Mercury content in roach (Rutilus rutilus L.) in circumneutral lakes--effects of catchment area and water chemistry

The environmental influence on the mercury content in roach (Rutilus rutilus L.) is investigated using partial least square regression on 46 environmental variables describing the land use in the catchment area, various catchment area and lake characteristics, lake water chemistry, and fish stock. The Hg content in the fish from the 78 investigated circumneutral lakes is heavily influenced by the land use in their surroundings. The boreal forest lakes possessed the highest Hg levels in roach, whereas fish from lakes surrounded by arable land had lower levels. The Hg levels also showed a negative relationship to the amount of dissolved ions and the total amount of nutrients in lake water. Lake pH did not have any significant influence on the Hg content in roach in these non-acidified lakes. The Hg levels in lakes influenced by large amounts of wetland were less well explained by the presently investigated environmental variables, which implies that the Hg burden in fish from this kind of lake is governed by other factors.     

Contemporary (1979-1988) and inferred historical status of headwater lakes in North Central Ontario, Canada

Fifty-six headwater Canadian Shield lakes were repetitively sampled from 1979 to 88 to determine their response to changes in acidic deposition of the period. Annual wet sulphate loadings varied between 38 and 83 meq m(-2), with highest deposition in the late 1970s followed by somewhat lower but variable deposition in the 1980s. Median pH of the lakes increased 0.42 pH units from 1979 to 1985 and decreased by 0.15 units between 1985 and 1988. Short water renewal times (x=1.1 y) promoted rapid equilibration. Since lake were so responsive to changes in SO4(2-) inputs, they were at or near steady state at all times. Comparison of predicted original pH and ANC with 1979 data indicate a median decline of 0.45 pH units and a loss of 34 microeq litre(-1). ANC. Four of 9 lakes were found to be historically fishless, based on the continued presence of Chaoborus americanus in sediment cores. The remaining five lakes historically had fish populations, but fish were not collected in 1979 when pH ranged betwen 4.6 and 5.3. By 1987, fish species were found in five of these lakes where pH had increased on average by 0.9 pH units. Our data indicate that water quality improvements could allow for the reinvasion or resumption of recruitment for a significant number of Ontario lakes.     

Lake variability: key factors controlling mercury concentrations in New York State fish

A 4year study surveyed 131 lakes across New York State beginning in 2003 to improve our understanding of mercury and gather information from previously untested waters. Our study focused on largemouth and smallmouth bass, walleye and yellow perch, common piscivorous fish shown to accumulate high mercury concentrations and species important to local fisheries. Fish from Adirondack and Catskill Forest Preserve lakes generally had higher mercury concentrations than those from lakes in other areas of the state. Variability between nearby individual lakes was observed, and could be due to differences in water chemistry, lake productivity or the abundance of wetlands in the watershed. We found the following factors impact mercury bioaccumulation: fish length, lake pH, specific conductivity, chlorophyll a, mercury concentration in the water, presence of an outlet dam and amount of contiguous wetlands.     

Mercury in fish in Swedish lakes

The aim of this work has been to try to obtain a picture of the past, present and future mercury situation in fish in Swedish lakes, to make an estimate of the number of lakes threatened by 'blacklisting', and to see if the data can be used to reveal anything about the impact of liming on the Hg content in pike. The register contains a broad set of data from 1456 lakes. The main results are as follows. Trend analyses indicate that the Hg content in 1-kg pike seems to increase with time. This is interesting since there has been a significant decrease in mercury emissions from Swedish industries during the last two decades. High Hg contents in 1-kg pike appear in a very characteristic pattern, linked to specific sources of Hg emission. The data indicate that old Swedish 'sins' are still causing a lot of problems. The factors governing the leakage of Hg from soils to water ought to be a very important topic for further studies. The Hg content in pike shows the highest correlation with the following parameters: Hg in surficial sediments, pH, distance from point source and water hardness, lake water alkalinity and conductivity, water retention time, size of drainage area and lake surface. A formula which provides the best possible degree of explanation (r2 = 0.78) has been derived. At present there are about 250 lakes 'blacklisted' in Sweden due to high Hg content in fish. Our data show that there are at least 9400 lakes that ought to be 'blacklisted' today. A successful liming operation will alter the chemical conditions in lakes and also decrease the Hg content in fish.     

Mercury in lakes and lake fishes on a conservation-industry gradient in Brazil

Mercury contamination in freshwater food webs can be severe and persistent, and freshwater fish are a major source of mercury contamination in humans. Northern hemisphere studies suggest that the primary pathway by which freshwater fish accumulate mercury is the food web, and that atmospheric deposition is the primary route by which mercury enters freshwater systems. Levels of atmospheric deposition are closely linked to proximity to sources of mercury emissions. These propositions have not been tested in the southern hemisphere. In this study, we measured mercury levels at three lakes in southern Brazil and assessed relationships between mercury in precipitation, lake water, sediment and fish tissues at sites close to (industrial and suburban areas) and distant from (protected conservation area) sources of mercury emissions. We also assessed relationships between mercury in fish species and their trophic habits. Mercury concentrations in sediment and lake water did not vary among lakes. In contrast, mercury in precipitation at the study lakes increased with proximity to industrial sources. Mercury in fish tissue generally increased along the same gradient, but also varied with trophic level and preferred depth zone. Atmospheric mercury deposition to these closed lakes may be directly linked to concentrations in fish, with surface-feeding piscivorous species attaining the highest concentrations.     

Use of rehabilitation experiments to understand the recovery dynamics of acid-stressed fish populations

We used rehabilitation experiments involving the stocking of 2 native sportfish, lake trout (Salvelinus namaycush) and smallmouth bass (Micropterus dolomieu), in combination with recent fish community surveys, to study the recovery dynamics of fish populations in acid-stressed lakes near Sudbury and Killarney, Ontario, Canada. Population recovery rates differed between the 2 species. Introduced lake trout did poorly in species-rich lakes and exhibited slower growth, lower survival and delayed recruitment. Smallmouth bass, in contrast, readily colonized species-rich lakes. The biomass of natural smallmouth bass recruits increased to reference lake levels within 5 years following water quality recovery and spawning by stocked fish, whereas the biomass of natural lake trout recruits remained well below reference levels 5-15 years after water quality recovery and spawning by adults occurred. We document introductions by anglers of smallmouth bass into acid-damaged lake trout lakes, including some lakes that did not contain bass prior to acidification. This range expansion of a warm-water species (bass) that can alter food-web structure and reduce the growth of a cold-water species (trout), illustrates the potential for the combination of climate warming and species introductions to greatly alter the biological recovery endpoints in acid-stressed lakes.     

Relationships between lake chemistry and calcium and trace metal concentrations of aquatic invertebrates eaten by breeding insectivorous waterfowl

Ca, P, Al, and trace metal (Cu, Ni, Zn, Cd, and Pb) concentrations were measured in several aquatic invertebrate taxa used as food by breeding insectivorous waterfowl, sampled from three sites in eastern Canada with widely varying water chemistry. Ca concentrations were highest in molluscs (snails and clams), averaging 200-300 mg g(-1) (shells included). Aquatic insects of varying sizes, life stages and habits (caddisfly larvae, dragonfly larvae, adult backswimmers, waterstriders, and whirligig beetles) had much lower mean Ca concentrations, ranging from about 0.6 mg g(-1) (beetles) to 1.8 mg g(-1) (caddisflies). Invertebrate-Ca concentrations decreased with increasing body mass for several taxa, with smaller and larger individuals providing similar absolute amounts of Ca. Ca concentrations in most aquatic insects (but not molluscs) were reduced under acidic, low Ca, high Al, low dissolved organic carbon (DOC) and/or low total phosphorus (TP) conditions. In stepwise multiple regressions, pH was consistently the main factor explaining variability in invertebrate-Ca, after controlling for the negative relationship between invertebrate-Ca and body mass for some taxa. Molluscs were absent from lakes below pH 5.3. In general, concentrations of P and metals in invertebrate taxa were not significantly correlated with lake pH. Levels of Al, Cd, or Pb were not sufficiently high to be considered toxic to potential consumers of these organisms. For waterfowl and other birds breeding in acid-stressed habitats and relying on aquatic invertebrates as a source of food, a reduced availability of dietary Ca is more likely than an increased exposure to toxic metals to negatively affect reproductive success, especially when other adverse effects of acidification (lower diversity of prey) are considered.     

Effects of acidic water on freshwater snails: results from a study of 1000 lakes throughout Norway

This paper summarizes some results from a monograph which focused on ecological and biogeographical aspects of the biota in Norwegian lakes and rivers with particular reference to the snail fauna. Field studies were made in the course of 20 summers within the period of 1953-1987. Acidic water is the main reason why snails are absent from some 4000 lakes in a low-calcium area in southern Norway ('Acid south'). This area is heavily influenced by acidic precipitation. Snails were not detected in lakes with pH below 5.2. In one lake, three snail species disappeared in parallel with acidification from pH 5.2 to 4.2. The calcium level modified the effect of low pH.     

Changes in the density, population structure, growth and reproduction of perch, Perca fluviatilis L. in two acidic forest lakes in southern Finland

The density of the perch populations decreased rapidly during the 1980s in two acidified lakes in southern Finland. In an acid-humic lake the density of the perch population decreased from 240 to 60 ha(-1) and in an acid clearwater lake from 150 to 20 ha(-1). In both lakes the rapid decrease was mainly due to failures in reproduction, although post-spawning mortality of adult fish was also recorded. The perch spawned annually in both lakes but according to the field exposures of egg strands, most of the eggs died before hatching. Nevertheless, some new recruits appeared in both lakes. These perch, born in 1983-1986, grew faster than those of the more abundant year-classes of the late 1970s, indicating a clear density-dependent growth response. According to the mortality estimate and the age distribution of perch, the latest year-classes that produced more than 100 adult perch ha(-1) were born in 1978 or 1979 in both lakes. It was calculated that the density of both populations was >500 ha(-1) in the early 1980s.     

Recolonization of acid-damaged lakes by the benthic invertebrates Stenacron interpunctatum,Stenonema femoratum and Hyalella azteca

The recolonization of acid-damaged lakes in Killarney Park, Canada is described for 3 species of benthic invertebrates; 2 mayflies (Stenonema femoratum, Stenacron interpunctatum) and an amphipod (Hyalella azteca). Synoptic surveys of 119 lakes for amphipods and 77 lakes for mayflies were conducted between 1995 and 1997 and defined pH thresholds of 5.6 for S. femoratum and H. azteca and pH 5.3 for S. interpunctatum. In an intensive study of 2 acid-damaged lakes and 2 reference lakes from 1997 to 2002, reestablishment of S. interpunctatum, S. femoratum and H. azteca occurred, when timing of the events could be estimated, less than 4-8 years after pH thresholds for specific taxa were reached. Dispersal of S. interpunctatum to all habitat patches within a lake was completed 3 years after recolonization was detected in the smallest lake (11 ha). It is anticipated that dispersal throughout the largest lake (189 ha) will take much longer. The time lag from estimated pH recovery to reestablishment and subsequent dispersal of mayflies to all suitable habitats within a lake was as much as 11 to 22+ years. The density of S. interpunctatum increased in the recovering lakes to levels higher than in reference lakes, but stable endpoints have not yet been reached during 6 years of monitoring.     

Pesticide and PCB levels in fish from Alberta (Canada)

Pesticide and PCB analyses were completed on fat and muscle samples of 750 fish collected from 11 major lakes and rivers in Alberta. Although phenoxy and organophosphate residues were always below detectable limits, traces of chlorinated pesticides and their derivatives, particularly DDE, DDD and chlordane, were detected in most fat samples. Methoxychlor was frequently found in goldeye from the North Saskatchewan River but not recorded in fish from any other lake or river. Its presence in goldeye, a highly migratory species, was probably a result of biting-fly control programs in the Saskatchewan part of the river. PCB levels exceeded 25 mg/kg in the fat of several species from the North Saskatchewan River but were generally lower in the other systems. Analysis of 160 sediment samples from the North Saskatchewan River revealed no point source of PCB contamination, with residues always <0.01 mg/kg dry weight.     

Monitoring of persistent organic pollutants (POPs): examples from Lake Vättern, Sweden

Lake V?ttern in southern Sweden is a large oligotrophic lake with high surface to catchment area ratio (ca. 0.4) and a water residence time of 60 years. The lake combines sensitivity to atmospheric POP-pollution with general susceptibility for effects of POPs and slow concentration decline. Time series, from the 1960s until 1996, of PCBs and DDT in fish and data on TCDD in sediment of L. V?ttern are presented and compared to other large lakes. The long time dataset of POPs in Arctic char (Salvelinus salvelinus) shows a significant annual decline of about 5% for PCBs and 13% for DDT. These rates are in agreement with other studies. Nutrients as well as biological effects and factors affecting the fate of POPs in oligotrophic lakes with long residence times are discussed. We propose that oligotrophic clear lakes are important ecosystems for monitoring POPs in biota.     

Reproductive status, blood chemistry, gill histology and growth of perch (Perca fluviatilis) in three acidic lakes

Perch (Perca fluviatilis L.) were sampled soon after spawning in three small acidic lakes (pH 4.3-6.1, Al(lab) 5-106 microg litre(-1), Ca2+ 0.01-0.08 mmol litre(-1)) and in one circumneutral lake (pH 5.9-6.4, Al(lab) 4-12 microg litre(-1), Ca2+ 0.06-0.07 mmol litre(-1)) in southern Finland. Due to the delayed spawning of perch in the acidic lakes, sampling in those lakes was performed later than in the reference lake. In spite of that, the gonadosomatic index (GSI) of males in all the acidic lakes was significantly greater than in the reference lake. Of the two lakes with similar low water pH, the effects on reproduction were more prominent in the lake with higher water Al content. The plasma Ca2+ concentrations of females in the acidic lakes were significantly smaller than in the females of the reference lake. The low female:male plasma Ca2+ ratio (1.0-1.32) depicted delay of spawning. Stress in perch in acidic water was also seen in elevated blood haematocrit values, especially in females. On the other hand, a low plasma Cl- level, a common response to acidic water in salmonids, was not detected in perch in the most acidic lakes. The amount of Al accumulated in the gill epithelium was highest in the most acidified lake with high Al concentration, but was also pronounced in a lake with low pH and low Al concentration.     

Critical loads for Finnish lakes: comparison of three steady-state models

Three simple steady-state water-chemistry models are used to calculate critical loads of sulfur for lakes in Finland. Because of the high concentrations of organic matter in Finnish lakes, the influence of organic anions on the calculation of critical loads has been given special attention. The first two methods are well known ion-balance methods which have been used in many previous lake-acidification studies. The third method, developed for this study, includes the numerical solution of equilibrium equations for organic anions, inorganic carbon species and inorganic monomeric aluminum. The original pH and aluminum concentration of the lakes are estimated with this model, and a method to estimate the original acid neutralizing capacity (ANC) by simulating a Gran-titration is also tested on the lake data. Uncertainty in the predictions is estimated by varying the most critical model parameters.     

Radium-226 in water, sediments, and fish from lakes near the city of Elliot Lake, Ontario, Canada

Ra-226 was measured by alpha-emission spectroscopy in water, sediments, and fish (tissues and gut contents), from five lakes in a watershed containing U mining and milling operations at Elliot Lake, Ontario, and from control lakes in an adjacent non-industrialized watershed. Ra-226 transfer parameters from lake water and sediments to fish tissues, and annual intakes by humans consuming fish, were estimated. Mean dissolved 226Ra levels ranged from approximately 76 mBq litre(-1) in water of the most affected lake, to < 10 mBq litre(-1) in control lakes. Levels in summer were consistently higher than in fall or winter; no consistent variation with depth was noted. Sediment levels ranged from approximately 3000 mBq g(-1) dry wt in one study lake to < 100 mBq g(-1) dry wt of sediment in control lakes. Bone 226Ra concentrations were higher than in muscle. The lake trout (Salvelinus namaycush), a predatory secondary consumer, had bone 226Ra levels (< 20 mBq g(-1) dry wt) that did not show significant site variation. In contrast, bottom feeding whitefish had significantly more 226Ra in bone tissue (to 38 mBq g(-1) dry wt in the lake whitefish, Coregonus clupeaformis, and 76 mBq g(-1) in round whitefish, Prosopium cylindraceum) in study lakes than in controls (< 20 mBq g(-1) dry wt). Ra-226 levels in lake trout muscle were low and showed erratic variation among lakes whereas levels in whitefish muscle did not vary significantly among study and control sites. Lake herring (= cisco, Coregonus artedii), a planktivorous fish taken only from Quirke Lake, had mean 226Ra levels of 18 and 1.4 mBq g(-1) dry wt in bone and muscle, respectively. Gut 226Ra levels, highest in lake trout from McCabe and Quirke Lakes (126 +/- 53, 64 +/- 44 mBq g(-1) dry wt, respectively), and just detectable in McCabe and Elliot Lake whitefish (24 +/- 2, 36 +/- 14 mBq g(-1) dry wt, respectively), were below detection in lake trout and whitefish from other lakes. Concentration ratios (CRs) of 226Ra from water to muscle ranged from 8 to 14 in lake trout, 7 to 14 in whitefish, and 4 to 6 in lake herring. The water to bone CRs varied from 81 to 142, 314 to 548, and 126 to 272 in the same species. CRs were always < 1 between sediments and fish tissues. Annual intake of 226Ra by human consumers of these fish was estimated to provide an effective dose of 0.003 mSv year(-1). This represents a small fraction of both natural background (> 2 mSv year(-1)) and the public dose limit (5 mSv year(-1)).     

Future distribution of Arctic char Salvelinus alpinus in Sweden under climate change: effects of temperature, lake size and species interactions

Novel communities will be formed as species with a variety of dispersal abilities and environmental tolerances respond individually to climate change. Thus, models projecting future species distributions must account for species interactions and differential dispersal abilities. We developed a species distribution model for Arctic char Salvelinus alpinus, a freshwater fish that is sensitive both to warm temperatures and to species interactions. A logistic regression model using lake area, mean annual air temperature (1961-1990), pike Esox lucius and brown trout Salmo trutta occurrence correctly classified 95?% of 467 Swedish lakes. We predicted that Arctic char will lose 73?% of its range in Sweden by 2100. Predicted extinctions could be attributed both to simulated temperature increases and to projected pike invasions. The Swedish mountains will continue to provide refugia for Arctic char in the future and should be the focus of conservation efforts for this highly valued fish.     

Correlates of mercury in fish from lakes near Clyde Forks, Ontario, Canada

Subsurface soils near Clyde Forks, Ontario, Canada, can have naturally high concentrations of mercury (Hg) from local geological sources. To investigate Hg in local aquatic food webs, Hg was measured in fish dorsal muscle (mainly yellow perch [YP] and pumpkinseed sunfish [PS]) and surface sediments from 10 regional lakes. Water chemistry, along with fork length, weight, and stable isotopes (delta15N, delta13C, delta34S) in fish were also measured. No lake sediments had elevated (>0.3microg/g dw) Hg, and average Hg concentrations in fish were not sufficiently high (<1microg/g dw) to be of concern for fish-eating wildlife. Variance in fish Hg was best explained by dietary carbon source (delta13C), and certain lake variables (e.g., pH for YP). PS with more pelagic feeding habits had higher delta34S and Hg than those with more littoral feeding habits. Potential biological linkages between fish Hg and delta34S, a parameter that may be related to the lake sulphate-reducing bacteria activity, requires further investigation.     

Fish status survey of Nordic lakes: effects of acidification,eutrophication and stocking activity on present fish species composition

The status of fish populations in 3821 lakes in Norway, Sweden and Finland was assessed in 1995-1997. The survey lakes were chosen by stratified random sampling from all (126 482) Fennoscandian lakes > or = 0.04 km2. The water chemistry of the lakes was analyzed and information on fish status was obtained by a postal inquiry. Fish population losses were most frequent in the most highly acidified region of southern Norway and least common in eastern Fennoscandia. According to the inquiry results, the number of lost stocks of brown trout (Salmo trutta), roach (Rutilus rutilus), Arctic char (Salvelinus alpinus) and perch (Perca fluviatilis) was estimated to exceed 10000. The number of stocks of these species potentially affected by the low alkalinity of lake water was estimated to exceed 11000. About 3300 lakes showed high total phosphorus (> 25 microg L(-1)) and cyprinid dominance in eastern Fennoscandia, notably southwestern Finland. This survey did not reveal any extinction of fish species due to eutrophication. One-third of the lakes had been artificially stocked with at least one new species, most often brown trout, whitefish (Coregonus lavaretus s.l.), Arctic char, rainbow trout (Oncorhynchus mykiss), pike-perch (Stizostedion lucioperca), grayling (Thymallus thymallus), pike (Esox lucius), bream (Abramis brama), tench (Tinca tinca) and European minnow (Phoxinus phoxinus). The number of artificially manipulated stocks of these species in Fennoscandian lakes was estimated to exceed 52000. Hence, the number of fish species occurring in Nordic lakes has recently been changed more by stockings than by losses of fish species through environmental changes such as acidification.     

Lake acidity and mercury content of fish in Darwin National Reserve, Russia

Darwin National Reserve is a protected natural area on the north-west shore of the Rybinsk Reservoir, 350 km north of Moscow. In June 1989, six lakes in the Reserve and the reservoir were surveyed to assess lake acidity and the mercury content of perch, Perca fluviatilis. Five were seepage lakes with no permanent inlets or outlets and one was a drainage lake with both an inlet and an outlet. The seepage lakes were acidic (mean pH 4.6-4.8) and varied in colour from 20 to 200 Hazen units. The drainage lake and reservoir were alkaline (mean pH 8.0-8.1) and colour spanned a similar range. The mean mercury content of perch dorsal epaxial muscle ranged from 0.5 to 1.1 microg g(-1) wet weight in the five acidic lakes and from 0.1 to 0.2 microg g(-1) in the alkaline lakes. Fish mercury content was negatively correlated with lake pH (r=-0.93, P=0.002) if all waters were considered together, and positively correlated with apparent colour (r=0.91, P=0.03) in the seepage lakes.     

Predicting change in fish mercury concentrations following reservoir impoundment

Fish mercury concentrations frequently increase after impoundment of a reservoir. Soil flooding releases organic matter and nutrients, providing food to bacterial communities that methylate inorganic mercury. Methylation and bioaccumulation are the primary pathways for mercury accumulation in fish. We investigated if changes in fish mercury concentrations could be predicted from the change in reservoir size. Data for three fish species, northern pike (Esox lucius), walleye (Stizostedion vitreum), and lake whitefish (Coregonus clupeaformis) from reservoirs in northern Manitoba and northern Quebec were used to evaluate four simple models of change in mercury with change in flooded area. For three additional fish species, all primary carnivores, the preferred model consisted of a single exponential enrichment term. This model successfully predicted two cases not used in model development-one with a large change in area and one with a small change in area. Models with good predictive skill can be developed when the underlying dynamics are known.