Geographic and temporal variation in the incidence of melanism in peppered moth populations in America and Britain

Journal of Heredity, Sep 1998

Industrial melanism in the peppered moth (Biston betularia is the classic example of observable evolution by natural selection. The phenomenon has been most thoroughly documented in Britain, where once rare melanic phenotypes increased in frequency in concert with increased levels of atmospheric pollution generated by industrial development. Populations of the American subspecies of the peppered moth (B. betularia cognataria) also experienced increases in melanisms, which are especially well documented in southern Michigan. Following a succession of governmental regulations designed to improve air quality, melanic phenotypes have been on the decline in moth populations in the United Kingdom and the United States. In both countries, the declines in melanism appear to be correlated primarily with reductions in atmospheric sulfur dioxide. Detailed studies of the evolutionary parallels have focused mainly on two locations: Caldy Common (near Liverpool) in England, and the George Reserve (near Detroit) in the United States. We have expanded the geographic range of American samples to include northeastern Pennsylvania and southwestern Virginia, and we have conducted a national survey in Britain to produce a current melanic frequency distribution map. The results clearly show that the decline in melanism is widespread in both countries, and the changes are generally consistent with decreases in levels of atmospheric SO2.

A PDF file should load here. If you do not see its contents the file may be temporarily unavailable at the journal website or you do not have a PDF plug-in installed and enabled in your browser.

Alternatively, you can download the file locally and open with any standalone PDF reader:

Geographic and temporal variation in the incidence of melanism in peppered moth populations in America and Britain

The Journal of Heredity Geographic and Temporal Variation in the Incidence of Melanism in Peppered Moth Populations in America and Britain B. S. Grant A. D. Cook C. A. Clarke D. F. Owen Industrial melanism in the peppered moth (Biston betularia) is the classic example of observable evolution by natural selection. The phenomenon has been most thoroughly documented in Britain, where once rare melanic phenotypes increased in frequency in concert with increased levels of atmospheric pollution generated by industrial development. Populations of the American subspecies of the peppered moth (B. betularia cognataria) also experienced increases in melanism, which are especially well documented in southern Michigan. Following a succession of governmental regulations designed to improve air quality, melanic phenotypes have been on the decline in moth populations in the United Kingdom and the United States. In both countries, the declines in melanism appear to be correlated primarily with reductions in atmospheric sulfur dioxide. Detailed studies of the evolutionary parallels have focused mainly on two locations: Caldy Common (near Liverpool) in England, and the George Reserve (near Detroit) in the United States. We have expanded the geographic range of American samples to include northeastern Pennsylvania and southwestern Virginia, and we have conducted a national survey in Britain to produce a current melanic frequency distribution map. The results clearly show that the decline in melanism is widespread in both countries, and the changes are generally consistent with decreases in levels of atmospheric SO2. - From the Department of Biology, College of William and Mary, Williamsburg, VA 23187- 8795 (Grant and Cook), the Department of Genetics and Microbiology, University of Liverpool, Liverpool, England (Clarke), and the School of Biological and Molecular Sciences, Oxford Brookes University, Oxford, England (Owen). We are greatly indebted to moth workers throughout the UK who very kindly sent us their data. Special thanks are owed to the following people who assisted us in various ways including providing additional data, permitting us to trap moths on their property, organizing the survey, photography, and correspondence: F. Clarke, H. Cook, K. Cooper, S. Lesevich, C. Owen, J. Owen, J. Thomas, S. Thompson, D. West, J. Woolridge, S. Woolridge, D. Zeveney, and T. Zeveney. Atmospheric reports were provided by several state agencies in Pennsylvania, Virginia, and Tennessee. C.A.C.’s work is supported by the Nuffield Foundation. D. F. Owen is deceased. Address correspondence to Dr. Grant at the address above or e-mail: . q 1998 The American Genetic Association 89:465–471 Parallel evolution occurs when two or more populations independently change in like manner to similar environmental challenges. Such are nature’s replicate ‘‘experiments,’’ and they provide rare opportunities for independent assessments of the potential factors promoting observed evolutionary responses. In this report we expand on the study of parallel evolutionary changes in melanic frequencies in moth populations in America and Britain (Grant et al. 1996) . The familiar half of the parallel begins with the classic example of observable evolution by natural selection: industrial melanism in the British peppered moth (Biston betularia) (reviewed by Berry 1990) . Industrial melanism has occurred in other species of Lepidoptera living in the vicinities of heavy industry ( Kettlewell 1973) , but none of these examples has received nearly the publicity that B. betularia has. Its recent history in this context is recounted in virtually all biology textbooks. The first melanic phenotype was discovered in 1848 near Manchester, England. Melanics are essentially jet black and differ markedly from common or ‘‘typical’’ phenotypes which are mottled gray, an appearance produced by patterns of white and black scales covering the body and wings. Intermediates (f. insularia) also exist and the range of pigmentation phenotypes results from multiple alleles at a single locus ( Lees and Creed 1977) . Our use of the term melanic in this report refers only to the fully pigmented extreme (f. carbonaria). By the turn of the last century over 90% of the specimens of peppered moths caught near Manchester were melanic. Similar observations were made in the vicinities of other industrial cities in Britain. The rapid rise and spread of the formerly rare melanic phenotype was first explained by Tutt (1896) as an example of natural selection. Tutt suggested that the typicals resembled bits of foliose lichen and when they rested by day on trees covered with lichens, they were overlooked by insectivorous birds. With the industrialization of a region the lichens nearby were killed as the tree surfaces became blackened with soot. Typicals in such polluted habitats fell easy prey to predators that use vision to locate resting moths. The melanics, however, were cryptic when at rest on blackened surfaces. As a result of selective predation by birds on the moths, Tutt advanced the idea that the melanics increased in frequency by natural selection. Another half century passed before Tutt’s hypothesis of selective predation was tested. In a series of experiments begun by Kettlewell (1955 , 1956), it has been demonstrated that birds capture and eat more conspicuous moths than cryptic moths. Consistent with this support for Tutt’s analysis have been well-documented reversals of melanic frequencies following modifications in industrial practices (Clarke et al. 1985) . The first of the British Clean Air Acts passed into law in 1956; this and subsequent legislation created ‘‘smokeless zones’’ which brought about significant reductions in atmospheric pollution. The particular pollutant most closely associated with industrial melanism in peppered moths is sulfur dioxide (Steward 1977) . That melanic phenotypes of peppered moths fared better in regions with high atmospheric SO2 appeared to establish a link to Tutt’s selective predation model. Many lichen species are apparently sensitive to pH changes (Seaward 1993) , and because atmospheric SO2 contributes to acid rain, the reduction of SO2 in a region should facilitate the recovery of lichens in the area. (A recent index to the literature on the effects of air pollution on lichens is provided by Henderson 1997.) Given that sequence of events, it should follow that typical phenotypes of the peppered moth would regain their crypsis at the expense of the melanics on resting surfaces lightened by lichen coverage. Grant and Howlett (1988) observed, however, that the best documented reversals in the frequencies of peppered moth phenotypes have occurred in regions where lichen coverage has remained sparse. For example, the Caldy Common site at West Kirby (18 km west of Liverpool) has been sampled each year since 1959, with almost 18,000 specimens taken to date. There melanics have fallen from a high of 93% to a current low near 7% (this report; for complete data through 1993, see Clarke et al. 1994) . Indeed, lichens are slowly returning to Caldy Common (Grant BS, personal observations), but the typical form of the peppered moth returned as the predominant phenotype before lichen coverage on trees was perceptible. Even now lichen coverage on trees in that region is sparse, yet the typicals exceed 93% on Caldy Common, a complete reversal in phenotype frequency since 1959. That major changes in melanic frequencies have occurred independent of lichen succession does not at all rule out selective predation by birds as an important factor in industrial melanism because potential resting surfaces used by moths have gradually changed in reflectance ( have become lighter) following the Clean Air Acts (Clarke et al. 1985) . However, nonvisual components of selection involving preadult viability differences between melanic and pale phenotypes likely complicate the dynamics of frequency changes in modified environments (Creed et al. 1980) . The less familiar half of parallel industrial melanism involving peppered moths was first reported by Owen (1961, 1962a,b) who surveyed populations of the American subspecies (B. betularia cognataria) near Detroit, Michigan. Melanic phenotypes were absent from collections of cognataria specimens taken from that region prior to 1929, but by 1959 they exceeded 90% in Owen’s samples. More recently, an ensuing parallel, the reversal in the frequency of melanics, occurred at that same Michigan location, dropping from above 90% to below 20% by 1994 (Grant et al. 1995) . To our knowledge this is the only documented example of ‘‘twoway’’ parallel evolution in which alleles at a single locus in geographically isolated populations of the same species have gone first in one direction, then in the other. The changes in the frequencies of the moth phenotypes in Michigan in both directions have occurred without any apparent changes in an abundant lichen flora in that region. It does appear, however, that the decline in the frequency of melanic peppered moths in Michigan also followed, as in Britain, regional reductions in atmospheric SO2 (Grant et al. 1996). Unfortunately the American subspecies of the peppered moth has received far less attention than its British counterpart. A 30 year gap separated the population samples taken in Michigan, and nothing has been reported for any other region of North America in recent years. We therefore extended our sampling range of American populations in 1996–1997 to include a wider geographic range. Specifically, here we report assessments of melanic frequencies in peppered moth populations in regions of Pennsylvania and Virginia for which there are published historical records. We also review levels of atmospheric SO2 reported for these regions to determine if geographic variation in the incidence of melanism in moth populations is consistent in direction with variation in levels of pollution. Finally, since it has been over a decade since a broad geographic survey of British peppered moth populations was last attempted (Cook et al. 1986) , we report here the results of a national survey begun in 1996 to bring the before-and-after picture of regional melanic frequencies up to date in the United Kingdom. Materials and Methods Trapping Procedures The standard procedures for trapping adult B. betularia from the wild employ either light traps or assembling traps. Light traps are of various designs, but the principle depends on ‘‘attracting’’ moths to a light source (usually a mercury-vapor lamp). Assembling traps employ caged females that draw males to their location by broadcasting pheromones. Details regarding these methods have been described elsewhere (e.g., Covell 1984; Robinson 1952) . While various traps are variously effective with respect to the size of samples, there is no difference between assembling traps and light traps with respect to the capture rate of melanics and typicals (Clarke et al. 1994). Sizes of samples vary enormously from year to year at the same locations, even when identical trapping methods are used; there is evidence that the fluctuations are cyclic (Grant et al. 1996; Manley 1988) . We employed both light and assembling traps, but most of the data we report were gathered using light traps. British Survey One of us ( D.F.O.) made a personal appeal to moth trappers throughout the United Kingdom to report their data from collections made in 1996 so that the current geographic distribution of melanic frequencies in moth populations could be compared to previous national surveys. The purpose of the comparison was to assess geographic variation in the incidence of melanism, and to determine if the decline in the frequency of melanics best documented at Caldy Common is widespread throughout the United Kingdom. American Sampling Locations We selected two regions to run moth traps in 1996 and 1997: northeastern Pennsylvania and the Virginias (southwestern Virginia and eastern West Virginia). Moth populations were sampled at three separate locations within each major region. The three locations in Pennsylvania were near Wilkes-Barre in the heart of the once-booming anthracite coal region. They include ‘‘Concrete City’’ (an abandoned coal-company village) in Nanticoke, within 2 km of an active coal mine; Jackson Township in the Back Mountain area, a rural setting 13 km from Nanticoke; and Catawissa, a farmland community along the Susquehanna River 60 km southwest of Wilkes-Barre. This general region was selected because it is the only area in the United States besides southern Michigan where melanic peppered moths had been documented at high frequencies (Manley 1988) . We sampled three locations within the region to determine if genetic subdivision with respect to melanism occurs among local populations occupying variously disturbed habitats separated by relatively short distances. The first two sites were sampled daily from May through August in 1996 and again in 1997. The Catawissa site was also sampled in both years but only intermittently. The Virginia locations included Tazewell, Virginia (146 km west of Roanoke), a rural mountainous site; Gary, West Virginia (27 km north of Tazewell), within the coal fields, 3 km from an active mine and loading site; and Monroe County, West Virginia (sampled briefly). The Tazewell and Gary sites were trapped nightly from 10 May to 16 August 1997. We selected this region because some historical records are available about melanic frequencies in peppered moth populations in the southern Appalachians (West 1977; West DA, unpublished data) , and, more importantly, to determine if geographic variation in the incidence of melanism exists among widely separated American populations during the same time period ( Virginia 1997 versus Pennsylvania 1997). The linear (air) distance between Tazewell, Virginia, and Wilkes-Barre, Pennsylvania, is almost identical to the distance each is from Pinckney, Michigan, the location trapped by Grant et al. (1996) . Essentially the interregional trapping centers form the vertices of an equilateral triangle with sides of approximately 650 km ( Figure 1). Regional Pollution Assessment A correlation between atmospheric SO2 and the frequency of melanics in peppered moth populations has been well established in Britain (Clarke et al. 1985; Lees 1981; Steward 1977) . Grant et al. (1996) showed a similar relationship in American peppered moths in Michigan. One goal of this study has been to compare levels of atmospheric SO2 in northeastern Pennsylvania and in southwestern Virginia to determine whether significant differences exist, and if so, to determine if the differences are consistent in direction with any differences we might find in the melanic frequencies of the moth populations living in those regions. Based on previous work, the prediction is that the incidence of melanism should be higher in moth populations living in regions with higher levels of SO2 than in regions where SO2 is low. Atmospheric data were kindly provided to us from the Pennsylvania Department of Environmental Protection, Bureau of Air Quality, and the Virginia Department of Environmental Quality, Air Monitoring Section. Both of these state agencies report air quality assessments to the U.S. Environmental Protection Agency. West Virginia does not operate SO2 monitoring stations south of Charleston; therefore we did not obtain information for that state. The closest SO2 monitoring facility to our Virginia trapping region was in Roanoke ( Vinton), 146 km to the east. For purposes of interregional comparisons between Virginia and Pennsylvania, we requested SO2 data gathered within a 150 km radius of Wilkes-Barre and calculated the average value to compare to the data recorded at Roanoke. Results British Survey Forty moth collectors very kindly sent us their 1996 trapping records in response to our appeal for information about B. betularia. Sample sizes ranged from 3 to 373. We selected from these only those reports based on sample sizes greater than 25. (We made one exception to this minimum cutoff to balance geographic representation.) In all, 18 sampling locations are included in our 1996 national survey. We compared our 1996 results, based on a total of 1642 moths with an average sample size of 91 moths per location, to the original national survey organized by Kettlewell between 1952 and 1956 (for complete data see Kettlewell 1973, Appendix C) . Kettlewell’s survey included many more sampling locations than does ours, so we extracted from his tabulations the data from the 1952–1956 locations that lie closest in position to those locations for which we have obtained data for 1996. The ‘‘paired’’ comparisons are listed in opposite columns in Table 1, ordered from north to south. The northernmost (Scotland) and southernmost ( Dorset) locations have remained unchanged with 0.0% (zero) melanics; at all other locations, the melanics have declined. Kettlewell’s national survey has been presented graphically by pie diagrams positioned on a map of the British Isles to illustrate geographic variation in the frequencies of the moth phenotypes. We have adapted this approach in Figure 2 to show the current geographic variation in melanic frequencies around Britain and to illustrate the before-and-after changes that have occurred around Britain over the past 40 years. This species completes only one generation per year in Britain. American Samples Of the 821 peppered moths caught in Pennsylvania and the Virginias in 1996– 1997, only 4 were of intermediate pigmentation (insularia-like). Intermediates are relatively rare in British populations as well (Clarke et al. 1994) . Effectively, only two phenotypes were common among our American samples: typicals and melanics. Table 2 lists the catch records at all locations as percentages of melanics among the total (N) B. betularia cognataria moths in the samples. The data for the 1952–1956 interval were extracted from Kettlewell (1973, Appendix C). The 1996 data were kindly contributed by those listed in the Data Source column. With the exception of West Kirby (near Liverpool), none of the recent trapping localities exactly match any of the localities included in the original surveys; therefore we selected the closest geographical matches for ‘‘paired’’ comparisons. The differences in the early and recent frequency distributions are highly significant (Wilcoxon’s signed-ranks test, P , .001). That the 1997 sample size in Pennsylvania is four times larger than the 1996 sample size (549 versus 137, respectively) does not reflect any difference in trapping efforts, techniques, personnel employed, locations, or time of year. Wide fluctuations in sample sizes are very common and, so far, inexplicable, but of relevance here, the apparent fluctuations in population size appear to be independent of recorded changes in melanic frequencies (see Grant et al. 1996) . Three distinct conclusions are clear from the results presented in Table 2: (1) melanic frequencies do not differ significantly among localities within regions during the same year; (2) melanic frequencies do differ significantly between regions during the same year ( North 1997 versus South 1997; x2 5 12.81, P , .001); and (3) melanics declined significantly in Pennsylvania between years from 17.5% in 1996 to 9.7% in 1997 (G 5 6.17, P , .025). Regional Pollution In Pennsylvania, 13 SO2 monitoring stations were within a 150 km radius of our trapping locations near Wilkes-Barre. Annual arithmetic means of atmospheric SO2 measured in micrograms per cubic meter of air were averaged over the 13 monitoring stations to produce a single value representing the region for each year over the past 10 years. These values are compared in Figure 3 to the annual arithmetic means of atmospheric SO2 recorded near Roanoke, Virginia, over the same 10 year period. Two patterns are evident: (1) levels of atmospheric SO2 have declined in both regions (significant by regression); and (2) levels of atmospheric SO2 are higher around Wilkes-Barre than at Roanoke for each of the 10 years compared (significant by Wilcoxon’s signed-ranks test, P , .005). Discussion Wherever melanic peppered moths have been documented at appreciably high frequencies in the United States and Britain during the last several decades, they are conspicuously less frequent today. The limited literature on melanism in peppered moth populations across continental Europe has been reviewed by Lees (1981) , and declines in melanism in The Netherlands have been detailed more recently by Brakefield (1990) . Curiously, not a single melanic specimen has ever been reported in the Japanese subspecies of the peppered moth, B. betularia parva (Asami and Grant 1995) . Chiefly, industrial G tests compare phenotypic distributions within regions and between years. Interregional differences (north versus south) are compared by contingency chi square. a x2. melanism in peppered moths has been most carefully studied in the United Kingdom. Britain The last national survey of melanic frequencies in populations of British peppered moths was taken in 1983–1984 through a program involving students in the Open University who were asked to collect moths over 5 day periods (Cook et al. 1986) . Although data were gathered from 190 locations, average sample sizes were very small (fewer than 10 moths per sample site). From the results the authors concluded that some localities experienced declines in melanic frequencies since previous surveys were conducted, while populations at other locations changed little. Much of their article was devoted to a particular location (Caldy Common) for which large, annual samples dating back to 1959 had been published by Clarke et al. (1985) . At the time of the Open University survey, the melanic frequency at Caldy Common was still above 60%. Since that time it has dropped to 7.14% (N 5 238) in 1997 (C.A.C.). ( For the record, insularia at Caldy Common in 1997 were 2.5% of the total catch.) With the notable exception of the Caldy Common site, surprisingly few long-term population assessments have been published for locations elsewhere in Britain. Of shorter-term studies, West (1994) documented a reduction in the frequency of melanics in northwest Kent from 78% in 1970 to below 24% by 1994. In addition, unpublished results communicated to us by avocational moth collectors indicate similar trends elsewhere. Particularly large samples, averaging over 300 specimens per year collected near Nottingham, document a 30% drop in the frequency of melanics from 59% in 1993 to 29% in 1997 (Cooper K, personal communication). Our formal survey of 1996 shows that the decline in melanism is to be found everywhere in Britain. The only exceptions are places where melanism has always been either absent or extremely rare. Nevertheless, geographic variation in the incidence of melanism still exists, and the remaining pockets of relatively high melanism reflect those regions where melanic frequencies had once approached monomorphic proportions. The current geographic differences may represent nothing more than different initial starting places in allele frequencies at various locations when selection favoring typicals began following regional habitat modifications, or they might be attributable to extant geographic variation in ecological conditions affecting selection. As SO2 has been implicated through multiple regression analysis as the most important of 13 environmental variables related to melanic frequencies in moth populations (Steward 1977) , it would be instructive to assess current geographic levels of atmospheric SO2 throughout the United Kingdom to determine if the correlation between SO2 and the incidence of melanism continues. The obvious has been pointed out repeatedly ( Berry 1990; Bishop et al. 1975; Lees 1981) that correlations do not establish causal relationships; yet, consistent correlations between two variables may guide us ultimately to their common cause. A direct relationship between atmospheric levels of SO2 and melanism in moths has yet to be established; an indirect relationship based on the succession of lichens is unfounded (Grant et al. 1996) . Clines in melanic frequencies across Britain have been attributed to a balance between selection within populations and migration among populations. Most efforts to assess dispersal have focused on distances adult moths might fly from points of release, and estimates vary in the range of several kilometers per night ( Bishop 1972; Brakefield and Liebert 1990) . Kettlewell (1973) also suggested that newly hatched caterpillars hanging from tree branches by threads of silk are caught in the wind and carried to new locations, much in the same way young spiders disperse. No quantitative data exist, so the degree to which larval ballooning contributes to gene flow among populations is unknown. America Parallel industrial melanism in American peppered moths in southern Michigan was first reported by Owen in a series of articles (Owen 1961, 1962a,b) , and a parallel decline in melanic frequency later followed reductions in atmospheric SO2 at the same Michigan location (Grant et al. 1996) . As in Britain, the phenomenon appears to be widespread. Manley (1988) recorded melanic frequencies in B. betularia cognataria in a rural area northeast of Klingerstown, Pennsylvania, between 1971 and 1986. For the first 7 years melanics at that location exceeded 50%. It is certainly possible that melanics in the region were once as common as they were in Michigan 10 years earlier, but unfortunately there are no published records for Klingerstown prior to Manley’s study. By 1986, the melanics had fallen to 38%. We resumed sampling the same general region 10 years later. The frequency of melanics in our 1996 sample was 17.5%, less than half the frequency of melanics Manley last reported. The difference in the numerical distribution of melanics to non-melanics in our 1996 sample and Manley’s 1986 sample is highly significant by G test (G1 5 19.30, P , .001). Although we did not trap moths at exactly the same site as did Manley, we are confident in our conclusion that melanics have declined significantly in the region. Our Catawissa site is approximately 15 km closer to Manley’s site than it is to either of our other two trapping sites in northeastern Pennsylvania. We intentionally selected our three sites for extreme local differences, at least to the human eye, in habitat modifications: (1) a coal mining region, with a very highly disturbed terrain surrounded by culm banks interspersed with stands of birch trees (Betula populifolia and B. nigra); (2) a rural, mountainous setting with denser woodlands of mixed stands of deciduous trees (oaks, hickories, and maples) and conifers (pines and hemlocks); and (3) open farmland fringed by mixed woodlands along a river valley. Yet, within the same summer, there were no significant differences in melanic frequencies among these intraregional habitats. It seems unlikely, therefore, that a frequency estimate for that general region of Pennsylvania based on samples from Nanticoke, Back Mountain, and Catawissa would differ significantly from Klingerstown in any given year. Additionally we have demonstrated in this study a significant decline in melanic frequencies at exactly the same three locations in Pennsylvania between 1996 and 1997 ( Table 2). The average melanic frequency in 1996 fell from 17.5% to 9.7% by 1997, a reduction of nearly 8% in a single year. Our trapping records and those of Owen (1962a) suggest that this species is bivoltine (two generations per season) at this latitude; nevertheless, the magnitude of the reduction in the frequency of melanics in a single year is extraordinary. Melanic peppered moths in southwestern Virginia have never been documented at frequencies above 5%. West (1977) sampled moth populations in Blacksburg and at the Mountain Lake Biological Station (MLBS) near Pembroke between 1966 and 1976 and showed the two locations averaged slightly but consistently different melanic frequencies, 1.2% compared to 2.9%, respectively. West (personal communication) continued to sample the MLBS population periodically until 1994, and one of us ( B.S.G.) also sampled the site on two separate years. In all, the MLBS site has been sampled 13 times over a 26 year period. The decline in melanics there from 4.2% in 1968 to effectively zero in recent years is significant by regression (b 5 20.123, t 5 23.244, P , .008, R2 5 0.49). In 1997 we trapped farther west in southwestern Virginia, closer perhaps to potential sources of pollution associated with the coal industry. The complete absence of melanics in our samples is consistent with West’s most recent results at MLBS. While we cannot conclude that melanics are now absent in the area, it is reasonable to conclude that they are uncommon. What accounts for the striking difference in the incidence of melanism in Virginia and Pennsylvania moth populations? We are left, again, with SO2 as a potential contributing factor, or at least a factor potentially linked to environmental conditions that affect Darwinian fitness among moth genotypes. The level of atmospheric SO2 in northeastern Pennsylvania averaged over the past 10 years is about double the southwestern Virginia level (paired samples t 5 17.93, df 5 9, P , .001). It is unfortunate, however, that the Virginia SO2 monitoring station was so far from our moth trapping sites. To gain a better picture of the region we solicited information from surrounding states. Only Tennessee maintained continuous SO2 monitoring stations within a 150 km radius of our Virginia trapping locations. We were surprised to learn that one of two stations near Kingsport reported SO2 levels consistently above Pennsylvania levels over the same 10 year period, while a second station within approximately 20 km of the first was consistently lower than Pennsylvania. Both, however, were significantly above Virginia levels recorded at Roanoke. The question then is, what levels of pollution do the moths living in the vicinities of our trapping sites actually experience? Perhaps the SO2 level at Tazewell lies somewhere in between the Kingsport and Roanoke levels. Fortunately another form of atmospheric pollution was monitored in Tazewell. In 1995 and 1996 Tazewell served as a monitoring station for PM10 (airborne particulate matter 10 mm or smaller in diameter). Although the components are different, SO2 and atmospheric particulates covary as both are common combustion by-products (see Figure 3 in Grant et al. 1996) . Although only 2 years of data are available for comparison, PM10 at Tazewell ranks third behind Pennsylvania and Tennessee for both years, which hints indirectly that SO2 at Tazewell is probably closer to Roanoke levels than to Kingsport levels. It would be very instructive to sample peppered moth populations in eastern Tennessee in the near future to determine if the frequency of melanics is higher there than in southwestern Virginia. If higher levels of SO2 provide some selective advantage to melanic moths, the prediction is that Tennessee should be more like the northern states of Michigan and Pennsylvania than it is to its neighboring southern state, Virginia. In conclusion, our study shows that the parallel declines in melanism in America and Britain reported by Grant et al. (1996) are widespread on both sides of the Atlantic, and, in general, are consistent in direction with reductions in atmospheric SO2. Perhaps, however, SO2 levels are now too low everywhere to influence melanism in moth populations, and the geographic variation that remains is a fleeting residue of history. Asami T and Grant BS , 1995 . Melanism has not evolved in Japanese Biston betularia (Geometridae) . J Lepid Soc 49 : 88 - 91 . Berry RJ , 1990 . Industrial melanism and peppered moths (Biston betularia ( L.)) . Biol J Linn Soc 39 : 301 - 322 . Bishop JA , 1972 . An experimental study of the cline of industrial melanism in Biston betularia ( L.) ( Lepidoptera) between urban Liverpool and rural North Wales . J Anim Ecol 41 : 209 - 243 . Bishop JA , Cook LM , Muggleton J , and Seaward MRD , 1975 . Moths, lichens and air pollution along a transect from Manchester to north Wales . J Appl Ecol 12 : 83 - 98 . Brakefield PM , 1990 . A decline of melanism in the peppered moth Biston betularia in The Netherlands . Biol J Linn Soc 39 : 327 - 334 . Brakefield PM and Liebert TG , 1990 . The reliability of estimates of migration in the peppered moth Biston betularia and some implications for selection-migration models . Biol J Linn Soc 39 : 335 - 341 . Clarke CA , Grant B , Clarke FMM , and Asami T , 1994 . A long term assessment of Biston betularia ( L.) in one UK locality (Caldy Common near West Kirby , Wirral), 1959 - 1993 , and glimpses elsewhere . Linnean 10 : 18 - 26 . Clarke CA , Mani GS , and Wynne G , 1985 . Evolution in reverse: clean air and the peppered moth . Biol J Linn Soc 26 : 189 - 199 . Cook LM , Mani GS , and Varley ME , 1986 . Postindustrial melanism in the peppered moth . Science 231 : 611 - 613 . Covell CV , 1984 . A field guide to the moths of eastern North America . Boston: Houghton Mifflin. Creed ER , Lees DR , and Bulmer MG , 1980 . Pre-adult viability differences of melanic Biston betularia ( L.) ( Lepidoptera) . Biol J Linn Soc 13 : 251 - 262 . Grant B and Howlett RJ , 1988 . Background selection by the peppered moth (Biston betularia Linn .) : individual differences . Biol J Linn Soc 33 : 217 - 232 . Grant B , Owen DF , and Clarke CA , 1995 . Decline of melanic moths . Nature 373 : 565 . Grant B , Owen DF , and Clarke CA , 1996 . Parallel rise and fall of melanic peppered moths in America and Britain . J Hered 87 : 351 - 357 . Henderson A , 1997 . Literature on air pollution and lichens XLV . Lichenologist 29 : 283 - 290 . Kettlewell HBD , 1955 . Selection experiments on industrial melanism in the Lepidoptera . Heredity 9 : 323 - 342 . Kettlewell HBD , 1956 . Further selection experiments on industrial melanism in the Lepidoptera . Heredity 10 : 287 - 301 . Kettlewell B , 1973 . The evolution of melanism . Oxford: Clarendon Press. Lees DR , 1981 . Industrial melanism: genetic adaptation of animals to air pollution . In: Genetic consequences of man-made change ( Bishop JA and Cook LM , eds). London: Academic Press; 129 - 176 . Lees DR and Creed ER , 1977 . The genetics of the insularia forms of the peppered moth, Biston betularia . Heredity 39 : 67 - 73 . Manley TR , 1988 . Temporal trends in frequencies of melanic morphs in cryptic moths of rural Pennsylvania . J Lepid Soc 42 : 213 - 217 . Owen DF , 1961 . Industrial melanism in North American moths . Am Nat 95 : 227 - 233 . Owen DF , 1962a . The evolution of melanism in six species of North American geometrid moths . Ann Entomol Soc Am 55 : 695 - 703 . Owen DF , 1962b. Parallel evolution in European and North American populations of a geometrid moth . Nature 195 : 830 . Robinson HS , 1952 . On the behaviour of night-flying insects in the neighbourhood of a bright source of light. Proc R Entomol Soc Lond A 27 : 12 - 21 . Seaward MRD , 1993 . Lichens and sulphur dioxide air pollution: field studies . Environ Rev 1 : 73 - 91 . Steward RC , 1977 . Multivariate analysis of variation within the insularia ''complex'' of the melanic moth Biston betularia . Heredity 39 : 97 - 109 . Tutt JW , 1896 . British moths. London: Routledge. West BK , 1994 . The continued decline of melanism in Biston betularia ( L.) ( Lep: Geometridae) in NW Kent . Entomol Rec 106 : 229 - 232 . West DA , 1977 . Melanism in Biston ( Lepidoptera: Geometridae) in the rural Central Appalachians . Heredity 39 : 75 - 81 . Received December 27, 1997 Accepted April 30 , 1998 Corresponding Editor: Stephen J. O'Brien

This is a preview of a remote PDF:

Grant, BS, Cook, AD, Clarke, CA, Owen, DF. Geographic and temporal variation in the incidence of melanism in peppered moth populations in America and Britain, Journal of Heredity, 1998, 465-471, DOI: 10.1093/jhered/89.5.465