BioOne sees sustainable scholarly publishing as an inherently collaborative enterprise connecting authorsnonprofit publishers academic institutions research libraries and research funders in the common goal ofmaximizing access to critical research

DISTRIBUTION OF SOREX CINEREUS AND SFUMEUS ON NORTH- AND SOUTH-FACINGSLOPES IN THE SOUTHERN APPALACHIANMOUNTAINSAuthor(s) M Patrick BrannonSource Southeastern Naturalist 1(3)299-306 2002Published By Eagle Hill InstituteDOI httpdxdoiorg1016561528-7092(2002)001[0299DOSCAS]20CO2URL httpwwwbiooneorgdoifull1016561528-7092282002290015B02993ADOSCAS5D20CO3B2

BioOne (wwwbiooneorg) is a nonprofit online aggregation of coreresearch in the biological ecological and environmental sciences BioOneprovides a sustainable online platform for over 170 journals and bookspublished by nonprofit societies associations museums institutions andpresses

Your use of this PDF the BioOne Web site and all posted and associatedcontent indicates your acceptance of BioOnersquos Terms of Use available atwwwbiooneorgpageterms_of_use

Usage of BioOne content is strictly limited to personal educational andnon-commercial use Commercial inquiries or rights and permissionsrequests should be directed to the individual publisher as copyright holder

SOUTHEASTERN NATURALIST2002 1(3)299-306

DISTRIBUTION OF SOREX CINEREUS AND SFUMEUS ON NORTH- AND SOUTH-FACING SLOPES

IN THE SOUTHERN APPALACHIAN MOUNTAINS

M PATRICK BRANNON 1

ABSTRACT ndash The distributions of masked shrews (Sorex cinereus Kerr) andsmoky shrews (S fumeus Miller) were examined in relation to environmentalconditions on opposing slopes in the southern Appalachian Mountains Arraysof drift fences and pitfall traps were established on 3 north-facing plots 3 south-facing plots and 6 streamside plots in the Gingercake Creek drainage of BurkeCounty North Carolina Shrews and invertebrates were collected on severalconsecutive nights each month in the autumn of 1996 and in the spring andsummer of 1997 (TN = 2544) Leaf litter moisture content daily high tempera-tures and volume of downed logs were measured South-facing plots weresignificantly warmer and drier than north-facing and streamside plots (p lt0001) and had the lowest percentage volume of heavily decomposed logs (p =002) Capture rates of S cinereus and S fumeus were significantly greater onmesic north-facing slopes and at streamside sites than on xeric south-facingslopes (p = 005 and 004) but biomass of invertebrate prey was similar betweenhabitats Smoky shrews were significantly positively correlated with percentagelitter moisture and heavily decomposed logs but negatively correlated withinvertebrate biomass Masked shrews were not significantly correlated with anyhabitat variable Results suggest that the distributions of these shrews arestrongly influenced by habitat heterogeneity at high-resolution scales

INTRODUCTION

Distributions of shrews are greatly influenced by environmentalmoisture (Kirkland 1991 Pagels et al 1994 Parmley and Harley 1995Spencer and Pettus 1966 Wrigley et al 1979) Because of their highmetabolic rates shrews experience relatively high respiratory waterlosses and may be unable to regulate such losses in xeric environments(Getz 1961) Moisture is also important in supporting diverse and abun-dant litter invertebrate fauna (Gist and Crossley 1975) that serve as thefood resource base for soricids (Kirkland 1991)

Masked shrews (Sorex cinereus Kerr) and smoky shrews (S fumeusMiller) are common soricids in the southern Appalachian mountainsthat occur primarily in moist forests with considerable structural micro-habitat such as logs and rocks (Brannon 2000 McCay et al 1998Mitchell et al 1997) They are most abundant in mesic northern hard-

1 Virginia Museum of Natural History 1001 Douglas Avenue Martinsville VA24112 pbrannonvmnhorg

Southeastern Naturalist Vol 1 No 3300

wood and cove hardwood communities but are relatively rare in xericoak-pine communities (Laerm et al 1999) Components of the forest-floor microhabitat may help to mediate humidity levels (Getz 1961Laerm et al 1999 McCay et al 1998 Pagels et al 1994)

However species may respond to components of their environmenton a scale of resolution finer than gross habitat differences (Dueser andShugart 1978) Habitat patches that seem homogenous at low resolu-tions may be extremely heterogeneous at high resolutions (Orrock et al2000) For example moisture temperature and associated vegetationcan differ greatly even between neighboring slopes (Matlack 1993Whittaker 1956) In the northern hemisphere north-facing slopes aretypically cooler and wetter than south-facing slopes because durationand intensity of sunlight exposure are reduced (Wales 1972) Conse-quently small mammal species with high moisture requirements maynot be distributed equally between opposing slopes (McComb andRumsey 1982) My objective was to determine if distributions of Scinereus and S fumeus in the southern Appalachian Mountains areaffected by variation in environmental moisture between north- andsouth-facing slopes

METHODS

Twelve 50- by 50-m plots were established in an approximately 15km2 area of the Gingercake Creek drainage of the Pisgah National ForestBurke County western North Carolina (35˚5530 81˚520) This areaconsisted of steep (20-38˚) north- and south-facing slopes separated by arelatively narrow riparian corridor that flowed roughly west to east Standages were about 55 years and elevations averaged 787 m

Plots were spaced a minimum distance of 100 m apart Three plotswere located on north-facing slopes and another 3 were located di-rectly opposite from them on neighboring south-facing slopes Woodyvegetation on north-facing slopes consisted primarily of white pine(Pinus strobus) chestnut oak (Quercus prinus) and red maple (Acerrubrum) with an understory of rhododendron (Rhododendron maxi-mum) South-facing slopes were dominated by white pine red maplechestnut oak and red oak (Q rubra) with a sparse understory ofmountain laurel (Kalmia latifolia) The remaining six plots were lo-cated within the riparian corridor and served as controls due to thepresence of a permanent water source Vegetation in streamside areasconsisted mainly of tulip poplar (Liriodendron tulipifera) black birch(Betula lenta) and eastern hemlock (Tsuga canadensis) with a denseunderstory of rhododendron

Drift fences with associated pitfalls were used to collect shrews Asingle Y-shaped array was installed in the middle of each plot consist-ing of a central pitfall surrounded by three other pitfalls spaced 3 m from

MP Brannon2002 301

the center Each 20-L pitfall (5-gal plastic bucket) was connected to thecentral pitfall by a 61 cm tall section of aluminum flashing Pitfalls werepartially filled with water to drown shrews quickly and prevent preda-tion within traps (Kirkland and Sheppard 1994) Pitfalls were openconcurrently for 5 to 7 consecutive nights each month from August toNovember 1996 and from March to August 1997 for a total of 2544 trapnights (TN) Trappability was assumed to be equal in all habitatsShrews were removed daily during each trapping period and depositedin the Appalachian State University mammal collection

Pitfalls also collected large numbers of epigeal invertebrates whichrepresented available prey Invertebrates were removed during eachtrapping period and collectively weighed to obtain total biomass foreach plot Invertebrate taxa that are not regularly eaten by shrews suchas millipedes were excluded

Environmental conditions of the forest floor at each plot were mea-sured for each trapping period Max-min thermometers were orientedwith slope and placed 5 cm off the forest floor near the center of eachplot to record daily high temperatures These values were used to calcu-late average maximum temperature per plot for each trapping period

A 025-m2 sample of leaf litter was collected from a randomlyselected point within each plot on the first and last day of each trappingperiod Samples were weighed immediately upon return to the lab (wetmass) dried at 100˚C for 24 h and reweighed (dry mass) Litter mois-ture content of each sample was calculated as wet mass minus dry massand expressed as a percentage of wet mass (Brannon 2000)

Because of their spongy texture logs in the advanced stages ofdecomposition may also serve as a source of environmental moisture forshrews (Brannon 2000) For each plot the diameter and length of everydowned log ge10 cm in diameter was measured to determine its volumeAny branches ge10 cm in diameter were treated as separate logs Stage ofdecomposition was classified for each log from class 1 for recentlyfallen logs with little evidence of decay to class 5 for extremely decom-posed logs (Maser et al 1979)

Differences in capture rates (number of captures per 100 TN) of Scinereus and S fumeus between plots in each of the three habitat typeswere examined using one-way analysis of variance (Zar 1984) One-wayANOVA was also used to examine differences in invertebrate biomassaverage maximum temperatures percentage litter moisture and per-centage volume of heavily decomposed (decay classes 4 and 5) logsbetween sites All percent values were arcsine-transformed prior toanalysis (Zar 1984) To assess their influence on shrew speciesrsquo relativeabundance habitat variables were correlated with shrew capture data foreach plot using Pearsonrsquos product moment correlation analysis (Ford etal 1997 Zar 1984)

Southeastern Naturalist Vol 1 No 3302

RESULTS

Sites differed significantly in both maximum temperatures (F2117 =2209 p lt 0001) and percentage litter moisture content (F2237 = 7934 plt 0001) South-facing slopes had the highest mean (plusmn 1 SE) maximumtemperatures (266 plusmn 06 EC) whereas north-facing slopes had thelowest (209 plusmn 05 EC) and those at streamsides were intermediate (228plusmn 06 EC) Mean percentage litter moisture was lowest on south-facingslopes (4727 plusmn 153) but similar between north-facing slopes (5671plusmn 180) and streamside habitats (5997 plusmn 109)

Percentage volume of heavily decomposed logs also differed signifi-cantly between sites (F29 = 511 p = 002) Mean percentage volume washigh on north-facing slopes (9692 plusmn 062) and at streamside sites (8878 plusmn499) but low on south-facing slopes (3037 plusmn 2148) Two of the threesouth-facing plots had no logs that were classified as heavily decomposed

No significant differences in invertebrate biomass were observedbetween sites (F29 = 064 p = 054) Mean biomass was 20779 plusmn 3458g at streamside sites 23741 plusmn 1697 g on north-facing slopes and31368 plusmn 2987 g on south-facing slopes

A total of 17 S cinereus and 32 S fumeus were collected on north-facing slopes whereas only 4 S cinereus and 10 S fumeus were cap-tured on south-facing slopes (Table 1) Capture rates (captures per 100TN) were significantly different between sites for both masked shrews(F29 = 354 p = 005) and smoky shrews (F29 = 410 p = 004) Meancapture rates of masked shrews were highest at north-facing plots (267plusmn 003) intermediate at streamside plots (157 plusmn 004) and lowest atsouth-facing plots (063 plusmn 003 Table 1) Only 33 of masked shrewcaptures occurred at each south-facing plot compared to 138 at eachnorth-facing plot and 81 at each streamside plot Mean capture ratesof smoky shrews were similar at mesic north-facing (503 plusmn 059) andstreamside plots (495 plusmn 124) but over 3 times less at xeric south-facingplots (157 plusmn 059 Table 1) Each south-facing plot yielded 32 ofcaptures of smoky shrews compared to 102 at each north-facing plotand 100 at each streamside plot

Smoky shrews (n = 105) were strongly positively correlated withpercentage litter moisture and heavily decomposed logs and signifi-cantly negatively correlated with invertebrate biomass (Table 2) There

Table 1 Summary of shrew captures in each habitat type in the study area in Autumn 1996and spring and summer 1997

Habitat Plots TN Sorex cinereus S fumeus Combined

n capture relative n capture relative n capture relativerate abundance rate abundance rate abundance

Streamside 6 1272 20 157 81 63 495 100 83 653 947North-facing slope 3 636 17 267 138 32 503 102 49 770 1119South-facing slope 3 636 4 063 33 10 157 32 14 220 320

MP Brannon2002 303

was no significant correlation between smoky shrews and maximumtemperatures Masked shrews (n = 41) showed no significant correlationwith any habitat variable (Table 2) although correlations with littermoisture and percentage of rotten logs were positive

DISCUSSION

Although many factors may influence distributions of soricids(Pagels et al 1994 Parmley and Harley 1995) environmental moisturemay be of the greatest importance to shrews because of their extremelyhigh water turnover rates (Getz 1961 Wrigley et al 1979) Maskedshrews and smoky shrews are most abundant in mesic forests and nearstanding water but are uncommon if not absent in dry habitats(Cudmore and Whitaker 1984 Getz 1961 Laerm et al 1999 Owen1984) North-facing slopes are generally cooler and moister than south-facing slopes (Wales 1972) and consequently yield greater numbers ofshrews (Laerm et al 1999 McComb and Rumsey 1982)

Greater sunlight exposure on south-facing slopes results in highertemperatures and evaporative drying of the leaf litter (Wales 1972) In thisstudy maximum temperatures increased gradually from north- to south-facing slopes but environmental moisture did not decrease along thisgradient Although temperatures were higher at streamside sites leaf littermoisture content in these areas was similar to that of cooler north-facingslopes and consequently had similar capture rates of shrews Streamsidehabitats often represent isolated zones of moisture in otherwise dryenvironments (Laerm et al 1999) because leaf litter in these areas is oftennot far above the water table (Sites 1978) Habitats such as riparian areasthat remain moist regardless of temperature provide favorable environ-mental conditions that facilitate the active movement of shrews

The primary importance of structural components of the microhabitatto shrews is their effect on humidity (Getz 1961 McComb and Rumsey1982 Pagels et al 1994 Parmley and Harley 1995) The percentage oflogs that were heavily decomposed was high on north-facing slopes andin streamside habitats in the study area but extremely low on south-facingslopes of the same stand age presumably because logs decompose moreslowly under xeric conditions (Abbott and Crossley 1982) Althoughmost logs can function as protective cover and as foraging sites (Loeb

Table 2 Correlation coefficients (r) between measured habitat variables and shrews in thestudy area in autumn 1996 and spring and summer 1997

Habitat variable Sorex cinereus (n = 41) S fumeus (n = 105) Combined (n = 146)

litter moisture 0443 0559 0490 Average maximum temperature -0254 -0484 -0481Invertebrate biomass -0267 -0873 -0785 heavily decomposed logs 0393 0553 0593

p le 005

Southeastern Naturalist Vol 1 No 3304

1993) those in the advanced stages of decomposition may be optimalmicrohabitats for shrews (Brannon 2000) because they contain highconcentrations of water (Maser et al 1979) and can serve as moisturerefugia during dry periods Shading from the dense understory of rhodo-dendron likely also contributed to favorable environmental moistureconditions for S cinereus and S fumeus on north-facing slopes and alongstreams (Laerm et al 1995 Pagels et al 1994)

Environmental moisture is also important to the distribution ofshrews because it may affect the availability of invertebrate prey (Gistand Crossley 1975 Kirkland 1991 Parmley and Harley 1995 Wrigleyet al 1979) Although in this study no significant differences in inverte-brate biomass existed between mesic and xeric habitats in moist envi-ronments prey may be more accessible because mobility of the shrews isenhanced (McCay 1996) If the litter is wet shrews may increase theiractivity on the forest floor where invertebrates are abundant but if thelitter is dry they may be restricted to small isolated patches of moisturewhere availability of prey may be reduced (Jaeger 1980) Althoughother prey items such as salamanders are abundant in mesic habitats ofthe southern Appalachian Mountains (Petranka et al 1993) they consti-tute only a very small percentage of the diets of most shrews (Hamilton1930) Greater numbers of shrews in mesic environments probably havemore to do with better environmental conditions for foraging associatedwith their ecophysiology than to an increased food supply

For small mammals associated with mesic environments speciespresence and abundance may be most predictable at a habitat scale witha high degree of resolution (Orrock et al 2000) The southern Appala-chian Mountains are comprised of a complex mosaic of habitats (Laermet al 1995 Laerm et al 1999 Pagels et al 1994 Whittaker 1956)Shrew species diversity is greatest at higher elevation sites where envi-ronmental conditions resemble those of more northern forests (Laerm etal 1999 Pagels et al 1994) Habitats that provide shaded moist condi-tions with complex structural microhabitats are especially important tothe distributions of shrews (Cudmore and Whitaker 1984 Ford et al1997 Kirkland 1991 Laerm et al 1999 Mitchell et al 1997 Parmleyand Harley 1995) My results are consistent with those of others (Getz1961 Spencer and Pettus 1966 Wrigley et al 1979) who found that theprimary factor governing the distributions of shrews is environmentalmoisture Moreover this study suggests that the distribution of Scinereus and S fumeus is strongly influenced by heterogeneity of mois-ture across landscapes such as between north- and south-facing slopes

ACKNOWLEDGMENTS

This study was supported in part by Appalachian State University GraduateSchool Graduate Student Association Senate and Biology Graduate StudentAssociation I thank R W Van Devender G L Walker and D B Meikle for

MP Brannon2002 305

their guidance during this project and J K H Brannon and many others forfield assistance I also thank J F Merritt J F Pagels and an anonymousreviewer for their suggestions for revision of the original manuscript

LITERATURE CITED

ABBOTT DT and DA CROSSLEY 1982 Woody litter decompositionfollowing clear-cutting Ecology 6335-42

BRANNON MP 2000 Niche relationships of two syntopic species of shrewsSorex fumeus and S cinereus in the southern Appalachian MountainsJournal of Mammalogy 811053-1061

CUDMORE WW and JO WHITAKER Jr 1984 The distribution of thesmoky shrew Sorex fumeus and the pygmy shrew Microsorex hoyi inIndiana with notes on the distribution of other shrews Proceedings of theIndiana Academy of Sciences 93469-474

DUESER RD and HH SHUGART 1978 Microhabitats in a forest-floorsmall mammal fauna Ecology 5989-98

FORD WM J LAERM and KG BARKER 1997 Soricid response to foreststand age in southern Appalachian cove hardwood communities ForestEcology and Management 91175-181

GETZ LL 1961 Factors influencing the local distribution of shrews Ameri-can Midland Naturalist 6567-88

GIST CS and DA CROSSLEY 1975 The litter arthropod community in asouthern Appalachian hardwood forest numbers biomass and mineral ele-ment content American Midland Naturalist 93107-122

HAMILTON WJ Jr 1930 The food of the Soricidae Journal of Mammalogy1126-39

JAEGER RG 1980 Microhabitats of a terrestrial forest salamander Copeia1980265-268

KIRKLAND GL Jr 1991 Competition and coexistence in shrews(Insectivora Soricidae) Pp 15-22 In JS Findley and TL Yates (Eds)The Biology of the Soricidae Special Publication of the Museum of South-western Biology University of New Mexico Albuquerque NM 91 pp

KIRKLAND GL Jr and PK SHEPPARD 1994 Proposed standard proto-col for sampling small mammal communities Pp 277-283 In JF MerrittGL Kirkland Jr and RK Rose (Eds) Advances in the Biology of ShrewsSpecial Publication of the Carnegie Museum of Natural History No 18Pittsburgh PA 458 pp

LAERM J E BROWN MA MENZEL A WOTJALIK WM FORD and MSTRAYER 1995 The masked shrew Sorex cinereus (Insectivora Soricidae)and red-backed vole Clethrionomys gapperi (Rodentia Muridae) in the BlueRidge province of South Carolina Brimleyana 2215-21

LAERM J WM FORD TS MCCAY MA MENZEL LT LEPARDO andJL BOONE 1999 Soricid communities in the southern Appalachians Pp177-193 In RP Eckerlin (Ed) Proceedings of the Appalachian Biogeogra-phy Symposium Special Publication of the Virginia Museum of NaturalHistory No 7 Martinsville VA 258 pp

LOEB SC 1993 The role of coarse woody debris in the ecology of southeast-ern mammals Pp 108-118 In JW McMinn and DA Crossley (Eds)Biodiversity and Coarse Woody Debris in Southern Forest USDA ForestService General Technical Report SE-94

Southeastern Naturalist Vol 1 No 3306

MASER C RG ANDERSON K CROMACK Jr JT WILLIAMS and REMARTIN 1979 Dead and down woody material Pp 78-95 In JW Tho-mas (Ed) Wildlife Habitats in Managed Forests the Blue Mountains ofOregon and Washington USDA Forest Service Handbook 553 512 pp

MATLACK GR 1993 Microenvironment variation within and among forestedge sites in the eastern United States Biological Conservation 66185-194

McCAY TS 1996 Response of masked shrews (Sorex cinereus) to precipita-tion in irrigated and nonirrigated forests American Midland Naturalist135178-180

McCAY TS J LAERM MA MENZEL and WM FORD 1998 Methodsused to survey shrews (Insectivora Soricidae) and the importance of forest-floor structure Brimleyana 25110-119

McCOMB WC and RL RUMSEY 1982 Response of small mammals toforest floor clearings created by herbicides in the central AppalachiansBrimleyana 8121-134

MITCHELL JC SC RINEHART JF PAGELS KA BUHLMANN andCA PAGUE 1997 Factors influencing amphibian and small mammalassemblages in central Appalachian forests Forest Ecology and Manage-ment 9665-76

ORROCK JL JF PAGELS WJ McSHEA and EK HARPER 2000 Pre-dicting presence and abundance of a small mammal species the effect ofscale and resolution Ecological Applications 101356-1366

OWEN JG 1984 Sorex fumeus Mammalian Species 2151-8PAGELS JF KL UTHUS and HE DUVAL 1994 The masked shrew

Sorex cinereus in a relictual habitat of the southern Appalachian mountainspp 103-109 In JF Merritt GL Kirkland Jr and RK Rose (Eds) Ad-vances in the Biology of Shrews Special Publication of the Carnegie Mu-seum of Natural History No 18 Pittsburgh PA 458 pp

PARMLEY D and D HARLEY 1995 The relative seasonal abundance ofshrews in two central Georgia deciduous woodlots Georgia Journal ofScience 5383-88

PETRANKA JW ME ELDRIDGE and KE HALEY 1993 Effects oftimber harvesting on southern Appalachian salamanders Conservation Bi-ology 7363-370

SITES JW 1978 The foraging strategy of the dusky salamanderDesmognathus fuscus (Amphibia Urodela Plethodontidae) an empiricalapproach to predation theory Journal of Herpetology 12373-383

SPENCER AW and D PETTUS 1966 Habitat preferences of five sympatricspecies of long-tailed shrews Ecology 47677-683

WALES BA 1972 Vegetation analysis of north and south edges in a matureoak-hickory forest Ecological Monographs 42451-471

WHITTAKER RH 1956 Vegetation of the Great Smoky Mountains Ecologi-cal Monographs 261-80

WRIGLEY RE JE DUBOIS and HWR COPELAND 1979 Habitat abun-dance and the distribution of six species of shrews in Manitoba Journal ofMammalogy 60505-520

ZAR JH 1984 Biostatistical Analysis 2nd ed Prentice-Hall IncEnglewood-Cliffs NJ 718 pp

Southeastern Naturalist Vol 1 No 3300

wood and cove hardwood communities but are relatively rare in xericoak-pine communities (Laerm et al 1999) Components of the forest-floor microhabitat may help to mediate humidity levels (Getz 1961Laerm et al 1999 McCay et al 1998 Pagels et al 1994)

However species may respond to components of their environmenton a scale of resolution finer than gross habitat differences (Dueser andShugart 1978) Habitat patches that seem homogenous at low resolu-tions may be extremely heterogeneous at high resolutions (Orrock et al2000) For example moisture temperature and associated vegetationcan differ greatly even between neighboring slopes (Matlack 1993Whittaker 1956) In the northern hemisphere north-facing slopes aretypically cooler and wetter than south-facing slopes because durationand intensity of sunlight exposure are reduced (Wales 1972) Conse-quently small mammal species with high moisture requirements maynot be distributed equally between opposing slopes (McComb andRumsey 1982) My objective was to determine if distributions of Scinereus and S fumeus in the southern Appalachian Mountains areaffected by variation in environmental moisture between north- andsouth-facing slopes

METHODS

Twelve 50- by 50-m plots were established in an approximately 15km2 area of the Gingercake Creek drainage of the Pisgah National ForestBurke County western North Carolina (35˚5530 81˚520) This areaconsisted of steep (20-38˚) north- and south-facing slopes separated by arelatively narrow riparian corridor that flowed roughly west to east Standages were about 55 years and elevations averaged 787 m

Plots were spaced a minimum distance of 100 m apart Three plotswere located on north-facing slopes and another 3 were located di-rectly opposite from them on neighboring south-facing slopes Woodyvegetation on north-facing slopes consisted primarily of white pine(Pinus strobus) chestnut oak (Quercus prinus) and red maple (Acerrubrum) with an understory of rhododendron (Rhododendron maxi-mum) South-facing slopes were dominated by white pine red maplechestnut oak and red oak (Q rubra) with a sparse understory ofmountain laurel (Kalmia latifolia) The remaining six plots were lo-cated within the riparian corridor and served as controls due to thepresence of a permanent water source Vegetation in streamside areasconsisted mainly of tulip poplar (Liriodendron tulipifera) black birch(Betula lenta) and eastern hemlock (Tsuga canadensis) with a denseunderstory of rhododendron

Drift fences with associated pitfalls were used to collect shrews Asingle Y-shaped array was installed in the middle of each plot consist-ing of a central pitfall surrounded by three other pitfalls spaced 3 m from

MP Brannon2002 301

the center Each 20-L pitfall (5-gal plastic bucket) was connected to thecentral pitfall by a 61 cm tall section of aluminum flashing Pitfalls werepartially filled with water to drown shrews quickly and prevent preda-tion within traps (Kirkland and Sheppard 1994) Pitfalls were openconcurrently for 5 to 7 consecutive nights each month from August toNovember 1996 and from March to August 1997 for a total of 2544 trapnights (TN) Trappability was assumed to be equal in all habitatsShrews were removed daily during each trapping period and depositedin the Appalachian State University mammal collection

Pitfalls also collected large numbers of epigeal invertebrates whichrepresented available prey Invertebrates were removed during eachtrapping period and collectively weighed to obtain total biomass foreach plot Invertebrate taxa that are not regularly eaten by shrews suchas millipedes were excluded

Environmental conditions of the forest floor at each plot were mea-sured for each trapping period Max-min thermometers were orientedwith slope and placed 5 cm off the forest floor near the center of eachplot to record daily high temperatures These values were used to calcu-late average maximum temperature per plot for each trapping period

A 025-m2 sample of leaf litter was collected from a randomlyselected point within each plot on the first and last day of each trappingperiod Samples were weighed immediately upon return to the lab (wetmass) dried at 100˚C for 24 h and reweighed (dry mass) Litter mois-ture content of each sample was calculated as wet mass minus dry massand expressed as a percentage of wet mass (Brannon 2000)

Because of their spongy texture logs in the advanced stages ofdecomposition may also serve as a source of environmental moisture forshrews (Brannon 2000) For each plot the diameter and length of everydowned log ge10 cm in diameter was measured to determine its volumeAny branches ge10 cm in diameter were treated as separate logs Stage ofdecomposition was classified for each log from class 1 for recentlyfallen logs with little evidence of decay to class 5 for extremely decom-posed logs (Maser et al 1979)

Differences in capture rates (number of captures per 100 TN) of Scinereus and S fumeus between plots in each of the three habitat typeswere examined using one-way analysis of variance (Zar 1984) One-wayANOVA was also used to examine differences in invertebrate biomassaverage maximum temperatures percentage litter moisture and per-centage volume of heavily decomposed (decay classes 4 and 5) logsbetween sites All percent values were arcsine-transformed prior toanalysis (Zar 1984) To assess their influence on shrew speciesrsquo relativeabundance habitat variables were correlated with shrew capture data foreach plot using Pearsonrsquos product moment correlation analysis (Ford etal 1997 Zar 1984)

Southeastern Naturalist Vol 1 No 3302

RESULTS

Sites differed significantly in both maximum temperatures (F2117 =2209 p lt 0001) and percentage litter moisture content (F2237 = 7934 plt 0001) South-facing slopes had the highest mean (plusmn 1 SE) maximumtemperatures (266 plusmn 06 EC) whereas north-facing slopes had thelowest (209 plusmn 05 EC) and those at streamsides were intermediate (228plusmn 06 EC) Mean percentage litter moisture was lowest on south-facingslopes (4727 plusmn 153) but similar between north-facing slopes (5671plusmn 180) and streamside habitats (5997 plusmn 109)

Percentage volume of heavily decomposed logs also differed signifi-cantly between sites (F29 = 511 p = 002) Mean percentage volume washigh on north-facing slopes (9692 plusmn 062) and at streamside sites (8878 plusmn499) but low on south-facing slopes (3037 plusmn 2148) Two of the threesouth-facing plots had no logs that were classified as heavily decomposed

No significant differences in invertebrate biomass were observedbetween sites (F29 = 064 p = 054) Mean biomass was 20779 plusmn 3458g at streamside sites 23741 plusmn 1697 g on north-facing slopes and31368 plusmn 2987 g on south-facing slopes

A total of 17 S cinereus and 32 S fumeus were collected on north-facing slopes whereas only 4 S cinereus and 10 S fumeus were cap-tured on south-facing slopes (Table 1) Capture rates (captures per 100TN) were significantly different between sites for both masked shrews(F29 = 354 p = 005) and smoky shrews (F29 = 410 p = 004) Meancapture rates of masked shrews were highest at north-facing plots (267plusmn 003) intermediate at streamside plots (157 plusmn 004) and lowest atsouth-facing plots (063 plusmn 003 Table 1) Only 33 of masked shrewcaptures occurred at each south-facing plot compared to 138 at eachnorth-facing plot and 81 at each streamside plot Mean capture ratesof smoky shrews were similar at mesic north-facing (503 plusmn 059) andstreamside plots (495 plusmn 124) but over 3 times less at xeric south-facingplots (157 plusmn 059 Table 1) Each south-facing plot yielded 32 ofcaptures of smoky shrews compared to 102 at each north-facing plotand 100 at each streamside plot

Smoky shrews (n = 105) were strongly positively correlated withpercentage litter moisture and heavily decomposed logs and signifi-cantly negatively correlated with invertebrate biomass (Table 2) There

Table 1 Summary of shrew captures in each habitat type in the study area in Autumn 1996and spring and summer 1997

Habitat Plots TN Sorex cinereus S fumeus Combined

n capture relative n capture relative n capture relativerate abundance rate abundance rate abundance

Streamside 6 1272 20 157 81 63 495 100 83 653 947North-facing slope 3 636 17 267 138 32 503 102 49 770 1119South-facing slope 3 636 4 063 33 10 157 32 14 220 320

MP Brannon2002 303

was no significant correlation between smoky shrews and maximumtemperatures Masked shrews (n = 41) showed no significant correlationwith any habitat variable (Table 2) although correlations with littermoisture and percentage of rotten logs were positive

DISCUSSION

Although many factors may influence distributions of soricids(Pagels et al 1994 Parmley and Harley 1995) environmental moisturemay be of the greatest importance to shrews because of their extremelyhigh water turnover rates (Getz 1961 Wrigley et al 1979) Maskedshrews and smoky shrews are most abundant in mesic forests and nearstanding water but are uncommon if not absent in dry habitats(Cudmore and Whitaker 1984 Getz 1961 Laerm et al 1999 Owen1984) North-facing slopes are generally cooler and moister than south-facing slopes (Wales 1972) and consequently yield greater numbers ofshrews (Laerm et al 1999 McComb and Rumsey 1982)

Greater sunlight exposure on south-facing slopes results in highertemperatures and evaporative drying of the leaf litter (Wales 1972) In thisstudy maximum temperatures increased gradually from north- to south-facing slopes but environmental moisture did not decrease along thisgradient Although temperatures were higher at streamside sites leaf littermoisture content in these areas was similar to that of cooler north-facingslopes and consequently had similar capture rates of shrews Streamsidehabitats often represent isolated zones of moisture in otherwise dryenvironments (Laerm et al 1999) because leaf litter in these areas is oftennot far above the water table (Sites 1978) Habitats such as riparian areasthat remain moist regardless of temperature provide favorable environ-mental conditions that facilitate the active movement of shrews

The primary importance of structural components of the microhabitatto shrews is their effect on humidity (Getz 1961 McComb and Rumsey1982 Pagels et al 1994 Parmley and Harley 1995) The percentage oflogs that were heavily decomposed was high on north-facing slopes andin streamside habitats in the study area but extremely low on south-facingslopes of the same stand age presumably because logs decompose moreslowly under xeric conditions (Abbott and Crossley 1982) Althoughmost logs can function as protective cover and as foraging sites (Loeb

Table 2 Correlation coefficients (r) between measured habitat variables and shrews in thestudy area in autumn 1996 and spring and summer 1997

Habitat variable Sorex cinereus (n = 41) S fumeus (n = 105) Combined (n = 146)

litter moisture 0443 0559 0490 Average maximum temperature -0254 -0484 -0481Invertebrate biomass -0267 -0873 -0785 heavily decomposed logs 0393 0553 0593

p le 005

Southeastern Naturalist Vol 1 No 3304

1993) those in the advanced stages of decomposition may be optimalmicrohabitats for shrews (Brannon 2000) because they contain highconcentrations of water (Maser et al 1979) and can serve as moisturerefugia during dry periods Shading from the dense understory of rhodo-dendron likely also contributed to favorable environmental moistureconditions for S cinereus and S fumeus on north-facing slopes and alongstreams (Laerm et al 1995 Pagels et al 1994)

Environmental moisture is also important to the distribution ofshrews because it may affect the availability of invertebrate prey (Gistand Crossley 1975 Kirkland 1991 Parmley and Harley 1995 Wrigleyet al 1979) Although in this study no significant differences in inverte-brate biomass existed between mesic and xeric habitats in moist envi-ronments prey may be more accessible because mobility of the shrews isenhanced (McCay 1996) If the litter is wet shrews may increase theiractivity on the forest floor where invertebrates are abundant but if thelitter is dry they may be restricted to small isolated patches of moisturewhere availability of prey may be reduced (Jaeger 1980) Althoughother prey items such as salamanders are abundant in mesic habitats ofthe southern Appalachian Mountains (Petranka et al 1993) they consti-tute only a very small percentage of the diets of most shrews (Hamilton1930) Greater numbers of shrews in mesic environments probably havemore to do with better environmental conditions for foraging associatedwith their ecophysiology than to an increased food supply

For small mammals associated with mesic environments speciespresence and abundance may be most predictable at a habitat scale witha high degree of resolution (Orrock et al 2000) The southern Appala-chian Mountains are comprised of a complex mosaic of habitats (Laermet al 1995 Laerm et al 1999 Pagels et al 1994 Whittaker 1956)Shrew species diversity is greatest at higher elevation sites where envi-ronmental conditions resemble those of more northern forests (Laerm etal 1999 Pagels et al 1994) Habitats that provide shaded moist condi-tions with complex structural microhabitats are especially important tothe distributions of shrews (Cudmore and Whitaker 1984 Ford et al1997 Kirkland 1991 Laerm et al 1999 Mitchell et al 1997 Parmleyand Harley 1995) My results are consistent with those of others (Getz1961 Spencer and Pettus 1966 Wrigley et al 1979) who found that theprimary factor governing the distributions of shrews is environmentalmoisture Moreover this study suggests that the distribution of Scinereus and S fumeus is strongly influenced by heterogeneity of mois-ture across landscapes such as between north- and south-facing slopes

ACKNOWLEDGMENTS

This study was supported in part by Appalachian State University GraduateSchool Graduate Student Association Senate and Biology Graduate StudentAssociation I thank R W Van Devender G L Walker and D B Meikle for

MP Brannon2002 305

their guidance during this project and J K H Brannon and many others forfield assistance I also thank J F Merritt J F Pagels and an anonymousreviewer for their suggestions for revision of the original manuscript

LITERATURE CITED

ABBOTT DT and DA CROSSLEY 1982 Woody litter decompositionfollowing clear-cutting Ecology 6335-42

BRANNON MP 2000 Niche relationships of two syntopic species of shrewsSorex fumeus and S cinereus in the southern Appalachian MountainsJournal of Mammalogy 811053-1061

CUDMORE WW and JO WHITAKER Jr 1984 The distribution of thesmoky shrew Sorex fumeus and the pygmy shrew Microsorex hoyi inIndiana with notes on the distribution of other shrews Proceedings of theIndiana Academy of Sciences 93469-474

DUESER RD and HH SHUGART 1978 Microhabitats in a forest-floorsmall mammal fauna Ecology 5989-98

FORD WM J LAERM and KG BARKER 1997 Soricid response to foreststand age in southern Appalachian cove hardwood communities ForestEcology and Management 91175-181

GETZ LL 1961 Factors influencing the local distribution of shrews Ameri-can Midland Naturalist 6567-88

GIST CS and DA CROSSLEY 1975 The litter arthropod community in asouthern Appalachian hardwood forest numbers biomass and mineral ele-ment content American Midland Naturalist 93107-122

HAMILTON WJ Jr 1930 The food of the Soricidae Journal of Mammalogy1126-39

JAEGER RG 1980 Microhabitats of a terrestrial forest salamander Copeia1980265-268

KIRKLAND GL Jr 1991 Competition and coexistence in shrews(Insectivora Soricidae) Pp 15-22 In JS Findley and TL Yates (Eds)The Biology of the Soricidae Special Publication of the Museum of South-western Biology University of New Mexico Albuquerque NM 91 pp

KIRKLAND GL Jr and PK SHEPPARD 1994 Proposed standard proto-col for sampling small mammal communities Pp 277-283 In JF MerrittGL Kirkland Jr and RK Rose (Eds) Advances in the Biology of ShrewsSpecial Publication of the Carnegie Museum of Natural History No 18Pittsburgh PA 458 pp

LAERM J E BROWN MA MENZEL A WOTJALIK WM FORD and MSTRAYER 1995 The masked shrew Sorex cinereus (Insectivora Soricidae)and red-backed vole Clethrionomys gapperi (Rodentia Muridae) in the BlueRidge province of South Carolina Brimleyana 2215-21

LAERM J WM FORD TS MCCAY MA MENZEL LT LEPARDO andJL BOONE 1999 Soricid communities in the southern Appalachians Pp177-193 In RP Eckerlin (Ed) Proceedings of the Appalachian Biogeogra-phy Symposium Special Publication of the Virginia Museum of NaturalHistory No 7 Martinsville VA 258 pp

LOEB SC 1993 The role of coarse woody debris in the ecology of southeast-ern mammals Pp 108-118 In JW McMinn and DA Crossley (Eds)Biodiversity and Coarse Woody Debris in Southern Forest USDA ForestService General Technical Report SE-94

Southeastern Naturalist Vol 1 No 3306

MASER C RG ANDERSON K CROMACK Jr JT WILLIAMS and REMARTIN 1979 Dead and down woody material Pp 78-95 In JW Tho-mas (Ed) Wildlife Habitats in Managed Forests the Blue Mountains ofOregon and Washington USDA Forest Service Handbook 553 512 pp

MATLACK GR 1993 Microenvironment variation within and among forestedge sites in the eastern United States Biological Conservation 66185-194

McCAY TS 1996 Response of masked shrews (Sorex cinereus) to precipita-tion in irrigated and nonirrigated forests American Midland Naturalist135178-180

McCAY TS J LAERM MA MENZEL and WM FORD 1998 Methodsused to survey shrews (Insectivora Soricidae) and the importance of forest-floor structure Brimleyana 25110-119

McCOMB WC and RL RUMSEY 1982 Response of small mammals toforest floor clearings created by herbicides in the central AppalachiansBrimleyana 8121-134

MITCHELL JC SC RINEHART JF PAGELS KA BUHLMANN andCA PAGUE 1997 Factors influencing amphibian and small mammalassemblages in central Appalachian forests Forest Ecology and Manage-ment 9665-76

ORROCK JL JF PAGELS WJ McSHEA and EK HARPER 2000 Pre-dicting presence and abundance of a small mammal species the effect ofscale and resolution Ecological Applications 101356-1366

OWEN JG 1984 Sorex fumeus Mammalian Species 2151-8PAGELS JF KL UTHUS and HE DUVAL 1994 The masked shrew

Sorex cinereus in a relictual habitat of the southern Appalachian mountainspp 103-109 In JF Merritt GL Kirkland Jr and RK Rose (Eds) Ad-vances in the Biology of Shrews Special Publication of the Carnegie Mu-seum of Natural History No 18 Pittsburgh PA 458 pp

PARMLEY D and D HARLEY 1995 The relative seasonal abundance ofshrews in two central Georgia deciduous woodlots Georgia Journal ofScience 5383-88

PETRANKA JW ME ELDRIDGE and KE HALEY 1993 Effects oftimber harvesting on southern Appalachian salamanders Conservation Bi-ology 7363-370

SITES JW 1978 The foraging strategy of the dusky salamanderDesmognathus fuscus (Amphibia Urodela Plethodontidae) an empiricalapproach to predation theory Journal of Herpetology 12373-383

SPENCER AW and D PETTUS 1966 Habitat preferences of five sympatricspecies of long-tailed shrews Ecology 47677-683

WALES BA 1972 Vegetation analysis of north and south edges in a matureoak-hickory forest Ecological Monographs 42451-471

WHITTAKER RH 1956 Vegetation of the Great Smoky Mountains Ecologi-cal Monographs 261-80

WRIGLEY RE JE DUBOIS and HWR COPELAND 1979 Habitat abun-dance and the distribution of six species of shrews in Manitoba Journal ofMammalogy 60505-520

ZAR JH 1984 Biostatistical Analysis 2nd ed Prentice-Hall IncEnglewood-Cliffs NJ 718 pp

MP Brannon2002 301

the center Each 20-L pitfall (5-gal plastic bucket) was connected to thecentral pitfall by a 61 cm tall section of aluminum flashing Pitfalls werepartially filled with water to drown shrews quickly and prevent preda-tion within traps (Kirkland and Sheppard 1994) Pitfalls were openconcurrently for 5 to 7 consecutive nights each month from August toNovember 1996 and from March to August 1997 for a total of 2544 trapnights (TN) Trappability was assumed to be equal in all habitatsShrews were removed daily during each trapping period and depositedin the Appalachian State University mammal collection

Pitfalls also collected large numbers of epigeal invertebrates whichrepresented available prey Invertebrates were removed during eachtrapping period and collectively weighed to obtain total biomass foreach plot Invertebrate taxa that are not regularly eaten by shrews suchas millipedes were excluded

Environmental conditions of the forest floor at each plot were mea-sured for each trapping period Max-min thermometers were orientedwith slope and placed 5 cm off the forest floor near the center of eachplot to record daily high temperatures These values were used to calcu-late average maximum temperature per plot for each trapping period

A 025-m2 sample of leaf litter was collected from a randomlyselected point within each plot on the first and last day of each trappingperiod Samples were weighed immediately upon return to the lab (wetmass) dried at 100˚C for 24 h and reweighed (dry mass) Litter mois-ture content of each sample was calculated as wet mass minus dry massand expressed as a percentage of wet mass (Brannon 2000)

Because of their spongy texture logs in the advanced stages ofdecomposition may also serve as a source of environmental moisture forshrews (Brannon 2000) For each plot the diameter and length of everydowned log ge10 cm in diameter was measured to determine its volumeAny branches ge10 cm in diameter were treated as separate logs Stage ofdecomposition was classified for each log from class 1 for recentlyfallen logs with little evidence of decay to class 5 for extremely decom-posed logs (Maser et al 1979)

Differences in capture rates (number of captures per 100 TN) of Scinereus and S fumeus between plots in each of the three habitat typeswere examined using one-way analysis of variance (Zar 1984) One-wayANOVA was also used to examine differences in invertebrate biomassaverage maximum temperatures percentage litter moisture and per-centage volume of heavily decomposed (decay classes 4 and 5) logsbetween sites All percent values were arcsine-transformed prior toanalysis (Zar 1984) To assess their influence on shrew speciesrsquo relativeabundance habitat variables were correlated with shrew capture data foreach plot using Pearsonrsquos product moment correlation analysis (Ford etal 1997 Zar 1984)

Southeastern Naturalist Vol 1 No 3302

RESULTS

Sites differed significantly in both maximum temperatures (F2117 =2209 p lt 0001) and percentage litter moisture content (F2237 = 7934 plt 0001) South-facing slopes had the highest mean (plusmn 1 SE) maximumtemperatures (266 plusmn 06 EC) whereas north-facing slopes had thelowest (209 plusmn 05 EC) and those at streamsides were intermediate (228plusmn 06 EC) Mean percentage litter moisture was lowest on south-facingslopes (4727 plusmn 153) but similar between north-facing slopes (5671plusmn 180) and streamside habitats (5997 plusmn 109)

Percentage volume of heavily decomposed logs also differed signifi-cantly between sites (F29 = 511 p = 002) Mean percentage volume washigh on north-facing slopes (9692 plusmn 062) and at streamside sites (8878 plusmn499) but low on south-facing slopes (3037 plusmn 2148) Two of the threesouth-facing plots had no logs that were classified as heavily decomposed

No significant differences in invertebrate biomass were observedbetween sites (F29 = 064 p = 054) Mean biomass was 20779 plusmn 3458g at streamside sites 23741 plusmn 1697 g on north-facing slopes and31368 plusmn 2987 g on south-facing slopes

A total of 17 S cinereus and 32 S fumeus were collected on north-facing slopes whereas only 4 S cinereus and 10 S fumeus were cap-tured on south-facing slopes (Table 1) Capture rates (captures per 100TN) were significantly different between sites for both masked shrews(F29 = 354 p = 005) and smoky shrews (F29 = 410 p = 004) Meancapture rates of masked shrews were highest at north-facing plots (267plusmn 003) intermediate at streamside plots (157 plusmn 004) and lowest atsouth-facing plots (063 plusmn 003 Table 1) Only 33 of masked shrewcaptures occurred at each south-facing plot compared to 138 at eachnorth-facing plot and 81 at each streamside plot Mean capture ratesof smoky shrews were similar at mesic north-facing (503 plusmn 059) andstreamside plots (495 plusmn 124) but over 3 times less at xeric south-facingplots (157 plusmn 059 Table 1) Each south-facing plot yielded 32 ofcaptures of smoky shrews compared to 102 at each north-facing plotand 100 at each streamside plot

Smoky shrews (n = 105) were strongly positively correlated withpercentage litter moisture and heavily decomposed logs and signifi-cantly negatively correlated with invertebrate biomass (Table 2) There

Table 1 Summary of shrew captures in each habitat type in the study area in Autumn 1996and spring and summer 1997

Habitat Plots TN Sorex cinereus S fumeus Combined

n capture relative n capture relative n capture relativerate abundance rate abundance rate abundance

Streamside 6 1272 20 157 81 63 495 100 83 653 947North-facing slope 3 636 17 267 138 32 503 102 49 770 1119South-facing slope 3 636 4 063 33 10 157 32 14 220 320

MP Brannon2002 303

was no significant correlation between smoky shrews and maximumtemperatures Masked shrews (n = 41) showed no significant correlationwith any habitat variable (Table 2) although correlations with littermoisture and percentage of rotten logs were positive

DISCUSSION

Although many factors may influence distributions of soricids(Pagels et al 1994 Parmley and Harley 1995) environmental moisturemay be of the greatest importance to shrews because of their extremelyhigh water turnover rates (Getz 1961 Wrigley et al 1979) Maskedshrews and smoky shrews are most abundant in mesic forests and nearstanding water but are uncommon if not absent in dry habitats(Cudmore and Whitaker 1984 Getz 1961 Laerm et al 1999 Owen1984) North-facing slopes are generally cooler and moister than south-facing slopes (Wales 1972) and consequently yield greater numbers ofshrews (Laerm et al 1999 McComb and Rumsey 1982)

Greater sunlight exposure on south-facing slopes results in highertemperatures and evaporative drying of the leaf litter (Wales 1972) In thisstudy maximum temperatures increased gradually from north- to south-facing slopes but environmental moisture did not decrease along thisgradient Although temperatures were higher at streamside sites leaf littermoisture content in these areas was similar to that of cooler north-facingslopes and consequently had similar capture rates of shrews Streamsidehabitats often represent isolated zones of moisture in otherwise dryenvironments (Laerm et al 1999) because leaf litter in these areas is oftennot far above the water table (Sites 1978) Habitats such as riparian areasthat remain moist regardless of temperature provide favorable environ-mental conditions that facilitate the active movement of shrews

The primary importance of structural components of the microhabitatto shrews is their effect on humidity (Getz 1961 McComb and Rumsey1982 Pagels et al 1994 Parmley and Harley 1995) The percentage oflogs that were heavily decomposed was high on north-facing slopes andin streamside habitats in the study area but extremely low on south-facingslopes of the same stand age presumably because logs decompose moreslowly under xeric conditions (Abbott and Crossley 1982) Althoughmost logs can function as protective cover and as foraging sites (Loeb

Table 2 Correlation coefficients (r) between measured habitat variables and shrews in thestudy area in autumn 1996 and spring and summer 1997

Habitat variable Sorex cinereus (n = 41) S fumeus (n = 105) Combined (n = 146)

litter moisture 0443 0559 0490 Average maximum temperature -0254 -0484 -0481Invertebrate biomass -0267 -0873 -0785 heavily decomposed logs 0393 0553 0593

p le 005

Southeastern Naturalist Vol 1 No 3304

1993) those in the advanced stages of decomposition may be optimalmicrohabitats for shrews (Brannon 2000) because they contain highconcentrations of water (Maser et al 1979) and can serve as moisturerefugia during dry periods Shading from the dense understory of rhodo-dendron likely also contributed to favorable environmental moistureconditions for S cinereus and S fumeus on north-facing slopes and alongstreams (Laerm et al 1995 Pagels et al 1994)

Environmental moisture is also important to the distribution ofshrews because it may affect the availability of invertebrate prey (Gistand Crossley 1975 Kirkland 1991 Parmley and Harley 1995 Wrigleyet al 1979) Although in this study no significant differences in inverte-brate biomass existed between mesic and xeric habitats in moist envi-ronments prey may be more accessible because mobility of the shrews isenhanced (McCay 1996) If the litter is wet shrews may increase theiractivity on the forest floor where invertebrates are abundant but if thelitter is dry they may be restricted to small isolated patches of moisturewhere availability of prey may be reduced (Jaeger 1980) Althoughother prey items such as salamanders are abundant in mesic habitats ofthe southern Appalachian Mountains (Petranka et al 1993) they consti-tute only a very small percentage of the diets of most shrews (Hamilton1930) Greater numbers of shrews in mesic environments probably havemore to do with better environmental conditions for foraging associatedwith their ecophysiology than to an increased food supply

For small mammals associated with mesic environments speciespresence and abundance may be most predictable at a habitat scale witha high degree of resolution (Orrock et al 2000) The southern Appala-chian Mountains are comprised of a complex mosaic of habitats (Laermet al 1995 Laerm et al 1999 Pagels et al 1994 Whittaker 1956)Shrew species diversity is greatest at higher elevation sites where envi-ronmental conditions resemble those of more northern forests (Laerm etal 1999 Pagels et al 1994) Habitats that provide shaded moist condi-tions with complex structural microhabitats are especially important tothe distributions of shrews (Cudmore and Whitaker 1984 Ford et al1997 Kirkland 1991 Laerm et al 1999 Mitchell et al 1997 Parmleyand Harley 1995) My results are consistent with those of others (Getz1961 Spencer and Pettus 1966 Wrigley et al 1979) who found that theprimary factor governing the distributions of shrews is environmentalmoisture Moreover this study suggests that the distribution of Scinereus and S fumeus is strongly influenced by heterogeneity of mois-ture across landscapes such as between north- and south-facing slopes

ACKNOWLEDGMENTS

This study was supported in part by Appalachian State University GraduateSchool Graduate Student Association Senate and Biology Graduate StudentAssociation I thank R W Van Devender G L Walker and D B Meikle for

MP Brannon2002 305

their guidance during this project and J K H Brannon and many others forfield assistance I also thank J F Merritt J F Pagels and an anonymousreviewer for their suggestions for revision of the original manuscript

LITERATURE CITED

ABBOTT DT and DA CROSSLEY 1982 Woody litter decompositionfollowing clear-cutting Ecology 6335-42

BRANNON MP 2000 Niche relationships of two syntopic species of shrewsSorex fumeus and S cinereus in the southern Appalachian MountainsJournal of Mammalogy 811053-1061

CUDMORE WW and JO WHITAKER Jr 1984 The distribution of thesmoky shrew Sorex fumeus and the pygmy shrew Microsorex hoyi inIndiana with notes on the distribution of other shrews Proceedings of theIndiana Academy of Sciences 93469-474

DUESER RD and HH SHUGART 1978 Microhabitats in a forest-floorsmall mammal fauna Ecology 5989-98

FORD WM J LAERM and KG BARKER 1997 Soricid response to foreststand age in southern Appalachian cove hardwood communities ForestEcology and Management 91175-181

GETZ LL 1961 Factors influencing the local distribution of shrews Ameri-can Midland Naturalist 6567-88

GIST CS and DA CROSSLEY 1975 The litter arthropod community in asouthern Appalachian hardwood forest numbers biomass and mineral ele-ment content American Midland Naturalist 93107-122

HAMILTON WJ Jr 1930 The food of the Soricidae Journal of Mammalogy1126-39

JAEGER RG 1980 Microhabitats of a terrestrial forest salamander Copeia1980265-268

KIRKLAND GL Jr 1991 Competition and coexistence in shrews(Insectivora Soricidae) Pp 15-22 In JS Findley and TL Yates (Eds)The Biology of the Soricidae Special Publication of the Museum of South-western Biology University of New Mexico Albuquerque NM 91 pp

KIRKLAND GL Jr and PK SHEPPARD 1994 Proposed standard proto-col for sampling small mammal communities Pp 277-283 In JF MerrittGL Kirkland Jr and RK Rose (Eds) Advances in the Biology of ShrewsSpecial Publication of the Carnegie Museum of Natural History No 18Pittsburgh PA 458 pp

LAERM J E BROWN MA MENZEL A WOTJALIK WM FORD and MSTRAYER 1995 The masked shrew Sorex cinereus (Insectivora Soricidae)and red-backed vole Clethrionomys gapperi (Rodentia Muridae) in the BlueRidge province of South Carolina Brimleyana 2215-21

LAERM J WM FORD TS MCCAY MA MENZEL LT LEPARDO andJL BOONE 1999 Soricid communities in the southern Appalachians Pp177-193 In RP Eckerlin (Ed) Proceedings of the Appalachian Biogeogra-phy Symposium Special Publication of the Virginia Museum of NaturalHistory No 7 Martinsville VA 258 pp

LOEB SC 1993 The role of coarse woody debris in the ecology of southeast-ern mammals Pp 108-118 In JW McMinn and DA Crossley (Eds)Biodiversity and Coarse Woody Debris in Southern Forest USDA ForestService General Technical Report SE-94

Southeastern Naturalist Vol 1 No 3306

MASER C RG ANDERSON K CROMACK Jr JT WILLIAMS and REMARTIN 1979 Dead and down woody material Pp 78-95 In JW Tho-mas (Ed) Wildlife Habitats in Managed Forests the Blue Mountains ofOregon and Washington USDA Forest Service Handbook 553 512 pp

MATLACK GR 1993 Microenvironment variation within and among forestedge sites in the eastern United States Biological Conservation 66185-194

McCAY TS 1996 Response of masked shrews (Sorex cinereus) to precipita-tion in irrigated and nonirrigated forests American Midland Naturalist135178-180

McCAY TS J LAERM MA MENZEL and WM FORD 1998 Methodsused to survey shrews (Insectivora Soricidae) and the importance of forest-floor structure Brimleyana 25110-119

McCOMB WC and RL RUMSEY 1982 Response of small mammals toforest floor clearings created by herbicides in the central AppalachiansBrimleyana 8121-134

MITCHELL JC SC RINEHART JF PAGELS KA BUHLMANN andCA PAGUE 1997 Factors influencing amphibian and small mammalassemblages in central Appalachian forests Forest Ecology and Manage-ment 9665-76

ORROCK JL JF PAGELS WJ McSHEA and EK HARPER 2000 Pre-dicting presence and abundance of a small mammal species the effect ofscale and resolution Ecological Applications 101356-1366

OWEN JG 1984 Sorex fumeus Mammalian Species 2151-8PAGELS JF KL UTHUS and HE DUVAL 1994 The masked shrew

Sorex cinereus in a relictual habitat of the southern Appalachian mountainspp 103-109 In JF Merritt GL Kirkland Jr and RK Rose (Eds) Ad-vances in the Biology of Shrews Special Publication of the Carnegie Mu-seum of Natural History No 18 Pittsburgh PA 458 pp

PARMLEY D and D HARLEY 1995 The relative seasonal abundance ofshrews in two central Georgia deciduous woodlots Georgia Journal ofScience 5383-88

PETRANKA JW ME ELDRIDGE and KE HALEY 1993 Effects oftimber harvesting on southern Appalachian salamanders Conservation Bi-ology 7363-370

SITES JW 1978 The foraging strategy of the dusky salamanderDesmognathus fuscus (Amphibia Urodela Plethodontidae) an empiricalapproach to predation theory Journal of Herpetology 12373-383

SPENCER AW and D PETTUS 1966 Habitat preferences of five sympatricspecies of long-tailed shrews Ecology 47677-683

WALES BA 1972 Vegetation analysis of north and south edges in a matureoak-hickory forest Ecological Monographs 42451-471

WHITTAKER RH 1956 Vegetation of the Great Smoky Mountains Ecologi-cal Monographs 261-80

WRIGLEY RE JE DUBOIS and HWR COPELAND 1979 Habitat abun-dance and the distribution of six species of shrews in Manitoba Journal ofMammalogy 60505-520

ZAR JH 1984 Biostatistical Analysis 2nd ed Prentice-Hall IncEnglewood-Cliffs NJ 718 pp

Southeastern Naturalist Vol 1 No 3302

RESULTS

Sites differed significantly in both maximum temperatures (F2117 =2209 p lt 0001) and percentage litter moisture content (F2237 = 7934 plt 0001) South-facing slopes had the highest mean (plusmn 1 SE) maximumtemperatures (266 plusmn 06 EC) whereas north-facing slopes had thelowest (209 plusmn 05 EC) and those at streamsides were intermediate (228plusmn 06 EC) Mean percentage litter moisture was lowest on south-facingslopes (4727 plusmn 153) but similar between north-facing slopes (5671plusmn 180) and streamside habitats (5997 plusmn 109)

Percentage volume of heavily decomposed logs also differed signifi-cantly between sites (F29 = 511 p = 002) Mean percentage volume washigh on north-facing slopes (9692 plusmn 062) and at streamside sites (8878 plusmn499) but low on south-facing slopes (3037 plusmn 2148) Two of the threesouth-facing plots had no logs that were classified as heavily decomposed

No significant differences in invertebrate biomass were observedbetween sites (F29 = 064 p = 054) Mean biomass was 20779 plusmn 3458g at streamside sites 23741 plusmn 1697 g on north-facing slopes and31368 plusmn 2987 g on south-facing slopes

A total of 17 S cinereus and 32 S fumeus were collected on north-facing slopes whereas only 4 S cinereus and 10 S fumeus were cap-tured on south-facing slopes (Table 1) Capture rates (captures per 100TN) were significantly different between sites for both masked shrews(F29 = 354 p = 005) and smoky shrews (F29 = 410 p = 004) Meancapture rates of masked shrews were highest at north-facing plots (267plusmn 003) intermediate at streamside plots (157 plusmn 004) and lowest atsouth-facing plots (063 plusmn 003 Table 1) Only 33 of masked shrewcaptures occurred at each south-facing plot compared to 138 at eachnorth-facing plot and 81 at each streamside plot Mean capture ratesof smoky shrews were similar at mesic north-facing (503 plusmn 059) andstreamside plots (495 plusmn 124) but over 3 times less at xeric south-facingplots (157 plusmn 059 Table 1) Each south-facing plot yielded 32 ofcaptures of smoky shrews compared to 102 at each north-facing plotand 100 at each streamside plot

Smoky shrews (n = 105) were strongly positively correlated withpercentage litter moisture and heavily decomposed logs and signifi-cantly negatively correlated with invertebrate biomass (Table 2) There

Table 1 Summary of shrew captures in each habitat type in the study area in Autumn 1996and spring and summer 1997

Habitat Plots TN Sorex cinereus S fumeus Combined

n capture relative n capture relative n capture relativerate abundance rate abundance rate abundance

Streamside 6 1272 20 157 81 63 495 100 83 653 947North-facing slope 3 636 17 267 138 32 503 102 49 770 1119South-facing slope 3 636 4 063 33 10 157 32 14 220 320

MP Brannon2002 303

was no significant correlation between smoky shrews and maximumtemperatures Masked shrews (n = 41) showed no significant correlationwith any habitat variable (Table 2) although correlations with littermoisture and percentage of rotten logs were positive

DISCUSSION

Although many factors may influence distributions of soricids(Pagels et al 1994 Parmley and Harley 1995) environmental moisturemay be of the greatest importance to shrews because of their extremelyhigh water turnover rates (Getz 1961 Wrigley et al 1979) Maskedshrews and smoky shrews are most abundant in mesic forests and nearstanding water but are uncommon if not absent in dry habitats(Cudmore and Whitaker 1984 Getz 1961 Laerm et al 1999 Owen1984) North-facing slopes are generally cooler and moister than south-facing slopes (Wales 1972) and consequently yield greater numbers ofshrews (Laerm et al 1999 McComb and Rumsey 1982)

Greater sunlight exposure on south-facing slopes results in highertemperatures and evaporative drying of the leaf litter (Wales 1972) In thisstudy maximum temperatures increased gradually from north- to south-facing slopes but environmental moisture did not decrease along thisgradient Although temperatures were higher at streamside sites leaf littermoisture content in these areas was similar to that of cooler north-facingslopes and consequently had similar capture rates of shrews Streamsidehabitats often represent isolated zones of moisture in otherwise dryenvironments (Laerm et al 1999) because leaf litter in these areas is oftennot far above the water table (Sites 1978) Habitats such as riparian areasthat remain moist regardless of temperature provide favorable environ-mental conditions that facilitate the active movement of shrews

The primary importance of structural components of the microhabitatto shrews is their effect on humidity (Getz 1961 McComb and Rumsey1982 Pagels et al 1994 Parmley and Harley 1995) The percentage oflogs that were heavily decomposed was high on north-facing slopes andin streamside habitats in the study area but extremely low on south-facingslopes of the same stand age presumably because logs decompose moreslowly under xeric conditions (Abbott and Crossley 1982) Althoughmost logs can function as protective cover and as foraging sites (Loeb

Table 2 Correlation coefficients (r) between measured habitat variables and shrews in thestudy area in autumn 1996 and spring and summer 1997

Habitat variable Sorex cinereus (n = 41) S fumeus (n = 105) Combined (n = 146)

litter moisture 0443 0559 0490 Average maximum temperature -0254 -0484 -0481Invertebrate biomass -0267 -0873 -0785 heavily decomposed logs 0393 0553 0593

p le 005

Southeastern Naturalist Vol 1 No 3304

1993) those in the advanced stages of decomposition may be optimalmicrohabitats for shrews (Brannon 2000) because they contain highconcentrations of water (Maser et al 1979) and can serve as moisturerefugia during dry periods Shading from the dense understory of rhodo-dendron likely also contributed to favorable environmental moistureconditions for S cinereus and S fumeus on north-facing slopes and alongstreams (Laerm et al 1995 Pagels et al 1994)

Environmental moisture is also important to the distribution ofshrews because it may affect the availability of invertebrate prey (Gistand Crossley 1975 Kirkland 1991 Parmley and Harley 1995 Wrigleyet al 1979) Although in this study no significant differences in inverte-brate biomass existed between mesic and xeric habitats in moist envi-ronments prey may be more accessible because mobility of the shrews isenhanced (McCay 1996) If the litter is wet shrews may increase theiractivity on the forest floor where invertebrates are abundant but if thelitter is dry they may be restricted to small isolated patches of moisturewhere availability of prey may be reduced (Jaeger 1980) Althoughother prey items such as salamanders are abundant in mesic habitats ofthe southern Appalachian Mountains (Petranka et al 1993) they consti-tute only a very small percentage of the diets of most shrews (Hamilton1930) Greater numbers of shrews in mesic environments probably havemore to do with better environmental conditions for foraging associatedwith their ecophysiology than to an increased food supply

For small mammals associated with mesic environments speciespresence and abundance may be most predictable at a habitat scale witha high degree of resolution (Orrock et al 2000) The southern Appala-chian Mountains are comprised of a complex mosaic of habitats (Laermet al 1995 Laerm et al 1999 Pagels et al 1994 Whittaker 1956)Shrew species diversity is greatest at higher elevation sites where envi-ronmental conditions resemble those of more northern forests (Laerm etal 1999 Pagels et al 1994) Habitats that provide shaded moist condi-tions with complex structural microhabitats are especially important tothe distributions of shrews (Cudmore and Whitaker 1984 Ford et al1997 Kirkland 1991 Laerm et al 1999 Mitchell et al 1997 Parmleyand Harley 1995) My results are consistent with those of others (Getz1961 Spencer and Pettus 1966 Wrigley et al 1979) who found that theprimary factor governing the distributions of shrews is environmentalmoisture Moreover this study suggests that the distribution of Scinereus and S fumeus is strongly influenced by heterogeneity of mois-ture across landscapes such as between north- and south-facing slopes

ACKNOWLEDGMENTS

This study was supported in part by Appalachian State University GraduateSchool Graduate Student Association Senate and Biology Graduate StudentAssociation I thank R W Van Devender G L Walker and D B Meikle for

MP Brannon2002 305

their guidance during this project and J K H Brannon and many others forfield assistance I also thank J F Merritt J F Pagels and an anonymousreviewer for their suggestions for revision of the original manuscript

LITERATURE CITED

ABBOTT DT and DA CROSSLEY 1982 Woody litter decompositionfollowing clear-cutting Ecology 6335-42

BRANNON MP 2000 Niche relationships of two syntopic species of shrewsSorex fumeus and S cinereus in the southern Appalachian MountainsJournal of Mammalogy 811053-1061

CUDMORE WW and JO WHITAKER Jr 1984 The distribution of thesmoky shrew Sorex fumeus and the pygmy shrew Microsorex hoyi inIndiana with notes on the distribution of other shrews Proceedings of theIndiana Academy of Sciences 93469-474

DUESER RD and HH SHUGART 1978 Microhabitats in a forest-floorsmall mammal fauna Ecology 5989-98

FORD WM J LAERM and KG BARKER 1997 Soricid response to foreststand age in southern Appalachian cove hardwood communities ForestEcology and Management 91175-181

GETZ LL 1961 Factors influencing the local distribution of shrews Ameri-can Midland Naturalist 6567-88

GIST CS and DA CROSSLEY 1975 The litter arthropod community in asouthern Appalachian hardwood forest numbers biomass and mineral ele-ment content American Midland Naturalist 93107-122

HAMILTON WJ Jr 1930 The food of the Soricidae Journal of Mammalogy1126-39

JAEGER RG 1980 Microhabitats of a terrestrial forest salamander Copeia1980265-268

KIRKLAND GL Jr 1991 Competition and coexistence in shrews(Insectivora Soricidae) Pp 15-22 In JS Findley and TL Yates (Eds)The Biology of the Soricidae Special Publication of the Museum of South-western Biology University of New Mexico Albuquerque NM 91 pp

KIRKLAND GL Jr and PK SHEPPARD 1994 Proposed standard proto-col for sampling small mammal communities Pp 277-283 In JF MerrittGL Kirkland Jr and RK Rose (Eds) Advances in the Biology of ShrewsSpecial Publication of the Carnegie Museum of Natural History No 18Pittsburgh PA 458 pp

LAERM J E BROWN MA MENZEL A WOTJALIK WM FORD and MSTRAYER 1995 The masked shrew Sorex cinereus (Insectivora Soricidae)and red-backed vole Clethrionomys gapperi (Rodentia Muridae) in the BlueRidge province of South Carolina Brimleyana 2215-21

LAERM J WM FORD TS MCCAY MA MENZEL LT LEPARDO andJL BOONE 1999 Soricid communities in the southern Appalachians Pp177-193 In RP Eckerlin (Ed) Proceedings of the Appalachian Biogeogra-phy Symposium Special Publication of the Virginia Museum of NaturalHistory No 7 Martinsville VA 258 pp

LOEB SC 1993 The role of coarse woody debris in the ecology of southeast-ern mammals Pp 108-118 In JW McMinn and DA Crossley (Eds)Biodiversity and Coarse Woody Debris in Southern Forest USDA ForestService General Technical Report SE-94

Southeastern Naturalist Vol 1 No 3306

MASER C RG ANDERSON K CROMACK Jr JT WILLIAMS and REMARTIN 1979 Dead and down woody material Pp 78-95 In JW Tho-mas (Ed) Wildlife Habitats in Managed Forests the Blue Mountains ofOregon and Washington USDA Forest Service Handbook 553 512 pp

MATLACK GR 1993 Microenvironment variation within and among forestedge sites in the eastern United States Biological Conservation 66185-194

McCAY TS 1996 Response of masked shrews (Sorex cinereus) to precipita-tion in irrigated and nonirrigated forests American Midland Naturalist135178-180

McCAY TS J LAERM MA MENZEL and WM FORD 1998 Methodsused to survey shrews (Insectivora Soricidae) and the importance of forest-floor structure Brimleyana 25110-119

McCOMB WC and RL RUMSEY 1982 Response of small mammals toforest floor clearings created by herbicides in the central AppalachiansBrimleyana 8121-134

MITCHELL JC SC RINEHART JF PAGELS KA BUHLMANN andCA PAGUE 1997 Factors influencing amphibian and small mammalassemblages in central Appalachian forests Forest Ecology and Manage-ment 9665-76

ORROCK JL JF PAGELS WJ McSHEA and EK HARPER 2000 Pre-dicting presence and abundance of a small mammal species the effect ofscale and resolution Ecological Applications 101356-1366

OWEN JG 1984 Sorex fumeus Mammalian Species 2151-8PAGELS JF KL UTHUS and HE DUVAL 1994 The masked shrew

Sorex cinereus in a relictual habitat of the southern Appalachian mountainspp 103-109 In JF Merritt GL Kirkland Jr and RK Rose (Eds) Ad-vances in the Biology of Shrews Special Publication of the Carnegie Mu-seum of Natural History No 18 Pittsburgh PA 458 pp

PARMLEY D and D HARLEY 1995 The relative seasonal abundance ofshrews in two central Georgia deciduous woodlots Georgia Journal ofScience 5383-88

PETRANKA JW ME ELDRIDGE and KE HALEY 1993 Effects oftimber harvesting on southern Appalachian salamanders Conservation Bi-ology 7363-370

SITES JW 1978 The foraging strategy of the dusky salamanderDesmognathus fuscus (Amphibia Urodela Plethodontidae) an empiricalapproach to predation theory Journal of Herpetology 12373-383

SPENCER AW and D PETTUS 1966 Habitat preferences of five sympatricspecies of long-tailed shrews Ecology 47677-683

WALES BA 1972 Vegetation analysis of north and south edges in a matureoak-hickory forest Ecological Monographs 42451-471

WHITTAKER RH 1956 Vegetation of the Great Smoky Mountains Ecologi-cal Monographs 261-80

WRIGLEY RE JE DUBOIS and HWR COPELAND 1979 Habitat abun-dance and the distribution of six species of shrews in Manitoba Journal ofMammalogy 60505-520

ZAR JH 1984 Biostatistical Analysis 2nd ed Prentice-Hall IncEnglewood-Cliffs NJ 718 pp

MP Brannon2002 303

was no significant correlation between smoky shrews and maximumtemperatures Masked shrews (n = 41) showed no significant correlationwith any habitat variable (Table 2) although correlations with littermoisture and percentage of rotten logs were positive

DISCUSSION

Although many factors may influence distributions of soricids(Pagels et al 1994 Parmley and Harley 1995) environmental moisturemay be of the greatest importance to shrews because of their extremelyhigh water turnover rates (Getz 1961 Wrigley et al 1979) Maskedshrews and smoky shrews are most abundant in mesic forests and nearstanding water but are uncommon if not absent in dry habitats(Cudmore and Whitaker 1984 Getz 1961 Laerm et al 1999 Owen1984) North-facing slopes are generally cooler and moister than south-facing slopes (Wales 1972) and consequently yield greater numbers ofshrews (Laerm et al 1999 McComb and Rumsey 1982)

Greater sunlight exposure on south-facing slopes results in highertemperatures and evaporative drying of the leaf litter (Wales 1972) In thisstudy maximum temperatures increased gradually from north- to south-facing slopes but environmental moisture did not decrease along thisgradient Although temperatures were higher at streamside sites leaf littermoisture content in these areas was similar to that of cooler north-facingslopes and consequently had similar capture rates of shrews Streamsidehabitats often represent isolated zones of moisture in otherwise dryenvironments (Laerm et al 1999) because leaf litter in these areas is oftennot far above the water table (Sites 1978) Habitats such as riparian areasthat remain moist regardless of temperature provide favorable environ-mental conditions that facilitate the active movement of shrews

The primary importance of structural components of the microhabitatto shrews is their effect on humidity (Getz 1961 McComb and Rumsey1982 Pagels et al 1994 Parmley and Harley 1995) The percentage oflogs that were heavily decomposed was high on north-facing slopes andin streamside habitats in the study area but extremely low on south-facingslopes of the same stand age presumably because logs decompose moreslowly under xeric conditions (Abbott and Crossley 1982) Althoughmost logs can function as protective cover and as foraging sites (Loeb

Table 2 Correlation coefficients (r) between measured habitat variables and shrews in thestudy area in autumn 1996 and spring and summer 1997

Habitat variable Sorex cinereus (n = 41) S fumeus (n = 105) Combined (n = 146)

litter moisture 0443 0559 0490 Average maximum temperature -0254 -0484 -0481Invertebrate biomass -0267 -0873 -0785 heavily decomposed logs 0393 0553 0593

p le 005

Southeastern Naturalist Vol 1 No 3304

1993) those in the advanced stages of decomposition may be optimalmicrohabitats for shrews (Brannon 2000) because they contain highconcentrations of water (Maser et al 1979) and can serve as moisturerefugia during dry periods Shading from the dense understory of rhodo-dendron likely also contributed to favorable environmental moistureconditions for S cinereus and S fumeus on north-facing slopes and alongstreams (Laerm et al 1995 Pagels et al 1994)

Environmental moisture is also important to the distribution ofshrews because it may affect the availability of invertebrate prey (Gistand Crossley 1975 Kirkland 1991 Parmley and Harley 1995 Wrigleyet al 1979) Although in this study no significant differences in inverte-brate biomass existed between mesic and xeric habitats in moist envi-ronments prey may be more accessible because mobility of the shrews isenhanced (McCay 1996) If the litter is wet shrews may increase theiractivity on the forest floor where invertebrates are abundant but if thelitter is dry they may be restricted to small isolated patches of moisturewhere availability of prey may be reduced (Jaeger 1980) Althoughother prey items such as salamanders are abundant in mesic habitats ofthe southern Appalachian Mountains (Petranka et al 1993) they consti-tute only a very small percentage of the diets of most shrews (Hamilton1930) Greater numbers of shrews in mesic environments probably havemore to do with better environmental conditions for foraging associatedwith their ecophysiology than to an increased food supply

For small mammals associated with mesic environments speciespresence and abundance may be most predictable at a habitat scale witha high degree of resolution (Orrock et al 2000) The southern Appala-chian Mountains are comprised of a complex mosaic of habitats (Laermet al 1995 Laerm et al 1999 Pagels et al 1994 Whittaker 1956)Shrew species diversity is greatest at higher elevation sites where envi-ronmental conditions resemble those of more northern forests (Laerm etal 1999 Pagels et al 1994) Habitats that provide shaded moist condi-tions with complex structural microhabitats are especially important tothe distributions of shrews (Cudmore and Whitaker 1984 Ford et al1997 Kirkland 1991 Laerm et al 1999 Mitchell et al 1997 Parmleyand Harley 1995) My results are consistent with those of others (Getz1961 Spencer and Pettus 1966 Wrigley et al 1979) who found that theprimary factor governing the distributions of shrews is environmentalmoisture Moreover this study suggests that the distribution of Scinereus and S fumeus is strongly influenced by heterogeneity of mois-ture across landscapes such as between north- and south-facing slopes

ACKNOWLEDGMENTS

This study was supported in part by Appalachian State University GraduateSchool Graduate Student Association Senate and Biology Graduate StudentAssociation I thank R W Van Devender G L Walker and D B Meikle for

MP Brannon2002 305

their guidance during this project and J K H Brannon and many others forfield assistance I also thank J F Merritt J F Pagels and an anonymousreviewer for their suggestions for revision of the original manuscript

LITERATURE CITED

ABBOTT DT and DA CROSSLEY 1982 Woody litter decompositionfollowing clear-cutting Ecology 6335-42

BRANNON MP 2000 Niche relationships of two syntopic species of shrewsSorex fumeus and S cinereus in the southern Appalachian MountainsJournal of Mammalogy 811053-1061

CUDMORE WW and JO WHITAKER Jr 1984 The distribution of thesmoky shrew Sorex fumeus and the pygmy shrew Microsorex hoyi inIndiana with notes on the distribution of other shrews Proceedings of theIndiana Academy of Sciences 93469-474

DUESER RD and HH SHUGART 1978 Microhabitats in a forest-floorsmall mammal fauna Ecology 5989-98

FORD WM J LAERM and KG BARKER 1997 Soricid response to foreststand age in southern Appalachian cove hardwood communities ForestEcology and Management 91175-181

GETZ LL 1961 Factors influencing the local distribution of shrews Ameri-can Midland Naturalist 6567-88

GIST CS and DA CROSSLEY 1975 The litter arthropod community in asouthern Appalachian hardwood forest numbers biomass and mineral ele-ment content American Midland Naturalist 93107-122

HAMILTON WJ Jr 1930 The food of the Soricidae Journal of Mammalogy1126-39

JAEGER RG 1980 Microhabitats of a terrestrial forest salamander Copeia1980265-268

KIRKLAND GL Jr 1991 Competition and coexistence in shrews(Insectivora Soricidae) Pp 15-22 In JS Findley and TL Yates (Eds)The Biology of the Soricidae Special Publication of the Museum of South-western Biology University of New Mexico Albuquerque NM 91 pp

KIRKLAND GL Jr and PK SHEPPARD 1994 Proposed standard proto-col for sampling small mammal communities Pp 277-283 In JF MerrittGL Kirkland Jr and RK Rose (Eds) Advances in the Biology of ShrewsSpecial Publication of the Carnegie Museum of Natural History No 18Pittsburgh PA 458 pp

LAERM J E BROWN MA MENZEL A WOTJALIK WM FORD and MSTRAYER 1995 The masked shrew Sorex cinereus (Insectivora Soricidae)and red-backed vole Clethrionomys gapperi (Rodentia Muridae) in the BlueRidge province of South Carolina Brimleyana 2215-21

LAERM J WM FORD TS MCCAY MA MENZEL LT LEPARDO andJL BOONE 1999 Soricid communities in the southern Appalachians Pp177-193 In RP Eckerlin (Ed) Proceedings of the Appalachian Biogeogra-phy Symposium Special Publication of the Virginia Museum of NaturalHistory No 7 Martinsville VA 258 pp

LOEB SC 1993 The role of coarse woody debris in the ecology of southeast-ern mammals Pp 108-118 In JW McMinn and DA Crossley (Eds)Biodiversity and Coarse Woody Debris in Southern Forest USDA ForestService General Technical Report SE-94

Southeastern Naturalist Vol 1 No 3306

MASER C RG ANDERSON K CROMACK Jr JT WILLIAMS and REMARTIN 1979 Dead and down woody material Pp 78-95 In JW Tho-mas (Ed) Wildlife Habitats in Managed Forests the Blue Mountains ofOregon and Washington USDA Forest Service Handbook 553 512 pp

MATLACK GR 1993 Microenvironment variation within and among forestedge sites in the eastern United States Biological Conservation 66185-194

McCAY TS 1996 Response of masked shrews (Sorex cinereus) to precipita-tion in irrigated and nonirrigated forests American Midland Naturalist135178-180

McCAY TS J LAERM MA MENZEL and WM FORD 1998 Methodsused to survey shrews (Insectivora Soricidae) and the importance of forest-floor structure Brimleyana 25110-119

McCOMB WC and RL RUMSEY 1982 Response of small mammals toforest floor clearings created by herbicides in the central AppalachiansBrimleyana 8121-134

MITCHELL JC SC RINEHART JF PAGELS KA BUHLMANN andCA PAGUE 1997 Factors influencing amphibian and small mammalassemblages in central Appalachian forests Forest Ecology and Manage-ment 9665-76

ORROCK JL JF PAGELS WJ McSHEA and EK HARPER 2000 Pre-dicting presence and abundance of a small mammal species the effect ofscale and resolution Ecological Applications 101356-1366

OWEN JG 1984 Sorex fumeus Mammalian Species 2151-8PAGELS JF KL UTHUS and HE DUVAL 1994 The masked shrew

Sorex cinereus in a relictual habitat of the southern Appalachian mountainspp 103-109 In JF Merritt GL Kirkland Jr and RK Rose (Eds) Ad-vances in the Biology of Shrews Special Publication of the Carnegie Mu-seum of Natural History No 18 Pittsburgh PA 458 pp

PARMLEY D and D HARLEY 1995 The relative seasonal abundance ofshrews in two central Georgia deciduous woodlots Georgia Journal ofScience 5383-88

PETRANKA JW ME ELDRIDGE and KE HALEY 1993 Effects oftimber harvesting on southern Appalachian salamanders Conservation Bi-ology 7363-370

SITES JW 1978 The foraging strategy of the dusky salamanderDesmognathus fuscus (Amphibia Urodela Plethodontidae) an empiricalapproach to predation theory Journal of Herpetology 12373-383

SPENCER AW and D PETTUS 1966 Habitat preferences of five sympatricspecies of long-tailed shrews Ecology 47677-683

WALES BA 1972 Vegetation analysis of north and south edges in a matureoak-hickory forest Ecological Monographs 42451-471

WHITTAKER RH 1956 Vegetation of the Great Smoky Mountains Ecologi-cal Monographs 261-80

WRIGLEY RE JE DUBOIS and HWR COPELAND 1979 Habitat abun-dance and the distribution of six species of shrews in Manitoba Journal ofMammalogy 60505-520

ZAR JH 1984 Biostatistical Analysis 2nd ed Prentice-Hall IncEnglewood-Cliffs NJ 718 pp

Southeastern Naturalist Vol 1 No 3304

1993) those in the advanced stages of decomposition may be optimalmicrohabitats for shrews (Brannon 2000) because they contain highconcentrations of water (Maser et al 1979) and can serve as moisturerefugia during dry periods Shading from the dense understory of rhodo-dendron likely also contributed to favorable environmental moistureconditions for S cinereus and S fumeus on north-facing slopes and alongstreams (Laerm et al 1995 Pagels et al 1994)

Environmental moisture is also important to the distribution ofshrews because it may affect the availability of invertebrate prey (Gistand Crossley 1975 Kirkland 1991 Parmley and Harley 1995 Wrigleyet al 1979) Although in this study no significant differences in inverte-brate biomass existed between mesic and xeric habitats in moist envi-ronments prey may be more accessible because mobility of the shrews isenhanced (McCay 1996) If the litter is wet shrews may increase theiractivity on the forest floor where invertebrates are abundant but if thelitter is dry they may be restricted to small isolated patches of moisturewhere availability of prey may be reduced (Jaeger 1980) Althoughother prey items such as salamanders are abundant in mesic habitats ofthe southern Appalachian Mountains (Petranka et al 1993) they consti-tute only a very small percentage of the diets of most shrews (Hamilton1930) Greater numbers of shrews in mesic environments probably havemore to do with better environmental conditions for foraging associatedwith their ecophysiology than to an increased food supply

For small mammals associated with mesic environments speciespresence and abundance may be most predictable at a habitat scale witha high degree of resolution (Orrock et al 2000) The southern Appala-chian Mountains are comprised of a complex mosaic of habitats (Laermet al 1995 Laerm et al 1999 Pagels et al 1994 Whittaker 1956)Shrew species diversity is greatest at higher elevation sites where envi-ronmental conditions resemble those of more northern forests (Laerm etal 1999 Pagels et al 1994) Habitats that provide shaded moist condi-tions with complex structural microhabitats are especially important tothe distributions of shrews (Cudmore and Whitaker 1984 Ford et al1997 Kirkland 1991 Laerm et al 1999 Mitchell et al 1997 Parmleyand Harley 1995) My results are consistent with those of others (Getz1961 Spencer and Pettus 1966 Wrigley et al 1979) who found that theprimary factor governing the distributions of shrews is environmentalmoisture Moreover this study suggests that the distribution of Scinereus and S fumeus is strongly influenced by heterogeneity of mois-ture across landscapes such as between north- and south-facing slopes

ACKNOWLEDGMENTS

This study was supported in part by Appalachian State University GraduateSchool Graduate Student Association Senate and Biology Graduate StudentAssociation I thank R W Van Devender G L Walker and D B Meikle for

MP Brannon2002 305

their guidance during this project and J K H Brannon and many others forfield assistance I also thank J F Merritt J F Pagels and an anonymousreviewer for their suggestions for revision of the original manuscript

LITERATURE CITED

ABBOTT DT and DA CROSSLEY 1982 Woody litter decompositionfollowing clear-cutting Ecology 6335-42

BRANNON MP 2000 Niche relationships of two syntopic species of shrewsSorex fumeus and S cinereus in the southern Appalachian MountainsJournal of Mammalogy 811053-1061

CUDMORE WW and JO WHITAKER Jr 1984 The distribution of thesmoky shrew Sorex fumeus and the pygmy shrew Microsorex hoyi inIndiana with notes on the distribution of other shrews Proceedings of theIndiana Academy of Sciences 93469-474

DUESER RD and HH SHUGART 1978 Microhabitats in a forest-floorsmall mammal fauna Ecology 5989-98

FORD WM J LAERM and KG BARKER 1997 Soricid response to foreststand age in southern Appalachian cove hardwood communities ForestEcology and Management 91175-181

GETZ LL 1961 Factors influencing the local distribution of shrews Ameri-can Midland Naturalist 6567-88

GIST CS and DA CROSSLEY 1975 The litter arthropod community in asouthern Appalachian hardwood forest numbers biomass and mineral ele-ment content American Midland Naturalist 93107-122

HAMILTON WJ Jr 1930 The food of the Soricidae Journal of Mammalogy1126-39

JAEGER RG 1980 Microhabitats of a terrestrial forest salamander Copeia1980265-268

KIRKLAND GL Jr 1991 Competition and coexistence in shrews(Insectivora Soricidae) Pp 15-22 In JS Findley and TL Yates (Eds)The Biology of the Soricidae Special Publication of the Museum of South-western Biology University of New Mexico Albuquerque NM 91 pp

KIRKLAND GL Jr and PK SHEPPARD 1994 Proposed standard proto-col for sampling small mammal communities Pp 277-283 In JF MerrittGL Kirkland Jr and RK Rose (Eds) Advances in the Biology of ShrewsSpecial Publication of the Carnegie Museum of Natural History No 18Pittsburgh PA 458 pp

LAERM J E BROWN MA MENZEL A WOTJALIK WM FORD and MSTRAYER 1995 The masked shrew Sorex cinereus (Insectivora Soricidae)and red-backed vole Clethrionomys gapperi (Rodentia Muridae) in the BlueRidge province of South Carolina Brimleyana 2215-21

LAERM J WM FORD TS MCCAY MA MENZEL LT LEPARDO andJL BOONE 1999 Soricid communities in the southern Appalachians Pp177-193 In RP Eckerlin (Ed) Proceedings of the Appalachian Biogeogra-phy Symposium Special Publication of the Virginia Museum of NaturalHistory No 7 Martinsville VA 258 pp

LOEB SC 1993 The role of coarse woody debris in the ecology of southeast-ern mammals Pp 108-118 In JW McMinn and DA Crossley (Eds)Biodiversity and Coarse Woody Debris in Southern Forest USDA ForestService General Technical Report SE-94

Southeastern Naturalist Vol 1 No 3306

MASER C RG ANDERSON K CROMACK Jr JT WILLIAMS and REMARTIN 1979 Dead and down woody material Pp 78-95 In JW Tho-mas (Ed) Wildlife Habitats in Managed Forests the Blue Mountains ofOregon and Washington USDA Forest Service Handbook 553 512 pp

MATLACK GR 1993 Microenvironment variation within and among forestedge sites in the eastern United States Biological Conservation 66185-194

McCAY TS 1996 Response of masked shrews (Sorex cinereus) to precipita-tion in irrigated and nonirrigated forests American Midland Naturalist135178-180

McCAY TS J LAERM MA MENZEL and WM FORD 1998 Methodsused to survey shrews (Insectivora Soricidae) and the importance of forest-floor structure Brimleyana 25110-119

McCOMB WC and RL RUMSEY 1982 Response of small mammals toforest floor clearings created by herbicides in the central AppalachiansBrimleyana 8121-134

MITCHELL JC SC RINEHART JF PAGELS KA BUHLMANN andCA PAGUE 1997 Factors influencing amphibian and small mammalassemblages in central Appalachian forests Forest Ecology and Manage-ment 9665-76

ORROCK JL JF PAGELS WJ McSHEA and EK HARPER 2000 Pre-dicting presence and abundance of a small mammal species the effect ofscale and resolution Ecological Applications 101356-1366

OWEN JG 1984 Sorex fumeus Mammalian Species 2151-8PAGELS JF KL UTHUS and HE DUVAL 1994 The masked shrew

Sorex cinereus in a relictual habitat of the southern Appalachian mountainspp 103-109 In JF Merritt GL Kirkland Jr and RK Rose (Eds) Ad-vances in the Biology of Shrews Special Publication of the Carnegie Mu-seum of Natural History No 18 Pittsburgh PA 458 pp

PARMLEY D and D HARLEY 1995 The relative seasonal abundance ofshrews in two central Georgia deciduous woodlots Georgia Journal ofScience 5383-88

PETRANKA JW ME ELDRIDGE and KE HALEY 1993 Effects oftimber harvesting on southern Appalachian salamanders Conservation Bi-ology 7363-370

SITES JW 1978 The foraging strategy of the dusky salamanderDesmognathus fuscus (Amphibia Urodela Plethodontidae) an empiricalapproach to predation theory Journal of Herpetology 12373-383

SPENCER AW and D PETTUS 1966 Habitat preferences of five sympatricspecies of long-tailed shrews Ecology 47677-683

WALES BA 1972 Vegetation analysis of north and south edges in a matureoak-hickory forest Ecological Monographs 42451-471

WHITTAKER RH 1956 Vegetation of the Great Smoky Mountains Ecologi-cal Monographs 261-80

WRIGLEY RE JE DUBOIS and HWR COPELAND 1979 Habitat abun-dance and the distribution of six species of shrews in Manitoba Journal ofMammalogy 60505-520

ZAR JH 1984 Biostatistical Analysis 2nd ed Prentice-Hall IncEnglewood-Cliffs NJ 718 pp

MP Brannon2002 305

their guidance during this project and J K H Brannon and many others forfield assistance I also thank J F Merritt J F Pagels and an anonymousreviewer for their suggestions for revision of the original manuscript

LITERATURE CITED

ABBOTT DT and DA CROSSLEY 1982 Woody litter decompositionfollowing clear-cutting Ecology 6335-42

BRANNON MP 2000 Niche relationships of two syntopic species of shrewsSorex fumeus and S cinereus in the southern Appalachian MountainsJournal of Mammalogy 811053-1061

CUDMORE WW and JO WHITAKER Jr 1984 The distribution of thesmoky shrew Sorex fumeus and the pygmy shrew Microsorex hoyi inIndiana with notes on the distribution of other shrews Proceedings of theIndiana Academy of Sciences 93469-474

DUESER RD and HH SHUGART 1978 Microhabitats in a forest-floorsmall mammal fauna Ecology 5989-98

FORD WM J LAERM and KG BARKER 1997 Soricid response to foreststand age in southern Appalachian cove hardwood communities ForestEcology and Management 91175-181

GETZ LL 1961 Factors influencing the local distribution of shrews Ameri-can Midland Naturalist 6567-88

GIST CS and DA CROSSLEY 1975 The litter arthropod community in asouthern Appalachian hardwood forest numbers biomass and mineral ele-ment content American Midland Naturalist 93107-122

HAMILTON WJ Jr 1930 The food of the Soricidae Journal of Mammalogy1126-39

JAEGER RG 1980 Microhabitats of a terrestrial forest salamander Copeia1980265-268

KIRKLAND GL Jr 1991 Competition and coexistence in shrews(Insectivora Soricidae) Pp 15-22 In JS Findley and TL Yates (Eds)The Biology of the Soricidae Special Publication of the Museum of South-western Biology University of New Mexico Albuquerque NM 91 pp

KIRKLAND GL Jr and PK SHEPPARD 1994 Proposed standard proto-col for sampling small mammal communities Pp 277-283 In JF MerrittGL Kirkland Jr and RK Rose (Eds) Advances in the Biology of ShrewsSpecial Publication of the Carnegie Museum of Natural History No 18Pittsburgh PA 458 pp

LAERM J E BROWN MA MENZEL A WOTJALIK WM FORD and MSTRAYER 1995 The masked shrew Sorex cinereus (Insectivora Soricidae)and red-backed vole Clethrionomys gapperi (Rodentia Muridae) in the BlueRidge province of South Carolina Brimleyana 2215-21

LAERM J WM FORD TS MCCAY MA MENZEL LT LEPARDO andJL BOONE 1999 Soricid communities in the southern Appalachians Pp177-193 In RP Eckerlin (Ed) Proceedings of the Appalachian Biogeogra-phy Symposium Special Publication of the Virginia Museum of NaturalHistory No 7 Martinsville VA 258 pp

LOEB SC 1993 The role of coarse woody debris in the ecology of southeast-ern mammals Pp 108-118 In JW McMinn and DA Crossley (Eds)Biodiversity and Coarse Woody Debris in Southern Forest USDA ForestService General Technical Report SE-94

Southeastern Naturalist Vol 1 No 3306

MASER C RG ANDERSON K CROMACK Jr JT WILLIAMS and REMARTIN 1979 Dead and down woody material Pp 78-95 In JW Tho-mas (Ed) Wildlife Habitats in Managed Forests the Blue Mountains ofOregon and Washington USDA Forest Service Handbook 553 512 pp

MATLACK GR 1993 Microenvironment variation within and among forestedge sites in the eastern United States Biological Conservation 66185-194

McCAY TS 1996 Response of masked shrews (Sorex cinereus) to precipita-tion in irrigated and nonirrigated forests American Midland Naturalist135178-180

McCAY TS J LAERM MA MENZEL and WM FORD 1998 Methodsused to survey shrews (Insectivora Soricidae) and the importance of forest-floor structure Brimleyana 25110-119

McCOMB WC and RL RUMSEY 1982 Response of small mammals toforest floor clearings created by herbicides in the central AppalachiansBrimleyana 8121-134

MITCHELL JC SC RINEHART JF PAGELS KA BUHLMANN andCA PAGUE 1997 Factors influencing amphibian and small mammalassemblages in central Appalachian forests Forest Ecology and Manage-ment 9665-76

ORROCK JL JF PAGELS WJ McSHEA and EK HARPER 2000 Pre-dicting presence and abundance of a small mammal species the effect ofscale and resolution Ecological Applications 101356-1366

OWEN JG 1984 Sorex fumeus Mammalian Species 2151-8PAGELS JF KL UTHUS and HE DUVAL 1994 The masked shrew

Sorex cinereus in a relictual habitat of the southern Appalachian mountainspp 103-109 In JF Merritt GL Kirkland Jr and RK Rose (Eds) Ad-vances in the Biology of Shrews Special Publication of the Carnegie Mu-seum of Natural History No 18 Pittsburgh PA 458 pp

PARMLEY D and D HARLEY 1995 The relative seasonal abundance ofshrews in two central Georgia deciduous woodlots Georgia Journal ofScience 5383-88

PETRANKA JW ME ELDRIDGE and KE HALEY 1993 Effects oftimber harvesting on southern Appalachian salamanders Conservation Bi-ology 7363-370

SITES JW 1978 The foraging strategy of the dusky salamanderDesmognathus fuscus (Amphibia Urodela Plethodontidae) an empiricalapproach to predation theory Journal of Herpetology 12373-383

SPENCER AW and D PETTUS 1966 Habitat preferences of five sympatricspecies of long-tailed shrews Ecology 47677-683

WALES BA 1972 Vegetation analysis of north and south edges in a matureoak-hickory forest Ecological Monographs 42451-471

WHITTAKER RH 1956 Vegetation of the Great Smoky Mountains Ecologi-cal Monographs 261-80

WRIGLEY RE JE DUBOIS and HWR COPELAND 1979 Habitat abun-dance and the distribution of six species of shrews in Manitoba Journal ofMammalogy 60505-520

ZAR JH 1984 Biostatistical Analysis 2nd ed Prentice-Hall IncEnglewood-Cliffs NJ 718 pp

Southeastern Naturalist Vol 1 No 3306

MASER C RG ANDERSON K CROMACK Jr JT WILLIAMS and REMARTIN 1979 Dead and down woody material Pp 78-95 In JW Tho-mas (Ed) Wildlife Habitats in Managed Forests the Blue Mountains ofOregon and Washington USDA Forest Service Handbook 553 512 pp

MATLACK GR 1993 Microenvironment variation within and among forestedge sites in the eastern United States Biological Conservation 66185-194

McCAY TS 1996 Response of masked shrews (Sorex cinereus) to precipita-tion in irrigated and nonirrigated forests American Midland Naturalist135178-180

McCAY TS J LAERM MA MENZEL and WM FORD 1998 Methodsused to survey shrews (Insectivora Soricidae) and the importance of forest-floor structure Brimleyana 25110-119

McCOMB WC and RL RUMSEY 1982 Response of small mammals toforest floor clearings created by herbicides in the central AppalachiansBrimleyana 8121-134

MITCHELL JC SC RINEHART JF PAGELS KA BUHLMANN andCA PAGUE 1997 Factors influencing amphibian and small mammalassemblages in central Appalachian forests Forest Ecology and Manage-ment 9665-76

ORROCK JL JF PAGELS WJ McSHEA and EK HARPER 2000 Pre-dicting presence and abundance of a small mammal species the effect ofscale and resolution Ecological Applications 101356-1366

OWEN JG 1984 Sorex fumeus Mammalian Species 2151-8PAGELS JF KL UTHUS and HE DUVAL 1994 The masked shrew

Sorex cinereus in a relictual habitat of the southern Appalachian mountainspp 103-109 In JF Merritt GL Kirkland Jr and RK Rose (Eds) Ad-vances in the Biology of Shrews Special Publication of the Carnegie Mu-seum of Natural History No 18 Pittsburgh PA 458 pp

PARMLEY D and D HARLEY 1995 The relative seasonal abundance ofshrews in two central Georgia deciduous woodlots Georgia Journal ofScience 5383-88

PETRANKA JW ME ELDRIDGE and KE HALEY 1993 Effects oftimber harvesting on southern Appalachian salamanders Conservation Bi-ology 7363-370

SITES JW 1978 The foraging strategy of the dusky salamanderDesmognathus fuscus (Amphibia Urodela Plethodontidae) an empiricalapproach to predation theory Journal of Herpetology 12373-383

SPENCER AW and D PETTUS 1966 Habitat preferences of five sympatricspecies of long-tailed shrews Ecology 47677-683

WALES BA 1972 Vegetation analysis of north and south edges in a matureoak-hickory forest Ecological Monographs 42451-471

WHITTAKER RH 1956 Vegetation of the Great Smoky Mountains Ecologi-cal Monographs 261-80

WRIGLEY RE JE DUBOIS and HWR COPELAND 1979 Habitat abun-dance and the distribution of six species of shrews in Manitoba Journal ofMammalogy 60505-520

ZAR JH 1984 Biostatistical Analysis 2nd ed Prentice-Hall IncEnglewood-Cliffs NJ 718 pp