COMPARISON OF TRAPS AND BAITS FOR CENSUSING SMALL IN NEOTROPICAL LOWLANDS

NEAL WOODMAN, ROBERT M. TIMM, NORMAN A. SLADE, AND TERRY J. DOONAN

Museum of Natural History and Department of Systematics and Ecology, University of Kansas, Lawrence, KS 66045 Present address of NW: Department of Biology, Southwestern College, 100 College Street, Winfield, KS 67156-2499 Present address of TJD: Florida Game and Fresh Water Fish Commission, Route 7, Box 440, Lake City, FL 32055

Snap-traps, live-traps, and baits affect the ability to capture small mammals, but few pre- vious studies have involved sampling communities of small mammals in tropical environ- ments. We tested differences in captures of small marsupials and by Victor snap- traps versus Sherman live-traps and by two types of bait in lowland rainforest at Reserva Cuzco Amaz6nico, southeastern . Snap-traps took ca. 3.5 times as many individuals as live-traps. Snap-traps also captured more (and more rare species), but we attribute this to more numerous captures overall because the relative proportions of species captured by the two traps generally were the same. Type of bait had little impact on our trapping results.

Key words: bait, community structure, Neotropics, Peru, Sherman traps, small mammals, snap-traps

Biologists long have recognized the tre- data covering several decades of species mendous diversity of organisms found in discovery in lowland, tropical rainforest at the tropics. Recent awareness of the rapid La Selva, Costa Rica, Timm (1994) con- disappearance of tropical rainforests (My- cluded that erroneous comparisons easily ers, 1988) has increased interest in under- are (and have been) made when investiga- standing and conserving this diversity (Wil- tors have an inadequate picture of mam- son, 1988, 1992). In recent years, the bio- malian community diversity because of in- logical community has attempted to charac- complete sampling. terize the diversity of particularorganisms at The majority of species of mammals are various tropical sites, and to compare and small (<0.5 kg), nocturnal, and secretive. contrast this diversity among sites (Gentry, Therefore, attempts to estimate mammalian 1990; McDade et al., 1994; Peters and Hut- community structure and population density terer, 1990). Comparisons among sites in the must include an effective regime for sam- tropics and also between tropical and tem- pling small mammals. There is a rich, albeit perate sites are valuable in understandingthe mostly anecdotal, literature discussing the structure and complexity of ecosystems, effectiveness of standard traps in sampling which in turn is essential for implementing populations of small mammals in temperate effective conservation efforts. regions. In many habitats, live-traps and Our ability to sample communities at snap-traps sample small commu- various sites accurately is critical to studies nities unevenly (Cockrum, 1947; Duran and of mammalian diversity. However, few re- Samz, 1973; Goodnight and Koestner, searchers have questioned the completeness 1942; Hansson and Hoffmeyer, 1973; Piz- of sampling at a particular site prior to mak- zimenti, 1979; Sealander and James, 1958; ing comparisons between sites. Based on Wiener and Smith, 1972), and studies have

Journal of Mammalogy, 77(1):274-281, 1996 274 February 1996 WOODMAN ET AL.-COMPARISON OF TRAPS AND BAITS 275

shown that the two types of traps may pro- provide a better understanding of the vide significantly different estimates of strengths and shortcomings of standard weight (Neal and Cock, 1969; Pizzimenti, sampling techniques for small mammals 1979), sex, or age structure (Galindo-Leal, and will assist future investigators in plan- 1990) of populations as well as communi- ning studies of biodiversity and community ties. Differences in captures also have been structure of small mammals in the tropics. shown between different models of snap- traps (Neal and Cock, 1969; Pizzimenti, MATERIALS AND METHODS 1979; Smith et al., 1971; Wiener and Smith, The Reserva Cuzco is and different sizes and de- study site, Amaz6nico, 1972) among located the north bank of the Rio Madre of and along signs live-traps (Dalby Straney, de Dios 14 km E of Puerto Maldonado,Tam- O'Farrell et 1976; Morris, 1968; al., 1994; bopata Province, Madre de Dios Department, Quast and Howard, 1953; Sealander and southeasternPeru (12o33'S, 69003'W). This is James, 1958; Slade et al., 1993). Addition- ca. 50 km west of the Bolivian border.Elevation ally, choice of bait used in trapping may of Cuzco Amaz6nico is ca. 200 m, and there is affect rates of capture significantly (Beer, little topographic relief. Vegetation in the ca. 1964; Buchalczyk and Olszewski, 1971; 10,000-ha reserve is typical of the region and Fowle and Edwards, 1954; Laurance, 1992; consists primarilyof evergreen, lowland, tropi- Patric, 1970; Rickart et al., 1991; Willan, cal forest. Following the Holdridgesystem (Tosi, Cuzco Amaz6nicois transitionalbetween 1986). the of data on 1960), Despite availability humid forest and forest.The effectiveness and biases of for tropical dry tropical traps captur- climate of the is seasonal with a small mammals in commu- region rainy ing temperate season from Octoberthrough March and a pro- nities et and references (Slade al., 1993, nounceddry season from April throughSeptem- therein), there have been few rigorous com- ber. Details of the Cuzco Amaz6nico site, in- parisons of effectiveness of traps or baits in cluding maps, were provided by Duellman and the lowland Neotropics. Koechlin (1991) and Woodmanet al. (1995). As part of the Neotropical Biological Di- Trapping was conducted along one trail (E versity Program (BIOTROP) of The Uni- trail) within each of two 500 by 500-m study versity of Kansas Museum of Natural His- zones for 12 consecutive days, after which traps were moved to a second trail within tory, and working in collaboration with the (U trail) each zone. We did this the season Museo de Historia Natural "Javier Prado," during dry 1989) and the wet sea- Lima, and the Missouri Botanical Garden, (June-July, again during son 1990). The zones were di- we had the to and test (January-March, opportunity develop vided into 20 by 20-m quadrats,and one trap- a for the known protocol surveying poorly ping station was placed within each of the 25 mammalian community at a lowland tropi- quadratsalong each trail.Trap stations consisted cal rainforest site-Reserva Cuzco Ama- of two Victor snap-type rat traps (8.7 by 17.5 z6nico-in the western Amazon Basin of cm) and two collapsible Sherman live-traps (8 extreme southeastern Peru. The original by 9 by 23 cm); one of each type of trap was goals of our fieldwork were to document placed on the ground and one of each was the mammalian diversity of the site and to placed above the ground (0.25-2.80 m). Addi- tional details on provide important specimens of relatively study zones, trappingmethods, and results were Woodman et al. rare species for systematic studies (Wood- provided by (1991, 1995). All were re- man et al., 1991). were collected captured Specimens moved, as voucher and de- in a fashion that prepared specimens, systematic provided quan- posited at the Museo de HistoriaNatural "Javier titative data regarding the structure of the Prado"or The Universityof KansasMuseum of small mammal community (Woodman et NaturalHistory. al., 1995). Here we compare effectiveness We used two types of bait; suet bait (finely of snap-traps and live-traps and two differ- ground beef suet, ground raisins, rolled oats, ent baits in the lowland tropics. Our results millet, grain sorghum,cracked corn, vanilla) and 276 JOURNAL OF MAMMALOGY Vol. 77, No. 1

TABLE 1.-Captures of small mammals at Cuzco Amazdnico by type of trap. Taxa are listed by total numbers of captures.

Dry season (1989) Wet season (1990) Species Snap Live Snap Live Total capito 69 11 83 24 187 Oryzomys nitidus 33 5 41 16 95 bicolor 13 4 36 4 57 Micoureus regina 12 12 18 3 45 Marmosops noctivagus 8 6 11 4 29 Proechimysa 14 1 7 5 27 Mesomys hispidus 3 4 8 1 16 Rhipidomysb 1 0 7 4 12 Oecomys robertic 4 0 3 0 7 Oryzomys yunganus 2 2 1 2 7 spinosus 2 0 3 1 6 Metachirus nudicaudatusd 5 0 0 0 5 Didelphis marsupialisd 0 0 2 1 3 Marmosops parvidens 2 0 1 0 3 Marmosa murina 0 0 2 0 2 0 0 0 1 1 1 0 0 0 1 Isothrix bistriatad 0 1 0 0 1 Sciurus ignitusd 1 0 0 0 1 Total 170 46 223 66 505

a Includes P. brevicauda, P. simonsi, and P. steerei. bA single, undescribed species (C. Tribe, pers. comm.). c We follow Musserand Carleton(1993) in treatingthe mid-sized Oecomysof the westernAmazon as 0. roberti,rather than 0. tapajinusas reportedby Woodmanet al. (1991). dOnly juveniles of these species were captured. peanut-butter bait (locally purchased peanut but- distribution. In all statistical analyses, a proba- ter and rolled oats). During the dry season, suet bility of P ? 0.05 was considered significant. bait was used for 6 days followed by 6 days of Because of small sample size, some analyses of peanut-butter bait. During the wet season, types individual species were limited to the most fre- of bait were alternated after every 2 nights, be- quently captured species, which we defined as ginning with suet bait. those with nine or more captures over both sea- We compared total numbers of captures in sons (Table 1). live-traps and snap-traps within each season to determine whether each type of trap provided RESULTS equal numbers of captures, the same community of animals, similar proportions of each of the Six species of marsupials and 15 species most-frequently captured species, or biases in of rodents were taken in live-traps and sex or body mass of captured individuals. We snap-traps set along the four trails in the made similar comparisons between the two two study zones. We captured a total of 505 of and we tested for interactions be- types bait, individuals (Table 1) during 9,600 trap- tween and bait. trap nights; our overall trap success was 5.3%. Observed frequencies of captures were com- A total of 216 individuals was captured dur- pared with distributions of traps and baits using ing the dry season and 289 in the rainy sea- chi-square goodness-of-fit tests (X2). For dichot- son. omous comparisons with n - 10 captures, we used chi-square tests with a single degree of In analyses of our trapping data, we com- freedom; when n < 10 we calculated probabil- bined three species of Proechimys (P. brev- ities of Type I errors directly from a binomial icauda, P. simonsi, and P. steerei), because February 1996 WOODMAN ET AL.-COMPARISON OF TRAPS AND BAITS 277

TABLE 2.-Captures of small mammals at Cuzco Amazdnico by type of bait. Taxa are listed by total numbers of captures.

Dry season (1989) Wet season (1990) Peanut Peanut Species Suet butter Suet butter Total Oryzomys capito 52 28 46 61 187 Oryzomys nitidus 22 16 24 33 95 12 5 17 23 57 Micoureus regina 15 9 14 7 45 Marmosops noctivagus 10 4 10 5 29 Proechimys 2 13 6 6 27 Mesomys hispidus 6 1 5 4 16 Rhipidomys 0 1 5 6 12 3 1 2 1 7 Oryzomys yunganus 3 1 1 2 7 2 0 2 2 6 Metachirus nudicaudatus 1 4 0 0 5 Didelphis marsupialis 0 0 2 1 3 Marmosops parvidens 1 1 1 0 3 Marmosa murina 0 0 2 0 2 Neacomys tenuipes 0 0 0 1 1 Oecomys superans 1 0 0 0 1 Isothrix bistriata 1 0 0 0 1 Sciurus ignitus 1 0 0 0 1 Total 132 84 137 152 505

they were difficult to distinguish ecologi- captures occurred in traps set on the cally, behaviorally, or morphologically, es- ground, and in each instance, individuals of pecially juveniles and subadults. the same species occupied both types of Type of trap.-More than three times as trap. Of 64 captures in live-traps during the many animals were captured in snap-traps dry season, 11 (14%) occurred while the (n = 393; 78%) as in live-traps (n = 112; snap-trap also was occupied. In six in- 22%), although the two types of traps were stances, both individuals were of the same distributed evenly between the zones and species. Thus, most (86-93.5%) captures in between seasons (Table 1). This pattern per- live-traps occurred when the accompanying sisted through both seasons. During the dry snap-trap was unoccupied. season, 170 (79%) animals were captured Type of bait.-During the dry season in snap-traps and 46 (21%) in live-traps (X2 there was an apparent significant difference = 72.001, d.f = 1, P < 0.01). During the between captures by the two baits; suet took rainy season, 223 (77%) individuals were more animals than did peanut butter (x2 = taken in snap-traps, and 66 (23%) were 10.180, d.f = 1, P < 0.01; Table 2). How- taken in live-traps (x2 = 85.291, d.f = 1, ever, because suet was used first for 6 days P < 0.01). of removal trapping, there were fewer ani- Because of the strong preference for mals available to be captured with peanut snap-traps, we checked whether captures in butter. Woodman et al. (1995) estimated a live-traps may have occurred only when the total population of 349 individuals available accompanying snap-trap was occupied. In along the trails in the dry season. In the 1st only three (6.5%) of the 46 captures in live- 6 days, suet caught ca. 38% of the estimat- traps during the rainy season, was the snap- ed number available, and in the subsequent trap also occupied; each of these double 6 days, peanut butter captured 39% of the 278 JOURNAL OF MAMMALOGY Vol. 77, No. 1 animals remaining. Thus, the baits actually from differential effectiveness for rare spe- were equally effective during this season. cies. This conclusion is reinforced by our finding Relative abundances of species.-Rela- no significant difference during the rainy tive abundances of the most frequently cap- season, when baits were alternated every 2 tured species taken in live-traps versus days; 137 animals were captured with suet snap-traps were not significantly different and 152 with peanut butter (x2 = 0.779, d.f. either with both seasons combined (X2 = = 1, P = 0.38; Table 2). Because they were 10.889, d.f = 7, P = 0.143), or in the rainy alternated every 2 days, comparison of baits season (x2 = 9.993, d.f. = 7, P = 0.190). was much more straight forward using the However, in the dry season there was a sig- wet-season data; hence, all additional tests nificant difference in the relative abun- of baits were restricted to data from the wet dances of species indicated by the two types season. of traps (x2 = 22.256, d.f. = 5, P < 0.01). Species richness.-More species were This difference was attributable mainly to captured in snap-traps than in live-traps. Of two species of mouse-opossums, Marmo- the 19 taxa captured, 17 were taken in snap- sops noctivagus and Micoureus regina, traps and 13 in live-traps (Table 1). Rare which tended to be captured nearly equally species of small mammals (five or fewer in both types of traps (Table 1). This was captures within a season) also were taken in contrast to other species, which had pro- more frequently in snap-traps than in live- portionally higher rates of capture in snap- traps. During the dry season, snap-traps traps. Our comparisons of relative abun- took seven rare species not captured in live- dances of the most common species trapped traps, and live-traps took one species not by suet versus peanut butter in the rainy captured by snap-traps. In the rainy season, season revealed no differences between three rare species were captured exclusively numbers of captures by the two baits (x2 = in snap-traps and one only in a live-trap. In 7.467, d.f = 7, P = 0.380). both seasons, rare species were captured ir- Body mass.-To test for size differences regularly throughout the 12-day trapping among animals captured, we compared the period along each trail (Woodman et al., distribution of classes of body mass (560, 1995). 61-140, and >140 g) between traps and be- Reduced species richness among animals tween baits without regard to species. There in live-traps could result from live-traps were no differences in classes of body mass capturing lower proportions of rare species sampled by snap-traps versus live-traps in or simply from fewer total captures (rare- either the dry (x2 = 3.252, d.f = 3, P = = faction sensu Sanders, 1968). To test these 0.350) or wet (x2 = 4.864, d.f. = 3, P hypotheses, we drew repeated random sam- 0.182) season, and no differences by baits ples, with replacement, from a population in the wet season (x2 = 1.878, d.f. = 5, P consisting of all of our 505 captures. We = 0.866). drew 500 samples each of sizes 112 and Sex.-There was no difference in sex de- 394 (the numbers of animals captured in tected between the populations sampled by live-traps and snap-traps, respectively). The snap-traps and live-traps in the dry season mean, median, and modal numbers of spe- (X2 = 0.088, d.f. = 1, P = 0.767). However, cies in our randomization experiment were in the wet season proportionally more 13 (SE = 0.065) for n = 112 and 17 (SE males (27% of all wet-season captures) than = 0.052) for n = 394. The match between females (15%) were captured in live-traps random drawings and our actual field cap- (X2 = 5.406, d.f = 1, P = 0.020). There tures suggests that the additional species was no difference in sex of animals sam- captured in snap-traps resulted solely from pled by the two baits in the wet season (x2 increased numbers of captures rather than = 0.757, d.f = 1, P = 0.384). February1996 WOODMANET AL.-COMPARISON OF TRAPS AND BAITS 279

DIscUSSION shy species may be revealed only after re- moval of some animals. The small mammal fauna at Cuzco Ama- Previous studies of a of baits in includes at least 23 of native variety z6nico species a number of habitats have shown that baits rodents and 9 of species marsupials (Wood- can affect numbers of individuals and the man et This is al., 1991). diversity compa- of small mammals rable to that seen at the few other sites in species composition cap- tured (Beer, 1964; and 01- the western Amazonian lowlands that have Buchalczyk szewski, 1971; Fowle and Edwards, 1954; been studied, such as Cocha Cashu mar- (9 Laurance, 1992; Patric, 1970; Rickart et al., 23 rodents-Pacheco et al., 1993), supials, 1991; Willan, 1986). Our two baits differed Pakitza (9 marsupials, 22 rodents-Pacheco in odor, appearance, and nutrient content, et al., 1993), and the lower Rio Cenepa but, in general, we found them to be equally drainage (9 marsupials, 19 rodents-Patton effective. et al., 1982). richness is than Species higher Adequately censusing a diverse commu- that in lowland faunas from other reported nity inhabiting a complex, three-dimension- such as Manaus in the Neotropical regions, al environment, such as is found in tropical central Amazon 16 ro- (7 marsupials, rainforests, requires a variety of techniques. dents-Malcolm, and Barro Colo- 1990), We repeat the recommendation made by rado 13 rodents-Glanz, (6 marsupials, other investigators (Pizzimenti, 1979; Seal- and La Selva 15 ro- 1990) (5 marsupials, ander and James, 1958) that an assortment in Central America. dents-Timm, 1994) of traps be used to account for possible bi- It seem to use may incongruous snap- ases of the species involved. traps or removal trapping for assessing bio- diversity for conservation purposes. How- ACKNOWLEDGMENTS ever, at Cuzco Amaz6nico, as at most trop- We thank R. Arana C., E. D. Hooper,C. Pa- ical sites, the small mammal fauna was so checo A., V. Pacheco, and C. A. Schmidt for in poorly known that it was necessary to col- their assistance the field. A. H. Gentry aided our of the associa- lect specimens for accurate identifications understanding vegetational tions in which we worked.J. PurisacaP., Direc- of species. Several rare, cryptic species ci6n GeneralForestal de Fauna,Ministerio de would have been identified in y incorrectly Agricultura,Lima, grantedpermits for our field the field had individuals not been preserved work. J. E. Koechlin, of Cuzco Amaz6nico as specimens. Previously, we estimated that Lodge, providedexcellent facilities and support we removed only 47-66% (Woodman et al., for our studies there. A. B. Luscombe and the 1995) of the small mammal fauna from the Asociaci6n de Ecologifay Conservaci6n(ECCO- narrow area sampled by our transects. If our Peru) assisted with logistics. W. E. Duellman's conclusion, that snap-traps and live-traps and P. S. Humphrey'sefforts in securingfunding did not differentially sample the fauna, is and aiding in logistics are most appreciated.E. C. and an reviewer correct and the difference in species rich- Birney anonymous provided helpful suggestions on an earlier draft of this ness was simply attributable to more cap- manuscript. Field work at Cuzco Amaz6nico tures in snap-traps, then one could obtain was supportedby National GeographicSociety the same results with but with the live-traps, Grant 4016-89 and the University of Kansas of expenditure more trapping effort. How- Museumof NaturalHistory. Portions of this pro- ever, this argument cannot be extended to ject were funded by the E. Raymond Hall En- include capture-and-release trapping. Nu- dowment Fund and the Glen C. Rinker Memo- merous studies have demonstrated that cer- rial Fund of the University of Kansas Museum tain individuals within populations (e.g., of NaturalHistory. Davis and Emlen, 1956; Joule and Camer- LITERATURECITED on, 1974; Smith, 1968) can dominate traps. BEER,J. R. 1964. Bait preferencesof some small Thus, subordinate individuals or more trap- mammals.Journal of Mammalogy,45:632-634. 280 JOURNAL OF MAMMALOGY Vol. 77, No. 1

BUCHALCZYK,T., ANDJ. L. OLSZEWSKI.1971. Behav- MYERS, N. 1988. Tropical forests and their species: ioural response of forest rodents against trap and going, going .. . ? Pp. 28-35, in Biodiversity (E. O. bait. Acta Theriologica, 16:277-283. Wilson, ed.). National Academy Press, Washington, COCKRUM,E. L. 1947. Effectiveness of live traps ver- D.C., 521 pp. sus snap traps. Journal of Mammalogy, 28:186. NEAL,B. R., ANDA. G. COCK. 1969. An analysis of DALBY,P. L., ANDD. O. STRANEY.1976. The relative the selection of small African mammals by two effectiveness of two sizes of Sherman live traps. break-back traps. Journal of Zoology (London), 158: Acta Theriologica, 21:311-313. 335-340. DAVIs,D. E., ANDJ. T. EMLEN.1956. Differential trap- O'FARRELL,M. J., ET AL. 1994. Use of a mesh live ability of rats according to size and age. The Journal trap for small mammals: are results from Sherman of Wildlife Management, 20:326-327. live traps deceptive? Journal of Mammalogy, 75: DUELLMAN, W. E., AND J. E. KOECHLIN. 1991. The 692-699. Reserva Cuzco Amaz6nico, Peru: biological inves- PACHECO,V., B. D. PATTERSON,J. L. PATTON,L. H. tigations, conservation, and ecotourism. Occasional EMMONS,S. SOLARI,AND C. E ASCORRA.1993. List Papers of the Museum of Natural History, The Uni- of mammal species known to occur in Manu Bio- versity of Kansas, 142:1-38. sphere Reserve, Peru. Publicaciones del Museu de DURAN,J. C., ANDR. W. SAMZ. 1973. Trap responses Historia Natural, Universidad Nacional Mayor de of small rodents to live traps and snap traps. Journal San Marcos, serie A, 44:1-12. of the Arizona Academy of Science, 8:170-171. PATRIC,E. E 1970. Bait preference of small mammals. FOWLE,C. D., ANDR. Y. EDWARDS.1954. The utility Journal of Mammalogy, 51:179-182. of break-back traps in population studies of small PATTON,J. L., B. BERLIN,AND E. A. BERLIN. 1982. mammals. The Journal of Wildlife Management, 18: Aboriginal perspectives of a mammal community in 503-508. Amazonian Peru: knowledge and utilization patterns GALINDO-LEAL,C. 1990. Live-trapping vs. snap-trap- among the Aguaruna Jivaro. Special Publications ping of deer mice: a comparison of methods. Acta Series, Pymatuning Laboratory of Ecology, Univer- Theriologica, 35:357-363. sity of Pittsburgh, 6:111-128. GENTRY,A. H. 1990. Four Neotropical rainforests. PETERS, G., AND R. HUTTERER.1990. Vertebrates in Yale University Press, New Haven, Connecticut, 627 the tropics. Alexander Koenig Zoological Research pp. Institute and Zoological Museum, Bonn, Germany, GLANZ,W. E. 1990. Neotropical mammal densities: 424 pp. how unusual is the community on Barro Colorado PIZZIMENTI,J. J. 1979. The relative effectiveness of Island, Panama? Pp. 287-313, in Four Neotropical three types of traps for small mammals in some Pe- rainforests (A. H. Gentry, ed.). Yale University ruvian communities. Acta Theriologica, 24: Press, New Haven, Connecticut, 627 pp. 351-361. GOODNIGHT,C. J., ANDE. J. KOESTNER.1942. Com- QUAST,J. C., ANDW. E. HOWARD.1953. Comparison parison of trapping methods in an Illinois prairie. of catch of two sizes of small-mammal live traps. Journal of Mammalogy, 23:435-438. Journal of Mammalogy, 34:514-515. HANSSON,L., ANDI. HOFFMEYER.1973. Snap and live RICKART,E. A., L. R. HEANEY,AND R. C. B. UTZUR- trap efficiency for small mammals. Oikos, 24:477- RUM. 1991. Distribution and ecology of small mam- 478. mals along an elevational transect in southeastern JOULE,J., ANDG. N. CAMERON.1974. Field estimation Luzon, Philippines. Journal of Mammalogy, 72:458- of demographic parameters: influence of Sigmodon 469. hispidus population structure. Journal of Mammal- SANDERS,H. L. 1968. Marine benthic diversity: a ogy, 55:309-318. comparative study. The American Naturalist, 102: LAURANCE,W. E 1992. Abundance estimates of small 243-282. mammals in Australian tropical rainforest: a com- SEALANDER,J. A., AND D. JAMES. 1958. Relative ef- parison of four trapping methods. Wildlife Research, ficiency of different small mammal traps. Journal of 19:651-655. Mammalogy, 39:215-223. MALCOLM,J. R. 1990. Estimation of mammalian den- SLADE, N. A., M. A. EIFLER, N. M. GRUENHAGEN,AND sities in continuous forest north of Manaus. Pp. 339- A. L. DAVELOS.1993. Differential effectiveness of 357, in Four Neotropical rainforests (A. H. Gentry, standard and long Sherman livetraps in capturing ed.). Yale University Press, New Haven, Connecti- small mammals. Journal of Mammalogy, 74:156- cut, 627 pp. 161. MCDADE,L. A., K. S. BAWA,H. A. HESPENHEIDE,AND SMITH, G. C., D. W. KAUFMAN, R. M. JONES, J. B. G. S. HARTSHORN.1994. La Selva: ecology and nat- GENTRY,AND M. H. SMITH. 1971. The relative ef- ural history of a Neotropical rain forest. University fectiveness of two types of snap traps. Acta Ther- of Chicago Press, Chicago, 486 pp. iologica, 16:284-288. MORRIS,R. D. 1968. A comparison of capture success SMITH,M. H. 1968. A comparison of different meth- between Sherman and Longworth live traps. The Ca- ods of capturing and estimating numbers of mice. nadian Field Naturalist, 82:84-87. Journal of Mammalogy, 49:455-462. MUSSER,G. G., ANDM. D. CARLETON.1993. Family TIMM,R. M. 1994. The mammal fauna. Pp. 229-237, Muridae. Pp. 501-755, in Mammal species of the in La Selva: ecology and natural history of a Neo- world: a taxonomic and geographic reference (D. E. tropical rain forest (L. A. McDade, K. S. Bawa, H. Wilson and D. M. Reeder, eds.). Second ed. Smith- A. Hespenheide, and G. S. Hartshorn, eds.). Univer- sonian Institution Press, Washington, D.C., 1206 pp. sity of Chicago Press, Chicago, 486 pp. February 1996 WOODMAN ET AL.-COMPARISON OF TRAPS AND BAITS 281

TosI, J. A., JR. 1960. Zonas de vida natural en el Peril. WOODMAN, N., N. A. SLADE, R. M. TIMM, AND C. A. Instituto Interamericano de Ciencias Agricolas de la SCHMIDT. 1995. Mammalian community structure in Organizaci6n de Estados Americanos, Boletin Tec- lowland, tropical Peru, as determined by removal nico, 5:1-271. trapping. Zoological Journal of the Linnean Society, WIENER,J. G., ANDM. H. SMITH.1972. Relative ef- 113:1-20. ficiencies of four small mammal traps. Journal of WOODMAN, N., ET AL. 1991. Annotated checklist of Mammalogy, 53:868-873. the mammals of Cuzco Amaz6nico, Peru. Occasion- WILLAN, K. 1986. Bait selection in laminate-toothed al Papers of the Museum of Natural History, The rats and other southern African small mammals. University of Kansas, 145:1-12. Acta Theriologica, 31:359-363. WILSON, E. O. 1988. Biodiversity. National Academy Submitted 11 October 1994. Accepted 5 April 1995. Press, Washington, D.C., 521 pp. . 1992. The diversity of life. The Belknap Associate Editor was Patricia W. Freeman. Press of Harvard University Press, Cambridge, Mas- sachusetts, 424 pp.