THESlB ’“ LIBRARY ! i' ' Michigan Sm: Uxivem‘ty This is to certify that the thesis entitled ARSENIC PROFILES IN SEDIMENTS AND SEDIMENTATION PROCESSES ALONG THE SLOPE OF A LAKE BASIN presented by Mehdi Siami has been accepted towards fulfillment of the requirements for PhD . Fisheries & degree 1}) Wildlife Q.®. “Mo/”Jig“ Major professor Date 20 July, 1981 _ . .. 1 ulllllllllll/llljlllIllllljllllllllllfllfllll '" MSU LIBRARIES _:—. RETURNING MATERIALS: Place in book drop to remove this checkout from your record. FINES will be charged if 550k is returned after the date stamped below. ARSENIC PROFILES IN SEDIMENTS AND SEDIMENTATION PROCESSES ALONG THE SLOPE OF A LAKE BASIN BY Mehdi Siami A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Fisheries and Wildlife 1981 ABSTRACT ARSENIC PROFILES IN SEDIMENTS AND SEDIMENTATION PROCESSES ALONG THE SLOPE OF A LAKE BASIN BY Mehdi Siami Lake Lansing, Michigan was treated with sodium arsenite for macrophyte control in 1957. Seven 1.5 m sediment cores taken on a line through the littoral zone to the deepest portion of the basin were analyzed for arsenic in 5 cm incre- ments. The objectives were to determine rates that sediment surfaces at different depths were returning to pre-treatment concentrations and to evaluate sedimentation processes affect- ing those rates. Arsenic concentrations going downward from the surface in each core increased to some maximum. Below the maximum, there was a recession to background concentrations. Depth of peak concentrations followed two patterns; three littoral cores showed peak arsenic at 0.13 m from the sediment surface; four cores from progressively deeper regions of the lake showed a regular decrease in peak depth from 0.32 m to 0.17 m. Magnitude of peak arsenic in each core increased with depth of Water from which the core was taken. This suggested that 1957 treatment arsenic quantitatively precipitated to Mehdi Siami the sediments as a function of depth of overlying water. Sediment accumulation rates were calculated. They were low in the littoral, highest at 3.75 m, and decreased going into deeper water. Particle—size sorting of sediments along the basin's slope was measured. This work suggested that sediments originated from wetland vegetation at the edge of the lake. Turbulent movement of water in the shallows caused suspension and down-slope movement of small particles. Fewer particles of wetland origin were available for sedimentation beyond the region of highest fallout (3.75 m), thus accounting for progressively lower sedimentation rates in deeper portions of the basin. In each sediment profile, there was a decline in arsenic from peak concentration to the 1980 sediment surface. Expo- nential curves were fit to these data. From them, the littoral sediment surface was predicted to reach pre-treatment concentration >100 years after treatment. Using this model, the pelagial sediment surface would return to background in 28 to 43 years. The latter rates are unrealistic; the rate of approach of deep sediments to background will be limited by the rate of approach of shallow sediments to pre-treatment arsenic concentrations. For Lake Lansing, that prediction is >100 years. DEDICATION To my wife, Lili ii ACKNOWLEDGMENTS I would like to thank Professor Clarence D. McNab who advised and encouraged me and provided invaluable assistance and time throughout this study. I wish to thank Drs. Nile R. Kevern, Bernard D. Knezek, and Peter G. Murphy, the other members of my committee. I would like to express my appreciation to my friend Robert P. Glandon, who spent many long hours discussing impor- tant aspects of the work and helping me edit the manuscript. Special thanks are due Ted R. Batterson who established in our laboratory the procedures used here for sampling, sample preparation, and arsenic analysis. His help with coring was appreciated. The talents of John R. Craig, Limno- logical Research Laboratory director, were of great value in design and construction of figures. Thanks go to my friends, George W. Knoecklein, Frederick C. Payne, and Bette J. Premo for providing encouragement and opportunities to complete this study. Finally, my deepest appreciation goes to my wife, Lili, and children, Saghar and Farshid, for their patience and encouragement throughout the course of study. This study was supported by funds provided by the U.S. Environmental Protection Agency, Clean Lakes Program, under Grant No. R80504601 and the Michigan Agricultural Experiment Station at Michigan State University. iii TABLE OF CONTENTS Page LIST OF TABLES .........................¢........... vi LIST OF FIGURES .................................... vii INTRODUCTION ........ ...... ... ..... ............ ..... 1 MATERIALS AND METHODS ............ ..... ... ..... ..... 6 RESULTS .. ....... .................. ...... .... ....... 20 DISCUSSION ......................................... 30 APPENDIX .............. .............. . .............. 46 LITERATURE CITED ................................... 56 iv Table A-3. LIST OF TABLES Calculated time for arsenic to reach background concentrations in surface sediments along the slope of the south basin of Lake Lansing ....... Organic and inorganic content of particulate and dissolved components, and density of surficial sediments along line AB of this study .......... Mean and one standard error of three replicates of percent organic matter in sediment cores from Lake LanSing along line AB 0 O O O O O O O O O O O O O O O O O O O 0 Concentrations of total As (ug As g.1 dry weight) with depth in sediment cores taken along line AB of the south basin of Lake Lansing ............. Concentrations of total As (ug As 9"1 dry weight) with depth in sediment cores taken along transect l of the south basin of Lake Lansing .. Concentrations of arsenic in macrophytes collected along line AB; July 1, 1980 .......... Concentrations of arsenic in macrophytes collected along transect 1; July 1, 1980 ....... The slope of the south basin of Lake Lansing ... Page 44 49 50 51 52 S3 54 LIST OF FIGURES Figure l. The Lake Lansing basin showing areas treated with sodium arsenite in 1957 (stippled), ug As 9’ of dry surficial sediments (from Batterson, 1980), and the position of the sampling transect (AB) used in this study ........................ Depth-volume curve for Lake Lansing, with tabled volumes for strata of the two deep holes and the lake as a whole ................ ........... ..... Temperatures (Co) in the south basin of Lake Lansing during 1978 ............................ Dissolved oxygen concentrations (mg 1-1) in the south basin of Lake Lansing during 1978 . ....... Shape of the south basin of Lake Lansing along line AB based on measurements of depth of water at metered distances from the edge of the lake. Bars indicate location of sediment core sampling .................................. ..... Concentrations of total-arsenic with depth in sediment cores taken along AB of the south basin of Lake Lansing .......... ... ................... Mean of percent dry weight contribution of each of four particles sizes constituting the surficial sediment of the south basin along line AB ..... .......... .......... ............... Regression of maximum arsenic concentrations found in cores from Lake Lansing on depth of the water column at coring stations corrected for sediment accumulation since 1957 ............... Sedimentation rates at the points of sampling along the slope of the Lake Lansing basin ...... vii Page 10 12 14 22 25 29 32 39 Figure Page 10. Expected change in arsenic concentration of surficial sediments as a function of the difference in arsenic concentration between sedimenting and base materials. Curves reflect extent of influence of mixing newly sedimented material with base sediments. Dashed line represents maximum influence of newly sedimented material on surficial sediment. Ratios are for new sediment: base sediment mixing ............ 41 viii INTRODUCTION Inorganic and organic compounds of arsenic exist as natural components of terrestrial systems and are detectable in nearly all soils (Peoples, 1975; Walsh and Keeney, 1972). Arsenic is naturally distributed in high levels in rocks and minerals that contain iron, sulfur, and manganese where the element can be concentrated up to 2,000 ppm (Fleischer, 1973). The arsenic content of soils is generally much lower, averaging 6 ppm and ranging from 1 to 40 ppm (Bowen, 1966; Vallee et a1., 1960). Where higher soil arsenic levels are found the source can usually be traced to anthropogenic acti- vities. Mining, disposal of industrial wastes and widespread use of arsenical pesticides can elevate arsenic to concen- trations several fold natural levels (Bishop and Chisholm, 1962; Vallee et a1., 1960). Natural concentrations of arsenic in marine waters are usually low, ranging from 0.15 l to 6.0 ug As 1'1 and averaging near 2 ug As 1- (Woolson, 1975). The highest concentrations in inland waters have been found in hot springs, like those in Nevada and Wyoming, where two examples showed arsenic levels of 2,300 and 500 ug l 1 As 1' (Hem, 1959). Ritchie (1961) reported 8,500 mg As 1‘ in a New Zealand hot spring. A survey of other fresh water systems in the U.S. revealed concentrations ranging from 10 1 ug As 1- to 140 ug As 1-1, with 76% of the 726 water samples 1 2 analyzed falling below 10 ug As 1-1 (Durum et a1., 1974). A survey of arsenic content of U.S. lakes placed 94% of the 1577 lakes sampled between 10 and 340 ug As 1-1 , with an average value of 60 ug As 1.1 (Kopp and Kroner, 1967). As with soil, high arsenic levels in lake water can often be attributed to human impact. Release of industrial and domes- tic waste, burning of fossil fuels, and application of arsen- ical pesticides are principal causes of artificially elevated arsenic in lakes and streams (Shapiro, 1971; Lis and Hopke, 1973; Aston et a1., 1975; Domogalla, 1926). Of particular interest here is the large scale use over the last several decades of sodium arsenite as an herbicide to control aquatic vegetation (Kobayashi and Lee, 1978; Ferguson and Gavis, 1972; Ruppert et a1., 1974; Mackenthun, 1950). This activity, coupled with the ensuing potential of acute, or more likely, chronic toxicity of arsenic to humans and non-target aquatic organisms, has generated interest in the fate of applied arsenic in lake systems (Bails and Ball, 1966; Cowell, 1965; Gilderhus, 1966; Crosby, 1966). Studies of the movement of arsenic in lakes or ponds treated with the herbicide sodium arsenite (Na2AsOZ) show that aqueous levels decrease within a period of weeks after application. The mechanism of this decline has generally been attributed to coprecipitation of arsenic from the water column with iron oxides, followed by sorption of the iron- arsenic complex by sediment particles (Crecelius, 1975; Sey- del, 1972; Sohacki, 1968; Mackenthun, 1964). This mechanism 3 is supported by observed increases of iron-associated arsenic in surficial sediments (Crecelius, 1975; Kanamori, 1965; Kobayashi and Lee, 1978). Because of the complex chem- istry of arsenic and unknown patterns of lake sedimentation, the fate of sediment arsenic has not been fully described. Most of the work toward this end has involved analysis of arsenic content with depth in single sediment cores, or in several cores taken from widely separated locations in a lake (Crecelius, 1975; Kobayashi and Lee, 1978). Arsenic profiles in single cores have been used to reflect the timing of arsenic loading to the sediments (e.g., Crecelius, 1975), and to suggest the change in potential of sediment arsenic to be recycled into overlying water. However, interpretation of results from a single or widely separated cores is limited in that it cannot be extended to develop a model for sediment arsenic distribution in the basin as a whole. In this study, a series of cores was taken along a line running from a wetland fringe,on Lake Lansing, Michigan, across the littoral zone of the lake, and down the pelagial slope to the deep plain of the lake. Batterson (1980) has shown that the vertical profile of arsenic in cores from the deep basin have an arsenic peak related to a single sodium arsenite treatment applied in 1957 for macrophyte con- trol. The Lake and Stream Improvement Section of the Michi- gan Department of Conservation treated areas with a total of 3800 liters of sodium arsenite in June of that year (Roelofs, 1958). This treatment resulted in an input of 2920 kilograms 4 of arsenic. Historical records indicate that this has been the sole arsenic treatment of the lake. Batterson (Ibid.) showed arsenic loading to the lake from atmospheric fallout and overland flow was negligible. He also demonstrated that the arsenic content of the upper 5 cm of sediments was 2 to 6 times pre-treatment concentrations found in deep portions of cores. For example, the arsenic content of pre-treatment sediments in the south basin was in the range of 17 to 20 ug g"1 dry weight; 46 pg As g.1 were found in surficial sedi- ments near shore and 125 ug As g-l just beyond 7 m contour (cf. Figure 1). The significance of high sediment arsenic levels stems from the potential of sediment to contribute soluble arsenic to overlying water. This was suggested when in 1978 the arsenic of the lower pelagial water of the south basin increased from 14 to 115 ug As 1.1 during a period of summer stratification. In his cores from the deep plain of the lake, Batterson (Ibid.) observed a recession in the arsenic con- centration from 1957 peak with the addition of recent sedi- ments to the lake bottom, and calculated the rate at which sediments of the pelagial plain were returning to the back- ground concentration. Since sedimentation rates and sediment mixing processes are likely different at different depths along the slope of the lake basin, he could not predict from his data the time necessary for sedimentation to ameliorate the effects of the 1957 treatment in the basin as a whole. The purposes of this study were: (1) determine the nature of 5 the arsenic profiles in littoral sediments and sediments of the pelagial slope, (2) to use these profiles to determine sedimentation rates and the degrees of mixing of new sedi- ments with base sediments at different depths in the lake, and (3) to use the profiles to predict the rates at which sediments at different depths would approach background, thus bringing sediment surfaces to pre-treatment arsenic concentra- tions. MATERIALS AND METHODS Lake Lansing is located approximately 5.6 kilometers northeast of the city of East Lansing, Michigan. The lake has a surface area of 1816 x 103 m2, mean depth of 2.3 m, and a maximum depth of 10 m. The littoral zone of the lake extends to the 3 m contour; 77% of the lake surface area lies over the littoral zone (Figure 1). The volume of the lake is approximately 4,124 x 103 m3 (Figure 2). The bathy- metry of Lake Lansing shows it is divided by a shallow bar into a north and a south basin. Each basin has a tendency to thermally stratify in the summer (Figure 3) and develop anoxic conditions in the lower pelagial regions. The ten- dency for oxygen loss is particularly evident in the south basin (Figure 4). The slope of the lake basin along transect line AB shown in Figure l was determined by gauging the depth of water at measured distances from the shore. This was accom- plished by lowering a plumb through augered holes in the ice cover. The plumb weight consisted of a 25 cm diameter disc to minimize error caused by sinking into the soft sediments. Seven cores were obtained along line AB during July to August, 1980. Since the sediments along the transect were loose and unconsolidated, they were sampled by freezing the sediment onto the exterior surface of tubing which extended 6 Figure l. The Lake Lansing basin showing areas treated with sodium arsenite in 1957 (stippled), ug As 9 of dry surficial sediments (from Batterson, 1980), and the position of the sampling transect (AB) used in this study. Figure 2. Depth-volume curve for Lake Lansing, with tabled volumes for strata of the two deep holes and the lake as a whole. 210 COO? .ww _¢. nXmN 0mm” owi max. n2u0_ 0304 22.3 .VN mm: mm; Imam ezoom £32 Nm” we. canon g.=om Amie. x was oE:_o> _ q _ d OOON - $7.0. x mEv OED—O) Eastm usdao (w) 11 Figure 3. Temperatures (Co) in the south basin of Lake Lansing during 1978. 12 mmm2m>oz mumOeoo mwmzuhdmm .5384 >42. .uza. >42 (‘1‘) H1d30 13 Figure 4. Dissolved oxygen concentrations (mg 1-1) in the south basin of Lake Lansing during 1978. mwm2w>oz mmmohbo 1329—.me Ema? >42. mzas >42 (W) H1d30 15 from the water surface and penetrated the sediments a known distance. Lengths of 5.08 cm o.d. thin-walled aluminum electrical conduit were used which were threaded and joined by couplings. Added to the water depth at each coring site was the length of the sediment core desired. Sections of tubing were then selected which would exceed that length by several feet to provide excess tubing above the water. The joints were not water-tight so silicone sealant was applied to the threads to accomplish this. The bottom of the sam- pling tube was stoppered and lowered into the water. Addi- tional lengths were added until the stoppered end was just above the sediment surface. When the last section was attached, the tube was carefully pushed into the sediments to the appropriate depth. After insertion into the sediments, pelletized dry ice was added to the end of the tube extend- ing above the water surface. The amount added was enough to freeze the sediments as well as a small portion of water above the sediment-water interface. Replenishment of dry ice was maintained at a rate to offset sublimation. Thirty minutes after the initial addition of dry ice, the samples were retrieved. As the tube was pulled out of the water the sections were uncoupled down to the frozen sample. After the sediment sample was removed from the lake, the unfrozen exterior layer was stripped away. The sample was then placed in plastic and the tube repacked with dry ice for transportation to the laboratory. In the laboratory, the dry ice was removed from the tube and replaced with tap 16 water. This melted the sediment in contact with the tubing and allowed for the tube to be pulled free. The frozen sample was then cut into 5 cm sections using an electric band saw. The exterior of each doughnut-shaped piece was rinsed with ion-free water and placed in a labeled plastic bag. There were two reasons for the rinsing: to wash away any contamination that might have resulted from the sectioning process or sediment contact with the aluminum tube, and to remove dislocated particles from the core surfaces. The frozen samples were then dried in a Napco model 630 forced air drying oven at 75°C for 72 hours and dry samples were ground with mortar and pestle. From each of the well-mixed ground samples approximately one gram of sediment was removed and dried at 105°C for 24 hours. The sample was then introduced into an acid-washed and pre- weighed two dram polyvial and weighed. After weighing, the polyvials were heat-sealed and taken to Michigan State Uni- versity's nuclear reactor facility for neutron activation analyses. For each group of samples that was irradiated there were three standards for quantifying the analyses. Two of the standards were obtained from the Natural Bureau of Standards and prepared for introduction to the polyvials ac- cording to the procedure recommended by the Bureau. These were Standard Reference Material 1645 (River sediment) and 1571 (Orchard leaves). The other standard was a 2 ml solu- tion containing 150 ug As ml-l. A Triga Mark I nuclear reactor was used for irradiation. l7 Thirty-seven sediment samples and three standards were intro- duced into a 40-position specimen rack that was rotated dur- ing irradiation to establish uniform flux for all sample positions. A flux rate of 1012 neutrons cm“2 sec was used. Sixteen to twenty hours following irradiation (allowing for 24 the partial decay of Na activity), the samples were counted for 1000 seconds live-time with a 76.2 cm3 active volume Ge (Li) detector having a relative efficiency of 15% and an energy resolution of 1.8 Kev FWHM at the 1.333 MeV photopeak of 60Co. The source-to-detector geometry was kept constant for all counts and the detector resolution was sufficient to 76 completely resolve the As peak (559 KeV) and the adjacent peak of 82 Br (554 KeV). The gamma-ray spectrum from each sample and standard was analyzed by a Canberra Series 80 multi-channel analyzer. This analyzer computed the peak net area which is the number of counts in a peak that are above an average background level. Standards and samples were corrected for decay during counting by the following equation: 0.693t l Ac=Ae Area corrected for decay between counting time of the sample and standards (net count) 76 where Ac A = Area of As (net count) Base of the natural logarithms = Half-life of 76As = 1584 minutes (I) ll 11’ t = Finishing time in minutes The mass of arsenic in the sample was derived using the time corrected counts of the standards. 18 Estimates were made of the mass of dry sediments in each 5 cm section of core. The volume of cores was obtained by determining the cross-sectional area of frozen sediments plus the sampling pipe. The cross-sectional area of the pipe was subtracted from the total area and the remainder was multiplied by the length of the core section (5 cm). The density of sediments in core sections was measured by a water displacement method. A known volume of ion-free water was added to a graduated cylinder. Longitudinal sections of frozen core material were placed in the cylinder. When the frozen piece thawed, volume in the cylinder was recorded. The volume of the piece of the core was calculated by sub- tracting the initial volume in the cylinder from the final volume. Contents of the cylinder were rinsed with ion-free water into a pre-dried and weighed aluminum tray and dried in an oven at 95°C to a constant weight. The weight of the dried material was divided by the calculated volume of the core fragment to obtain an estimate of the density of the 3 in the fragment multi- sediments in the fragment; 9 DW cm‘ plied by the volume of 5 cm core sections yielded the total dry weight in core sections. Additional analyses of the physical features of cores were made during this study. The results of this work are included here as Appendix Table 1. In February of 1981, surficial sediment samples were collected along line AB for particle size fractionation. Samples were collected with an Ekman dredge at points corres- ponding to those from which cores had been taken. Each 19 sample was mixed thoroughly and three 200 ml volumes were withdrawn. These replicates were dried to a constant weight at 95°C. In addition, three 200 ml subsamples from each dredge were rinsed through a stack of U.S. Standard Sieves; numbers 10, 20, and 50 (pore sizes 2, 0.833, and 0.227 mm, respectively) were used. A measured volume of tap water was used for the rinse. Particles retained by each sieve were emptied into pre-weighed aluminum trays and dried at 95°C to constant weight. The weight of particles passing through the smallest sieve (0.227 mm) was calculated by subtracting the sum of the dry weights of the larger size classes from the dry weight of whole samples. The data obtained were used to calculate the percent dry weight contribution of each particle size fraction in the surficial sediments along the slope of the lake basin. RESULTS The profile of the south basin along transect AB showed two zones with distinct gradients (Figure 5). Within the first zone extending from shore to approximately 105 m lake- ward, the basin gradually declined to 2 m below the lake surface. The slope of the basin in this region was approx- imately 1:50. Within the second zone, which extended from the edge of this shallow shelf to a point about 50 m lake- ward, the basin dropped 5 m; the slope increased nearly five- fold. The basin profile suggests that there was an extensive shallow region where the sediment surface was subject to wind-generated water movement that could resuspend previously sedimented materials. Since aquatic plant cover tends to stabilize the sediments, resuspension processes may be most intense during periods when submersed plant biomass is low, as during spring overturn. However, potential for contact between these sediments and moving water is higher than in the deep region of the basin. The vertical bars of Figure 5 indicate the position and depths of cores taken in this study. Core sampling was concentrated in the portion of the basin where slope was greatest. One would expect sedimenta- tion and mixing processes to change most rapidly in that region of the lake where influence of water movement changes rapidly (Hutchinson, 1957). 20 21 Figure 5. Shape of the south basin of Lake Lansing along line AB based on measurements of depth of water at metered distances from the edge of the lake. Bars indicate location of sediment core sampling. 22 ON N CON oo— AEV m< «000:9; 0:20 0.52» 50.; 00:03.0 oo. 0! ON. oo. 00 ow 0* ON O NMVPID‘DNOO) ( w ) mdop mom 23 The concentration of total arsenic with depth in each sediment core taken along the slope of the south basin is present in Figure 6. It is obvious from a comparison of the curves that the total amount of arsenic was different in cores from different depths in the lake; for example, cores from the shallows of the lake contained less arsenic than cores from the pelagial region. Arsenic concentrations going downward in each core increased to some maximum point. The maximum was followed by a recession to background levels in deep portions of the cores. The position of the arsenic trace with respect to the abscissa, and the magnitude of peak arsenic in each core, increased with the depth of the water from which the core was taken. The depth of occurrence of peak concentrations followed two patterns in the series of cores. The three shallow cores showed peak arsenic at 0.13 m from the sediment surface. The four cores of the deeper regions of the basin showed a regu- lar decrease in depth of occurrence of peak arsenic concen- tration as a function of increasing water depth. Figure 6 shows that the rate of recession from the peak arsenic con- centration to the sediment surface increased with depth of water, or perhaps more significantly, with the magnitude of the arsenic peak. This has resulted in a convergence of the surficial arsenic concentrations in the five deepest cores to a range of 84-92 ug As g.1 dry weight. Arsenic profiles were obtained for cores taken along transect 1 in Figure l, as well as along line AB. However, stations on transect l 24 Figure 6. Concentrations of total-arsenic with depth in sediment cores taken along line AB of the south basin of Lake Lansing. 25 _ _ _ _ _ _ _ _ _ _ _ _ OmN OON O¢N ONN OON Om. Ow. OS ON_ 00. Om OO O? ON O 29o; be To 259.4 oi _ — ON. 00. ON ow O¢ ON O ( m ) l“(1190 26 were used during 1978 and 1979 in intensive sampling programs for water chemistry and zooplankton and benthic populations. The data from transect 1 is included here in Appendix Tables 3 and 4. In general, the depth profiles for arsenic along transect 1 showed patterns similar to those described for line AB. Portions of individual profiles appear to have been badly disturbed by sampling. Decreases in arsenic from peak concentrations toward the surface in each core suggests that newly deposited sedi- ments of relatively low arsenic concentration were burying the arsenic introduced in 1957. The low relief topography of the land surrounding the lake, the relatively small size of the drainage basin, the absence of appreciable stream flow to the lake, and the lack of eroding beaches (Batterson, 1980) argue for the position that inorganic soil materials do not contribute substantially to the buildup of sediments. They appear to accumulate from the breakdown of vegetation from residential shorelines and wetlands around the lake (Knoecklein, 1981) and from submersed plant remains. Wet- lands dominate the shoreline of the south basin of Lake Lansing (Figure l). The shoreward origin of the transect used in this study was located at the edge of a wetland. That in-shore sediments originated in the wetland was sug- gested by the common occurrence of macroscopic fragments of plant tissues in the shallows. Currents generated by wind action on the lake were expected to sort out particles on the sediment surface in relation to their size and the 27 velocity of the currents in a manner that is well known for streams (Wetzel, 1975). The percent dry weight contribution of each of four particle sizes constituting the surficial sediments along line AB are presented in Figure 7. Regular changes occurred in the largest and the smallest size categories. The largest particles (> 2000 u) were made up of fibrous fragments of the wetland vegetation. The largest size class made up 40.6% of the sediments by weight at 0.15 m water depth. This size category dropped to 1.3% at 3.75 m. At depths greater than 3.75 m the change in this size was less than 1%. In con- trast, the smallest particle size (< 227 u) made up 30.4% of the sediments at 0.15 m water depth, and increased to 75% at 2.75 m. This size class did not change appreciably beyond 2.75 m. The data of Figure 7 show that particles of the sizes measured were sorted by currents primarily in depths of 2.75 m and less; size distribution was essentially the same for depths of 3.75 m and more. There may have been significant differences in particle size distribution within the smallest size class (< 227 u) between the four deepest stations; if so they would not be evident with the techniques used here. 28 Figure 7. Mean of percent dry weight contribution of each of four particles sizes constituting the surficial sediment of the south basin along line AB. ) 2000 p. E g 2000 - 833 p. 833 - 227 p, 29 WW .\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\‘ ml 5.75 .\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\‘ WW :Ezfzfz ”I :1:?:? 5m — ......... ......... ooooooooo ......... nnnnnnnnn ......... ......... D e, ........................... nnnnnnnnnnnnnnnnnnnnnnnnnnnnnn ------------------------------- ooooooooooooooooooooooooooooooo ooooooooooooooooooooooooooooooo ooooooooooooooooooooooooooooooo ooooooooooooooooooooooooooooooo IOO - 90- 60- 50- 40- 3o_ 20— IO— 0 l l o o on Is iqblaM Kilp '010‘ *0 % 7.20 6.75 0.34 L75 2.75 3.75 4,75 0J5 Depth of water overlying the sediments (m) DISCUSSION The profiles of arsenic concentration in the cores taken along line AB in the south basin of Lake Lansing (Fig- ure 6), show that the peak concentrations increase as the depth of water overlying the sediments increased. This sug- gests that 1957 treatment arsenic quantitatively precipitated to the sediment surface as a function of depth of overlying water. That relationship is presented in Figure 8. The horizontal scale in the figure was obtained using estimated sedimentation rates at sampling points along the slope of the basin; these are discussed later in this section. Part of the scatter in this relationship may result from decreases in the magnitude of arsenic peaks since 1957 due, for example, to the transport of precipitated arsenic from the sediment surface downward in the sediment profile. Three conditions must be met for the relationship in Figure 8 to be accepted as valid. Arsenic sprayed over weed-beds at the time of lake treatment must have become well mixed in the volume of water in the lake before precipitation. Some mechanisms for the removal of arsenic from the water column were required. Once on the sediment surface, arsenic must have remained relatively immobile, thus allowing the maximum concentration in each core profile to represent the depth in the sediments of the 1957 sediment surface. 30 31 Figure 8. Regression of maximum arsenic concentrations found in cores from Lake Lansing on depth of the water column at coring stations corrected for sediment accumulation since 1957. Peak sediment Arsenic concentration 32 Y = 41.29x + 7.53 r2 = 0.95 I l l l i l J l 2 3 4 5 6 7 Estimated depth of water column in i957 (m) 33 Evidence for extensive mixing of arsenic in the lake prior to fallout on the sediments comes from the work of Batterson (1980). His core taken from the north deep basin of Lake Lansing had an arsenic profile and peak arsenic con- centrations remarkably similar to his core from the deep portion of the south basin. This was observed even though the distances from 1957 treatment areas to his coring areas were substantially different. Batterson's surficial sedi- ment data is given in Figure l of this paper. They show that arsenic concentrations in surface sediments along six tran- ects in the lake decreased from the shallows to deep water 1n the same manner observed along line AB. The prediction from the relationship presented here between peak arsenic concentration and depth of overlying water in 1957 (Figure 8) is that peak arsenic concentrations marking the time of treat- ment in 1957 occur beneath the surface sediments sampled by Batterson over most of the lake bottom. That Lake Lansing was likely well mixed after the arsenic treatment in July of 1957 is further suggested by the weak thermal stratification that exists in the lake in summer. An example of this is shown here in Figure 3. Vertical temperature differences occur only in the small volume of water over the deep holes in the lake. High south and southwest winds in summer tend to prevent the development of a stable metalimnion. The principal mechanisms involved in the loss of inor- ganic arsenic from the water column of lakes has received considerable attention in the literature. Ferguson and Gavis 34 (1972) suggest that arsenite, As (III), tends to be oxidized -to:arsenate, As (V), in aerobic water. Arsenite is most likely to exist as the anion HAsO42-. Chemically similar to phosphate, it can be absorbed, occluded or precipitated with hydrous ferric oxides. Kobayashi and Lee (1978) studied accumulation of arsenic in sediment of five Wisconsin lakes treated with sodium arsenite. They found a strong coefficient of correlation between arsenic and iron in the sediments of Lake Mendota. They concluded that iron controls arsenic levels in the water column through sorption of arsenate by ferric hydroxides, followed by precipitation to the sedi- ment. Crecelius (1975), studying the geochemical cycle of arsenic in Lake Washington, found a strong correlation be- tween sediment iron and arsenic (r2 = 0.94). He suggested that arsenic is associated with the iron phase which causes a major portion of arsenic to be removed from Lake Washing- ton water and accumulated in the sediments. Seydel (1972) studied the distribution and circulation of arsenic through water, organisms, and sediments of Lake Michigan. She sug- gested that accumulation of the arsenic in the sediments up to 28.8 ppm was due to the coprecipitation of arsenic with iron. In snaerobic water of a hypolimnion or in anaerobic sediments, arsenate tends to be reduced to arsenite (Ferguson and Gavis, 1972). Ferguson and Anderson (1974) reported that 2- at low Eh in the presence of sulfide (S ), arsenite should be effectively removed from the water column as insoluble 35 sulfides. The experiments of Batterson (1980) lead to the conclusion that iron controlled the solubility of inorganic arsenic in aerated freshwater systems, while sulfide con- trolled the solubility in anoxic systems. Because of these mechanisms, significant quantities of soluble arsenic are expected only where the redox status permits oxidized sulfur and reduced iron to exist simultaneously. Batterson (1980) showed that these conditions can occur in the hypolimnion of Lake Lansing; for example, arsenic increased from 14 to 115 ug 1.1 in deep water of the south basin in a two-week period in the summer of 1978. However, he showed the conditions were short-lived and were not typical. Similar elevations in hypolimnetic arsenic were not observed in the winter, spring or summer of 1979. This discussion argues for the position that arsenic, well mixed in the volume of Lake Lan- sing, would fall out on the sediment surface and tend to stay there as insoluble compounds of iron or sulfur. A question can arise as to the immobility of the arsenic peak deposited on the sediments as a result of the 1957 treatment. Carighan and Flett (1981) showed that phosphorus in lake sediments could migrate upward and accumulate near the mud-water interface. In spite of the similarity between arsenic and phosphorus chemistry, an important difference is that phosphorus does not combine with sulfide as arsenic does. Crecelius et a1. (1975) found that the concentration of total arsenic was high in the surface sediment of Puget Sound in Washington and dropped to background levels of 10 ppm with 36 the depth in the core. They suggested that high arsenic at surface sediment was the result of a recent additional input of arsenic from a large copper smelter. In the same study, sediment accumulation rates were determined by the lead-210 technique. They showed that the arsenic level started to increase in the cores at the time when the copper smelter started to operate. Crecelius (1975) also found that the position of peak concentrations of arsenic for five different locations in Lake Washington varied with sedimentation rate. In areas with lower sedimentation rates, peak concentration occurred at a shallower depth in sediment cores. Kobayashi and Lee (1978) studied accumulation of arsenic in sediments of lakes treated with sodium arsenite. Arsenic profiles were developed for cores from five lakes. They used sedimentation rate for eutrophic lakes in the study area from Bartleson (1970) and showed that the depth of peak con- centration corresponded to treatment time. In two lakes (Big Cedar and Pewaukee) with the same sedimentation rates, the difference in depth of peak concentration was due to time of treatment. From these considerations, it is con- cluded that arsenic deposited on Lake Lansing sediments fol- lowing treatment in 1957 has been relatively immobile. Sediment accumulation rates along line AB can be cal- culated using the depth of peak arsenic concentration in each core to represent the 1957 sediment surface. The density of core sections (g DW cm-3) was used with depth of the peaks to express sedimentation rate in units of g DW rn-2 yr-l; these 37 data are presented in Figure 9. Net sedimentation rates were low in shallow portions of the‘lake. The rate was highest at 3.75 m, and diminished from this maximum going into deeper water. Particle-size sorting of sediments along line AB has been demonstrated in this study (Figure 7). It is postu- lated that sediments originate largely from fragmenting vege- tation of the wetland at the edge of the lake. Wind-induced turbulent movement of water in the shallows causes suspension and down-slope movement, particularly of small particles. The region of highest sedimentation rate corresponds to the point along the slope of the basin where the mean turbulent energy of water is diminished rapidly with sudden increase in depth (Wetzel, 1975). Fewer numbers of particles of marsh origin are available for sedimentation beyond the region of highest fallout, thus accounting for progressively lower sedimentation rates in portions of the south basin deeper than 3.75 m. In each sediment profile presented in this study, there was a decline in arsenic from the peak concentration to the 1980 sediment surface. It is proposed that the rates of these declines are a function primarily of sedimentation rates, concentration of arsenic in sedimenting materials, concentration of arsenic in base sediments, and the degree to which sedimenting materials are mixed with base sediments. The relationship between these factors is expressed in Figure 10. An underlying assumption of this figure is that diffusion processes are not important in establishing observed arsenic 38 Figure 9. Sedimentation rates at the points of sampling along the slope of the Lake Lansing basin. 39 mNm mfim LEO 35663 33:26 .203 do 530 mks—V oh.» mNN 05. 05.0 d d L4 S 9 m" w o m N CO. H O u 08 1 w 000. 9 00¢ \I 6 000 P M oom M n. cos M. 00m m . com 2 oh 1 000. I ( Figure 10. 40 Expected change in arsenic concentration of surficial sediments as a function of the difference in arsenic concentration between sedimenting and base materials. Curves reflect extent of influence of mixing newly sedimented material with base sediments. Dashed line represents maximum influence of newly sedimented material on surficial sediment. Ratios are for new sediment: base sediment mixing. Expected change In surficial sediment [As] (no 9" I I50 I40 I30 l20 IIO 80 70 50 40 30 20 I0 41 " l l l J l l l l I l l l 1 l // I0:I ” / r— // 5:, / _ / / u / // 2:1 I / / #- ‘Gé/ — é / M s“ / i- 635/ s°/ / . P / 1.2 / ~— / / - / / I-‘5 _ / // -- MO 1 0 IO 20 30 40 50 60 70 80 90 l00 IIO I20 I30 I40 I50 ISO A [As] (m o" i 42 profiles. The horizontal axis of Figure 10 represents the differ- ence in arsenic concentration between sedimenting and base materials. This scale can be used independently of the arsenic concentrations of these materials. For example, if base sediment has 100 ug As 9"1 and sedimenting material 50, or base 1100 and sedimenting material 1050, or base 50 and sedimenting material 0, then all of these conditions are represented by the same point on the abscissa (A [As] 50 ug 9-1). To facilitate use of the figure, it is best to be con- sistent by subtracting sedimenting arsenic concentration from that of base material. Note that when sedimenting arsenic concentration is lower than that in base material, the expected change would be a negative value. The vertical axis marks the expected change in surficial arsenic concentration following sedimentation and mixing of the materials under consideration. This value is negative when sedimenting arsenic concentration is lower than base concentration, and positive when it is higher. This scale can also be used independent of the arsenic concentrations in the sediment materials. The curves of the figure represent the degree of mixing at the sediment surface under consideration. Mixing is viewed as a ratio; for example, when the degree of mixing of new sediments with base sediments is low, the ratio is high. In general, this situation is likely in lakes where turbulent flow is diminishing and suspended materials 43 fall out on a sediment surface that is not exposed to appreciable turbulence. As shown in Figure 10, the surface arsenic concentration is expected in this case to be heavily influenced by the concentration of arsenic in sedimenting materials. The family of curves given in Figure 10 illus- trates a range of cases. The shapes of recession curves in cores of this study from peak arsenic concentrations to the arsenic concentra- tions of 1980 sediment surfaces suggest that the relationships of Figure 10 were operative along line AB since 1957. These declines can be described as exponential decreases in arsenic concentration with distance from the depth of the peak. Using the core data, exponential coefficients were calculated. Employing the described function, which takes the general form y = aebx where y = [As] and x = depth, the estimated times for sediment surfaces to reach background were calcu- lated. The time element was obtained by using sedimentation rates estimated from the depths of peak concentration due to 1957 treatment. These data are presented in Table 1. The results in the last column of the table show that shallow sediments are expected to take a long period of time (> 100 yrs.) to reach pre-treatment background. The surface of these sediments, covered primarily by materials of fring- ing wetland and submersed macrophyte origin, could experience relatively high mixing due to the action of waves and wind- generated currents. Mixing arsenic-bearing surface sediments with newly sedimented materials of lower concentration slows <44 .umnu >3 pwospou ma wumefiumm amen menu mo zomusoon ecu “wo>uso Hmwucmcomxw omen» u0m mucofiowwmmoo :flmuno ou mHanfim>m wuw3 mucflom camp wows» >~co« .mmuoo mo mcofluuom moot CH unedumuucoocoo mo mcmwzH mm me.e .eo.o .em.me oee.o me He ~m.e me.e em em.o .mo.o .mm.mm mee.o me me he.m me.m oe mm.o .eo.e .e~.ee mmH.H OH He ~o.m me.e me Am.o .me.e .HH.- mae.H NH em so.e me.m eooax owm.o .mo.o .me.om mem.o an em em.~ me.~ .OOHA mmm.o .mo.o .e~.mm mem.o ex me em.a me.H .eoxx mmm.o .mo.o .me.~m mem.o m on em.o me.o mu» my .a .m Hiu> E0 Him 0: to m: E E ccsoumxomm exam H emmx omen ou xmmd mo>uso chucmcoaxm new :ONDMHDEsooc oflcmmu< owcmmud noun: umumz Eoum mEfle mucwfiowwuwou :oflmmmuowm ucoeflpmm Hondouoxomm mommusm mo nummo «G gamma _ .mcwmccq exam «0 canon fiasco on» no odon on» macaw murmepwm commusm cw mcofiunuucmocoo pcooumxomn semen on oficomum uOm wee» owumHSOch .H manna 45 the.process of burying 1957 arsenic. The data of Table 1 further suggest that sediments in the deep portion of the lake would return to background much faster than shallow sediments. As an explanation for this, it is proposed that sediments from shallow water with relatively low arsenic con- centrations have been a dominant source of new material for deep sediments since the time of treatment. If these were mixed poorly in deep water with the heavily contaminated base sediments there, the arsenic concentration of the deep sedi- ment surfaces would recede rapidly toward the level of the incoming materials. Poor mixing of surface sediments by tur- bulence is expected in deep portions of the lake. It must be noted that by this model, the rate of approach of deep sedi- ments to background arsenic concentration will be limited by the rate of approach of shallow sediments to background. Be- cause of this, it is proposed that the years to reach back- ground calculated for deep sediments and given in Table 1 are unrealistic. If shallow sediments are predicted to reach background in > 100 years, then on the assumptions of this dissertation, a similar length of time can be predicted for deep sediments as well. The data of this study, and the assumptions used to examine them, provide the framework for an experimental approach to answer a question of considerable ecological importance. The processes involved in burying contaminants in lake sediments, and the time required to accomplish this are not generally known. APPENDIX 46 47 During February of 1981,.cores were taken from transect AB from the same depth as those used for total arsenic analysis. The top 5 cm section was removed from each frozen core and cut longitudinally. Density of the surficial sedi- ments as well as bulk density of constituent particles was measured by a water displacement method. The results are given in Appendix Table 1. A known volume of distilled water was added to a grad- uated cylinder; weight of cylinder and water was measured and volume of water was recorded. Longitudinal sections were placed in the cylinder. When the frozen piece thawed, weight and volume of the cylinder and contents were measured and water temperature recorded. The weight of core section, consisting of particulate and dissolved solids plus core water, was obtained from the increase in weight. The volume of core section was calculated by subtracting the initial volume in the cylinder from the final volume. Contents of the cylinder were rinsed into a pre-dried and weighed alum- inum tray and dried in an oven at 95°C to a constant weight. The weight of dry material was divided by the calculated volume of core section to obtain the density of particulate and dissolved solids in the core section. The weight of the core water was calculated by subtracting the weight of dry solids from the weight of solids plus core water. The volume of core water was calculated from the weight of the water corrected for density at the temperature at which the weight was measured. To obtain the volume of solids, the 48 the volume of core water was subtracted from the volume of solids plus core water. The bulk density of solids was cal- culated by dividing solids weight by solids volume. Core total solids density and bulk total solids density were corrected for dissolved solids. To measure the dissolved solids component of core water, a frozen core fragment was thawed on a glass fiber filter and the water was drawn through its 0.5 pores. A measured volume of filtrate was placed in a pre-washed aluminum tray and dried in an oven at 95°C to a constant weight. The resultant dry weight of dissolved solids was divided by the filtrate volume to obtain the concentration of dissolved solids in the core water. The mass of dissolved solid was subtracted from estimates of the core total solids density and from the bulk total solids density. The percent-ash was determined in dissolved solids fractions of cores. After weighing, the contents of alum- inum trays containing dried core filtrate were combusted at 550°C for one hour (APHA, 1976). The weight of the residue was used to calculate ash-weight; weight loss on ignition was taken to represent organic material. 49 .Hmfluoums oumasowuumm >2 pwflmoooo AoEsHo> waofluumm pom mEsHo> umum3v mESHo> ucwswpmm unmoummm .wESHo> waowuumm Hecwmfluo uo E0 Mom unmflmz who mHOHuumdm m .N manna xwpcmmmm Eouw :mxmu mous>H ma.m ~N.~ «.mm av.m v.~ mnmn.~ tho.o mmao.o h.m~ onwo.o «who.o mn.w mn.m mm.H m.v~ mm.n m.~ mmon.~ vevo.o mmao.o m.mm vomo.o mmeo.o mn.m nv.o mm.o w.~v Hm.o m.m vvmv.m Hovo.o wmmo.o o.vm mmmo.o nono.o mh.¢ m¢.o mm.o o.ov om.o m.m Hmao.~ Hevo.o mamo.o m.mm mmmo.o Homo.o mh.m m~.o om.o m.vm mm.o m.m ommm.a mmmo.o oamo.o m.vm mome.o HHoo.o mn.~ ill: III: III: mm.o m.m «mom.a mmmo.o mmao.o m.~m mmmo.o mmmo.o mh.H Hm.o mm.o v.He mm.o h.m vnmv.H owmo.o onHo.o o.~m ommo.o mmm.o mn.o m~.o wH.o H.mm no.0 m.v mvnH.H vmao.o wmmo.o m.mo oamo.o mHmo.o ma.o Ammmwuwmv Ammwwuwmv uwwwcz Ammmwawwv mhwwo> Amwfimcww Ammflmmmv Ammwwmmv uwwwmz Ammmwmmv uhmmwm mwo SMMWD oficmmuocH oacmmuo owcmmuo awn m N . DaemouocH Uflcmmuo wwcmmuo wwo wmuos uwumz ZOHBUgm OM>AOmmHO "LOU—“804mm NBSDUHBmdm .wpsum own» no md mafia mcoHc mucoEflpmm Hmfloflmusm mo >uwmcop can .mucocomeoo po>H0mmflp can mDMHBOAHHma mo acovcoo owcomuocw one owcmmuo .Hic manna 50 .muoo: vm nouwm OOmOH um pocflmuno mos usmflm3 5H0 .unmflwS Owcmmuo mo mumefiumm an we coxmu mm: coflumsneoo >n ucowm3 mo mmoH “moo: H MOM 000mm ou wusmomxo an pmumsneoo mums monEmm .muoo mo cofluomm some Eoum :wxmu oum3 memEmminsm mouse H Nv.owm.mm IIIIIIIII hm.0Hw.mm HH.OHm.hm mv.owh.mv vv.OHN.Hv vH.oHH.mm mb.OHm.ov on me vo.owm.Hm va.onm.mv mm.onm.mm o~.owa.hm om.ow~.ov mv.owo.wm mo.oum.mm no.HHv.mm mv ov Ho.owh.mm om.oum.mm mm.on.Nm hH.0HN.hm mH.0HH.ov ve.owo.wm mm.owm.mm NH.owm.mm ow mm hH.cflv.mm ma.ouh.mm wH.owm.vm mm.owm.mm ma.owm.mm mv.oah.mm mm.0#m.mm mv.onm.vm mm om om.o«a.mm mm.onh.mm vo.o«v.vm mo.ouh.mm ma.owo.mm mm.ow~.mm mo.owo.mm ma.owm.¢m om mm vm.owo.mm ha.oho.mm NN.¢HH.vm mm.onv.mm mv.oum.mm mm.owo.om HN.OHo.mm bN.owm.mm mm om hm.o«m.mm NN.ono.mm NH.ouv.vm m~.o«m.vm mm.ouo.hm mm.0flN.Nm mm.o«m.vm mm.ouv.Nv 0N ma vm.o»w.mm om.ono.mm No.0um.vm em.OHH.¢m mm.owa.vm HN.OHm.mm mv.owm.~m o~.0Hc.hm ma 0H om.owm.mm mH.OHv.mm no.onm.vm Ha.owm.mm Ha.onm.mm hm.owv.am mH.ch.Hm oc.owm.mm 0H m mm.oan.m~ vm.owm.m~ ma.cwo.vm vo.owm.mm vm.oww.vm om.oww.~m NM.OHo.~m mm.o«a.mw m o mh.m mh.m mh.v mh.m mh.N mh.H mh.o oo.o AEOV OomeSm “Ev noum3 ocewauo>o mo cameo ucmfiwpmm Eouw cameo H.m< mafia anode mcflmsmq oxen Eoum mouoo unwefloom a“ Hounds ofinmmuo unwound mo mmucofifldou owns» mo uouuo pumpcmum oco one new: .N l< THEME 51 in NH NH NH NH N oN oNH NNH in NH NH u- oH N NH NNH NNH in NH NH HH NH N HH NNH NNH u- NH HH N HH oH oH NNH oNH NH NH oH HH NH NH NH oNH NNH NH NH oH .. N N NH NNH oNH NH NH oH NH HH N NH oNH NHH NH NH N NH NH N HH NHH oHH NH NH HH HH NH N HH oHH NoH NH NH N oH oH N oH NoH ooH NH NH N oH N oH N ooH NN NH NH N oH N NH oH NN oN oN NH oH NH oH NH N oN NN NN NH NH NH HH NH N NN oN NN NH HH NH oH NH N oN NN oe NH N NN HH NH N NN oN Ne NH NH NN N NH N oN NN NN NN NN NN NH NH oH NN oN NN No eN oN NH NH oH oN NN NN Ne HN NN HN NH N NN oN es in NN Ne NN NH N NN Ne Ne NN Ne oN eN oN HH Ne oe Ne NN eN HNH NN NN NH oe NN oN NN HHH HoN NN HN NH NN oN eoH NoH NNN NNH Ne oN NH oN NN NNN NNN NNN NNH NN NN NH NN oN NNN NNH NNH NHH HN NN NN oN NH NNH NNH NHH NN NNH NN oe NH oH NNH HNH NoH HN NN HN NN oH N HN NN HN eN NN No eN N o NN.N NN.N NN.N NN.N NN.N NN.H NN.o Isoc ooouhoo LEV acoENme unfixHHo>o Houm3 mo gamma ucoENpmm 3oHoo suooo mo mm chH mcoHc coxcu muuoo unmENpom :N nudop :qu AucmHmB Nun A .mcfimccq oxen mo cflmmn canon ecu to ma may m4 Hmuou mo mcofiumnucoocoo .mn< manna 52 NN NH NH NH .. oH oH u- oNH NNH NH NN NH N i- N N N NNH NNH NN NN NH NH in N N oH NNH NNH NH NH NH NH .. HH oH HH NNH oNH oN HN NH NH .. oH N oH NNH NNH .NH NN NH NH -1 N N oH NNH oNH NH oN NH NH NH HH N N oNH NHH NH HN NH NH NH NH N HH NHH oHH NH NH HH oH NH NH N HH oHH NNH NH NN NH HH NH oH NH HH NNH ooH NN NN NH NH NH oH NH oH ooH NN NN NN HH HH .. N NH oH NN oN NN NH oH NH NH N NH HH oN NN NN NN NN NH oH N NH N NN oN NN NN NN NH HH N NH HH oN NN NN Ne NH N NH NH NH oH NN oN NN Ne NH N NH NH NH N oN NN .. Ne NH HH NH NN NH oH NN oN HN NN NH NH NH NN NH N oN NN oN HN NN HH NN NN NN HH NN oN NN NN NN NH oN NN NH N oN Ne NN NN NN NN Ne NN NH N Ne oe NN eN NN NN Ne NNH oN N oe NN NN NN NNH HNH NN NNH oN NH NN oN HoH NN NHN --- NN NNH NH oN oN NN NNH HNN NNH oNH NoH NN NN NN NN oN NNN NNN oNH oNH NNH NN HN NN oN NH oHN NNN NNH NNH NNH NN oN NN NH oH NNH NNH NNH NNH oNH NN NN NN oH N NHH NHH oNH NoH NNH oN NN HN N o N.N N.N N.N N.N N.N N.N N.H N.N lees ooNNHsm HEN ucoeNpom mcHxHHo>o nouns mo gamma unoEprm soHon spoon H uoomcmuu macaw :mxmu mouoo unmeflpmm :N cameo nufi3 Ausmww3 who .mcfimccq oxen mo :Nnmn :uzom on» no um u< 0:. m4 Hmuou mo mcoNumuucoocoo .vud uHacB H 53 .Acowuo>auoo couusocv monEom ucoeflpoo HON omonu ou HoHNENm moHopoooud Luflz mdlaouou HON pouxHoco poo .pcsoum .poflup ouoz coNuooHHoo ucoHQ coco mo monEoolnom oowne H ouocm Eoum No.~ N NN.NN mucon oHonz .A Esososop EsNmemoNosoo m.mio.~ E oHHimoa oHocm Eouu vn.~ H NN.NN mucoam oHon3 .dmo nonomosouom o.~lm.a E moaion ouocm cococon A.xnow2. Eoum NN.N H NN.NN 35.3 30:3 £32358 ooeoNN N.H-NN s 3.8 oHocm mo.va N NN.NH muoou poo moEoNN:H Eouw o wo>moH poo moHoHqu .uwd erodes noxdzz m.ono E omio NN.NH w NN.NN muoou poo moEONHnH o mo>ooH pco mEoum .4 UNNokmNozmxe oxdm& mm.m N Hm.vv muoou o mo>ooH poo mEoum .A BNNBoNNoo Esxxumn o oopo oxoa pcoauoz mm.~ w vm.mH muoou cN£UN3 o mo>ooH can mEouw .H omsoowNoo Essxumu o E mated 30 mmonM a: mwmod moNoodm SMMWQ cowuoooq an Hm noun: .ommH .H zaob Hm< ocNH mcoHo pouooHHoo mouhnmouooe CN owcomuo mo ncowuouucoocoo .mia oases 54 mouspoooum nuflz mmlaouou new pou>Hono pan .pcoouo .poNHQ ouo3 :oNuooHHoo ucoam coco mo monEomlnom oouce .HcoNuo>Huoo couusocv moHQEom acoENUom How omonu ou HoHNENm H ouosm Eoum mm.v N mo.vm mucoam oHonS .4 Esososop EsNNNxdoNoxob m.~|o.m E oomloma ouosm conocoad A.xnoflzv Eouw Nh.m N om.mm mucoam oHozz owoxopcxuo oopomm o.~uo.H E omatooa onoso AHHNSSBV Eouw no.¢ « av.mm mucoam oHon3 omsoNxaon oneeb o.H|m.o E ooauom upNEsom poo xumom NN.N N NN.NN NosoHo oHoss x.oHHHzN NNNNaoNN NeNse oHosm om.n H hm.mH mpoou use ooEOanu Eoum o mo>ooa one moHoNuom .uH< stomps seemsz m.ono E omlo mn.ma H wm.ow muoou o mo>ooa can mEoum .H omsoomNoo Esxxumn o ompo oxoa om.¢ H om.mH muoou can moEoNNSH o mo>ooa can mEoum .A omNokmuozmxo UNQNN pcoauo3 ow.m H NN.NH muoou :«nuw3 o mo>ooH one mEoum .H owseocho Ezsxumu o E maioH a mmmaou m: owned moNoomm nhmwo :oNuoooq 3 HI 4 u an Houoz .omma .H >H5h “a uoomcouu ocean pouooaaoo mouxnmoHooE EN oficomuo mo mEONuoHucoocoo .mic oHnoB 55 Table A-7. The slope of the south basin of Lake Lansing. TRANSECT 1 LINE AB Distance Depth Distance Depth from shore cm from shore cm 0 0 0 0 so 84 35 84 100 93 55 123 130 109 65 140 160 138 > 75 158 190 192 85 191 200 260 95 191 210 344 105 271 220 391 110 320 230 421 115 . 392 _ 240 455 120 429 250 497 125 477 260 549 130 504 270 606 135 540 280 676 145 604 290 711 155 665 300 730 165 702 310 750 185 730 210 723 LITERATURE CITED LITERATURE CITED American Public Health Association. 1976. Standard Methods for the Examination of Water and Wastewater, 14th ed. American Public Health Association, Washington, D.C. 1193 pp. Aston, S.R., I Thornton, and S.J. Webb. 1975. Arsenic in stream sediments and waters of south west England. Sci. Total Environ. 4:347-358. Bails, J.D. and R.C. Ball, 1966. Response of pond metabolism to sodium arsenite. Mich. Acad. Sci. Arts and Let. 11: 193-208. Batterson, Ted R. 1980. Arsenic in Lake Lansing, Michigan. PhD. Dissertation. Michigan State University, East Lansing, Michigan. 79 pp. Bishop, R.F. and D. Chisholm. 1962. Arsenic accumulation in Annapolis Valley orchard soils. Canad. Jour. Soil Sci. 42:77-80. Bowen, H.J.M. 1966. Trace Elements in Biochemistry. Academic Press, New York. 250 pp. Carigman, R. and R.J. Flett. 1981. Postdepositional mobility of phosphorus in lake sediments. Limnol. Oceanogr. 26: 361-366. Cowell, B.C. 1965. The effects of sodium arsenite and silvex on the plankton populations in farm ponds. Trans. Amer. Fish. Soc. 94:371-377. Crecelius, Eric A. 1975. The geochemical cycle of arsenic in Lake Washington and its relation to other elements. Limnol. Oceanogr. 20:441-451. Crecelius, Eric A., Michael H. Bothner, and Roy Carpenter. 1975. Geochemistries of arsenic, antimony, mercury, and related elements in sediments of Puget Sound. Environ. Sci. Technol. 9:325-333. Crosby, D.C. and R.K. Tucker. 1966. Toxicity of aquatic herbicides to Daphnia magna. Science.154:289-290. 56 57 Dupree, H.K. 1960. The arsenic content of water, plankton, soil and fish from ponds treated with sodium arsenite for weed control. Proc. 14th Ann. Conf. SE Assn. Game and Fish Comm. 132-137. Durum, W.H., J.D. Hem, and 8.6. Heidel. 1971. Reconnaissance of Selected Minor Elements in Surface Waters of the United States. Geological Survey Circular 643. USDI, Washington, D.C. Domogalla, B.P. 1926. Treatment of algae and weeds in lakes at Madison, Wisconsin. Eng. News-Record. 97:850-870. Erik, R.C. and C. Nan-Kwang. 1981. Fluxes of arsenic, lead, zinc and cadmium to Green Bay and Lake Michigan. Environ. Sci. Technol. 15:553-558. Ferguson, John F. and Jerome Gavis. 1972. A review of the arsenic cycle in natural waters. Water Res. 6:1259-1274. Ferguson, John F. and Marc A. Anderson. 1974. Chemical forms of arsenic in water supplies and their removal, p. 133- 158. In Alan J. Rubin (ed.), Chemistry of water supply, treatment, and distribution. Ann Arbor Science Publishers Inc., Ann Arbor, Michigan. Fleischer, M. 1973. "Cycling and Control of Metals: Proceed- ings of an Environmental Resources Conference," National Environmental Research Center, Cincinnati. Gilderhus, P.M. 1966. Some effects of sublethal concentra- tions of sodium arsenite on bluegills and the aquatic environment. Trans. Amer. Fish. Soc. 95:289-296. Hem, J.D. 1959. Study and interpretation of the chemical characteristics of natural waters. Geological Survey Water-Supply Paper 1473. U.S. Govt. Print. Off. 123 pp. Hutchinson, G. Evelyn. 1957. A Treatise on Limnology. I. Geography, Physics, and Chemistry. John Wiley and Sons, Inc., New York. 1015 pp. Kanamori, S. and K. Sugawara. 1965. Geochemical study of arsenic in natural waters, Part 2, Arsenic in river water. Rep. Research Lab., Faculty of Science, Nagoya Univ. 46-57. Knoecklein, George W. 1981. The vegetation and hydrology of a lakeside wetland. Master's thesis. Michigan State University. 38 pp. Kobayashi, Shigeru and G. Fred Lee. 1978. Accumulation of arsenic in sediments of lakes treated with sodium ar- senite. Environ. Sci. Technol. 12:1195-1200. 58 KopP: J.F. and R.C. Kroner. 1967. "Trace Metals in Waters of the United States," U.S. Dept. of Interior. Lawrence, J.M. 1958. Recent investigations on the use of sodium arsenite as an algacide and its effects on fish production in ponds. Proc. 11th Ann. Conf. SE Assn. Game and Fish Comm. pp. 281-287. Lis, Steven A. and Philip K. Hopke. 1973. Anomalous arsenic concentrations in Chautauqua Lake. Environ. Letters. 5:45-51. Mackenthun, Kenneth M. 1950. Aquatic weed control with sodium arsenite. Sewage Ind. Wastes. 22:1062-1067. Peoples, S.A. 1975. Review of arsenical pesticides. In E.A. Woolson, ed. Arsenical Pesticides (ACS Symposium Series, No. 7). American Chemical Society, Washington, D.C., pp. 1-12. ' Ritchie, J.A. 1961. Arsenic and antimony in some New Zealand thermal waters. N.Z.J. Sci. 4:218-229. Roelofs, Eugene W. 1958. The effect of weed removal on fish and fishing in Lake Lansing. Department of Fisheries and Wildlife, Michigan State University, East Lansing, Michigan. 10 pp. Ruppert, D.F., P.K. Hopke, P. Clute, W. Metzger, and D. Crowley. 1974. Arsenic concentrations and distribution in Chautauque Lake Sediments. J. Radioanal. Chem. 23: 159-169. Sayre, W.W., H.P. Guy, and A.R. Chamberlain. 1963. Uptake and transport of radionuclides by stream sediments. Geol. Surv. Professional Paper 433-A. U.S. Govt. Print. Off. 33 pp. Seydel, Ilse S. 1972. Distribution and circulationrof arsenic through water, organisms and sediments of Lake Michigan. Arch. Hydrobiol. 71:17-30. Shapiro, Joseph. 1971. Arsenic and phosphate: Measured by various techniques. Science. 171:234. Sohacki, Leonard P. 1968. Dynamics of arsenic in the aquatic environment. Ph.D. Dissertation. Michigan State Univer- sity, East Lansing, Michigan. 94 pp. Surber, E.W. 1929. The utilization of sloughs in the upper Mississippi Wildlife and Fish Refuge as fish ponds. Trans. Amer. Fish Soc. 59:106-113. 59 Vallee, B.L., D.D. Ulmer and W.B. Wacker. 1960. Arsenic toxicology and biochemistry. AMA. Arch. of Ind. Health. 21:132-151. Wagemann, R., N.B. Snow, D.M. Rosenberg, and A. Lutz. 1978. Arsenic in sediments, water and aquatic biota from lakes in the vicinity of Yellowknife, Northwest Territories, Canada. Arch. Environ. Contam. Toxicol. 7:169-191. Walsh, L.M. and D.R. Keeney. 1975. Behavior and phytotoxicity of inorganic arsenicals in soils. In E.A. Woolson, ed. Arsenical Pesticides (ACS Symposium Series, No. 7). American Chemical Society, Washington, D.C., pp. 35-52. Wetzel, Robert G. 1975. Limnology. W.B. Saunders Co., Phila- delphia. 743 pp. Woolson, E.A. 1975. Bioaccumulation of arsenicals. In E.A. Woolson, ed. Arsenical Pestici-es (ACS Symposium Series, No. 7). American Chemical Society, Washington, D.C., pp. 97-107. MICHIGAN STATE UNIV. LIBRARIES iliillWilWWillHillWIINIWHIlllliliiliiilmill 31293103903898