Niche Relationships of Two Syntopic Species of Shrews, Sorex Fumeus and S

American Society of Mammalogists

Niche Relationships of Two Syntopic Species of Shrews, Sorex fumeus and S. cinereus, in the Southern Appalachian Mountains Author(s): M. Patrick Brannon Source: Journal of Mammalogy, Vol. 81, No. 4 (Nov., 2000), pp. 1053-1061 Published by: American Society of Mammalogists Stable URL: http://www.jstor.org/stable/1383372 . Accessed: 22/08/2013 13:42

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

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

American Society of Mammalogists is collaborating with JSTOR to digitize, preserve and extend access to Journal of Mammalogy.

http://www.jstor.org

This content downloaded from 128.109.48.4 on Thu, 22 Aug 2013 13:42:39 PM All use subject to JSTOR Terms and Conditions Journal of Mammalogy, 81(4):1053-1061, 2000

NICHE RELATIONSHIPS OF TWO SYNTOPIC SPECIES OF SHREWS, SOREX FUMEUS AND S. CINEREUS, IN THE SOUTHERN APPALACHIAN MOUNTAINS

M. PATRICK BRANNON*

Departmentof Biology, AppalachianState University,Boone, NC 28608 Present address: VirginiaMuseum of Natural History, Martinsville, VA 24112

The smoky shrew (Sorexfumeus) and the masked shrew (S. cinereus) are common soricids in mature southern Appalachian hardwood forests. To better understand the role of body size and niche relationships in these syntopic species, 12 50- by 50-m plots were established in the Pisgah National Forest of western North Carolina. Trapping was conducted from August through November 1996 and from March through August 1997 using Y-shaped drift fences with associated pitfalls. Prey items and microhabitat components were measured to examine correlations with abundance of shrew species. Total shrew captures (n = 176) included 105 (59.7%) Sorex fumeus and 41 (23.3%) S. cinereus. Smoky shrews were significantly larger than masked shrews in both mass and body length. Stepwise multiple regression analyses determined that a combination of litter moisture, class 5 coarse woody debris (CWD), and number of invertebrates and mountain dusky salamanders (Desmo- gnathus ochrophaeus) was the best predictor of S. fumeus abundance (R2 = 92.8%), whereas S. cinereus abundance was best explained by a combination of litter moisture, leaf-litter depth, class 3 CWD, and invertebrate size (R2 = 57.7%). Microhabitat niche breadth (MB) of S. cinereus (2.11) was narrower than that of S. fumeus (2.27). Linear discriminant function analysis revealed significant ecological separation between the 2 shrew species (D2 = 0.62), despite high levels of microhabitat niche overlap (MO = 65.7%). The larger body size of smoky shrews may provide an advantage in that it can use parts of the microhabitat that are inaccessible to its smaller congener, thereby reducing interspecific competition.

Key words: Appalachianmountains, body size, niche, resource partitioning,shrews, Sorex

Body size is believed to be an important, fore a 2nd less common species of the same organizing factor in the community struc- size class is present (Fox and Kirkland ture of many closely related coexisting ver- 1992; Kirkland 1991). Interspecific com- tebrate species (Asplund 1974; Bowers and petition is the most likely mechanism re- Brown 1982; Dickman 1988; Krzysik sponsible for these nonrandom associations 1979). In multispecies communities of (Fox and Kirkland 1992). Ecological sepa- shrews (Insectivora: Soricidae), species sort ration in shrews may be achieved through into 3 general size classes: large (>10 g), differential exploitation of common re- medium (5-10 g), and small (<5 g). These sources by species of different size classes communities are not random assemblages (Churchfield and Sheftel 1994; Ellenbroek but rather appear to follow a species-assem- 1980; Hawes 1977; Terry 1981; Yashino rule that bly predicts that each of the 3 size and Abe 1984). Larger body size may con- niches and (Kirkland Snoddy 1999) should vey a competitive advantage by facilitating be filled with a common be- single species use of a wider range of prey items and high- * Correspondent: [email protected] er-quality microhabitats inaccessible to

1053

This content downloaded from 128.109.48.4 on Thu, 22 Aug 2013 13:42:39 PM All use subject to JSTOR Terms and Conditions 1054 JOURNAL OF MAMMALOGY Vol. 81, No. 4 more diminutive congeners (Dickman meson 1949) due to their coexistence in a 1988; Fox and Kirkland 1992), resulting in variety of habitats and their relative simi- a broader niche breadth (Churchfield 1991; larity in body size and diet (Hamilton 1930; Churchfield and Sheftel 1994; Wilson Huggins and Kennedy 1989; Whitaker and 1975). Cudmore 1987; Whitaker and French 1984) Although microhabitat selection by sori- make these species interesting subjects for cids is strongly influenced by environmen- studies of niche relationships. I tested hy- tal moisture and protection from predators potheses that in commonly occurring, syn- (Getz 1961; Seagle 1985), differences can topic species pairs of Sorex members sort reflect foraging behavior and diet as deter- into 2 distinct body size classes, larger body mined by body size (Churchfield 1991). size conveys an advantage in terms of re- The most common difference in foraging source accessibility, and ecological separa- activity among syntopic shrews is stratifi- tion between species is determined by body cational segregation of the forest floor size and differential use of foraging micro- (Churchfield and Sheftel 1994; Ellenbroek habitat. 1980; Terry 1981; Yashino and Abe 1984). MATERIALS AND METHODS Although shrews have generalized diets because of their high metabolic rate (Aitchi- Research was conducted in the Gingercake Creek of the National son 1987; Hamilton 1930), the larger mem- drainage Pisgah Forest, ber of a is often more fossorial Burke County, western North Carolina. This species pair area was characterized and consume a of by steep (20-38') slopes may greater proportion with numerousstreams and such as earthworms seepages. Vegetation large hypogeal prey, consisted primarilyof mixed hardwoodswith an Churchfield and Sheftel (Churchfield 1991; understoryof Rhododendronand mountainlau- 1994; Kirkland 1991; Whitaker and Cud- rel (Kalmia latifolia). Stand age is about 55 more 1987; Whitaker and French 1984). years old, according to a United States Forest Shrew diversity should therefore depend on Service CISC database,and elevations averaged availability of different foraging microhab- about 787 m (757-818 m). A more detailed de- itats (Dickman 1988; Fox and Kirkland scription of the habitat is available in Brannon 1992). An understanding of how forest- (1997). Twelve 50- were floor resources are partitioned among syn- by 50-m plots established, with lines and topic species pairs is important in determin- positioned boundary parallel per- pendicularto prevailing contours. In the center ing niche relationships of shrews and struc- of each plot was a drift fence array consisting ture of soricid assemblages. of 3 arms of 3.0-m-long by 61-cm-tall aluminum was to determine of re- My goal patterns flashing arrangedin a "Y" with one 20-1 pitfall source use in 2 of congeneric species (5-gallon plastic bucket) located at each end and shrews (Sorex sp.) that co-occur in the anotherat the centralintersection. Each of the 4 higher elevations of the southern Appala- pitfalls was partiallyfilled with water to drown chians. Soricid assemblages in this region capturesand thus preventpredation within traps. tend to consist of 4 species of decreasing In addition to effectively capturing shrews size: the northern short-tailed shrew (Blar- (Kirklandand Sheppard 1994), that method of was ina brevicauda), smoky shrew (Sorex fu- trapping useful for sampling prey items such as surface-activemacroinvertebrates meus), masked shrew (S. cinereus,) and (Dick- man 1988; 1986) and salamanders(Gib- pygmy shrew (S. hoyi), although occasion- Ryan bons and Semlitsch 1981). A 50- 50-cm there are others et al. by ally (Ford 1997; board, elevated about 10 cm with nails, was Laerm et al. The masked shrew and 1999). placed over each pitfall to preventaccumulation smoky shrew are closely related species that of rain and leaves. Pitfalls were closed with are abundant throughout their ranges and tight-fittingplastic lids when not in use. are broadly sympatric. The potential for in- Trapping was conducted for 6 consecutive terspecific intolerance (Cawthorn 1994; Ja- days each month from August to November

This content downloaded from 128.109.48.4 on Thu, 22 Aug 2013 13:42:39 PM All use subject to JSTOR Terms and Conditions November2000 BRANNON-SHREW NICHE RELATIONSHIPS 1055

TABLE1.-Nine resource variables used for backwards-elimination, stepwise multiple regression and linear discriminant function analyses.

Resourcevariable Description INVERTN Total numberof invertebrates INVERTG Average invertebratesize (mass in grams) DESMOG Total numberof Desmognathusochrophaeus HRBCVR Percentageherbaceous cover CWD3 Volume of decay class 3 coarse woody debris CWD4 Volume of decay class 4 coarse woody debris CWD5 Volume of decay class 5 coarse woody debris LITDEPTH Leaf-litterdepth (cm) LITMOIST Percentagemoisture content of leaf litter

1996 and from March to August 1997. No col- were used to determine relative amounts of lecting was done from December 1996 to Feb- CWD within plots using a line-intercept method. ruary 1997 because of site inaccessibility. On a Diameter and length of any fallen log cm daily basis during each trapping period, verte- in diameter at the point of intersection along_10 the brates were removed from traps and taken im- transect were recorded. Any branches cm mediately back to the lab, where they were mea- in diameter were treated as separate logs._10 Stage sured, identified to species, and fixed in 10% of decomposition of each log also was ranked formalin. All vertebrate specimens were later from class 1 CWD for recently fallen trees with transferred to 70% ethanol and deposited in the support points intact and little evidence of decay collections of Appalachian State University. to class 5 CWD for extremely decomposed, soft, Because most shrews take whatever prey they moist logs that were partially buried by soil and encounter within the constraints imposed by forest litter (Maser et al. 1979; Petranka et al. their body dimensions (Churchfield 1991), all 1994). Decomposition classes differ in their eco- invertebrate taxa were considered potential food logical function (Maser et al. 1979). Because items. Invertebrates were strained from pitfalls classes 1 and 2 logs are suspended above on the last day of each trapping period. Total ground, have not undergone significant decay, number of invertebrates for each plot was count- and are uncommon in mature forests, they were ed, and individuals were identified to order and considered to be unimportant as foraging micro- weighed to obtain the average prey size (mass habitat (Petranka et al. 1994). in grams). Although shrews are known to also Herbaceous cover, leaf-litter depth, and litter consume salamanders (Hamilton 1930, 1940; moisture content were measured every trapping Whitaker and Cudmore 1987), only the moun- period because of seasonal variability. At 5-m tain dusky salamander (Desmognathus ochro- intervals along the middle transect, a 1-m2 quad- phaeus) was considered as potential prey be- rat was established for a total of 10 quadrats per cause it is small and terrestrial and lacks noxious plot. Percentage cover of herbaceous plants skin secretions of other, less palatable species <0.5 m high was estimated visually for each (Brodie et al. 1979). Stomach contents of shrews quadrat and averaged to obtain percentage cover were not examined. per plot. Leaf litter was defined as nonwoody Microhabitat features (Table 1) to be mea- material and woody stems <10 mm in diameter. sured were selected on the basis of their proba- Average litter depths for each plot were mea- ble importance to shrews in providing environ- sured by simply pressing a metric ruler through mental moisture, prey abundance, and protective the litter to the A-horizon of the soil at each of cover. Those included volume of coarse woody the 10 quadrats. To obtain measurements of litter debris (CWD; fallen logs), percentage herba- moisture, a 0.25-m2 sample was collected ran- ceous cover, and leaf-litter depth and moisture domly from each plot on 3 days during each content (Doyle 1987; Dueser and Shugart 1978; trapping period. Samples were weighed imme- Lee 1995; Seagle 1985; Yahner 1986). Three 50- diately on return to the lab (wet mass), dried at m transects were established at 12.5-m intervals 100TC for 24 h, and reweighed (dry mass). Litter across contours within each plot. Those transects moisture content was calculated as wet mass mi-

This content downloaded from 128.109.48.4 on Thu, 22 Aug 2013 13:42:39 PM All use subject to JSTOR Terms and Conditions 1056 JOURNAL OF MAMMALOGY Vol. 81, No. 4 nus dry mass and expressed as a percentage of means. Mahalanobisdistances were used to test wet mass. if population means showed significant segre- Differences in mass and body length (exclud- gation along the resource gradientby referring ing tail length) between shrew species were ex- the equation amined using a 1-way analysis of variance. Mi- + nn22 crohabitat niche breadth (MB) for each species n, n - p - D2 was calculated as (n, + n2 - 2)p + n2) )\nl to an F-distributionwith v, = p and v2 = n1 + MB = > - p, Inpj n2 - p - 1 degrees of freedom and with n, and n2 representingthe sample size of each of the 2 where was the proportion of the total number p,j shrew species, respectively (Johnson and Wich- of for i at j (Yah- captures species captured plot ern 1992). ner 1986). Microhabitatniche overlap (MO) between species pairs was calculatedas RESULTS MO= 1 - - 0.5 Ipxj PjI During 2,544 trap-nights, 176 shrews were captured. Of that total, 105 (59.7%) where and were the of total pxj pyj proportions were Sorexfumeus, and 41 (23.3%) were S. captures of species x and species y in plot j, cinereus. The remaining soricids consisted respectively (Schoener 1968). of 15 (8.5%) Sorex hoyi, 14 (8.0%) Blarina Combinationsof resource variables (Table 1) brevicauda, and 1 rock shrew, Sorex variation in shrew abun- (0.6%) may explain species Other small mammals collected in- dance more than can an individual dispar. adequately cluded 192 white-footed mice predictor.The importanceof each combination (Peromyscus 115 woodland mice was determinedby regressingthe numberof in- leucopus), jumping dividuals capturedper plot (dependentvariable) (Napeozapus insignis), 16 red-backed voles on subsets of predictors(independent variables), (Clethrionomys gapperi), 11 golden mice using backwards-elimination,stepwise multiple (Ochrotomys nuttalli), and 2 chipmunks regression(Minitab, Inc. 1996). Thatmethod be- (Tamias striatus). Pitfall traps also captured gan with a model containing a combinationof 259 salamanders (65 Desmognathus och- all predictors and systematically eliminated rophaeus), 4 frogs, and 1 snake. A total of those variablesthat contributedleast to the over- 9,486 invertebrates were collected. The all regression equation. At each step, an F-sta- most abundant taxa in decreasing order tistic for each predictorremaining in the model were Orthoptera, Coleoptera, was calculated,and the variablewith the greatest Hymenoptera, larvae, amount of errorwas removed. That was equiv- lepidopteran Thysanura, spiders, alent to excluding the variablewith the smallest millipedes, centipedes, earthworms, and partial correlation.That procedure was contin- snails, with each plot yielding similar pro- ued until additionalremoval of a predictorre- portions of each taxon. sulted in a significant reductionin the percent- Sorex fumeus and S. cinereus differed in age of variationexplained by the overall equa- mass (F = 675.01, d.f = 1, 144, P < 0.01) tion (R2). and body length (F = 283.14, d.f. = 1, 144, To verify ecological separationbetween spe- P < 0.01). Mean mass (?1 SE) was 8.47 the 9 variableswere en- cies, original predictor ? 0.77 g for S. fumeus and 4.32 ? 0.85 g tered into a linear discriminantfunction analysis for S. cinereus. Mean was 55.3 Inc. That body length (Minitab, 1996). procedurepredicted ? 3.8 mm for S. and 46.5 ? 2.5 based on a set of continu- fumeus species membership mm for S. cinereus. ous variables,with g = numberof groups (spe- Microhabitat niche breadth of S. cinereus cies) and p = number of predictors. Stepwise models sequentially extracted those orthogonal (MB = 2.11) was narrower than that of S. variables most capable of separatingspecies by fumeus (MB = 2.27). Despite high micro- maximizing among- to within-groupssums of habitat niche overlap (MO = 65.7%), some squares and providedthe Mahalanobisdistances resource partitioning was determined to ex- (D2) Orsample-squared distances between group ist between the 2 species. A combination of

This content downloaded from 128.109.48.4 on Thu, 22 Aug 2013 13:42:39 PM All use subject to JSTOR Terms and Conditions November2000 BRANNON-SHREW NICHE RELATIONSHIPS 1057

TABLE2.-Results of backwards-elimination,stepwise multiple regression analysis for 2 shrew species.

Resource Partial Regression Species variable (x) F coefficient constant R2 Sorex fumeus LITMOIST 25.50 1.04 -31.38 0.928 CWD5 18.15 -2.02 INVERTN 27.67 -0.01 DESMOG 6.10 -0.27 S. cinereus LITMOIST 2.53 0.34 -64.43 0.577 LITDEPTH 6.92 0.11 CWD3 0.48 1.61 INVERTG 2.53 4.51

litter moisture, class 5 CWD, and number and Sheftel 1994; Fox and Kirkland 1992; of invertebrates and mountain dusky sala- Kirkland 1991). manders was the best predictor of S. fumeus In multispecies communities, partitioning abundance (R2 = 92.8%), whereas abun- of the forest-floor microhabitat occurs along dance of S. cinereus was best predicted by a vertical gradient in relation to differences a combination of litter moisture, leaf-litter in shrew body size (Kirkland 1991). Al- depth, class 3 CWD, and invertebrate size though there is a tendency for small species (R2 = 57.7%; Table 2). Linear discriminant to take smaller, epigeal (litter-dwelling) function analysis showed that means of the prey and for large species to take larger, microhabitat niches for the 2 species were more hypogeal (soil-dwelling) prey significantly separated (D2 = 0.62, F = (Churchfield and Sheftel 1994; Kirkland 1.92, d.f = 9, 136, P < 0.05), with 69.2% 1991), diet of shrews is not correlated clear- of the original observations correctly clas- ly with body size (Churchfield and Sheftel sified. 1994). Shrews have very generalized diets because of their high energy demands and DIscusSION rates of food consumption (Aitchison Morphological and behavioral differenc- 1987), which largely reflect variable avail- es can act to reduce interspecific competi- ability of different prey types (Churchfield tion between syntopic congeners. There 1991; McCay and Storm 1997). Despite may be a limit on how similar in size 2 co- high levels of dietary overlap in terms of occurring species can be and still avoid taxonomic categories, shrews differ in the competitive exclusion. A ratio of 1.3 for proportions that they consume each cate- length and 2.0 for body mass has been sug- gory (Churchfield 1991; Hamilton 1930; gested as an estimate of the degree of dif- Whitaker and Cudmore 1987; Whitaker and ference necessary for 2 species to coexist French 1984). Size and location of prey syntopically in high numbers (Hutchinson may be more important than taxonomic and MacArthur 1959). These values corre- group in reducing interspecific competition spond almost exactly to the size difference (Churchfield 1991; Kirkland 1991). ratios observed for S. fumeus and S. ciner- Diminutive species such as S. cinereus eus in this study (1.2 for body length and are better equipped to feed on tiny, surface- 2.0 for mass). Such differences in body siz- and litter-active invertebrates, while larger, es facilitate coexistence of syntopic shrew more robust shrew species may be semifos- species by allowing differential exploitation sorial (Kirkland 1991; Ryan 1986; Whitak- of foraging microhabitat and associated er and French 1984). While some dietary prey items (Churchfield 1991; Churchfield overlap exists between smoky shrews and

This content downloaded from 128.109.48.4 on Thu, 22 Aug 2013 13:42:39 PM All use subject to JSTOR Terms and Conditions 1058 JOURNAL OF MAMMALOGY Vol. 81, No. 4 masked shrews (Hamilton 1930; Whitaker (Ash 1995; Gist and Crossley 1975; Lee and French 1984), S. fumeus consumes con- 1995), logs that are in the later stages of siderably more large prey, such as earth- decay may be optimal microhabitats be- worms and centipedes, than does S. ciner- cause they also can offer a stable micro- eus (Whitaker and Cudmore 1987), which environment refuge during periods of pro- reflects a partially hypogeal foraging mode longed drought (Jaeger 1980; Maser et al. (Kirkland 1991). Smoky shrews also con- 1979). The spongy texture of extremely de- sume a greater proportion of salamanders composed logs may enable burrowing by (Hamilton 1930). Although small body size larger shrews while serving as both a mois- may preclude diminutive species from ef- ture reservoir and a source of a wide variety ficiently handling and consuming large prey of soil-dwelling invertebrates (Maser et al. (Getz 1961; McCay and Storm 1997), the 1979; McComb and Rumsey 1982) and sal- reverse is much less true (Wilson 1975). amanders (Jaeger 1980; Petranka et al. Thus, the smaller S. cinereus is limited 1994). more greatly by prey size than the larger S. Because of their high levels of rainfall, fumeus, whereas the latter may only be lim- moderate temperatures, and abundant leaf ited by prey abundance. Increased body size litter, southern Appalachian hardwood for- may provide an advantage over smaller spe- ests generally are more productive than cies by allowing use of a wider range of many other forests in terms of invertebrate foraging microhabitats and food items biomass (Gist and Crossley 1975). Larger (Dickman 1988). species of shrews have greater per capita Although dietary analyses were not per- food requirements than smaller species. Nu- formed in this study, correlations with mi- merical dominance of shrew species may crohabitat suggest that differences in for- therefore be determined in part by habitat aging modes do exist between smoky productivity (Hanski 1994; Kirkland and shrews and masked shrews based on body Snoddy 1999). Assuming no difference in size. Shrews require microhabitats with foraging efficiency between species of dif- high levels of environmental moisture be- ferent sizes, larger species should dominate cause it affects their water balance and in higher-quality habitats and vice versa. availability of prey (Getz 1961; Kirkland Typically, more generalized species are nu- 1991; Wrigley et al. 1979). Species exploit merically dominant in multispecies com- the best foraging microhabitats available to munities of soricids (Churchfield 1991; minimize energy expenditure and risk of Laerm et al. 1999). predation (Barnard and Brown 1987; Han- In this study, S. fumeus was more abun- ski 1989; Seagle 1985; Yahner 1986). Al- dant than S. cinereus and had a broader though both species of shrews actively for- niche breadth, which may be indicative of age on the forest floor and in leaf litter, es- its larger size and ability to utilize a greater pecially during periods of moderate rainfall variety of microhabitats and prey types. (Brannon 1997; Kirkland et al. 1998; Vick- Conversely, the diminutive size of S. ciner- ery and Bider 1978), in this study only S. eus restricts its use of optimal foraging mi- fumeus abundance was associated with crohabitats and prey items in this habitat class 5 CWD. Similarly, McCay et al. type (Laerm et al. 1999) and may make it (1998) found smoky shrews to selectively competitively inferior to its larger congener use microhabitat structures such as logs and (Dickman 1988; Fox and Kirkland 1992; rocks, and Yahner (1986) found a negative Hawes 1977), resulting in a more com- relationship between logs and masked pressed niche breadth (Churchfield and shrews. Although other forms of CWD and Sheftel 1994). However, if moisture and in- a deep layer of leaf litter can provide mois- vertebrate prey are generally abundant, spe- ture, abundant prey, and protective cover cies may share a large proportion of the

This content downloaded from 128.109.48.4 on Thu, 22 Aug 2013 13:42:39 PM All use subject to JSTOR Terms and Conditions November 2000 BRANNON-SHREW NICHE RELATIONSHIPS 1059 available resources without the danger of BARNARD,C. J., ANDC. A. J. BROWN.1987. Risk-sen- sitive foraging and patch residence time in common competitive exclusion. Considerable niche shrews, Sorex araneus L. Animal Behaviour 35: overlap between congeners, as in this study 1255-1257. between S. and S. cinereus, of- BOWERS,M. A., AND J. H. BROWN.1982. Body size fumeus may and coexistence in desert rodents: chance or com- with reduced ten be correlated competition. munity structure? Ecology 63:391-400. Maximum tolerable overlap in niches BRANNON, M. P. 1997. Microhabitat factors influencing should be lower in intensely competitive shrew diversity in southern Appalachian deciduous forest. M.S. thesis, Appalachian State University, situations than in environments with a sur- Boone, North Carolina. plus of resources or low demand:supply ra- BRODIE, E. D., JR., R. T. NOWAK, AND W. R. HARVEY. tios It should be remem- 1979. The effectiveness of antipredator secretions (Pianka 1972). and behavior of selected salamanders that shrews are not the against bered, however, only shrews. Copeia 1979:270-274. abundant forest-floor predators to use these CAWTHORN,J. M. 1994. A live trapping study of two resources (Jaeger 1980; Petranka et al. syntopic species of Sorex, S. cinereus and S.fumeus, in southwestern Pennsylvania. Pp. 39-43 in Ad- 1994). vances in the biology of shrews (J. E Merritt, G. L. Results of this study are consistent with Kirkland, Jr., and R. K. Rose, eds.). Special Publi- the idea that size is an fac- cation of the Carnegie Museum of Natural History, body important 18:1-458. tor in the of soricid communities Pittsburgh structuring CHURCHFIELD,S. 1991. Niche dynamics, food resourc- (Churchfield 1991; Fox and Kirkland 1992; es, and feeding strategies in multispecies commu- Kirkland 1991). Larger size may facilitate nities of shrews. Pp. 23-34 in The biology of the Soricidae (J. S. Findley and T. L. Yates, eds.). Spe- use of some prey items and foraging micro- cial Publication of the Museum of Southwestern Bi- habitats that are inaccessible to smaller con- ology, University of New Mexico, Albuquerque. (Dickman 1988). Coexistence of CHURCHFIELD,S., ANDB. I. SHEFTEL.1994. Food niche geners and in a shrews is therefore on overlap ecological separation multi-species syntopic dependent community of shrews in the Siberian taiga. Journal different-sized species exploiting measur- of Zoology (London) 234:105-124. ably different parts of their common envi- DICKMAN,C. R. 1988. Body size, prey size, and community structure in insectivorous mammals. Ecology ronment, or distinct microhabitat niches. 69:569-580. DOYLE,A. T 1987. Microhabitat separation among ACKNOWLEDGMENTS sympatric microtines, Clethrionomys californicus, Microtus oregoni, and M. richardsoni. The Ameri- I am greatly indebtedto R. W. Van Devender can Midland Naturalist 118:258-265. for his guidance duringthis project.I also thank DUESER, R. D., AND H. H. SHUGART, JR. 1978. Micro- D. B. Meikle, G. L. Walker,J. K. H. Brannon, habitats in a forest-floor small mammal fauna. Ecol- and the numerous persons who provided field ogy 59:89-98. assistance. for this was in ELLENBROEK,E J. M. 1980. Interspecific competition Funding project given in the shrews Sorex araneus and Sorex minutus (Sor- part by the AppalachianState University Grad- icidae, Insectivora): a population study of the Irish uate School, GraduateStudent Association Sen- pygmy shrew. Journal of Zoology (London) 192: ate, and Biology GraduateStudent Association. 119-136. This study was conductedunder North Carolina FORD, W. M., J. LAERM, AND K. G. BARKER. 1997. Sor- Wildlife icid response to forest stand age in southern Appa- Resources Commission Scientific Col- lachian cove hardwood communities. Forest Ecolo- lecting PermitsNC-96-ES-11 and NC-97-ES-39 gy and Management 91:175-181. and United States Forest Service Special Use Fox, B. J., ANDG. L. KIRKLAND,JR. 1992. An assem- Permit 2104-03. bly rule for functional groups applied to North American soricid communities. Journal of Mam- 73:491-503. LITERATURE CITED malogy GETZ,L. L. 1961. Factors influencing the local distri- ArrCHISON,C. W. 1987. Review of winter trophic re- bution of shrews. The American Midland Naturalist lations of soricine shrews. Mammal Review 17:1- 65:67-88. 24. GIBBONS, J. W., AND R. D. SEMLITSCH.1981. Terrestrial ASH, A. N. 1995. Effects of clear-cutting on litter pa- drift fences with pitfall traps: an effective technique rameters in the southern Blue Ridge Mountains. for quantitative sampling of animal populations. Castanea 60:89-97. Brimleyana 7:1-16. ASPLUND,K. K. 1974. Body size and habitat utilization GIST, C. S., AND D. A. CROSSLEY, JR. 1975. The litter in whiptail lizards (Cnemidophorus). Copeia 1974: arthropod community in a southern Appalachian 695-703. hardwood forest: numbers, biomass and mineral el-

This content downloaded from 128.109.48.4 on Thu, 22 Aug 2013 13:42:39 PM All use subject to JSTOR Terms and Conditions 1060 JOURNAL OF MAMMALOGY Vol. 81, No. 4

ement content. The American Midland Naturalist LAERM,J., W. M. FORD,T. S. McCAY, M. A. MENZEL, 93:107-122. L. T LEPARDO,AND J. L. BOONE.1999. Soricid com- HAMILTON,W. J., JR. 1930. The food of the Soricidae. munities in the southern Appalachians. Pp. 177-193 Journal of Mammalogy 11:26-39. in Proceedings of the Appalachian biogeography HAMILTON,W. J., JR. 1940. The biology of the smoky symposium (R. P. Eckerlin, ed.). Special Publication shrew (Sorex fumeus fumeus Miller). Zoologica 25: of the Virginia Museum of Natural History, Mar- 473-492. tinsville 7:1-258. HANSKI, I. 1989. Habitat selection in a patchy environ- LEE, S. D. 1995. Comparison of population character- ment: individual differences in common shrews. An- istics of three species of shrews and the shrew-mole imal Behaviour 38:414-422. in habitats with different amounts of coarse woody HANSKI,I. 1994. Population biological consequences debris. Acta Theriologica 40:415-424. of body size in Sorex. Pp. 15-22 in Advances in the MASER, C., R. G. ANDERSON, K. CROMACK, JR., J. T. biology of shrews (J. E Merritt, G. L. Kirkland, Jr., WILLIAMS, ANDR. E. MARTIN.1979. Dead and down and R. K. Rose, eds.). Special Publication of the woody material. Pp. 78-95 in Wildlife habitats in Carnegie Museum of Natural History, Pittsburgh, managed forests: the Blue Mountains of Oregon and 18:1-458. Washington (J. W Thomas, ed.). United States De- HAWES,M. L. 1977. Home range, territoriality, and partment of Agriculture Forest Service Handbook ecological separation in sympatric shrews, Sorex va- 553:1-512. grans and S. obscurus. Journal of Mammalogy 58: MCCAY, T. S., J. LAERM,M. A. MENZEL,AND W. M. 354-367. FORD.1998. Methods used to survey shrews (Insec- HuGGINS,J. A., ANDM. L. KENNEDY.1989. Morpho- tivora: Soricidae) and the importance of forest-floor logic variation in the masked shrew (Sorex cinereus) structure. Brimleyana 25:110-119. and the smoky shrew (S. fumeus). The American MCCAY, T. S., ANDG. L. STORM.1997. Masked shrew Midland Naturalist 122:11-25. (Sorex cinereus) abundance, diet, and prey selection HUTCHINSON,G. E., ANDR. H. MACARTHUR.1959. A in an irrigated forest. The American Midland Natu- theoretical ecological model of size distributions ralist 138:268-275. among species of animals. The American Naturalist MCCOMB,W. C., ANDR. L. RUMSEY.1982. Response 93:117-125. of small mammals to forest clearings created by her- JAEGER,R. G. 1980. Microhabitats of a terrestrial forest bicides in the central Appalachians. Brimleyana 8: salamander. Copeia 1980:265-268. 121-134. JAMESON,E. W., JR. 1949. Some factors influencing the MINITAB,INC. 1996. Minitab reference manual. Release local distribution and abundance of woodland small 11.1 for Windows. Sowers Printing Company, Leb- mammals in central New York. Journal of Mam- anon, Pennsylvania. malogy 30:221-235. PETRANKA, J. W., M. P. BRANNON,M. E. HOPEY,AND JOHNSON,R. A., ANDD. W WICHERN.1992. Applied C. K. SMITH.1994. Effects of timber harvesting on multivariate statistical analysis. Prentice Hall, En- low elevation populations of southern Appalachian glewood Cliffs, New Jersey. salamanders. Forest Ecology and Management 67: KIRKLAND,G. L., JR. 1991. Competition and coexis- 135-147. tence in shrews (Insectivora, Soricidae). Pp. 15-22 PIANKA,E. R. 1972. r and K selection or b and d se- in The biology of the Soricidae (J. S. Findley and lection? The American Naturalist 106:581-588. T. L. Yates, eds.). Special Publication of the Museum RYAN, J. M. 1986. Dietary overlap in sympatric pop- of Southwestern Biology, University of New Mexi- ulations of pygmy shrews, Sorex hoyi, and masked co, Albuquerque. shrews, Sorex cinereus, in Michigan. The Canadian KIRKLAND, G. L., JR., AND P. K. SHEPPARD. 1994. Pro- Field-Naturalist 100:225-228. posed standard protocol for sampling small mammal SCHOENER, T. W. 1968. The Anolis lizards of Bimini: communities. Pp. 277-283 in Advances in the bi- resource partitioning in a complex fauna. Ecology ology of shrews (J. E Merritt, G. L. Kirkland, Jr., 49:704-726. and R. K. Rose, eds.). Special Publication of the SEAGLE,S. W. 1985. Patterns of small mammal micro- Carnegie Museum of Natural History, Pittsburgh 18: habitat utilization in cedar glade and deciduous for- 1-458. est habitats. Journal of Mammalogy 66:22-35. KIRKLAND,G. L., JR., P. K. SHEPPARD,M. J. SHAUGH- TERRY,C. J. 1981. Habitat differentiation among three NESSY,JR., AND B. A. WOLESLAGLE.1998. Factors species of Sorex and Neurotrichus gibbsi in Wash- influencing perceived community structure in near- ington. The American Midland Naturalist 106:119- arctic forest small mammals. Acta Theriologica 43: 125. 121-135. VICKERY,W. L., ANDJ. R. BIDER.1978. The effect of KIRKLAND,G. L., JR., ANDH. W. SNODDY.1999. Bio- weather on Sorex cinereus activity. Canadian Jour- geography and community ecology of shrews nal of Zoology 56:291-297. (Mammalia: Soricidae) in the northern Appalachian WHITAKER,J. O., JR., ANDW. W. CUDMORE.1987. Food mountains. Pp. 167-175 in Proceedings of the Ap- and ectoparasites of shrews of south central Indiana palachian biogeography symposium (R. P. Eckerlin, with emphasis on Sorex fumeus and Sorex hoyi. Pro- ed.). Special Publication of the Virginia Museum of ceedings of the Indiana Academy of Science 96: Natural History, Martinsville 7:1-258. 543-552. KRZYSIK,A. J. 1979. Resource allocation, coexistence, J. W. FRENCH. WHITAKER, O., JR., ANDT. 1984. Foods and the niche structure of a streambank salamander of six species of sympatric shrews from New Bruns- community. Ecological Monographs 49:173-194. wick. Canadian Journal of Zoology 62:622-626.

This content downloaded from 128.109.48.4 on Thu, 22 Aug 2013 13:42:39 PM All use subject to JSTOR Terms and Conditions November 2000 BRANNON-SHREW NICHE RELATIONSHIPS 1061

WILSON,D. S. 1975. The adequacy of body size as a YASHINO,H., AND H. ABE. 1984. Comparative study niche difference. The American Naturalist 109:769- on the foraging habits of two species of soricine 784. shrews. Acta Theriologica 29:35-43. WRIGLEY, R. E., J. E. DuBoIS, AND H. W. R. COPLAND. 1979. Habitat, abundance, and distribution of six species of shrews in Manitoba. Journal of Mammal- Submitted 24 November 1998. Accepted 24 January ogy 60:505-520. 2000. YAHNER, R. H. 1986. Microhabitat use by small mammals in even-aged forest stands. The American Mid- land Naturalist 115:174-180. Associate Editor was Robert K. Rose.

This content downloaded from 128.109.48.4 on Thu, 22 Aug 2013 13:42:39 PM All use subject to JSTOR Terms and Conditions