A Tracking Technique to Locate Small Mammals at Low Densities

A TRACKING TECHNIQUE TO LOCATE SMALL MAMMALS AT LOW DENSITIES

R. BOONSTRA, M. KANTER, AND C. J. KREBS Division of Life Sciences, ScarboroughCampus, Universityof Toronto, 1265 Military Trail, Scarborough,Ontario, Canada M1C 1A4 (RB, MK) Departmentof Zoology, The Universityof British Columbia, 6270 UniversityBoulevard, Vancouver,British Columbia, Canada V6T 2A9 (CJK)

To optimize our livetrappingof low-density populations of lemmings (Dicrostonyx kilangmiutak) and voles (Microtus oeconomus) (0.2-0.4 animals/ha) in the Arctic tundra, we developeda powdered-slidetracking technique to determineif burrowswere occupied.Paint- ed microscopeslides coveredwith talcumpowder on a thin oil base were placed at all burrows on trappinggrids, and only those burrowsshowing activitywere trapped. Our capturesuccess per trap tripled with this method in comparisonwith standardgrid-stake trapping,and our effort per grid decreased by >50% after all burrowswere initially located. This technique also should be useful in similar arid environmentssuch as deserts, for obtaining relative indices of density, and for following toe-clipped animals.

Key words: Dicrostonyx, Microtus, tracking techniques, Arctic

Low-density populations of small mam- long lasting 1-2 years). Some burrowsin mals are difficult to study because the the Churchill,(_ Manitoba, region appear to amount of data obtained per unit effort is last > 15 years (R. J. Brooks, pers. comm.). low (e.g., Mihok et al., 1988). Hence, there Though in some cases it is obvious that a is a tendency to study populations at times burrowis occupied because of fresh digging when they are abundant or only in good- (Brooks and Banks, 1971), radiotelemetry quality habitat. Two major problems con- and livetrappingindicate that burrowsoften front an investigator working on low-den- were occupied without any observablefresh sity populations. First, it may be difficultto sign, especially when newly weaned young locate the few active sites in a wide expanse are present in the population. We needed a of vacant habitat. This problem is com- simple tracking technique to indicate the pounded when there may be ample evi- presence of animals. Three methods gen- dence (e.g., burrow sites, runways)that the erally have been used to record the foot- animals have been there, but because of a printsof small mammals on trackingboards: recent population decline most sites are in- a smoked-paper technique (Johns, 1979; active. Second, even after active sites have Justice, 1961; Marten, 1972; Sheppe, 1965), been located, frequent movements may metal plates coated with liquid talcum pow- make repeated live capture difficult. Both der in silicone (Brown, 1969), and white problems plagued us in a study of the col- tiles, partially covered with a mixture of lared lemming, Dicrostonyx kilangmiutak, printingink and mineral spirits (Lordet al., and the tundra vole, Microtus oeconomus. 1970). With all of these, the production of Burrows of lemmings and voles are ob- tracking boards can be time consuming, vious throughout the Arctic tundra, and messy, and produce a strong odor (Lord et lemming burrowsin particularappear to be al., 1970). We developed a simple tracking

J. Mamm., 73(3):683-685, 1992 683 684 JOURNALOF MAMMALOGY Vol.73, No. 3 technique to identify active burrows and, and were marked with surveyor's flags. Slides when combined with livetrapping,we max- were then placed inside the entranceto each bur- imized our efficiency in studying small row (the width of the slide being slightlyless than mammals in the Arctic tundra of northern the diameter of the burrow)and left overnight. Canada. The next day, all slides were removed and those sites with positive sign receiveda Longworthtrap MATERIALS AND METHODS and a flag of a different color. On permanent trappinggrids, those burrowswithout active sign Since 1987, we have conducted population retained the original flag, and thus there were studies of collared and tundravoles at lemmings flagsof two colors presentat active burrowsand Pearce Northwest Canada Point, Territories, one at inactive burrows.All burrowscould 12204 This site is on the Arctic flag (69049'N, I'W). be reassessed with slides coast and is characterized again rapidly powdered by coarse, shallow, priorto a On removalar- calcareoussoils a subsequenttrapping. slightly supporting Dryas-Salix eas, flagswere removed from inactive burrows. arctica discontinuous tundra (Ritchie et al., 1987).We establishedlivetrapping grids in a row- RESULTS AND DISCUSSION and-columnfashion with trappingstations spaced 30.5 m apart.All gridswere trappedweekly from In 1987, the trappingeffort was labor in- Junethrough August. In 1987, we used only stan- tensive, occupying one person's time for 2 dard livetrappingtechniques, placing one Long- days/week, yet was unproductive. Trapping worth 5 m from trap 5 everysecond station. Traps success during the first 6 weeks (number of were moved to the adjacent station in alternate animals session weeks so that area caught/trap/2-daytrapping the entire was trappedevery x was low = = n = 2 weeks. 100) (I 9.0%, SE 1.9, In we In 1988, because of extremely poor trap suc- 13). subsequent years, trapped only at sites as cess the previousyear, we abandonedthe scheme active determined by the pow- of trappingat grid stakesand insteadtrapped dered-slide trackingtechnique. only at active burrow sites. These were deter- We presentonly the data collectedin 1989 mined by a powdered-slidetechnique we devel- on four grids to indicate the efficacyof this oped that was simple, inexpensive, and rapid, technique.These gridsaveraged 23.3 ha (SE and one in which the trackingsurfaces, slides, fit = 2.7; range, 18.6-28.4) in size. Though the nicely inside small-mammal burrows. Unfrost- initial effort to locate all burrowswas time ed, glass microscope slides (2.5 by 7.5 cm) were consuming 18.0 h/grid;SE = 2.8; range, on both sides a (X= spray painted with black, matte 14-26), of enamel Slides were then coated with a thin subsequent placing powdered paint. slides at burrowsites that had been filmof oil, into of talcum already light dipped ajar pow- located reduced the effort > 6 der, and prior to being taken to the field, were by 50% ( = SE = n these stored in microscope slide boxes containing 25 h/grid; 1.2; = 3; range, 4-8; or 100 positions;we found that the smaller,plas- data were recorded for only three grids). tic slide boxes were more convenient in the field. These three grids had an average of 418.8 We found that the best oil producinga uniform burrows/grid(SE = 54.4; range, 325-575), film on the slide was machinist's oil (Tellus Oil of which 26.6% (SE = 3.8; range, 10-37) 10, Shell Company). Oil containing detergents were deemed active, and each of which re- (e.g., standardmotor oils) were less satisfactory ceived a livetrap. The averagetrapping suc- as caused on the slide and resulted they beading cess increased to 31.9% (SE = 3.8; n = 7; in an uneven talcum coat. A film of oil was ap- and hence was > 3 x that ob- a soft cloth in The range, 14-42) plied by saturated the oil. tained with We same was used for new slides and for simple grid-staketrapping. procedure assumed that the densities of small mam- soiled slides brought in from the field. In one stroke, the cloth both cleaned the soiled slides mals in these 2 years were comparable, and and applied a new oil film. One person was able that the lower success in 1987 was related to prepareup to 300 slides/h. to poor trap placement. Population demog- All burrowswithin the study areawere located raphy for the other years (1988-1990) was by a systematic search at the start of the season similar, with low densities occurring every August1992 BOONSTRAET AL.--TRACKING SMALL MAMMALS AT LOWDENSITIES 685 summer (0.2-0.4/ha), and we suggest that ACKNOWLEDGMENTS 1987 was a replicate of subsequent years. We thank A. Kenney, J. Serensits, D. Cluff, Thus, lemmings and voles at these low den- and D. Gamier for providing assistance in the sities and with low vegetative cover do not Arctic, the NaturalSciences and EngineeringRe- appear to seek out the traps, and trapping search Council of Canada and the Polar Conti- at active burrow sites is the most efficient nental Shelf Projectof the Departmentof Energy method to study them. and Mines, Canada,for financialassistance, and There are a number of reasons why trap- R. J. Brooksfor criticalcomments on the manu- ping success still remained at about one- script. third of the traps set at active burrow sites. First, though most burrow systems usually LITERATURECITED contained only one animal, most burrow ANDRZEJEWSKI,R., H. FEJGIN,AND A. LIRO. 1971. had two or more entrancesand Trappabilityof trap-proneand trap-shybank voles. systems traps Acta Theriologica, 16:401-412. were placed at each of these. Second, some BOONSTRA, R., C. J. KREBS, AND M. KANTER. 1990. animals were trap-shy (Andrzejewskiet al., Arctic ground squirrel predation on collared lem- Tanaka, as mings. CanadianJournal of Zoology, 68:757-760. 1971; 1980; Tanton, 1965) they BROOKS,R. J., ANDE. M. BANKS.1971. Radio-track- were known to be in the burrow because ing study of lemminghome range.Communications they carried a radiocollar from a previous in BehavioralBiology, 6:1-5. capture et al., 1990). some BROWN,L. E. 1969. Field experimentson the move- (Boonstra Finally, ments ofApodemussylvaticus L., using trappingand burrows probably were misclassified as ac- trackingtechniques. Oecologia (Berlin), 2:198-222. tive because other factors, such as wind CAUGHLEY, G. 1977. Analysis of vertebrate popula- over the slide or spiders tions. John Wiley and Sons Ltd., Toronto, Canada, moving grass large 234 pp. crawling over the slide, made scratches on JOHNS,A. D. 1979. A comparative assessment of the slides. methods of individual trackingwithin a population We have used this to of Microtusagrestis (Mammalia: Muridae). Journal technique simply of Zoology (London), 189:333-338. indicate presenceor absence of animals pri- JUSTICE,K. E. 1961. A new method of measuring or to livetrapping,and thus all demographic home ranges of small mammals. Journal of Mam- information (densities, survival, recruit- malogy, 42:462-470. LORD, R. D., A. M. VILCHES, J. I. MAIZTEGUI, AND C. ment) was still based on actual capture of A. SOLDINI.1970. Tracking board: a relative census animals. The techniqueoptimized our effort technique for studying rodents. Journal of Mam- in this low-density environment. This tech- malogy, 51:828-829. MARTEN,G. G. 1972. Censusing mouse populations nique also should be valuable in other open by means of tracking.Ecology, 53:859-867. areas such as sagebrush flats, arid grass- MIHOK,S., T. LAWTON,AND B. SCHWARTZ. 1988. Fates lands, or other area in which burrow and movements of meadow voles (Microtuspenn- any sylvanicus)following a populationdecline. Canadian sites or runways are readily obvious. The Journalof Zoology, 66:323-328. technique could be used independently of RITCHIE, J. C., K. A. HADDEN, AND K. GAJEWSKI. to relative indices of 1987. Modern pollen spectra from lakes in Arctic livetrapping give pop- western Canada. Canadian Journal of Botany, 65: ulation density (Caughley,1977). A number 1605-1613. of studies have reported using powdered- SHEPPE,W. 1965. Characteristics and uses of Pero- techniques to myscus trackingdata. Ecology, 46:630-634. tracking individually identify TANAKA,R. 1980. Controversial problems in ad- toe-clipped animals (Brown, 1969; Justice, vanced researchon estimating population densities 1961; Sheppe, 1965). The thin coat of tal- of small rodents.Researches on PopulationEcology, cum on the slides provides an ideal surface Supplement,2:1-67. TANTON,M. T. 1965. Problems of live-trapping and for recordingthe fine detail of footprintsand population estimation for the wood mouse, Apode- hence should prove useful for these pur- mus sylvaticus(L.). The Journalof Animal Ecology, 34:1-22. poses. In conjunction with toe clipping, the powdered-slidetechnique could provide an Submitted10 June 1991. Accepted9 October1991. indication of the presence and number of unmarked individuals in the population.