An update on the lake acidification mitigation project (LAMP)

Limestone addition is a management tactic to mitigate acidic conditions in lakes. The Lake Acidification Mitigation Project is a long-term integrated project to study the ecological effects of liming and develop the technical information necessary for an effective liming program. Three Adirondack lakes have been limed: Cranberry Pond, a small, 60-day hydraulic residence time fishless lake; Woods Lake, a larger fishless lake with longer residence time (214 days); and Little Simon Pond, a large lake (160 ha) containing populations of brook trout, lake trout and other fishes, with a residence time of 450 days. Woods Lake has been limed twice and the stocked brook trout fishery has been maintained; Cranberry has been limed once and then allowed to reacidify with loss of the stocked fishery. Little Simon Pond showed no ill effects of liming on the extant populations. In general, short term evaluation of ecological effects has shown no major deleterious effects of liming. Methods for predicting reacidification are extremely accurate and use of different limestone particle-size fractions is recommended to achieve longer term in-lake neutralization.     

Changes in the fish community of a limed lake near sudbury,Ontario: Effects of chemical neutralization or reduced atmospheric deposition of acids?

Nelson Lake, a moderately acidic (pH 5.7), metal-contaminated (Cu 22 μg L^?1; Zn 18 ug L^?1) lake, 28 km from the smelters at Sudbury, had a degraded fish community in the early 1970's, with lake trout (Salvelinus namaycush) scarce, smallmouth bass (Micropterus dolomieui) extinct, and the littoral zone dominated by the acid-tolerant yellow perch (Perca flavescens). Liming of the lake in 1975–76 increased pH to 6.4, and decreased metal concentrations. Chemical conditions have remained relatively stable in the 10 yr following base addition. Initially, it appeared that neutralization produced dramatic changes in the resident fish community. Yellow perch abundance declined rapidly after neutralization, lake trout abundance increased to the extent that 3.26 kg ha^?1 were caught in the winter of 1980, and reintroduced smallmouth bass reproduced and established a large population. However, these changes in the fish community can not be directly attributed to liming, as water quality and the sport fisheries of an unlimed nearby lake also improved. Reduced emissions from Sudbury smelters were responsible for improvements in the untreated lake. Recovery of the lake trout population in Nelson Lake appears to have begun prior to liming. Of the lake trout sampled during the 1980 winter fishery, 65.8% were present prior to the chemical treatment. Predation by lake trout was the likely cause of the perch decline. Our results suggest that chemical conditions producing population level responses in fish have abrupt thresholds and that neutralization of lakes above these thresholds may not produce distinguishable effects.     

Ecological consequences of invasive lake trout on river otters in Yellowstone National Park

The introduction of lake trout (Salvelinus namaycush) to Yellowstone Lake in Yellowstone National Park has contributed to a significant decline in the endangered Yellowstone cutthroat trout (Oncorhynchus clarki bouvieri), leading to concern over the persistence of this subspecies but also to piscivorous predators in this community. We assessed the impact of lake trout on a key piscivore, the river otter (Lontra canadensis) in two lakes in Yellowstone National Park. Yellowstone Lake continues to support a native cutthroat trout population, although the recent introduction of lake trout has dramatically impacted the cutthroat trout population. Nearby Lewis Lake has an entirely introduced fish fauna of lake trout, brown trout (Salmo trutta), and Utah chub (Gilia atraria) but lacks cutthroat trout. Analysis of otter scat from Yellowstone Lake implicated trout (lake or cutthroat trout) as a major prey item (57% of scat), whereas stable isotopes identified longnose suckers (Catostomus catostomus) as the primary prey there (58% of diet). By contrast, scat from Lewis Lake implicated minnows, presumably Utah chub, as the primary prey for otters occupying that lake (86%), while stable isotopes implicated brown trout (64%) over both lake trout and Utah chub (36% combined). Our data establish the importance of alternative prey to otters and suggest that lake trout-induced reductions in cutthroat trout may not be catastrophic for otter populations here. These data do not necessarily exonerate lake trout, as their impact on other species, most notably grizzly bears (Ursus arctos) and piscivorous birds, has been documented to be substantial, and further data on the nutritional value of alternate prey are required to confirm or refute a working hypothesis that otter populations will continue to thrive in the face of the lake trout invasion.     

Extreme acidification of a lake in southern Norway caused by weathering of sulphide-containing bedrock

In 1986 Lake Langedalstjenn in southern Norway was a weakly acidified lake with a pH of 5.2–5.6, and an average concentration of SO₄ of 330 μeq L^?1. The total Al concentration varied between 10 and 20 μeq L^?1 (expressed as Al³⁺). The lake supported populations of brown trout and perch and had supplied about 100 people with drinking water until the late 1980's. During 1986–1989, a dramatic change in the water chemistry occurred because of blasting of and weathering of sulphidic gneisses in the watershed. The oxidation of sulphide to sulphate (sulphuric acid) caused an increase in the SO₄ concentration of the draining stream of up to ≈ 4800 μeq L^?1. Weathering and/or cation exchange of Ca and Mg neutralized approximately 52% of the protons from the sulphuric acid production, while about 46% were consumed by mobilization of aluminium and iron. Nevertheless, about 2% of the hydrogen ions from the sulfuric acid were still present, which resulted in a stream pH of 4.0. In the lake, the pH was 4.4, and the concentrations of all major cations and anions were significantly lower than in the heavily affected stream. Mixing of the stream water with lake water, formation of aluminium-sulphate complexes and coprecipitation of Ca may explain the resulting concentrations of major ions in the lake.     

The chemical responses of acidic Woods Lake,NY to two different treatments with calcium carbonate

A study was conducted to compare and contrast two approaches to neutralize an acidic lake by CaCO₃ addition (Woods Lake, Adirondack Mountains, New York, U.S.A.; 42°52′ N, 71°58′ W); direct water column treatment and treatment of both the water column and sediment. The latter treatment strategy was designed to release a slow, diffusive Ca²⁺. flux across the sediment/water interface and thereby delay the rate of reacidification. Both applications satisfactorily increased the acid neutralizing capacity (ANC) and Ca²⁺ of the lake, temporarily buffering the water column pH against acidic inputs from the watershed. The water column/sediment application involved a greater (50% more) dose in order to accomplish sediment treatment and resulted in a prolonged period of positive ANC in the water column. However, both treatments neutralized approximately the same quantity of runoff and equivalence of acidic inputs. Water column/sediment treatment may be an effective approach to help delay reacidification in shallow lakes of moderate hydraulic retention times. Unfortunately, many acidic Adirondack lakes have very short hydraulic retention times (<0.5 yr) and may not be neutralized for adequate durations by either water column or water column/sediment CaCO₃ treatments.     

Base neutralizing capacity of sediments from an acidic lake

The base neutralizing capacity (BNC), or alkalinity consumption, of acidic lake sediments may influence the amount of neutralizing agent required to neutralize a lake if the sediment BNC is large relative to the BNC of overlying waters. The extent ofin situ sediment BNC in acidic Bowland Lake (pH 5.0) was inferred by (1) measuring the loss of Ca-45 to acidic sediments from labeled lake water neutralized with CaCO₃, and (2) measuring exchangeable Ca in sediments collected prior to and following neutralization of Bowland Lake with calcite (CaCO₃). The sediment BNC derived from the Ca-45 radiolabeling experiment was 0.01 mg CaCO₃ g^?1 w wt. The mean losses of Ca-45 from the aqueous phase of neutralized and untreated sediment/water mixtures were not significantly different. The mean pH of both neutralized and untreated mixtures decreased to 4.0 during the incubation, possibly because of oxidation of reduced sediments. Sediment BNC estimates derived from literature data for several lakes may be overestimated because of the inclusion of anoxic sediments containing significant amounts of reduced Fe. There was no significant difference in exchangeable Ca between sediments from untreated Bowland Lake and sediments collected 10 m after whole-lake neutralization indicating that little of the supplied alkalinity had been lost to the sediments. Hence,in situ sediment BNC was probably small in Bowland Lake.     

A Scandinavian model used to predict the reacidification of limed lakes in Nova Scotia and Ontario,Canada

A Scandinavian reacidification model was used to predict the reacidification of three limed lakes in the Sudbury area, Ontario, and Sandy Lake in Nova Scotia. A prediction was made also for the future reacidification of Bowland Lake, 70 km north of Sudbury, limed in 1983. The reacidification of a neutralized lake is dependent on the continuing dilution of the dissolved calcium carbonate, and for a limited period of time, the dissolution of calcite from the bottom. The test of the model on the Canadian laked shows that the model may be a useful tool for lake liming design and planning.     

Primary production in a eutrophic acid lake

Total P concentrations, chlorophyll concentrations, and phytoplankton production were investigated bi-weekly in Tibbs Run Lake, Monongalia County, West Virginia, from March 1977 to March 1978. Mean H⁺ concentration in the lake was 25.1 μeq 1^?1 (pH 4.6). The acidic condition of the lake is attributed to inputs of acid via precipitation (mean H⁺ concentration of the bulk precipitation was 79 μeq 1^?1, pH 4.1), and the low buffering capacity of the watershed (bedrock composition of sandstone). Effect of the watershed is shown by the net retention of imput of P (ca. 26%) and H⁺ (ca. 68%). Total P loading to the lake was 0.495 g P m^?2 yr^?1. The single inflow accounted for 95% of the total loading while bulk precipitation accounted for the remainder. Mean summer chlorophyll concentration was 22.2 mg m^?2. Phytoplankton production expressed volumetrically as aP-vol-x value was 9.78 mg C m^?3 h^?1. Regression analysis indicated that H⁺ do not affect chlorophyll concentrations or phytoplankton production but rather that P limits algal biomass. Trophic status of Tibbs Run Lake based on a P budget model, chlorophyll concentration, and volumetric production all indicate that the lake is meso-eutrophic.     

Recovery from acidification in Lake Örvattnet,Sweden

Lake Örvattnet has been monitored extensively for both chemical and biological variables since 1967. The lake acidified during the 1960's and pH was mostly below 5 throughout the 1970's. Due to the acidification, peat moss (Sphagnum spp.) expanded over the lake bottom and the only surviving fish species was perch (Perca fluviatilis), but it experienced reproduction problems. In the mid 1980's, the Sphagnum cover collapsed, and by 1989 it had almost disappeared. There has been a clear recovery of the perch population. Recovery of the lake is also recorded by diatom assemblages in the lake sediment. Diatom-inferred pH increased from 4.7 to 4.9. The development of measured lake-water pH is unclear, but acid episodes in spring have become less severe. By 1993, atmospheric sulphate deposition had decreased by 30–40% in this area of Sweden compared to the late 1960's. Lake-water sulphate concentrations have decreased by ～30% since the 1960's. Nitrogen deposition has increased over the last decades, but is not yet contributing to lake acidification. No major land-use changes have occurred and changes in hydrology cannot explain the observed changes in chemistry and biology. We ascribe the recent recovery in the lake to reduced deposition of sulphate. In conclusion, Lake Örvattnet has begun to recover from acidification.     

Diatoms as indicators of the rate of lake acidification

Autecological studies of diatoms as pH indicators have opened the way to estimating a lake's past pH on the basis of its diatom species composition and relative abundance. Estimating the rate of lake acidification from its sediment subfossil diatoms is possible when these subfossils can be identified and accurately enumerated in the surface sediments of 20 to 30 lakes. Once this is done the diatoms down the length of the sediment core of one or more of these lakes can be enumerated and the pH inferred at each depth. This technique holds considerable promise in assessing the temporal impact of acid precipitation for acid-sensitive lakes. When loga values were regressed against observed pH for 28 lakes located north of Lake Superior, a significant (P < 0.01) correlation (r = 0.89) resulted. Downcore diatom stratigraphy for one of these lakes indicated that its pH had dropped from 6.2 to 5.2 over the last 20 yr while a second lake had dropped from a pH of 7.1 to 5.2 over the last 30 yr.     

Lake and watershed neutralization strategies

The Experimental Watershed Liming Study (EWLS) evaluated the application of CaCO₃, to a forested watershed to mitigate the acidification of surface water. During October 1989, 6.9 Mg CaCC₃/ha was applied by helicopter to two subcatchments of about 50% (102.5 ha) of the Woods Lake watershed area. The EWLS team investigated the response to treatment of soils (chemistry and microbial processes), vegetation, wetland, stream and lake waters, and phytoplankton and fish, and applied the Integrated Lake Watershed Acidification (ILWAS) model in predicting a watershed treatment duration of up to 50 years. Observations showed a gradual change in pH, acid neutralizing capacity (ANC) and Ca²⁺ in the water column; direct lake additions of CaCO₃ (three different times) were characterized by abrupt changes following base addition and subsequent rapid reacidification. Moreover, the watershed treatment eliminated the snowmelt acidification of the near-shore region of the lake observed during direct lake treatments. Positive ANC water in the tributary and near-shore area improved conditions for fish reproduction and for a viable fish population. Budgets for 12-month periods before and after the watershed treatment showed that the lake shifted from a source of ANC to a sink due to retention of elevated inputs of Ca²⁺ from the watershed CaCO₃ application.     

Dose-response relationships for the addition of limestone to lakes and ponds in the Northeastern United States

A titration based model (DeAcid) has been used to predict treatment dose and times for reacidification for CaCO₃-treated lakes in the Living Lakes, Inc. (LLI) aquatic liming program. Water quality constitutents (pH, ANC and Ca) were used to measure the effectiveness of the dosing model and reacidification rates. Data from 22 lakes or ponds in 5 northeast states have been collected since June 1986. With few exceptions, pH and ANC values ranged from 4.5 to 6.5 and ?30 to +65 ueq L^?1, respectively, in untreated sites and 6.5 to 7.8 and 120 to 300 ueq L^?1, respectively, in sites approximately 30 days after treatment. Changes in Ca concentration levels have been used to evaluate the utility of the dose model for treatment of both inland and coastal waters. For coastal, seepage lakes application of a single-box mass transfer model to observed post-liming changes in ANC and Ca adequately simulates lake response.     

Development of methodology for predicting reacidification of calcium carbonate treated lakes

As part of the EPRI funded Lake Acidification Mitigation Project(LAMP), an ongoing study to evaluate the feasibility of acid lake liming, two models (WAM and ALaRM) have been tested for two calibrated Adirondack Mountain study lakes(Woods and Cranberry). A priori predictions of reacidification rates based on Ca carbonate application and historical hydrological data are presented. WAM (Watershed Acidification Model) is a deterministic model that is capable of simulating the movement of water and chemical constituents through a watershed. WAM generates indata in the form of hydrologic and alkalinity to ALaRM (Acid Lake Reacidification Model), a general mass balance model developed for the temporal prediction of the principal chemical species in both the water column and sediment pore water of small lakes and ponds. A matrix of runs to determine model sensitivity to input loadings, kinetic coefficients, and sediment dosage levels indicate that reacidification of the water column is most sensitive to variation in hydrologic loading followed by variation in sediment dosage levels. Baseline estimates (initial water column alkalinity between 400 to 500 μeqL^?1) indicate that reacidification to near zero alkalinity occurs after a time period equivalent to approximately two to three average hydraulic retention times.     

Effects of liming and acid surface water on the mayfly Leptophlebia vespertina in Lake Hovvatn

The acidified Lake Hovvatn have been limed in 1981, 1987, 1989 and 1991. After the first liming the lake reacidified close to the prelimed condition. The reliming, which started in 1987, was planned to maintain the pH at a relatively high level for the lake. A detailed monitoring of pH and temperature was performed at depth 0.5, 1, 1,5 and 5 m since spring 1993. Quantitative samples of benthic invertebrates were taken in spring and fall in 1977 and regularly at the same seasons from 1981 at depth 0.5, 2, 5 and 10 m. A reference lake, Lille Hovvatn have been sampled with the same procedure since 1988. The acid tolerant mayfly Leptophlebia vespertina responded quickly to the first liming with a 20 times increase in density after a few months. However, the densities rapidly decreased during the first years of reacidification. The lime treatments in 1987 and 1989, resulted in a second peak in density in 1990. After this, the densities have been reduced in spite of generally good water quality in Hovvatn. During fall the density increase was significant at 0.5 m depth in 1990, at 2 m depth in 1989, 1990 and 1992 and at 5 m depth in 1988 to 1990. No significant increase was observed in the limed localities during spring. It is concluded that acid surface water, prior to ice break, affect the food resources to L. vespenina and reduce the population at all depths during spring and in the littoral zone in fall.     

Responses of perch,Perca fluviatilis L., from an acidic and a neutral lake to acidic later

Responses to low pH of perch, Perca fluviatilis, from a naturally acid and a neutral lake were compared by 24 hr exposures to pH 4.6, 4.1 and 3.8 and by 72 hr exposures to pH 4.5. Plasma osmolality and plasma concentrations of Na and chloride decreased in fish from both lakes during acid exposures. Significant differences between the populations were observed at pH 4.1 and 4.5. Hematocrits of the fish from the acid lake increased rapidly and at higher pH compared with those of fish from the neutral lake. This was interpreted as an adaptation to their normal acidic environment, connected with the maintenance of red cell oxygen affinity. The perch from the acid lake maintained their muscle water balance at lower pH better than did the fish from the neutral lake.     

Improved growth in stunted brown trout (Salmo trutta L.) after reliming of Lake Hovvatn,Southern Norway

The chronically acidic Lake Store Hovvatn and the adjoining pond Pollen in southernmost Norway were limed in March 1981. The two locations were stocked with brown trout (Salmo trutta L.) at low and high densities in Hovvatn and Pollen, respectively. After 6 yr of reacidification, the locations were relimed in July 1987. Growth depression during the reacidification process in spite of low fish densities and superabundance of food was observed in Lake Store Hovvatn. Three months after reliming, a substantial growth response was found in trout from Lake Store Hovvatn; Mean annual length increment was 68% higher than that of the preceding year. In Pollen, reliming had no apparent effect on growth. In both populations reliming caused increased swimming activity measured as an increase in CPUE-values. These results show that the growth response to liming depends on population density and food availability. Moreover, the results indicate that the food conversion rate of the trout is negatively affected in acid waters.     

Electrode pH error,seasonal epilimnetic pCO2, and the recent acidification of the maine lakes

Based on laboratory experiments involving standard pH protocols, on statistical analysis of data from the Eastern Lake survey, and on independent field evidence (1985–86), recent (post 1975) electrometric pH measurements for Maine lakes are reevaluated for accuracy. A standard pH protocol used in Maine up until September 1985 can introduce bias of ca. ?0.3 to ?0.4 pH units for representative epilimnetic samples with low CO₂ concentrations (near atmospheric equilibrium). The bias arises out of the very long electrode equilibration times for such samples. A new protocol using these same Ross combination electrodes provides pH estimates accurate within 0.03 pH units for a wide range of northeastern environmental samples, and independent of sample DOC and conductance. The new pH measurements, when paired with the earliest colorimetric measurements from the same Maine lakes, indicate that epilimnetic pH's may have increased significantly (by 0.1 to 0.2 units) since the pioneering lake surveys of G. P. Cooper (1938–44). Earlier, often-cited, reports documenting historical pH reductions in Maine lakes are apparently vitiated by analytical error in many modern electrode pH measurements. The error also affects the calibration of paleo-ecological models of recent lake acidification in New England.     

Metal concentration and calcification of bone of white sucker (Catostomus commersoni) in relation to lake pH

The concentrations of 17 elements were determined in the bone of white suckers (Catostomus commersoni) netted from 5 acid (pH range 4.8 to 5.8) and 2 circumneutral (pH=6.2 and 6.3) lakes in south-central Ontario. The bone Ca:P dry weight ratios were similar (2.0:1) for all fish populations except those of George Lake (pH=4.8) which showed a significantly lower Ca:P ratio (1.9:1, P < 0.05). Magnesium was also lower in the bone of these fish and in fish from 2 other acid lakes. Only bone Ba and S concentrations in the 7 fish populations correlated significantly to lake pH (R=?0.9 and R=?0.8, respectively, P < 0.05). Bone Mn concentrations correlated to dissolved lake Mn concentrations (R=0.8, P < 0.05), and was 7 fold greater in the bone of fish from George Lake and 2 fold greater in King Lake (pH=5.0) fish, vs fish from the 2 circumneutral lakes. Bone Zn was significantly greater in white sucker from George Lake, and tended to be higher in this species from King Lake, compared to all other fish populations. Bone concentrations of Fe, Cu, Ni and A1 showed no apparent trends among the 7 fish populations. Cd, Co, Cr, Mo, V and Be were not detected. The occurrence of a reduced Ca:P ratio coincident with the highest concentrations of Mn, Zn and Ba in the bone of fish from the most acidified environment suggests that increased metal concentrations which occur in surface waters coincident with lake acidification may affect bone calcification.     

Effects of lake liming on phytoplankton production in acidic Adirondack lakes

Phytoplankton community characteristics were monitored prior to and following CaCO₃ addition to two small, highly acidic lakes (Cranberry Pond and Woods Lake) and one larger, less acidic lake (L. Simon Pond). Data were also collected from a control site (Dart's Lake) exhibiting chemical characteristics similar to the pretreatment conditions observed at the experimental sites. In the two small, most acidic lakes, base addition was associated with higher chlorophyll levels during the first summer following treatment. Woods Lake was maintained at a circumneutral pH for 3 yr and exhibited increased phytoplankton abundance throughout the posttreatment period. In contrast, Cranberry Pond reacidified within 1 yr following based addition. Reacidification was accomplished by a decrease in lake chlorophyll levels to pre-treatment levels. At the larger, less acidic lake (L. Simon Pond), liming was associated with lower chlorophyll levels during the first summer after treatment. Reductions in chlorophyll levels at L. Simon Pond reflect the absence of the spring phytoplankton peak and a decrease in phytoplankton growth below the depth of the thermocline. At Cranberry Pond, annual differences in phytoplankton production did not correspond to changes in lake acidity and phytoplankton abundance. Productivity in Woods Lake exhibited an increasing trend during the 3 yr following treatment. Interpretation of treatment effects on productivity was confounded by high between-year variability at the control site.     

Fish species distribution in relation to lake acidity in Ontario

Survey data from approximately 2,900 Ontario lakes were used to examine the distribution of 12 fish species with respect to lake pH. Yellow perch (Perca flavescens), white sucker (Catostomus commersoni), brook trout (Salvelinus fontinalis), pumpkinseed (Lepomis gibbosus), fathead minnow (Pimephales promelas) and redbelly dace (Phoxinus eos) proved to be tolerant of low pH (pH <6.4). Lake trout (Salvelinus namaycush), common shiner (Notropis cornutus), blacknose shiner (Notropis heterolepis), lake whitefish (Coregonus clupeaformis), walleye (Stizostedion vitreum vitreum) and northern pike (Esox lucius) showed limited distribution in low pH lakes (pH <6.4). The limited distribution of lake whitefish and northern pike likely relates to zoogeographic factors rather than their intolerance to low pH. Lake trout and common shiner occur in lakes susceptible to, and receiving high loadings of acidic deposition. However, without historical data, we cannot determine if their decreased distribution in low pH systems is due to intolerance to low pH.