Effect of Timber Harvesting on Southern Appalachian Salamanders | WFS 493, Papers of Animal Biology

Download Effect of Timber Harvesting on Southern Appalachian Salamanders | WFS 493 and more Papers Animal Biology in PDF only on Docsity! Effects of Timber Harvesting on Southern Appalachian Salamanders JAMES W. PETRANKA MATTHEW E. ELDRIDGE KATHERINE E. HAIFY Department of Biology University of North Carolina Asheville, NC 28804, U.S.A. Abstract: We compared the species richness and abundance o f salamanders on six recent clearcuts (< 10 years old) with that o f salamanders on 34 mature forest stands (>50 years old) in southern Appalachian forests in western North Caro- lina~ U.XA. Catches o f salamanders f rom plots in mature forest stands were about f ive times higher than those on recent clear6"u~ Almost all species and major taxonomic groups o f salamanders were adversely affected by timber re- moval Mean number o f species collected per p lo t was about twice as great in mature forest stands as in cleam'utg Anal- yses o f stand age versus salamander catch for 47plots indi- cate that 50-70 years are required for populations to return topredlsturbance levels fo l lowing cutting. We conservatively estimate that clearcutting in U.X national forests in western North Carolina results in a loss o f nearly 14 million sala- manders annually. I t also i$ chronically reducing regional populat ions by more than a quarter o f a billion sala- manders (9%) below that which could be sustained i f ma- ture forests were not cut. Efectos de la tala del bosque sobre las salamandras en el sur de los Apalaches R e s u m e n : Nosotros comparamos ia riqueza de especies y abundancia de ias salamandras en seis recientes cortas to- tales de bosque (<10 afu~s) con la de rodales maduros (>50 a~gs) en los bosques del sur de los Apalache~ en el oeste de Carolina del Nort¢ Estados Unidog La captura de salaman- dras por p lo t en bosques de rodales maduros f ue aproximd- damente cinco veces mayor queen aquellos cortados recidn- temente Casi todas las especies y los mayores grupos taxo- n6micos de salamandras fueron adversamente afectados por ia tala La media del namero de especies capturada por p lo t rue aproximddamente dos veces mayor en los rodales de bosques maduros q u e e n los cortados recid, nteraente Los andllsts de edad del rodal versus captura de salamandras pare 47 plots indican que entre 50-70 a~os son necesartos para que las poblaciones retornen a los niveles prevtos a la perturbaci6n ocasionada por la tala del bosque Nosotros estimamos en una forma conservadora que la tala de los Bosques Nacionales en el oeste de Carolina del Norte trae como consecuencia una p6~,dida de alrededor de 14 millones de salamandras anualmente La tala tambi~a estd reduciendo cr6ntcamente las poblaciones a nivel re. gional en rods de un cuarto de billones de salamandras (9%) por debajo del nivel que podrfa set sostenido por los bosques maduros si no hubieran sido cortado~ Paper submitted October 2L 1991; revised manuscript accepted Apm 9, 1992. 363 Conservation Biology Volume 7, No. 2, June 1993 364 Elgecls of Timber Halvesti~ on SalRmanders PetranRa et al. Introduction Salamanders are important ecological components of many forest ecosystems in North America. In mesic for- ests in the eastern U.S., salamanders are often the most abundant group of vertebrates in both numbers and bio- mass (Burton & Likens 1975eg 1975/7' Hairston 1987). Salamanders also play important roles in food webs, where they prey upon small invertebrates and serve as a food source for an array of larger predators (Pough et al. 1987; Corn & Bury 1989). Despite their importance in many forest systems, sala- manders have often been neglected in forest manage- ment studies (Bury et al. 1980; Pough et al. 1987). Stud- ies in the Pacific N o r t h w e s t indica te that many salamander species are adversely affected by t imber harvesting (see Bury & Corn 1988; Raphael 1988; Wel- sch & Lind 1988; Corn & Bury 1989; Welsh 1990). Relatively few studies have been conducted in the east- ern United States (Bennett et al. 1980; Enge & Marion 1986; Blymer & McGinnes 1977; Pough et al. 1987; Ash 1988; Buhlmann et al. 1988), and most are difficult to interpret because of lack of replication or pseudorepli- cation (Hurlbert 1984). Nonetheless, these studies col- lectively suggest that t imber harvesting is detrimental to salamanders in eastern forests. The southern Appalachians have an extraordinarily rich and and abundant salamander fauna that in many respects is unparalleled worldwide. As many as 35 spe- cies belonging to five families occur in the Appalachian region of Nor th Carolina a l o n e (Conant & Collins 1991). In addition, the local biomass of salamanders in southern Appalachian forest communities often exceeds that of all other vertebrate predators combined (Hair- ston 1987). Prior to the 1960s, t imber harvesting often involved intense selective cutting in which all but a few large trees were removed from timbered tracts. Since the 1960s, clearcutting has almost completely replaced selective cutting as the preferred method of timber har- vesting by the U.S. Forest Service in the southern Appa- lachians. Although most clearcuts are relatively small (typically < 1 0 - 1 2 ha), they are often cut in larger blocks that are separated by narrow belts of uncut forest. Almost no published data are available on the impact of clearcutting on southern Appalachian salamanders other than that of Ash (1988), who found that cutting completely eliminated a local population of Plethodon jordanL Here, we report on the effects of clearcutting on salamanders in western North Carolina. We also pro- vide data on the recovery times of local populations following t imber removal, and on the regional impact of t imber removal on salamander abundance. Methods Salamanders were sampled be tween May 16 and August 8, 1991, from 47 sample plots in and adjoining the Craggy Mountains, Pisgah National Forest, Buncombe County, North Carolina (Fig 1). The work was part of a collaborative effort to document and monitor long-term changes in biodiversity in the southern Appalachians. The study area is located about 25 km northeast of Asheville and encompasses about 6000 ha of mostly mixed mesophytic deciduous forests (Fig 1). Forty-one sample sites were selected from equidistant points es- tablished by randomly placing a grid on a topographic map of the study area. These stands were selected to provide an unbiased estimate of the relative abundance and diversity of salamanders in different communities within the Craggy Mountains. The stands varied in age from 19 to 120 years and ranged in elevation from 817 to 1667 m. Six clearcuts be tween 2 and 10 years old were also sampled to determine the effects of t imber harvesting on species abundance and diversity (Fig. 1). These ranged in elevation from 969 to 1280 m. At all sites, a 50 × 50 m plot was established parallel and perpendicular to prevailing contours. Plots were centered on permanent plot markers, and each was sam- pled once by walking roughly parallel transects and turning all movable rocks, logs, bark, and other surface objects that could prov ide cover for salamanders. Cracks and crevices in rock outcrops were inspected for crevice-dweUing salamanders, and unsubmerged stones and logs in streams or seepages w e r e t u rned and searched. Search time varied from 1.33 to 4 people- ,, = \ t I I t B°ar~tsville e ° • • • • e e • ~ • e e Z ' Ikm \ (LLS~ Figure 1. Location o f the general s tudy area in west- ern North Caroling Closed circles indicate the loca. t ion o f p lo ts a long grid coordinates Open circles indicate the location o f clearcut~ Conservat ion B i o l o g y Volume 7, No. 2 , J u n e 1995 Petranka et al. Effects of Timber Harvesting on Salamanders 367 regrowth (p = 0.001 ), wi th stand age explaining about 37% of the among-plot variation in total catch. For stands more than 70 years old, total catch of sala- manders was independent of stand age (p = 0.63). To- tal number ofDesmognathus (p = 0.05; r 2 = 0.15) and total number of Plethodon (p = 0.0004; r 2 = 0.35) captures also correlated significantly with age for stands less than 70 years old, but not for stands more than 70 years old (p = 0.93 for Plethodon; p -- 0.25 for Des- mognathus). Maximum catch pe r plot occur red in stands 51 -70 years old, and was slightly less in older plots. General trends for total Plethodon caught and total Desmognathus caught were similar (Fig. 3) and suggest that salamander communit ies require about 5 0 - 70 years to recover to preharvest levels. This estimate may be conservat ive for c learcut sites because it is based on combined data for clearcut and selectively cut sites. Complete estimates of recovery rates on clearcut stands are not available because most clearcut sites are currently less than 30 years old. Discussion Impact of Clearcutting on Local Populations and Communities We found that clearcutting strongly depletes local pop- ulations of salamanders and reduces local species rich- hess. We estimate that about 75-80% of salamanders in mature stands are lost following t imber harvesting by clearcutting. This estimate is conservative because it in- chides plots as old as 10 years that may have undergone partial recovery. Although the fate of salamanders on recent clearcuts is unclear, we assume that most died following t imber removal and that few salamanders dis- persed to surrounding forests. Most southern Appalachian salamanders are sensitive to environmental disturbances that modify prevailing temperature , humidity, or soil mois ture regimes be- cause adults lack lungs and exchange gases almost en- tirely by cutaneous respiration. Because their skin must be kept moist to facilitate gas exchange, adults generally restrict their activity to moist forest-floor microhabitats and are active on the ground surface only at night when relative humidities are high. Adult plethodontid sala- manders may rapidly dehydrate if microhabitats beco me too dry (see Duellman & Trueb 1986). Clearcutting de- grades forest-floor microhabi ta ts for salamanders by eliminating shading, reducing leaf litter, increasing soil- surface temperature, and reducing soil-surface moisture (Bury 1983; Ash 1988; Raphael 1988; Welsch 1990). Consequently, it is likely that most animals died f rom physiological stress following the removal of trees from sites. Increased sedimentat ion and general deterioration of s tream quality may also have contr ibuted to the de- d i n e of species wi th aquatic larval stages (Corn & Bury 1989). Dispersal of plethodontid salamanders from plots fol- lowing cutting is unlikely because adults have small home ranges and are strongly philopatric. Experimen- tally displaced Plethodon and Desmognathus species readily home back to their place of capture, and show no tendency to disperse away from home ranges when disturbed or handled (Duellman & Trueb 1986; Hair- ston 1987). Although certain plethodontids, such as Desmognathus and Eurycea, move seasonally to and from breeding sites, they are highly sedentary during most of the year. Because we relied on surface counts to estimate rel- ative population size, it is possible that the reduct ion in numbers that we observed on clearcuts does not reflect true population declines. One alternative explanation for the apparent decline in salamanders on clearcuts is that decreases in surface moisture following t imber re- moval forced salamanders to move into subsurface re- treats during the day. If this were the case, relatively few animals would be expec ted to be taken in dayt ime searches. However, studies by Ash (1988) based on nighttime searches of clearcut sites indicate that this is not the case. Ash (1988) intensively studied the effects of clearcutting on P. jordani on Rich Mountain South near Highlands, North Carolina, and noted a near com- plete elimination of this species the second summer af- ter cutting. After the fourth summer, no salamanders were found on cut plots during nightt ime searches. Changes associated with the decline of P. jordani in- cluded elimination of most shading during the first sum- mer and a significant increase in the amount of bare soil. Our findings are consistent with those of Ash (1988) and others (Blymer & McGinnes 1977; Bury 1983; Enge & Marion 1986; Pough et al. 1987; Bury & Corn 1988; Corn & Bury 1989), which indicate that logging signif- icantly reduces amphibian species abundance and diver- sity. Impact of Clearcutting on Regional Populations Because most individuals of forest-floor species such as Plethodon jordant and Desmognathus ochrophaeus are u n d e r g r o u n d at any g iven t ime, d a y t i m e sur face searches usually uncover only a small percentage of the existing population. Direct counts and mark-recapture studies show that many species in Appalachian forests occur at very high densities. Large Plethodon species like P. jordani and P. glutinosus typically have densities of 0.2-0.9 animalsdm 2 of forest floor, while densities of Desmognathus species can be much higher on rock faces and near streams and seepages (Table 2). These estimates are conservative because they exclude larval stages of Desmognathus as well as young Plethodon that spend their first year or so after hatching under- ground (Hairston 1983). Based on data provided in Table 2, we conservatively Conservation Biology Volume 7, No. 2, June 1993 368 Effects of Timber H~rvesting on S~nanders Petranica et al. Table 2. Estimated densities of salamander species based on mark-recapture or direct counts of populations in the Appalachian Mountains. Species Location Number/m 2 Source P. jordani southern Blue Ridge 0.18 P. jordani Great Smoky Mountains 0.26 P. jordani Balsam Mountains 0.33 P. jordani Great Smoky Mountains 0.86 P. jordani Howards' Knob (Boone, N.C.) 0.50 P. jordani (average of all sites) 0.43 P. glutinosus Great Smoky Mountains 0.23 P. cinereus Blackrock Mountain, Virginia 2.2 D. ochrophaeus Howards' Knob (Boone, N.C.) 2.05 D. ochrophaeus southern Blue Ridge 18--41 * D. ochrophaeus southern Blue Ridge 6--7* D. ochrophaeus Nantahaia Mountains 0.70** Ash 1988 Ash 1988 Ash 1988 Merchant 1972 Howard 1987 Merchant 1972 Jaeger 1980 Howard 1987 Huheey & Brandon 1973 Tilley 1980 Hairston 1986 * Rock- facepopu la t ions whose densities are generally much higher than popu la t ions f o u n d on the fores t f l o o r ** Est imate o f forest- f loor densi ty based on repeated removals during a single n ight Remova l o f 0. 7 animals /ra 2 d id no t s igni f icant ly reduce the n u m b e r o f an ima l s emerging f r o m underground retreats, so the actual ly density was p re sumab ly f a r greater. Al l sites are in North Carolina except where noted estimate that Ple thodon species collectively average 0.5 animals/m 2 of forest floor in southern Appalachian for- ests. In the Craggy Mountains, Ple thodon species con- stituted only 48% of all salamanders collected (Table 1). Thus, a conservative estimate is that there is at least one salamander/m 2 of forest floor or about 10,000 sala- manders/ha in mature forest stands in the Craggy Moun- tains. This value seems a reasonable first-order approx- imation of average densities of salamanders in the southern Appalachians as a whole, although in optimal habitats such as mesic cove forests or stream banks, local densities may be much higher. Howard (1987) estimated 22,608 salamanders/ha in mesic forests near Boone, North Carolina. However, densities of sala- manders on dry ridgetops and low-elevation forests with sandy soils are presumably much lower. Timber harvesting by clearcutting of national forests in western North Carolina has averaged 1,709 ha per year between 1981 and 1990 (personal communication from Ed Brown of the U.S. Forest Service). Assuming an 80% loss of resident animals following cutting and an average density of 10,000 salamanders/ha, we estimate that clearcutting on U.S. Forest Service lands in western North Carolina has eliminated an average of 13.7 million salamanders annually in recent years. Although the ab- solute number of animals being lost annually is substan- tial, it constitutes only about 0.34% of the estimated total number of salamanders found in national forests in western North Carolina. (This estimate is based on the percent of national forest holdings cut annually, after correcting for 20% survival in clearcuts). Despite the fact that the annual loss of animals is less than 0.5%, significant reductions in regional popula- tions could occur because of the long recovery period required for populations to return to predisturbance levels following timber harvesting. An estimate of the long-term effects of cutting on southern Appalachian populations can be obtained by using data on current stand age together with age-specific recovery rates de- rived from the regression of total salamander catch ver- sus stand age. Using the regression model for stands less than 70 years old, we estimate that if forests continue to be cut at 1981-1990 rates, regional populations will be chronically reduced by about 8.5%, or 267 million an- imals below the numbers which could be sustained in mature forests. This estimate assumes that 80% of the salamanders are lost after removing timber, that 0.34% of the total salamanders on national forests are elimi- nated annually by cutting, and that populations fully re- cover in 50 years. This reduction percentage is similar to that which has occurred historically during the last 50 years, based on a current estimate that 16% of U.S. Forest Service lands in western North Carolina are less than 50 years old (data provided by E. Brown, U.S. For- est Service). Although these estimates are intended only as rough approximations, they provide a feel for the magnitude of losses related to timber harvesting. Stiven and Bruce (1988) provided evidence that tim- ber harvesting may influence the genetic diversity of local populations of black-bellied salamanders in the southern Appalachians. Our data suggest that the deple- t ion of local populat ions of terrestrial species by clearcutting may be of sufficient magnitude to produce bottlenecks that would significantly alter genetic diver- sity. There are now conflicting theoretical views as to how population bottlenecks affect genetic diversity. Some have argued that bottlenecks should lower ge- netic diversity (Nei et al. 1975), while others have taken the opposite stance (Bryant et al. 1986; Goodnight 1987). Any resolution of the problem will require a more complete understanding of how patch dynamics, the scale of disturbance, and local population dynamics interact to influence regional genetic diversity. We consider the chronic depletion of populations in national forests in western North Carolina by more than a quarter of a billion animals (9%) to be significant from Conservation Biology Volume 7, No. 2 , June 1993 Petmik~ et al. Effects of Timber Harvesting on Salamanders 369 a regional perspective, particularly when one considers that a significant port ion of the eastern U.S. has been deforested since its colonization by Europeans. These losses seem to be sustainable in the sense that none of the species studied is in eminent danger of extinction. However, the chronic deple t ion of salamanders is clearly contributing to a decline in the general health of south Appalachian forest communities, particularly at the local level were salamander communities are se- verely perturbed during timber removal. The estimated average reduction of 9% in numbers does not reflect the fact that the intensity of logging varies regionally. Timber removal tends to be concen- trated in areas with moderate Mopes that are outside of designated wi lderness or rec rea t ion areas. Conse- quently, chronic reductions of salamander numbers in many regions far exceeds 9%. Whether alternative har- vesting methods would reduce those losses is uncertain. Selective cutting would presumably have less impact on local salamander populations because the loss of shade and leaf litter following cutting would be reduced. In order to harvest a specific volume of timber, however, selective cutting requires that more acreage be cut than when clearcutting. Exactly how this tradeoff would ul- timately affect salamander abundance remains to be de- termined. Acknowledgments Special appreciation is extended to Sally Browning, Chris Ulrey, and Henry McNab for their assistance dur- ing the course of the study, and to Joe Pechmann and an anonymous reviewer for constructive comments on our w o r e This work was funded in part by Contract 11-118 of the National Forests in North Carolina. H ~ Cited Ash, A. 1988. Disappearance of salamanders from clearcut plots. Journal of the Elisha Mitchell Scientific Society 104:116--122. Bennett, S. H.,J. W. Gibbons, andJ. GlanviUe. 1980. Terrestrial activity, abundance, and diversity of amphibians in differently managed forest types. American Midland Naturalist 103:412- 616. Blymer, M.J., and B. McGinnes. 1977. Observations on possi- ble detrimental effects of clearcutting on terrestrial amphihi- arts. Bulletin of the Maryland Herpetological Society 13:79-83. Bryant, E.H., S.A. McCommas, and L.C. Combs. 1986. The effect of an experimental bottleneck upon quantitative genetic variation in the housefly. Genetics 114:1191-1211. Bub.lmann, K. A., C. A. Pague, J. C. Mitchell, and R. B. Glasgow. 1988. Forestry operations and terrestrial salamanders: tech- niques in a study of the Cow Knob salamander, Plethodon punctatug Pages 38-44 in R. C. Szaro, K. E. Severson, and D. IL Patton, editors. Management of amphibians, reptiles, and mam- mals in North America. Technical Report RM- 166. U.S. Depart- ment of Agriculture, Forest Service, Rocky Mountain Forest and Range Experiment Station, Fort Collins, Colorado. Burton, T. M., and G. E. Likens. 1975at Energy flow and nutri- ent cycling in salamander populations in the Hubbard Brook experimental forest, New Hampshire. Ecology 56:1068-1080. Burton, T. M., and G. E. Likens. 1975b. Salamander populations and biomass in the Hubbard Brook experimental forest, New Hampshire. Copeia 1975:541-546. Bury, R.B. 1983. Differences in amphibian populations in logged and old-growth redwood forest. Northwest Science 57:167-178. Bury, R.B., and P. Corn. 1988. Responses of aquatic and streamside amphibians to timber harvest: a review. Pages 165- 181 in ILJ. Raedeke, editor. Streamside management: riparian wildlife and forestry interactions. Contribution No. 59. Insti- tute of Forest Resources, University of Washington, Seattle, Washington. Bury, R.B., H.W. Campbell, and N.J. Scott. 1980. Role and importance of nongame wildlife. Pages 197-207 in D. L Flath and J. R. Torres, editors. Transactions of the 45th North Amer- ican Wildlife and Natural Resources Conference, 1980. Wild- life Management Institute, Washington, D.C. Conant, 1L, and J. T. Collins. 1991. A field guide to reptiles and amphibians of eastern and central North America. Houghton Mifflin, Boston, Massachusetts. Corn, P. S., and R. B. Bury. 1989. Logging in western Oregon: responses of headwater habitats and stream amphibians. Forest Ecology and Management 29:39-57. Duellman, W. E., and L. Trueb. 1986. Biology of amphibians. McGraw-Hill, New York, New York. Enge, K. M., and W. Marion. 1986. Effects of clearcutting and site preparation on herpetofauna of a north Florida llatwoods. Forest Ecology and Management 14:177-192. Goodnight, C.J. 1987. On the effect of founder events on epi- static genetic variance. Evolution 41:80-91. Hairston, N. G., Sr. 1983. Growth, survival, and reproduction of Plethodon jordani: tradeoffs between selection pressures. Copeia 1983:1024-1035. Halrston, N.G., St. 1986. Species packing in Desmognathus salamanders: Experimental demonstration of predation and competition. American Naturalist 127:266-291. Halrston, N. G., St. 1987. Community ecology and salamander guilds. Cambridge University Press, Cambridge, England. Howard, T.A. 1987. Population and biomass estimates in four species of terrestrial plethodontid salamanders. M.S. thesis. Ap- palachian State University, Boone, North Carolina. Huheey, J. E., and R.A. Brandon. 1973. Rock-face populations of the mountain salamander, Desmognathus ochropbaeu~ in North Carolina. Ecological Monographs 43:59-77. Conservation Biology Volume 7, No. 2, June 1993