A Perspective on Declining Amphibian Populations Author(S): Andrew R

Herpetologists' League

Chicken Little or Nero's Fiddle? A Perspective on Declining Amphibian Populations Author(s): Andrew R. Blaustein Reviewed work(s): Source: Herpetologica, Vol. 50, No. 1 (Mar., 1994), pp. 85-97 Published by: Herpetologists' League Stable URL: http://www.jstor.org/stable/3892877 . Accessed: 03/10/2012 18:54

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at . http://www.jstor.org/page/info/about/policies/terms.jsp

. JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact [email protected].

Herpetologists' League is collaborating with JSTOR to digitize, preserve and extend access to Herpetologica.

http://www.jstor.org Herpetologica, 50(1), 1994, 85-97 ? 1994 by The Herpetologists'League, Inc.

CHICKEN LITTLE OR NERO'S FIDDLE? A PERSPECTIVE ON DECLINING AMPHIBIAN POPULATIONS

ANDREW R. BLAUSTEIN Department of Zoology, 3029 Cordley Hall, Oregon State University, Corvallis, OR 97331-2914, USA

THERE have been numerous recent re- ology (Blaustein et al., in press). In my ports suggesting that the populations of New Orleans talk and in Blaustein et al. amphibian species in a wide array of geo- (in press), we came to several of the same graphic regions and habitats have appar- conclusions presented by Pechmann and ently declined or have experienced range Wilbur (1994). Some of the points I make reductions (e.g., reviews by Blaustein and below are similar to those in Blaustein et Wake, 1990; Hayes and Jennings, 1986; al. (in press), and I liberally quote from Honegger, 1978; Phillips, 1990; Vitt et al., that paper. 1990; Wake, 1991; Wake and Morowitz, Pechmann and Wilbur (1994) bring a 1991; specific examples are Beebee et al., healthy skepticism to certain arguments of 1990; Bradford 1989, 1991; Clarkson and the amphibian decline problem. More- Rorabaugh, 1989; Corn and Fogleman, over, the questions that they pose on eval- 1984; Heyer et al., 1988; Kagarise Sher- uating evidence for population declines man and Morton, 1993; Semb-Johannson, should help biologists to focus on key as- 1989; Tyler and Davies, 1985). Pechmann pects of the amphibian decline problem. and Wilbur (1994) evaluate the literature I agree with two of the general conclusions on amphibian declines and (1) stated that presented by Pechmann and Wilbur "few would dispute the fact that habitat (1994): (1) habitat destruction and other destruction and other anthropogenic ef- anthropogenic effects have reduced or fects have reduced or eliminated many eliminated many populations of amphib- populations of amphibians as well as other ians as well as other taxa and (2) there are taxa", (2) suggested that the declines of not enough long term data to state con- "isolated, protected amphibian popula- clusively, for most species, that their pop- tions" cannot be unequivocally separated ulation declines are unusual. from natural population fluctuations, (3) I do not, however, agree with several suggested that the "implication (italics other points made by Pechmann and Wil- mine) of these reports, however, is that bur (1994). For example, in my opinion, declines and disappearancesof amphibian and contrary to the views of Pechmann populations represent a distinct phenom- and Wilbur (1994), previous reviews of the enon that goes beyond the general biodi- amphibian decline problem did not imply versity crisis", and (4) questioned whether that it was a phenomenon distinct from or not "amphibiansare sensitive indicators the general biodiversity crisis. Further- of environmental stresses". more, despite the assertion of Pechmann In this paper, I present my perspective and Wilbur (1994), I believe that under on the amphibian decline problem and of- certain conditions, amphibians are good fer my response to points made in the pa- bioindicators of environmental stress. per by Pechmann and Wilbur (1994). I first presented some of the ideas in this TALES FROM CANTERBURY AND paper at the symposium on amphibian de- IRVINE AND THE clines at the 1990 meeting of The Her- TENOR OF THE TIMES petologists' League in New Orleans. My After hearing numerous anecdotal re- talk in New Orleans formed the basis of a ports in conversation, and listening and paper that is in press in Conservation Bi- reading some substantive reports in sci- 85 86 HERPETOLOGICA [Vol. 50, No. 1 entific sessions,many participantsin the Moreover, Wilbur, as quoted in Barinaga FirstWorld Congress of Herpetology,held (1990), stated that "The data are anec- in Canterbury,England, in 1989 began dotal, but it's so well repeated they cer- questioningwhether the reportsof am- tainly are believable." phibian declines representeda general The meetings in Canterbury and Irvine trend.In February1990, the NationalRe- provided a good deal of anecdotal (and searchCouncil held a workshopin Irvine, some empirical) evidence that possibly Californiathat addressed further the issues amphibian populations in various parts of concerningdeclining amphibianpopula- the world were in decline. This was the tions.Several reports covering the general general picture at the time-right or consensusof the Irvine participantswere wrong. Since these conferences, persons published(e.g., Barinaga,1990; Blaustein have changed or modified their stance. As and Wake,1990; Phillips, 1990; Vitt et al., more information comes in, the general 1990; Wake and Morowitz,1991). These picture regarding the amphibian declines reportswere all very similarbecause they may change. However, the Irvine confer- summarizedthe Irvine meeting. Pech- ence served as a springboard for action. mann and Wilbur (1994) criticize these Long-term studies of amphibian popula- reportsthroughout their paper as if they tions were called for. Hypotheses were were original refereed scientific articles, constructedand, in some cases, predictions which they were not. from these hypotheses are being tested. AfterCanterbury and Irvine,the overall atmospherewas that there might be a LAG TIME problem concerning the populationsof Pechmann and Wilbur (1994) state that amphibiansin variousparts of the world. "information on some widely-cited cases The publishedreports on the Irvine con- [on amphibian declines] . .. is not available ferencecorrectly conveyed the generalat- in the primary technical literature." This mosphereat the time.Phillips (1990) wrote, is no surprise. The lag-time from research "By the end of the Februaryworkshop- to print takes several years. Much of the the firstto focuson the amphibiandecline- work stimulated by the Irvine conference the scientistsagreed that anecdotalevi- is just now appearing in the refereed, dence indicatesthat at least some of the mainstream literature. Some of the pieces world's 5130 amphibianspecies are de- to the amphibian decline puzzle are be- cliningat an alarmingrate." Blaustein and ginning to fit. For example, based on his- Wake (1990) stated that "The decline in torical accounts, museum records, and in- populationsand shrinkagein geographical tensive searches, Fellers and Drost (1993, rangeof severalof thesespecies exemplify in press for over a year) concluded that the types of decline that are apparently populations of the Cascades frog (Rana occurring throughout the world." The cascadae) in northern California have ex- consensusfrom the Irvineconference that hibited precipitous declines for more than amphibiandeclines may have begun in the 15 years. Blaustein et al. (in press; in re- 1970'sand that the accumulationof evi- view for eight months) identified a species dence suggeststhe possibilityof a global of pathogenic fungus that may be one cause decline, was harshlycriticized by Pech- for the unusual mortality of toad eggs in mann and Wilbur(1994). Yet, Vitt et al. the central Cascade Range of Oregon. The (1990),in a paperco-authored by Wilbur, fungus is circumglobular in distribution (cited by Pechmannand Wilburonly on and is commonly found on stocked fishes. page 76 and not cited or criticizedalong It can be carried on amphibians as they with the otherreviews on the Irvinemeet- disperse to other locales, and it may be ing),stated that "certainamphibians, even very important in the overall amphibian in habitatsthat appearto be pristine,are decline picture. Kagarise Sherman and disappearingat an alarmingrate, and the Morton (1993) recently documented declines are widespreadand have been changes over 20 years in the sizes of breed- particularlyserious since the late 1970's." ing populations of Yosemite toads (Bufo March 1994] HERPETOLOGICA 87 canorus). Other empirical papers address- 1988). Numerous threatened species exist ing the amphibian decline problem will in all plant and animal groups, including appear shortly. Some investigators (myself amphibians (e.g., IUCN, 1990; McNeely included) are waiting for enough long-term et al., 1990). data to accumulate before publishing their results. ARE AMPHIBIANSDECLINING IN Pechmann and Wilbur (1994) argue that PRISTINEHABITATS? the reviews of the amphibian decline prob- Pechmann and Wilbur (1994) question lem (primarily reviews of the Irvine con- reports suggesting that "some of the de- ference) did not, in general, attempt "a clines and apparent extinctions have taken comprehensive or random survey of the place in isolated, seemingly pristine ar- status of amphibian populations, but rath- eas." Too many biologists (myself includ- er emphasize declines and disappeared) have probably used the term pristine ances." Again, this is no surprise. The re- too loosely when describing population ports criticized by Pechmann and Wilbur processes in amphibians from locations (1994) were reviews of the Irvine confer- where the habitat has not obviously been ence. They were not intended to provide altered or destroyed. For example, Wilbur, comprehensive reviews of anything. Nev- as quoted by Barinaga (1990), stated that ertheless, at least one report (Blausteinand "habitat destruction is probably the dom- Wake, 1990) did address the issue of spe- inant thing going on-its not the whole cies and areas that do not seem to be af- story, because we have a lot of pristine fected by "declines", by illustrating sev- (italics mine) areas where [amphibian] eral examples of populations that were not populations are going down the tubes as known to be in decline. well" (see also quote above from Vitt et al., 1990). IS THE AMPHIBIAN DECLINE I agree with the more recent statement PROBLEM A DISTINCT PHENOMENON? by Pechmann and Wilbur (1994) that Pechmann and Wilbur state that the "some areas appear protected and pristine "implication (italics mine) of these reports but are not so from an amphibian's point [from the Irvine Conference] ..., is that of view." It is very difficult (probably im- declines and disappearancesof amphibian possible) to find truly pristine habitat. Hu- populations represent a distinct phenom- man-induced habitat alterationor destruc- enon that goes beyond the general biodi- tion permeates all regions of the world. versitycrisis." This "implication"may have Habitat destruction or alteration may oc- been a result of the tenor of the times cur, for example, through overt devasta- which included some outrageous media tion of rain forests or old-growth stands. coverage [including a feature in the World However, some changes may be more sub- Weekly News (17 April 1990) showing tle. Environmental changes may occur aliens taking frogs en masse to another when exotic species (perhaps with patho- planet to feed their kind-an unsubstan- gens) are introduced into areas inhabited tiated report from the "gray" literature]. by amphibians. Conditions may change as In the reports on the Irvine conference, I pollutants, not perceived by humans, are simply do not see this implication. None introduced into the atmosphere. of the reportscited by Pechmann and Wil- Pechmann and Wilbur (1994) state that bur (1994) state that the amphibian de- "Reviews of declining amphibian popu- cline problem represents a distinct phe- lations have not clearly separated declines nomenon from the overall biodiversity for which human activities are an obvious crisis. It is obviously one part of the overall culprit from the others." Most of the re- biodiversity crisis. Rates of extinction have views in question (e.g., Barinaga, 1990; accelerated in recent times, in many cases Blaustein and Wake, 1990; Phillips, 1990; because of human interference that has Wake and Morowitz, 1991) were sum- damaged suitable habitat (e.g., McNeely maries of the Irvine conference. They were et al., 1990; Simberloff, 1986; Wilson, not intended to go into detail about causes. 88 HERPETOLOGICA [Vol. 50, No. 1

However, in general, they did present all during a given period of time, or if it did the potential causes discussed at the Irvine become extinct, whether it recolonized the meeting. area within the time span, it must be monitored for at least one complete turnover THE "GRAY" LITERATURE of all individuals of that species at that Pechmann and Wilbur (1994) state that particularsite. Connell and Sousaalso sug- "recent perturbations reported in other gested that to demonstrate if a population taxa share much in common with those is stable, it must be monitored for at least reported in amphibians." They argue that one turnover. Connell and Sousa (1983) much of the information on declines in surveyed the literature of long-term stud- organisms ranging from black abalone to ies of organismsthat fit their criteria. Their dolphins, to sea urchins and amphibians, unbiased search found no such studies at comes from the so-called "gray literature". all on amphibians-none showing a de- I agree that one must be cautious about clining, increasing, or stable population. using literature from unrefereed papers A decade after Connell and Sousa (1983) and personal communications. However, published their paper, long-term studies Pechmann and Wilbur (1994) are incon- of amphibianpopulations remain rare.This sistent in their arguments about using such lack of long-term data is surprising, be- information. For example, almost all of the cause anuran amphibians have been used information that they cite, in the begin- in numerous studies as model vertebrates ning of their paper, on the leopard frog for studying complex life cycles, aquatic declines that seemed to have begun in the community structure and mating patterns 1960'sand lasted into the 1970's,were from (e.g., reviewed by Duellman and Trueb, unrefereed literature or personal com- 1986; Olson et al., 1986; Wilbur, 1980). munications. Although Pechmann and Limited information on the long-term dy- Wilbur stated that there are no hard data namics of amphibian populations makes it that leopard frog populations have recov- difficult to evaluate whether recently ob- ered, based solely on personal communi- served declines in population density are cations, primarilyfrom supply houses,they unusual in magnitude or duration. More- state that "R. pipiens populations appar- over, due to the paucity of long-term data ently have recovered from the decline, and on amphibians, it is obviously difficult to that bumper crops of frogs were observed distinguish between natural fluctuations during 1991 and 1992". With all the media and those caused by humans. attention regarding amphibian declines, Nevertheless, there have been several representativesof supply houses may have long-term studies of amphibian declines a vested interest in saying that everything published since the paper by Connell and is fine. The information from Pechmann Sousa (1983) appeared. I briefly review and Wilbur (1994) on the population re- some of these studies below. Some of the covery of Bufo cognatus, in Cleveland Co. accounts I discuss below are similar in de- Oklahoma, the site where Bragg (1960) scription to those in Blaustein et al. (in reported their decline over 30 years ago, press). Most of those studies that reported was based solely on one personal com- declines were not discussed in any detail munication. by Pechmann and Wilbur (1994). A study by Pechmann et al. (1991) carefully monitored the breeding population LONG-TERM STUDIES OF sizes of four amphibian species at one site AMPHIBIAN POPULATIONS in South Carolina for 12 years. Pechmann In my New Orleans talk, I detailed the et al. (1991) showed that the populations views of Connell and Sousa (1983) con- of three species fluctuated and one species cerning the evidence needed to judge eco- increased over that time span. logical stability and persistence. Connell Semb-Johansson's(1989) study of com- and Sousa (1983) argued that to demon- mon toads (Bufo bufo) on islands off the strate that a population became extinct Norwegian coast is an excellent example March 1994] HERPETOLOGICA 89 of a long-term population study of an am- tively stable. The gradual loss in numbers phibian.Toads were monitoredfor 24 years of P. shenandoah, due to interspecific (1966-1989). Their numbers declined dra- competition, may yet provide evidence of matically from 1966-1975 and have re- the elusive "natural" extinction event. mained low. Declines of natterjack toads, based pri- Corn and Fogleman's (1984) study doc- marily on range reduction data collected umented the local extinction of leopard over 20 years (e.g., Banks and Beebee, 1987; frogs (Rana pipiens) in Colorado across a Beebee, 1977; Beebee et al., 1990), have number of sites. In this study, six popu- also been reported (see discussion below). lations of R. pipiens were examined for 10 Thus, of the seven long-term studies dis- years (1973-1982). Reproductive failure cussed above, two (Berven, 1990; Pech- was seen in 1973 at one site and by 1981 mann et al., 1991) showed species that were no individuals of R. pipiens were seen at generally fluctuating in numbers, four any site. Rana pipiens was absent from the (Banks and Beebee, 1987 and Beebee et area at the end of the study. Rana pipiens al., 1990; Corn and Fogleman, 1984; Ka- older than four of five years are rare (Le- garise Sherman and Morton, 1993; Semb- clair and Castanet, 1987). Thus, these pop- Johansson, 1989) showed declining popu- ulations were followed for almost two lations, and one (Jaeger, 1980) showed the turnovers. decline of one species and a possible stable Kagarise Sherman and Morton (1993) population in the other. Of course, it is documented the population changes over possible that the populations in decline may 20 years in Bufo canorus at Tioga Pass, be undergoing natural fluctuations in California. Comprehensive surveys of numbers; years with strong recruitment breeding aggregations were made from may be followed by years of poor repro- 1971-1982 and less systematic observa- duction during which the population de- tions were taken from 1983-1991 at Tioga clines. With a year or two of successful Pass. Six additional populations in north- reproduction by older individuals (Bufo ern California were also monitored from canorus can live for over 30 years: Goin 1973-1990. At the largest breeding pools et al., 1978), the populations could rapidly at Tioga Pass, the populations declined rebound from their current low levels. about nine-fold from 1974-1982. The In their summary of the available long- mean number of toads found in daily term data, Pechmann and Wilbur (1994) searches also declined during the 20-year seem to contradict themselves by stating period. Similar declines at the other sites that "long-term census data on amphibians were reported. are too few to warrant generalizations Berven's(1990) seven year study of wood about their variability" when in the very frogs (Rana sylvatica) illustratesthat their next sentence they generalize that "Those populations turn over about every 2-3 data that are available suggest that pop- years. The populationsof adult wood frogs ulations of amphibians range from the in Berven's study showed erratic interan- highly variable to the highly stable." Their nual fluctuations largely due to variation latter statement seems inaccurate, es- in juvenile recruitment. Declines in pop- pecially in light of the long-term data (some ulations of adult wood frogs were followed of which reflected declines) that I pre- by sharp increases in one pond and rela- sented above. tively low but stable numbers in another. In addition to the long-term studies cit- A 14-year study by Jaeger (1980; see also ed above, populations of other amphibian Jaeger, 1970), largely overlooked from the species have disappeared from portions of population perspective, showed that the their historical ranges (without concomi- Shenandoah salamander (Plethodon she- tant shifts in their ranges) and have failed nandoah) has been declining (and contin- to reestablish at such sites for periods lon- ues to do so; R. Jaeger, personal commu- ger than their estimated maximum life nication) probablydue to competition with span. I will briefly discuss range reductions P. cinereus whose populations are rela- in three species with the caveat that much 90 HERPETOLOGICA [Vol. 50, No. 1 of the data concerning them have not yet cadae are being accumulated by a number been published in refereed journals. of workers in Oregon, Washington, and Until the mid 1970's,the red-legged frog California. These data will soon be avail- (Rana aurora) was extremely abundant in able in the mainstreamliterature. The data the Willamette Valley of Oregon, a rela- for at least one species, R. cascadae, in tively large valley (160 by 60 km) bounded California, have been recently published by the Coastal and Cascade Mountain (Fellers and Drost, 1993). ranges (Blaustein and Wake, 1990; Nuss- Information on range reduction, some- baum et al., 1983). Rana aurora is now times in conjunction with demographic extremely rare, and breeding populations data, has resulted in many species being have not been observed for at least 20 years placed on regional threatened lists. For ex- in the Willamette Valley (Blaustein and ample, 15% of the amphibian species in Wake, 1990; R. M. Storm, personal com- the western United States are listed as can- munication). didates for threatened species status (see The western spotted frog (Rana pre- Walls et al., 1992, and references therein). tiosa) was abundant throughout Washing- In the Pacific Northwest, this is of special ton and Oregon until the mid-1970's (Mc- concern, because 54% of the native am- Allister and Leonard 1990; Nussbaum et phibian species in Oregon are listed as sen- al., 1983), but it has become extremely rare sitive, 46% in Washington state are listed in the western portion of its range (Mc- in the special concern category, and 29% Allister and Leonard, 1990, 1991; Nuss- are listed as threatened in Idaho (Walls et baum et al., 1983; Stebbins, 1985). Popu- al., 1992). Of course, the listing of some lations of R. pretiosa are exceptionally rare species is questionable because of the lack west of the Cascade Mountains in Wash- of long-term population data. ington, and they have not been found west of the Cascade Mountains in Oregon for THREE FAMOUS CASES at least 23 years (McAllisterand Leonard, There are several cases concerning the 1990, 1991; Nussbaum et al., 1983). One amphibian decline problem that have re- specimen, tentatively identified as R. pre- ceived a great deal of attention. I will brief- tiosa, was found near Olympia, Washing- ly discuss three of them to illustrate the ton in 1990 (K. R. McAllister and B. Leo- importance of being cautious when inter- nard, personal communication). No other preting reports of the population status of specimen has been found at that site (K. amphibians. These reports involve two R. McAllister and B. Leonard, personal rather unique species and the long-term communication). Thus, this species has study of Pechmann et al. (1991) (see Bari- been missing from about one-third of its naga, 1990; Blaustein and Wake, 1990; range since the mid-1970's. Phillips, 1990; Tangley, 1990; Wake, 1991). The intensive search of historical ac- The following accounts are liberally taken counts and museum records and searches from Blaustein et al. (in press). at 16 sites led Fellers and Drost (1993) to The gastric brooding frog (Rheobatra- conclude that populations of the Cascades chus silus) was discovered in 1973 in rel- frog (Rana cascadae) in northern Califor- atively undisturbedareas of the Conondale nia have exhibited a precipitous decline and Blackall Ranges about 160 km north for more than 15 years. Local extinctions of Brisbane,Australia (Fanning et al., 1982; of R. cascadae in Oregon have also been Liem, 1973). This species could have un- reported (Blaustein and Wake, 1990). locked many of the mysteries of physiol- However, based on surveys of numerous ogy and digestion because of its habit of sites since the Irvine meeting, population swallowing and brooding its young in its declines of R. cascadae in Oregon do not stomach (Tyler and Carter, 1981). The de- seem to be as severe as they are in Cali- cline of this species began in the late 1970's fornia (Blaustein, unpublished data). and it has not been found in nature since Information on the population dynam- 1979 (Tyler, 1991; Tyler and Davies, 1985). ics of R. aurora, R. pretiosa, and R. cas- The fact that this species was only recently March 1994] HERPETOLOGICA 91 discovered suggests that it may have al- because of their conclusion that "there is ways been rare and/or had a highly lo- no evidence that the declines in amphibian calized and little explored geographic dis- populations observed in other locations tribution. Although this species seems to have occurred in populations at Rainbow have gone extinct, it is not unusual for Bay." This study was conducted at only some species to go undetected for years. one site in an area relatively protectedfrom The golden toad (Bufo periglenes) is a human interference. Caution should be sexually dimorphic species endemic to taken not to generalize about amphibian Costa Rica (Crump et al., 1992). The males populations in other regions based on one are brilliant orange whereas the females study site. It is possible that populationsof are brightly mottled. Moreover, larval even the same species reported to be fluc- golden toads have the rare habit of being tuating by Pechmann et al. (1991) might facultative non-eaters(Crump, 1989). Each be declining or increasing in another area. year from the early 1970's through 1987 Although the authors were careful not to the toads emerged from underground to generalize about other populations, an breed in the spring (Crump et al., 1992). "implication" could easily be conjured up In 1987, more than 1500 individuals were (as did some of the media) from this study observed by Crump et al. (1992) but re- that most reportsof declines are really only cruitment was nearly zero. From 1988- the result of natural population fluctua- 1990, only 11 adult toads were found tions. (Crump et al., 1992). While it appears that the number of golden toads has drastically REGIONAL DIFFERENCES IN dwindled, it is also possible that adults of POPULATION DYNAMICS B. periglenes are estivating below ground The population processesand patch dy- in response to unfavorable weather con- namics of amphibians may vary regionally ditions, and may emerge when conditions (see examples reported in Blaustein and become more favorable for breeding Wake, 1990). For example, in the south- (Crump et al., 1992). Some species in the eastern United States where Pechmann et same family as B. periglenes can live more al. (1991) conducted their study, there is than 30 years (Duellman and Trueb, 1986), an extremely diverse and abundant am- and many toad species within the same phibian fauna. In certain areas, such as the genus can live for more than 10 years Savannah River site, there are relatively (Bowler, 1977). Thus, populations of the dense amphibian populations and a great golden toad can probably persist through deal of continuous suitable habitat (see dis- several years of poor recruitment. Like the cussionin Jacksonet al., 1989; Wake, 1991). gastric brooding frog, the golden toad was If populations go locally extinct in this re- only recently discovered (Savage, 1966), gion, it seems likely that the probability of so its population dynamics have not been recovery will be higher than in certain oth- studied in detail and are poorly under- er regions (Wake, 1991). In other localities, stood. such as some regions in mountainouswest- Caution should be taken when declaring ern North America, where several species that either the golden toad or the gastric appear to have diminishing ranges, many brooding frog are extinct. In both cases, species are found in habitats that are lo- there is no concrete evidence that the ob- calized or fragmented and opportunities served declines are atypical, nor is the cause for recolonization may be much lower of the declines known. Because there are (Wake, 1991). no long-term data for these species, we cannot reject the possibility that, under ARE AMPHIBIANS "CANARIES"? natural conditions, these species often un- Most reviews of the amphibian decline dergo large fluctuations in numbers. problem suggested that amphibians are The study by Pechmann et al. (1991) good bioindicatorsof environmentalstress. has received a great deal of attention, be- They have permeable, exposed skin (not cause it was a careful long-term study and covered by tough scales, hair, or feathers) 92 HERPETOLOGICA [Vol. 50, No. 1 and eggs (not covered by hard or leathery studies champion amphibians as bioindi- shells) that may readily absorb substances cators of toxic substances ranging from from the environment. The complex life heavy metals to pesticides (e.g., Birge et cycles of many species potentially exposes al., 1979, 1983; Cooke, 1981; Fonovich de them to both aquatic and terrestrial en- Schroederand Pechen de D'Angelo, 1991; vironmental agents. This view was ampli- Jayaprakashand Madhyastha,1987). After fied by statements made by Vaughn Shoe- compiling more than 250 references on the maker (at Irvine as cited in Blaustein and effects of toxicants on amphibians, Power Wake, 1990), a highly regarded amphib- et al. (1989) stated that "Amphibians are ian physiologist, by the keynote talk given particularly sensitive to metals and acidi- by Henry Wilbur in New Orleans (as stat- fication. They are considered to be useful ed by Pechmann and Wilbur, 1994), and indicator species for measuring the effects by Vitt et al. (1990) who noted that am- of local changes in environmentalstudies." phibians were "harbingers of decay"'. Indeed, anuran embryos and tadpoles I agree with Pechmann and Wilbur have become "models" for testing the tox- (1994) that amphibians "aremore sensitive icity of numerous chemicals (e.g., Bantle in some cases and less sensitive in others, et al., 1992; Cooke, 1981). As Cooke (1981) and there is a considerablevariation among stated "Forsuch work, tadpoleshave many species and toxicants." This statement advantages",including characteristicmor- would probably be true for organisms in phological changes that occur in the pres- almost any taxon. Moreover, I agree that ence of particular pollutants. The char- it is a "quantum leap from most toxico- acteristicmorphological changes that occur logical studies to projecting impacts on during development when amphibian lar- populations and communities in the field" vae are exposed to chemical pollutants is (Pechmann and Wilbur, 1994). I am not the basis of an assay and Atlas of Abnor- sure how to quantify a "good" bioindi- malities in the Frog Embryo Teratogenesis cator. However, I do believe that Pech- Assay-Xenopus (FETAX) program (Ban- mann and Wilbur (1994) are selling am- tle et al., 1992). The FETAX program was phibians short when they state that they initiated (by James N. Dumont and co- were "not aware of any evidence available workers of Oak Ridge National Labora- to substantiate"that amphibianswere good tory) because of increased costs and con- bioindicators and that the "toxicological cern over using mammalian models. literature does not support a general state- Originally, the assay and atlas was intend- ment that amphibians are a relatively sen- ed to be used primarily for embryos and sitive group." Under many conditions, am- larvae of Xenopus. This system is now be- phibiansmay be as good as other organisms ing used by several government agencies as bioindicators of environmental stress. to test the effects of chemical pollutants This is one reason why they are being used on development in North America ranids with increasing frequency in this regard. and bufonids. The FETAX programis cur- Their effectiveness as bioindicators, like rently investigating the possibility of using members of other taxa, depends upon the amphibians in the field to assay chemical situation and the species in question. pollutants. Thus, the U.S. Environmental Protec- Acidification is one proposed cause for tion Agency, the U.S. Fish and Wildlife amphibian declines (Dunson et al., 1992). Service, and the Canadian Wildlife Ser- With regard to acid tolerance, Pechmann vice have all published extensive reports and Wilbur (1994) state that "most am- and bibliographies on the use of amphib- phibians are comparatively tolerant of ians as bioindicators,especially in the con- acid" citing Pierce (1985) and Mierle et text of toxicology (see for examples, Bantle al. (1986). In fact Pierce (1985) only tested et al., 1992; Power et al., 1989). Moreover, 14 anurans and two salamandersand stat- there are many papers in the mainstream ed that "most amphibian species [in his literature that have assessed the effects of study] are relatively acid tolerant." More- toxicants on amphibians. Many of these over, Mierle et al. (1986) briefly reviewed March 19941 HERPETOLOGICA 93 the acid tolerances of only six species of hanced in conditions of relatively low pH amphibians showing that some were more because of reduced predation by a sala- tolerant than others. There is an extensive mander predator. Warner et al. (1993) literature on the effects of acidity on am- showed that at relatively high pH, larval phibians (see symposium in Journal of Hyla femoralis caused decreased survival Herpetology 26:349-442, 1992, and ref- and an increase in the larval period of Hyla erences therein). Certain studies show that gratiosa. some amphibians are very sensitive to pH, In summary, in some cases, amphibians other studies show that some species are may make excellent canaries, in other sit- relatively tolerant to pH. uations (for example, in coal mines), ca- Differential sensitivity to pH has also naries are probably better. The sensitivity been documented in numerous organisms of amphibians to certain pollutants (e.g., from other taxa. For example, in his review those causing increased acidification) may of acid precipitation in aquatic ecosys- have significant ecological consequences. tems, Haines (1981) showed differential Finally, I agree with Pechmann and Wil- sensitivity to pH in bacteria, fungi, algae, bur (1994) that there may be no analyses insects, fish, and amphibians. Similarly, comparing population changes in amphib- Roff and Kwiatkowsky (1977) showed dif- ians to those in other taxa to substantiate ferential sensitivity to pH in zooplankton. the fact that "human impacts fall more It is very difficult to generalize how pH heavily upon amphibiansthan other taxa". affects a group of organisms. Some species within a particular taxon may be sensitive, IS THE SKY FALLING? others may be tolerant, depending upon Is ROME BURNING? conditions and the life history stage. In the broad sense, the sky is falling. Many amphibian species are sensitive The world is losing an unprecedented enough to acidification that they can be number of species in all taxa per year, affected at the population and community primarily due to habitat destruction and levels. For example, Beebee et al. (1990) alteration (McNeely et al., 1990; National suggested that acidification due to pollu- Science Board, 1989; Wilson, 1988, 1992). tion has adversely affected populations of As rain forests and old-growth stands are natterjack toads (Bufo calamita) in Brit- destroyed, as wetlands are filled in or pol- ain. They stated that "Natterjack toad de- luted as exotic species, or chemicals that clines have therefore almost certainly been adversely affect our atmosphere, are in- strongly influenced by water acidifica- troduced, suitable habitat for many or- tions, and this may account for the sub- ganisms, including amphibians, are being stantial number of extinctions at heathland destroyed. sites." They further suggested that even What should be the responseto the num- small decreases in pH have substantial ef- ber of anecdotal accounts and some recent fects on natterjack populations. Most prob- long-term studies suggesting amphibian ably, other organisms sympatric with nat- declines? On the one hand, it is essential terjack toads were also affected by that rigorous census studies of a represen- increased acidity. I am not suggesting that tative sample of amphibian populationsbe natterjack toads were affected more than initiated worldwide as a means of assessing other organisms. But in this case, at least, the directions, magnitudes, and agents of increased acidification seems to have changes in their numbers. How much in- caused a decline in populations of the nat- formation is needed before one can decide terjack toad. whether special efforts should be under- Several studies have shown that pH can taken to protect or restore populationsthat affect competition and predation between are declining? The conservative approach amphibians. These effects can potentially of withholding intervention until extinc- alter community structure. For example, tion rates are conclusively demonstrated Sadinsky and Dunson (1992) showed that to be unusually high might result in an survival of larval Rana sylvatica was en- unacceptable loss of populations or entire 94 HERPETOLOGICA [Vol. 50, No. 1 species. The opposite approach may give habitat to their former condition can then mistaken conclusions that a global decline be undertaken. is occurring when populations are simply Pechmann and Wilbur (1994) pose the exhibiting normal ranges of fluctuations. question, "What should be done?" and es- This could waste resources and political sentially call for continued monitoring. capital. Therefore, one must balance the They state that there is "no doubt that risk of lost credibility, which might seri- there have been numerous reductions and ously compromise future conservation ef- losses of amphibian populations world- forts in this and other arenas, against the wide because of human impacts" that in- cost of failing to respond to a potentially clude habitat destruction, pollution, and serious environmental crisis. the introductionof exotic species. Yet, they As stated in Blaustein et al. (in press), do not even suggest that tests, experimen- clearly an effort should be made to initiate tally, or otherwise, are needed for key hy- long-term monitoring programs for a broad potheses, until more long-term data are array of amphibian species, populations, available. For certain regions, there are habitats, and geographic regions. With such unquestionablyenough baseline data (both long-term records, one could evaluate (1) anecdotal and empirical) to generate spe- which species have, on average, more vari- cific hypotheses and to test them. able populations per turnover, (2) what the Testing key hypotheses can probably average probability of local extinction is best be done by implementing field ex- per turnover, (3) if local populations re- periments. For example, as referenced by cover from extinction, how long on aver- Pechmann and Wilbur (1994), several re- age recovery takes, and (4) what the spatial ports have suggested that the introduction scale is of local extinctions. All of these of bullfrogs (Rana catesbeiana) has caused comparative statements require a common the demise of some amphibian populations relative time scale: e.g., a population turn- in certain regions in western North Amer- over, so that differences in longevity of ica. One paper cited by Pechmann and individuals do not confound the compar- Wilbur was a review in a refereed journal isons. of the bullfrog problem (Hayes and Jen- Without such long-term data, one can- nings, 1986). The other two references that not unambiguously state that amphibian they cite (see Pechmann and Wilbur, 1994) populations are suffering unusual declines. were technical reports. Bullfrogs may in- The absence of information, however, is deed be causing the demise of certain am- not license to remain indifferent to the phibian populations, but the evidence for potential crisis. Among the examples of this remains equivocal. To my knowledge, long-term studies discussed above are sev- there have been no published papers in eral that have demonstrated recent rapid, the mainstream literature that tested this and sometimes widespread, declines or ex- hypothesis using field experiments. One tinctions of amphibian populations. In my could manipulate bullfrogs experimental- view, there is a sufficient number of such ly, in the field. Bullfrog tadpoles or adults cases to warrant investigation of potential could be added to some ponds where other links to human-caused environmental deg- amphibians are present but which had no radation. bullfrogs. If populations of other amphib- In my opinion, only if rigorous sampling ian species in ponds where there were no studies and ongoing assessments of the hy- bullfrogs (controls) did better than those pothesis of decline proceed simultaneous- where bullfrogs were added, the hypoth- ly can the decision to intervene be made esis that bullfrog presence is detrimental on solid scientific grounds. A case for in- to other amphibians is supported. In other tervention will be most convincing if the ponds, bullfrogs could be removed. If pop- observed demographic changes can be ulations of amphibians where removals linked to specific human-induced alter- have taken place do better than popula- ations of the habitat. Efforts to halt the tions where bullfrogs remained, the hy- declines and restore populations and their pothesis is supported. Of course, the ex- March 1994] HERPETOLOGICA 95 periments must be conducted in an ians, in certain regions, are declining. unbiased manner and with an adequate Whether or not specific declines are nat- number of replicates for the results to be ural or human-induced must be carefully meaningful. Additionalexperiments would assessed. (5) Like members of other taxa, be necessary to determine the mecha- under certain conditions, amphibians are nism(s) (e.g., competition, predation) by good bioindicatorsof environmental stress. which bullfrogs hamper populations of Their overall effectiveness as bioindicators other amphibian species. may differ with the type of stress, species, In fact, several studies are in progress and life history stage. (6) Long-term mon- that are incorporatingfield experiments in itoring, in conjunction with experimental the investigation of the effects of bullfrogs tests of key hypotheses, preferably in the on native frogs. For example, using field field, is warranted. (7) Criticizing reviews experiments and detailed observations, of meetings should result in a stiff fine. Sarah J. Kupferberg of the University of Acknowledgments. -Discussions with J. Connell California, Berkeley, is investigating the and W. Sousa helped me to formulate my ideas on role of bullfrog tadpoles on tadpoles of population dynamics. I thank D. Formanowicz, Jr. Rana boylii and Hyla regilla in the Coast G. Hokit, J. Kiesecker, R. Mason, B. Menge, D. Wake, Range of California. Joseph M. and S. Walls for critically reviewing the paper. J. Kiesecker, Pechmann provided useful comments on this reply. is conducting a similar study in Oregon. Assistance was provided by P. Marlowe, V. Stern- Hence, within 2-5 years, there should be wood, C. Sternwood, E. Mars, and B. Ohls. I thank some insight into the impact of bullfrogs R. Jaeger and H. Mushinsky for inviting me to write on amphibians. this paper. Financial support was kindly provided by the National Geographic Society and the National Pechmann and Wilbur (1994) suggest Science Foundation (BSR-9024880). that certain "hypothetical explanations" (e.g., increased ultraviolet radiation) may LITERATURE CITED be unworthy of investigation at this time BANKS, B., AND J. C. BEEBEE. 1987. Factors influ- because their study "distractsus from . .. encing breeding site choice by the pioneering am- seriousknown problems such as rain forest phibian Bufo calamita. Holarc. Biology 10:14-21. destruction or BANTLE, J. A., J. N. DUMONT, R. A. FINCH, AND G. stocking or exotic fish." LINDER. 1992. Atlas of Abnormalities: A Guide Quick dismissalof certain hypothetical ex- for the Performance of FETAX. Oklahoma State planations is close to scientific advocacy. University Press, Stillwater, Oklahoma. Justbecause there is scant knowledge about BARINAGA,M. 1990. Where have all the froggies certain agents that could potentially harm gone? Science 247:1033-1034. BEEBEE,T. J. C. 1977. Environmental change as a amphibian populations (as well as popu- cause of Natterjack toad (Bufo calamita) declines lations from other taxa), this does not mean in Britain. Biol. Conserv. 11:87-102. that one should not investigate them. BEEBEE,T. J. C., R. J. FLOWER,A. C. STEVENSON, S. T. PATRICK, P. G. APPLEBY, C. FLETCHER, C. CONCLUSIONS MARSH, J. NATKANSKI, B. RIPPEY, AND R. W. BATTARBEE. 1990. Decline of the Natterjack toad (1) The amphibian decline problem is Bufo calamita in Britain: Paleoecological, docu- part of the overall crisis in biodiversity. (2) mentary and experimental evidence for breeding Habitat destruction and habitat alteration site acidification. Biol. Conserv. 53:1-20. are undoubtedly the single most important BERVEN, K. A. 1990. Factors affecting population fluctuations in larval and adult stages of the wood cause for declines of species in all taxa, frog (Rana sylvatica). Ecology 71:1599-1608. including amphibians. (3) There probably BIRGE, W. J., J. A. BLACK,AND A. G. WESTERMAN. is no such thing as pristinehabitat. (4) There 1979. Evaluation of aquatic pollutants using fish are not enough long-term data on am- and amphibian eggs as bioassay organisms. Pp. 108- phibian populations to assess 118. In S. W. Neilsen, G. Migaki, and D. G. Scar- the overall pelli (Eds.), Animals as Monitors of Environmental significance of the amphibian decline Pollutants. National Academy of Sciences, Wash- problem. However, long-term studies are ington, D.C. appearing with increasing frequency and BIRGE,W. J., J. A. BLACK,A. G. WESTERMAN,AND in conjunction with some range reduction B. A. RAMEY. 1983. Fish and amphibian embryos-A model system for evaluating teratoge- data, there certainly is enough evidence to nicity. Fund. App. Toxicol. 3:237-242. suggest that populations of some amphib- BLAUSTEIN, A. R., D. G. HOKIT, R. K. O'HARA, AND 96 HERPETOLOGICA [Vol. 50, No. I

R. A. HOLT. Pathogenic fungus contributes to am- Introduction to Herpetology. W. H. Freeman, San phibian losses in the Pacific Northwest. Biol. Con- Francisco. serv.: In press. HAINES, T. A. 1981. Acidic precipitation and its BLAUSTEIN, A. R., AND D. B. WAKE. 1990. Declin- consequences for aquatic ecosystems: A review. ing amphibian populations: A global phenomenon? Trans. Am. Fish. Soc. 110:669-707. Trends Ecol. Evol. 5:203-204. HAYES, M. P., AND M. R. JENNINGS. 1986. Decline BLAUSTEIN, A. R., D. B. WAKE, AND W. P. SOUSA. of ranid frogs in western North America: Are bull- Amphibian declines: Judging stability, persistence frogs (Rana catesbeiana) responsible? J. Herpetol. and susceptibility of populations to local and global 20:490-509. extinctions. Conserv. Biol.:In press. HEYER, W. R., A. S. RAND, C. A. GONCALVEZDE BOWLER, J. K. 1977. Longevity of reptiles and am- CRUZ, AND 0. L. PEIXOTO. 1988. Decimations, phibians in North American Collections. Herpetol. extinctions, and colonizations of frog populations Circ. 6:1-32. in southeast Brazil and their evolutionary impli- BRADFORD, D. F. 1989. Allotopic distribution of cations. Biotropica 20:230-235. native frogs and introduced fishes in high Sierra HONEGGER,R. E. 1978. Threatened Amphibians Nevada lakes of California: Implication of the neg- and Reptiles in Europe. European Committee for ative effect of fish introductions. Copeia 1989:775- the Conservation of Nature and Natural Re- 778. sources-Council of Europe. Akademische Ver- . 1991. Mass mortality and extinction in a lagsgesellschaft, Wiesbaden, Germany. high-elevation population of Rana muscosa. J. Her- IUCN. 1990. IUCN Red List of Threatened Ani- petol. 25:174-177. mals. IUCN-The World Conservation Union, BRAGG, A. N. 1960. Population fluctuation in the Cambridge. amphibian fauna of Cleveland county, Oklahoma JACKSON, M. E., D. E. SCOTT, AND R. A. ESTES. 1989. during the past twenty-five years. Southwest. Nat. Determinants of nest success in the marbled sala- 5:165-169. mander (Ambystoma opacum). Can. J. Zool. 67: CLARKSON, R. W., AND J. C. RORABAUGH. 1989. 2277-2281. Status of leopard frogs (Rana pipiens complex) in JAEGER,R. G. 1970. Potential extinction through Arizona and southeastern California. Southwest. competition between two species of terrestrial sal- Nat. 34:531-538. amanders. Evolution 24:632-642. CONNELL, J. H, AND W. P. SOUSA. 1983. On the . 1980. Density-dependent and density in- evidence needed to judge ecological stability or dependent causes of extinction of a salamander persistence. Am. Nat. 121:789-824. population. Evolution 34:617-621. COOKE, A. K. 1981. Tadpoles as indicators of harm- JAYAPRAKASH,R. I., AND M. N. MADHYASTHA. 1987. ful levels of pollution in the field. Environ. Poll. Toxicities of some heavy metals to the tadpoles of (Series A) 25:123-133. frog, Microhyla ornata (Dumeril and Bibron). Tox- CORN, P. S., AND J. C. FOGLEMAN. 1984. Extinction icol. Let. 36:205-208. of montane populations of the northern leopard KAGARISE SHERMAN, C., AND M. L. MORTON. 1993. frog (Rana pipiens) in Colorado. J. Herpetol. 18: Population declines of Yosemite toads in the eastern 147-152. Sierra Nevada of California. J. Herpetol. 27:186- CRUMP, M. L. 1989. Life history consequences of 198. feeding versus non-feeding in a facultatively non- LECLAIR, R., AND J. CASTANET. 1987. A skeleto- feeding toad larva. Oecologia 78:486-489. chronological assessment of age and growth in the CRUMP, M. L., F. R. HENSLEY, AND K. L. CLARK. frog Rana pipiens Schreber (Amphibia, Anura) from 1992. Apparent decline of the golden toad: Un- Southwestern Quebec. Copeia 1987:361-369. derground or extinct? Copeia 1992:413-420. LIEM, D. S. 1973. A new genus of frog of the family DUELLMAN, W. E., AND L. TRUEB. 1986. Biology Leptodactylidae from SW Queensland, Australia. of Amphibians. McGraw-Hill, New York. Mem. Queens Mus. 16:459-476. DUNSON, W. A., R. L. WYMAN, AND E. S. CORBETT. MCALLISTER, K. R., AND B. LEONARD. 1990. 1989 1992. A symposium on amphibian declines and Progress Report-Past Distribution and Current habitat acidification. J. Herpetol. 26:349-352. Status of the Spotted Frog in Western Washington. FANNING, J. C., M. J. TYLER, AND D. J. C. SHEARMAN. Washington Department of Wildlife, 1989 Progress 1982. Converting a stomach to a uterus: The mi- Report. croscopic structure of the stomach of the gastric . 1991. 1990 Progress Report-Past Distri- brooding frog Rheobatrachus silus. Gastroenter- bution and Current Status of the Spotted Frog in ology 82:62-70. Western Washington. Washington Department of FELLERS, G. M., AND C. A. DROST. 1993. Disap- Wildlife, 1990 Progress Report. pearance of the Cascades frog Rana cascadae, at MCNEELY, J. A., K. R. MILLER, W. V. REID, R. A. the southern end of its range. Biol. Conserv. 65: MITTERMEIER, AND T. B. WERNER. 1990. Con- 177-181. serving the World's Biological Diversity. IUCN, FONOVICH DE SCHROEDER, T. M., AND A. M. PECHEN WRI, CI, WWF-US, The World Bank, Washing- DE D'ANGELO. 1991. Dieldrin effects on phos- ton, D.C. pholipid and phosphoinositide metabolism in Bufo MIERLE, G., K. CLARK, AND R. FRANCE. 1986. The arenarum oocytes. Comp. Biochem. Physiol. 98C: impact of acidification on aquatic biota in North 287-292. America: A comparison of field and laboratory re- GOIN, C. J., 0. B. GOIN, AND G. R. ZUG. 1978. sults. Water, Air, Soil Poll. 31:593-604. March 1994] HERPETOLOGICA 97

NATIONAL SCIENCE BOARD. 1989. Loss of Biolog- STEBBINS, R. C. 1985. Western Reptiles and Am- ical Diversity: A Global Crisis Requiring Interna- phibians. Houghton Miffin, Boston. tional Solutions. Report NSB-89-171. National Sci- TANGLEY, L. 1990. Croaked. Earthwatch, October ence Foundation, Washington, D.C. 1990:8-10. NUSSBAUM, R. A., E. D. BRODIE, JR., AND R. M. STORM. TYLER, M. J. 1991. Declining amphibian popula- 1983. Amphibians and Reptiles of the Pacific tions-A global phenomenon? An Australian per- Northwest. University Press of Idaho, Moscow, Ida- spective. Alytes 9:43-50. ho. TYLER, M. J., AND D. B. CARTER. 1981. Oral birth OLSON, D. H., BLAUSTEIN, A. R., AND R. K. O'HARA. of the gastric brooding frog Rheobatrachus silus. 1986. Mating pattern variability among western Anim. Behav. 29:280-282. toad (Bufo boreas) populations. Oecologia 70:351- TYLER, M. J., AND M. DAVIES. 1985. The gastric 356. brooding frog Rheobatrachus silus. Pp. 469-470. PECHMANN,J. H. K., D. E. SCOTT,R. D. SEMLITSCH, In G. Grigg, R. Shine, H. Ehmann (Eds.), The Bi- J. P. CALDWELL,L. J. VITT, AND W. GIBBONS. ology of Australasian Frogs and Reptiles. Surrey 1991. Declining amphibian populations: The Beatty, Chipping Norton, New South Wales, Aus- problem of separating human impacts from natural tralia. populations. Science 253:892-895. VITT, L. J., J. P. CALDWELL, H. M. WILBUR, AND D. PECHMANN,J. H. K., AND H. M. WILBUR. 1994. C. SMITH. 1990. Amphibians as harbingers of de- Putting declining amphibian populations into per- cay. BioScience 40:418. spective: Natural fluctuations and human impacts. WAKE, D. B. 1991. Declining amphibian popula- Herpetologica 50:65-84. tions. Science 253:860. PHILLIPS, K. 1990. Where have all the frogs and WAKE, D. B., AND H. J. MOROWITZ. 1991. Declin- toads gone? BioScience 40:422-424. ing amphibian populations-A global phenome- PIERCE, B. A. 1985. Acid tolerance in amphibians. non? Findings and recommendations. Report to BioScience 35:239-243. Board on Biology, National Research Council, on POWER, T., K. L. CLARK, A. HARFENIST, AND D. B. workshop in Irvine, California 19-20 February 1990; PEAKALL. 1989. A Review and Evaluation of the reprinted 1991. Alytes 9:33-42. Amphibian Toxocological Literature. Technical WALLS, S. C., A. R. BLAUSTEIN, AND J. J. BEATTY. Report Series No. 61, Canadian Wildlife Service. 1992. Amphibian biodiversity of the Pacific ROFF, J. C., AND R. E. KWIATKOWSKI. 1977. Zoo- Northwest with special reference to old-growth plankton and zoobenthos communities of selected stands. Northwest Environ. J. 8:53-69. northern Ontario lakes of different acidities. Can. WARNER, S. C., J. TRAVIS, AND W. A. DUNSON. 1993. J. Zool. 55:899-911. Effect of pH variation on interspecific competition SADINSKY, W. J., AND W. A. DUNSON. 1992. A mul- between two species of hylid tadpoles. Ecology 74: tilevel study of effects of low pH on amphibians of 183-194. temporary ponds. J. Herpetol. 26:413-422. WILBUR, H. M. 1980. Complex life cycles. Ann. SAVAGE,J. M. 1966. An extraordinary new toad Rev. Ecol. Syst. 11:67-93. (Bufo) from Costa Rica. Revista Biol. Trop. 14:153- WILSON, E. 0. (Ed.). 1988. Biodiversity. National 167. Academy Press, Washington, D.C. SEMB-JOHANSSON, A. 1989. Padden (Bufo bufo)- 1992. Diversity of Life. Harvard University Et stebarn inorsk zoologi. Fauna 42:174-179. Press, Cambridge, Massachusetts. SIMBERLOFF, D. 1986. The proximate causes of extinction. Pp. 259-276. In D. M. Raup and D. Ja- Accepted: 10 September 1993 blonski (Eds.), Patterns and Processes in the History Associate Editor: Henry Mushinsky of Life. Dahlem Konferenzen, 1986, Springer-Ver- lag, Berlin.