3675_C013.fm Page 253 Friday, June 22, 2007 1:53 PM 13 Conserving Vernal Pool Amphibians in Managed Forests Phillip G. deMaynadier and Jeffrey E. Houlahan CONTENTS Introduction............................................................................................................254 Natural History......................................................................................................255 A Complex Mosaic....................................................................................255 Philopatry and Movement Ecology...........................................................256 Vernal Pool–Forestry Relationships......................................................................257 Vernal Pool Basin Relationships...............................................................257 Physical Integrity.............................................................................257 Hydrology.........................................................................................258 Water Quality...................................................................................259 Forest Canopy Relationships.....................................................................260 Clearcutting......................................................................................260 Partial Harvesting and the Canopy Continuum...............................261 Forest Floor Relationships.........................................................................263 Forest Litter......................................................................................264 Coarse Woody Debris (CWD).........................................................265 Conservation Recommendations...........................................................................265 Habitat Management Guidelines for Preharvest Planning........................268 Habitat Management Guidelines for Harvest Operations.........................269 Vernal Pool Depression....................................................................269 Vernal Pool Protection Zone............................................................271 Vernal Pool Life Zone.....................................................................272 Summary................................................................................................................273 Acknowledgments..................................................................................................274 References..............................................................................................................275 253 3675_C013.fm Page 254 Friday, June 22, 2007 1:53 PM 254 Science and Conservation of Vernal Pools in Northeastern North America INTRODUCTION Arguably the most conspicuous of natural elements in the Northeast are trees and for good reason, as the region ranks among the most forested in North America. All northeastern states and provinces are more than half forested, with forests in northern portions of the region dominating as much as 85% (New Brunswick) to 89% (Maine) of the landscape (Smith et al. 1994; Natural Resources Canada 2006). And yet it has not always been this way, with the region among the few in the world that is actually more forested today than it was 100 years ago (Foster 1995). Paradoxically, while northeastern landscapes today contain much of their natural forests they also host some of the densest, longest-settled human populations in North America. Starting in the early 1700s Europeans cleared much of the region’s forests for agriculture. However, beginning in the mid-1800s, farming declined across the Northeast with the result that abandoned pastures and fields were again reclaimed by forests whose legacy persists to this day (Irland 1982). The ebb, recovery, and present dominance of forestland profoundly influenced the historic distribution and status of the region’s wildlife (Litvaitis 1993; Wilcove 1999). Species requiring expanses of mature, mast-producing forest declined to a level of near-extinction (e.g., wild turkey, Meleagris gallopavo) or total extinction (e.g., passenger pigeon, Ectopistes migratorius). Other wide-ranging animals (e.g., timber wolf, Canis lupus, and woodland caribou, Rangifer tarandus caribou) were exposed to unsustainable levels of hunting and persecution, aggravated by the decline and fragmentation of previously remote forestlands. Fortunately, the region’s forest-dwelling frogs, toads, and salamanders neither required exceptionally large blocks of continuous forest nor attracted much hunting pressure or exploitation, with the result that our amphibian fauna remains mostly intact. More so perhaps than any other region in North America, the Northeast’s forests and the fate of the wildlife found here lie in private hands, by ownership or lease (i.e., Crown lands, Canada). As much as 90% of the commercial forest land in New England, for example, is privately owned (Irland 1982; Smith et al. 1994). Further- more, most of the region’s private woodlands are subject to varying levels of forest management intensity, ranging from light firewood harvesting to large-scale com- mercial pulp and saw log production. Indeed, with private forestlands dominating the natural landscape, timber management is by definition the region’s most wide- spread land-use practice. Understanding forestry–wildlife relationships and, for our purposes, forestry–amphibian relationships, in particular, is thus critical if biologists, foresters, and landowners are to make informed decisions when affecting the quality and extent of managed forest habitat in the Northeast. Although still lagging behind that of other vertebrate groups, an impressive body of knowledge is accumulating on the effects of forest management practices on amphibians, mostly in North America (see reviews by deMaynadier and Hunter 1995 and Welsh and Droege 2001). Still, research on the specific effects of timber man- agement on vernal pools and their fauna remains limited. This is of special concern in the Northeast where a high proportion (27 spp. or ~56%) of the salamander, frog, and toad fauna frequent vernal pool ecosystems for breeding, development, foraging, and hibernation (Chapter 7, Semlitsch and Skelly). Furthermore, vernal pools are 3675_C013.fm Page 255 Friday, June 22, 2007 1:53 PM Conserving Vernal Pool Amphibians in Managed Forests 255 nearly ubiquitous in northeastern forest landscapes where densities range as high as 13.5 pools/km2 (35 pools/mi2; Calhoun et al. 2003). As such, we suggest that it is nearly impossible to avoid at least incidental impacts, positive or negative, on habitat for vernal pool breeding wildlife during the course of most timber harvesting oper- ations in northeastern forests. The purpose of this chapter is to introduce readers to specific aspects of the biology of pool-breeding amphibians relevant to forest management and to provide a review of the limited, but growing body of research concerning vernal pool amphibian responses to common forest harvesting practices. Additionally, we use the literature on forestry–vernal pool relationships to inform the development of specific Habitat Management Guidelines for conserving pool-breeding amphibians in managed landscapes. Our premise is that forest management, if practiced in an ecologically sensitive manner, is among the most compatible consumptive land uses for conserving important elements of vernal pool habitat. NATURAL HISTORY Although amphibians are probably the most abundant vertebrate group in northeast- ern forests (Burton and Likens 1975; Hairston 1987), their small size, nocturnal activity, and often fossorial habits make them relatively inconspicuous and difficult to study. Consequently, most forest and wildlife managers are more familiar with the ecology of birds, mammals, and fish that has dominated investigations of forest wildlife relationships to date (Gibbons 1988). We provide a brief introduction to the ecology of vernal pool-breeding amphibians, focusing on aspects of their natural history relevant to forest ecosystems and their management. A COMPLEX MOSAIC Most pool-breeding amphibians have complex life cycles (Wilbur 1980), beginning as aquatic eggs, hatching to gilled larvae, and metamorphosing into terrestrial, lunged adults within a few weeks (e.g., eastern spadefoot toad, Scaphiopus holbrookii) to a few months (e.g., mole salamanders, Ambystoma spp.) of hatching. The habitat mosaic required to host self-sustaining populations of pool-breeding amphibians is also complex and generally comprised of (1) temporary to semipermanent pools lacking fish (for adult breeding and larval development), (2) terrestrial foraging, resting, and overwintering sites, often spatially removed from breeding pools, and (3) a mostly forested matrix permeable to migrating adults and dispersing juveniles. The ecology of the aquatic phase of pool-breeding amphibians is relatively well- studied (reviewed by Duellman and Trueb 1994, Alford 1999) with important effects on larval fitness and performance documented from both biotic (mainly competition and predation; Wilbur 1980; Hairston 1987), and abiotic factors (mainly temperature, water chemistry, and hydroperiod; Pechmann et al. 1989; Babbitt et al. 2003; Well- born et al. 1996). Foresters and land managers have the potential to most directly affect the latter, through impacts and manipulations to physical basin integrity and marginal forest vegetation. Tree harvesting decreases pond shading and increases water temperature, often accelerating larval amphibian development, of potential 3675_C013.fm Page 256 Friday, June 22, 2007 1:53 PM 256 Science and Conservation of Vernal Pools in Northeastern North America benefit for some species (e.g., spring peepers, Pseudacris crucifer) and detriment for others (e.g., marbled salamanders, Ambystoma opacum) (Skelly et al. 2002, 2005). The effects of forestry activities on pond hydroperiod are potentially signif- icant but also less predictable (see Vernal Pool Basin Relationships). Moving to a terrestrial environment places adult amphibians under a different set of constraints that are complicated by ectothermy, permeable skin, and small size. The moist, permeable skin of most adult, pool-breeding species serves as a partial respiratory organ (Stebbins and Cohen 1995), increasing their vulnerability to microhabitat drying. This skin also readily absorbs desiccants and toxins from the surrounding environment (Frisbie and Wyman 1991). Small size and linear proportions (in salamanders) contribute to a high surface area to volume ratio, further increasing the risk of adult, and especially juvenile, desiccation. In addition, amphib- ians are ectothermic in a region characterized by large intra-annual variation in temperature, including extended periods of subzero winter temperatures and hot, drying summer temperatures. Amphibians must thus respond to two pressing prob- lems: how to prevent or cope with the potential freezing of internal body fluids, and how to forage, migrate, and otherwise stay active on the forest floor during periods of high temperatures and low relative humidity. As we will explore further, many pool-breeding amphibians avoid the problem of freezing and desiccation by selecting shaded, forested habitats that contain deep, moisture-trapping litter and woody cover, often with abundant small mammal burrows that provide access below the frostline (Faccio 2003; Regosin et al. 2003). The challenge to foresters is to conserve the integrity of these and other important elements of forest structure during the course of timber management. PHILOPATRY AND MOVEMENT ECOLOGY Investigations of amphibian breeding-site fidelity suggest that most pool-breeding amphibians are highly philopatric (Sinsch 1990; Smith and Green 2005). Berven and Grudzien (1990) found that all adult wood frogs were faithful to their first breeding pond and that 82% of juveniles were faithful to their natal pool. Vasconcelos and Calhoun (2004) documented similar breeding pool fidelity by adult wood frogs (88% female return rates; 98% male), and spotted salamanders (Ambystoma macu- latum; 100% female and male). Additionally, nearly 100% breeding pond site fidelity has been documented for adult eastern newts (Notophthalmus viridescens) (Gill 1978). Site fidelity of this magnitude underscores the importance of avoiding seem- ingly small-scale disturbances to high value breeding pools that may have lasting impacts on localized amphibian populations. Successful upland migrations among breeding pools, summer foraging habitats, and overwintering locales are critical to meet the complex seasonal habitat require- ments of adult pool-breeding amphibians. Both adult migration and juvenile dis- persal through terrestrial ecosystems are complicated by desiccation risk for the reasons outlined above, and evidence is growing that several pool-breeding amphib- ians select for moist, shaded forest conditions during seasonal movements (deMaynadier and Hunter 1998, 1999; Rothermel and Semlitsch 2002; Patrick et al. 2006). To develop pool-specific habitat management guidelines we suggest that data 3675_C013.fm Page 257 Friday, June 22, 2007 1:53 PM Conserving Vernal Pool Amphibians in Managed Forests 257 on adult migration of pool-breeding amphibians is more informative than juvenile dispersal for two reasons. First, natal dispersal occurs over exceptional scales that often exceed forest management and ownership boundaries — e.g., over 1 km (0.6 mi) for Amystomatids (Funk and Dunlop 1999; Pechmann et al. 2001) and 2.5 km (1.6 mi) for wood frogs (Berven and Grudzien 1990) — complicating efforts to design management prescriptions around discrete elements (e.g., breeding pools). Secondly, relatively little is known about the habitat preferences of juvenile amphib- ians during their dispersal phase, and it is possible that an intensively managed forest matrix is more forgiving to short-term dispersal movements than to the more sed- entary home-range and migration movements of adults of the same species (Gibbs 1998; deMaynadier and Hunter 2000; Marsh et al. 2004). Although more work is needed, knowledge of the habitat preferences and adult migration distances of pool- breeding amphibians in the Northeast is growing (Chapter 7, Semlitsch and Skelly; Color Plate 17) and serves as a foundation for our spatial recommendations for timber management planning around vernal pools. VERNAL POOL–FORESTRY RELATIONSHIPS Given the complex life cycle of pool-breeding amphibians, it is important to consider potential effects that forest harvesting can have on the characteristics of both aquatic (breeding) and terrestrial (nonbreeding) habitat. To inform specific management recommendations for the conservation of both these habitats we organized our review around three elements that together define the local quality and extent of pool- breeding amphibian habitat in managed forest landscapes: the vernal pool basin and shoreline, overstory canopy cover in the surrounding forest, and the structure and condition of the forest floor. VERNAL POOL BASIN RELATIONSHIPS Land use activities located directly within the vernal pool basin, and its associated riparian nursery habitat, can have important effects on the breeding success and long-term population viability of pool-breeding amphibians. Specifically, forestry practices can affect pool basin habitat characteristics through changes to the physical integrity of the depression, pool hydrology, water quality, and riparian tree canopy cover and composition. Physical Integrity It is well established that disturbance by heavy machinery can have lasting impacts on forest ecosystems (Martin 1988; Turcotte et al. 1991). Harvesting operations in the pool itself, even during winter, can remove basin or riparian vegetation that helps provide egg attachment sites, shade, and organic material to the vernal pool’s detrital foodchain. Furthermore, the basin of many pools is extremely heterogeneous, offer- ing varied moisture and temperature conditions from the development of hummock topography, hardwood leaf litter wells, sphagnum moss, and accumulations of coarse woody debris. These moisture-trapping structures provide refuge to the eggs, larvae, 3675_C013.fm Page 258 Friday, June 22, 2007 1:53 PM 258 Science and Conservation of Vernal Pools in Northeastern North America metamorphs, and adults of various pool-breeding amphibians, reptiles and inverte- brates at different times of the year and yet are readily compromised by heavy machinery operating in the pool basin. Hydrology Precautions should be taken to avoid harvest activities that alter pond hydroperiod — a key driver of amphibian and invertebrate community composition (reviewed by Wellborn et al. 1996 and Semlitsch 2003). Management activities that contribute to shortened hydroperiods can lead to pond-wide desiccation and mortality, or force amphibian larvae to develop more quickly — generally causing smaller body size at metamorphosis, decreased survival, lower female fecundity, and delays in first reproduction (Howard 1978; Semlitsch et al. 1988; Berven 1990). Alternatively, an artificially extended hydroperiod increases pond habitat suitability for predatory fish and invertebrates (e.g., odonata, coleoptera), and potentially competition and preda- tion by other amphibians more closely associated with permanent waters (e.g., green frogs, Rana clamitans; bullfrogs, R. catesbiana). We briefly review three pool basin forestry activities with potential to affect hydroperiod, including soil disturbance (compaction and rutting), road construction, and local tree removal. Forest soils, particularly when wet, are vulnerable to rutting and compaction by heavy machinery used for harvesting and extraction (Nugent et al. 2003; Horn et al. 2004). Deep forest floor ruts that intersect pool basins can alter normal overland drainage patterns, artificially increasing or decreasing pool water-holding capacity depending on local topography (J. Houlahan, personal observation). Machinery- created ruts in the surrounding forest floor proximate to breeding pools can affect the success of adult breeding migrations by potentially impeding or redirecting salamander movements (Means et al. 1996) or attracting frogs and salamanders to lay eggs in artificial rut pools that often dry prematurely and constitute ecological traps (DiMauro and Hunter 2002; P. deMaynadier, personal observation). Roads are often an unavoidable byproduct of timber harvesting and have impor- tant ecological impacts on streams, lakes, and smaller wetlands. In fact, it has been suggested that harvest roads have potentially larger impacts on hydrology than tree removal (Lockaby et al. 1997; Cornish 2001), with ditches, culverts, and road surfaces able to change the direction and speed of overland sheetflow. Recently, Gomi et al. (2006) reported that forest roads contribute significantly to detritus and sediment accumulation in streams. The effects of roads on vernal pool hydrology is not well-studied, but it seems reasonable to suggest that, given their generally small size and shallow depth, vernal pools may be especially sensitive to sheetflow alter- ation and sedimentation. Indeed, preliminary results suggest a negative correlation between forest road density and amphibian species richness in small New Brunswick ponds (Jacobs and Houlahan, in preparation). Finally, there is growing evidence that forest tree removal can influence the water table dynamics and hydroperiod of local wetlands. Generally, forest harvesting in forested bottomlands results in elevated water tables in the first few years after clearcutting due to decreased evapotranspiration (Sun et al. 2001; Pothier et al. 2003). More recently evidence suggests that although deforestation tends to increase water 3675_C013.fm Page 259 Friday, June 22, 2007 1:53 PM Conserving Vernal Pool Amphibians in Managed Forests 259 table elevation immediately after harvesting, the long-term effects are subtle and temporally variable, with higher water tables during the nongrowing season and lower water tables during the growing season (Bliss and Comerford 2002). Martin et al. (2000) also showed increases in water yield immediately after strip-cutting that lasted for three to six years, but these increases were followed by more than 20 years of below-average water yield due to the regeneration of rapid-growing pioneer species (e.g., Prunus spp., Populus spp.) with higher transpiration rates. Vernal pools are particularly vulnerable to changes in water table elevation since water loss (and presumably water gain) is a function of wetland perimeter to area ratio, a measure that is relatively high in small wetlands (Millar 1971). Water Quality One of the common effects of forest harvesting in or near the pool basin is diminished canopy cover over the breeding pool. Recent evidence suggests that amphibian species richness, growth, and development are lower in heavily shaded pools (Skelly et al. 2002, 2005), likely due to lower water temperatures and lower food quality (Halverson et al. 2003; Skelly et al. 2005). However, the conservation implications are complicated by the fact that several amphibian pool-breeding specialists (e.g., marbled salamanders; Chapter 7, Semlitsch and Skelly) and invertebrate specialists (e.g., fairy shrimp, Eubranchipus spp.; Ossman and Hanson 2002) thrive, or are found more often, in shaded pool locales. Perhaps less intuitive, but no less real, are the impacts to vernal pool water quality that result from watershed-scale forest management practices. Since the early experiments at Hubbard Brook (Bormann et al. 1968), it has become axiomatic that forest harvesting increases sedimentation and nutrient transport (Lamontagne et al. 2000) and lowers surface water quality in clear-cut watersheds (Martin et al. 2000). The evidence indicating that increased nutrient inputs are significant for lakes sug- gests that they are almost certainly significant for small, shallow vernal pools as well. Additionally, intensive forest management often includes conversion of mixed stands to predominately softwood plantations, and these changes in tree community composition can impact water quality. For example, Ito et al. (2005) found that watersheds dominated by coniferous trees exported more dissolved organic carbon (DOC) than those dominated by deciduous trees and that DOC levels were higher in lakes in conifer-dominated watersheds. Conversely, nitrate export to lakes is higher in deciduous forest-dominated watersheds (Ito et al. 2005). This suggests that inten- sive site conversion to softwood-dominated stands may result in higher DOC con- centrations and lower nutrient levels in vernal pools. Interestingly, Waldick et al. (1999) found that amphibian species richness is lower and community composition different in ponds surrounded by conifer plantations than ponds surrounded by natural mixedwood forests. Chemicals are widely used in forestry in the Northeast including insecticides such as carbaryl, tebufenozide, and Bacillus thuringiensis, and herbicides like glyphosate and triclopyr (Chapter 11, Boone and Pauli). In most cases, these pesti- cides are applied aerially, making vernal ponds particularly vulnerable to contami- nation because of limited regulatory restrictions on aerial spraying around temporary 3675_C013.fm Page 260 Friday, June 22, 2007 1:53 PM 260 Science and Conservation of Vernal Pools in Northeastern North America water bodies and the difficulty of delineating buffer zones around small, widely distributed pools. Carbaryl is especially persistent under acidic conditions and has been demonstrated to have indirect effects on salamander populations. Tebufenozide, which has recently replaced fenitrothion as the insecticide of choice for controlling spruce budworm (Choristoneura fumiferana), has not been shown to have large effects on organisms in aquatic systems (Pauli et al. 1999) although there is evidence that it depresses cladoceran (often called daphnia) species richness (Kreutzweiser et al. 2004), a valuable prey source for pool-breeding fauna. Glyphosate is one of the most widely applied herbicides in forestry (Woodburn 2000), and it has, until recently, been considered relatively benign because it acts on a metabolic pathway that is only found in plants. However, commercial formulations include surfactants that prevent glyphosate from “beading up” and rolling off plant leaves, and these may be more toxic than glyphosate itself (Giesy et al. 2000). Generally, the use of surfactants should be avoided in proximity to high-value vernal pools as they facil- itate absorption through the moist, permeable skin of amphibians and their ecological effects remain largely unknown. FOREST CANOPY RELATIONSHIPS Clearcutting One of the most defining structural elements in any forest is its canopy, and the response of amphibian communities to canopy presence or absence has been rela- tively well-studied. Most North American studies examining the effects of clearcut- ting, for example, report significantly lower overall abundance (deMaynadier and Hunter 1995). More interestingly for our purposes, some amphibian groups are more sensitive to intensive canopy removal than others. A detailed review of only those studies including data for pool-breeding species native to the Northeast reveals a striking pattern (Figure 13.1). Specifically, the ratio of median abundance for each of four northeastern seasonal pool-breeding specialists was several times (3.7–5.5) greater in control stands than in clearcut stands, and exceeded the same ratio for North American amphibian taxa generally. Mole salamanders as a group (Ambystoma spp.), and spotted salamanders in particular, appear to be especially sensitive to large-scale canopy removal (Figure 13.1). The ratio for spotted salamanders is notable in that it exceeds that previously reviewed for Plethodontidae (5.0 times; deMaynadier and Hunter 1995), a family of terrestrial, lungless salamanders gener- ally considered among the most sensitive of amphibian taxa to forest management and other practices that alter forest floor microclimate (Welsh and Droege 2001). An in-depth study of forest clearcutting and edge effects on a community of 14 species of amphibians in Maine yields further support for the premise that pool- breeding amphibians rank among the most sensitive of northeastern amphibian taxa to intensive harvest practices (deMaynadier and Hunter 1998). Specifically, in con- structing a “management sensitivity index” composed of the ratio of abundance of relative forest interior to clearcut captures, three of the four species identified as most sensitive to the effects of complete canopy removal — redback salamanders (Plethodon cinereus), wood frogs, spotted salamanders, and blue-spotted salamanders (Ambystoma 3675_C013.fm Page 261 Friday, June 22, 2007 1:53 PM Conserving Vernal Pool Amphibians in Managed Forests 261 6.00 o 5.00 ati R e 4.00 c n a nd 3.00 u b A n 2.00 a di e M 1.00 0.00 S. holbrooki A. opacum R. sylvatica Ambystoma A. maculatum (n=3) (n=3) (n=5) (n=7) (n=5) FIGURE 13.1 The median ratio of abundance from mature vs. clearcut forest for several pool- breeding amphibian taxa characteristic of the Northeast. The horizontal line indicates the same ratio for amphibian taxa generally across North America (n = 18 studies; from deMaynadier and Hunter 1995). Ratios were calculated from all datasets that permitted estimates of the relative magnitude difference in captures for mature (control) versus clearcut (treatment) stands. Inde- pendent estimates were calculated from mean captures, absolute totals, or frequency data, depending upon the reference consulted. (Contributing sources include: Bennett et al. 1980; deMaynadier and Hunter 1998; Enge et al. 1986; Grant et al. 1994; Mitchell et al. 1997; Patrick et al. 2006; Ross et al. 2000; Waldick et al. 1999. With permission.) laterale) — were northeastern pool-breeding specialists. Furthermore, the effects of intensive canopy removal extended beyond the boundaries of harvested stands with edge effects reducing pool-breeding amphibian abundance levels at distances of 25–35 m (82–115 ft) into adjacent unmanaged forests. Partial Harvesting and the Canopy Continuum In contrast to clearcutting, the effects of partial harvesting and uneven-aged forestry practices are less understood, despite their greater frequency of use in most north- eastern forests (Seymour 1995). Because the shade cast by standing trees in a partially harvested stand can have beneficial effects on forest floor microhabitats (e.g., shaded logs are significantly cooler and moister than unshaded logs; Heatwole 1962), it is likely that partial cutting practices have relatively less impact on sedentary species that spend most of their life on the forest floor. However, the question remains: how much canopy can be removed during partial harvesting before terres- trial pool-breeding populations respond with significant declines? Although there is a growing body of knowledge on the effects of partial canopy disturbance on amphibians generally (Pough et al. 1987; Mitchell et al. 1996; Messere and Ducey 1998; Sattler and Reichenbach 1998; Brooks 1999; Grialou et al. 2000; Moore et al. 2002; Knapp et al. 2003; Renken et al. 2004; Karraker and Welsh 2006), specific results for pool-breeding amphibians of the Northeast are limited (Ross et al. 2000; Patrick et al. 2006). 3675_C013.fm Page 262 Friday, June 22, 2007 1:53 PM 262 Science and Conservation of Vernal Pools in Northeastern North America In our view, the most illuminating investigation to date on the effects of partial harvesting on vernal pool amphibians comes from northern Pennsylvania (Ross et al. 2000) where 47 managed hardwood forest stands were selected to represent a continuum of canopy closure from a nearly intact overstory (unharvested for over 70 years), to a wide range of partially intact canopy stands (selection and diameter- limit harvests), to complete absence of any residual overstory (recent clearcuts). In this manner, the investigators were uniquely prepared to document potential thresh- olds in population responses to increasing levels of canopy removal associated with varying intensities of forest management. We reanalyzed the raw data from Ross et al. (2000) and found just such patterns (Figure 13.2A,B). When examining sala- manders alone (12 spp. total), a clear trend emerges of increasing abundance with increasing basal area and canopy cover (Figure 13.2A), with a potential threshold of canopy cover at ~45–50%, beyond which salamander abundance levels increase dramatically. More interestingly for our purposes, a similar pattern and threshold (at ~50–55% canopy cover) persists when only a restricted sample of pool-breeding salamander specialists is examined (Ambystoma jeffersonianum, A. opacum, A. mac- ulatum, Hemidactylium scutatum; Figure 13.2B). More fieldwork is needed to attempt replication of these results for other pool-breeding taxa and other forest types, but it appears that several highly characteristic pool-breeding amphibians of the Northeast are sensitive to harvesting practices that remove overstory canopy levels below a level of approximately 50%. Other studies examining the effects of partial harvesting have yielded results consistent with the patterns above, with few if any significant impacts documented in the abundance of resident salamander populations following light intensity canopy 600 d) 500 n a st 400 o./ n s ( 300 r e nd 200 a m a 100 al S 0 0 20 40 60 80 100 % Overstory Cover FIGURE 13.2A Relationship between salamander abundance (12 spp) and overstory canopy closure for 47 forest stands in northeastern Pennsylvania (Pearson’s coefficient of correlation = 0.63; P<0.05). The arrow indicates a potential canopy threshold of ~45–50%, below which salamander abundance remains consistently low. (Data reanalyzed from Ross et al. 2000. With permission.)