BioOne sees sustainable scholarly publishing as an inherently collaborative enterprise connecting authors, nonprofit publishers, academic institutions, researchlibraries, and research funders in the common goal of maximizing access to critical research.

AN ALTERNATIVE FIELD TECHNIQUE FOR ESTIMATING DIVERSITY OFSMALL-MAMMAL POPULATIONSAuthor(s): Michale J. Glennon , William F. Porter , and Charlotte L. DemersSource: Journal of Mammalogy, 83(3):734-742. 2002.Published By: American Society of MammalogistsDOI: http://dx.doi.org/10.1644/1545-1542(2002)083<0734:AAFTFE>2.0.CO;2URL: http://www.bioone.org/doi/full/10.1644/1545-1542%282002%29083%3C0734%3AAAFTFE%3E2.0.CO%3B2

BioOne (www.bioone.org) is a nonprofit, online aggregation of core research in the biological, ecological, andenvironmental sciences. BioOne provides a sustainable online platform for over 170 journals and books publishedby nonprofit societies, associations, museums, institutions, and presses.

Your use of this PDF, the BioOne Web site, and all posted and associated content indicates your acceptance ofBioOne’s Terms of Use, available at www.bioone.org/page/terms_of_use.

Usage of BioOne content is strictly limited to personal, educational, and non-commercial use. Commercial inquiriesor rights and permissions requests should be directed to the individual publisher as copyright holder.

734

Journal of Mammalogy, 83(3):734–742, 2002

AN ALTERNATIVE FIELD TECHNIQUE FOR ESTIMATINGDIVERSITY OF SMALL-MAMMAL POPULATIONS

MICHALE J. GLENNON,* WILLIAM F. PORTER, AND CHARLOTTE L. DEMERS

State University of New York, College of Environmental Science and Forestry, 1 Forestry Drive,Syracuse, NY 13210 (MJG, WFP)

Adirondack Ecological Center, 6312 Route 28N, Newcomb, NY 12852 (CLD)

Sampling of populations of small mammals has traditionally focused on use of live traps,which are often costly and labor intensive. We tested track tubes as an alternative techniquefor assessing populations of small mammals. Track tubes record footprints of small mam-mals and allow determination of their presence or absence without live capture. We com-pared results from livetrapping against data from track tubes on 5 sites over 1-week periodsin June 1999 and June 2000. Correlations between indices of abundance from the 2 tech-niques were significant in 1999 (rs 5 0.656, P , 0.001) and 2000 (rs 5 0.715, P , 0.001).Using track tubes we distinguished footprints of 4 species. We were not able to distinguishPeromyscus from Clethrionomys gapperi; species of Sorex could not be distinguished. Incomparison with livetrapping, track tubes are inexpensive, are much less labor intensivefor researchers, and can be run simultaneously on several sites. The technique has goodpromise where investigators seek only to identify composition and relative abundance ofsmall-mammal species.

Key words: abundance, livetrapping, richness, small mammals, track tubes

We tested an alternative technique to as-sess population attributes of small-mammalspecies by using track tubes to record foot-prints (Drennan et al. 1998). Studies ofsmall mammals often involve livetrappingfor determining population parameters (Ca-rey and Johnson 1995; Fitzgibbon 1997;Morris 1984; Songer et al. 1997; Yahner1992). Commonly, study areas are sampledwith a grid or transect of live traps spacedat particular intervals, with $1 traps placedat each station. Livetrapping of animals isoften intended, in part, to gain informationon species composition in an area (Chase etal. 2000; Loeb 1999; Menzel et al. 1999).These data may be vital in answering par-ticular questions. However, effort requiredto collect data is often substantial. Investi-gators recommend that trapping be con-ducted for a minimum of 4–5 nights (Olsen

* Correspondent: mjglenno@syr.edu

1975; Steele et al. 1984) at each location todepict accurately the dynamics of smallmammals inhabiting the area. Live traps arealso relatively expensive ($10–25 per trapdepending on type and manufacturer). Live-trapping efforts, consequently, require asignificant investment of time, energy, andfunds depending on the scope of the re-search.

Track tubes represent an alternative tolivetrapping by providing some of the sameinformation but at greatly reduced effortand cost. Such track approaches have beenused frequently for assessing furbearerabundance (Barrett 1982; Taylor and Ra-phael 1988; Zielinski and Truex 1995) andoccasionally in investigations of smallmammals (Drennan et al. 1998; MaBee1998; Van Apeldoorn et al. 1993). Tracktubes record presence of species of smallmammals through footprints left on a track

GLENNON ET AL.—SMALL-MAMMAL TRACKINGAugust 2002 735

surface inside the tubes. Track tubes are in-expensive (,$2.00 per tube), are light-weight, allow for simultaneous sampling ofsmall-mammal populations over several ar-eas, and can reduce risk of exposure of re-searchers to hantavirus because handling ofanimals is not required (Mills et al. 1995).Track tubes need only be examined onceevery 48 h. This technique makes broad-scale sampling of species of small mam-mals more feasible.

We tested track tubes against a more tra-ditional livetrapping approach on 5 areas innorthern New York. Our goals were to de-termine whether track tubes allow for esti-mates of relative abundance comparablewith those obtained from livetrapping andwhether track tubes detect the same diver-sity of small-mammal species as those cap-tured in live traps.

MATERIALS AND METHODS

Study area.—Our study was conducted atHuntington Wildlife Forest in Newcomb, NewYork, in the central Adirondack Park (448009N,748139E). Adirondack Park is 2.5 3 106 ha inarea and is characterized by numerous hills,lakes, and rounded mountains with a climax for-est of spruce, fir, and northern hardwoods. Ele-vations range from 30 to 1,600 m. HuntingtonWildlife Forest encompasses 6,000 ha of re-search forest operated by College of Environ-mental Science and Forestry, State University ofNew York. This property consists mainly of2nd-growth northern hardwoods (72%), such asAmerican beech (Fagus grandifolia), yellowbirch (Betula alleghaniensis), white birch (B.papyrifera), sugar maple (Acer saccharum), andred maple (A. rubrum). The remaining 28% iscomposed of mixed hardwood–softwood andsoftwood stands with common species, includ-ing white pine (Pinus strobus), balsam fir (Abiesbalsamea), red spruce (Picea rubens), and east-ern hemlock (Tsuga canadensis—Tierson et al.1985). Soils are mostly glacial tills, and topog-raphy is mountainous. Average annual precipi-tation is 102 cm with 307 cm of annual snowfall.Average temperatures range from 208C in Julyto 288C in January (R. L. Masters, in litt.).

We sampled small mammals by use of live-trapping and track tubes on 5 sites in Huntington

Forest: Electric Fence, Hare Area, Maple Sale,Natural Area, and Sucker Brook. These sitesvary in terms of forest-management history butare all representative of the northern-hardwoodcomplex characterizing the central Adirondacks.The Natural Area is a stand of old-growth north-ern hardwoods from which no timber has beenextracted, with the exception of a small amountof white pine and spruce in the early 1900s. Theother 4 study sites have been subjected to anumber of different harvest treatments and arerepresentative of several different age- and size-class structures.

Live traps.—We conducted livetrapping ofsmall mammals on the 5 study sites over 4 con-secutive days and nights during June 1999 and2000. Livetrapping has been conducted on thesesites for 11 years as part of the Adirondack LongTerm Ecological Monitoring Program atHuntington Wildlife Forest (Adirondack Ecolog-ical Center, in litt.). Each site consisted of a 7-by-7 trapping grid, with 20-m spacing betweentrap stations (1.44-ha grid). Traps used wereTomahawk wire cage traps (40.6 by 12.7 by 12.7cm) and Sherman traps (7.6 by 8.9 by 22.8 cm).We placed the 2 trap types at alternate stationson the grid, resulting in a total of 25 Tomahawktraps and 24 Sherman traps. We conducted trap-ping during 4 nights and 4 days for a total of392 trap nights at each site. Traps were baitedwith a mixture of peanut butter, rolled oats, andfood-grade paraffin (H. B. Underwood, pers.comm.). Paraffin was used as a bait additive be-cause of its advantage in reducing melting of thebait and facilitating preparation of traps andtrack tubes. A small amount of polyester fill wasplaced in each of the Sherman traps for use asnesting material.

We examined traps every morning and everyevening over a 5-day period. Captured animalswere examined to determine sex and reproduc-tive condition, weighed, and marked with metalear tags (type 1005-1; National Band & TagCompany, Newport, Kentucky) or by fur clip-ping for species lacking external ear pinnae orwith pinnae too small to mark.

Track tubes.—We adapted the design of tracktubes from Drennan et al. (1998). Tubes wereconstructed from 2 pieces of plastic rain guttercut into 30.5-cm sections and taped at 1 sidewith duct tape (Fig. 1). Tubes were approxi-mately 11.5-cm wide at the widest point and 15cm high. Tubes were taped along 1 edge only,

736 Vol. 83, No. 3JOURNAL OF MAMMALOGY

FIG. 1.—Track tube constructed from plasticrain gutter, showing Tamias striatus with bait,from Huntington Wildlife Forest, Newcomb,New York, 1999.

to allow for ease of storage and transport whennot in use. Edges that were not taped were at-tached with large binder clips when tubes wereplaced in the field. Track plates provided a tracksurface inside the tubes. We constructed trackplates from aluminum strips (30.5 by 7.6 cm)with contact paper attached to them. Clear con-tact paper (sticky side up) provided a track sur-face and was attached to white paper for a per-manent record of tracks from each tube. Clearcontact paper is analogous to a clear plastic filmand allowed tracks to be seen when white paperwas attached. We attached contact paper to alu-minum strips by folding it under the 2 short endsand leaving the sticky side facing up. The con-tact paper was not sticky enough to hindermovement or harm animals in any of the fieldtrials we observed. Felt squares (7.6 cm2) wereplaced at both ends of the track plate; theyserved as ink pads and were attached directly tothe contact paper. Bait (the same as that used forlivetrapping) was placed in the center of eachtrack plate, and the plate was then placed in thecenter of each track tube. Animals stepped ontothe felt squares upon entering tubes and, in trav-eling through, transferred ink onto the surface ofthe contact paper. Ink was made from a combi-nation of carbon black and paraffin oil in a 1:1ratio (MaBee 1998; Van Apeldoorn et al. 1993).

We placed track tubes at 49 live-trap locationson each grid during the week after conclusionof livetrapping. In 1999, track tubes were placedat all sites after a 1-week period after the con-clusion of livetrapping during which we were

not active. In 2000, no such lag was possiblebecause of logistical constraints, and tubes wereplaced at trap locations after only a 2-day periodof no activity. We did not consider reversing theorder of livetrapping and track-tube activities,both because of logistical constraints and be-cause the livetrapping effort was part of a long-established monitoring program with less flexi-ble scheduling options. We believed that a 5-dayperiod of supplemental food addition to thesesites would not be enough time for small-mam-mal abundance to rise significantly in responseto added food. Therefore, we do not believe thattrapping affected population numbers of smallmammals. Lacki et al. (1984) found that, after 4weeks of supplemental food addition, chip-munks did not alter the size of their home rangesin response to added food.

Track tubes were placed at each location onMonday, examined and rebaited on Wednesday,and examined again and removed on Friday. Werevisited each site after 48 h to collect tracks andreplace track plates. A 48-h period allowed fora reasonable number of tracks to be collected butnot so many that individual species could not bedistinguished. When collecting track plates, weremoved felt pads from each end of the trackplates and placed a strip of unlined white paperon the contact paper surface. The strips of clearcontact paper (with white paper attached in thefield) served as a permanent record of tracksfrom each tube.

We developed a reference collection of foot-prints of local species and used that as a guideto determine the species responsible for tracksobtained from track tubes. We collected foot-prints from animals captured during trapping.Those captured were released through a double-length track tube, thereby creating a record offootprints for each known species. We obtained5–20 track samples for each species to studyvariability of footprints.

Prints from track tubes were interpreted to de-termine the species sampled by track tubes.Most prints could be determined to the level ofspecies. We were able to distinguish footprintsof Blarina brevicauda, Glaucomys sabrinus,Tamiasciurus hudsonicus, and Tamias striatus.Several other species could not be distinguishedfrom one another because of extreme similarityof print characteristics. These included Napaeo-zapus insignis and Zapus hudsonius, Peromys-cus and Clethrionomys gapperi, and species of

GLENNON ET AL.—SMALL-MAMMAL TRACKINGAugust 2002 737

FIG. 2.—Relationship between livetrappingand track-tube indices for all sample sites com-bined for a) 1999 and b) 2000, at HuntingtonWildlife Forest, Newcomb, New York.

Sorex. Sorex occurring in Adirondack mountainsinclude masked (S. cinereus), smoky (S. fu-meus), water (S. palustris), and pygmy (S. hoyi)shrews; masked and smoky shrews are mostcommon. These species are distinguished easilyfrom short-tailed shrew by size of footprints, butinterspecifically they are too similar in footprintcharacteristics to be distinguished. Track-tubedata were therefore placed in the following spe-cies categories: (1) B. brevicauda, (2) G. sabri-nus, (3) N. insignis or Z. hudsonius (or both),(4) Peromyscus or C. gapperi (or both), (5) So-rex, (6) Tamiasciurus hudsonicus, and (7) Tam-ias striatus.

Analysis.—We assigned data from live cap-tures to the same categories as those used fortrack tubes, to compare directly between the 2techniques. As a result, N. insignis and Z. hud-sonius were combined into 1 class, Peromyscusand C. gapperi were combined, and species ofSorex were combined into 1 class. Live capturesof P. maniculatus and P. leucopus were not de-termined to a species level because of similarityof morphological characteristics.

We used total number of track tubes contain-ing a print of a particular species, summed overthe entire sample, as a track-tube index (Dren-nan et al. 1998). We did not make an effort todetermine if .1 individual had been responsiblefor tracks in a particular tube. We used totalnumber of captures of a particular species as theestimate from livetrapping data for comparisonwith the track-tube index. Richness was definedas total number of species at each location.

Because of nonnormal distributions of eachindex, we used a Spearman rank correlation toassess relationships between track-tube and live-trap indices in both years. Data were normalwhen considered separately for each site or spe-cies; therefore, we used Pearson correlation co-efficients to assess the relationship between live-trap and track-tube indices both within sites andwithin species across all sites.

We also tested data from track tubes againstpopulation estimates from live captures calcu-lated using the closed captures model of pro-gram MARK (White and Burnham 1999). Weused the closed captures model because our datafit this model best among the available options.The track-tube index was correlated with abun-dances estimated from MARK, which are basedon mark–recapture data from livetrapping.Abundances based on mark–recapture informa-

tion from livetrapping were calculated for chip-munks using the program MARK during 1999and 2000. Chipmunks were the only species forwhich a sufficient number of captures existed tocalculate a population estimate.

RESULTS

Live-trap and track-tube indices werepositively correlated in 1999 (rs 5 0.656, P, 0.001, n 5 35; Fig. 2a) and 2000 (rs 50.715, P , 0.001, n 5 35; Fig. 2b). Stronglinear correlations were evident betweenlive-trap and track-tube indices across allsites and species in both years.

Pearson correlations between live-trap

738 Vol. 83, No. 3JOURNAL OF MAMMALOGY

TABLE 1.—Comparison of population esti-mates for small mammals from livetrapping andtrack tubes, as indicated by Pearson correlationcoefficients (r) between livetrapping and popu-lation indices from track tubes within sites, allspecies combined. Data were collected on 5study sites in the central Adirondack Park, NewYork, during 5-day sampling periods in June1999 and June 2000.

Location

1999

r P

2000

r P

Electric FenceHare AreaMaple SaleNatural AreaSucker Brook

0.8630.9460.9220.7390.943

0.0130.0010.0030.0590.002

0.9900.8540.9980.9000.955

0.0010.0150.0010.0060.001

TABLE 2.—Sample sizes and Pearson correlation coefficients (r) between livetrapping and track-tube population indices within species, across all sites. Data were collected on 5 study sites in thecentral Adirondack Park, New York, during 5-day sampling periods in June 1999 and June 2000.Total number of trapnights was 1,960.

Species

1999

Sample size

Livetrap-ping

Tubetrack-

ing r P

2000

Sample size

Livetrap-ping

Tubetrack-

ing r P

Blarina brevicaudaGlaucomys sabrinusNapaeozapus insignis and/or Zapus hudsoniusPeromyscus and/or Clethrionomys gapperiSorexTamiasciurus hudsonicusTamias striatus

822

18123

033

468

461

17164141

59414

0.61420.250

0.9900.921

0.7850.876

0.2770.6880.0020.030

0.1220.056

214

1513

359

112

19134

1615

422

20.57820.250

0.9060.2340.6050.4770.577

0.3130.6880.0380.7070.2860.4220.314

and track-tube data within sites were strongin both 1999 and 2000 (Table 1). Strengthof correlations between the 2 indices withinindividual species was highly variable andprobably largely affected by sample size inboth years (Table 2).

Correlations between MARK abundancesand track-tube indices were high in 1999 (r5 0.816, d.f. 5 3, P , 0.092, n 5 5) butmuch lower in 2000 (r 5 0.467, d.f. 5 3,P , 0.428, n 5 5), largely because of theextremely high capture rate of chipmunksin Maple Sale. When Maple Sale was re-moved from the analysis, correlation coef-

ficient was much higher (r 5 0.951, d.f. 52, P , 0.049, n 5 4).

All species groups described previouslywere recorded by both live traps and tracktubes, with the exception of American mar-ten (Martes americana), which was detect-ed only in live traps (Table 3). Nonetheless,martens do make use of track tubes and oc-curred in other locations that were not apart of this study (M. J. Glennon, in litt.).A total of 10 species were detected by livetraps and 9 species by track tubes.

DISCUSSION

We drew 2 main conclusions from thetest of the track-tube technique on 5 sitesduring summers of 1999 and 2000. First,data from track tubes produce estimates ofrelative abundance that are similar to thosecalculated from livetrapping. Second, tracktubes detect the same species as those cap-tured by live traps.

Estimates of relative abundance fromtrack tubes were similar to those based onlive captures. Number of animals capturedin live traps was significantly related tonumber of animals recorded by track tubes,when all sites and species were consideredtogether. These relationships were strongerwhen considered within each sampling site,indicating that track tubes may be an effec-

GLENNON ET AL.—SMALL-MAMMAL TRACKINGAugust 2002 739

TABLE 3.—Species recorded by livetrapping and using track tubes on 5 study sites in the centralAdirondack Park, New York, during 5-day sampling periods in June 1999 and June 2000.

Species

1999

Livetrapping

n SE

Tube tracking

n SE

2000

Livetrapping

n SE

Tube tracking

n SE

Blarina brevicaudaClethrionomys gapperiGlaucomys sabrinusMartes americanaNapaeozapus insignisPeromyscusSorex cinereusTamiasciurus hudsonicusTamias striatusZapus hudsoniusUnknown

8285

21

1738

33468

1

4.236.540.4002.442.09

4.1215.91

0

46164a

1

17b

164a

141c

5941417b

134

2.487.560.20

3.167.566.656.585.633.164.27

26124913

359

0.250.970.200.240.580.6600.40

22.91

11134a

2

19b

134a

16c

1542219b

38

1.206.750.40

3.116.751.161.058.663.110.75

a C. gapperi and Peromyscus are combined in 1 category.b N. insignis and Z. hudsonius are combined in 1 category.c All Sorex are combined in 1 category.

tive way of assessing differences in small-mammal communities between differentstudy locations or habitat types.

We did not make an adjustment for effortin determining an index of abundance foreither technique, to provide the most directcomparison between the 2 methods. Al-though livetrapping data are often correctedfor effort (i.e., catch per unit effort—Wilsonet al. 1996), amount of effort for both tech-niques was relatively similar. In both in-stances there were 49 traps or tubes avail-able during each sampling period, and bothwere sampled over periods of 4 days and 4nights. Although live traps represent ‘‘sin-gle-catch’’ traps and track tubes do not,track tubes suffer to some degree from thesame reduced effort caused by single-catchtraps (i.e., sprung traps reduce the totalnumber of traps left available). For exam-ple, because individual animals cannot bedistinguished by footprint characteristics,once a particular species has been recordedin a tube, the record counts for a positiveidentification of the presence of that speciesin each tube. Additional individuals of thesame species add no further information tothe index.

When considering each species separate-

ly, the relationship between track-tube andlive-trap indices was highly variable. In1999 most species occurred in the same rel-ative proportions with the use of each tech-nique. The exceptions were shrews, whichoccurred in very low abundances. In 2000the abundance of jumping mice was strong-ly correlated between the 2 techniques, butabundances of several other species werenot.

Low trappability of some species underthis particular design may be responsible,in part, for the poor agreement betweentrack-tube and live-trap indices for partic-ular species. Shrews are difficult to capturein Sherman live traps compared with spe-cies such as red-backed voles and chip-munks, and they often suffer high mortalityin traps (Little and Gurnell 1989; Yunger etal. 1992). Additionally, clumped distribu-tions of flying squirrels may require differ-ent trapping grid designs for effective sam-pling of these species. Modifications of ourlivetrapping design may have been neces-sary to reflect more accurately the abun-dance of these groups.

Chipmunks are another likely cause forthe poor agreement between track-tube andlive-trap indices, especially in 2000. Chip-

740 Vol. 83, No. 3JOURNAL OF MAMMALOGY

munks were the species captured most fre-quently in live traps in both years of thisstudy. Chipmunks are captured readily,which may prevent other species from be-ing caught because of the single-catch na-ture of live traps. Though chipmunks arediurnal and, therefore, have activity patternsdifferent from those of other species, theyare frequently detected in live traps thathave already been examined by researchersduring evening checking and have been leftopen for the night session (C. L. Demers,in litt). Some individuals may be willing toalter their diurnal behavior patterns to somedegree in order to obtain additional food.Track tubes may be more effective for sam-pling species that occur in low abundance,in those situations where the presence ofchipmunks may affect the ability to captureother species. Although other small-mam-mal species are known to travel throughtrack tubes even after baits have been re-moved, the likelihood that they will do somay be reduced. At 1 site (Maple Sale),which consistently had higher capture ratesfor chipmunks than did the other 4 locations(C. L. Demers, pers. comm.), number oflive captures of chipmunks was higher thannumber recorded in track tubes in 1999. Inall other sites and in both years, the oppo-site was the case; track tubes generally re-cord higher rates of abundance becausethey are not single-catch traps and oftencontain tracks of .1 species of small mam-mal.

Chipmunks are also the likely cause oflow correlations between MARK estimatesand track-tube estimates in 2000. The over-whelming number of chipmunk captures atthe Maple Sale site makes it an outlier andcauses the relationship between live andtrack-tube indices to break down becauselive captures of chipmunks were dispropor-tionately high at that site. The cause of dis-proportionately high capture rates in thissite is not known. When the Maple Sale sitewas removed from the analysis, correlationbetween the track-tube index and MARKestimates increased from r 5 0.467 to r 5

0.951. In this instance, low correlations be-tween the 2 indices might be expected be-cause of more accurate sampling of thosespecies not often detected by live traps. Oneassumption of the closed captures model ofprogram MARK is that all animals have anequal probability of capture (Otis et al.1978). Our alternate placements of traptypes may have affected our ability to meetthis assumption and affected MARK esti-mates of chipmunk abundance. Chipmunksare caught more often in Tomahawk trapsthan in Sherman traps (C. L. Demers, inlitt.), and therefore an individual whosehome range encompasses a Tomahawk trapmay have a higher likelihood of capturethan would one whose home range encom-passes a Sherman trap. Despite this poten-tial problem, however, agreement betweenchipmunk abundances calculated byMARK and those estimated from tracktubes was generally good.

The 2nd main conclusion of this study isthat track tubes recorded the same speciesas those captured using traditional livetrap-ping techniques. The only exception to thispattern was the American marten, which isnot common in live traps or track tubes butenters both (M. J. Glennon, in litt.). All oth-er species were sampled using both tech-niques, some more effectively by tracktubes.

The results of our study were consistentwith those of Drennan et al. (1998). Dren-nan et al. (1998) also found that relativeabundance of small mammals, as indexedby track stations, was significantly correlat-ed with independent measures of populationsize based on livetrapping. These research-ers were able to take advantage of 2 differ-ent ongoing studies, resulting in larger sam-ple sizes than we were able to obtain. Theresults of our study agree with theirs andhave increased the scope of the techniqueby examining small-mammal species in ad-dition to sciurids. Drennan et al. (1998)were able to use a mixture of alcohol andcarpenter’s chalk sprayed onto aluminumplates for obtaining footprints. Precipitation

GLENNON ET AL.—SMALL-MAMMAL TRACKINGAugust 2002 741

and moisture conditions in our study areapreclude such an approach. Use of ink andcontact paper works well in moist environ-ments and allowed us to obtain a permanentrecord of all tracks.

For rare species, track tubes probably donot represent an ideal assessment technique.Additionally, lack of data on sex, age, re-productive condition, and other morpholog-ical characteristics may preclude use oftrack tubes for many applications. Popula-tion estimates based on track tubes cannotincorporate information on mark–recapturehistory of individual animals. Relativeabundance estimates from track tubes may,therefore, be less robust than those basedon mark–recapture data.

Advantages of track tubes may outweighdisadvantages in many situations. Tracktubes are considerably less expensive thanlive traps. A single track tube costs approx-imately $1.50, whereas Sherman live trapscost $$8.25 each (H. B. Sherman Traps,Inc., Tallahassee, Florida) and Tomahawktraps $$21.00 (Tomahawk Live Trap Com-pany, Tomahawk, Wisconsin). In additionto low cost, track tubes are lightweight, du-rable, cause no mortality of study animals,and allow for sampling of several sites overa short time period. Though deciphering oftracks adds to the time required for con-ducting assessments with track tubes, over-all effort is still reduced relative to thatwhich would be required for livetrappingbecause several sites may be sampled si-multaneously.

The results of our study indicate that spe-cies of small mammals are detected equallyby track tubes and live captures. This tech-nique was examined as a potential methodfor assessing small-mammal richness andrelative abundance in different habitat typesin the central Adirondacks. Track tubes canbe effective for assessment of presence, ab-sence, and relative abundance between sitesin a broad-scale study. Therefore, tracktubes may provide an alternative samplingtechnique for small mammals for research-ers interested only in documenting species

occurrence and relative proportions in asite.

ACKNOWLEDGMENTS

Funding and facilities for this research wereprovided by New York State Department of En-vironmental Conservation, Adirondack Ecolog-ical Center, and Roosevelt Wild Life Station atthe State University of New York, College ofEnvironmental Science and Forestry. M. Baum-flek, K. Clarke, F. DeSantis, Z. Hart, J. Oldroyd,A. Peck, and E. Speith aided in collection oftrack-tube data. P. Salmon provided the track-tube photograph. An earlier version of this man-uscript was reviewed by J. F. Merritt.

LITERATURE CITED

BARRETT, R. H. 1982. Smoked aluminum track platesfor determining furbearer distribution and relativeabundance. California Fish and Game 69:188–190.

CAREY, A. B., AND M. L. JOHNSON. 1995. Small mam-mals in managed, naturally young, and old-growthforests. Ecological Applications 5:336–352.

CHASE, K. M., B. W. KRISTAN, J. A. LYNAM, V. M.PRICE, AND T. J. ROTENBERRY. 2000. Single speciesas indicators of species richness and composition inCalifornia coastal sage scrub birds and small mam-mals. Conservation Biology 14:474–487.

DRENNAN, J. E., P. BEIER, AND N. L. DODD. 1998. Useof track stations to index abundance of sciurids.Journal of Mammalogy 79:352–359.

FITZGIBBON, C. D. 1997. Small mammals in farmwoodlands: the effects of habitat, isolation, and sur-rounding land-use patterns. Journal of Applied Ecol-ogy 34:530–539.

LACKI, M. J., M. J. GREGORY, AND P. K. WILLIAMS.1984. Spatial response of an eastern chipmunk pop-ulation to supplemented food. American MidlandNaturalist 111:414–416.

LITTLE, J., AND J. GURNELL. 1989. Shrew captures androdent field studies. Journal of Zoology 218:329–331.

LOEB, S. C. 1999. Responses of small mammals tocoarse woody debris in a southeastern pine forest.Journal of Mammalogy 80:460–471.

MABEE, T. J. 1998. A weather-resistant tracking tubefor small mammals. Wildlife Society Bulletin 26:571–574.

MENZEL, M. A., W. M. FORD, J. LAERM, AND D. KRI-SHON. 1999. Forest to wildlife opening: habitat gra-dient analysis among small mammals in the southernAppalachians. Forest Ecology and Management 114:227–232.

MILLS, J. N., ET AL. 1995. Guidelines for working withrodents potentially infected with hantavirus. Journalof Mammalogy 76:716–722.

MORRIS, D. W. 1984. Patterns and scale of habitat usein two temperate-zone, small mammal faunas. Ca-nadian Journal of Zoology 62:1540–1547.

OLSEN, R. W. 1975. Length of trapping period in pop-ulation studies. Journal of Mammalogy 56:696–697.

742 Vol. 83, No. 3JOURNAL OF MAMMALOGY

OTIS, D. L., K. P. BURNHAM, G. C. WHITE, AND D. R.ANDERSON. 1978. Statistical inference from capturedata on closed animal populations. Wildlife Mono-graphs 62:1–35.

SONGER, M. A., M. V. LOMOLINO, AND D. R. PERAULT.1997. Niche dynamics of deer mice in a fragmented,old-growth-forest landscape. Journal of Mammalogy78:1027–1039.

STEELE, B. B., R. L. BAYN, JR., AND C. VAL GRANT.1984. Environmental monitoring using populationsof birds and small mammals: analyses of samplingeffort. Biological Conservation 30:157–172.

TAYLOR, C. A., AND M. G. RAPHAEL. 1988. Identifi-cation of mammal tracks from sooted track stationsin the Pacific Northwest. California Fish and Game74:4–15.

TIERSON, W. C., G. F. MATTFELD, R. W. SAGE, JR., AND

D. F. BEHREND. 1985. Seasonal movements andhome ranges of white-tailed deer in the Adiron-dacks. Journal of Wildlife Management 49:760–769.

VAN APELDOORN, R., ET AL. 1993. Footprints of smallmammals: a field method of sampling data for dif-ferent species. Mammalia 57:407–422.

WHITE, G. C., AND K. P. BURNHAM. 1999. ProgramMARK: survival estimation from populations ofmarked animals. Bird Study 46 Supplement:120–138.

WILSON, D. E., F. R. COLE, J. D. NICHOLS, R. RUDRAM,AND M. S. FOSTER (EDS.). 1996. Measuring and mon-itoring biological diversity: standard methods formammals. Smithsonian Institution Press, Washing-ton, D.C.

YAHNER, R. H. 1992. Dynamics of a small mammalcommunity in a fragmented forest. American Mid-land Naturalist 127:381–391.

YUNGER, J. A., R. BREWER, AND R. SNOOK. 1992. Amethod for decreasing trap mortality of Sorex. Ca-nadian Field-Naturalist 106:249–251.

ZIELINSKI, W. J., AND R. L. TRUEX. 1995. Distinguish-ing tracks of marten and fisher at track-plate stations.Journal of Wildlife Management 59:571–579.

Submitted 22 May 2001. Accepted 12 December 2001.

Associate Editor was John G. Kie.