Management of Amphibians, Reptiles, and Small Mammals in North America

Home , Abies concolor, Acer glabrum, Picea engelmannii

Abstract.-We measured habitat components for Ha bitat Requirementsof New two state-listed endangered salamanders in New Mexico's Endangered Mexico in 1986 and 1987. ~0thspecies are restricted to mesic environments within high-elevation, mixed Salamanders1 coniferous forests. Steep slope and high elevation were the most useful variables for predicting the occurrence of Jemez Mountains salamanders and Cynthia A. Ramotnik2and Norman J. Sacramento Mountain salamanders, respectively. Scott, Jr.3 Although the discriminant models show some predictive value in detecting salamanders based on habitat variables, we believe that the best survey technique is ground-truth surveys in wet weather. A better fit of the discriminant models might be obtained by including variables not measured e.g., fire and logging history, and soil characteristics. We offer interim management guidelines as a result of our analysis.

Two of the three species of salaman- the U.S. Forest Service (USFS) and ders that occur in New Mexico are changes in timber practices have restricted to coniferous forests at prompted concern about the effect of high elevations. The Jemez Moun- logging on these salamanders (Scott tains salamander (Plethodon neomexi- et al. 1987, U.S. Fish & Wildlife Serv- canus) (fig. 1) is known only from ice 1986). Most of the range of each north-central New Mexico at the species occurs on National Forest southern terminus of the Rocky (NF) lands, and the close association Mountains (Reagan 1972). The Sacra- of these salamanders with mixed co- mento Mountain salamander (Aneides niferous forests may make them vul- hardii) (fig. 2) occurs in the Capitan nerable to some forest-management and Sacramento Mountains in south- practices. In 1985, both species were central New Mexico (Williams 1976). placed under review as potentially Figure 2.-Sacramento Mountain These lungless salamanders, with threatened or endangered species salamander (Aneides hardii). Photo by Stephen Corn. small body sizes and terrestrial juve- under the Federal Endangered Spe- nile development, are restricted to cies Act (Ramotnik 1986, Staub 1986). In 1986, the U.S. Fish & Wildlife mesic environments. Lowe (1950) As a result, an interagency commit- Service (USFWS) contracted with the suggested that both species are rel- tee was established to identify data USFS to study these species on NF icts of the mid-Tertiary Rocky Moun- and management needs and develop lands. The primary objectives were tain fauna. strategies to address these needs. to survey for salamanders in plan- In 1975, both species were listed ning units under consideration for by the state of New Mexico as endan- future logging operations and to gered due to their restricted distribu- characterize salamander habitats us- tion (Hubbard et al. 1979).Since ing habitat components that are 1980, increases in timber harvest by meaningful and useful to USFS biolo- gists and land managers. This infor- 'Paper presented at symposium, Man- agement of Amphibians, Reptiles, and mation would be used to assess po- Srn all Mammals in North America. [Flag- tential salamander habitat from maps staff, AZ,July 7 9-2 7 7 988). or aerial photos, thereby reducing 2MuseumSpecialist, U.S. Fish & Wildlife the need to inventory areas by Service, National Ecology Research Center, ground-truth assessment. 7300 Blue Spruce Drive, Fort Collins, CO 80524. In this paper, we characterize habitats of Jemez Mountains sala- 3Zoologist,US. Fish & Wildlife Service, manders and Sacramento Mountain National Ecology Research Center, Mu- Figure I .-Jemez Mountain salamander seum of southwestern Biology, Universify of (Plefhodon neomexicanus). Photo by salamanders based on general site New Mexico, Albuquerque, NM 87 13 1. Stephen Corn. characteristics and surf ace cover items that could serve as refugia for subalpine forest association, Engel- and Raphael and Rosenberg (1983). salamanders. We use a multivariate mann spruce, Douglas-fir, and white The areas of four classes of cover analysis of habitat characteristics that fir are the most common trees. Aspen items (rock, bark, fine woody debris, describes areas with and without and Rocky Mountain maple are and coarse woody debris) were esti- salamanders, and present manage- found to a lesser extent. Aspen mated visually. We further divided ment guidelines as a result of this groves, talus fields, and open m.ead- coarse woody debris (CWD) into analysis. ows are present at higher elevations. three decay classes, adapted from a Annual precipitation in the Jemez five-class scheme for rating decom- Mountains ranges from 400-550 mm position of Douglas-fir logs (Franklin Study Areas (Castetter 1956) and is slightly higher et al. 1981). To emphasize differences in the Sacramento Mountains. Much between decay classes, we combined We studied the Jemez Mountains of the precipitation falls between July classes 1 and 2 (CWDl), and classes 3 salamander within the Santa Fe NF and September (Kunkel1984). and 4 (CWD3), and placed the most in the Jemez Mountains (Los Alamos, decayed logs, class 5, in a third cate- Rio Arriba, and Sandoval Counties, gory (CWD5). New Mexico), which are located ap- Methods Aspect was taken with a magnetic proximately 100 km north of Al- compass at 10,30, and 50 m. Com- buquerque (fig. 3). The Jemez Moun- We conducted fieldwork in the sum- pass readings were assigned to one tains are volcanic in origin and are mers of 1986 and 1987 (Jemez Moun- of four aspect classes where 316-45" = underlain by volcanic rock, ash, and tains: 28 July-14 August 1986,29 north-facing; 46-135" = east-facing; pumice. The predominant feature in June-11 July 1987,24 August5 Sep- 136-225" = south-facing; and 226-315" the area is the volcanic caldera, the tember 1987; Sacramento Mountains: = west-facing. Percent slope was de- Valle Grande, around which the 22 August-10 September 1986,s-20 termined with a clinometer, and per- mountains lie. Fieldwork on the Sac- June 1987; 20 July-1 August 1987). cent canopy cover was estimated ramento Mountain salamander was These dates included the surface ac- with a spherical densiome ter conducted in the Sacramento Moun- tivity periods of Jemez Mountains (Lemmon 1956). Both measurements tains, within the Lincoln NF, Otero salamanders (Reagan 1972) and Sac- were recorded at 10-m intervals. All County, New Mexico fig. 3). Volcanic ramento Mountain salamanders readings were made along the intrusions occur within the Paleozoic (Williams 1976). transect and averaged for the strata of the Sacramento Mountains. Transects were established in for- Elevations in the Jemez Mountains ested areas; most were located in range from 2130-3410 m, and from planning units selected by USFS per- 2290-3600 m in the Sacramento sonnel. Within these areas, locations Mountains. of transects were selected from topo- Habitat types within these eleva- graphic maps to sample a variety of tional ranges occur within the Rocky topographic aspects. South-facing Mountain upper montane (2290-2900 slopes were not searched in the m) and subalpine (2900-3660 m) for- Jemez Mountains due to the diffi- est association (Castetter 1956). The culty in locating salamanders on upper montane forest association these slopes (Ramotnik 1988). To en- (Shelford 1963) is characterized by sure having sites occupied by sala- mixed coniferous forests dominated manders, we visited known localities by white fir (Abies concolor), Douglas- or areas where salamanders had re- fir (Pseudotsuga menziesii), Engelmann cently been found. A small number spruce (Picea engelmannii), and blue of sites outside planning units were spruce (Picea pungens). Deciduous chosen from topographic maps. components include quaking aspen We established 100-m2transects (2 (Populus frmuloides), Rocky Moun- m x 50 m) oriented uphill from near tain maple (Acer glabrum), oak (Quer- the bottoms of slopes. Our transect is cus spp.), New Mexico locust (Robinia modified from area-constrained neomexicana), and oceanspray (Holo- searches, a technique developed by discus dumosus). Ponderosa pine others, e.g., Bury (1983), Bury and (Pinus ponderosa) stands predominate Corn (this volume), Bury and Ra- Figure 3.-Distribution of Jemez Mountains salamanders (Plethodon neomexicanus) at the lower elevations, particularly phael (1983), Campbell and Christ- and Sacramento Mountain salamanders on south-facing slopes. Within the man (1982), Raphael (this volume), (Aneides hardii) in New Mexico. transect. Numbers sf white fir and mot transformed because values were timal separation between groups us- Douglas-fir were pooled in a single distributed normally. ing linear transformations of the in- class ('TFIR), as were Engelmann and A stepwise variable entry proce- dependent variables based on vari- blue spruce (TSPRUCE),and Pinus dure (STEPDISC) selected the "best ables selected by the stepwise pruce- sgp. (TPINEj. Numbers of trees set" of habitat variables to discrimi- dure. The Mahalanobis distance be- within tree classes were counted in a nate between groups and reduced tween group means was tested using 20-rn x 50-m plot centered over the the complexity of the original vari- an F-statistic. transect. Twenty-three measured and able set. &cause the models selected Predictive discriminant analysis derived variables were used in the by STEPDISC are not necessarily the (PDA) (Williams 1983) (DISCRIM) analyses (table 1). best possible models (§AS Institute was used to test the discriminatory We determined numbers of sala- 9nc 1982), cross-validation was ac- power of the variables selected by manders on transects by searching all complished by using canonical analy- DDA. We used chi-square analysis to cover items manually or with potato sis (CANDISC) or descriptive dis- compare cover i tern use (of the four rakes. The locations of salamanders criminant analysis (DDA) (Williams classes) to availability and to com- in other than the four classes of cover 1983). DDA attempts to establish op- pare aspects of transects with and items also were recorded. When a salamander was found, we recorded r snou t-vent length (distance from tip Table 1 .-Description of measured and derived habitat variables used in of snout to anterior edge of vent), habitat selection analysis of two species of New Mexico salamanders. sex, and dimensions and type of cover i tem. For coniferous logs, we also recorded salamander position relative to the log (in, under, or under bark) and decay class (modified from Corn and Bury, in press, Ra- phael and Rosenberg 1983). These data were usd to calculate densities of salamanders on transects and to determine cover item use by salamanders. We acquired additional data on cover item use by salamanders by locating salamanders in areas on both sides of the transects.

Statistical Analysis (sticks) (mZ) Data for transects with and without ROCK Estimate of amount of surface rock (m2) salamanders were pooled separately. SLOPE Average percent slope measured with a cli- We calculated descriptive statistics nometer (mean, standard error, range) for habitat variables in the two groups 50-m x 20-m plot and used a one-way analysis of vari- SFlR Number of small fir (~20cm dbh) ance to compare transformed vari- MFlR Number of medium fir (20-50 cm dbh) ables between groups. Size classes of LFl R Number of large fir (>50 cm dbh) fir and spruce were compared be- TFlR SFlR + MFlR + LFlR tween the two groups with a t-test. SSPRUCE Number of small spruce (40 cm dbh) The following transformations MSPRUCE Number of medium spruce (20-50 crn dbh) were applied to stabilize the variance LSPRUCE Number of large spruce (>50 cm dbh) TSPRUCE SSPRUCE + MSPRUCE + LSPRUCE of the habitat variables (Snedecor TASPEN Number of aspen (all sizes) and Cochran 1967) and to increase TNOD Number of non-oak deciduous (all sizes) the probability of a normal distribu- TOAK Number of oak (all sizes) tion: arcsine (SLOPE CANOPY); TPlNE Number of pine (all sizes) square root + 0.5 (tree densities); and TSNAGS Number of snags (all sizes) log + 0.5 (cover items). Elevation was without salamanders. The Statistical also did not differ significantly be- many pine and large spruce trees Analysis System computer package tween the two groups of transects (X2 containing salamanders, to shallow (SAS, Version 5) was used for all = 0.28, df = 2, P > 0.90). The amount slopes with few pine or large spruce analyses (SAS Institute Inc 1982). Sig- of CWDl was similar between trees without salamanders. SLOPE nificance levels were set at P < 0.05 groups but amounts of CWD3 and had the highest discriminating power unless otherwise indicated. CWD5 were higher on transects with (r2= 0.73). PDA correctly classified salamanders. Although no south-fac- 91 % of the 33 transects without sala- ing slopes were searched, propor- manders and 80% of the 10 transects Results tions of other aspects occupied by with salamanders. salamanders were not different from The 10 transects and additional Jemez Mountains Salamander the proportions of total aspects searches produced 148 Jemez Moun- searched (X2 = 1.3, df = 2, P > 0.50). tains salamanders; the type of cover Salamanders (N = 28) were present Three of the original 20 variables item was known for all but one sala- on 10 of 43 transects (23%)with a were selected by the stepwise vari- mander. Ninety-six percent (141 / mean density of 3/ 100 m2in occu- able entry procedure for inclusion in 147) of salamanders were distributed pied areas. One hundred twenty the descriptive discriminant model: among the four major cover classes salamanders were found in areas off SLOPE, TPINE, and LSPRUCE (table as follows: CWD, 100 (68%);ROCK, the transects. Transects with sala- 3). Subsequent analysis by DDA re- 40 (27%);FWD, 1 (1%).No salaman- manders occurred on significantly tained these variables. The resultant ders were found under BARK. Three steeper slopes and at lower eleva- discriminant function explained 38% salamanders (2%) were found on tions than transects without salaman- of the between-group variance; how- transects under surface litter and ders (table 2). Analysis of size classes ever, it did not have significant three salamanders (2%)were found of fir and spruce showed no signifi- power in discriminating between under aspen logs. The frequency of cant differences between transects groups (F = 2.34, P = 0.09). This func- salamanders associated with CWD with and without salamanders. Pro- tion describes a multivariate gradient by decay class was CWDI--4%; portions of decay classes of CWD that ranges from steep slopes with CWD3--66%; CWD5--30%. Of 28 salamanders found on transects, 24 salamanders were associated with one of the four classes of cover items. Because of the small sample size, we were unable to determine a correla- tion between cover item availability and use.

Sacramento M~untain Salamander

Salamanders (N = 233) were present on 26 of 80 transects (33%)with a mean density of 6/100 m2 in occupied areas. We located 387 sa'laman- ders in areas off the transects. Transects with and without salamanders differed in several respects: transects with salamanders occurred at significantly higher elevations, on shallower slopes, and had higher numbers of spruce and lower numbers of pine than transects without salamanders (table 4). Analysis of size classes of fir and spruce revealed that densities of large fir and all size classes of spruce were significantly higher on transects with salamanders (LFIR: t = 3.38, P = 0.001; SSPRUCE: t The frequency of salamanders associ- without salamanders, it did identify = 2.85, P = 0.008; MSPRUCE: t = 2.56, ated with CWD in the three decay steep slopes as the most useful vari- P = 0.016; LSPRUCE: t = 3.04, P = classes was CWD1-13%; CWD3- able in determining the occurrence of 0.003) (fig. 4). Although the total 62%;CWD5-25%. Of 233 salaman- Jemez Mountains salamanders. It is amount of CWD on transects with ders found on transects, 209 sala- possible that steep slopes contain and without salamanders was not manders were associated with one of more interstitial spaces in the soil significantly different, there was sig- the four classes of cover items. Ex- than do shallower slopes. The soils of nificantly more CWD5 on transects amination of cover item availability steep slopes may be less compacted with salamanders (X2 = 6.93, df = 2, P and use for these salamanders re- than those of more gentle slopes due > 0.05). The proportions of transects vealed that salamanders are associ- to the combined effects of gravity, by aspect did not differ between the ated with some cover items dispro- and movement of water and soil. As two groups (X2 = 3.83, df = 3, P > portionate to their availability (X2 = a consequence of steep slope and the 0.10). 59.9, df = 3, P < 0.001). In particular, presence of underlying volcanic rock Because numbers of the three size Aneides was found in association characteristic of the Jemez Mountains classes of spruce were significantly with FWD proportionately less fre- (Burton 19821, spaces within this ma- higher on transects with salaman- quent than expected, and used well- ders, we substituted TSPRUCE for decayed and moderately decayed SSPRUCE, MSPRUCE, and logs to a greater extent than expected LSPRUCE in subsequent analyses. A (X2 = 62.1, df = 2, P < 0.001). stcpwise variable entry procedure selected eight of the original 20 variables for inclusion in the descriptive Discussion discriminant model (table 5). Subse- quent DDA kept all but three Jemez Mountains Salamander (SLOPE, CWD1, and TAPSEN) in the model. The resultant discriminant While canonical analysis did not dis- function explained 49% of the be- criminate between transects with and tween-group variance and had significant power in discriminating between groups (F = 6.87, P < 0.0001). This function can be interpreted ecol- ogically to describe a gradient that ranges from low elevations with many pine, few spruce and large fir, and infrequent CWD5 without salamanders, to higher elevations, few pine, many spruce and large fir, and abundant CWD5 that contain salamanders. ELEV had the highest discriminating power (r2= 0.64). PDA correctly classified 96% of the 54 transects without salamanders and 58% of the 26 transects with salamanders. The 26 occupied transects and additional searches produced 620 Sac- ramento Mountain salamanders. Ninety-five percent (589) were distributed among the four major cover classes as follows: CWD, 377 (64%); ROCK, 127 (22%);BARK, 58 (10%); and FWD, 27 (4%). Fourteen salamanders (2%)were found under aspen logs and 17 salamanders (3%) were above or below surface litter. SPRUCE

-1 nb 11C.I "1 n LI1 V II "1 %I1 1L1 LC "1 &I IL IXVLRY Mountains. The largest concentra- tions of P. neomexicanus have been / WlTH SALAMANDERS I found in association with talus slopes 1 WITHOUT SALAMANDERS (Whitford and Ludwig 1975, Clyde Jones pers. comm.), which are also important to many other western Ple- thodon (Brodie 1970). Other plethodontids are virtually restricted to areas with a loose rocky soil (Aubry et al. 1987, French and Mount 1978, Herrington and Larsen 1985, Jaeger 1971). The variables selected by canonical analysis showed some predictive value. A1 though three transects without salamanders were rnisclassified by PDA as transects with salamanders, Plethodon was found in areas adjacent to the transects. The two transects misclassified as transects without salamanders had FIR values for TPINE and LSPRUCE closer to values usually associated with transects without salamanders. Because a larger percentage of transects without salamaders were correctly classified by PDA, these three variables may better describe the conditions under which salamanders are absent from an area, rather than describing favorable conditions under which they would occur. The limited discriminatory and predictive power of the variables se-

f- Table 5.-Correlations of habitat variables with discriminant scores for transects with and without Sac- ramento Mountain salamanders.

Mnemonic DFl

(20 CM 20-60 CM 60 CM ELEV 0.55 TSPRUCE 0.42 TPI NE -0.47 D.B.H. CWD5 0.44 LFlR 0.34 CWDl -0.05 Figure 4.-Comparisons of average size classes (d.b.h.) of spruce and fir on transects with and without Sacramento Mountain salamanders. Boxes indicate 95% confidence intervals SLOPE -0.06 for the mean. Levels of significance indicated by asterisks are 0.05 (+)and 0.005 (*+). TASPEN -0,02 \ / lected by multivariate techniques the distributions of up to 10 amphibi- Aneides is often present where the may reflect our inability to reliably ans in southeastern New York were best habitat predictors indicate they and consistently detect the presence significantly influenced by soil pH should not occur. While high-eleva- of Plethodon at a site. We believe that and moisture (Wyman 1988). tion, wet, north-facing slopes with a our ability to detect salamanders is Salamanders also may be absent mature mixed-conifer forest do har- fairly good and repeatable, but we from a given site for reasons other bor Aneides, salamanders are also realize that environmental factors than unsuitability of habitat. For ex- found less predictably in areas that can influence the relative numbers of ample, access to a particular area by may be drier and more exposed than salamanders. During repeated visits salamanders may be impossible due the model would indicate. With the to the same sites, Plefhodon was more to the unsuitability of the area that exception of elevation, the ranges of abundant when we searched under surrounds it, e.g., dry, open field. Or, habitat variables on transects occu- wet conditions, and other studies a climatic event may have eliminated pied by salamanders are not strik- have reported a significant correla- salamanders from a given area with- ingly different from those on plots tion between movement and activity out sufficient time occurring for them without salamanders (table 4). This of salamanders, and precipitation to recolonize the site. overlap may be due to factors not (Barbour et al. 1969, Kleeberger and measured, e.g., fire and logging his- Werner 1982, MacCullough and tory, and it may show an ability of Bider 1975). Low densities and Sacramento Mountain salamanders to persist after habitats patchiness of P. neomexicanus popula- Salamander have been altered. tions also can hinder detection of the animal. In comparison with densities The variables selected by canonical of red-backed salamanders, P. cin- analysis were able to discriminate be- Management Guidelines ereus, (0.9-2.2 individuals/m2; Heat- tween transects with and without wole 1962, Jaeger 1980), our density salamanders. However, these vari- Our data show that, despite some estimates for Jemez Mountains sala- ables had limited predictive value. predictive power of the habitat vari- manders are extremely low (0.03 in- Although a larger percentage of ables, the level of uncertainty in pre- dividuals/m2).Although Williams transects without salamanders were dicting salamander occurrence may (1972) reported estimates of Jemez correctly classified by PDA, there is preclude their use by the USFS. At Mountains salamanders ten times still a one-in-five chance of being this time, we feel the best survey greater than ours, he noted that their wrong in predicting that salaman- technique for salamanders is ground- distribution was spotty. ders are absent from a site. For most truth surveys in wet weather during A better fit to a discriminant management decisions, this level of the activity season of each species. model might be obtained by includ- uncertainty will not be acceptable, Under proper conditions, both spe- ing variables that we did not meas- and ground-truth searches will have cies are easy to find and relatively ure, e.g., fire and logging history and to be made. unskilled persons can be quickly soil characteristics (moisture, pH, High elevation was the best pre- trained to survey habitats. Our im- and compaction). Williams (1976) dictor of the presence of Sacramento pression was that Plefhodon was suggested that logging may have Mountain salamanders (table 5). more difficult to survey, because it eliminated Jemez Mountains sala- Weigmann et al. (1980) also found tended to retreat underground dur- manders from part of Peralta Canyon significantly more Sacramento ing dry periods. Aneides, however, due to dry conditions resulting from Mountain salamanders on transects can usually be found even during ex- removal of most of the canopy. How- at higher elevations. The higher ele- tended dry periods. ever, there was no documentation vations of the Sacramento Mountains Our attempts to explain the ab- that salamanders occurred at the site experience greater rainfall, cooler sence of salamandersfrom a given prior to logging. Soil characteristics, temperatures, and lower area, i.e., potential difficulty of de- which can be affected by fire and log- evapotranspiration rates than the tecting all salamanders present, and ging practices (Childs and Flint 1987, lower elevations and therefore may low density or patchy distribution of DeByle 1981, Krag et al. 19861, also be more hospitable to plethodontid populations, may overlook the possi- can influence the distribution of ple- salamanders. The low critical ther- bility that absence is not solely due to thodontid salamanders, that occupy mal maximum of Aneides probably unsuitable habitat. Absence does not the soil-litter interface. Plethodon cin- reflects adaptations to the low tem- necessarily mean avoidance, but may ereus was excluded from 27% of for- peratures characteristic of their mi- be due to insufficient time for the est habitat in eastern deciduous for- crohabi tat (Whitford 1968) and may animal to recolonize an area, or inac- ests because of low soil pH (Wyrnan restrict salamanders to high eleva- cessibility of a suitable area due to and Hawksley-Lescault 1987), while tions. unsuitable habitat surrounding it. In lieu of specific recommenda- Other studies provide some evi- Discovely of Larch Mountain sala- tions, the USFS needs interim man- dence for negative effects of logging manders Plethodon Zarselli in the agement guidelines to protect the on amphibian populations (Bennet et central Cascade Range of Wash- salamanders from population de- al. 1980, Blymer and McGinnes 1977, ington. Biological conservation clines. We suggest the following Bury 1983, Gordon e t al. 1962, Her- 42:147-152. steps: rington and Larsen 1985, Pough et al. Earbour, Roger W., James W. Har- 1987, Ramotnik 1988, Staub 1986, and din, James P. Schafer, and Michael Salamander surveys should Williams 1976) and we suspect that J. Harvey. 1969. Home range, be made on specific sale ar- intensive loggng, slash removal, and movements, and activity of the eas as early in the planning burning will reduce or eliminate dusky salamander, Desmognathus process as possible. The populations of Plethodon neomexica- fuscus. Copeia l969:293-297. USFS could maintain a team nus and Aneides hardii. Only intensive Bennet, Stephen H., J. Whitfield Gib- of seasonal employees for observations of salamander popula- bons, and Jill Glanville. 1980. Ter- such surveys and for other tions throughout the logging cycle restrial activity, abundance and activities related to endan- will provide the information needed diversity of amphibians in differ- gered species. to make management recommenda- ently managed forest types. tions. These studies are in progress, American Midland Naturalist To the extent possible, inten- but may require years before defini- 103:412-416. sive logging operations (i.e., tive results are available to assess the Blymer, Michael J. and Burd S. clearcuts, seed-tree cuts, trac- effects of logging on Plethodon and McGinnes. 1977. Observations on tor logging) should not be Aneides. possible detrimental effects of conducted in areas occupied clearcutting on terrestrial amphibi- by salamanders. Cable log- ans. Bulletin of the Maryland Her- ging in winter, when the Acknowledgments petological Society 13:79-83. ground is frozen and the Brodie, Edmund D., Jr. 1970. West- salamanders are under- We thank the following U.S. Forest ern salamanders of the genus Ple- ground, is probably the least Service personnel: Santa Fe National thodon: systematics and geo- damaging activity. In com- Forest-R. Alvarado, D. Delorenzo, graphic variation. Herpetologica parison, tractor logging on and M. Morrison; Lincoln National 24:468-516. wet soils can compact the Forest-R. Dancker, D. Edwards, S. Burton, Barry W. 1982. Geologic evo- soil to such a degree that Lucas, J. Peterson, and D. Zaborske; lution of the Jemez Mountains and salamanders cannot use it. and L. Fisher, Regional Office. Much their potential for future volcanic of the funding was provided by the activity. Los Alamos National Modifications of current US. Forest Service (Southwestern Laboratory, LA-8795-GEOL. Los practices, such as leaving Region). Alamos, New Mexico. 31 p. slash where it falls or leaving Field personnel included M. J. Al- Bury, R. Bruce. 1983. Differences in as much canopy as possible, tenbach, R. R. Beatson, A. Bridegam, amphibian populations in logged help prevent the soil surface R. B. Bury, C. Campbell, S. Corn, T. and old growth redwood forest. from drying out and will H. Fritts, B. E. Smith, and M. C. probably benefit salaman- Northwest Science 57:167-178. Tremble. S. Stefferud (Endangered ders. Bury, R. Bruce and Martin G. Ra- Species, U.S. Fish & Wildlife Service) phael. 1983. Inventory methods and C. Painter (Endangered Species Because current timber har- for amphibians and reptiles. p. Program, New Mexico Department vest schedules will inevitably 416-419. In Renewable resource of Game & Fish) were welcome field lead to younger-aged stands inventories for monitoring companions. S. Corn provided pho- with few or only small changes and trends. J. F. Bell and tographs. K. Aubry, K. Buhlmann, S. downed logs, a mix of young T. Atterbury, eds. Oregon State Corn, C. K. Dodd, and C. Painter and old logs should be main- University, Corvallis. provided helpful criticism of earlier tained to ensure short-term Campbell, Howard W. and Stephen drafts. and long-term habitat com- P. Christman. 1982. Field tech- ponents. Old logs provide niques for herpe tofaunal commu- cover to Aneides and Pletho- nity analysis. p. 193-200. In Herpe- Literature Cited don, while younger logs are tological communities. Norman J. potential sources of cover in Scott, ed. U.S. Fish & Wildlife Aubry, Keith B., Clyde M. Senger, future years. Service, Wildlife Research Rep. 13. and Rodney L. Crawford. 1987. Castetter, Edward, F. 1956. The vege- tation of New Mexico. New Mex- Plethodon larselli Bums. Biological Stebbins and Riemer, Jemez ico Quarterly 26:256-285. Conservation 34:169-179. Mountains salamander. Submitted Childs, S. W. and L. E. Flint. 1987. Ef- Hubbard, John P., Marshall C. Con- to the U.S. Fish & Wildlife Service, fect of shadecards, shelterwoods, way, Howard Campbell, Gregory Region 2, Office of Endangered and clearcuts on temperature and Schmitt, and Michael D. Hatch. Species, Albuquerque, NM. 55 p. moisture environments. Forest 1979. Handbook of species endan- Ramotnik, Cynthia A. 1988. Habitat Ecol. and Manage. 18:205-217. gered in New Mexico. New Mex- requirements and movements of Corn, P. Steven and R. Bruce Bury. In ico Department of Game & Fish. Jemez Mountains salamanders, press. Sampling terrestrial am- Jaeger, Robert G. 1971. Moisture as a Plethodon neomexicanus. M.S. the- phibians and reptiles. In Popula- factor influencing the distributions sis, Colorado State University, tion Monitoring Techniques for of two species of terrestrial sala- Fort Collins. 84 p. Wildlife in Pacific Northwest For- manders. Oecologia 6: 191-207. Raphael, Martin G, and K. V. Rosen- ests. A. B. Carey and L. F. Ruggi- Jaeger, Robert G. 1980. Microhabitats berg. 1983. An integrated ap- ero, eds. USDA Forest Service of a terrestrial forest salamander. proach to wildlife inventories in General Technical Report. Pacific Copeia 1980:265-268. forested habitats. p. 219-222. In Northwest Research Station, Port- Kleeberger, Steven R. and J. Kirwin Renewable resource inventories land, Oregon. Werner. 1982. Home range and for monitoring changes and DeByle, Norbert V. 1981. Clearcut- homing behavior of Plethodon cin- trends. J. F. Bell and T. Atterbury, ting and fire in the larch/Douglas- ereus in northern Michigan. Copeia eds. Oregon State University, fir forests of western Montana-a 1982:409-415. Corvallis. multifaceted research summary. Krag, R., K. Higginbotham, and R. Reagan, Douglas P. 1972. Ecology General Technical Report INT-99. Rothwell. 1986. Logging and soil and distribution of the Jemez USDA Forest Service, Intermoun- disturbance in southeast British Mountains salamander, Plethodon tain Forest and Range Experiment Columbia. Canadian Journal of neomexicanus. Copeia 1972:486-492. Station. 73 p. Forest Research 163345-1354. SAS Institute Inc. 1982. SAS user's Franklin, J. F., K. Cromack, Jr., W. Kunkel, Kenneth E. 1984. Tempera- guide: statistics, 1982 ed. Cary, Denison, A. McKee, C. Maser, J. ture and precipitation summaries North Carolina: SAS Institute Inc. Sedell, F. Swanson, and G. Juday. for selected New Mexico locations. 584 p. 1981. Ecological characteristics of Climate report, Office of State Cli- Scott, Norman J. Jr., Cynthia A. Ra- old-growth Douglas-fir forests. matologist, New Mexico Depart- motnik, Marilyn J. Altenbach, and General Technical Report PNW- ment of Agriculture, Las Cruces, Brian E. Smith. 1987. Distribution 118. USDA Forest Service, Port- New Mexico. and ecological requirements of land, Oregon. 48 p. Lemmon, Paul E. 1956. A spherical endemic salamanders in relation French, Thomas W. and Robert H. densiometer for estimating forest to forestry management. Summary Mount. 1978. Current status of the overstory density. Forest Science of 1987 activities. Part 2: Santa Fe Red Hills salamander, 2:314-320. National Forest. Final report sub- Phaeognathus hubrichti Highton, Lowe, Charles H., Jr. 1950. The sys- mitted to the USDA Forest Serv- and factors affecting its distribu- tematic status of the salamander ice, Albuquerque, New Mexico. tion. Journal of the Alabama Plethodon hardii, with a discussion 33 p. Academy of Science 49:172-179. of biogeographic problems in Shelford, Victor E. 1963. The ecology Gordon, Robert E., James A. MacMa- Aneides. Copeia 1950:92-99. of North America. University of 11- hon, and David B. Wake. 1962. MacCullough, Ross D. and J. R. linois Press, Urbana. 610 p. Relative abundance, microhabitat Bider. 1975. Phenology, migra- Snedecor, George W. and William G. and behavior of some Southern tions, circadian rhythm and the ef- Cochran. 1967. Statistical methods. Appalachian salamanders. Zool- fect of precipitation on the activity 6th ed. The Iowa State University ogica 47:9-14. of Eu ycea b. bislineata in Quebec. Press, Ames, Iowa. 593 p. Heatwole, Harold. 1962. Environ- Herpetologica 31 :433-439. Staub, Nancy. 1986. A status survey mental factors influencing local Pough, F. Harvey, Ellen M. Smith, of the Sacramento Mountain distribution and activity of the Donald H. Rhodes, and Andres salamander, Aneides hardii, with an salamander, Plethodon cinereus. Collazo. 1987. The abundance of assessment of the impact of log- Ecology 43:460-472. salamanders in forest stands with ging on salamander abundance. Herrington, Robert E. and John H. different histories of disturbance. Final report submitted to the U.S. Larsen. 1985. Current status, habi- Forest Ecol. and Manage. 20:l-9. Fish and Wildlife Service, Region tat requirements and management Ramotnik, Cynthia A. 1986. Status 2, Endangered Species Program, of the Larch Mountain salamander report let hod on neomexicanus Albuquerque, New Mexico. 42 p. US. Fish & Wildlife Service. 1986. Distribution and ecological requirements of endemic salamanders in relation to forestry management. Part 2: Santa Fe National Forest. Report submitted to the US. Forest Service, Albuquerque, New Mexico. 40 p. Weigmann, Diana L., Mark Hakkila, Karl Whitmore, and Richard A. Cole. 1980. Survey of Sacramento Mountain salamander habitat of the Cloudcroft and Mayhill dis- tricts of the Lincoln National For- est. Final report submitted to the US. Forest Service, Lincoln Na- tional Forest. 45 p. Whitford, Walter G. 1968. Physiologi- cal responses to temperature and desiccation in the endemic New Mexico plethodontids, Plethodon neomexicanus and Aneides hardii. Copeia 1968:247-251. Whitford, Walter G. and John Ludwig. 1975. The biota of the Baca geothermal site. Report for Union Oil Co.-Geothermal Divi- sion. 140 p. Williams, Byron K. 1983. Some observations on the use of discriminant analysis in ecology. Ecology 64:1283-1291. Williams, Stephen R. 1972. Repro- duction and ecology of the Jemez Mountains salamander, Plethodon neomexicanus. M. S. thesis, Univ. of New Mexico, Albuquerque. 98 p. Williams, Stephen R. 1976. Compara- tive ecology and reproduction of the endemic New Mexico plethodontid salamanders, Plethodon neomexicanus and Aneides hardii. Ph. D. thesis, University of New Mex- ico, Albuquerque. 152 p. Wyman, Richard L. 1988. Soil acidity and moisture and the distribution of amphibians in five forests of southcentral New York. Copeia 1988:394-399. Wyman, Richard L. and Dianne S. Hawksley-Lescault. 1987. Soil acidity affects distribution, behavior, and physiology of the salamander Plethodon cinereus. Ecolow 68:1819-1827.