur at the sedimentwater interface has received little attention in models of lake acidification although Schnoor et al. (1983) included an in-lake neutralization term in their "trickle-down" model. The ability of sediments to neutralize acidity may be particularly important for lakes that have a small ratio of watershed area to lake surface area and receive most of their inflow directly from precipitation. These "Type I" lakes (Schnoor et al. 1983), which are particularly susceptible to acidification, are common in northern Wisconsin and north-central Florida. For example, the 13 soft water lakes in the Trail Ridge Region of north-central Florida surveyed by Brezonik et al. (1983b) 12

PAGE 20

13 have a mean watershed arearlake surface area of 7.0 (range = 1.4 to 20) and a mean pH of 5.2 (Table 3-1). However, the role of precipitation in the water budgets of these lakes is even more important than indicated by the watersheds ake area ratios since much of the precipitation fal 1 ing on the sandy soi 1 s of this region passes directly to the regional water table. The objective of this chapter is to examine the potential for sediment neutralization of acid in softwater lakes. The addition of acid or base to sediment-water slurries was used to determine the potential magnitude of inorganic neutralization mechanisms under completely mixed conditions. Other experiments were designed to simulate natural sediment-water interactions. Chemical changes that occur during subsurface seepage were evaluated by passing synthetic groundwater through columns of littoral sediment. Acidification experiments involving laboratory sediment water microcosms and littoral mesocosms were used to examine the capacity of sediments to neutralize inputs of acidity to the overlying lake water. Finally, analyses of sediment pore waters were used to evaluate the extent of sediment neutralization processes J_n situ The principal site for this phase of the investigation was McCloud Lake. McCloud Lake is an ideal site for the study of in-lake neutralization mechanisms since the lake receives nearly all of its water directly from precipitation (Chapter 4). Furthermore, the lake has become more acidic during the past 14 years. Water chemistry data collected during 1968-69 (Brezonik and Shannon 1971) and 1978-79 (Brezonik et al. 1983b) show that the pH of McCloud Lake has dropped

PAGE 21

14 c_> >p +-> Ln • •r— ^ • t-> 10 O E CO CO LD 00 4-> 3 O CO Ln CM LO CO I— 1 CM c o \ O) C 00 E O 3 O S• OJ to Q. X > s3 c: O o c to •r(U t-> r— _I o o ra (0 ^ o O yn to N ^ CT LO I— 1 CO 1 1 — — QJ r— 1 S3 (-> +J ro 3 (£> o CO LO I— 1 Z m ID o CM .—1 o CVJ o c Q. • LD LO Ln Ln +J Q. c 0) 0) jC Q E 4-> Q. • C -o 0) •r~ 0) "O CO to (/I o o VD 1 Ln o CM 00 c C CT) o c ro CM CO CO CO I— 1 o O) V ^ to S E • 3 fO to o '4(O 00 LD O o t— 1 Ln SQ) .—1 in CO CM r— 1 QJ (/) 3 SCM c 3 u C/) fO o N (/) (O x_ ra • • • CQ z. Llto to to o O) u. LiLl. •r— rE s_ 3 3 o 4o ft Ll_ s_ (0 o ft to s_ o o o o ft • • • o o o o O o to o ro O o C_) n3 o E E 3 E O +-> n3 J= to to to ra n3 c c o £ >^ ro (T3 C u +J +-) +J n3 <— 1 o 3 3 ra to 1 _l o. Q. Q. C_) > > > T3 C CO o n3 u sO O to > n3 ro c o C_5 ra -C ra c: ro ro O -a o o 13 Q _l <: s: <: C3 (O J3

PAGE 22

15 from 5.0 to 4.6. The littoral sediments are generally sandy, interspersed with pockets of peat-like material while the profundal sediments are a soft, highly organic ooze that extend to a depth of approximately 6 meters. Sediments from four other lakes in the Trail Ridge group and from three lakes in northern Wisconsin were included in the batch neutralization experiments (Table 3-1). Like McCloud Lake, these lakes lack well-defined inlets or outlets and have water with little buffering capacity (alkalinity <200 meq/L) and pH values < 7.0. Mechanisms for Sediment Buffering Cation exchange and mineral dissolution are undoubtedly responsible for much of the acid neutralization of watersheds (Henriksen 1980; Thompson 1982). In the wel 1 -weathered soils of Florida, kaolinite and gibbsite are the major clay minerals. Both minerals neutralize acidity during congruent dissolution (Stumm and Morgan 1981); kaol inite also has a 1 i mi ted cation exchange capacity (CEC) of 1 ess than 10 meq/100 g resulting from isomorphic substitution and from edge charges. In the sensitive watersheds of northern Wisconsin, the dissol ution of fel dspars and other cl ay mineral s may neutral ize acidity. Organic matter, which has a CEC up to 300 meq/100 g, may contribute substantially to the total cation exchange capacity of sandy soils (Yuan et al. 1967) and must be a major component of the total exchange capacity of lake sediments, which often have an organic matter content over 80 %. Humic acids in lakes are known to precipitate upon acidification, a process that consumes protons and may account for the increased clarity observed in acidified lakes (Aimer

PAGE 23

16 et al. 1978). Conversely, precipitated humic acids may buffer inputs of base, such as lime, that may be added to neutralize acidity. There is some evidence that these reactions may result in the depletion of cations in surficial sediments of recently acidified lakes. Norton et al. (1981) found lower zinc concentrations near the surface of sediment cores of 20 lakes in northeastern U.S. and Norway and suggested that the zinc mobilization was caused by cation exchange or solubilization of ZnS. Kahl et al. (1982) reported evidence of accelerated leaching of Ca^"*", Mg^"*", Mn^"^, and Zn'*'^ in surficial sediments of three ponds in Maine and proposed that the trend may be the result of recent acidification. Neither of these studies (Norton et al. 1981 or Kahl et al, 1982) related the depletion of cations to the hydrogen ion buffering capacity of the sediments. Cation exchange was found to be of minor importance in the neutralization of Lake 223 in the Experimental Lakes Area (ELA) of western Ontario during a whole-lake acidification project (Cook et al. 1982). However, Lake 223 was acidified to only pH 5.1 and it is not known whether cation exchange may have played a more significant role in buffering at lower pH level s. Although the role of sulfate reduction in pH-buffering of sediments is well-known (Nriagu and Hem 1978), this process has received little attention as a neutralization mechanism in acid-sensitive lakes. The one exception is the study by Cook et al. (1982) which found sulfate reduction to be a major neutralizing mechanism in an experimentally acidified lake. Over a five-year period approximately half of the H2SO4 added to ELA 223 was neutralized by sulfate reduction in the littoral sediments and the anaerobic hypolimnion. This

PAGE 24

17 work clearly suggests the potential for sulfate reduction as a neutralizing mechanism in lakes receiving acid precipitation. Although sulfate reduction in culture is inhibited at pH levels below 5 (Zinder and Brock 1978), this process undoubtedly occurs in neutral microzones in the environment. Several reports indicate that sulfate reduction occurs at significant rates in acidic peat bogs. Hemond (1980) reported that 77% of the $04^" entering Thoreau's Bog, Mass., was retained and suggested that di ssimi 1 atory reduction may have accounted for approximately half of this loss. Collins et al. (1978) isolated and enumerated sul fate-reducing bacteria, incl uding Desul fovibrio desul furicans from a bog that had nominal pH levels of 3.0-4.6 but gave no indication of the activity of the organisms in the bog. Rippon et al. (1980) concluded, from $04^" and profiles and ^^504^experiments, that sulfate reduction was a significant process in a Danish bog (pH 4.8-5.2). These reports clearly suggest that sulfate reduction may be a significant process in acidic environments, including moderately acidic lakes. Sulfate adsorption can occur by two mechansims. Reversible, or non-specific adsorption occurs when 504^" acts as a counterion on a positively charged surface (Hsu 1977). In some watersheds, reversible adsoption results in seasonal stabilization of stream sulfate levels. During summer, when precipitation is relatively acidic, soil surfaces become positively charged and SO^^is adsorbed. During snowmelt, more neutral water containing less sulfate passes through the soil; surfaces are neutralized and sulfate is desorbed. This process has been in watersheds of the Integrated Lake-Watershed Study (ILWAS) in

PAGE 25

18 the Adirondack Mountains (Chen et al. 1979) and in the Birkeness watershed in Norway (Christopherson and Wright 1981). Irreversible sulfate adsorption (also called specific adsorption) involves a substitution of SO^"^ into the inner sphere of a metal hydroxide surface. The mechanism proposed by Rajan (1978) results in the displacement of two hydroxide ions for each sulfate ion adsorbed Al Al / ^OH ^0 ^0 2 + 2H^ + so/ > 2HoO + \ .OH ^0^ 0 Al^ Al (3-1) Irreversible sulfate adsorption results in a net (permanent) accumulation of sulfate and protons in a watershed. Johnson et al. (1980) reported sulfate accumulation rates up to 9.3 kg/ha-yr for a North Carolina watershed, although many watersheds show a balance between sulfate input and output. Sulfate adsorption has been correlated with iron and aluminum content, clay content and the presence of sesquioxides, and it is inversely correlated with organic content (Johnson et al. 1980). Specific adsorption occurs only under acidic conditions and is not a significant process above pH 6-7. Reactions of the nitrogen cycle also consume or generate protons in sediment environments (Figure 3-1). Since approximately one-third of the acidity in Florida precipitation is HNO3 (Brezonik et al. 1983) these reactions are potentially significant in controlling the pH of softwater lakes. In forest and bog ecosystems, most of the nitrate in

PAGE 26

19 PLANT R-C-R NH. R organic nitrogen H NO, 2HSOIL OH NHI 2 R-C-R H" 2H" NH. I R organic nitrogen NH/ NO, Source: Impact Assessment Group, 1983 Figure 3-1. Effect of nitrogen cycle on proton balance.

PAGE 27

20 precipitation is retained (Kerekes et al. 1982; Hemond 1980; Wright and Johannessen 1980). Retention of nitrate by assimilation into organic matter or by denitrification has the same effect on pH since both reactions consume one H"*" per N03' ion consumed. Although the assimilation of NH4''" results in the production of H"*", nitrate deposition exceeds ammonium deposition in Florida and the net effect of terrestial retention of inorganic nitrogen species is the consumption of protons. The effect of nitrogen cycling on the pH of softwater lakes has received little attention. In addition to nitrate and ammonium assimilation, mineralization of organic nitrogen and other reactions of the nitrogen cycle, such as nitrification, could potentially alter lake pH. Although some of these reactions-notably nitrification and denitrification— are inhibited at pH levels below 5 in laboratory cultures (NAS 1978), ample evidence indicates that they may occur in acidic environments (see Chapter 5). Methods Sediment-Water Batch Experiments Initial experiments to determine the neutral izing capacity of sediments were conducted by adding 25 g of wet sediment, 100 mL of distil led water, and 1.0 mL of chloroform to 10 Erlenmeyer flasks. Each flask received a single dose of acid (0.1 N H2SO4) or base (0.1 N NaOH) and was placed on a shaker table for one week. Upon termination of the experiment, water from each flask was filtered and analyzed for pH, Ca^"*", Mg^"^, Na"^, K"^, Si(0H)4, and total aluminum. In later experiments, conducted with sediments from Wisconsin lakes, sequential doses of acid or base were added daily, after first determining that the results were similar to those of the single dose method. Several

PAGE 28

21 other modifications included the use of LPE bottles rather than glass Erlenmeyer flasks and the use of 20 g of sediment. After each experiment, the sediment in each flask was dried to 103 C to determine dry weight, and then ashed at 500 C to determine the volatile solids content. The neutralizing capacity of the sediments was calculated on a dry weight basis: NC^= 100 X [V^Ng^^j 10 P^i-10P% (Vf)]/W (3-2) where = cumulative volume of acid at i^^^ addition, mL, N, = normality of acid or base, pRq = original pH (no acid or base), pH^ = final pH, = volume of water in f 1 ask(incl uding water content of wet wet sediment) mL, W = dry weight of sediment, g, and NC^= neutralizing capacity of sediment at the i*" acid or base addition, meq/100 g. The exchange of metal cations also was computed on a dry weight basis in order to calculate the contribution of each metal to the total buffering capacity. The potential for sulfate adsorption by McCloud Lake sediments was evaluated in a separate batch experiment in which a single dose of 0.1 N H2SO4 (0.1 to 1.0 mL) was added to each of ten LPE bottles containing 20 g sediment and 50 mL DDW. Since chloroform could potential ly alter sulfate adsorption and/or sulfate analysis, the experiment was conducted with and without the addition of 1.0 mL chloroform. The entire experiment was also conducted with a series of bottle blanks that received no sediment. The bottles were placed on a shaker table for a week; upon termination of the experiment water from each bottle was filtered and analyzed for sulfate and pH. Sulfate

PAGE 29

22 adsorption was evaluated by comparing sulfate recovered with sulfate added, after accounting for sulfate levels in controls (no acid added) Seepage Column Experiment An upflow column study was conducted to simulate the flow of groundwater through McCloud Lake littoral sediment. The synthetic groundwater used in this experiment (Table 3-2) represented soil water from a depth of 190 cm at a site adjacent to Lake McCloud that had been repeatedly irrigated with simulated pH 3.0 rain (J. Byers, University of Florida, pers. comm.). This soil water was circumneutral (pH 6.6) and contained appreciable alkalinity (estimated by ion balance) The synthetic groundwater was passed through three 20 cm intact columns of McCloud Lake littoral sediment at a flow rate of "40 mL/day, which corresponded to the highest seepage rates observed during 1981. Three additional columns received the same groundwater acidified to pH 3.5 with H2SO4 (designated "low pH" treatment). A third set of three columns received groundwater acidified to pH 6.2 (through week six) and then acidified to pH 4.0 to observe the effects of rapid acidification. Following a one-month period of pH stabilization, the eluate from each column was collected weekly and analyzed for sulfate, chloride, and nitrate (ion chromatography), alkalinity, major cations, and pH. Microcosm Experiment The influence of sediments on the composition of overlying lakewater was simulated in 10 L microcosms containing 1.0 L of

PAGE 30

23 o ^ 1— •IC 3 O re -i-> -o O) E s2: o 00 o a. s; o o -p •1ta -o -O 3 3 O +-> S00 CT) c 3 4-> CO c o o IT) ro "51CM I— 1 CO CSJ CO CO ro .— 1 I— 1 CO ID o I— 1 CVJ o IX) 1—1 00 o 00 I— 1 I— 1 o o 00 ro CSJ o CVJ o CVJ o 1— 1 1— 1 ro o CVJ CVJ CVJ oo <: ro Lf> <£) oo <— 1 CVJ >— 1 Lf) CD CO <— I t— ( o o oo CVJ CM OO o Lf) CTi oo ro oo 1—1 CVJ o oo Ln o CVJ CVJ CTl oo 00 >* o o oo Ln oo V£) OO I— 1 1—1 .—1 oo Lf) o + CVJ O + I I OO CVJ O ^ I o o t— oo o 00 o (/) -t-> o Q(-> l/l 0) so <+a OJ Q. •r<++J Ci. o OJ •rO O X fO cu E o _I MCD E scu c: 4-> fO C3

PAGE 31

24 McCloud Lake littoral sediment plus 9.0 L of synthetic Lake McCloud water (based on 1981 data). Following a one week period of pH stabilization, duplicate microcosms were acidified to pH 5.0, 4.5, 4.0, and 3.5 using 1.0 N H2SO4; an additional set of duplicates was used as controls. Repeated additions of acid were required to maintain the desired pH levels. Acid inputs were based on pH values determined immediately before each addition; the vol ume of H2SO4 added on each date was sufficient to reach the desired pH level in the absence of neutralization. Samples collected after 4U days (prior to the addition of labelled algae and the beginning of the decomposition experimentsee Chapter 5) were filtered and analyzed for alkalinity, sulfate and chloride, major cations, and pH. Components of buffering were inferred by comparing the chemical composition of the microcosms at 40 days with the initial composition. Littoral Mesocosms A mesocosm experiment conducted in the littoral region of Lake McCloud to evaluate the effects of pH alterations on biological processes (see Chapter 5) also served as a preliminary experiment on insitu pH buffering. During March, 1981, three 4.0 m diameter enclosures (designed after Landers 1979) were placed in the lake at a depth of rJi meter and anchored firmly to the bottom with wooden stakes to enclose an area of 12 m^. One enclosure was acidified to pH 3.6 with 1.0 L aliquots of 0.72 N H2SO4 while a second received 1.0 L aliquots of 0.1 N NaOH to raise its pH to 5.6. A third bag received no acio or base and served as a control. Acid or base was added weekly with the intention of reaching the desired pH level within one month. Continued additions of acid or base were required throughout the 26 week

PAGE 32

25 study to reach or maintain the desired pH levels. Duriny this experiment, samples were analyzed weekly for pH, major cations, sulfate, chloride, and major nutrients. Pore Water Prof i les Littoral and profunda] cores were obtained from McCloud Lake using a Livingston-type corer during Febuary, 1983. Three cores (3.8 cm interior diameter) from each site were dissected into 2.b cm segments in a glove box that had been purged with 02-free nitrogen gas. Individual segments were placed into centrifuge tubes, treated with 0.1 mL 2 N zinc acetate and 0.06 mL NaOH to precipitate ZnS (APHA 1981), and centrifuged for 20 minutes at 15,000 rpm. The supernatant was filtered and analyzed for sulfate, chloride, and nitrate by ion chromatography. Resul ts and Discussion Sediment-Water Batch Studies Magnitude of neutral ization Results of the sediment batch studies are shown as buffering capacity curves for the McCloud Lake littoral and profundal sediments (Figure 3-2). Each point in Figure 3-2 represents the neutralizing capacity for a given pH, as calculated from equation 3-2. The profundal sediment had far more neutralizing capacity than the littoral sediment, as seen in the much steeper slope for the profundal sediment. To facilitate comparisons of neutralizing capacities among sediments, neutralizing capacities were computed over defined pH ranges. Thus, the NC4^5_jj^q is the neutralizing capacity, expressed as meq/100 g, over the pH range 4.5 to 5.0 and the NC5^[j_5^5

PAGE 33

26

PAGE 34

27 neutralizing capacity between pH 5.0 and pH 5.5. As seen in Table 3-3, the profundal sediments typically have NC4^5_5^q values of 8-10 meq/100 g while the littoral sediments, which are considerably more sandy, have much lower buffering capacities (NC^ g.g^Q =0.6-1.2 meq/100 g). Although there are not enough data points for sediments having intermediate levels of volatile solids to permit formal regression analysis, it is clear that the volatile solids content, or some factor associated with volatile solids content (such as clay fraction) is strongly related to the buffering capacity of sediments. Cation exchange Changes in cation concentrations during titration of the McCloud Lake sediment (Figure 3-3) suggest that ion exchange was the major buffering mechanism. Although mineral dissolution and precipitation could account for these results, minerals containing these cations are probably not present in McCloud Lake. Furthermore, the H'^ neutralization occurred rapidly and was essentially complete within a few hours. This behavior is typical of cation exchange, whereas mineral dissolution, especially of clay minerals, is generally much slower. When NaOH was added, protons on sediment exchange sites were replaced by cations from solution (primarily Ca^"*" and Mg^"*"), and their solution concentrations decreased. When H2SO4 was added, base cations on sediments surfaces were replaced by protons, and cation concentrations in solution increased. Calcium and magnesium exchange accounted for over 50% of the total buffering capacity of McCloud Lake sediments throughout most of the pH range (Figure 3-4). Exchange of Na"^ was significant at pH levels over 6 when NaOH was the source of OH" in the experiment. Thus, some of the Na"^ added displaced protons on the

PAGE 35

28 Table 5-3. Buffering capacity of some Florida and Wisconsin lake sediments. Acid neutralizing capacity meq/100 g Volatile pH pH solids Sand Dry weight Sediment 4.5-5.0 b.0-5.5 % dry wt % dry wt % wet wt Littoral Anderson-Cue, Fla. 1.15 1.0 18.6 30.2 45.8 Adaho Ma. U.D 0.5 9.0 41 7 47.6 Johnson, Fla. 0.6 2.1 11.3 38.0 46.2 Geneva, Fla. 1.1 12.4 27.5 37.0 McCloud, Fla. 1.0 0.5 10.4 31.2 Profunda! McCloud, Fla. 8.3 2.2 73.3 0 8.6 Clara, Wis. 8.7 3.3 62.6 0 5.8 McGrath, Wis. 10.3 7.0 37.1 0 3.2 Sand, Wis. 9.0 7.0 67.3 0 5.4

PAGE 36

29 Figure 3-3. Cation concentrations versus pH in batch neutralization experiment with McCloud Lake profunda! sediment.

PAGE 37

30 Figure 3-4. Contribution of cations to total neutralizing capacity of McCloud Lake profundal sediment.

PAGE 38

31 sediment surfaces, which contributed to the neutralization of the added OH". At pH 6.0, Na"*" exchange accounted for nearly 50% of the total neutralization capacity. Concentrations of dissolved aluminum increased significantly upon acidification, but aluminum solubilization contributed less than 10% to the total neutralizing capacity (even assuming a charge of +3) throughout the pH range 4.0-7.0. Final ly, a portion of the buffering could not be attributed to cation exchange or mineral dissolution. This component increased to 30% of the total neutralizing capacity at pH 6.6 and probably resulted from the dissolution of humic acids. This hypothesis is supported by the observation that the sediment-water solutions became visibly colored at pH levels over 6.0. Experiments with sediments from other lakes in the Trail Ridge Region and from northern Wisconsin lakes show that exchange of Ca^"*" and Mg" is the major mechanism of pH buffering in all of these lakes. Calcium exchange accounted for 41-72% of the total buffering in these lakes, while Mg^"*" exchange accounted for 20-40% of the buffering in the Florida lakes but only 8-9 % of the total buffering in the three Wisconsin lakes (Table 3-4). Aluminum solubility Although dissolution of aluminum accounted for less than 20% the total neutralizing capacity of the Florida lakes (again assuming a charge of +3/mole), aluminum solubility is of interest because aluminum toxicity is a major probl em associated with 1 ake acidification. Previous studies (Johnson et al. 1981; Driscoll et al. 1982) have shown that Al^"*" solubility is controlled by gibbsite, although several

PAGE 39

32 Table 3-4. Components of buffering capacity for selected Florida and Wisconsin lakes titrated to pH 4. % of total neutralizing capacity Final 2+ ?+ + + Lake pH Ca^ Mg^ Na K kV ECa+Mg Johnson 4.46 35.7 39.3 1.2 2.5 75.0 Geneva 4.83 63.3 24.7 2.3 11.1 88.0 Anderson-Cue 4.07 46.5 22.0 1.2 18.4 68.5 Adaho 3.93 46.2 40.0 <.l 2.6 86.2 McCloud: center 1 ittoral 3.96 3.90 41.2 71.8 20.9 20.7 0.5 .25 .25 15.5 62.1 92.5 Clara 3.61 44.0 8.9 1.1 1 .2 52.9 Sand: center 3.72 49.6 8.3 0.5 1 .3 57.9 McGrath 3.44 43.3 7.6 0.2 1 .0 50.9

PAGE 40

33 other aluminum-bearing minerals, including kaolinite, amorphous aluminum hydroxide, and feldspars potentially could control aluminum solubility. Nordstrom (1982) have recently shown that several sul fatebearing minerals, notably jurbanite (Al (SO4 )(0H)) and alunite (K{S04)(A1 )3(0H)g), may control aluminum solubility in highly acidic water such as mine drainage. Total aluminum concentrations in the batch experiments with Florida sediments increased with increasing acidity below pH 5 but were relatively constant (10"^ to 10"^ M) above pH 5 (Figure 3-5). Also shown in Figure 3-5 are total aluminum concentrations (sum of hydrolysis species) predicted from the solubility of gibbsite (log K = -33.2 from May et al. 1979) and from the solubility of kaolinite (log K=-38.7, Stumm and Morgan 1981). Since silica levels remained nearly constant thoughout the pH range, the mean concentration of 44 uM was used to compute kaolinite solubility. These data clearly show that total aluminum in the sediment titrations was generally above the levels predicted from the dissolution of these two minerals. The most likely reason for the discrepancy between measured total aluminum and predicted solubility is that the measured total aluminum includes organic complexes and fluoride complexes that were not accounted for in these solubilitiy calculations. Driscoll et al. (1982) have shown that organic Al complexes accounted for an average of 44% of the total monomeric aluminum and that Al-F complexes accounted for 29% of the total monomeric aluminum in a group of Adirondack lakes. Fluoride complexes are not likely to be important in McCloud Lake because the lakewater contains very little fluoride (<0.003 meq/L). However, water in the experimental flasks often became visibly colored at pH

PAGE 41

34 PH Figure 3-5. Aluminum concentration versus pH in sediment neutralization experiments for five softwater Florida lakes.

PAGE 42

35 levels above 5, indicating the solubilization of humic acids. These humic acids undoubtedly formed Al complexes that kept total aluminum levels in solution far above levels predicted on the basis of inorganic complexes. Without further information on the extent of organic complexes, the control of Al solubility cannot be determined precisely. Solubility calculations fortheTrail Ridge lakes, using data from Brezonik et al. (1983b), were conducted with the assumption that total aluminum represented free Al^"*". These calculations show that several minerals, including kaolinite, gibbsite, and alunite, may control aluminum solubility. The possibility of alunite precipitation is particularly interesting, since this process would be a sink for sulfate as well as a H'''-neutral ization mechanism. However, alunite precipitation would be accompanied by decreased concentrations of K"*" and SO4 in the batch studies and these ions were conservative for the McCloud Lake sediments. Sul fate adsorption The sulfate adsorption experiment showed that sulfate recovery was at least 100% of sulfate added for both littoral and profundal sediments (Figure 3-6); the addition of chloroform to the bottles had virtually no effect on sulfate recovery in either blanks or treatment bottles. Although it is not clear why sul fate recovery was usually slightly greater than 100%, this experiment shows that sulfate adsorption or other abiotic sul fate-reduci ng processes, such as the precipitation of alunite or other sul fate-containing mineral (See discussion of aluminum solubility), were unimportant as neutralizing mechanisms within the pH range of this experiment (4.0-5.5).

PAGE 43

36 2.5r 2.0 1.5 3 ^ 1.08 0.5 / '/ / / / ^ A = Littoral v= Littoral + CHCL3 B= Profuneal = Profundal + QICL7 0^ 0.5 1.0 1.5 2.0 2.5 [S(^'] ADDED. tiEo/L Figure 3-6. Sulfate recovery in batch acidification of McCloud Lake sediments.

PAGE 44

37 Seepage Column Experiment Neutralization of H'^ occurred in the upflow column experiment to the extent that the eluate from al 1 three sets of columns had pH values between 5.0 and 7.0 following the stabilization period (Figure 3-7). Changes in the composition of major ions in the column eluates (Figure 3-8) occurred in all three sets of columns. In the high pH columns (inflow pH = 6.6), concentrations of Ca^"*", Mg^"*", HCO3", SO^^", and NO3" decreased, but pH remained nearly constant. During weeks 5-8, the reduction in divalent cations (Ca^^ and Mg^"^) of 0.20 meq/L was similar in magnitude to the reduction in alkalinity (0.18 meq/L). These results can be accounted for by the following mechanism: nHC03+ + nH+-X — -> nH2C03 + M-X (3-3) The inference to be made is that alkalinity and divalent cations leached from the soil by acidic precipitation are removed from seepage water as it passes into the lake through the littoral sediments. In the 1 ow pH col umns (inf 1 ow pH = 3.5), nearly al 1 (99%) of the influent protons were neutralized by the sediment. Calcium, which was initially retained by the columns (weeks 5-7), later was leached from the columns (Figure 3-8). Although magnesium was leached from the column throughout the study, the extent of leaching increased by week 14. Calcium retention nearly balanced magnesium leaching during weeks 5-8; cation exchange was therefore unimportant in pH-buf f eri ng. Of much greater significance was the reduction in sulfate levels. The average loss of sulfate (0.410 meq/L) during weeks 5-8 accounted for 86% of the sum of proton loss plus alkalinity gain. Although sulfate adsorption potentially could account for the observed results, sulfate

PAGE 45

38

PAGE 46

39 O I 1 C\J o Q. 3 O CO O + 00 CM o + CO to 0) o I CVJ o 00 CO o o I 3 o c cu T3 CU I/) I— +-> (C C CU O E -(-> -r+J s•r o cu CO I CO cu scn ID O + o o Lf) in o o I + o CD o o I + o o in o I n/D3W 'N0IiVMN33N0D NI 39NVH3

PAGE 47

40 adsorption was found unimportant in the sulfate adsorption study (see Figure 3-6). Biological sulfate reduction is a more likely mechanism; sulfate entering the columns was either reduced to and lost to the atmosphere or precipitated in the columns as FeS. Both sulfate reduction and cation exchange were important H"*"buffering mechanisms in the intermediate pH columns. During weeks 56, a loss of sulfate (0.14 meq/L) was exactly balanced by a loss of cations and the pH remained constant. Upon further acidification of the inf 1 ow to pH 4.0 during week 6, the 1 oss of sul fate increased to 0.31 meq/L and was accompanied by a comparable increase in the loss of protons. Thus, even with rapid acidification the pH of the eluate remained above 5.0. By week 14 however, sulfate reduction was diminished and the pH of the eluate decreased slightly. Concentrations of Na"^, K"*", or CI" were unchanged in al 1 three sets of the columns. These results support the conclusion from the batch studies that monovalent ions do not contribute to ion exchange buffering in the sediments. It is interesting to note that NO3" concentrations were decreased by 90 % in all columns. This reduction could be the result of either nitrate assimilation or dissimi 1 atory nitrate reduction. Although both reactions consume protons (see Figure 3-1), the level of NO3" in the column inflows was so low (< 40 ueq/L) that the effect of nitrate loss mechanisms on eluate pH was minor. SedimentWater Microcosms The McCloud Lake sediment-water microcosms exhibited a substantial capacity to neutralize acid added to the overlying water.

PAGE 48

41 Maintenance of the acidified microcosms at the desired pH levels required continued inputs of H2SO4 during the 20-week experiment. As seen in Figure 3-9, predicted pH values, calculated from the cumulative acid additions, were far lower than measured pH levels. By the end of the experiment over 90% of the acid added to each microcosm had disappeared (Figure 3-9). An analysis of major ions during week 7 (prior to the addition of ^^N-labelled algae to begin the decomposition experiment described in Chapter 5) was used to determine the components of pH buffering. Alkalinity was not measured due to time constraints but was inferred from the ion balance of other major constituents. As seen in Table 3-5, cation exchange was responsible for over 60% of the total buffering in all of the microcosms. Calcium exchange accounted for 45-60% of the buffering, while magnesium exchange comprised 15-25% of the total buffering. Sodium and potassium concentrations changed little and were unimportant in buffering. Sulfate reduction accounted for 25-40% of the total buffering in the acidified microcosms, although sulfate concentrations actually increased slightly (^15%) in control microcosms. This small increase may have resulted from sulfate oxidation but also may have been the result of experimental error. In the acidified microcosms, cation exchange plus sulfate reduction accounted for 81% to 104% of the observed buffering. As with the groundwater seepage experiment, the importance of sulfate reduction as a buffering mechanism increased with decreasing pH and increasing sulfate concentration. Littoral Mesocosms The littoral mesocosms also exhibited a substantial capacity for pH buffering (Figure 3-10). During the first 14 weeks of treatment.

PAGE 49

42 0 5 10 15 20 WEEK Figure 3-9. pH of acidified sediment-water microcosms.

PAGE 50

43 — ^ > — c r~ (U iM" d) <+o o o .— •-> Q. m ra E E •-> u 3 O O 00 u +-> I •Ic io +-> -r5C +-> O 3 O ^ oI •rC SO 4-> •O 3 CNJ C7) + CM O CM >^ o OO o0) •rC so O 3 O jQ o•rC so +-> -r•4-> O 3 cr cu I •Ic so +-> -ra c +-> o 3 o ^ (U h M_J cr <4C O) o -rE l/l DIICO Wl o o + c_) — 1 ^ I o SE O (/) •rO s: cj <£> (Ti CO CO O ro IT) CM 00 O CD CTl i-H in O CO o o t— 1 CO LD in o o CO o o CM O O o o o o o o O o o un CM CM CM — r 1 1 t— 1 1 — t 00 I-H vo 1—1 00 I-H O o CM o ^ CM O .—1 CM o ID .—1 o o 1 o 1 o 1 o 1 o 1 cr> o .— 1 1 — 1 CM CO CM cn 1 — 1 in CM VD in I-H cn o O un o CM CM o CM o o o o o o CM CO f — 1 00 ID ID o uo un CM in 00 CM cn CM .— I CM in o CM un o o O o o o CO O ID CO UD in 1 — 1 CTl CM CM CM CO 1—1 CvJ 1 r— 1 VD CM x—t 00 o o un o 1 — 1 CM O O O o o O CM o CO in CTl CTl VD ID un I-H CO 1 — 1 r-H CD CT> CM I-H o in o in O un CO un un

PAGE 51

44

PAGE 52

45 0.46 meq H"''/L was added to the acidified mesocosm, and 80% was neutralized. While the areal buffering rate of the acidified mesocosm (1.2 meq/m^-yr) during the first 14 weeks was comparable to the areal buffering rate in the pH 3.5 laboratory microcosms (1.3 meq/m^-yr), the nature of the buffering was different. Cation exchange, which accounted for 65% of the buffering in the pH 3.5 microcosms, was apparently unimportant in buffering of the mesocosms. Although calcium levels fluctuated considerably during the mesocosm study, there was no sustained increase in calcium or magnesium with acidification (Figure 3-11). The neutralization of added protons was paralleled by a loss of added sulfate, suggesting that sulfate reduction probably was responsible for the neutralization (Figure 3-12). In the neutralized mesocosm, repeated additions of NaOH failed to cause a sustained increase in pH (Figure 3-10). As in the acidified mesocosm, calcium levels fluctuated considerably, but the expected decrease of calcium and magnesium concentrations did not occur (Figure 3-11). Sodium, however, did disappear, and its rate of disappearance closely paralleled the neutralization of added OH" (Figure 3-13). The loss of Na"*" is disturbing because Na"^ was conservative in the batch studies below pH 5. The sodium loss may possibly have occurred as the result of water exchange between the mesocosm and the lake, although it is unlikely that the rapid disappearance of OH" observed during the first month of base additions occurred as the result of water exchange. By week 14, the loss of sodium was equivalent to 71% of the 1 OSS of OH".

PAGE 53

46 Figure 3-11. Calcium and magnesium concentrations in McCloud Lake littoral mesocosms.

PAGE 54

47

PAGE 55

sson %

PAGE 56

49 Precautions taken to prevent mixing of lakewater with the water in the experimental enclosures included anchoring the bottom with stakes and providing a floating collar to minimize wave action, but some exchange undoubtedly occurred. Furthermore, inputs of groundwater and precipitation were not determined and presumably had some effect on the composition of water in the enclosures, even during the relatively short experimental period. In future experiments, it would be useful to use an inert tracer (such as bromide) in control mesocosms in order to ascertain the extent of water exchange between the mesocosms and the open lake. Pore Water Profiles The pore water profiles show that sulfate reduction occurs not only in laboratory microcosms but in Lake McCloud itself. As seen in Figure 3-14, sulfate pore water concentrations for the three profundal cores decreased from 150 ueq/L in the overlying water to < 15 ueq/L below 10 cm. The flux of sulfate to the sediment in this part of the lake was calculated from Fick's law: F = Dc(Ci Ci+i)/z (3-4) where D^. = diffusion coefficient, cm^/sec, C^= concentration at depth i, mg/L, and C^-^-^ = concentration at depth i+1. According to Lerman (1978), the effective diffusion coefficient, D(., is approximately equal to D^Q^, where Dq is the bulk diffusion coefficent and (D is the porosity. Li and Gregory (1974) reported that Dq for SO^^" = 8.9 X 10"^ cm^/sec at 18 C. If the porosity is estimated from the water content (0.9), Dc is 7.2 x 10"^ cm^/sec. From d = 0 to d = 10 cm, the concentration gradient is 136 ueq/L, so the flux rate of SO4 to the profundal sediment is approximately 225

PAGE 57

50 A. PROFUNDAL Core 1 Core 2 Core 3 Lake water Solid line represents mean 50 ICQ 150 200 SULFATE CONCENTRATION, UEQ/L 1x1 o B. LITTORAL o = Core 1 A = Core 2 Core 3 100 200 300 400 SULFATE CONCENTRATION, UEQ/L Figure 3-14. Sulfate profiles in pore water of McCloud Lake sediments, February, 1982.

PAGE 58

51 meq/m^-yr. This rate of sulfate reduction is 54% of the wet-only deposition of 420 meq/m^-yr for the annual period May, 1981-April, 1982. Individual pore water profiles in the littoral sediment pore water were very irregular (Figure 3-14), with concentrations ranging from 56 to 320 ueq/L. These data suggest that distinct pockets of sulfate reduction and oxidation may occur in the sandy, littoral sediments. The irregularity of sulfate reduction was confirmed by the irregular pattern of mottling seen on silver-coated rods inserted in sediment 24 hours prior to collection of the littoral cores. While some reduction and reoxidation may occur in individual core segments, the ratio of SO4 /CI" in the pore water profile is nearly constant with depth (1.21) and is almost identical to the SO4 /CI" ratio in the lake water (1.24). Chloride concentrations in the pore water were comparable to lakewater concentations but higher than concentrations in the more dilute in-seepage (see Chapter 4), suggesting that water was moving out of the lake when these cores were taken. The movement of water may have obliterated sulfate profiles that formed during the summer, resulting in the irregular profiles observed. Thus, while this pore water study confirms the occurrence of sulfate reduction in the profunda! sediments, there appears to be little evidence of sulfate reduction occurring in the littoral zone. Concl usions These experiments demonstrate that reactions at the sedimentwater interface are potentially important in neutralizing inputs of H"*" to McCloud Lake. The major mechanisms of H"*" neutralization observed

PAGE 59

52 were cation exchange and sulfate reduction. Cation exchange, in which Ca^"*" and Mg^"^ were displaced by H"*" on sediment surfaces, provided a neutralizing capacity of up to 10 meq/100 in the profunda! sediments. Exchange of Na"*" and K"*" was unimportant, as was sulfate adsorption. Although aluminum was solubilized in the batch experiments, aluminum sol ubi 1 ity was relatively unimportant as a neutralizing mechanism. Evidence of sulfate reduction was obtained in the seepage column experiments, in both the mesocosm and microcosm experiments, and in the pore water profiles. This mechanism is interesting because it means that sulfate is not a conservative ion in softwater lakes, as is commonly believed, and because sulfate reduction is a proton-consuming process that may contribute to the neutralization of acid precipitation.

PAGE 60

CHAPTER 4 McCLOUD LAKE MASS BALANCE Introduction In the previous chapter, results of lab experiments showed that McCloud Lake sediments are capable of neutralizing acid by several mechanisms. Cation exchange, particularly the replacement of divalent cations (Ca^"*" and Mg^"*") by protons resulted in the neutralization of up to 10 meq/100 g dry sediment. Evidence of sulfate reduction was obtained by groundwater seepage and sediment-water microcosm experiments and from in situ pore water profiles of sulfate. The objective of this chapter is to evaluate the magnitude of inlake neutralization processes in McCloud Lake by constructing mass balances for major ions. Based on the results from the previous chapter, two hypotheses were generated: 1) cation exchange at the sediment-water interface results in the enrichment of the overlying water with base cations, particularly Ca^"*" and Mg and 2) sulfate reduction in the sediments reduces the quantity of sulfate in the overlying water and neutralizes an equivalent amount of acidity. A third hypothesis is that assimilation of N03~ by phytopl ankton and macrophytes, which consumes protons (Brewer and Goldman 1976), is a significant sink for acidity. 53

PAGE 61

54 Mass balances of major ions entering and leaving McCloud Lake during the period September, 1981-August, 1982 were used to evaluate these hypotheses. Fluxes measured or estimated in the mass balances included wet precipitation, aerosol and gaseous dry deposition, and groundwater seepage to and from the lake. Further evidence of neutralizing mechanisms was obtained by evaluating historical water chemistry data for McCloud Lake col lected during 1968-69 (Brezonik and Shannon 1971), 1978-79 (Brezonik et al. 1983b), and 1981-82 (present study) Background Mass balance models have been widely used to evaluate the role of watersheds in neutral izing acid precipitation (Johnson et al. 1981; Galloway et al. 1980; Thomson 1982). Cation exchange and mineral dissolution have been recognized as the most important neutralizing mechanisms in watershed soils, and several regional scale sensitivity models are based exclusively on mineral weathering reactions (e.g., Henriksen 1980). Assimilation of nitrate by vegetation in watersheds may also contribute to the neutralizing capacity of forested watersheds. As noted in Chapter 3, the assimilation of nitrate produces hydroxide ions that consume protons. Several studies have shown that most of the nitrate entering forested watersheds is retained with an equivalent number of protons (see Impact Assessment Group 1983). Finally, sulfate adsorption may be an important neutralizing process in some watersheds, particularly those with sesquioxide-rich subsoils (Johnson et al. 1980). In addition to permanent (ligand exchange) sulfate adsorption, reversible adsorption is believed to account for

PAGE 62

55 short-term stabilization of SO4 concentrations in watersheds (Chen et al. 1979; Chri stopherson and Wright 1981). Few mass balance models have been compiled for sensitive, softwater lakes and the role of in-lake processes on the H"*" balance of lakes is not well understood. However, several recent reports suggest that ani on-consuming processes may be important in regulating lake pH. The potential role of sulfate reduction in neutralizing inputs of acid precipitation to softwater lakes was elucidated by Schindler et al. (1980) and Cook and Schindler (1983). Following the artificial acidification of Lake 223, these authors found that the reduction of sulfate in the anaerobic hypolimnion and littoral sediments accounted for most of the in-lake neutralization in the first two years of acidification (Schindler et al. 1980) and consumed 50% of the total sulfate input during the five years following acidification (Cook and Schindler 1983). Kilham (1982) reported that acid precipitation may actually increase the alkalinity of non-sensitive lakes. His hypothesis is that the protons in precipitation are neutralized by cation exchange and mineral weathering. Since the consumption of acidic anions (NO3" and S04'^") al so consumes protons (or produces OH", i.e. al kal inity), the addition of nitric and sulfuric acids to non-sensitive watersheds can result in an increase in alkalinity production. Kilham (1982) demonstrated that nearly all of the NO3" and 65% of the SO^^" entering the watershed of eutrophic Weber Lake, Michigan, was retained by the lakewatershed system and concluded that nitrate assimilation and sulfate reduction were sufficient to neutralize acid precipitation. Acid precipitation also appeared to have increased the mineralization rate

PAGE 63

56 in the watershed and may have caused a doubling of alkalinity observed over the past 30 years. Kelley et al. (1982) also postulated that hypo! imnetic OH" production resulting from anion-consuming processes may be sufficient to neutralize inputs of acid precipitation to eutrophic lakes. Finally, in what appears to be one of the only mass balance models reported for an undisturbed, softwater lake, Wright and Johannessen (1980) found that 30% of the H''" entering Langtern Lake, Norway (6 keq/km^-yr) was neutralized together with 20% (9 keq/km^-yr) of the SO4 These authors concluded that sulfate reduction may have accounted for the observed H"*" and SO^^" depletion. Methods Wet Precipitation Wet-only precipitation was collected using an Aerochem-Metrics Model 101 sampler placed on a raft in the center of McCloud Lake. This sampler is designed so that the wet sample bucket is open only during precipitation events and is tightly sealed with a lid between events to prevent evaporation and contamination by birds, insects, and other debris. Samples were collected monthly and analyzed for pH, major ions, and nutrients. Dry Deposition Ambient air concentrations of aerosols (SO^^", NH^"^) and gases (SO2, NO2, and HNO3) were measured during 20 24-hour collection periods during August and September, 1981. Dry deposition rates of these constituents were computed using the equation (Chamberlain and Chadwick 1953)

PAGE 64

57 F = Vd X (4-1) where F= areal deposition rate, Vjj = deposition velocity, and = concentration at a reference height, usually one meter. Dry deposition rates of other aerosol constituents (Ca^"*", Mg^"*", K"*", Na"*", and CI") were estimated from two northern Florida data bases. Edgerton (unpublished data) recently conducted a study of dry deposition throughout northern Florida that included one remote site similar to McCloud Lake. The study of Brezonik et al. (1983b) included wet and dry deposition measurements for four inland Florida sites and bulk deposition measurements for 25 sites, including four near McCloud Lake (Gainesville, Jasper, Hastings, and Waldo). Two methods were used to estimate dry deposition of aerosol constituents at McCloud Lake from these data. First, the ratios of wet to total deposition for the four wet/dry col lectors were used to estimate dry deposition at McCloud Lake from measured wet deposition. The second method was to subtract McCloud Lake wet deposition from bulk deposition measured at the four 1 ocal bul k col 1 ectors. Seepage Seepage flow to and from McCloud Lake was measured using 22 seepage meters (Fellows and Brezonik 1981) placed along six transects perpendicular to the shoreline (Figure 4-1). Flows were measured once a month. When the direction of flow was towards the lake, samples of the inflow groundwater were collected for analyses of major ions and nitrogenous species. Flux rates of water and ions were calculated by dividing the lake into five concentric rings having boundaries that paralleled the shoreline and were equidistant from adjacent seepage

PAGE 65

Figure 4-1. McCloud Lake seepage meter transects.

PAGE 66

59 meters along each transect. The outer three rings were further divided into 6 subdivisions whose boundaries were equidistant from adjacent transects. Thus, a total of 20 regions were formed, each represented by one, or, for the two inner regions, two seepage meters (Figure 41). The flux of constituent i was computed from the equation n Fi = j X Qj (4-2) j=l ^ where j = concentration of constituent i in seepage region j, Qa'= flow in region j. (Sign convention: "+" = inflow; ^ "-" = outflow). Groundwater Wei 1 s Two to four wells (2 diameter PVC pipes) were placed in the ground along each transect (see Figure 1-1) in order to permit sampling of the local groundwater. During the second half of 1982, samples from these wells were collected for analyses of major ions and nutrients. Lake Storage A bathymetric map of McCloud Lake was used to construct stagevolume and stage-area relationships. (See Figure 1-1). Stage measurements were made 2-6 times per month and the mean stage reading was used to compute lake volume and area for the month. Water samples col lected monthly at 1-meter interval s near the center of the lake were analyzed for major ions and nutrient species. Since the lake never stratified, storage calculations were based on mean concentrations of samples collected at 3-4 depths. Evaporation Pan evaporation measured at Lisbon, Gainesville, and Lake City (CI imatological Abstracts 1981 and 1982) was used to compute a distance-weighted mean representative of evaporation at McCloud Lake:

PAGE 67

60 E^^ = E(D/di X Ei)/E (D/d^.) (4-3) 1=1 i=i where E^^j^, = distance-weighted mean of pan evaporation, cm/month, = distance to station i, D = total distance, i.e., d^ and E^= pan evaporation at station i, cm/month. A pan coefficient of 0.70 was used to estimate lake evaporation from the Class A pan evaporation data (Linsley et al. 1975). Results and Discussion Water Budget Components of the McCloud Lake water budget are shown in Figure 4-2 (see Table A-1). To check the accuracy of the water balance, monthly precipitation (P), evaporation (E), and net seepage (G) were used to model the change in storage (AS) and the new storage (S) for each month, beginning with the measured storage in August, 1981: AS^ = E^ + G (4-4) Si = Si_i + ASi (4-5) where i = month. Although the computed S of ten di ffered from the measured S for individual months (Figure 4-2), computed S and AS agree well with measured values for the entire year. At the end of the year, computed storage was only 2% less than the measured storage, and the computed annual change in storage (+ 17.45 x 10"^ m^) was 15% less than the measured change in storage of 20.47 x 10"^ m^. The difference between predicted and measured AS was only 3.6% of the total water input (precipitation + in-seepage). Precipitation accounted for 90 % of the total water input, while in-seepage accounted for the remaining 10%. It should be noted that

PAGE 68

61 the contribution of seepage to the water budget was far less than suggested by the watershed:! ake surface area ratio of 20:1. Because of the sandy soils, gentle slopes, and a lack of defined stream channels, precipitation falling on most of the watershed passes directly to a regional water table rather than to a lake or outflow stream. Although precipitation during the model year was 6% above the 40 year mean, the previous year was very dry and had an annual precipitation of only 57% of the 40 year mean. Thus, seepage during the first seven months of the model year was outward, resulting in recharge of a depleted water table. During the last five months, continued precipitation and a rising water table resulted in seepage flows into the lake (Figure 4-2). Seepage occurred mainly in the sandy littoral area. As shown in Figure 4-3, seepage rates during both outflow periods (e.g., October, 1981) and inflow periods (e.g., April, 1982) diminished with distance from shoreline. This was particularly evident during the inflow period. During April, 1982, for example, the mean flow rate decreased from 1.1 cm/d at 3 meters from shore to nearly 0 cm/d at 25 m (Figure 4-3). However, since much of the total lake area was in the central low-seepage zone, the central region of the lake was significant in terms of total seepage flow. Thus, 72% of the total outflow but only 4% of the total inflow occurred in the 71% of the lake within the interior of the 17.5 m contour. The water residence time, based on total inflow (precipitation + seepage), was 1.7 years. Most of the water entering the lake evaporated (63% of the total inflow) and only 16% left the lake via outseepage (the remaining inflow resulted in a change in storage). The

PAGE 69

62 12 10 o CO 8 s: CO o 6 1 4 LU LU ? o > U • 10 o CO 8 2^ CO o T-l 4 LlJ 2 o u O 4 CO 2 s: CO 0 o 1—1 -2 LU -4 _l o > -6 10 o 5 CO 0 CO o .—1 -5 LlJ -10 e: ZD —1 -15 OA PRECIPITATION EVAPORATION CHANGE IN STORAGE 1 r__J— J 1 Measured j Predicted Sept. Nov. Jan. Mar. May July Oct. Dec. Feb. Apr. June Aug. 1981 1982 Figure 4-2. Water balance for McCloud Lake, September, 1981, to August, 1982.

PAGE 70

63 >3-0.5 o -1.0 OCTOBER, 1981 (FLOW FROM LAKE) Solid line shows mean 10 20 30 40 DISTANCE FROM SHORE, M 50 2.0 I1.5 >1.0 0.5 0.0 -0.5 APRIL, 1982 (FLOW TO LAKE) Sol id 1 ine shows mean 10 20 30 40 DISTANCE FROM SHORE, M 50 Figure 4-3. Seepage flow versus distance from shore.

PAGE 71

64 residence time based on outflow was 9.6 years. Since volatilization of substances other than water entering the lake is minimal or nonexistent, the outflow-based water residence time should be the residence time for conservative ions. Chemistry of McCloud Lake. Past and Present Data collected during 1968-1969 (Brezonik et al. 1969), 1978-79 (Brezonik et al. 1983b), and in the current study (Tables A-2 and A3) allow an evaluation of historical water chemistry trends for McCloud Lake. Before discussing these data, several problems concerning the evaluation of historical water chemistry trends must be considered. The most serious problem with comparing data collected over a span of 14 years is that certain analytical methods have changed significantly. While methods for analyzing cations were similar in all three studies (cal cium, magnesium, sodium, and potassium were analyzed by flame atomic adsorption), methods for collecting pH samples and for analyzing sulfate and chloride have changed. In the 1968-69 study, pH samples were collected in glass D.O. bottles "to prevent CO2 transfer with the atmosphere" (Brezonik et al. 1969, pg. 83). Although Kramer and Tessier (1982) have pointed out that soft glass bottles may contribute 20 100 ueq/L alkalinity, the same authors concluded that the use of glass containers would not be be a problem in studies where analyses were conducted immediately following sample collection. In the study of Brezonik et al. (1969), prompt (same day) analysis was done, as inferred by the stated concern for carbon dioxide transfer. Thus, comparisons of the 1968-69 pH data with the data from the two more recent studies (in which pH samples were

PAGE 72

65 collected in distilled water-soaked linear polyethylene (LPE) bottles and analyzed within eight hours) probably are valid. The most significant methodological change involves sulfate analyses. Sulfate was analyzed by the turbidimetric method in the 1968-69 study (APHA 1971) which, in addition to being an awkward technique, has a detection limit of only 1 mg/L. The reliability of the reported mean sulfate concentration of 2.2 mg/L is therefore questionable. Sulfate analyses in 1978-79 and in the current study were done by the automated methyl thymol blue technique, modified to give a detection limit of 0.1 mg/L. Methods of chloride analysis also changed during the 14 year observation period. Chloride in the 196869 study was determined by mercuric nitrate titration, using the "low level" modification (APHA 1971), while in the recent studies chloride was determined by the automated ferricyanide procedure (APHA 1981), also modified to give a sensitivity of 0.1 mg/L. Both techniques, however, are generally free of interferences and considered reliable. Although concentrations of chloride in McCloud Lake are so low that only 1 mL of titrant would be required to reach the endpoint with the mercuric nitrate titration, small bore burets were used to assure that smal 1 vol umes coul d be del i vered accuratel y. The excess of measured anions in the 1968-69 study (10% of the sum of ions) indicates that analytical errors were a problem (Table 41). Ion balances were more exact in the 1978-79 study (mean error = 5%) and in the current study ( mean error = 3.5 %). A second problem in evaluating historical trends in acidification is that McCloud Lake undergoes substantial changes in volume as a result of long-term variations in precipitation. Between September,

PAGE 73

66 t 1981, and August, 1982, for example, the volume increased by 20% because of increased precipitation following several drought years. These volume fluctuations cause changes in the concentrations of dissolved substances by dilution or evaporative concentration that may obfuscate long-term changes in water quality. One way to eliminate changes caused by dilution/concentration effects is to normalize the concentrations of dissolved ions to the concentration of a conservative ion such as chloride. Both concentrations and ion/chloride ratios are shown in Table 4-1. As shown in Table 4-1, McCloud Lake has become more acidic during the past 14 years. Hydrogen ion concentrations have nearly doubled from 14 ueq/L (pH 4.9) in 1968-69 to 32 ueq/L (pH 4.5) in 1982, while sulfate concentrations appear to have increased from 104 ueq/L in 1968-69 to 142 ueq/L in 1978-79 and 173 ueq/L in 1981-82 (Table 4-1). Reported values of sulfate in 1968-69 may be erroneously high due to analytical problems associated with the turbidimetric method (discussed above), and sulfate levels may have increased more than these data indicate. The excess of measured anions in the 1968-69 data suggests that reported values of one or more anions were too high or that reported cation values were too low. Increased acidification has apparently resulted in enhanced leaching of base cations. Calcium concentrations in McCloud Lake have doubled from 30 ueq/L in 1968-69 to 66 ueq/L in 1981-82, while magnesium levels have increased by 34%, from 47 ueq/L in 1968-69 to 63 ueq/L in 1981-81. These data must be interpreted with caution because of the poor ion balance associated with the 1968-69 data. The

PAGE 74

67 Table 4-1. Chemical composition of McCloud Lake, 1968 to present. 1968-69^ 1978-79'^ 1981-82 Cone. uea/L Ci/Cl^ Cone. uea/L Ci/Cl*^ Cone. iieo/L Ci/Cl 14 0.08 19 0.13 32 0.19 Ca^-^ 30 0.18 47 0.32 66 0.39 47 0.28 51 0.35 63 0.37 6 0.04 15 0.10 6 0.04 Na^ 122 0.73 121 0.83 153 0.90 SO4" 104 0.62 142 0.98 173 1.02 CI 167 1 .00 145 1 .00 170 1.00 219 253 320 2A" 271 287 343 % Error 10.6 6.3 3.5 ^Brezonik and Shannon (1971). ^Brezonik et al (1983b). Ratio of ion i to chloride.

PAGE 75

68 calcium data are particularly suspect, since ratios of Ca /Mg'^ have changed significantly since 1968-69. Although the Ca^"*"/Mg^''' ratios in 1978-79 and 1981-82 were close to 1.0, the ratio in 1968-69 was only 0.6. No other investigators have reported such a shift in cation ratios associated with acidification, and it is not clear to what extent the observed trend reflects analytical errors. The early calcium data are particularly suspect since calcium analysis is susceptible to interferences and because an increase in the early calcium values would produce Ca^'*"/Mg^''' ratios closer to 1.0. Chloride concentrations were relatively stable during this period. Although the mean chloride concentration in 1978-79 (145 ueq/L) was slightly lower than the 1968-69 mean (167 ueq/L), the current mean chloride concentration (170 ueq/L) is nearly identical with the 196869 mean. These data indicate that dilution or concentration resulting from variations in relative rates of precipitation and evaporation cannot account for the observed differences in H"*", SO^^", Ca^"*", or Mg Sodium and potassium levels have changed little since the first study; this is consistent with the observation that these ions were nearly conservative in the sediment titration experiments (Chapter 3). Concentrations of major cations and anions in the lake during 1981-82 are shown in Figures 4-4 and 4-5. Constituents which are derived primarily from precipitation, including sodium, chloride, and sulfate, fluctuated considerably during the study as a reflection of variations in lake volume. Sodium concentrations, for example, increased from 125 ueq/L during October, 1980, when the lake volume was 157 X 10-^ m^ to 229 ueq/L by December, 1981 when the volume had decreased to 120 x 10 m'^. As the lake volume increased, sodium

PAGE 76

69

PAGE 77

70

PAGE 78

71 levels declined. By the end of the study, the lake volume had increased to 148 X 10-^ while the sodium levels declined to 139 ueq/L. Concentrations of other base cations were relatively stable. Although calcium concentrations decreased from 112 ueq/L to --^ 60 ueq/L during the first three months for some unknown reason, levels were relatively stable throughout the remainder of the study. Magnesium levels were even more stable; concentrations in the two-year study consistently were within _+10% of the mean value. Potassium levels were consistently less than 10 ueq/L and the observed variablity at these low levels (4.1 to 8.5 ueq/L) may reflect analytical errors. Hydrogen ion concentrations show a moderate response to lake volume. Levels increased from 25 ueq/L in October, 1980, to 44 ueq/1 in January, 1981, then decreased to 25 ueq/L by August, 1982. Dry Deposition Measured concentrations of gases (SO2, NO2, and HNO3) and aerosol constituents (NH^"^, NO3", and SO^^"), shown in Table 4-2, were used to compute fluxes of these constituents using equation 4-1. The accuracy of this approach depends on the selection of appropriate deposition velocities (v^^). Deposition velocities depend on the nature of the gas or aerosol, the deposition surface, and meteorological conditions. Published values of v^j for a given substance thus may vary over several orders of magnitude. However, when only one type of surface is considered (in this case, water) the range of deposition velocities decreases. For examp 1 e, whi 1 e pub 1 i shed va 1 ues of v^ f or SO2 including all surfaces and conditions span four orders of magnitude (see

PAGE 79

72 Table 4-2. Atmospheric concentrations and fluxes of nitrogen and sulfur species at McCloud Lake site, AugustSeptember, 1982. SO4" nhJ HNO3 NO2 ^ Mean^concentration, uq/m^ (n=20) 2.7 0.4 0.5 3.0 6.0 Deposition Low velocity (v.), High cm/ sec 0.2 0.6 0.2 0.6 0.5 1.5 0.5 1.5 0.5 1.5 Flux, Low eq/ha-yr High 40 120 14 43 12 36 150 450 210 620 ^Determined at Gainesville.

PAGE 80

73 Sehmel 1980), deposition velocities for water surfaces under most conditions range from 0.4 to 2.2 cm/sec. Garland (1978) concluded that the mean v^j for SO2 is approximately 0.8 cm/sec; several researchers have used values of 0.5 to 1.0 cm/sec (Joranger et al. 1980, Edgerton 1981). For this study, the SO2 flux was estimated using a range of v^ value of 0.5-1.5 cm/sec. Relatively few deposition velocities have been reported for NO2, NH3, and HNO3 (Soderlund 1981). However, since these compounds are highly soluble in water, their deposition to wet surfaces is probably limited by surface resistance (Fowler 1980), and their deposition velocities are probably similar to those for 502Thus, in this study the deposition velocities of these nitrogenous species was estimated as 0.5-1.5 cm/sec. The value of v^j for sulfate aerosols is less well known. Garland (1978) reported that most of the sulfate aerosol has a diameter of 0.1 to 1.0 urn and has a mean deposition velocity of 0.1 cm/sec. However, based on a comparison between measured sulfate deposition in dry buckets (Aerochem-Metri cs collector) and ambient air concentrations, Edgerton (unpublished data) has concluded that the mean deposition velocity for sulfate aerosol in Florida is 0.4 cm/ sec. Accordingly, values of 0.2-0.6 cm/sec were used to estimate the deposition of sulfate aerosol in this study. The dry deposition of other aerosol constituents (Ca^"*", Mg^"*", K^"*", Na"^, CI", and NO3") and of NH^"^ and S0^^~ was estimated using two northern Florida data bases. Edgerton (unpublished data) has recently collected dry deposition data from four sites in northern Florida using an Aerochem-Metrics wet/dry collector (Table 4-3). These data show considerable variability in deposition rates among sites.

PAGE 81

74 Table 4-3. Dry deposition at five northern Florida sites, 1982. Annual dry deposition, eq/ha-yr Site Ca2+ Na^-^ CT Gainesville 115 70 12 76 60 oT.. MugusLine /t 50 12 90 bU Cross City 649 30 12 73 66 Archibald 95 36 12 89 105 Panhandle 33 17 12 37 40 Corrected for sea salt influence. Source: Edgerton, unpublished data.

PAGE 82

75 reflecting local anthropogenic influences. However, the Panhandle site closely resembled the McCloud Lake site in its remoteness from urban areas and its sandy soils and pine-forested surroundings. Deposition rates at this collector were the lowest observed and represent reasonable minimum deposition rates for McCloud Lake. Deposition rates of Ca^"*", Mg^"^, and K"^ at this site were used as lower limits of deposition of these constituents McCloud Lake. The study of Brezonik et al. (1983b) included wet/dry deposition measurements at three inland Florida sites (Gainesville, Belle Glade, and Apopka) as wel 1 as bul k precipitation data for 25 sites, incl uding four in the vicinity of McCl oud Lake (Gai nes v i 1 1 e, Jasper, Waldo, and Hastings). Two methods were used to estimate dry deposition at the McCloud site from these data. First, the ratios of wet to total deposition at the three inland wet/ dry collectors (Table 4-4) were used to estimate dry deposition from measured wet deposition at McCloud Lake. This method should produce accurate estimates of dry deposition for substances that have uniform wet/total deposition ratios, such as chloride (wet/total ratios = 0.55-0.59), sodium (wet/ total ratios = 0.55-0.60), potassium (wet/total ratios = 0.55-0.60), nitrate (wet/total ratios = 0.68-0.74), and magnesium (wet/total ratios = 0.40-0.45). For substances such as calcium (wet/total ratios = 0.290.54), sulfate (wet/total ratios = 0.71-0.88), and ammonium (wet/total ratios = 0.73-0.88) this method is less reliable. The second method of estimating dry deposition was to subtract the measured wet depositon at McCloud Lake from the bulk deposition measured

PAGE 83

76 o o Q. in re (J iQ. o 0) • 2 c: ~— o I— -•-) re re -P !-> o •>+J Q. •rI/) U 3 CU CO i. SQ. (U > >> OJ o 2 I M C (U •2 CO E CD O £ •r^T3 re c o o cu -r> CO •Io +-> Q. re (u r— X) cu re 1/1 CO cu <_) II ;aX o UJ C/) I I CO o CVJ + + re o re CX3 CO 00 O CO o +-> re o o U3 CO I — LD C\l in en LO O U3 CM Lf) 00 Lf) o o ID en en o o o o o o t— 1 1 — 1 1 — 1 re re c SSre 3 cu ^ +J cu +-> 1— s_ "O 1 — Z3 re 3 O O > -rCD co E re Stl CU cu CD C 00 Q. re re •r— v ^ O re Q. cu ct CO

PAGE 84

77 at the four regional collectors (Jasper, Gainesville, Waldo, and Hastings) during the 1978-79 study (Table 4-5). Dry deposition rates estimated from these two data bases are shown in Table 4-6. For each method, some of the measured deposition rates were excluded because they reflect extreme conditions that are not representative of conditions at McCloud Lake. Calcium deposition, in particular, appears to be strongly influenced by local anthropogenic influences, and data from three sites were excluded. Dry deposition of calcium at Cross City (Table 4-3) reflected the proximity of a school playground, whereas bulk deposition of calcium at Gainesville was influenced a parking lot and city streets and bulk deposition of calcium at Jasper reflected phosphate mining activity. Dry deposition of sodium and chloride deposition at St. Augustine and bulk deposition of these ions at Hastings and Jasper reflected a strong sea-salt influence that was not representative of McCloud Lake, and these data were not used. Finally, ammonium deposition at Jasper seemed excessively high (283 eq/ha-yr). Although the reason for the high ammonium deposition is not known, this value was rejected as an outlier. Dry deposition of alkalinity in the study of Brezonik et al. (1983b) was inferred from the difference between the sum of measured cations and the sum of measured anions (Table 4-5). From Table 4-5 it appears that the inferred alkalinity was approximately 0.6 times the calcium deposition. To estimate alkalinity deposition in this study, this ratio was used in conjunction with estimated calcium deposition. Dry deposition of NO2, SO^ HNO3, and alkalinity were used to compute an "effective dry deposition" for protons. This computation

PAGE 85

78 1—1 I 00 4J ro o so sZ3 o 44-> to I/) O Q. LO I "O I to cr o a. -a ca 1 CO CO 00 0 OvJ 0 CM CM I— ( CM CM ^ O to + CM CD + CM to O ro ro ro 00 00 CM .—I LO CO 00 0 0 1^ 0 LT) I— 1 LT) I— 1 0 CM I— 1 1 — ( CM CM cn LO .— 1 CD WD I— t CM ro CO ro ro O 00 00 CO LO cn o cn CM cn LO cn LO LO CM 00 cn CO LO 1 — t r-H 00 0) •1— > LO 01 0 c C3l +-> 0 CO CO (O to to to CD 3: ro 00 cn to o N Ol iO iO oo

PAGE 86

79 assumes that the gases dissolve in water and undergo the fol lowing reactions: SO2 + + 1/2 O2 > + ZH"^ (4-6) 2NO2 + 1/2 O2 + H2O > 2N03" + ZH^ (4-7) HNO3 > H"^ + N03~ (4-8) The effective dry deposition of H"*" (D^+) was thus Dh+ = Dso2 + Dno2 + Dhno3 Dal k (4-9) where D5Q2....Da ] |^ = deposition rates of subscripted substances, eq/ha-yr. The last two columns in Table 4-6 show dry deposition rates used in the McCloud Lake model. These values bracket the measured or estimated dry deposition rates for north Florida. Sodium and chloride were further constrained to occur in a ratio of 0.88:1 (the ratio of these ions in seawater--see Stumm and Morgan 1981), although this constraint required very 1 ittle modification of observed deposition rates (Table 4-6). As shown in Table 4-6, the estimated deposition of Na"*" and CI" varied by a factor of three, while estimates of other aerosol constituents varied by factors of three to ten. Estimates of gaseous deposition varied by a factor of three, reflecting the use of V(j values from 0.5-1.5 cm/sec. Deposition rates for NH4'^ and SO^^" estimated from aerosol concentrations (Table 4-2) fel 1 within the range of estimates made from bulk and dry collector data. Dry deposition constituents were assigned to three groups: 1) sea-salt components (Na"^ and CI"), 2) anthropogenic aerosols ($04^", NO3", Ca^"^, Mg^"^, K"^, and alkalinity), and 3) gases (SO2, HNO3, and NO2). To compile the McCloud Lake mass balances, constituents within each group were assigned the upper or lower estimates of dry deposition

PAGE 87

80 +J XI TJ •1OJ O ^ QO ^ 3^ X5 "O GJ ^ (/) o o fO l_) •1-o 3 •4-> J3 C SI/) o o — i 1 1 Q.) -4—^ Ll_ %3 CO jd c/) •>> t/) O 4-* QJ CIS Oi-4-^ ID O *'~ "^"^ S "O E lO CT 1 o f-\ \J to c: ^ 3 OJ E (J O O ^ (/) ij ^ C J >^ +-> 4-* *o O C ~0 CO QJ +-> O •4— J > CT3 -M +-> E CD >— • ^ tA o LU -M O /1 1 +-* \^ CJ fr\ •VJ .J Q O (/) •r— 0) 1 H — L) '' *' CO _C t/) O -M 4—' QJ O "O w >^ iO 1^ iO 1 — -f— Q Li_ o o. c ^ 'rCJ 1 — --> iE •! — 3 O CO CO 1 O o (T3 +-> Q. 0> GJ OJ ^ 2 "O C 0) 1 3 !-> cu -M CO lo c 03 o <_> o o LO LO o o o o o o CM LO o CM I— 1 1— 1 CM t— 1 LO o o LT) o O o o CO CM 1 — 1 00 OO ,—1 CM LO CM oo CO LO o <— ( t— 1 1 1 00 CM I— 1 t— 1 Ln 00 CM ?^ 1 — 1 1 CM 1 1 CO CM ro l£) PO I— t CO I— 1 o U •o o CO CTi ^ O LO CM LO .—1 00 O 00 r— ( CM I— 1 o 1 1 t— 1 1 00 00 00 I — 1 oo o CM 1 — ( o 00 oo LO t— 1 LO CO I— 1 t— 1 .—1 r— 1 + + I CM CNJ + I ,-M ^1 OOf (O CT)+ (Oi — <—) f—) -rosr^zooozz o o u O) c -l-J a CO OJ 3 o a> 3 3 <: O •-> •r— +j o st>0 o OJ X3 c LO ro CO r> S!a OJ TJ 3 Q. 0) Q. to T3 C na 3 3 "D (J \ X C 00 OJ o > cn +- to c S0) OJ (U c: +-> Q. C7> •1— to 00 T3 ns n3 na LU t3 n: a o D

PAGE 88

81 simultaneously. Thus, eight scenarios were constructed representing combinations of high and low sea salts, high and low anthropogenic aerosols, and high and low gases. All combinations of aerosol deposition (i.e., high sea-salt + low anthropogenic aerosols, low sea-salt + low anthropogenic aerosols, high sea-salt + high anthropogenic aerosols, and high sea-salt + low anthropogenic aerosols) had ion balances (equation 2-1) within 5%. Since deposition of gases contributed to the effective H"*" deposition and the deposition of alkalinity reduced the effective H"*" deposition (equation 4-9), the higher estimate of H"*" deposition was computed using the high estimate of gas deposition and the low estimate of aerosol deposition. Conversely, the lower estimate of H"*" deposition was obtained using the lower estimate of gas deposition and the higher estimate of alkalinity deposition (Table 4-6). Wet Precipitation Wet-only precipitation at McCloud Lake, like precipitation throughout northern Florida (Brezonik et al. 1983b), was acidic (volume-weighted mean pH: 4.5). The source of H"*" in wet precipitation can be inferred from the anion composition. The occurrence of chloride and sodium in a ratio of 0.85:1 in wet precipitation clearly suggests a sea-salt origin for these ions rather than an HCl source for chloride. The NO3' concentration in wet precipitation, 9.8 ueq/L, could account for no more than 34% of the 29 ueq H"*'/L. Sulfuric acid was therefore the primary source of protons in wet precipitation: the SO4 concentration of 29 ueq/L was adequate to account for 100% of the protons in wet precipitation.

PAGE 89

82 (U T3 O r— o u I/) 0) o c: o -t-> ta +-> u O) s(U to o o +-> >> (O I M >— 1 t-> 3 O) cC 1 -a I— 1 c 00 o o en O) •>Si-H -M +-> >, rO -r1 E CO (0 •-O ^ -M CL-\ E to O) CT o LiJ T3 O) _J +-> Q. (V 00 ^ 4J 4-> n3 >•-1— -l-> C rO O 4J +-> C 01 0) 3 O c o o cr 3. 3 +-> to o o .— 1 I— 1 o o x—< CO IT) £) I— 1 o 1 O CO o in ro I— 1 o O 1 CVJ 1 CO LO (X) I I CO o CNJ CM •-I o ir> uo o IT) I CM .— t CM CO 00 I I cn o ID (X) o CM I ID I o 00 LD 00 I o CM Lf) LD CM un 00 o o CO I o CM CM o I CO o CM I o o o o o cn o 1^ O CM LO ID O O O O CM O ID O ID O O LO ro CM I— I CM LD ro o o o CO ID o o CM LD o o o LD o CO o CM m LD ro CO CM ro + + CM CM +1 f ns cr)+ 10 I — in o s z o 00 I 00 o I o LO o ID CM O oo CO ro I I ro 00 CTi CO I I o I— ( ID o 00 1 o ro o o o o LO CM 'dr-* LD o LO o ro CM CM t— 1 1—1 1 o 1 LO 1 o LO CO CM <— 1 o o ro <-< CM CM o o O CM o o t-H CM O O CM <^ ^ r-l O LO 00 CM 00 O LO 10 -(-> o >> p (0 CO +-> to (-> I/) (0 ro o to cn c SCD +J CD 3 <+1 oo (O CM CM ^ 1 oo CM O 1 oo + =t3 O O o o 2: o 00 00 OO 5 cu u CU to OJ o to

PAGE 90

83 Wet precipitation accounted for 60-82% of the chloride, 59-81% of the sodium, 62-93% of the nitrate, 75-88% of the ammonium, and 37-93% of the sulfate deposition to McCloud Lake. Gaseous deposition was an additional major source of sulfate and nitrate. Sulfur dioxide (SO2) deposition, estimated to be 150-450 eq/ha-yr (expression on equivalent basis assumes conversion to SO^^' at lake surface), was 25-40% of the total sulfur deposition. The adsorption of NO2 at the lake surface contributed 210-620 eq N03"/ha-yr, or 57-70% of the total NOx (NO3 + NO2 + HNO3). Interestingly, although $04^" was the dominant acidic anion in wet-only precipitation, the dry deposition loading of NO^ was equal or greater than the dry deposition of $04^" + SO2 (Table 4-7). Dry deposition of H"*" (equation 4-10) was a major source of protons to McCloud Lake. Although the range of estimated values was large (250-1090 eq/ha-yr), these estimates show that deposition of gaseous species (primarily SO2 and NO2) may be an important source of protons to McCloud Lake, and may account for 54-71% of the total proton deposition. Groundwater Chemistry Table 4-8 shows water quality data for the 12 wells sampled during May-October, 1982 (arithmetic means) and for the 1982 inseepage (flow-weighted means). These data show that precipitation passing through the soils surrounding McCloud Lake was altered by several mechanisms, even though these soils are extremely sandy and permeable. First, evaporation concentrated the precipitation, resulting in increased concentrations of conservative ions such as chloride and sodium. Based on observed chloride values, well water

PAGE 91

84 Table 4-8. Chemistry of groundwater at McCloud Lake. Seepage Meters Well Water Constituent Cone. uea/L c./ci E.F. Cone. iieo/L C^CI E.F. 1.9 0 05 0 007-0 03 \J m \J\J i \J \J\J 1.9 0.02 0.003-0.01 Ca^-^ 92 L. • >J\J 1 ^ -7 Q 204 2.27 1.4 -7.6 51 1 30 2 6 -6 5 90 1.00 2.0 -5.0 Na" 34 0 87 1 0 91 1.01 1.0 20 0.50 1.3 -5.0 37 0.41 1.0 -4.0 36 0.92 0.2 -0.3 49 0.54 0.1 -0.2 n ~ L 1 on 1.00 90 1.00 nhJ 3 0.09 0.1 -0.2 NO4 0.10 .02 .07 4.9 .05 0.01 -0.03 HCO3 ^ 123 3.14 280 3.11 Inferred from ion balance.

PAGE 92

85 was concentrated by a factor of 4 while in-seepage was concentrated by a factor of A/3 (Table 4-8). The fact that in-seepage was less concentrated than well water suggests that seepage flows to the lake originated as precipitation that f el 1 very near the 1 ake shore and passed rapidly through the soil and into the lake. Groundwater further from the lake apparently remains in the soil longer and becomes more concentrated by further evaporation. Both seepage and wel 1 water had pH values between 5.5 and 6.0. Neutralization of precipitation acidity probably occurs as the result cation exchange that produces both alkalinity and dissolved Ca^"*" and 2+ Mg This enrichment can be seen by comparing ion/Cl~ ratios in the seepage and wel 1 water with the ion/Cl" ratios in the total (wet + dry) precipitation using an enrichment factor (EF): EF = (Cj/CI-K (4-10) where (C^-/Cl")j = ratio of ion i to CI" in compartment j; (C^/Cl") = precipitation ion/Cl" ratio An enrichment factor of < 1 indicates a depletion of the ion resulting from chemical precipitation or assimilation, while an enrichment factor > 1 indicates enrichment from mineral dissolution or mineralization. An EF of 1.0 indicates that the ion is conservative. As seen in Table 4-8, calcium was enriched by a factor of 1.58.0 in both the well water and in-seepage, while magnesium was enriched by a factor of 2.6-6.5 in the seepage and 2.0-5.0 in the wells. Alkalinity, calculated by inference from the ion balance, roughly balances Ca Mg in both seepage and well water. Potassium may also have been enriched in the seepage and well waters, although when the lower estimates of K"*" dry deposition were used to compute the EF,

PAGE 93

86 potassium appeared to be conservative in both compartments (Table 4-8). Sulfate was depleted in both seepage water (EF = 0.2 0.3) and in well water (EF = 0.1-0.2). Sulfate adsorption or biological assimilation was probably responsible for the sulfate depletion in the well water although sulfate reduction could deplete sulfate if the well water became anaerobic. Sulfate reduction undoubtedly caused the sulfate depletion in the seepage water, since this process occurred in the laboratory seepage experiment (Chapter 3). Both nitrate and ammonium were depleted in the subsurface waters. Much of this depletion undoubtedly occurs as the result of assimilation of these nutrients by terrestial vegetation and soil microorganisms. These results are consistent with reports that inorganic nitrogen is retained in forest soils (Impact Assessment Group 1983). Denitri f ication in littoral in-seepage also may account for some of the nitrate depletion; laboratory seepage column experiments (Chapter 3) also showed a dramatic reduction of nitrate in the eluate. Fate of Major Ions in McCloud Lake Data on wet and dry precipitation, seepage flows, and changes in storage were used to evaluate the fate of ions entering McCloud Lake. Several approaches were used to evaluate in-lake sinks and sources of ions. Annual mass balances were compiled for major ions and nitrogen species using the equation dM/dt = P + G,-, G,^^ + S (4-11) where dM/dt = change in storage, eq, P = precipitation (wet + dry) inputs, eq/yr.

PAGE 94

87 '^in'^out flroundwater inputs and outputs, respectively, Tir/yr, and S = sink (negative value) or source (positive value), eq/yr. This approach has the advantage of not requiring steady state conditions. However, evaluation of the sink/ source term is sensitive to errors that result determining smal 1 differences between large numbers. This problem is discussed in greater detail with respect to the chloride mass balance below. Sinks and sources of major ions also were evaluated by comparing the ion/chloride ratios in the inflow (wet + dry precipitation + inseepage) to ion/chlorides ratios in the lake. Since chloride is conservative in aquatic systems, the ion/chloride ratio of a substance entering the lake can change only occur as the result of sinks or sources of that substance within the lake. The magnitude of the sink for substance i (S^) can be evaluated as: Si = Lj/Lpi [Cj], /[Cl]| X 100 (4-12) M^kl where Li r] = annual input of substance i and chloride, eq/yr, [CiJL, [C1]l = concentration of substance i and of chloride in lake, meq/L, and Si = sink/source term as per cent of input. Although this approach assumes that the lake is at steady state, it does not depend on measured changes in storage and is less sensitive to errors in input load. Input loads and mean storage were also used to compute the residence time (T) for various substances entering McCloud Lake, where: T = mean storage, eg (4-13) annual input, eq/yr The residence time of conservative substances such as chloride and sodium is the same as the water residence time based on outflow.

PAGE 95

88 Shorter residence times indicate that a substance has an internal sink, while longer residence times indicate an internal source. Chloride Most of the chloride entering McCloud Lake came from wet + dry precipitation (Table 4-9); groundwater seepage accounted for < 25% of the total chloride input. Although the mass balance equation for the 1981-82 model year indicated a source of 4960-5370 eq/yr, chloride was undoubtedly conservative and the calculated source reflects a problem with the mass balance calculation. Since the mass balance equation (equation 4-11) computes the sink/source term as the difference between inputs, outputs, and the change in storage, it is subject to errors that accrue from estimating smal 1 differences between large numbers. The change in storage term is the most likely source of error. During the model year, storage of chloride increased by 4470 eq (20% of the mean storage), compared with an annual precipitation input of only 1120-1520 eq. However the change in storage term is subject to errors in both chloride measurement and estimation of water volume. The error term can be calculated from the equation: EdM/dt = Ci*Evi Ec2*C2 ^2*^2 (^-l^) ^C1*^V1 1^C2*^V2 where ^HM/dt ^ error in change in storage for a given constituent, VI, V2 = initial and final volume, C1,C2 = initial and final concentration, ^V1^V2 ^ error in initial and final estimates of volume, and ^C1^C2 error in initial and final estimates of concentration. If each measured term had an error of only 5% of the true value, and if initial and final storage were approximately equal, the potential error in dM/dt could be as high as 25% of the mean storage. Although

PAGE 96

89 0) OJ o •rOJ 00 B q: -p o lO •!(/I I/) o o E •-r— O +-> &5 Mio O >, ^ Ol E CVJ 1 CO in I — 1 1 — ( Lf) I— 1 I— 1 CM 1— ( o o 1 LO 1 T— 1 1 CvJ 1 1 1 C\J r— 1 00 .— 1 CO 00 .— 1 o o o OJ o 3 o o o cn c: z. cu to o o -t-> I CO CD CL O) 00 00 oo cr <] 2: to >> +-> 3 cr o a; cr s>> Icr O OJ cu S+-> 3 +-> 1/1 c o o o OJ OJ CO o 00 OJ I f o CO OJ ro OJ y3 lO CTl 1 LD CTl CTl CTl CTl cn CO 1 CO 1 (T> 1 1 CO LO 00 00 CTl o o o Uf) ID o 1—1 .— 1 UO o CO 00 CO — 1 CM 1—1 r— 1 1 i 1 1 1 OJ CO Ln CO CO CO I— 1 CM — 1 o o o o o o o CvJ CM ID CO Ln CM CO CTl OJ CO .— ( + 1 + 1 1 + 1 o o o o o o <^ CM CTl ID CTl CO to 00 OJ C\J o CO o CM o o o o o o O o TM 1 — 1 00 CO CO CO CO f— t o 00 o CO 00 I r—t CM I O OJ CO o O o O o o o o o ID o o 1 — CO CO O 1 — 1 n CM CO 00 CM CM CM o 00 o CO o o o o o O o o o CO — 1 1 Ln 1 I— 1 :! o o o O 1 o 1 o 1 o 1 o 1 o o in o o CM 1—1 >— 1 o CTl Ln 1—1 CO 1 — t 1 — I CO 1 — 1 I h CM CO o z oo + -M CO (-3 + -NJ C71 I CO o I c o +-> CO 3 cr (V 0) o E (C CO J3 to 1/5 CO CO 3 C C CO O sT3 CO to II CM 1—1 1 c OJ o (J S+-> 3 CD o 3 t/) cr
PAGE 97

90 this example represents a worse-case scenario (all errors in the same direction), it illustrates the type of problem encountered in using a mass balance model for a substance that has a large storage relative to inputs during the modeling period. For chloride, with a residence time of 12.0-15.4 years in McCloud Lake, a much longer modeling period would be needed to accurately estimate the sink/source term. Sodium Precipitation was also the major source of sodium to McCloud Lake, accounting for 77% of the total input. The mass balance equation indicates a sink of 1360 to 2220 eq/yr, but this is < 10% of the mean storage and is probably insignificant for reasons discussed above. Calculations based on Na''"/Cl' ratios indicated a small sink (11% of annual inputs), suggesting that sodium is nearly conservative. The residence time for sodium in McCloud Lake, 13.5 to 17.2 years, is very near that of chloride. Sul fate Wet + dry precipitation accounted for over 90% of the total inorganic sulfur ($04^" + SO2) to McCloud Lake (Table 4-9). Both ion ratios and the annual mass balance indicate that the lake was a sink for sulfate. Although the mass balance yields a sink of 230 to 2820 eq/yr, the magnitude of this sink is small relative to the lake storage of 24,360 eq and therefore subject to the type of errors discussed above. The presence of a sink is confirmed, however, by ion ratio calculations which indicate that 39-73% of the inorganic sulfur entering the lake is lost internally and by the calculated residence time of only 4.1 to 7.4 years.

PAGE 98

The most likely mechanism for an internal sulfate sink is biological sulfate reduction at the sediment-water interface. This hypothesis is strongly supported by the sulfate pore water profiles (Figure 3-14), which show a distinct gradient of sulfate in the profundal sediments, and by the seepage column experiments, which showed a loss of sulfate in the groundwater eluate (see Chapter 3). These data alone do not conclusively prove that biological sulfate reduction is responsible for the observed sulfate sink since end products (FeS or H2S) have not been identified. On the other hand, sulfate adsorption did not occur to an appreciable extent in McCloud Lake sediments (Chapter 3), and it is unlikely that other mechanisms such as biological assimilation or chemical precipitation of sul f ate-beari ng minerals (Nordstrom 1982) could account for the observed sink. Furthermore, the organic-rich sediment (over 90% volatile solids in the profundal sediments) and warm temperatures (annual mean = 20 C) are conducive to sulfate reduction in McCloud Lake. Although there is some evidence that sulfate reduction may be inhibited below pH 5 (Zinder and Brock 1978), the pH of McCloud Lake sediments is generally above 5, perhaps as a result of the pH-buffering effect of the sulfate reducers. Sulfate reduction has been observed in other acidic environments including peat bogs (Rippon et al. 1980) and experimentally acidified Lake 223 (Kelley et al 1982). The evidence for sulfate reduction in McCloud Lake suggests that sulfate reduction may be an important sink for sulfate in acidic seepage lakes.

PAGE 99

92 Cal cium, magnesium, and potassium Mass balance calculations and comparisons of ion/Cl" ratios of inputs and lakewater indicate that McCloud Lake is a sink for all three cations. Ion ratio calculations, for example, show that 53-77% of the calcium, 7-45% of the magnesium, and 78-92% of the potassium entering the lake is lost through an internal sink. Potential sinks, such as biogenic accumulation or precipitation of inorganic solids, are unlikely to be important in McCloud Lake. Potassium generally is assumed to be conservative. Potassium concentrations are rarely affected by phytoplankton blooms (Wetzel 1975), K"*" was essentially nonreactive in the neutralization experiments described in Chapter 3, and there are few potassium-containing minerals that would precipitate in a softwater lake. Chemical precipitation of calcium and magnesium is also unlikely in this lake. It therefore appears that inaccuracies in the estimated inputs of these cations are responsible for the apparent sinks. Of the input components, wet deposition is undoubtedly the most accurate. Dry deposition for these elements was estimated only to within a range of approximately one order of magnitude, although the values used undoubtedly bracket the actual dry deposition rates. The term most likely subject to errors is in-seepage. As shown in Table 4-9, in-seepage was a major source of calcium (39-64% of the total loading), magnesium (44-58% of the total loading), and potassium (29-62% of the total). Furthermore, a large (50%) error in calculated in-seepage would be nearly undetectable in the water balance, since in-seepage only accounted for 10% of the total inflow to the lake. For example, a 50% reduction in in-seepage would change the computed endof-year water storage by only 3% but would decrease the total calcium

PAGE 100

93 loading by as much as 32% and increase the residence time to 4.4-8.2 years. It is apparent that more intensive sampling of seepage would be required to obtain accurate estimates of fluxes of these cations in McClould Lake. It should be noted that seepage was a minor source of sulfate, protons, chloride, sodium, nitrate, and ammonium, and relatively large errors in seepage loadings would have minimal effect of budgets of these substances. Nitrogen speci es Wet + dry precipitation accounted for nearly al 1 (over 95%) of the inorganic nitrogen input to McCloud Lake. The total NO3" load (including NO2 and HNO3 adsorption at the lake surface) was 1940 to 4610 eq/yr, about five times that of NH^"^ (450-910 eq/yr). Both species of nitrogen were removed from the water column by in-lake mechanisms. The mass balance indicated an annual sink of 1350-1820 eq for NH^"^ (several times the mean storage of 603 eq) and NH^'^'/Cl" ratios show that 86-96% of the NH^"^ entering McCloud Lake was removed. Nitrate had an annual sink of 2140-4810 eq during the model year, compared with a mean storage of 670 eq. Ion ratio calculations indicated that 97-99% of the input nitrate disappeared via an internal sink. The residence times for these substances is appoximately one year. Biological assimilation by phytopl ankton, bacteria, and macrophytes and the accumulation of organic nitrogen in the sediment are probably the most important mechanisms for removing inorganic nitrogen from the water column of McCloud Lake. Additional loss of NH^"^ may occur as the result of adsorption to sediment surfaces and additional

PAGE 101

94 loss of nitrate may occur via denitrification. A detailed discussion of the nitrogen cycle in McCloud Lake is presented in Chapter 5. Proton balance Nearly al 1 of the protons entering McCloud Lake come from wet precipitation or are the the result of gaseous deposition of SO2, NO2, and HNO3 directly to the lake surface. Seepage inflows contributed 480 eq/yr of alkalinity (which was considered in calculations a negative proton contribution) that neutralized 6-13% of the protons entering the lake via precipitation. The lake was a sink for protons. The mass balance and the ion ratios both indicate that 86-94% of the input protons were lost by an internal sink. Potential mechanisms to remove H"*" in McCloud Lake include sulfate reduction, nitrate assimilation, and cation exchange. Sulfate reduction, which accounted for a 1 oss of 39-73% of the input sul fate, or 1290 to 4390 eq/yr (based on ion ratio calculations), consumed an equivalent amount of protons and was therefore a major sink for protons in McCloud Lake. Nitrogen transformations were also important proton sinks. Assimilation of NH4''" by organisms results in the release of protons (1 eq H"^ released/eq NH^"^ assimilated) to the environment, while assimilation or denitrification of NO3" consumes protons (1 eq H"*" consumed/eq NO3" consumed). Since nitrate input was approximately five times that of ammonium input, the net result was a consumption of 1880-4550 eq NO3" minus 400 to 870 eq NH^"^, or as much as 1010-4150 eq H^/yr. Cation exchange of protons entering via precipitation with cations absorbed on sediments may be a significant mechanism— as suggested by results of laboratory experiments in Chapter 3— but the mass balance

PAGE 102

95 developed in this study was not sensitive enough to evaluate the magnitude of this process.

PAGE 103

CHAPTER 5 DECOMPOSITION AND NITROGEN CYCLING IN McCLOUD LAKE Introduction Grahn et al. (1974) proposed that acidification of lakes induces a "sel f-ol igotrophication" process in which decomposition and nutrient regeneration are retarded, resulting in decreased productivity in the water column. Although synoptic studies often show a decrease in nutrient levels with decreasing pH (Aimer et al. 1974; Crisman et al. 1980), it not clear whether decreased pH causes a decrease in nutrient regeneration or whether the relationship between low nutrient levels and low pH occurs because nutrient-poor lakes also are most susceptible to acidification. Few studies have attempted to quantify the relationship between pH and decomposition rates, and even less work has been directed towards the effects of pH on the mineralization and cycling of nutrients. The objective of this chapter is to examine the effects of low pH on decomposition and nitrogen mineralization in McCloud Lake. Littoral sediment-water mesocosms (limnobags) were used to observe effects of pH alterations on nutrient levels and to conduct studies of sediment metabolism. Laboratory microcosms, with and without sediment, were used to determine effects of pH on nutrient regeneration under more closely controlled conditions. Finally, measurements of all nitrogenous inputs and in-lake nitrogenous species were used to evaluate historical trends, to determine sources and sinks of nitrogen 96

PAGE 104

97 in the lake, and to evaluate the role of atmospheric nitrogen inputs in lake productivity. Background Effects of Low pH on Decomposition Relatively few studies have been conducted to evaluate the role of pH on the decomposition process in lakes, and the results of published studies are mixed. Studies of the effects of pH on metabolic rates in culture generally indicate a decrease in respiration with decreasing pH (Traeen 1980; McKeown et al. 1968). Traaen (1980) however, found that the lag phase observed in degradation of simple substrates (glucose and glutamic acid) disappeared at pH values down to 4.0 when the microbial seed used in his studies was accl imated at the treatment pH, even though respiration rates in the exponential growth phase were still somewhat lower at pH 4 than at pH 7. Lower pH seems to have a greater effect on the decomposition of more complex substrates, such as tree leaves containing lignin and cellulose, than on simple substrates (Traaen 1980; Leivestad et al. 1976). This occurs for two reasons. First, a relatively small number of organisms can use cellulose, lignin, and other complex substrates, and many of these organisms are pH-sensiti ve. Second, the microbial degradation of leaves and other structurally complex material in aquatic systems is facilitated by macroinvertibrates and insects that scrape and shred the coarse material, making it amenable to microbial attack. There is good evidence that adverse effects of low pH on these organisms inhibits degradation of leaves and woody material at low pH (Traaen 1980; Hal 1 et al. 1980).

PAGE 105

98 Decomposition at the sediment surface of lakes may be relatively unaffected by increased acidity of the overlying water. Gahnstrom et al. (1980) found little correlation between lake pH and sediment oxygen uptake by 10 cores from seven Swedish lakes (pH 4.5-6.8), although lime treatment of L. Hogsjon (from pH 4.7 to pH 6.6) increased sediment respiration by 30%. Andersson et al. (1978) found that differences in oxygen consumption or CO2 production between acidified (pH 3.1) and control (pH 4.5) cores disappeared after 75 days, although glucose turnover times were still longer in the acidified cores. Both Andersson et al. (1978) and Gahnstrom et al. (1980) hypothesized that the amelioration of pH effects on benthic respiration results from neutralization effects of the sediments. For example, Andersson et al. (1978) found that while the difference in pH of the overlying water in the acidified and control cores was 1.5 units, the difference in sediment pH (2 cm depth) was only 0.5 units after the 75 day incubation period. Decreased pH also may affect in-lake decomposition through indirect mechanisms. Gahnstrom et al. (1980) postulated that changes in the rate of sedimentation of organic material may alter sediment respiration rates, and Grahn et al. (1974) suggested that Sphagnum mats which often grow in acidic lakes may inhibit diffusion across the sediment-water interface. It is also plausible that the increased compaction that seems to accompany acidification of sediments (observed in the sediment titration experiments) may result in diminished diffusion. Whole lake manipulations indicate that decomposition and nutrient cycling are not greatly affected by moderate pH alterations. Schindler et al. (1980) found no changes in levels of dissolved

PAGE 106

99 inorganic carbon or in nutrient levels in Lake 223 following acidification to pH 5.1 and concluded that evidence for reduced decomposition was lacking. Schindler et al. (1980) noted, however, that Lake 223 lacked the coarse particulate shredders that may be most affected by increased acidity. Dillon et al. (1979) found no change in the organic content of the sediments or in aqueous phosphorus concentrations of three Ontario lakes that were limed to raise their pH levels from 4-5 to 7-8. Hultberg and Andersson (1982) reported no change in phosphorus concentrations in three Swedish lakes that had been limed over a period of seven years. Effects of Low pH on Nitrogen Cycling The effects of acidification on rates of nutrient cycling in aquatic systems have received very little attention, and the little work that has been done has dealt almost exclusively with phosphorus. Effects of acidification on the nitrogen cycle have been studied in acidic soils and peat bogs, and these studies provide a basis for generating hypotheses on the effects of acidification on nitrogen cycling in aquatic systems. Ammonifi cation Other than the brief study by Leivestad et al. (1976), which reported that ammonifi cation of peptone in culture was inhibited below pH 5, there appear to be no studies of pH effects on ammoni f ication in aquatic systems. Alexander (1980) postulated that ammoni fi cati on would not show marked pH sensitivity because a wide variety of bacteria and fungi can carry out the process. The study of ammonification is complicated by the fact that NH4''" can be adsorbed by soils and

PAGE 107

100 sediments and can be converted to nitrate or be reassimi 1 ated. For example, Francis (1981) observed greater NH^"*" accumulation in a pH 4.6 soil than in the same soil that had been acidified to pH 2.6 or neutral ized to pH 7.0. Nitrate production, however, was greater in both the acidified and the neutralized soil. Nitri f ication The growth of autotrophic nitrifiers ( Nitrosomonas and Nitrobacter ) is inhibited in culture below pH 5 (NAS 1978). Although Alexander (1980) also concluded that nitrification in soils is general ly inhibited below pH 5, Lee and Stewart (1978) reviewed numerous reports of nitrification in acidic soils and concluded that soil acidity per se does not result in cessation of nitrification. The variability of pH effects on nitrification is illustrated by the study of Klein et al. (1982), who recently examined nitrification rates in acidic (pH 3.6-4.1) soils from the Adirondacks by adding simulated rain (pH 3.6, 4.1, and 5.6) to lab columns. While pH 3.6 rainfall inhibited nitrification in two soils, it stimulated nitrification in the third. Although rates of nitrification varied among treatments, nitrification occurred in all soils and at all rainfall pH levels. In a separate experiment, NH4"^ additions to flasks containing these soils failed to stimulate NO3" production, leading these authors to postulate that heterotrophic nitrification, in which organic N is converted directly to N03~, may have been responsible for the observed nitrification. Denitrif ication Several studies have reported that denitrif ication rates decrease below pH 7-8 (Noninik 1956; Focht 1974) and that the end product shifts

PAGE 108

101 from N2 to (Focht 1974; Blackmer and Bremner 1978). However, as with nitrification, denitrif ication has also been reported in acidic soils and bogs (Ekpete and Cornfield 1965; Martin and Holding 1978). At low pH chemodenitrification may occur by autodecomposition of NO2 and by reactions of NO with aromatic compounds (Broadbent and Clark 1965; Bremner and Nelson 1968). The significance of these reactions in natural soils and lake sediments is unknown. Nitrogen fixation The most important agents of nitrogen fixation in aquatic systems are blue-green algae. Numerous studies have shown that populations of blue-green algae are reduced at pH levels below 5 (Brock 1973; Leivestad et al. 1976; Aimer et al. 1978; Yan and Stokes 1978), and it is reasonable to conclude that nitrogen fixation is also diminished below pH 5. Heterotrophic (bacterial) nitrogen fixation may occur in lakes under anaerobic conditions (MacGregor and Keeney 1973, Brezonik and Harper 1969), and the effects of pH on these organisms has not been studied. However, nitrogen fixation rates appear to be inversely related to inorganic N:P ratios (Vanderhoef et al. 1974) and the softwater lakes that are susceptible to acidification tend to have high N:P ratios. Consequently, nitrogen fixation is not likely to be an important process in these lakes, regardless of pH. Methods Littoral Mesocosms Three in-lake mesocosms (cylindrical enclosures) were placed in the littoral zone of McCloud Lake to evaluate the effects of pH on biological and chemical processes. The 4m di ameter enc 1 osures

PAGE 109

102 designed after Landers (1979), were placed in the lake in March, 1981 at a depth of one meter. Fol 1 owing a month of stabi 1 ization, acid (0.72 N H2SO4) was added to lower the pH of one enclosure to 3.6 while base was added to a second enclosure to raise its pH to 5.6. A third enclosure received no acid or base and served as a control. Weekly additions of acid or base were made in an attempt to reach target pH levels within one month. While the low pH was attained within a month, frequent additions of acid were required to maintain a pH of 3.6. The pH of the neutralized enclosure barely changed within the first month, and repeated additions of base were required to maintain a pH of approximately 5. The process of sediment neutralization believed to be responsible for the observed resistance to pH alteration has been discussed in Chapter 3. Throughout the first 16 weeks of the experiment, samples were collected weekly from each mesocosm and from the adjacent littoral water for analyses of pH, major ions, and major nutrients (N and P species). Dissolved oxygen (D.O.) and temperature were measured in the field with a YSI D.O. meter. Chemical analytical methods are described in Chapter 2; methods for sampling and analyses of bacteria, zooplankton, phytop 1 ankton and chlorophyll are described by Crisman et al. (1983). During late July and August, 1981, three experiments were conducted to measure diurnal fluxes of O2, NH^"*", and soluble reactive phosphate (SRP) across the sediment-water interface of each mesocosm. Clear plexiglass chambers (30 x 20 x 20 cm) were placed on the bottom of each mesocosm and sealed by inserting the bottom edge '-'5 cm into the sediment. A pumping system (Figure 5-1) was used to withdraw samples for measurements of D.O. (by Winkler titration in experiment

PAGE 110

103

PAGE 111

104 1) or to circulate water across an in-line D.O. electrode (experiments 2 and 3). In al 1 three experiments samples were withdrawn for nutrient analyses during each 2-5 hour interval. In the second and third experiments, D.O. and nutrient analyses were also made in the main water column of each mesocosm in order to determine whol e-mesocosm fluxes. Gross productivity, respiration, and net productivity were calculated from diurnal O2 flux curves using the procedure outlined by Hall and Moll (1978). Ammonium and phosphate fluxes were obtained by integrating the area under the time-flux curves. Microcosm Experiments Water-only and sediment-water microcosm experiments were conducted to evaluate decomposition and nitrogen cycling under more controlled conditions. Water-only microcosms were constructed by fi 1 1 ing BOD bottles with McCloud 1 akewater and dried, homogenized substrate (pi ankton tow materi al or benthi c macrophyte ti ssue). After pH adjustment (with H2SO4 or NaOH) and the addition of a microbial seed, five bottles at each pH level (2.6, 3.6, 4.6, 5.6, and 6.6) were sealed and incubated for three weeks. Dissolved oxygen measurements were made each week using a stirring D.O. electrode and at the end of the incubation period using a Winkler titration. Samples for nutrient analyses (NH4''" and SRP) were also collected at the end of the experiment. The three mL of water removed from each bottle for nutrient analyses was replaced with DDW prior to oxygen analysis. The microbial seed in the first experiment was prepared by adding 10 mL sediment + 10 mL 1 akewater to large test tubes and adjusting the pH of the

PAGE 112

105 water to treatment pH levels. After a one-week acclimation period, 1 mL of overlying water was withdrawn and added to each BOD bottle. The same procedure was used in the second experiment, except that 1 mL activated sludge from a neutral-pH environment was added to the test tubes containing sediment and water prior to pH adjustment. Nitrogen regeneration was also evaluated in sediment-water microcosms that contained one liter of littoral sediment and nine liters of synthetic lakewater. Following a one-week stabilization period, the pH in al 1 microcosms was 5.8-6.1. Treatment pH levels (3.5, 4.0, 4.5, 5.0, and control -"^5.6) were attained by adding acid (1 N H2SO4) to duplicate microcosms. Repeated additions of acid were required over the 15-week experiment to maintain desired pH levels, apparently because of neutralization of the added acid by the sediments (see Chapter 3). To measure mineralization products directly, ^^N-labelled algae was added to each microcosm. To obtain the labelled algae, Mougeotia sp. was grown on WC medium with ^^KN03 substituted for 50% of the NaN03. The dense growth obtained in three weeks was collected by passing the entire contents of the culture aquaria through 200 u Nitex screen, then through 22 u Nitex screen. The collected algae was then homogenized using the pulse cycle on a Waring blender. The resulting slurry was divided into five beakers and adjusted to treatment pH levels with 0.1 N H2SO4. Three fifty mL aliquots from each pH level were poured into smaller beakers and allowed to sit overnight. The last step was taken to assure that any nutrient release due to the shock of rapid pH change would occur prior to adding the algae to the microcosms. Individual aliquots of the algae were then reconcentrated with the Nitex screen and added directly to the microcosms; one beaker

PAGE 113

106 of algal concentrate at each pH level also was saved to determine dry weights. The microcosms were kept in the dark to observe nutrient regeration rates; samples for nutrient analyses and isotope ratio analysis were collected throughout the next 105 days. Nitrogen Mass Balance A mass balance of nitrogen in McCloud Lake was compiled for 1982. Methods for measuring wet precipitation and seepage fluxes and for estimating dry deposition of NH4'*', NO3", NO2, and HNO3 were described in Chapter 4. Atmospheric deposition of organic nitrogen was estimated using ratios of organic N/inorganic N observed at other north Florida sites (Brezonik et al. 1983b) together with inorganic N deposition data for the McCloud site. Results Enclosure Experiment Concentrations of nitrogen species in the littoral mesocosms are plotted in Figure 5-2. The most striking observation in this experiment is the large peak in NO3" that followed a peak in NH^''" in the acidified mesocosm. The increase in NH^"*" occurred during the seventh through tenth weeks and was accompanied by a decrease in D.O. to <75% saturation (compared with 100% saturation throughout the rest of the experiment). Visual signs of decomposition including increased turbidity (week 10) and floating macrophyte strands and insect castes (week 11) were also observed. Ammonium concentrations peaked at 0.180 mg N/L during week 10 and then declined during the next three weeks while NO3" rose to 0.212 mg N/L by week 13. This sequence of events strongly suggests a classic ammonification-nitrification sequence, and

PAGE 114

107 0 b 5 0 ? ac o 0. c / O MII4-11 fiO^-ii y?y i^^Q — ty—Vi r T T "1 0 3 0.2 0 6 0.6 ; 2 3 i 5 6 ; 8 9 10 11 1? 13 U IS 16 17 le • Oryanic fl inl.ol ) OtlH^-tl 1 2 5 6 7 9 10 11 12 13 \i IS 16 17 0.0 0.3 0.2 > 0. 1 S 0.0 1.0 0.6 J 0.6 w O.i < S 0.2 o 0.0 9 10 II i; 13 14 IS 16 17 18 J 1 1 0.0 0.3 0.? 0.1 S 1.2 1.0 0.8 o 0.6 a: 0.4 O 0.0 D. LITTORAL 20N£ Organic N Onh,-n ^ NOj-N 1 2 3 0.3 0.2 J I I I I 0.1 0.0 Figure 5-2. Concentrations of organic nitrogen, nitrate, and ammonium in McCloud Lake littoral mesocosms and adjacent littoral zone.

PAGE 115

108 is notable because nitrification often is considered to be inhibited below pH 5 (see discussion above). Levels of inorganic nitrogen were relatively stable in the other mesocosms. Nitrate concentrations increased briefly following the initial addition of base to the neutralized mesocosm but then declined and remained relatively constant. Ammonium concentrations in the control mesocosm showed a moderate increase during weeks 7-8, when algal strands appeared white and were apparently decomposing. Relatively stable inorganic N levels in the two less acidic mesocosms and in the littoral water suggest that nutrients released during mineralization were reassimi 1 ated rapidly and did not accumulate. Alternatively, the accumulation of NH^"*" at low pH may have been the result of decreased NH4''" adsorption by the surface sediments. Differences in levels of total organic nitrogen (TON), NH^"*", and NO3" in the mesocosms and the littoral zone were compared during weeks 5 through 20 using a paired t-test (Mendenhall and Ott 1972). Average levels of TON were similar in all three mesocosms (0.363, 0.343, and 0.304 mg N/L at pH 5, 4.6, and 3.6, respectively), but they were somewhat lower than TON levels in the adjacent littoral zone (mean= 0.420 mg N/L). The mean differences were significant for littoral versus pH 4.5 (p = 0.02) and littoral versus pH 3.6 (p = 0.09). The most likely explanation for lower TON levels in the mesocosms is decreased turbulence, which reduced the amount of suspended material in the water column. Ammonium concentrations were highest in the pH 3.6 mesocosm (mean = 0.061 mg N/L), largely because of the decomposition event discussed above. The paired t-test shows significant differences in NH4+ concentrations between pH 5 and pH 3.6 (p=0.07) and

PAGE 116

109 between the littoral zone and pH 3.6 (p<0.01). Nitrate levels were also highest in the low pH mesocosm (mean = 0.056 mg N/L) but were similar in the other mesocosms (mean = 0.010 mg N/L at pH 5 and 0.012 mg N/L at pH 4.6) and in the littoral zone (mean = 0.019 mg N/L). Differences between nitrate levels in the acidic mesocosm and the other treatments and the 1 ittoral zone were al 1 significant at the 0.10 level or better (Table 5-1). Measurements of oxygen and nutrient fluxes at the sediment surface revealed distinct diurnal cycles for D.O. and NH^"'' (Figure 5-3), but SRP concentrations remained so low in the sediment chambers that flux rates could not be determined. During all three 24-hour experiments O2 levels increased during the day as the result of photosynthesis and decreased during the night when only respiration occurred. Ammonium fluxes were opposite: NH^^ fluxes were negative during the day (i.e., net removal from the water) and positive at night (i.e., net gain by the water). Presumably this pattern occurred because NH4''" assimilation by macrophytes and periphyton was higher during the daytime (Toetz 1971), resulting in a net loss of NH4''' from the overlying water. At night, reduced rates of NH^"*" assimilation in conjunction with continued mineral izat ion of organic nitrogen resulted in positive NH^"^ fluxes. Daily respiration in the chambers was computed by connecting the last pre-dawn measurement with the first post-sunset measurement, as suggested by Hal 1 and Mol 1 (1978). The rationale for this is that respiration rates increase during the daytime, as evidenced by the fact that post-sunset respiration rates are nearly always higher than

PAGE 117

no Table 5-1. Statistical analysis of differences in nitrogen species in littoral mesocosms during 12 weeks following pH adjustment. Treatment pH pairs ^ 0 VS 4.0 0 VS 6.0 4.D VS 0.0 L VS b C 1 A C 0 L VS 4 0 1 L VS 0.0 Organic N d 0.019 0.059 0.040 -0.057 -0.076 -0 .116 Sd 0.191 0.167 0.244 0.138 0.097 0 .209 *d 0.34 1.22 0.57 -1.43 2.71 1 .92 P >0.10 >0.10 >0.10 >0.10 0.02 0 09 NH^ d -0.007 -0.034 -0.027 -0.016 -0.010 0 018 ^d 0.035 0.057 0.058 0.018 0.038 0 063 t -0.64 -2.07 -1.61 -4.26 -0.91 -0 99 P >0.10 0.10 0.07 <0.01 >0.10 >0 10 NO" d 0.003 -0.030 -0.032 -0.021 -0.024 0 009 Sd 0.009 0.054 0.051 0.020 0.016 0 039 t 1.12 1.92 2.17 3.64 5.30 0. 80 P >0.10 0.09 0.05 <0.01 <0.01 >0. 10 All concentrations in mg/L. d = mean difference; S. = standard deviation; t, = test statistic for paired t-test; P = probability level.

PAGE 118

Ill pre-dawn rates. Areas representing respiration (R) and gross primary productivity (GP) are illustrated in Figure 5-3. Net primary production (NP) was calculated as the difference bewteen 6P and R. Positive and negative NH^'*' fluxes were calculated by simply integrating the areas above and below the baseline ( Figure 5-1). Respiration rates in the three experiments (Table 5-2) averaged 1.37, 1.29 and 1.05 g/m^-d at pH 5, 4.6, and 3.6, respectively, and did not differ significantly among treatments (based on paired ttests). Mean gross production was nearly identical at pH 5 (1.30 g/m^-d) and at pH 4.6 (1.29 g/m^-d) but was slightly lower at pH 3.6 (1.05 g/m -d). The paired t-test showed that differences between treatments were marginal ly significant for pH 3.6 versus pH 4.6 (p = 0.10). Net production rates were very low in all chambers and represented a small portion of the daily gross production. Ratios of gross photosynthesis to respiration (P/R ratio) were close to 1.0 during all observations except one (August 25, pH 5). These data indicate that photosynthesis and respiration were essentially balanced at all pH levels during these experiments. Respiration rates determined in the mesocosm sediments were compared with sediment respiration rates determined for other oligotrophic lakes (Gundersson et al 1980; Andersson et al. 1978; Hayes and MacAuley 1959) to elucidate the relationship between pH and sediment respi ration. The data in Tabl e 5-3 incl ude the only two 1 akes from the study of Hayes and MacAuley (1959) that could be considered clear and oligotrophic (based on data from Hayes and Anthony 1958). The data of Andersson et al. (1978) include sediment respiration rates determined for acidified and non-acidified cores from L. Tjarnesjon

PAGE 119

112 I CD X ZD 0.2 r0.1 0.0 C3 S -0-1 o -0.2 A. DIURNAL OXYGEN FLUX Night' Respiration \\\\\\\1 Gross primary production DC 4.0 3.0 2.0 ^ 1.0 0.0 -1.0 5-2.0 -3.0 B. DIURNAL AMMONIUp FLUX Night ^ \\V\\) Positive flux Negative flux I I Day 9 11 pm 3 5 am 7 9 TIME OF DAY 11 3 5 pm Figure 5-3. Benthic diurnal fluxes of oxygen and ammonium in the acidified mesocosm, August 24-25, 1981.

PAGE 120

113 Table 5-2. Oxygen and ammonium fluxes in littoral mescocosms. 2 O2 Fluxes, g/m -d Sediment Whole Mesocosm Date pH GP R NP P/R GP R NP P/R 7/25 5 4.6 3.6 1.80 1.29 0.94 1.73 1.19 0.82 0.07 0.10 0.12 1.04 1.08 1.15 8/10 5 4.6 3.6 1.66 1.39 1.09 1.74 1.43 0.97 -0.08 -0.03 0.12 0.95 0.98 1.12 3.36 4.74 2.51 3.62 5.54 2.19 -0.26 -0.80 0.32 0.93 0.86 1.15 8/25 5 4.6 3.6 0.43 1.20 1.12 0.65 1.27 1.24 -0.22 -0.07 -0.12 0.66 0.94 0.90 2.97 2.98 2.35 3.18 3.51 2.67 -0.21 -0.53 -0.32 0.93 0.85 0.88 NH^ Fluxes, mg/m -d Sediment Date pH + Net + 7/25 5 25.2 15.6 9.6 4.6 8.0 11.7 -3.7 3.6 6.6 10.9 -4.3 8/10 5 24.4 11.7 12.7 105.6 52.5 53.2 4.6 17.3 10.6 6.7 32.6 63.3 -30.7 3.6 10.3 14.2 -3.9 32.7 36.8 4.1 8/25 5 14.8 1.3 13.5 26.7 11.7 15.0 4.6 4.7 5.5 -0.8 10.0 8.8 1.2 3.6 4.8 6.2 -1.4 9.6 8.7 0.9 Whole Mesocosm Net

PAGE 121

114 Table 5-3. DO uptake rates for oligotrophic lake sediments. ^2 0.. at Lake pH Depth T, C 2 g/m -d 20 C Bluff, Hayes & MacAuley (1959) 5.3 11 .188 0.393 Grand, Hayes & MacAuley (1959) 6.4 11 .264 0.552 L. uaras>jun uaiinbtroiTi cu ai ^lyou^ ( 3/79) 4.5 4.6 4.6 3 18 J 18 6 6 7 / 7 .305 .411 OIL .569 0.962 1.296 n Qi 1.65 L. riuicii \C/ 1 1 j uaiiiioLiuiii cl ai. (1980) D O J 1 o / L. nogsjon \L\j/i/) bannstroiTi ei, ai. (1980) & 7 1 t> 1 n iU t/u 1 07 L. ijaiiicojuii \0/ / 0 ) uaririoLruiii cL ai* (1980) n 3 U C o 1 n iU toy 1 nn 1 ^tnv"a MooHon (Fy/7P,\ Gahnstrom et al (1980) '^ 1 HO 1 n U Do L. jUUra ndi> Leva L Lcil \u//yj Gahnstrom et al (1980) b d o O 1 o 12 .411 0. 79 L. Skarsjon (6/75) Gahnstrom et al (1980) 5.2 8 12 .377 0.73 L. Tjarnesjon, Andersson et al (1978) 4.5 3.1 10 10 .52 .52 1.18 1.18 McCloud Lake mesocosms: pH'v^S 7/25 8/10 8/25 4.8 M.6 4.6 i 30 31 31 1.73 1.74 0.65 0.76 0.83 0.31 pH 4.6 7/20 8/10 8/25 4.5 4.4 4.4 30 31 31 1.19 1.42 1.27 0.52 0.68 0.61 pH 3.6 7/20 8/10 8/25 3.7 3.8 3.7 1 30 31 31 0.82 0.97 1.24 0.39 0.46 0.59 ^See equation 5-1.

PAGE 122

115 fol lowing a 75 day accl imation period. Since sediment respiration rates were determined at a variety of temperatures in these studies, the published rates were converted to respiration rates at 20 using the, equation (Metcalf and Eddy, 1972): K2 = K^exp^c^'I'l '''2) (5-1) where Kj,K2 = respiration rates at temperatures and T2 respectively, and t(= temperature correction factor. The value of 0.082 used for t^in these calculations was based on a review of literature on factors affecting benthic respiration (Baker 1980) and corresponds to a Qj^g ^ 2.23. Figure 5-4 shows that there is little relationship between pH and respiration for the lakes studied; a st rai ght1 i ne linear regression for these data has a correlation coefficient of only 0.09. These data suggest that pH does not greatly affect sediment oxygen consumption in lakes. Whole mesocosm GP and R, determined on August 10 and August 24 (Ogburn, unpublished data) show no differences in GP or R among treatments (Table 5-2). It is interesting to note that benthic 6P and R generally were less than 50% of the whole mesocosm GP and R, even though the depth of the mesocosms at the time of these experiments was less than 1 m. These data suggest that phytopl ankton and the periphyton and associated fauna attached to the polyethylene limnobags contributed substantially to the metabolism of the mesocosms. There was no discernible pH trend in the ratio of sediment/whole mesocosm GP or R. Benthic ammonium fluxes showed a clear pH trend. Positive fluxes were greater at pH 5 (mean = 22 mg/m^-d) than at pH 4.6 (mean of NH^

PAGE 123

116 3.0 4.0 5.0 6.0 7.0 PH Figure 5-4. Sediment oxygen utilization at 20 C.

PAGE 124

117 = 10 mg/m -d) or pH 3.6 (mean = 7 mg/m^-d). The paired t-test indicates that the differences were significant between pH 5.6 and pH 4.6 (p = 0.07) and between pH 5 and pH 3.6 (p = 0.03). Negative NH4'^ fluxes were similar in all three mesocosms and generally lower than the positive fluxes (Table 5-2). Net NH^"*" fluxes were positive at pH 5 (mean = 12 mg/m^-d) and at pH 4.6 (mean = 0.7 mg/m^-d) but negative at pH 3.6 (mean = -3.2 mg/m^-d). These differences were significant for pH 5 versus pH 4.6 (p = 0.05) and for pH 5.6 versus pH 3.6 (p < 0.01). Positive fluxes for the whole mesocosms also were higher at pH 5 (66 mg/m^-d) than at pH 4.6 or pH 3.6 (mean = 21 mg/m^-d in each), while negative fluxes were relatively uniform (Table 5-2). These differences were, however, not significant at the p = 0.10 level. Ammonium Adsorption Effects of acidification on phosphorus adsorption and precipitation have been studied (Aimer et al. 1978), but there is no mention in the literature on the effects of acidification on NH4'*' adsorption. It is reasonable to postulate that NH^"*" is desorbed from sediment surfaces at low pH because surfaces are less negatively charged at low pH. To determine whether desorption occurs, NH^"*" levels were measured at each pH level of a batch acidification experiment (Chapter 3) using McCloud littoral sediment. The results (Figure 5-5) show that NH^"*" desorption is substantial below pH 5. At pH 4.94 (control), the average ammonium concentration was 0.033 mg N/L, and the concentration increased to 0.140 mg N/L at pH 3.57. Since biological activity was inhibited by the the addition of chloroform, cation exchange is the only reasonable mechanism to account for these results. It can be inferred from these results that lake acidification may result in

PAGE 125

118 'N0IiVyiN33N03 ^HN + ^HN

PAGE 126

119 enhanced NH^'*' re 1 ease i f the pH of the surf icial sediments a1 so decreased. As pH levels approach the pK;^ for NH^'*' (8.3), an increasing fraction of the total NH3 + NH^"*" is present as NH3, a neutral species that does not undergo cation exchange. At pH 7, only 5% of the total NH^"*" + NH3 exists as NH3, while at pH 8.3, 50% exists as the nonionized species that would be found in solution rather than on sediment surfaces. Thus, as pH in the experimental bottles increased from 5.0 to 8.0, ammonium levels increased from less than 0.05 mg N/L to over 0.12 mg N/L (Figure 5-5). Microcosm Experiments Water-only microcosms In the initial decomposition experiment, pH-adjusted McCloud sediment was used as seed to facilitate microbial decomposition of dried plankton. Measurements of dissol ved oxygen at the the end of week 3 show that maximum O2 uptake occurred at pH 4.5. Uptake was significantly lower at pH 2.6 and at pH 6.6 (Figure 5-6A). Apparently, the one-week acclimation period was not sufficient to grow bacteria with higher pH optima. The experiment was repeated using a mixed seed developed from McCloud sediment and activated sludge (pH '^1). The results of this experiment (Figure 5-6B) show that O2 uptake increased in the pH range 2.6 to 4.6 but remained constant in the pH range 4.6 to 6.6. The addition of bacteria from a neutral -pH environment (activated sludge) thus broadened the pH range of optimum oxygen uptake, presumably reflecting the mixture of bacteria having different

PAGE 127

120 5.0 4.0 3.0 CM O CD Q O CQ 2.0 1.0 1.0 2.6 A. SEED SOURCE: McCLOUD LAKE SEDIMENT Mean and 95% confidence interval shown. 3.6 4.6 PH 5.6 6.6 o CD Q O CQ 4.0 3.0 ^ 2.0 1.0 0.0 2.6 B. SEED SOURCE: ACTIVATED SLUDGE + McCLOUD LAKE SEDIMENT Mean and 95% confidence interval shown 3.6 4.6 5.6 6.6 pH Figure 5-6. Three-week BOD versus pH with McCloud Lake sediment seed (A), and activated sludge + sediment seed (B).

PAGE 128

121 pH optima. The results suggest that bacteria from the surficial sediment of McCloud Lake are well-adapted to their environmental pH. Sediment-water microcosms Ammonium concentrations in the sediment-water microcosms fluctuated considerably (Figures 5-7 to 5-11) but did not increase significantly following the addition of labelled algae in any of the microcosms with a pH > 4. At pH 3.5, a modest rise in NH^"*" levels did occur (Figure 5-7). Concentrations in both pH 3.5 aquaria increased to^ 0.140 mg N/L by day 17, then gradual ly decl ined. Isotope ratio analyses of microcosm NH4'*" conducted on days 2, 9, 22, and 38 showed that very little of the NH^"*" in the water at any time originated from the labelled algae (Table 5-4). The maximum observed ^^NH^"*" concentrations in the microcosms were at pH 3.5 (0.010 mg ^^N/L in replicate A and 0.014 mg ^^N/L in replicate B); ^^NH^"^ levels in the other microcosms never exceeded 0.010 mg/L. Since the 68 mg algae added to each microcosm with a 50% label contributed r^-O.lS mg organic N/L, the NH^"*" in the water column always was less than 8% of the organic N added. These results are consistent with those of Nichols and Keeney (1973) who reported little ammonium release following the addition of herbicide-killed macrophytes to sediment-water microcosms. Thus, most of the increase in NH^'*' at pH 3.5 was apparently due to the continued release of sediment-bound NH4''". As noted earlier, frequent additions of acid were required to maintain the low pH levels in this experiment, and the release of sediment-bound NH4''' continued during at least the first 30 days following the addition of labelled algae. These results are consistent with the finding of NH4''' desorption in the sediment acidification experiments.

PAGE 129

122 o 1/H 9W 'N0IiVyiN33N03 n/N 9W 'N0IiVyiN3DN03

PAGE 130

123 1/N 9W 'N0IiVyiN33N03 1/N 9W 'N0IiVyiN33N03

PAGE 131

124 o OOOOOOOO OOOOOOOO SCT) OOOOOOOO OOOOOOOO -r^/N 9W 'NOIiVyiNBONOO n/N 9W 'N0IiVyiN33N03

PAGE 132

125

PAGE 133

126 B If) o OOOOOOOO 3 oooooooo t~^y3Ln^roc>j— to ai t-.(i3Ln^ooc\j>— (O • • •• • ... .1oooooooo Ut. oooo oooo VN 9W 'NOIiVMNBDNOa VN 9W N0IiVyiN33N03

PAGE 134

127 15 Table 5-4. Concentrations of NH.-N in sediment-water microcosms. Values in mg N/L pH and Replicate Day 2 Day 9 Day 22 Day 38 3.0 A n nriQ U UU3 n mn 0.002 3 C D 0.0 D U Ult 0 010 0.005 /I n A
PAGE 135

128 Levels of NO3" increased in all microcosms. Since no nitrate was added, and NO3" does not accumulate in sediments by adsorption, it is clear that nitrification occurred at all pH levels. At pH 3.5, NO3" levels increased from 0.123 mg N/L at day 1 to 0.368 mg N/L by day 29 in replicate A and from 0.063 mg N/L at day 1 to 0.210 mg N/L by day 60 in replicate B. Although nitrate levels increased significantly at pH 3.5 and 4.0, the increases were more pronounced at pH 5, where concentrations increased from < 0.100 mg N/L to over 0.400 mg N/L within the first month (Figure 5-10). At pH 5.7 (control), NO3' levels were already high (-^0.45 mg N/L) when the labelled algae were added, but increased further to over 0.60 mg N/L within the first month (Figure 5-12). Isotope ratio analyses were not conducted for the NO3" in the microcosms, and consequently it is not possible to determine the fraction of N03~ produced that originated from the labelled algae. It is likely that some of the NO3" originated from sediment-bound NH^"*" that was oxi di zed fol 1 owi ng re 1 ease to the water. Regard 1 ess of the ammonium source, it is clear that nitrate was produced, even at pH 3.5. This confirms the observation of nitrification in the McCloud Lake littoral mesocosms and demonstrates that nitrification can occur in highly acidic lakes. Further work is needed to demonstrate whether the process is mediated by pH-tolerant heterotrophs or whether autotrophic nitrifiers carry out the process in relatively neutral sediments.

PAGE 136

129 McCloud Lake Nitrogen Balance Precipitation Methods of measuring inorganic nitrogen deposition were discussed in Chapter 3. To estimate the input of total nitrogen entering McCloud Lake, an estimate of organic N was needed. Brezonik et al. (1983b) reported that organic N deposition in bulk precipitation was 36-64% of the inorganic N deposition at 4 north Florida sites (Gainesville, Waldo, Jasper, and Hastings). These ratios were used to estimate organic N deposition to McCloud Lake from the bulk deposition (wet + aerosol) inorganic N (Table 5-5). Wet deposition accounted for 20-48% of the total deposition of NO3" and HH^'^. Since concentrations of both inorganic species were inversely correlated with monthly precipitation volume (r = -0.53 for NOj" and -0.47 for NH^"*"), monthly loadings were fairly uniform (Figure 5-12). Gaseous deposition of NO2 and HNO3 (3.1-9.2 kg/ha-yr) was an important source of N03~ to the lake, accounting for approximately 40% of the atmospheric nitrogen input. Aerosol NH4''' deposition was a minor source of nitrogen, accounting for only 0.1 0.3 kg N/ha-yr. Deposition of organic N, estimated to be 1.2-3.6 kg N/ha-yr, was comparable with bulk (wet + dry) ammonium deposition. Bulk precipitation of total N (Table 5-5) at McCloud Lake (estimated from wet + aerosol deposition to be 4.4 9.3 kg/ha-yr) was similar to the state average of 7.7 kg/ha-yr (Brezonik et al. 1983b) and to the 1968 bulk deposition at nearby Anderson-Cue Lake (Brezonik et al. 1969). While bulk total N deposition has not changed appreciably in 14 years, bulk NO3" deposition appears to have increased from 1.7 kg/ha-yr to 2.1-3.3 kg/ha-yr.

PAGE 137

130 Table 5-5. Deposition of nitrogenous species at McCloud Lake. kg/ha-yr nhJ NOTON TN McCloud Lake, 1982 Wet only Aerosol Est. bulk Gaseous 1.0 0.1-1.4 1.1-2.4 2.0 0.1-1.3 2.1-3.3 2.9-8.7 (NO2) 0.2-0.5 (HNO3) 1.2-3.6 4.4 -9.3 Total 1.1-1.4 5.2-12.5 1.2-3.6 7.5-17.5 Ander<;nn-rijp IQfiR Bulk 1.7 1.7 2.9^ 6. 3 State ave., 1978-79 Bulk 7. 7 ^Nine-month total

PAGE 138

131 Sept. Oct. Nov. Dec. Jan. Feb. Mar. Apr. May June July Aug. 1981 MONTH 1982 NO, I I I I I I I I I I X Sept. Oct. Nov. Dec. Jan. Feb. Mar. Apr. May June July Aug, 1981 MONTH 1982 Figure 5-12. Wet-only deposition of inorganic nitrogen to McCloud Lake.

PAGE 139

132 McCloud Lake nitrogen Concentrations of organic N and inorganic species fluctuated considerably during 1981-1982, although a seasonal pattern is not clear (Figure 5-13). The distinct rise in nitrate levels during the fall and winter suggests that nitrification may have occurred, particularly since increasing NO 3" levels paral 1 el ed a decl ine in NH4''". However, NO3" deposition during the period October, 1981 -January 1982 was estimated to be 7-20 kg, or 0.06-0.17 mg N/L lakewater, and could account for the observed rise if algal uptake of nitrate were minimal. Data in Table 5-6 indicate that the TN/TP ratio in 1981-82 was 40:1 while the ratio of soluble inorganic nitrogen to soluble reactive phosphate (SIN:SRP) was 33:1. Since a SIN:0P ratio of 20:1 generally is recognized as an indication of phosphorus limitation (Porcella and Bishop 1975), it can be concluded that algal growth in McCloud Lake is phosphorus limited. Nutrient data from 1968-69, 1978-79, and 1981-82 (Table 5-6) reveal no significant trends in TON, NH^'*', NO3", total P, or SRP. If acidification has caused a decreased in nutrient levels, the trend is not yet discernible in McCloud Lake. This observation is consistent with the studies of Schindler et al. (1980), Hultberg and Andersson (1982), and Di 1 1 on et al (1979) which report little or no change in nutrient levels following artificial pH manipul ations in lakes. Mass bal ance As noted in Chapter 3, 98-99% of the nitrate and 95-98% of the ammonium entering McCloud Lake comes from atmospheric deposition. Organic nitrogen in seepage was not measured, but since seepage flows

PAGE 140

133 1.0 9 NDJ FM AM JJAS ON DJFM AM JJ A 1980 1981 1982 DATE Figure 5-13. McCloud Lake nitrogen species.

PAGE 141

134 Table 5-6. McCloud Lake nutrient concentrations, 1968-1982. Concentrations as mg/L 1968^ 1978*^ 1981-82 Total organic N 0.420 0.243 0.363 + NHJ N 0.105 c 0.184^ 0.070 NO3 N 0.041 0.063 Total N 0.566 0.427 0.484 Organic P 0.006 0.011 0.012 PO" P 0.006 0.005 0.004 Total P 0.012 0.016 0.012 TN/TP 47 27 40 SIN/OP 24 37 33 ^Brezonik et al (1969). Brezonik et al (1983b). Total inorganic nitrogen.

PAGE 142

135 were <1Q% of the total water input, inputs of organic N via seepage were assumed to be negligible. Inorganic nitrogen entering the lake was rapidly immobilized. Ammonium has a residence time of 0.7-1.4 years and nitrate has a residence time of 0.1-0.3 years (Table 5-7). Ammonium probably is immobilized by biological assimilation and sedimentation as organic N. Nitrate is immobi 1 ized by biol ogical assimi 1 ation, but al so may be lost via denitrification. Several lines of evidence support the view that denitrification may occur in McCloud Lake. First, nitrate was almost completely removed from the eluate in the groundwater seepage experiments (Chapter 3). Second, pore water profiles in both the littoral and profundal sediments (methods described in Chapter 4) show a dramatic N03~ gradient in the upper 5 cm (Figure 5-14), suggesting that nitrate diffuses into the surficial sediments and is transformed. Finally, NO3" levels in the sediment-water microcosms decreased after reaching peak levels (Figures 5-7 to 5-11). Although these data clearly suggest that nitrate is transformed in the sediments, it is not certain whether denitrification is the mechanism responsible. Assimilation by macrophytes, algae, or bacteria in the surficial sediments could cause the observed depletion of nitrate. However, while denitrification is a pH-sensitive process, other pH-sensitive processes, including sulfate reduction and nitrification, can occur in McCloud Lake sediments. Thus, low pH of the overlying water is not a^ priori evidence that denitrification could not occur. Although these data suggest that denitrification may occur, identification of end products (N2O or N2) would be needed to provide conclusive evidence.

PAGE 143

136 Table 5-7. McCloud Lake nitrogen budget. no;, Nh!, Organic N, Kg N/yr Kg N/yr Kg N/yr Precipitation 26.7-63.6 5.9-12.4 6.0-18.1 Wet 10.4 5.2 Dry 16.3-53.2 0.77.2 Aerosol 0.76.4 0.77.2 Gas 15.6-46.8 Seepage In 0.5 0.4 ? Out 1.1 1.2 ? Storage 9.4 8.4 55.9 kg -4.4 -13.8 T, yr 0.10.3 0.71.4 3.19.3

PAGE 144

LAKE I 1 1 1 ^—/P 0 1 2 3 4 14 NITRATE CONCENTRATION, UM Figure 5-14. Nitrate in McCloud Lake sediment pore waters.

PAGE 145

138 It can be seen that organic N in McCloud Lake has a much longer residence time (3.1-9.3 years) than the inorganic species. This undoubtedly occurs because inorganic N is assimilated in the water column, resulting in sustained high levels of organic N. The significance of atmospheric nutrient loading in maintaining productivity can be seen by comparing atmospheric inputs of N and P to McCloud Lake to established nutrient loading criteria (Vol 1 enweider 1968, 1975; Shannon and Brezonik 1972). Current phosphorus loadings to the lake are approximately one half or less of the loading that would be required to sustain mesotrophic conditions (Table 5-8). Nitrogen loadings, in contrast, are at or near the minimum mesotrophic loading rates. These data support the contention, based on inorganic N/P ratios, that lake productivity is phosphorus limited, and illustrate the significance of atmospheric nutrient deposition as a source of nutrients to seepage lakes in this region. Although this study has focused on the impact of acidification, atmospheric deposition is also significant from a standpoint of eutrophication. Since McCloud Lake receives enough nitrogen through atmospheric deposition to sustain mesotrophic conditions, the addition of small amounts of phosphorus would result in enhanced productivity. For example, according to the Vol 1 enweider (1975) loading criterion, the addition of 0.2 mg P/m -yr would result in mesotrophic conditions. For this 5 hectare lake, 0.2 mg P/m -yr corresponds to the addition of only 10 kg phosphorus, an input that would occur with the additon of domestic waste from a mere 5 individuals assuming an input of 2 kg/capita-yr (Vol 1 enweider 1968). While McCloud Lake is protected from residential development.

PAGE 146

139 Table 5-8. Atmospheric deposition of nitrogen and phosphorus relative to loading criteria. Minimum McCloud mesotrophic atmospheric Criterion Units loading input Phosphorus Vollenweider (1968) g/m^-yr 0.044 0.027 Shannon & Brezonik (1972) g/m^ -yr 0.022 0.010 Vollenweider (1975) g/m^-yr 0.10-0.11 0.027 Nitrogen Vollenweider (1968) g/m^-yr 0.67 0.7-1.8 Shannon & Brezonik (1972) g/m^-yr 0.86 0.3-0.7

PAGE 147

140 extensive development is occurring on nearby lakes and can be expected to cause enhanced productivity of these lakes. Thus, while atmospheric deposition of acidic nitrogen species may be largely neutralized by incorporation into organic matter, atmospheric nitrogen inputs may contribute to enhanced algal productivity. Concl usions Although Grahn et al. (1974) and others have postulated that lake acidification results in diminished nutrient regeneration, this study and other recent studies indicate that the effect of acidification on decomposition and nutrient regeneration may be minimal. Evidence to support this conclusion include the following: 1) There is no demonstrated reduction in sediment oxygen utilization with decreasing pH. There were no significant differences in sediment respiration rates among the littoral mesocosms (pH 3.5-5). Data compiled from other studies also fail to show a pH trend. 2) The water-only microcoms showed that oxygen uptake was higher at pH 4.5 than at higher or lower pH levels when McCloud littoral sediments were used as a seed, indicating that microorganisms in the surficial sediments are adapted to the lake pH. 3) Nitrification, which is often assumed to cease below pH 5, occurred in the littoral mesocosms and in the sediment-water microcosms at a pH of 3.5, although rates appeared to be reduced below pH 4.5-5. Further study is needed to determine whether the process is mediated by autotrophic nitrifiers living in the relatively neutral sediments or whether the process is mediated by pH-tolerant heterotrophs.

PAGE 148

141 4) Denitrification may also occur in McCloud Lake. Nitrate was depleted in in situ littoral and profundal sediment pore waters. Over 90% of the nitrate was removed from the eluates in the groundwater seepage experiment, and nitrate levels decreased from peak levels in the microcosm experiment. 5) Concentrations of inorganic and organic nitrogen and phosphorus species in McCloud Lake do not appear to have changed significantly over the past 14 years, despite a decrease in pH of nearly 0.5 units. The accumulation of NH^^"*" in the acidified littoral mesocosms and in the acidified sediment-water microcosms suggests that some ammonium regeneration occurs at pH levels down to 3.5. This accumulation in the water column may result from decreased NH^"*" adsorption at low pH (demonstrated in the sediment titration experiments) or because of reduced assimilation or nitrification. The diurnal sediment flux studies show that positive NH4"*" fluxes decreased with decreasing pH, but is not clear whether this occurred because of decreased ammonification at low pH or because of reduced assimilation by the benthic macrophyte community. Atmospheric inputs accounted for over 95 % of the total nitrogen input to the lake. Inorganic nitrogen entering McCloud Lake was rapidly immobilized. This is significant because it results in a net consumption of protons in the lake, as discussed in Chapter 3. Atmospheric inputs of nitrate, but not of total nitrogen, appear to have increased somewhat during the past 14 years. Atmospheric nitrogen inputs are currently sufficient to maintain mesotrophic conditions.

PAGE 149

CHAPTER 6 CONCLUSIONS Sediment Neutralization Laboratory experiments conducted with sediment-water slurries showed that sediments from softwater lakes can neutralize additions of acidity. The acid neutralizing capacity of profunda! sediments is as high as 20 meq/100 g in the pH range 4.5-5.5 and is correlated with the organic content of the sediments. Cation exchange is the major abiotic mechanism of H'*'-neutral i zation. Sulfate adsorption, protonation of humic acids, and dissolution of aluminum minerals were unimportant as neutralization mechanisms. Evidence of sulfate reduction was obtained in the seepage column experiments and from in situ sediment pore water analyses in McCloud Lake, and this process is probably an important H'^-consumi ng mechanism in McCloud Lake sediments. McCloud Lake Mass Balance McCloud Lake receives 90% of its water from precipitation; the remaining 10% of the inflow is groundwater seepage. Nearly all of the sulfate, nitrate, ammonium, sodium, chloride, and protons in McCloud Lake enter via precipitation. An internal sink (sulfate reduction) consumed 37-73% of the sulfate entering the lake. Approximately 90% of the ammonium and nitrate entering the lake was consumed by internal sinks. Biological assimilation is undoubtedly the most important sink for inorganic nitrogen species. Additional losses of ammonium may 142

PAGE 150

143 occur by adsorption to sediment surfaces and additional losses of nitrate may occur by denitrif icati on. Most (86-94%) of the protons entering the lake were removed by internal sinks, primarily sulfate reduction and assimilation or denitrif ication of nitrate. Nitrogen Cycling and Decomposition Although decompositon is believed to be retarded in acidic lakes, there was no evidence for reduced sediment oxygen utilization in the littoral mesocosms in the pH range 3.6 to 5. Data compiled from other studies also fail to show a pH trend in sediment oxygen utilization. Water-only microcosms showed that oxygen utilization was highest at pH 4.5 when McCloud Lake sediment was used as a seed, indicating that bacteria in the lake are well-adapted to acidic conditions Nitrification, which is often assumed to cease below pH 5, occurred in the littoral mesocosms and in the sediment-water microcosms at pH levels as lowas 3.5, although nitrification rates were reduced below pH 4.5-5.0. Nitrification probably occurs in the surficial sediments, where pH levels are higher than lake pH levels. Denitrifi cation, which is al so bel ieved to be inhibited bel ow pH 5, occurred in the seepage experiment and in the in situ sediments. Concentrations of nutrient species do not appear to have changed significantly during the past 14 years, although the lake pH has decreased by 0.5 units. This observation is consistent with the results of several whole-lake manipulation experiments and suggests that nutrient levels in lakes may not decrease appreciably with moderate acidification.

PAGE 151

BIBLIOGRAPHY Alexander, M. 1980. Effects of acidity on microorganisms and microbial processes in soil. Jjk Hutchinson, T. C, and M. Havas (eds.). Effects of Acid Rain on Terrestial Ecosystems Plenum Press, New York. Aimer, B., W. Dickson, C. Ekstrom, and E. Hornstrom. 1978. Sulfur pol 1 ution and the aquatic ecosystem. Uk_ Nriagu, J. 0., (ed.) Sul fur vn the Environment. Part II: Ecological Impacts John Wi 1 ey and Sons, New York. Andersson, G., S. Gleischer, and W. Graneli. 1978. Influence of acidification on decomposition processes in lake sediment. Verh. Internat. Verein. Limnol. 20: 802-807. APHA, 1971. Standard Methods for the Exami ni nati on of Water and Wastewater, 8th ed^, American PuETTc Health Association, Washington, D.C. APHA, 1981. Standard Methods f or the Exami nati on of Water and Wastewater, 15th ed., American PubTTc Health Association, Washington, D.C. Baker, L. A. 1980. Predicted Limnology of the Ridges Basin Reservoir M.S. Thesis, Utah State University, Logan, Utah. Blackmer, A. M., and J. M. Bremner. 1978. Inhibitory effect of nitrate on reduction of N2O to No by soil microorganisms. Soi 1 Biology and Biochemistry 10: 187-191. Bremner, J. M., and D. W. Nelson. 1968. Chemical decomposition of nitrite in soils. Trans. 9th Internat. Cong. Soil Science II: 495503. Brewer, P. G. and J. C. Gol dman. 1976. Al ka 1 i ni ty changes generated by phytopl ankton growth. Limnol. Oceanogr. 21: 108-117. Brezonik, P. L., L. A. Baker, R. W. Ogburn, III, E. S. Edgerton, and T. L. Crisman. 1983a. Ecological Effects of Acid Precipitation on a Sensitive Softwater Lake in Florida first annual report, USEPA-NCSU APP-007-02-1980. Brezonik, P. L., and C. L. Harper. 1969. Nitrogen fixation i some lacustrine environments. Science 164: 1277-1279. Brezonik, P. L., C. 0. Hendry, Jr., E. S. Edgerton, R. L. Schulze, and T. L. Crisman. 1983b. Acidi ty. Nutrients, and Minerals in 144

PAGE 152

145 Atmospheri c P reci pi tat i on over F 1 orida: Deposition Patterns, Mechani sms, ancTEcol ogical Effects USEPA #805560, Corvallis, Ore. Brezonik, P. L., W. H. Morgan, E. E. Shannon, and H. D. Putnam. 1969. Eutrophication Factors in North Central Florida Lakes Fla. Water Resources Research Center Publ. No. 5, Gainesville, Fla. Brezonik, P. L., and E. E. Shannon. 1971. Trophic State Characterizati on of Lakes i n North Central Fl orida, FT orida Water Resources Research Center Publication # 13, Gainesville, Fla. Broadbent, F. E., and F. E. Clark. 1965. Deni tri f i cati on. In: Bartholomew, W. V., and F. E. Clark (eds.). Soil Nitrogen Amer. Soc. Agronomy, Madison, Wis. Brock, T. 1973. Lower pH limit for the existence of blue-green algae: evolutionary and ecological implications. Science 179: 480-483. Chamberlain, A. C. 1980. Dry deposition of sulfur dioxide. In: Shriner, D. S., C. R. Richmond, and S. E. Lindberg (eds.), Atmospheri c Sul fur Deposition: Environmental Impact and Human Heal th Effects Ann Arber Science, Ann Arbor, Mich. Chamberlain, A. C, and R. C. Chadwick. 1953. Deposition of airborne radioiodine vapor. Nucl eonics 8: 22-25. Chen, C. W., S. A. Gherini, and R. A. Goldstein. 1979. Modeling the lake acidification process. Ijn: Proc. Workshop on Ecological Effects of Acid Precipitation Central Electricity Research Laboratory, Surry, England. Chri stopherson, N. and R. F. Wright. 1981. Sulfate budget and a model for sulfate concentration in stream water at Birkeness, a small forested catchment in southernmost Norway. Water Resour. Res. 17 : 371-389. CI imatological Abstracts, 1981 Prepared by U. S. Weather Bureau, Wash! ngton D. C. CI imatol ogical Abstracts. 1982 Prepared by U. S. Weather Bureau, Washington, D.C. Cogbill, C. V. and G. E. Likens. 1974. Acid precipitation in northeastern United States. Water Resour. Res. 10: 1133-1137. Collins, V. G., B. T. D'Sylva, and P. M. Latter. 1978. Microbial populations in peat. In: Perkin and Heal (eds.). Ecology of British Moors and Montane GrassTands Ecol. Studies, 27: 113-125. Cook, R. B., and D. W. Schindler. 1983. The bi ogeochemi stry of sulfur in an experimentally acidified lake. Uk Hallberg, R. (ed.). Environmental Biogeochemi stry Ecol. Bull. 35:115-127.

PAGE 153

146 Cook, R. B., D. W. Schindl er, and C. A. Kel ly. 1982. Neutral ization of Hydrogen Ion i n an Experimental 1 y Acidi f ied Lake: A1 kal inity Production in the Anoxic Hypolimnion Lamont-Doherty Geological Observatory Contribution No. 0000. Crisman, T. L., R. W. Bienert, J. A. Foran, M. A. Gunn, P. Schuerman, N. Gourlie, R. A. Garren, M. W. Binford, R. W. Ogburn, and P. L. Brezonik. 1983. Current and Projected Effects of Acid Precipitation on the Biota of Florida Lakes first annual report, USEPA APP-007-021980. Crisman, T. L., R. L. Schulze, P. L. Brezonik, and S. A. Bloom. 1980. Acid precipitation: the biotic response in Florida lakes. j_n: Drablos, D., and A. Tollan (eds.), Proc. Int. Conf. Ecol. Impact Acid Precip. SNSF project, Oslo, Norway. Dickson, W. 1978. Some effects of the acidification of Swedish lakes. Verh. Internat. Verein Limnol 20: 851-856. Dillon, P. J., N. D. Yan, W. A. Scheider, and N. Conroy. 1979. Acidic lakes in Ontario, Canada: characterization, extent and responses to base and nutrient additions. Arch. Hydrobiol Be1th. 13: 317-336. Dow Chemical Corporation. Operation and Maintenance Manual for the Dionex Ion Chromatograph Dovland, H., E. Joranger, and A. Semb. 1976. Deposition of air pollutants in Norway, jji: Braake, F. H.(ed.), Impact of Acid Precipitation on Forest and Freshwater Ecosystems of Norway SNSF project, Oslo, Norway. Driscoll, C. T., J. P. Baker, J. J. Bisogni, and C. L. Schofield. 1982. Aluminum speciation and equilibria in dilute acidic surface waters of the Adirondack region of New York. Presented at the Symp. Acid Precip. Am. Chem. Soc. Ann. Meeting, Las Vegas, Nev. Edgerton, E. S. 1981. A Mass Balance Atmospheric Sulfur Model for Florida M. S. Thesis, University of Florida, Gainesville. Ekpete, D. M., and A. H. Cornfield. 1965. Effect of pH and addition of organic materials on denitrification losses from soil. Nature 208: 1200-1202. Fellows, C. R., and P. L. Brezonik. 1980. Seepage flows into Florida lakes. Water Resources Bui 1. 16: 635-641. Focht, D. D. 1974. The effect of temperature, pH, and aeration on the production of nitrous oxide and gaseous nitrogen a zero order kinetic model Soil Science 118: 173-179.

PAGE 154

147 Fowler, D. 1980. Removal of sulfur and nitrogen compounds from the atmosphere in rain and by dry deposition. _I_n: Drablos, D., and A. Tol Ian (eds.), Proc. Int. Conf. Ecol. Impact Acid Precip. SNSF project, Oslo, Norway. Francis, A. J. 1981. Effects of acidic precipitation and acidity on soil microbial processes. Submitted to Water, Air. Soil Poll. Gahnstrom, G. G. Andersson, and S. Fleischer. 1980. Decomposition and exchange processes in an acidified lake sediment. In: Drablos, D. and A. Tol Ian (eds.), Proc. Int. Conf. Ecol. Impact Acid Precip. SNSF project, Oslo, Norway. Galloway, J. N., B. J. Cosby, Jr., and G. E. Likens. 1979. Acid precipitation: measurement of pH and acidity. Limnol. Oceanogr. 24: 1161-1165. Galloway, J. N., C. L. Schofield, G. R. Hendrey, E. R. Altwicker, and D. E. Troutman. 1980. An analysis of lake acidification using annual budgets. j_n: Drablos, D, and A. Tollan (eds.), Proc. Int. Conf. Ecol Impact Acid Precip ., SNSF project, Oslo, Norway. Garland, J. A. 1978. Dry and wet removal of sulfur from the atmosphere. Atmos. Environ. 12: 349-362. Gjessing, E. T., A. Henriksen, M. Johannessen, and R. F. Wright, 1976. Effect of acid rain on freshwater chemistry. In: Braake, F. H. (ed.). Impact of Aci d Precipitation on Forest and Freshwater Ecosystems of Norway SNSF project, Oslo, Norway. Grahn, 0., H. Hultberg, and L. Lander. 1974. 01 igotrophication -a self-accelerating process in lakes subjected to excessive supply of acidic substances. Ambio 3: 93-94. Hall, R. J., G. E. Likens, S. B. Fiance, and G. R. Hendry. 1980. Experimental acidification of a stream in the Hubbard Brook Experimental Forest, New Hampshire. Ecol ogy 61: 976-989. Hall, R. J. and J. J. Moll. 1978. Primary productivity measurement in aquatic systems. J_n: Likens, G. E. (ed.). Primary Productivity of the Biosphere Int. Biol. Studies, Cornell U., Ithaca, N.Y. Hayes, F. R., and E. H. Anthony. 1958. Lake water and sediment. I: characteristics and water chemistry of some Canadian east coast lakes. Limnol. Oceanogr. 3: 299-307. Hayes, F. R., and M. A. MacAulay. 1959. Lake water and sediment V: oxygen consumed in water over sediment cores. Limnol Oceanogr. 4: 291-298. — Hendrey, C, K. Baalstrud, T. Traaen, and M. Laake. 1976. Acid precipitation: some hydrobiol ogical changes, Ambio 5: 224-227.

PAGE 155

148 Hemond, H. F. 1980. Biogeochemistry of Thoreau's Bog, Massachussetts. Ecol. Monogr. 50: 507-526. Henriksen, A. 1980. Acidification of freshwatersa large scale titration. J_n: Drabl os. P.. and A. To 1 1 an (eds.) Proc. Conf. Ecol Impact Acid Precip., Oslo, Norway. Hsu, P. H. 1977. Kaolinite and serpentine minerals. In: Dixon, J. B., and S. B. Weed (eds.). Minerals in the Soil Environment Soil Science Society of America, Madison, Wis. Hul tberg, H., and I. B. Andersson. 1982. Limi ng of acidi fied 1 akes: induced long-term changes. Water, Soil, Air Poll. 18: 311-331. Hutchinson, T. C, W. Gizyn, M. Havas, and V. Zobens. 1978. Effect of long-term lignite burns on arctic ecosystems at the Smoking Hills, N. W. T. jji: Hemphill, D. D. (ed.) ,Trace Substances in Environmental Health XII, A Symposium U. of Missouri, Columbia. Impact Assessment Group. 1983. Impact Assessment Final Report Submitted to the coordinating committee, U. S. Canada memorandom of intent on transboundary air pollution. Johnson, D. W., J. W. Hornbeck, J. M. Kelley, W. T. Swank, and D. E. Todd. 1980. Regional patterns of soil sulfate accumulation: relevance to ecosystem sulfur budgets. In: Shriner, D. S., C. R. Richmond, and S. E. Lindberg (eds.). Atmospheric Sulfur Deposition: Environmental Impact and Health Effects Ann Arbor Science, Ann Arbor, Mich. Johnson, N. M., C. T. Driscoll, J. S. Eaton, G. E. Likens, and W. H. McDowell. 1981. 'Acid rain', dissolved aluminum and chemical weathering at the Hubbard Brook Experimental Forest, New Hampshire. Geoch. Cosmich. Acta 45: 1421-1437. Joranger, E., J. Schaug, and A. Semb. 1980. Deposition of air pol 1 utants in Norway. J[_n: Drablos, A. and A. To 1 1 an (eds.), Proc. Int. Conf. Ecol. Impact Acid Precip., SNSF project, Oslo, Norway. Kahl J. S., S. A. Norton, and J. S. Williams. 1982. Chronology, magnitude, and pal eol imnol ogical record of changing metal fluxes related to atmospheric deposition of acids and metals in New England. Presented at Symp. Acid Precip., Am. Chem. Soc. Ann. Meeting, Las Vegas, Nev. Kelly, C. A., J. W. M. Rudd, R. B. Cook, and D. W. Schindler. 1982. The potential importance of bacterial processes in regulating rate of lake acidification. Limnol. Oceanogr. 27: 868-882. Kerekes, J., G. Howel 1 S. Beauchamp, and T. Pol lock. 1982. Characterization of three lake basins sensitive to acid precipitation. Int. Revue ges. Hydrobiol. 76: 679-694.

PAGE 156

149 Kilham, P. 1982. Acid precipitation: its role in the alkalization of a lake in Michigan. Limnol. Oceanogr. 27: 856-867. Klein, T. M., J. P. Kreitinger, and M. Alexander. 1982. Nitrate formation in acid forest soils from the Adirondacks. Submitted to Soi 1 Sci. Soc. Am. J. Kramer, J., and A. Tessier. 1982. Acidification of aquatic systems: a critique of chemical approaches. Environ. Sci. Technol 16: 606A615A. Landers, D. H. 1979. A durable, reusable enclosure system that compensates for changing water levels. Limnol. Oceanogr. 24: 991994. Lee, J. A., and G. R. Stewart. 1978. Ecological aspects of nitrogen assimilation. Ijn: Ecological Aspects of Nitrogen Assimilation Adv. Bot. Res. 6: 1-43. Leivestad, H., G. Hendrey, I. P. Muniz, and E. Snekvik. 1976. Effects of acid precipitation on freshwater organisms. _Iji: Braake, F. H. (ed.). Impact of Acid Precipitation on Forest and Freshwater Ecosystems, of Norway SNSF project, Oslo, Norway. Lerman, A. 1978. Geochemical Process: Water and Sediment Envri ronments John Wiley and Sons, New York. Li, Y. H., and S. Gregory. 1974. Diffusion of ions in sea water and in deep-sea sediments. Geochem. Cosmich. Acta 38: 703-714. Linsley, R. K., M. A. Kohler, and J. L. H. Paulus. 1975. Hydrol ogy for Engineers, 2nd ed. McGraw-Hill Book Co., New York. MacGregor, A. N., and D. R. Keeney. 1973. Acetylene-reduction of anaerobic nitrogen fixation by sediments of selected Wisconsin lakes. J. Environ. Qual 2: 438-440. Martin, N. J., and A. J. Holding. 1978. Nutrient availability and other factors limiting microbial activity in the blanket peat. In: Perkin and Heath (eds.), Ecol ogy of British Moors and Montane Grasslands Ecol. Studies 27: 113-125. May, H. M., P. A. Helmke, and M. L. Jackson. 1979. Gi bbsi te sol ubi 1 ity and thermodynamic properties of hydroxy 1 -a 1 uminum ions in aqueous solution at 25*^0. Geoch. Cosmich. Acta 43: 861-868. McKeown, J. J., A. H. Bennect, and G. M. Locke. 1968. Studies on the behavior of benthal deposits of wood origin. JWPCF 40: R333-R353. Mendenhall, W., and L. Ott. 1972. Understanding Statistics Duxbury Press, Belmont, Calif.

PAGE 157

150 Metcalf and Eddy, Inc. 1972. Wastewater Engineering McGraw-Hill Book Co., New York. Moss, B. 1973. The influence of environmental factors on the distribution of freshwater algae: an experimental study II: the role of pH and the carbon dioxide bicarbonate system. J. Ecol. 61: 157-177. NAS. 1978. Nitrates: An Envi ronmental Assessment Report prep, by Panel on Nitrates, Coordinating Committee for Scientific and Technical Assessments of Environmental Pollutants, National Academy of Sciences, Washington, D.C. Nichols, D. S., and D. R. Keeney. 1973. Nitrogen and phosphorus release from decaying water milfoil. Hydrobiol ogica 42: 509-525. Nommik, H. 1956. Investigations on denitri f ication. Acta. Agr. Scand. 6: 195-228. Nordstrom, D. K. 1982. The effect of sulfate on aluminum concentrations in natural waters: some stability relations in the system AI2O3-SO3at 298 K. Geochem. Cosmichim. Acta 681-692. Norton, S. A., C. T. Hess, and R. B. Davis. 1981. Rates of accumulation of heavy metals in preand postEuropean sediments in New Engl and 1 akes. J_n* Eisenreich, S. J. (ed.). Atmospheric Pol 1 utants in Natural Waters Ann Arbor Science, Ann Arbor, Mich. Norton, S. A., J. S. Williams, D. W. Hanson, and J. N. Galloway. 1980. Changing pH and Meta 1 s Level s i n Streams and Lakes i n the Eastern United States Caused by Acidic Precipitation USEPA Report. Nriagu, J. 0., and J. D. Hem. 1978. Chemistry of pol 1 utant sul fur in natural waters. J_n: Nriagu, J. 0. (ed.). Sulfur in the Environment, Part II: Ecological Impacts John Wiley and Sons, New York. Porcella, D. B., and A. Bruce Bishop. 1975. Comprehensive Management of Phosphorus Water Pollution Ann Arbor Science, Ann Arbor, MI. Rajan, S. S. S. 1978. Sulfate adsorbed on hydrous alumina, ligands displaced, and changes in surface charge. Soi 1 Sci. Soc. Am. 42: 3944. ; Rippon, J. E., R. A. Skeffington, M. J. Wood, K. A. Brown, and D. J. A. Brown. 1980. Hydrogen, sulfur, and nitrogen budgets in soils and catchments. In: Drablos, D., and A. Tollen (eds.), Proc. Int. Conf. Impact Acid Precip SNSF project, Oslo, Norway. Schindler, D. W., R. B. Wageman, R. B. Cook, T. Ruszcynski, and J. Prokopowich. 1980. Experimental acidification of Lake 223, Experimental Lakes Area: background data and first threee years of acidi f ication. Can. J. Fish. Aquatic Sci. 37: 342-354.

PAGE 158

151 Schnoor, J. L., J. M. Eilers, and G. E. Glass. 1983. Acidification of lakes: chemical weathering rates as a key factor. Submitted to Nature Scofield, C. L. 1976. Acid precipitation: effects on fishes. Ambi o 5: 228-230. Sehmel, G. A. 1980. Particle and gas dry deposition: a review. Atmos. Environ. 14: 983-1011. Shannon, E. E., and P. L. Brezonik. 1972. Relationships between lake trophic status and N and P loading rates. Environ. Sci. Technol. 6: 719-725. Soderlund, R. 1981. Dry and wet deposition of nitrogen compounds. In : Clark, F. E., and T. Rosswall (eds.), Terrestial Nitrogen Cycles: Processes, Ecosystem Stratigies, and Management Impacts Ecol. Bull. No. 33, Swedish Natural Science Research Council, Stockholm. Stumm, W. and J. J. Morgan. 1981. Aquatic Chemistry, 2nd ed. Mi 1 eyInterscience, New York. Thompson, M. E. 1982. The cation denudation rate as a quantitative index of sensitivity of eastern Canadian rivers to acidic atmospheric precipitation. Water, Air, Soi 1 Pol 1 18: 215-226. Toetz, D. W. 1971. Diurnal uptake of NO3 and NH^ by Ceratophyl 1 um periphyton community. Limnol Oceanogr. 16: 819-822. Traaen, T. S. 1980. Effects of acidity on decomposition of organic matter in aquatic environments. ]jn: Drablos, D., and A. Tollan (eds.), Proc. Int. Conf. Ecol Impact Acid Preci p. SNSF project, Oslo, Norway. USEPA, 1974. Methods for Chemical Analysis of Mater and Wastes U.S. Environmental Protection Agency, Cincinnati, Ohio. Vanderhoef, L. N., Chi-Ying Huang, and R. Musil. 1974. Nitrogen f i xati on(acety 1 ene reduction) by phytopl ankton in Green Bay, Lake Michigan, in relation to nutrient concentrations. Limnol. Oceanogr. 19: 119-125. Vol 1 enweider, R. A. 1968. Scientific Fundementals of the Eutrophication of Lakes and Fl owing Waters, with Particul ar Reference to Nitrogen and Phosphorus as Factors in Eutrophication OECD, Paris. Vol 1 enweider, R. A. 1975. Input-output models with special reference to the phosphorus loading concept in limnology. Schwei z. Z. Hydrobiol. 37: 53-84. Wetzel, R. G. 1975. Limnol ogy Saunders Co., Philadelphia.

PAGE 159

152 Wright, R. F. 1982. Norwegian models for surface water chemistry: an overview. Presented at Symp. Acid Precip., Am. Chem. Soc. Ann. Meeting, Las Vegas, Nev. Wright, R. F., and M. Johannessen. 1980. Input-output budgets at gauged catchments in Norway. J_n: Drablos, D., and A. Tol Ian (eds.), Proc. Int. Conf. Ecol. Impact Acid Precip. SNSF project, Oslo, Norway. Yan, N. D., and P. Stokes. 1978. Phytopl ankton of an acidic lake, and its response to experimental alterations in pH. Env. Conservation 5: 93-100. Yuan, C. K., N. Gammon, Jr., and R. G. Leighty. 1967. Relative contribution of organic and clay fractions to cation-exchange capacity of sandy soils. Soi 1 Science 104: 123-128. Zinder, S. H., and T. D. Brock. 1978. Microbial transformations of sulfur in the environment. Ijn: Nriagu, J. 0. (ed.), Sul fur in the Environment, Part II: Ecological Impacts John Wiley and Sons, New York.

PAGE 160

APPENDIX: McCLOUD LAKE HYDROLOGIC AND CHEMICAL DATA Table A-1. McCloud Lake water budget, September, 1981, to August, 1982. Month PrecipiEvapoSeepage Volume Change in Area Change in tation^ 10 ^ m"* ration 10^ m-^ 10^ m3 3 3 10-^ m-^ volume 10-^ m-^ 3 2 10-^ m^ area ^ 10-^ m^ Sept., 1981 1.66 5.01 -2.89 126.01 -2.36 50.30 -0.35 Oct. 1.50 4.45 -3.05 123.50 -2.51 49.97 -0.33 Nov. 4.89 2.75 -1.42 123.21 -0.29 49.90 -0.07 Dec. 1.28 2.05 -2.14 120.50 -2.71 49.35 -0.55 Jan., 1982 9.24 2.85 -1.55 121.26 0.76 49.70 -0.35 Feb. 4.84 2.85 -0.94 121.26 0.00 49.35 -0.35 March 5.50 4.14 -1.31 117.45 -3.81 48.70 -0.65 April 9.94 5.06 2.12 127.75 10.30 51.23 2.53 May 7.90 6.34 2.31 130.44 2.69 50.97 -0.26 June 12.23 6.35 1.09 134.60 4.15 51.60 0.63 July 7.38 6.28 0.39 143.88 9.28 53.10 1.50 August 8.93 5.73 2.66 148.85 4.97 53.77 0.67 153

PAGE 161

Table A-2. Major ions in McCloud Lake, October, 1980, to August, 1982. All values in ueq/L Month^ Mg2+ Na^ S04^CI October, 1980 112 66 7.4 126 154 November 24.5 118 66 6.9 121 Ill 150 December 27.0 102 66 9.7 128 — 163 January, 1981 16.7 58 64 6.5 120 — Febuary 78.5 79 66 5.4 114 — March 27.8 47 65 5.6 116 133 — Apri 1 30.9 48 62 6.4 126 138 — May 30.8 64 62 9.0 138 152 June 28.2 36 66 8.5 136 137 July 29.3 197 August 27.5 167 September 30.4 64 63 6.4 189 184 167 October 35.2 52 66 6.8 199 231 170 November 38.1 58 63 5.7 179 198 185 December 39.6 50 69 6.5 229 214 189 January, 1982 43.8 48 69 5.2 188 206 174 Febuary 33.2 48 66 4.2 181 207 174 March 28.5 61 65 4.1 166 191 159 April 30.5 53 65 4.7 170 175 155 May 35.6 62 66 4.8 163 210 205 June 25.2 58 68 7.1 147 155 177 July 23.0 56 67 6.9 137 138 137 August 25.1 51 67 7.0 139 159 173 ^Values represent means of all depths.

PAGE 162

155 Table A-3. Physical data and nutrients in McCloud Lake, October, 1980 to August, 1982. Month^ Conducti vity, uS/cm @ 20 C nutrient data in mg/L Temp., TKN NH, NO3" Total Orthophosphorus phosphorus Oct., 1980 0.532 0.130 0.031 0.007 0.007 Nov. 52 17.1 0.144 0.021 0.107 0.007 0.006 Dec. 0.238 0.010 0.039 0.007 0.007 Jan. 1981 53 0.160 0.009 0.054 0.007 Feb. 18.0 0.183 0.053 0.068 0.009 0.003 March 56 21.8 0.502 0.030 0.010 0.007 April 25.1 0.386 0.016 0.043 0.018 0.005 May 69 27.1 0.480 0.099 0.039 0.007 0.005 June 30.2 0.351 0.067 0.037 0.007 0.002 July 0.326 0.040 0.009 0.004 August 0.489 0.137 0.044 0.010 0.003 Sept. 50 27.0 0.409 0.135 0.044 0.032 0.003 Oct. 24.0 0.278 0.119 0.055 0.011 0.003 Nov. 18.0 0.537 0.107 0.108 0.011 0.001 Dec. 44 15.0 0.442 0.063 0.118 0.005 0.001 Jan. 1982 47 17.0 0.461 0.020 0.145 0.007 0.001 Feb. 21.0 0.140 0.011 0.143 0.001 0.001 March 43 25.0 0.564 0.054 0.058 0.017 0.003 April 41 25.0 0.865 0.098 0.080 0.008 0.001 May 37 24.0 0.637 0.084 0.063 0.031 0.010 June 43 29.0 0.619 0.063 0.051 0.018 0.003 July 42 31.0 0.780 0.019 0.015 0,023 0.004 August 40 32.0 0.251 0.023 0.007 0.011 0.002 ^Values represent means of all depths.

PAGE 163

156 Table A-4. Chemistry of McCloud Lake wet-only precipitation. May, 1981, to September, 1982. All chemical data in ueq/L Na"^ SO^^ci Month Precip.H"^ 0-i_ Ca2+ Mg 2+ cm/mo May, '81 2.1 28.6 4.5 1.6 0.90 June 17.3 38.2 5.0 4.8 0.50 July 9.0 64.4 8.5 3.0 0.90 August 12.3 74.8 10.0 4.2 0.30 Sept. 3.3 32.4 14.2 4.9 0.70 Oct. 3.0 35.5 9.0 7.7 0.80 Nov. 9.8 23.4 2.5 3.6 0.40 Dec. 2.6 24.5 5.0 4.0 0.90 Jan., '82 : 18.6 22.4 3.5 2.1 1.40 Feb. 9.8 20.9 4.0 3.5 1.00 March 11.3 24.0 2.5 2.7 0.50 April 19.4 30.2 7.0 4.5 0.70 May 15.5 33.9 1.1 2.6 0.20 June 23.7 17.0 2.0 1.5 0.40 July 13.9 33.7 3.6 2.2 1.30 August 16.6 35.7 3.3 2.4 0.40 Sept. 27.6 44.0 5.6 2.6 0.70 4.6 15.7 15.1 11.2 5.3 31.9 15.7 13.1 12.7 23.1 12.7 10.0 7.0 5.2 10.9 7.9 9.2 23.7 32.5 79.4 63.8 61.2 43.3 25.8 28.3 24.0 20.0 26.7 43.6 24.8 15.6 28.5 28.1 35.5 7.5 18.9 20.1 13.5 6.3 36.7 17.5 14.9 15.5 29.6 15.5 11.0 6.5 4.5 11.0 8.5 9.9 NO311.9 22.5 34.0 29.4 19.6 22.1 8.6 8.6 6.4 10.0 5.0 6.4 10.5 6.4 15.5 16.6 22.3 NH4^ 13.3 16.0 27.7 13.6 18.1 7.1 4.3 6.4 5.0 5.7 4.3 10.7 2.8 2.4 4.9 0.9 5.6

PAGE 164

157 Table A-5. Chemical composition of in-seepage to McCloud Lake, April, 1982, to September, 1982. Month^ Ca2^ Mg2-^ Na+ 504^ cr NO3NH4"' April, '82 0.40 94 47 32 23 31 38 8.7 1.3 May 1.66 72 47 23 11 25 31 2.5 6.4 June 0.71 123 67 43 18 41 39 3.2 1.8 July 1.30 48 34 19 22 18 26 4.6 4.2 August 0.88 93 54 53 27 52 57 2.1 2.9 Sept. 1.55 98 42 54 27 58 56 3.1 2.0 ^ Values represent means for all seepage stations.

PAGE 165

158 Table A-6. Ambient air concentrations at McCloud Lake, AugustSeptember, 1982. A]] v*^ 1 lip^ i n V u 1 U CO III uy/ III Date SO/i^8/13 3.46 0 5f)9 n ??7 N/ A N/A 11/ n 8/15 1.27 0.154 0.186 N/A N/A 11/ n 8/17 2.56 0.471 0.343 1 59 8/19 1.24 0.194 0.175 1 4? 7 4? 8/20 2.07 0.410 0.203 1.13 5.14 8/22 1.51 0.297 0.150 2.71 5.14 8/25 3.99 0.102 1.511 2.58 10.28 8/26 1.91 0.075 0.632 2.97 1.71 8/28 6.87 1.816 0.191 3.57 4.00 8/29 1.90 0.148 0.639 1.52 4.57 8/31 3.28 0.192 1.341 4.34 4.57 9/1 3.09 0.248 0.686 1.55 10.28 9/10 2.03 0.204 0.764 4.57 13.70 9/12 3.57 0.671 0.309 7.02 5.71 9/15 1.91 0.248 0.136 2.10 2.86 9/17 2.06 0.258 0.272 2.93 6.28 9/18 2.16 0.552 0.215 3.62 9.71 9/20 4.12 0.672 1.003 8.78 3.43 9/21 3.20 0.430 0.216 0.79 1.71 9/22 2.30 0.179 0.302 1.12 7.42

PAGE 166

BIOGRAPHICAL SKETCH Lawrence Alan Baker was born on March 30, 1951 in Louisvil le, Kentucky. He was rai sed in New Marti ns v i 1 1 e, West Virginia, on the banks of the Ohio River. After graduating from Penn State University with a B.S. in biol ogy in 1 973, he worked briefly for the Pol 1 ut ion Control Division of the City of Houston, Texas, Health Department, then returned to graduate school where he first studied biology, and then environmental engineering. In 1981, he received a M.S. in environmental engineering from Utah State University and began doctoral studies at the University of Florida. Soon after completing his Ph.D., he will marry Nancy Ann Rodenborg and will begin working as a post-doctoral fellow at the University of Minnesota. 159

PAGE 167

I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. Jj^. Heaney, Chairifian •rofessor of Environmental Engineering Sciences I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. Professor of Environmental Engineering Sciences I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. G. R. Best Research Scientist of Environmental Engineering Sciences I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. Prdffessor of Environmental Engineering Sciences

PAGE 168

I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate in scope and quality, as a dissertation for the degree of Doctor of Philosophy. Professor of Soil Science This dissertation was submitted to the Graduate Faculty of the College of Engineering and to the Graduate School, and was accepted as partial fulfillment of the requirements for the degree of Doctor of Philosophy. April 1984 Dean, College of Engineering Dean for Graduate Studies and Research


xml version 1.0 encoding UTF-8
REPORT xmlns http:www.fcla.edudlsmddaitss xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.fcla.edudlsmddaitssdaitssReport.xsd
INGEST IEID EAVDGX6AU_YXXEUY INGEST_TIME 2014-12-23T19:15:30Z PACKAGE AA00026655_00001
AGREEMENT_INFO ACCOUNT UF PROJECT UFDC
FILES