Rich Lakes Research Articles

Natural inactivation of phosphorus by aluminum in preindustrial lake sediments

Noncalcareous phosphorus (P)‐rich lake sediments typically release P associated with iron hydroxide [Fe(OH)3] during the development of hypolimnetic anoxia. High concentrations of aluminum hydroxide [Al(OH)3] in such sediments (e.g., in aluminum [Al]‐treated lakes) can prevent the P release. Here we show that sediment ability to bind P can naturally develop during lake history because of changes in the Al(OH)3 concentration and the Al(OH)3 to Fe(OH)3 ratio in sediment. We reconstructed the development of sediment Psorption characteristics through the Late Glacial and Holocene periods on the basis of sequential fractionation analysis of P, Al, iron, calcium, and magnesium in a ~14,000‐yr‐long sediment record, fresh settling seston, and bedrock from Plešné Lake, Czech Republic. The most significant change occurred at the Late Glacial—Holocene transition (~10,000 yr before present), when the watershed became forested and soil erosion decreased. The Late Glacial sediment was rich in mineral detritus, derived from the watershed till and bedrock, had most of its P associated with Fe(OH)3 (37%) and calcite and/or apatite (42%), and was probably able to release P during anoxia. In contrast, the Holocene sediment was highly organic, the P release during anoxia was likely negligible, and >90% of P was associated with Al(OH)3. This Al(OH)3 originated from photochemical liberation of Al from dissolved organically‐bound Al (Alo) exported to the lake from soils. Because the photochemical mechanism was a more efficient source for Al(OH)3 than for Fe(OH)3, the sediment became a P trap by the beginning of the Holocene. The ability of the sediment to immobilize P increased further during the anthropogenic acidification era because of elevated terrestrial export of ionic Al. Similar sediment fractionation results in cores from six lakes in Maine suggest that this P‐immobilizing mechanism is a general process that can occur in lakes with high Alo inputs.

Degradation of water quality in Lough Neagh, Northern Ireland, by diffuse nitrogen flux from a phosphorus‐rich catchment

Annually resolved fossil records of nitrogen (N) inputs (as sedimentary δ15N, N content), aquatic production (δ13C, C content), and algal abundance and gross community composition (pigments, nonsiliceous microfossils) from Lough Neagh, Northern Ireland (NI), were compared with annual records of climatic variability, atmospheric and urban nutrient loading, whole‐catchment nutrient budgets, and limnological monitoring data to identify the unique effects of N on the eutrophication of a phosphorus (P)‐rich lake during ca. 1933‐1995. Cluster analysis revealed two major biostratigraphic zones. Zone I (ca. 1933–1955) was characterized by moderate lake production, as inferred from low concentrations of most fossil pigments and reduced δ15N signatures but elevated δ13C values and chlorophyte microfossil concentrations. In contrast, Zone II (ca. 1955‐1995) exhibited greatly increased contents of 15N, N, C, and algal pigments, combined with strongly reduced δ13C ratios and chlorophyte fossil abundance, a pattern consistent with recent severe eutrophication. Overall, microfossils of diazotrophic cyanobacteria were most abundant during the transition period between zones (ca. 1955‐1964). Regression analysis revealed that past N influx to the lake (as δ15N; r2 = 0.916, p < 0.0001), colonial cyanobacterial abundance(as myxoxanthophyll; r2 = 0.837, p < 0.0001), and total algal standing crops (as b‐carotene; r2 = 0.388, p < 0.0001) were all strongly correlated to agricultural inputs of N to NI farmland, weakly correlated to P inputs to NI farmland (r2 δ15N = 0.503, p < 0.0001; r2cyanobacteria = 0.296, p < 0.0001; r2total algae = 0.046, p < 0.05), and uncorrelated to most measures of climatic variability and atmospheric or urban nutrient inputs. Thus, degradation of water quality during the 20th century resulted from excessive loading of diffuse N to the lake from P‐rich agricultural lands.

High Heterotrophic Bacterial Production in Acidic, Iron-Rich Mining Lakes

The acidic mining lakes of Eastern Germany are characterized by their extremely low pH and high iron concentrations. Low concentrations of CO2 in the epilimnion due to the low pH and reduced light transmission due to dissolved ferric iron potentially limit phytoplankton primary production (PP), whereas dissolved organic carbon (DOC) may promote heterotrophic production of bacteria (HP). We, therefore, tested whether HP exceeds PP in three lakes differing in pH and iron concentration (mean pH 2.3-3.0, 23-500 mg Fe L(-1)). Bacterial biomass and HP achieved highest values in the most acidic, most iron-rich lake, whereas PP was highest in the least acidic lake. HP was often higher than PP (ratio HP/PP up to 11), indicating that planktonic PP was not the main carbon source for the bacteria. HP was not related to PP and DOC, but HP as well as bacterial biomass increased with decreasing pH. Light stimulated the formation of ferrous iron, changed the DOC composition, and increased the HP in laboratory experiments, suggesting that iron photoreduction caused DOC degradation. This may explain why we found the highest HP in the most acidic and most rich lake. Overall, the importance of bacteria in the cycling of matter and as a basis for the whole food web seemed to increase in more acidic lakes with higher iron concentrations.

UV-light induced mineralization of organic matter bound chlorine in Lake Bjän, Sweden––a laboratory study

Surface water and aqueous solutions of isolated organic matter from a humic rich lake in southern Sweden were exposed to artificial UV radiation to investigate the UV light induced influence on organic matter bound chlorine in natural systems. It was found that the photodegradation of organic matter bound chlorine was more pronounced than the photodegradation of organic carbon. After 120 h of irradiation of the isolated organic matter, only 35% of the initial organochlorine was still in the solution compared to about 70% of the dissolved organic carbon (DOC). A similar result was obtained for unfractionated surface water. Furthermore, our results indicate that the loss of organic chlorine was mainly due to a mineralization of organic chlorine into chloride ions. The total decrease of organic chlorine after 120 h was 32 μg Cl org l −1, of which the major part disappeared in the initial irradiation phase. A similar increase was observed in the chloride concentration (34 μg Cl − l −1).

Interactions among epilimnetic phosphorus, phytoplankton biomass and bacterioplankton metabolism in lakes of varying submerged macrophyte cover

The effect of submerged macrophytes on interactions among epilimnetic phosphorus, phytoplankton, and heterotrophic bacterioplankton has been acknowledged, but remains poorly understood. Here, we test the hypotheses that the mean summer phytoplankton biomass (chlorophyll a): phosphorus ratios decrease with increased macrophyte cover in a series of nine lakes. Further, we test that both planktonic respiration and bacterioplankton production increase with respect to phytoplankton biomass along the same gradient of increasing macrophyte cover. Increased macrophyte cover was associated with a lower fraction of particulate phosphorus in epilimnia, with total particulate phosphorus declining from over 80% of total phosphorus in a macrophyte free lake to less than 50% in a macrophyte rich lake. Phytoplankton biomass (chlorophyll a) too was lower in macrophyte dominated lakes, despite relatively high levels of total dissolved phosphorus. Planktonic respiration and bacterioplankton production were higher in macrophyte rich lakes than would be expected from phytoplankton biomass alone, pointing to a subsidy of bacterioplankton metabolism by macrophyte beds at the whole lake scale. The results suggest that the classical view of pelagic interactions, which proposes phosphorus determines phytoplankton abundance, which in turn determines bacterial abundance through the production of organic carbon, becomes less relevant as macrophyte cover increases.

Seasonal variation in mortality of brown trout (Salmo trutta) in an acidic aluminium-rich lake

We have studied the seasonal variation in aluminium toxicity in caged brown trout (Salmo trutta), during one year (October 1994 to September 1995) in a small acidic aluminium rich lake, Lake Nepptjern (ANC -22.6 μeq l-1, pH 5.2, total Al 400 μg l-1). Trout from two and three different year classes were exposed each month to the lake water for 48 h. Fish were placed in keepnets located in the middle of the lake, at 2 m depth. Fish mortality and water physico-chemistry were monitored during the exposures. The concentration of inorganic monomeric aluminium in the water was approximately 300 μg l-1 in average, and the water was acutely toxic to the fish. The observed mortality varied throughout the year, and was highest during spring and summer. During spring, the small fish were more sensitive to the toxic water than larger fish, while the opposite was the case during summer. Water temperature and fish length could explain most of the seasonal variation in mortality. Statistical analyses indicated that water acid neutralising capacity (ANC) and the amounts of total organic carbon (TOC) and silicon in the water also could explain some of the variation in mortality. Variation in other physico-chemical parameters, however, such as silicon, TOC and ANC could only explain the variation in mortality to a limited extent. The mechanism for the temperature dependent mortality is discussed, and we suggest that the dependence of water O2-solubility and fish metabolism upon temperature is of importance. The difference in mortality between small and large fish is discussed in terms of the gill area/body weight ratio, and it seems to be that small fish suffer more from diffusive ion loss having a larger relative gill area than larger fish. Large fish, on the other hand, have a lower relative maximum oxygen uptake than small fish and will suffer more under conditions where aluminium is accumulated on the gill surface.

Dynamics of habitat distribution in breeding mallards: assessing the applicability of current habitat selection models

The applicability of ideal free, ideal despotic and ideal preemptive habitat selection models to explain dynamics of habitat distribution of breeding mallards (Anas platyrhynchos) was explored. Data from 35 lakes studied between 1985 and 2000 were used to examine overall habitat distribution of breeding pairs, breeding success in different habitats, within‐year order of habitat occupation, and density dependence of habitat distribution and breeding success. Two habitat types, rich and poor, were defined based on the structure and luxuriance of shore vegetation; each lake belonged to one or the other of the habitat types. Breeding pairs used the rich habitat more than expected, breeding density also being higher there than in the poor habitat. Both average brood density and breeding success were higher in the rich habitat than in the poor. Breeding success was not density dependent, neither when analysed separately for the habitat types, nor in the study area in general. Within season, arriving mallard pairs did not occupy rich lakes earlier than poor lakes. An isodar analysis based on between‐year variation of the breeding density in rich and poor habitats revealed that habitat distribution of breeding pairs was not density dependent. By contrast, density in the rich habitat increased and proportional use of the poor habitat decreased with increasing overall population density, i.e. the rich habitat got increasingly crowded. None of the habitat selection models considered was applicable to explain the dynamics of habitat distribution of breeding mallards.

Role of phosphorus and nitrogen for bacteria and phytopankton development in a large shallow lake

Nutrient (P and N) enrichment experiments in small enclosures (20 l) were carried out to determine P and/or N limitation of bacterioplankton in Lake Vortsjarv. The specific interest of the study was to test if it is possible to detect nutrient `physiological' or growth (rate) limitation of bacterioplankton and competition for nutrients (N and P) with phytoplankton in generally nutrient rich lake. Thymidine and leucine incorporation; leucine aminopeptidase, β-D-glucosidase and alkaline phosphatase activity, total count of bacteria, chlorophyll a concentration and primary production as well as the concentrations of different chemical forms of N and P were followed during 4–5 days of the experiment. To address the question of the interactions between nutrients, bacterio- and phytoplankton, experimental and seasonal data sets were included in the analyses. Phosphorus (P) had a positive effect on bacterioplankton in enclosure experiments in June 1997; no effects of nutrients were found in September 1996, while in May 1996, P affected mainly the phytoplankton. On the seasonal scale, the development of bacterioplankton was connected to primary production, total phosphorus and temperature. In enrichment experiments, bacterioplankton was mainly related with primary productivity but the possible importance of bacterial grazers could be presumed. Thus, no evidence was found for nutrient growth limitation and/or competition for N and/or P, rather bacterioplankton depended on organic food supply originating from phytoplankton.

A Simple Carbon Cycle Model of Lake Caohai, Guizhou Province

The δ 13 C values of the dissolved inorganic carbon (DIC), the organic matter of surface sediments and the aquatic plant cellulose were measured in Lake Caohai, Guizhou Province. The δ 13 C values (vs PDB) of DIC, organic matter of surface sediments and aquatic plant cellulose, range from -3.70‰ to -10.60‰, from -20.50‰ to -21.60‰, from -16.10‰ to -17.40‰ respectively. Then A simple carbon cycle model is built through the mass balance calculation. All results indicate the photosynthesis respiration of aquatic plants and the decomposition of organic matter are the decisive factors to the carbon isotopic compositions of the lake system as Lake Caohai, which is a shallow, oxygen rich lake with abundant aquatic plants.

Individual foraging behaviour indicates resource limitation: an experiment with mallard ducklings

The linkage between individual behaviour and population processes has recently been emphasized. Within this framework we studied the effect of resource limitation on the behaviour of mallard (Anas platyrhynchos) ducklings in boreal lakes. One group of 12 human-imprinted ducklings (4-16 days old) were taken to 11 "rich" lakes, i.e., with a relatively high concentration of total phosphorus in the water, and the other group of 12 ducklings to 11 "poor" lakes to forage for a period of 6 h. During this, a time budget study lasting 5 min was done of each of the 12 ducklings. In the rich lakes, ducklings fed significantly more and moved less than in the poor ones. This difference was particularly striking in above-surface feeding. Variation in foraging performance was associated with change in body mass of the ducklings: the less distance the ducklings moved and the more they fed above water, the more they gained mass. Earlier results had suggested that at least some of the boreal wetlands that lack duck broods year after year (70% of the total in one study) do so because they do not harbour enough food. Hence, it is possible that mallard populations are resource-limited at the brood stage during the breeding season.

RESPONSE OF MALLARD DUCKLINGS TO VARIATION IN HABITAT QUALITY: AN EXPERIMENT OF FOOD LIMITATION

Occurrence of Mallard (Anas platyrhynchos) pairs and broods was studied on 86 boreal lakes in two areas in south Finland during 1988–1997 (35 lakes) and 1989–1996 (51 lakes), and field experiments were conducted to determine whether food limitation is an important factor regulating population densities. In general, pairs and broods used and avoided the same lakes, but in both study areas, the proportion of lakes that were unoccupied every year was higher for broods (71% and 69%) than for pairs (26% and 31%). We hypothesized that lakes without breeding Mallards, and especially broods, were too poor to raise young in, and we tested that in field experiments on 22 lakes in boreal Sweden. Based on concentration of total phosphorus in the water, these lakes were divided into two groups: 11 “poor” lakes and 11 “rich” lakes. Mallard ducklings imprinted on humans were used to address brood-stage food limitation by studying mass change. Thirty-one individually marked ducklings were divided into three experimental groups: 12 ducklings were used in poor lakes, 12 ducklings in rich lakes, and 7 ducklings were used as controls receiving no experimental treatment in the field. Each group was studied in 11 daytime trials on 11 different lakes, except the control group which was kept in a pen with free access to food. Ducklings of all groups spent nights together as a big 31-individual group in the same pen and with free access to food. Ducklings foraging on poor lakes gained significantly less body mass than those feeding on rich lakes. At night, when ducklings of both groups had equal and free access to food, ducklings that had been feeding in a poor lake that day gained more mass than those that had been feeding in a rich lake. Overall, ∼95% of daily mass gain was accounted for by daytime gain in the control group, the corresponding percentage was 35% in the rich-lake experimental group and 11% in the poor-lake experimental group. Thus, we have observational, as well as experimental, evidence to conclude that many boreal lakes are inferior breeding habitat for Mallards, especially during the brood stage. Our results suggest (1) that it is important to recognize at what stage of the yearly reproductive cycle food limitation may occur, (2) that brood stage may be more important than nesting stage for possible population limitation and regulation in breeding Mallards, and (3) that the reason why so many lakes in the boreal region lack breeding Mallards may be food limitation.

Individual foraging behaviour indicates resource limitation: an experiment with mallard ducklings

The linkage between individual behaviour and population processes has recently been emphasized. Within this framework we studied the effect of resource limitation on the behaviour of mallard (Anas platyrhynchos) ducklings in boreal lakes. One group of 12 human-imprinted ducklings (4-16 days old) were taken to 11 "rich" lakes, i.e., with a relatively high concentration of total phosphorus in the water, and the other group of 12 ducklings to 11 "poor" lakes to forage for a period of 6 h. During this, a time budget study lasting 5 min was done of each of the 12 ducklings. In the rich lakes, ducklings fed significantly more and moved less than in the poor ones. This difference was particularly striking in above-surface feeding. Variation in foraging performance was associated with change in body mass of the ducklings: the less distance the ducklings moved and the more they fed above water, the more they gained mass. Earlier results had suggested that at least some of the boreal wetlands that lack duck broods year after year (70% of the total in one study) do so because they do not harbour enough food. Hence, it is possible that mallard populations are resource-limited at the brood stage during the breeding season.

Cadmium sources and exchange rates for Chaoborus larvae in nature

Although freshwater insects are known to accumulate trace metals in the laboratory from both water and food, the relative importance of metal sources for these animals, as well as the rate at which they take up and eliminate their metal, has not been measured in nature. We describe a novel in situ approach that allowed us to determine that trophic transfer is the main source of cadmium for larvae of a common lake‐dwelling animal, the phantom midge Chaoborus punctipennis. We transferred C. punctipennis larvae from a low‐cadmium to a high‐cadmium lake, where they were exposed in 64‐μm‐mesh mesocosms to the prevailing high‐Cd concentrations in water and to various quantities of prey collected from the Cd–rich lake. Our experimental design ensured exposure of C. punctipennis larvae to realistic Cd concentrations in water and in a natural mixture of prey types. Our results indicate that larvae take up their Cd mainly from prey. Thus models of metal dynamics and effects on these invertebrates are likely to be more realistic if they include food as a metal source. Using the same mesh mesocosm design, we also determined that C. punctipennis larvae transferred from a high‐Cd to a low‐Cd lake lost their Cd slowly. Combining our information on Cd uptake and loss from C. punctipennis allowed us to model Cd exchange between this insect and its surroundings.

Phytoplankton Activities in Hypersaline, Anoxic Conditions. II—Photosynthetic Activity of some Sulphide Adapted Cyanobacterial Strains Isolated from Solar Lake, Taba, Egypt

Five cyanobacterial strains namely Oscillatoria limnetica, Aphanotheca stagnina, Synechococcus aeruginosus, Spirulina tenissima and Dactylococcopsis salina were isolated from an anaerobic sulphide rich hypersaline Solar Lake, Taba- Egypt. At oxygenic conditions Spirulina was the most tolerant strain to sulphide concentrations, while Aphanotheca and Synechococcus were the most sensitive. Aphanotheca have higher anoxygenic photoassimilation activity under a broad range of sulphide concentrations. The response of the five cyanobacterial strains to sulphide under oxygenic and anoxygenic conditions depends on the organism, time of exposure, presence or absence of DCMU and also on sulphide concentration.

Long-term phytoplankton changes in a shallow hypertrophic lake, Albufera of Valencia (Spain)

A long-term phytoplankton study was carried out in the Albufera of Valencia, a shallow hypertrophic lake (surface area 21 km2, mean depth 1 m, total inorganic nitrogen load 155 g m-2 y-1, total inorganic phosphate load 15 g m-2 y-1) from 1980 to 1988. The lake functions as a reservoir for the surrounding rice cultivation. From 1940's to 1988, its phytoplankton assemblage has been altered from a mesotrophic to a hypertrophic character, as consequence of the increasing pollution. For 1980–88, annual variations in the phytoplankton were less pronounced than seasonal changes. The hypertrophic and morphometric features of the lake favoured the stability of the phytoplankton assemblage and chlorophyll a levels during the study period. Seasonal and horizontal distribution of the total phytoplankton abundance and biomass were highly influenced by the hydrological cycle of the lagoon. Compared with other shallow nutrient rich lakes, the Albufera of Valencia is similar to the shallow hypertrophic lakes of the Netherlands.

The effects of naturally occurring high and low molecular weight inorganic and organic species on the yolk-sack larvae of Atlantic salmon ( Salmo salar L.) exposed to acidic aluminium-rich lake water

Yolk-sac larvae of Atlantic salmon ( Salmo salar L.) were exposed to water from two acidic, aluminium rich lakes, Nepptjern and Lomstjern. The average pH of the lakes during the experiments were 4.85 and 5.24. The concentrations of total monomeric aluminium (Al a) were high in both lakes, i.e. 640 ± 22 and 451 ± 16 μg Al 1 −1, but the total amount of organic carbon (TOC) differed significantly (1.89 ± 0.04 and 11.6 ± 0.1 mg Cl −1). The larvae were also exposed to cation exchanged and hollow-fiber ultrafiltered water from both lakes. Based on the hollow-fiber ultrafiltration, the toxic aluminium species present were as the inorganic, low molecular weight fraction (mol. wt. < 3 000 Da). Within this fraction, the highest mortality rates [LT 50=24.6±0.5 h ( n = 6)] were observed in the lake water having the highest amounts of the hydrolysis products, i.e. Al 3+, Al(OH) 2+ and Al(OH) 2 +. The aluminium-fluorides, AlOHF +, AlF 2+, and AlF 2 + were also found to be toxic, but less toxic [LT 50=49.6±1.1 h ( n=4)] than the aluminium hydrolysis products. Removal of high molecular weight species including inorganic and organic Al species, did not affect the toxicity or the mortality rates. No acute toxicity was observed when the larvae were exposed to the cation exchanged eluates. Thus, the organic aluminium-complexes seemed to have no acute toxic effect, at least at the concentration levels tested.

The short term response to eutrophication abatement

We compare results of a new model for predicting the short term inter annual changes in chlorophyll-a (chl-a) in lakes after reductions in total phosphorus (TP) to predictions made by least squares regression models. In the new method, slopes of chl-a/TP graphs (both axes in mg · m−3) are depicted in frequency diagrams and used to extract information on the expected, short term chl-a/TP response. The short term response for nine shallow (< 10 m deep) and nutrient rich lakes to changes in TP was found to be: Chl-a = 0.49 · TP + 17.3, and for nine deep, P-limited lakes: Chl-a = 0.08 · TP + 3.5. If the TP-reduction is known to be greater than 10 mg · m−3, the expected slope increases to 0.58 for shallow lakes and to 0.26 for deep lakes. The slope, 0.58, is 8% lower than the slope for the long term response calculated by regression for the shallow lakes. For deep lakes the slope, 0.26, is 2 to 3 times higher than that calculated by regression, indicating that reductions in TP for deep lakes give greater effects than least squares regression equations suggest. We have also calculated the reduction in TP which will give about 80% probability that a reduction in chl-a will be observed next year. For shallow, P-limited lakes this reduction is about 30 mg · m−3 (5% of average initial in-lake TP concentration), and for deep lakes about 14 mg · m−3 (35% of average initial in-lake TP concentration).

Methylmercury in some swedish surface waters

AbstractThree lakes and catchment runoff in the Lake Gårdsjön watershed on the Swedish west coast have been sampled for methylmercury during one year. The analyses indicated that high outputs of methylmercury were present during autumn in flushing water systems with short retention time, whereas smaller outputs and variations occurred in lake waters with longer retention time.Thus, methylmercury concentration were measured as 0.16 to 0.41 ng/L (humic rich lake), 0.12 to 0.26 ng/L (limed lake), &lt;0.1 ng/L (acid clearwater lake with long retention time) and 0.13 to 0.28 ng/L (acid runoff from a forest).A flux estimate from one forested catchment (3.6 ha) gave a yearly output of 5.2 mg and the specific output was estimated as 144 ± 70 mg/km2/y.The measured methylmercury concentration in this study accounted for only 6 to 13% of the total dissolved organic bound mercury in the water samples.

Effects of road deicing salt on chloride levels in four adirondack streams

A 2 yr study was conducted in 1987 and 1988 to determine the effects of road salt run-off from New York State highway 28N on chloride levels in four tributary streams to Rich Lake, located in the central Adirondack Mountains of New York. All four study streams exhibited significant increases in chloride concentrations at sampling stations located 50 and 100 m downstream from the highway. Chloride concentrations in downstream samples were as much as 31 times higher than comparative upstream samples. Elevated chloride levels continued throughout the 6 mo period following the termination of winter salt applications. A water chemistry profile developed for Rich Lake revealed a slight increase in chloride content at lower depths.

Biogenic silica as an estimate of siliceous microfossil abundance in Great Lakes sediments

Biogenic silica concentration (BSi) in sediment cores from the Great Lakes is evaluated as an estimate of siliceous microfossil abundance. A significant linear relationship was found between measured BSi and diatom valve abundance for sediment cores from the Bay of Quinte, Lake Ontario, Lake Erie, Lake Michigan and Lake Superior and between measured BSi and diatom biovolume for Lake Erie, Lake Michigan, and Lake Superior but not for Lake Ontario. Diatom silica predicted from diatom species abundance and an estimated silica content per cell in the Lake Erie cores accounted for 117% and 103% of measured BSi, respectively. By contrast, predicted diatom silica could only account for 28% of measured BSi in the Lake Michigan core and only 25% in the Lake Superior core. A few large diatoms with a large silica content per cell comprised a major portion of predicted diatom silica in all cores. The discrepancy between chemically measured BSi and the silica predicted from diatoms in the Lake Michigan and Lake Superior cores was partially due to the inability of the regression model, used to estimate diatom silica content, to account for different degrees of silicification in the diatom asemblages from the more dissolved silica rich Lake Michigan and Lake Superior.