Acid Deposition: Long-Term Trends (1986)

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

8 Fish Population Trends in Response to Surface Water Acidification Terry A. Haines INTRODUCTION The most widely reported consequence of acid deposition is the reduction or elimination of fish populations in response to surface water acidification (Overrein et al. 1980, Haines 1981, Altshuller and Linthurst 1984, Dillon et al. 1984). To demonstrate conclusively that fish population trends in a body of water are related to atmospheric acid deposition, temporal associations between surface water chemistry and fish populations are required It must be shown that the water body formerly supported a viable fish population, either self-sustaining or hatchery-maintained; that one or more fish species formerly present have been reduced or eliminated; that the water body is more acidic now than when f iSh were present and that the increased acid level was not caused by local factors; and that other adverse factors are either absent or unimportant. The number of available data sets meeting these criteria is very small, and most have not been published in the peer-reviewed literature. A larger number of data sets exist that demonstrate spatial associations between surface water chemistry and fish populations. In some cases, laboratory or field experiments serve to confirm the spatial associations between fish species with water chemistry parameters linked to acidification (e.g., pH and aluminum). In other cases temporal data on fish populations are available, but there are no temporal data on water quality; or the converse may be true. Other than temporal association studies with data on both fish populations and water quality, the strongest cases for demonstrating adverse effects of acidification on fish population trends are those that combine spatial data 300 1

301 with one or more of the above types of evidence (Magnuson et al. 1984). These studies are also very scarce. In this chapter I review available North American data sets that purport to relate fish population status or trends to acid deposition. LABORATORY INVESTIGATIONS Laboratory experiments have confirmed that chemical parameters associated with acidification, such as pH, dissolved organic carbon, levels of aluminum, and calcium affect the survival of fish. These laboratory studies have been reviewed in Altshuller and Linthurst (1984) and Haines (1981). Exposing fish to various pH levels reveals intra- specific and interspecific differences in acid tolerance (Tables 8.1 and 8.2). The environment, however, is much more complex than this, and acid tolerance is also affected by other factors. Varying concentrations of toxic trace metals, such as aluminum and cadmium, can act synergistically with acid, while divalent cations, especially calcium, and organic ligands can act antago- nistically with acid and trace metals. In general, reduced pH and increased trace metal concentrations are toxic to fish, and this toxicity is reduced by increased levels of dissolved organic carbon and divalent cations. Toxicity varies in a complex manner with changing propor- tions of the toxic and mediating factors, fish life his- tory stage, and fish species. Given the complexity of the interactions it is generally not possible to predict the response of a fish species population to acidification in the absence of considerable experimental evidence. Such evidence is currently available for only a few fish species, such as brook trout (Salvelinus fontinalis) and white sucker (Catostomus commersoni), although data are now being collected on additional species. F IELD EXPERIMENTS A few researchers have studied the effects of acidification on fish using manipulation experiments. Hall et al. (1980) acidified a small stream in New Hampshire, but the stream did not support a permanent fish population. A few adult brook trout were present in the acidified reach during part of the year, but they did not exhibit morphological symptoms of acid stress.

302 D Cal C) C) - Cal ·_ X Em C) o U: o X U: .~ - U, _' - U) ._ 00 m 04 Ct =. . . Cal ~ . . O _ . . Go at . ~ rid . . O _ ~ 00 O ~ o C) . Cal . V) ct ~ ~ ~ ~ 04 0C ~ . ~ , ~ _ ~ ~ ~ ._, ~ - 0 0 to O ~ 00 O _ _ to _ ( - I _ ~ — O _ 00 0 ~ ' 00 ~ \0 ~ ~ O O O ~ ED ~ _ _ ~ DO ~ CJ~ ~ on I'll c~) Cal (~} ~ Us \~} ~ ~ O _ C - ] I I 00 00 ~ _ ~ _ _ CJ~ — ~t ~ — — — ~ ~ a~ ~ — — V V — — — — ~ ~ ~ ~ v v - ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ E ~ ~ e~~ os o~ E 0 ~ ~ 0 0 0 0 _ O O u~ u~ ~ ~ u~ ~ O O r~ ~ o ~ ~— o —~ c~ ~ ~ ~ — 3 _ ~o,, ~ 8 ~ ~ ~ _ * * * * * * * * * * ~ '_ O ~ 3 ~0 eC O ~ ~ ~ ~, ._ C) - o v o 1— r~ ._ ~ _ _ C — ,,, oOo ~ _ Eo =, ax _ 0 ~ C) ~ ~ C) ~ ~t 5 3 ~ ~ c,4 ,, Ce c ~ ~ c :` V 3 C: 0 3 3 E c ~ :3 ~ ~ m ~ ~ ._ _ ~ ~ =:._.~_ E _ V _ 3 ~ 0 04 _ ._ v _~ _ v~ ~ ~ ~ _ _ c' ~ E =: ~ .c ~ ~ X ~ \O ~ ~ 21 I . ~ ~o ~ ~ ~ s ~ ~ o ~ ~ ~ D ~ ~ ~ ~ "O Z o D V) ~ ~ c~

303 Ct U. Cal .O x En - . ~ o U: LM U. .~ C40 - g V: C) a, Do m C) Ct ~ C) ~ C) ~ ~ ~ ~ ~ of 8 8 ~ ~ ~ _ ~ oo o to . . . . Do o . . . . . . ~ in ~ ~ o ~ ~, o . U~ C~ V) o ~ o o oo o _ _ o o - oo o ~ ° o o ~ ~ ~ o o o ~ ~ ~ '~ oo O O ~D oo ~ ° o o ~ oo o _ ~ o 0 0 0 0 0 _ ~ ~ o o o o _ _ _ _ ~ b4 ~ ~ ~ ~ ~ t4 ~ 8 =° ~ ,, ~c' 8 oo - ° >` 8 ~ ~ o — - — _ ~ _ * * * * * * .. .. .. .. .. ~ .~ e =- e e~ ~ ~ m ~ m ~ ~ ~ ax ~ t— Ct ~ 3 _ ._ s~ - Ct - o - ~4 ._ C~ U3 .= x C) ·- ~ ~ C~ ~ Ce ~ — Ct .Cq ~ ~ ~ ~ _ ~ t~ ;~. (e s ~ ~ ~ o.o ~= ~ U) 3 ~ ~ ~ ~ Ct ~ ~ ~q C) sc . , :r == oo - U, s~ . Ct ~ C) ~ . . . ~ o V,

304 Mills (1984) acidified Lake 223 of the Experimental Lakes Area, Ontario, by direct addition of sulfuric acid. The lake pH was reduced from 6.49 (1976) to 5.17 (1983). Lake trout (Salvelinus namaYcush), white sucker (Catostomus commersoni), fathead minnow (Pimephales promelas), and slimy sculpin (Cottus cognates) were abundant at the start of the experiment, and pearl dace (Semotilus margarita) were rare. Fathead minnow and slimy sculpin abundance declined in 1979 (pH 5.64), lake trout declined in 1980 (pH 5.59), and white sucker declined in 1981 (pH 5.16). Pearl dace first increased rapidly in 1980, following the decline of fathead minnow, then declined in 1982 (pH 5.11). - ~ -I fully reproduced in 1982. NO clan species success- Adults of some species survived DUt were In generally poor condition because of the loss of forage species. It is difficult, however, to extrapo- late these results to areas that receive acidic deposi- tion. Inasmuch as the acid was added directly to the lake, the dissolved organic carbon and trace metal (especially aluminum) concentrations in the lake may not reflect those in lakes in which the watershed also receives acid inputs. SPATIAL ASSOC IATION S South Central Pennsylvania Personnel from the Institute for Research on Land and Water Resources, Pennsylvania State University, have conducted a number of studies in the Laurel Hill area of south central Pennsylvania. In one, researchers conducted a spatial association study in 61 streams on Laurel Hill (W. Sharpe, Pennsylvania State University, personal communication, 1985; Figure 8.1). For 12 of the streams on-site surveys identified significant cultural disturbances in the watershed that were expected to affect water quality, such as agricultural, highway, and industrial discharge, and these streams were therefore excluded from further consideration. The remaining streams were located in watersheds with only minor land use activities. These streams were surveyed for water chemistry and fish population (by electrofishing a 100-m section) and were divided into three groups: those with reproducing fish populations (young-of-the-year fish of nonhatchery origin present), those with remnant fish populations (only 2-year classes), and those with no fish (Table 8.3).

305 79o158 0 5 10 kilometers 4oo 1 5 Ligonier 17 Donega~ ° 2 1 Johnstown ~O~Ddi 3 1 79°1 5' ~3 - - _ Stream O Basin Boundary . 4ooool Reproducing cold water fishery absent (12 Remnant cold water fishery remaining ( 2 year classes ) (4) Confounding cultural impacts (12) Reproducing cold water fishery present (3 FIGURE 8.1 Location and fish population status of 61 streams on Laurel Hill, south central Pennsylvania. Source: W. Sharpe, Pennsylvania State University. The 33 streams that contained reproducing fish populations all were of pH 6.0 or higher, and aluminum concentrations were less than 100 ~g/L. Of the 12 streams lacking reproducing fish populations, only one (Beaver Dam Run, pH 6.22) had a pH above 6.0. Eight of the 12 had pH values less than 5.0 and aluminum

306 TABLE 8.3 Fish Population Status and Water Chemistry of 49 Streams on Laurel Hill, South Central Pennsylvania Fish Population Number Mean pH Mean Alkalinitya Mean Aluminumb Status of Streams (range) (range) (range) ReproducingC 33 6.73 139 21 (6.0-7.2) (27-340) (5-97) Remnants 4 6.36 88 49 (5.7-7. 1) (6-262) (18-100) Not reproducinge 12 4.97 7 454 (4.3-6.2) (0-43) (15-1 ,000) a Units are microequivalents per liter. b Units are micrograms per liter total filterable. C Young-of-the-year fish present. Only two year-classes present. eNo young-of-the-year fish present. SOURCE: W. Sharpe, Pennsylvania State University. concentrations of 400 ug/L or more. The 4 streams with remnant fish populations were chemically quite variable, with pH values ranging from 5.7 to 7.1. Of the streams that contained reproducing fish populations, 10 contained only one species (brook trout), 18 contained two species (brook trout plus mottled sculpin in 17, brook trout plus blacknose dace in 1), 4 streams contained three species (brook trout, mottled sculpin, and brown trout in 2, and rainbow trout in place of brown trout in the remaining 2), and 1 stream contained four species (the three salmonids plus mottled sculpin). Fish population status in these streams is thus generally related to stream chemistry. Although there are exceptions, the more acidic streams are generally those with sparse or no fish population. This does not prove cause and effect, although other explanations involving factors other than acidification have been ruled out. Vermont Langdon (1983, 1984) conducted a spatial association study of fish populations and water chemistry in 29 Vermont lakes. The lakes selected for the fisheries survey were chosen from a larger set of lakes (Clarkson 1982) and represented the lowest alkalinity lakes in the

307 survey. _ Fish were collected with experimental (variable mesh size) gill nets, baited minnow traps, and seines. Two of the lakes were fishless, and 27 contained 1 to 10 fish species. For this report I conducted a stepwise multiple- regression analysis for physical and chemical factors that could potentially affect the number of fish species in these lakes. The variables considered were pH, specific conductance, color, sum of divalent cations (calcium and magnesium), surface area, maximum depth, and elevation. The only significant (p < 0.05) variables were pH-(p = 0.0001), sum of divalent cations (p = 0.026), and surface area (p = 0.028). The regression equation is , ~ ~ log(number of species + 1) = -1.33 + 0.41 (pH) -0.004 (sum of divalent cations) + 0.0004 (surface 2 area); r = 0.62, p = 0.0001, n (number of lakes) = 29. I also compared fish abundance, expressed as catch per day (a standard unit of effort), with the same physical and chemical variables. Similar to the above analysis the three statistically significant variables of fish abundance were pH (p = 0.048), sum of divalent cations (p = 0.006), and surface area (p = 0.070). The regression equation is log(catch per unit effort + 1) = -1.17 + 0.57 (pH) -0.023 (sum of divalent cations) + 0.001 (surface area); r2 = 0.51, p = 0.001, n = 29. Generally, smaller lakes have lower pH and contain fewer numbers of fish species and individuals. Sur- prisingly, the relationship with divalent cations was reversed; larger lakes were lower in divalent cations than expected. Attempts to partition the variability between pH and surface area were unsuccessful. Apparently these two variables are highly correlated. The lakes surveyed were generally not affected by cultural disturbances. Most contained no permanent structures in the watershed, although five contained seasonal dwellings and one contained year-round dwellings. Many of the lakes, however, were stocked with salmonid fish, but only one lake contained stocked salmonids alone. All the lakes were accessible to fishing and, fishing pressure was not believed to be excessive for any lake.

308 The previous presence of fish populations in the two lakes that are now fishless is equivocal. One lake is known to have contained fish previously but may have been affected by beaver activity that blocked fish access to a spawning stream. There are no reliable data to confirm that the other lake ever contained fish, but there are anecdotal records that suggest that it did (Langdon 1983). Maine Haines (1985) conducted a spatial association survey of water chemistry and fish population in 22 lakes in Maine. The lakes were all low in color and contained no human habitation or other recent land disturbance in the watershed. None had been stocked, reclaimed, or otherwise manipulated, and they were sufficiently remote from vehicle access as to make casual introduction of fish species or intensive fishing pressure unlikely. A survey team visted each lake at least three times. The team collected water samples in spring, summer, and fall to analyze for pH, alkalinity, specific conductance, color, and all major ions. Fish populations were surveyed once, in summer, using a standard collecting protocol. Two experimental gill nets (6 ft deep containing five 25-ft long panels of 0.5-, 0.75-, 1-, 1.25-, and 1.5-inch mesh square measure) were set from late afternoon until midmorning the following day. One net was set in shallow water and one in the deepest area. In addition, six standard wire mesh (0.25 inch square measure) minnow traps were set in various locations in the littoral area for the same time period. The team counted and identified the species of all fish collected. A stepwise multiple-regression analysis was conducted following the same procedure used with the Vermont spatial association data. For number of fish species the significant variables were sum of divalent cations (p = 0.045), and surface area (p = 0.042). important variable was pH (p = 0.21). equation (including pH) is log(number of fish species + 1) = -0.659 The next most The regression + 0.003 (sum of divalent cations) + 0.008 (surface 2area) - 0.124 (pH); r = 0.66, p = 0.0002, n = 22.

309 Fish abundance was also analyzed as for Vermont. The only significant variable was sum of divalent cations (p = 0.066). The next most important variable was pH (p = 0.23). The regression equation (including pH) is log(catch per unit effort + 1) = -1.27 + 0.011 (sum of divalent cations) + 0.353 (pa); r2 = 0.53, p = 0.0007, n = 22. The number of significant variables for the Maine data was less than for the Vermont data, but the same variables were identified as most important in both cases. In the Maine lakes the relationship of divalent cations was direct, as expected, rather than inverse. In general, small lakes were more acidic, lower in divalent cations, and contained fewer numbers of species and individuals. Attempts to partition the variability among pH, divalent cations, and area were unsuccessful, apparently because of the high degree of correlation among these variables. Wisconsin Several authors have reported results of fish popula- tion and water chemistry surveys in northern Wisconsin lakes. Rahel and Magnuson (1980) reported no relation- ship between lake pH and number of fish species present in lakes that were similar in size and habitat. Con- versely, the same authors (Rahel and Magnuson 1983) reported that cyprinids and darters were generally absent from Wisconsin lakes of pH ~ 6.2. However, oligotrophic lakes with pH near 7.0 also lacked acid-sensitive fish species because of species interactions and biogeo- graphical factors. Wiener et al. (1984) surveyed fish populations in two groups of lakes in northern Wisconsin. All lakes were seepage lakes, but one group was acidic with pH levels near 5.0 and the other consisted of lakes with near- neutral pH levels. The acidic group contained signifi- cantly fewer fish species than the near-neutral group, and the absent species consisted primarily of cyprinids and darters. However, it is doubtful that these lakes have been acidified by atmospheric deposition. (See Chapter 7.) These data confirm that fish species distribution is affected by water chemistry factors related to acidification but do not demonstrate trends related to atmospheric deposition.

310 TEMPORAL ASSOCIATIONS Ontario Lakes located in the LaCloche Mountain region of Ontario constitute one of the best North American data sets documenting the adverse effects of acid deposition on fish populations, although the major source of acid is certainly the sulfur dioxide emissions from metal smelters located 65 km to the northeast in Sudbury, Ontario (Beamish 1976). Acidification has occurred with suf- ficient rapidity that declines and in some cases extinctions of fish populations have been documented since the late 1960s (Beamish and Harvey 1972, Beamish 1974a,b, Beamish et al. 1975, Harvey 1975, Beamish 1976, Beamish and Van Loon 1977, Harvey and Lee 1982). The acidic deposition is accompanied by increased deposition of certain heavy metals (e.g., copper, nickel), which may also be toxic to fish (Beamish 1976, Beamish and Van Loon 1977). Of the 212 lakes in the LaCloche Mountains, 68 have been surveyed for fish populations (Harvey 1975). Of these, 38 lakes are known or suspected to have had reductions in the number of fish species, and 54 fish populations are known to have been lost (Harvey and Lee 1982). The number of fish species per lake was signifi- cantly (p = 0.005) correlated with lake pH, even after correction for differences in lake area. Explanations for declines in fish populations other than changes in water chemistry related to acid deposition have been ruled out (Altshuller and Linthurst 1984). Adirondack Mountains, New York The first reports of acid-deposition-induced acidifi- cation of lakes and concomitant reductions in fish populations in North America were for the Adirondack Mountain region (Schofield 1965, 1973, 1976a,b). Several recent survey reports (Pfieffer and Festa 1980; Colquhoun et al. 1984) summarized recent data concerning water chemistry and fish populations in lakes within the Adirondack ecological zone. Schofield (1976b) surveyed 40 lakes that had been initially surveyed in the 1930s. In the 1930s three lakes had pH < 5.0 and no fish, and one lake with pH between 6.0 and 6.5 also had no fish. In 1975, 19 of these 40 lakes had pH < 5.0 and had no fish. Two additional lakes with pH between 5.0 and 5.5

311 TABLE 8.4 Summary of Fish Distribution by pH Class for 289 Waters with Concurrent Fish Surveys and Water Chemistry (1975-1982) Fish Population Status pH Classification < 5.0 5.0 to 6.0 > 6.0 All water samples 33 93 163 Waters without fish 13 5 6 Waters with only nontrout species 5 11 35 Waters with trout/salmonid species 15 77 122 SQURCE: Colquhoun et al. (1984~. also lacked fish. Thus 17 lakes apparently lost fish populations during the interval between the 1930s and 1975. Pfeiffer and Festa (1980) concluded that at least 180 lakes in the Adirondack zone had lost fish populations as a result of acidification. However, the data sup- porting this conclusion were not published. Colquhoun et al. (1984) presented a summary of pH and fish population status data for 289 Adirondack lakes (Table 8.4). Baker and Harvey (1985) recently completed an exhaustive analysis of all available water chemistry and fish population status data for Adirondack lakes. To evaluate changes in fish population status over time they developed semiquantitative indicators of population status that incorporate the uncertainties associated with varying sampling techniques. These indicators were then used to test hypotheses by means of ordered classification statistics. Because fish sampling procedures have varied significantly over time, quantitative indicators such as catch per unit effort and population presence/absence cannot be used. Indicators were developed to express stocking data, the quantity of the survey data, the status of the population of 14 common fish species, and the fish community's status as a whole. Each indicator consisted of an ordered classification variable that assigned a numerical scale to qualitative ratings. For example, the quality of fish survey data was assigned a classification variable ranging from O to 8, as follows:

312 Number of Years Rating Surveyed Number of Years between First and Last Survey O O - 1 1 — 2 2 <10 3 2 >10 4 3 <10 5 3 >10 6 4 >10 7 >5 10-20 8 >5 >20 The classifications were made by a single experienced biologist to ensure uniformity and were done blind, i.e. without knowledge of the lake identity. Fish population data were classified with no knowledge of lake chemistry or acidity status. Ordered classifications are used in cases in which different levels of some factor can be recognized but not quantified with a standard unit of measurement on a continuous scale (Kleinbaum and Kupper 1978). Standard statistical tests are available for the classified data. The data assessment involved three phases: (1) evalu- ating fish population status, (2) evaluating possible explanations for observed population declines or losses, and (3) analyzing statistical associations between decline or loss of fish populations and chemical factors related to acidity. Of the 604 lakes that had sufficient data for rating fish community status, 49 (8.1 percent) were rated with high probability as having populations reduced or eliminated as the result of acidification. An additional 64 (10.6 percent) lakes were so rated with marginal probability. A number of lakes did not have previous survey data to demonstrate changes in fish communities over time but currently have either no fish or only sparse populations of acid-tolerant species. The estimated number of Adirondack lakes with fish communities adversely affected by acidification varies from 50 to 200-300, depending on the degree of certainty accepted by the evaluator. The numbers of populations of the 14 species of fish that were investigated and that were apparently lost as a result of acidification are summarized in Table 8.5.

313 TABLE 8.5 Numbers of Populations Classified by Ratings of Population Status and Likely Adverse Effects of Acidification Apparently as a Result No Evidence Apparently of Acidification for toss of Unrelated to Marginal Adequate Species Population Acidification Evidence Evidence Brook trout 409 86 114a 98 Lake trout 68 25 10 8 Rainbow trout 54 34 2 Lake whitefish 28 24 1 Smallmouth bass 92 42 2 Largemouth bass 49 12 1 Chain pickerel 29 12 0 White sucker 336 66 10 Brown bullhead 420 74 13 13 Pumpkinseed sunfish 240 104 7 Yellow perch 185 80 2 Golden shiner 213 83 18 12 Creek chub 147 100 7 Lake chub 6 12 0 aIncludes 71 lakes considered typical brook trout habitat, with only recent survey data available and no fish caught. SOURCE: Baker and Harvey (19851. Fish community status ratings were significantly correlated with lake pH (Figure 8.2). The lake pH levels were not significantly different (p > 0.05) for community status ratings 0, 1, and 2 and for community status ratings 3, 4, and 5, but the two groups were significantly (p < 0.01) different from each other. Although this correlation does not prove cause and effect, it lends support to the hypothesis that declines or losses in fish populations in these lakes are the result of acidification. An investigation of the relationship between lake pH and fish community status (Figure 8.3) indicated that lakes with pH 5.6 or greater had a high probability of having a healthy fish com- munity, whereas lakes with pH below 5.6 had a decreasing probability of having a healthy fish community. Eight New York lakes have been identified for which there are several sources of information about trends in chemistry and biota (Table 8.6). Measurements of water chemistry and observations of fish population status have been made for these lakes, which have been cored for analysis of sediment diatoms and trace elements. (See

314 6.4 6.2 6.0 I 5.8 Q a: UJ 5.6 Ad 6 Lo 5.4 5.2 5.0 4.8 _ t (234) -1 1 |(79) (45) 1 . ( 1 4) ~ ( 1 6) rl I I I I (8) 1 0 1 2 3 4 5 FISH COMMUNITY STATUS FIGURE 8.2 (Opposite) Mean pH (and 95 percent confidence interval) for lakes with fish community status ranked 0 to 5. Mean pH based on average value over 5-year period centered on year of last fish survey. Number of lakes in parentheses. Source: Baker and Harvey (1985).

315 Community Status Key Community Characteristics Community appears "healthy. n Any changes in species composition through time appear random and are apparently due to normal fluctuations. For lakes with only recent survey data available, species diversity is high and/or characteristic for the Adirondacks. Species considered acid sensitive are present. One or two species have apparently declined in abundance and/or disappeared from the lake. These species are not, however, considered particularly acid sensitive (relative to other species in the lake), and it is unlikely that population declines or losses resulted from acidification. One or two species have apparently declined in abundance and/or disappeared from the lake. The species affected may be acid sensitive (relative to other species in the lake); thus the community decline may be a result of acidification. Neither the evidence for loss of populations nor the indications of the potential influence of acidification are particularly strong, however. A number of populations have maintained constant or increased levels of abundance over time. Several species have disappeared from the lake. These species are expected to be acid sensitive relative to other species remaining in the lake, suggesting the possibility that community decline may be a result of acidification. All species or the majority of species have disappeared. Any species left (e.g., brown bullhead, yellow perch) are expected to be acid tolerant. No questions regarding sampling techniques, although the absence of species is confirmed by only one sample. All species have disappeared. No fish caught in two or more samples (in 2 or more years). No questions regarding sampling techniques.

316 1.1 0.8 m 0.7 6 m o Cat 0.5 c, 0~4 us at: cat 0.6 0.3 0.2 0.1 0.0 , ~ .., -1 /' . /' ~ .. ·e ~ · ~ ~ , t— I I I I I I I · / / 4.0 4.4 4.8 5.2 5.6 6.0 6.4 6.8 pH (5 year mean) FIGURE 8.3 Predicted probability (and 95 percent confidence interval (dashed lines)) that status = 1 (fish community "healthy") as a function of mean 5-year pH. Status = 1 for lakes with fish community status rated O or 1. Source: Baker and Harvey (1985). Chapter 9.) Lakes Sagamore and Panther exhibit little or no evidence of acidification. Fish populations reflect only local management activities and are otherwise unchanged. Woodhull Lake chemistry is ambiguous, but the fish population may be experiencing acid stress as demonstrated by the disappearance of smallmouth bass and a recent decline in lake trout abundance. However, brook trout also disappeared, which cannot be attributed to acidification, lake whitefish are still present, and declines in lake trout abundance may reflect changes in hatchery introductions. Honnedaga, Big Moose, Woods, Upper Wallface, and Deep lakes, however, also show con- comitant declines in pH and fish populations. Disappear- ance of fish generally lags behind pH decline by a few years. This may be variation in the dating of pH decline or may result from survival of adult fish for several years after reproduction ceases.

317 TABLE 8.6 Comparisons of Water-Chemistry and Fish-Population Trends with Trends Predicted from Analysis of Sediment Diatomsa Diatom Stratigraphy Present Present/Background Onset of Laker pH pH pH Change Woodhull 5.6 6.0/6.1 No obvious trend Fish Population Histories Sagamore 5.6 6. 3/6.1 Panther 6.2 6.1/6.4 Honnedaga 4.8 5. 2/6. 1 6 species in 1931: X spe- cies collected plus 3 re- ported in 1954. X species in 1974-1 9X 1: brook trout last collected 1954 and smallmouth bass last col- lected 1974, but lake trout and lake whitefish still present. No obvious 9 species in 1933, 8 spe- trend cies in 1976: no changes . . . In any mayor species. Fish population eradi- cated in 195 1 and re- stocked: no changes in fish survival. Lake trout became rare and emaciated in early 1950s, last collected in 1954: brook trout de- clined in mid-1960s and were rare by the mid- 1 970s. Lake trout reported pres- ent previously but never collected; formerly known for large brook trout, brook trout abun- dant in 1961. scarce since 1966, and all of hatchery origin; golden shiner last collected in 1966. 10 species in 1948, 5-6 species since 1962; smallmouth bass and lake whitefish last collected in 1948; lake trout last col- lected in 1966. Brook trout present, mod- erately abundant in 1954, absent in 1964; no survi- val of stocked fish since 1964. Brook trout moderately abundant in 1963-1968. absent in 1975. Slight steady decline Unclear Woods 4.7 4.8/5.2 Not dated Big Moose 4.9 4.9/5.8 ~ 1950 Deep 4.7 4.8/5.0 1940-1950 Upper 5.0 4.7/5.1 1945-1955 Wallface a Data are also available for a ninth lake. Little Echo, but because it is a colored bog-pond, it is not considered here. b Background pH values determined from pre-1850 estimates. SOURCES: Diatom data are from Chapter 9 of this report; water-chemistry and fish- population data are from Baker and Harvey (1985) and D. Webster, Cornell University, . . persona communication.

318 Although Woods, Deep, and Upper Wallface lakes experienced relatively small decreases in pH (0.4, 0.4 units, respectively), the sensitivity of fish to acidity increases markedly over the range of pH from about 5.2 to 4.8. This contention is supported by the observation that many lakes in New York and New England with pHs of 5.0 to 5.2 currently support fish, while lakes with lower pHs generally do not. A contributing factor may be the increase in dissolved aluminum, a metal toxic to fish, at pH values lower than about 5 owing to the decrease in dissolved organic chemical species that can effectively bind (and thus immobilize) aluminum at higher pHs (Davis et al. 1985). South Central Pennsylvania An intensive study of precipitation chemistry, stream chemistry, and fish populations was conducted during February 1981, on four streams in the Laurel Hill, Pennsylvania, area, three of which were included in the survey discussed in the section of this chapter on spatial associations (Sharpe et al. 1984). Three of the streams (McGinnis Run, Linn Run, and Card Machine Run) were acidic, with pH declining that month to <5.0 (Figure 8.4), and contained only a few stocked trout. One stream (Wildcat Run) was well buffered (minimum pH 5.6) and contained self-sustaining populations of mottled sculpin, brook trout, and rainbow trout. Historical fisheries data document the presence of brook trout in three of the streams (Wildcat, McGinnis, and Card Machine) before 1930 and the survival of stocked fish in Linn Run in 1932. Fish kills at fish-rearing facilities on two of the acidic streams (Card Machine Run and Linn Run) were recorded in the 1960s and of adult stocked fish in all three streams in the 1970s. Two of the acidic streams (Linn Run and McGinnis Run) contain a spruce bog in the headwaters, but the bog covers less than 1 percent of the area of either stream basin, and comparing main- stream (bog influenced) with tributary stream (not bog influenced) chemistry indicated no significant differences attributable to organic acids. An in situ fish toxicity bioassay was conducted on one acidic stream (McGinnis Run) and on the nonacidic stream (Wildcat Run) (Sharpe et al. 1983). Trout fry survived only 4 to 9 days in the acidic stream but lived for the duration of the experiment (36 days) in the nonacidic

319 6.8 6.4 6.0 4.8 j 4.4 _ 6.8~ 6.4: 6.0: 5.6 5.2 4.8 1 4.411 I Q Wiidcat Run 1\ I 1 1 —;~k Discharge pH l %. , __'~————~————_ +_ _ _— I 10 12 14 16 18 20 February McGinnis Run _` r\ 1 1 22 ^' ,, `` \ J '' '` 1 ' - ' 1 ----1~ ~=; f=~= ~o 10 12 14 16 18 20 22 24 _ 5 _ 4 _ 3 _ 2 68 L 6.4 _ 6.0 _ 5 6 _ _ 5.2 4.8 February Linn Run ~ ~W ,~_! 'J \ ': _ 4.4 ___~-l —-—~ -1 ~- 10 12 14 16 18 20 22 24 February 6.8 6.. 6.0 5.6 5.2 4.8 4.4 1 ~3 2 1 Card Machine Run _ 5 _ 4 _ 3 t —-—J—~ ~—- - -—H—- - - -t-~l I ~ O 10 12 14 16 18 20 22 24 February DATE - v, J - LU G I 4 ~ FIGURE 8.4 Changes in pH and discharge (in liters per second per hectare) over time for four streams on Laurel Hill, south central Pennsylvania. Source: Sharpe et al. (1984).

320 TABLE 8.7 Results of 36-Day In Situ Bioassay for Brook, Brown, and Rain- bow Trout Fry in Two Streams on Laurel Hill, South Central Pennsylvania Species Survival Time (Days) Low pHa High pHb Rainbow fry 4 Brown fly Brook fly aLow-pH stream: pH, 4.8-5.9; aluminum concentration, 0.18-1.1 mg/L. bHigh-pH stream: pH, 6.1-7.0; aluminum concentration, 0.01-0.14 mg/L. SOURCE: Shape et al. 1983. 9 9 36 36 36 stream (Table 8.7). These results confirm that the streams that are now without fish cannot support viable fish populations because of the acidic conditions. Inasmuch as these streams formerly did support viable populations (either self-sustained or long-term survival of hatchery fish) it is concluded here that acidification has caused the disappearance of fish from these streams. North Central Massachusetts A temporal association study of water chemistry and fish species distribution was conducted in tributaries to the Millers River, Massachusetts (Figure 8.5) (D. Halliwell, Massachusetts Division of Fisheries and Wildlife, personal communication, 1985). Fifteen streams that had been surveyed in 1953 and 1967 were surveyed again in 1983 (Table 8.8). Fish surveys were conducted by using a fish toxicant (rotenone) in 1953 and by electrofishing in 1967 and 1983. Species were identified but individuals were not enumerated in 1967. Stream pH was measured calorimetrically in 1953 and 1967 and electrometrically in 1983. Streams with pH > 6.0 in 1983 generally had relatively stable fish communities and pH over time (Table 8.8). One stream, Riceville Brook, lost three species over the study interval, but the other streams lost no more than two species. Of the streams that had pH < 6.0 in 1983, most lost three or more species, although one (Osgood) was unchanged and one (Mormon Hollow) lost only two species. These streams generally had substantial pH reductions over time (>1 unit), with most of the decline occurring between 1953 and

321 - 1 ___' J tar U. - _~%,%\,,1 L:.'' :. _:>'' ',~''4,~ - lr 1~- ,r ,_ \ - . ~ u ~ ~ ~ ,, ~ _ 1 1 ~ 1 C 3 at ¢': ill-' A_ I `8A!d i \ U] a' U] US S U] to X Sot to ~5 U] 3 a' .,, ~q .,, o oo C' H

322 ._ Cal - .~ o U. ._ Ct ._ A: ._ so Ct U) ._ Cal V) .~ o so By o .~ o U: U) ~ o U) ._ o V Ct Do . oo m t: ~Z ex: 3 ~ - ° ~ m m S°- o m Eb C,3 0 Ct¢t m ~ .~ -.~ =: ~ m° ~ 00 00 r~ ~o 00 00 V~ C~ 00 Ct _ ~ I I ~ 00 _ c~ + T T — ~-S~ c— ++ + + ~ ~ ~ 00 ~ ~ _ ~ ~D— ~ O O _ _ ~ + + + + + + + ~ ~ ~ u~ oo oo ~ ~ ~ ~ ~ - - - ~ ~ ~ c~ ~ ~ ~ - - = ~ ~o Mo ~ - + + ++ o ~ - ~ ~ ~ u~ ~ ~ o ~ ~ r— 1— - , ~ oo — — — o~` ~ — - ) - c~ ~ c~ - - oo ~ - t - - oo - ) <) =' t— - ~ ~ - ~o ++ ++ ~ o - , - , ~ _ ', ~ ~ ~ ~t ~ ~D ~ — - —1 1 0 0 ~ C~ — t_ oo T O C~ — ~ r— O C~ — 00 — ~ _ _ + +++++ O~ ~ ~S, —~ — ~ O ~ ~ ~ ~ O ~ _ - , _ ~ — ~ C~ ~ ~ O ~ — _ ~ ~ \~C) — 2 ~ O ~ . O ~C ~ 5 ~ e ° _ ~V ~

323 a .= Cot - U3 o o ._ , oo oo oo oo Cot Do - Ct V) _ + ~ + _ ~ ~ oo ~ oo ~ _ To — Cal - _ oo o — - o~ o _ + + + + _ ~ ~ oo oo on _ o o + UP — + + - - - ~ - ~ ~ - + _ o or _ in, _ ~ _ ~ _ ~ o oo ~ ~ _ S o ~ Cal ~ A O ~ C

324 Us Cal U. I.., — Cal Z o o ~ C`3 m so ~ .= _` o C) _' 00 00 m Cal =o 00 00 00 Us 00 ~ + Us - E ~ O Cal ~ ~ ~ ~ ~ ~ Do ~ Rae ~ ~ ~ ~ ~ ~ . . . . . . . . . . . . . . . Us us o 0 ~ ~ 0 us 0 0 1 Vet ~ ~ Us ~ ~ ~ 1 If ~ 1 ~ ~ ~ 1 ~ ~ ~ ~ ~ ~ ED ~ ~ ~ ~ ~ ~ ~ <~~ (~ ED C<) ~)—) ~ ~— ~ ~ ~ ~ O O == 1 — 1 =~= 1 ~ 1 °° oo - oo ~ ~ ~ ~ ~ ~ o oo ~ o ~ ~ - - - c~ - ~ ~ ~ - - ~o oo + cr oo - ~ 00 C`1 ~ 1 oo ~ 3 c,, ~ ~ o A ~ ~ ~ Y ~ 3 E .= C~ Cq .= e~ . _ Ct U) s~ s: _ o~ C~ Ct 3 C~ C) ._ r' - .° Ct o . ~S U) o C~ ._ C) C~ :: ._ 0C 0~= Ct ~ ._ ·_ _ ~ ._ ~_ C~ S~ C~ ~.O C) ~ ~S ;> U ._ . Cd ~ ~ O ._ U. Ct - U3 o C~ ._ U. e~ ._ .= C~ VO ~: Ct C~ .^ Ct - ~0 _ _ - ' U) O ~ U: C~ (D Ct .o ~ ~ ~ O _ ~ - - c, 3 .o.3 e~ ~ ~— ~ C~ ~ ~ .= ~ ~ =,~ .~= ~ o '~ U) s~ _ ~ . ~ Ct 't: ~ X Z ~ ._ D C} .0 U, ._ ·fi . . o V)

325 1967. Inasmuch as pH was measured calorimetrically both times, the difference seems unlikely to be a result of methodological differences. (See Chapter 7.) Two streams, Wilder and Templeton, had pH < 5.0 in 1983 and had lost all fish species by 1967. These streams were of different character than the others surveyed, being of lower gradient (i.e., drop in elevation per stream mile), and higher color (organic acids), and con- taining primarily warm-water fish species. Nevertheless, pH apparently has declined and fish communities have been lost in these streams. There is no apparent source of Acid other than atmospheric deposition. Nova Scotia Watt et al. -(1983) compared temporal trends in angler harvest of Atlantic salmon with current water chemistry in 22 rivers in Nova Scotia. Angler harvest data were available from 1936 to the present and were nearly continuous. Earlier data exist, but the methodology used is not comparable with that used subsequent to 1936. Previous (1954-1955) water chemistry data of high quality (multiple pH values measured electrometrically) were available for five rivers. Twelve of the r i Ye c now Hal pH > 5.0 (Table 8.9). Three of these have previous pH data available; two (St. Marys and Musquodobit) have not changed significantly, and one (Medway) declined 0.8 unit. Ten of the twelve rivers had no significant change in salmon harvest, one (Gold) increased significantly, and one (Musquodobit) decreased significantly although its pH did not change appreciably. Ten of the rivers now have pH < 5.0. Two of these (Clyde and Tusket) have previous pH data and both declined in pH, but statistical tests of the decline could not be performed because of inadequate data. (There is only a single historical value in one case, and there was a change in sample collection site in the other.) Nine of the ten rivers with low pH exhibit a significant decline in angler harvest of salmon, which generally began in the 1950s; the other's decline (Lisomb) was not significant. The rivers included in this comparison were carefully selected to eliminate any that were affected by other pollution sources, fish stocking, or change in physical barriers to fish migration. It is presumed that fish from rivers with geographical proximity have similar migratory patterns and are subjected to similar levels ~ ~ .. ~ ~ ~

326 oqleH sanest ably ~U!N qruoos!q ~~snL u~o~oual~e~ u, PI ._ o Q cd so ct 4_ u, 50 8 - .O Cal Car .s C: Do m PHI :) alpp!W 9!8U~ tuel~uI used Keypad toqoponbsny~ l!~0d !~ qleH d!9S (ppona juntas cuno:[ ullo~nG u°d lo nemodse~ P1°O sped IS cry to ~ _ On . . 1 ~ - ~ _ ~ 0 00 0 00 ~ car in, ~ _ ~~D c car cry 0 t_ O At D Hi 1 - ~0 ~ C~ . ~ O O O . . to 1 ~ O . . _ cry 1 is) V) ~ 1 1 1 ~ . . — cry 1 ~ Car ~ C' O ~ 1 Us ' ~ ~ c ' 1 + ~ ~ c~ ~ er c c, c~ ~ c ' ~ o ~ o ~ 1 o o = o - + c~ ~D ~ '~ o ~ c~ c~ L~ e ~ ~ ~ 'e u, ~ - e e ~ - e~ - - e c: o cn ~ o o c~ c~ e 'e ) C) . ~D ~ ~ o ~ ~ v: ) ~ ~z z ~ ~o ~ c~ o - - ~ c~ - ;^ ~ - ~ - c) · ~ ~ - oC · . o 0 ;o° 0 L~ ~ o o o o c~ x v v v v ~ cL, ~ s:~ o ~ ~ ~ 04 s v, - ~ oo a~

327 and sources of mortality away from the home rivers Other studies indicate a significant mortality of early feeding fry of Atlantic salmon in a Nova Scotia river with pH < 5.0 (Lacroix et al. 1983) SUMMARY . . The hypothesis that increased acidity of surface waters caused by long-range atmospheric transport has reduced or eliminated fish populations in northeastern North America was evaluated by examining fishery survey data. The number of statistically valid data sets located was remarkably low. The strongest evidence in support of the hypothesis consists of temporal association data from Adirondack Mountain lakes and Nova Scotia rivers. Both data sets clearly demonstrate declines in acid-sensitive fish species populations over the past 20 to 40 years. Limited water chemistry data indicate that the lakes and streams in question are currently more acidic than they were previously, and fish population status is clearly correlated with present pH. Waters that are now acidic (pH < 5.0) support few or no fish populations. A temporal association study in Massachusetts streams provides similar results, although the number of streams is small and the data are of lesser quality. The LaCloche Mountain area data set is of high quality but does not reflect effects from long-range transport. The remaining data sets consist largely of spatial associations of surface water chemistry and fish popula- tions, supported in some cases by temporal association data or field experiments. Data from Pennsylvania streams, Vermont lakes, and Maine lakes demonstrate that fish population status is related to present water chemistry. Generally, waters with summer pH less than 5.0 to 5.5 support few or no fish populations. Limited data suggest that at least some of these waters formerly supported fish. Limited field experimental data demonstrate that fish will not now survive in these acidic waters and that addition of acid to surface waters will eliminate fish populations. Although these data demonstrate a relationship between surface water acidity and fish population status, they cannot be used to determine the source of the acidity. Critics have suggested that other factors could reduce or eliminate fish populations. These factors include chemical pesticides use, change in fish hatchery

328 production, change in angler pressure, and increased beaver activity Direct effects of human activities (obstruction of fish migrations by dams, degradation of water quality by agricultural and urban runoff, municipal sewage, and industrial wastes, for example) greatly reduced fish populations in accessible waters in the colonial and post-Civil War periods. However, remote lakes were not directly affected by these factors, and while lumbering and subsequent burning of the watersheds undoubtedly affected fish populations in less accessible lakes, these factors generally were most important in the late 1800s and early 1900s. Concern for declining fish resources resulted in developing artificial propagation of fish in hatcheries and led to the founding of the American Fisheries Society to advance knowledge concerning fish resources (Kendall 1924, Thompson 1970). Early attempts to supplement fish populations by introducing hatchery-reared fish generally failed and fell into disfavor, however. For example, Smallwood (1918) documents the failure of hatchery introduction in Lake Clear, New York. Lake Clear now has a pH > 7.0 and contains at least five species of fish (Colquhoun et al. 1984). Beaver activity can either enhance or degrade fish populations, depending on the particular circumstances. Beaver were reintroduced into the Adirondack Mountains of New York beginning in 1905 and by the 1920s had reached population densities high enough to raise concern about beaver damage to various resources (Johnson 1927). Chemical pesticides have been used in remote areas of northeastern North America for control of spruce budworm and blackfly populations, with detrimental effects on fish populations (Burdick et al. 1964, Anderson and Everhart 1966, Elson 1967, Kerswill and Edwards 1967, Locke and Havey 1972). Organochlorine compounds are generally no longer used for these purposes, and most affected fish populations have recovered (Dean et al. 1979). Analysis of brook trout from a series of remote lakes of varying pH in northern New England failed to detect significant organochlorine residues in any fish (Haines 1983). It is impossible to rule out factors other than acidification as important in the decline or loss of fish populations except by intensive, case-by-case investiga- tions. Such investigations have seldom been made, and in fact the quantity of data on which to base estimations of trends in fish populations is exceedingly sparse. How-

329 ever, a few such data sets do exist, and in the cases cited here the involvement of factors other than acid~- fication in the decline of fish populations appears to be negligible. REFERENCES Altshuller, A. r and R. Linthurst, eds. 1984. The Acidic Deposition Phenomenon and Its Effects: Critical Assessment Review Papers. Vol. II. Effects Sciences. - EPA-600/8-83-016BF. U.S. Environmental Protection Agency. Anderson, R., and W. Everhart. 1966. Concentrations of DOT in landlocked salmon (Salmo salar) at Sebago Lake Maine Trans. Am. Fish. Soc. 95:160-164. Baker, J. 1981. Aluminum toxicity to fish as related to acid precipitation and Adirondack surface water quality. Ph.D. dissertation. Cornell University, Ithaca, N.Y. Baker, J., and H. Harvey. 1985. Critique of acid lakes and fish population status in the Adirondack region of New York State. Draft final report for NAPAP Project E3-25. U.S. Environmental Protection Agency. Beamish, R. 1972. Lethal pH for the white sucker, Catostomus commersoni (Lacepede). Trans. Am. Fish. Soc. 101:355-358. Beamish, R. 1974a. Growth and survival of white suckers (Catostomus commersoni) in an acidified lake. J. Fish. Res. Bd. Can. 31:49-54. Beamish, R. 1974b. Loss of fish populations from unexploited remote lakes in Ontario, Canada as a consequence of atmospheric fallout of acid. Water Res. 8:85-95. Beamisn, R. 1976. Acidification of lakes in Canada by acid precipitation and the resulting effects on fishes. Water Air Soil Pollut. 6:501-514. Beamish, R., and H. Harvey. 1972. Acidification of the La Cloche Mountain lakes, Ontario, and resulting fist mortalities. J. Fish. Res. Bd. Can. 29:1131-1143. Beamish, R., and J. Van Loon. 1977. Precipitation loading of acid and heavy metals to a small acid lake near Sudbury, Ontario. J. Fish. Res. Bd. Can. 34:649-658. Beamish, R., W. Lockhart, J. Van Loon, and H. Harvey. 1975. Long-term acidification of a lake and resulting effects on fishes. Ambio 4:98-102.

330 Brown, D. 1981. The effects of various cations on the survival of brown trout, Salmo trutta at low pHs. J. Fish. Biol. 18:31-40. Burdick, G., E. Harris, H. Dean, T. Walker, J. Skea, and D. Colby. 1964. The accumulation of DOT in lake trout and the effect on reproduction. Trans. Am. Fish. Soc. 93:127-136. Clarkson, B. 1982. Vermont acid precipitation program winter lake surveys 1980-1982. Vermont Department of of Water Resources and Environmental Engineering, Montpelier. Colquhoun, J., W. Kretzer, and M. Pfeiffer. 1984. Acidity status update of lakes and streams in New York State. New York State Department of Environmental Conservation Report WM (P-83(6/84)). Albany. Craig, G., and W. Baksi. 1977. The effects of depressed pH on flagfish reproduction, growth and survival. Water Res. 11:621-626. Davis, R., D. Anderson, and F. Berge. 1985. Loss of organic matter, a fundamental process in lake acidification: paleolimnological evidence. Nature 316:436-438. Daye, P., and E. Garside, 1975. Lethal levels of pH for brook trout, Salvelinus fontinalis (Mitchill). Can. J. Zool. 53:639-641. Daye, P., and E. Garside. 1976. Histopathologic changes in surficial tissues of brook trout, Salvelinus fontinalis (Mitchill), exposed to acute and chronic levels of pH. Can. J. Zool. 54:2140-2155. Dean, H., J. Skea, J. Colquhoun, and H. Simonin. 1979. Reproduction of lake trout in Lake George. N.Y. Fish Game J. 26:188-191. Dillon, P., N. Yan, and H. Harvey. 1984. Acidic deposition: effects on aquatic ecosystems. CRC Critical Reviews in Environmental Control 13(3):167-194. Dively, J., J. Mudge, W. Neff, and A. Anthony. 1977. Blood Po2, Pco2 and pH changes in brook trout (Salvelinus fontinalis) exposed to sublethal levels of - acidity. Comp. Biochem. Physiol. 57A:347-351. Edwards, D., and T. Gjedrem. 1979. Genetic variation in survival of brown trout eggs, fry, and fingerlings in acidic water. Research Report 16, Acid Precipitation-- Effects on Forest and Fish Project. Sur Nedb~rs Virkning Pa Skog Og Fisk (SNSF), Aas, Norway.

331 Edwards, D., and S. Hjeldnes. 1977. Growth and survival of salmonids in water of different pH. Research Report 10, Acid Precipitation--Effects on Forest and Fish Project. Sur Nedb~rs Virkning Pa Skog Og Fisk (SNSF), Aas, Norway. Elson, P. 1967. Effects on wild young salmon of spraying DOT over New Brunswick forests. J. Fish. Res. Bd. Can. 24:731-767. Falk, D., and W. Dunson. 1977. The effects of season and acute sub-lethal exposure on survival times of brook trout at low pH. Water Res. 11:13-15. Haines, T. 1981. Acidic precipitation and its consequences for aquatic ecosystems: a review. Trans. Am. Fish. Soc. 110:669-707. Haines, T. 1983. Organochlorine residues in brook trout from remote lakes in the northeastern United States. Water Air Soil Pollut. 20:47-54. Haines, T. 1985. U.S. Fish and Wildlife Service, Orono, Maine, Unpublished data. Hall, R., G. Likens, S. Fiance, and G. Hendrey. 1980. Experimental acidification of a stream in the Hubbard Brook Experimental Forest, New Hampshire. Ecology 61:976-989. Harvey, H. 1975. Fish populations in a large group of acid-stressed lakes. Verh. Int. Ver. Limnol. 19:2406-2417. Harvey, H., and C. Lee. 1982. Historical fisheries changes related to surface water pH changes in Canada. Pp. 45-55 of Acid Rain/Fisheries, R. Johnson, ed. Bethesda, Md.: American Fisheries Society. Johnson, C. 1927. The beaver in the Adirondacks: its economics and natural history. Roosevelt Wild Life Bull. 4:501-641. Johnson, D. 1975. Spawning behavior and strain tolerance of brook trout (Salvelinus fontinalis) in acidified water. M.S. thesis. Cornell University, Ithaca, N.Y. Kendall, W. 1924. The status of fish culture in our inland public waters, and the role of investigation in the maintenance of fish resources. Roosevelt Wild Life Bull. 2(3):205-351. Kerswill, C., and H. Edwards. 1967. Fish losses after forest sprayings with insecticides in New Brunswick, 1952-62, as shown by caged specimens and other observations. J. Fish. Res. Bd. Can. 24:709-729. Kleinbaum, D., and L. Kupper. 1978. Applied Regression Analysis and Other Multivariable Methods. Boston, Mass.: Duxbury Press.

332 Kwain, W. 1975. Effects of temperature on development and survival of rainbow trout, Salmo gairdneri, in acid waters. J. Fish. Res. Bd. Can. 32:493-497. Lacroix, G., D. Gordon, and D. Johnson. 1983. Water chemistry and ecology of fish populations in streams of the Westfield drainage in Nova Scotia. Pp. 1-10 of 1983 Workshop on Acid Rain, R. Peterson and H. Hord, eds. Can. Tech. Rep. Fish. Aquat. Sci. 1213. Langdon, R. 1983. Fisheries status in relation to acidity in selected Vermont lakes. Montpelier, Vt.: Vermont Agency of Environmental Conservation Report. Langdon, R. 1984. Fishery status in relation to acidity in selected Vermont lakes, 1983. Montpelier, Vt.: Report of the Department of of Water Resources and Environmental Engineering, Agency of Environmental Conservation. Lloyd, R., and D. Jordon. 1964. Some factors affecting the resistance of rainbow trout (Salmo gairdneri Richardson) to acid waters. Int. J. Air Water Pollut. 8:393-403. Locke, D., and K. Havey. 1972. Effects of DOT upon salmon from Schoodic Lake, Maine. Trans. Am. Fish. Soc. 101:638-643. Magnuson, J., J. Baker, and F. Rahel. 1984. A critical assessment of effects of acidification on fisheries in North America. Phil. Trans. R. Soc. London Ser. B 305:501-516. McDonald, D., H. Robe, and C. Wood. 1980. The influence of calcium on the physiological responses of the rainbow trout, Salmo gairdneri, to low environmental pH. J. Exp. Biol . 88 :109-131 . Menendez, R. 1976. Chronic effects of reduced pH on brook trout (Salvelinus fontinalis). J. Fish. Res. Bd. Can. 3 3: 118-123 . Mills, J. . 1984. Fish population responses to experimental acidification of a small Ontario lake. Pp. 117-131 in Early Biotic Responses to Advancing Lake Acidification. G. Hendrey, ed. Boston, Mass.: Butterworth. Mount, D. 1973. Chronic effect of low pH on fathead minnow survival, growth and reproduction. Water. Res. 7: 989-993. Overrein, L., H. Seip, and A. Tollan. 1S80. Acid precipitation--Effects on forest and fish. Research Report 19, Acid Precipitation--Effects on Forest and Fish Project. Sur Nedb~rs Virkning Pa Skog Og Fisk (SNSF), Aas, Norway.

333 Packer, R., and W. Dunson. 1972. Anoxia and sodium loss associated with the death of brook trout at low pH. Comp. Biochem. Physiol. 41A:17-26. Pfeiffer, M., and P. Festal 1980. Acidity status of lakes in the Adirondack region of New York in relation to fish resources. New York State Department of Environmental Conservation Report FW-P168 (10/80) Albany, N.Y. Rahel, F., and J. Magnuson. 1980. Fish in naturally acidic lakes of northern Wisconsin, U.S.A. Pp. 334-335 of Ecological Effects of Acid Precipitation, D. Drably and A. Tollan, eds. Acid Precipitation-- Effects on Forest and Fish Project. Sur Nedb~rs Virkning Pa Skog Og Fisk (SNSF), Aas, Norway. Rahel, R., and J. Magnuson. 1983. Low pH and the absence of fish species in naturally acidic Wisconsin lakes: inferences for cultural acidification. Can. J. Fish. Aquat. Sci. 40:3-9. Robinson, G., W. Dunson, J. Wright, and G. Mamolito. 1976. Differences in low pH tolerance among strains of brook trout (Salvelinus fontinalis). J. Fish. Biol. 8:5-17. Schofield, C. 1965. Water quality in relation to survival of brook trout, Salvelinus fontinalis (Mitchill). Trans. Am. Fish. Soc. 94:227-235. Schofield, C. 1973. The ecological significance of air-pollution-induced changes in water quality of dilute-lake districts in the northeast. Trans. N.E. Fish Wildl. Conf. Pp. 98-111. Schofield, C. 1976a. Lake acidification in the Adirondack Mountains of New York: causes and consequences (abstract). P. 477 in Proceedings of the First International Conference on Acid Precipitation and the Forest Ecosystem, L. Dochinger and T. Seliga, eds. General Technical Report NE-23. U.S. Department of Agriculture Forest Service. Schofield, C. 1976b. Acid precipitation: effects on fish. Ambio 5:228-230. Sharpe, W., W. Kimmel, E. Young, and D. DeWalle. 1983 In-situ bioassays of fish mortality in two Pennsylvania streams acidified by atmospheric deposition. Northeast Environ. Sci. 2:171-178 Sharpe, W., D. DeWalle, R. Leibfried, R. DiNicola, W. . Kimmel, and L. Sherwin. 1984. Causes of acidification of four streams on Laurel Hill in southwestern Pennsylvania. J. Environ. Qual. 13:619-631.

334 Smallwood, W. 1918. An examination of the policy of restocking the inland waters with fish. Am. Naturalist 52:322-352. Swarts, F., W. Dunson, and J. Wright. 1978. Genetic and environmental factors involved in increased resistance of brook trout to sulfuric acid solution and mine acid polluted waters. Trans. Am. Fish. Soc. 107:651-677. Thompson, P. 1970. The first fifty years--the exciting ones. Pp. 1-2 of A Century of Fisheries in North America. N. Benson, ed. Special Publication 7. Washington, D.C.: American Fisheries Society. Watt, W., C. Scott, and W. White. 1983. Evidence of acidification of some Nova Scotian rivers and its impact on Atlantic salmon, Salmo salar. Can. J. Fish. . _ Aquat. Sci. 40:462-473. Wiener, J., P. Rago, and J. Eilers. 1984. Species composition of fish communities in northern Wisconsin lakes: relation to pH. Pp. 133-145 of Early Biotic Responses to Advancing Lake Acidification, G. Hendrey, ed. Vol. 6 of Acid Precipitation Series. Boston, Mass.: Butterworth.