.......................... Reports .

Population ecology andreproduction of the white-eared opossum

Didelphis albiventris (Mammalia, Marsupialia)in an urban environment of Brazil

Nn..TON CARLOS CÁCERES

Departamento de Zoologia, Universidade Federal do Paranâ, Curitiba, PR 81531-990, Brasil

The seasonal variation in numbers of individuaIs, causes for death, breeding season andlitter size of the whíte-eared opossum, Didelphis albiventris, were studied in two small ur-ban forest fragments of southern Brazil. In order to trap opossums, 30 live traps were usedonce a week in a forest fragment (5 ha) between February 1995 and January 1996. In addi-tion, 10 traps were used twice a month in another area (2.5 ha) between November 1996 andFebruary 1997. IndividuaIs trapped were sexed, aged, and released. Females with litter hadtheir pouch young counted. The bait used was banana with codfish liver oil. As results afteran effort of 1770 traps set, 37 opossums were captured (20 females and 17 males) with thesex ratio being significantly equal (p > 0.40). Juvenile individuaIs were captured seasonallyduring the wet season (spring and summer). Since juveniles represented the most commonindividuaIs trapped, total abundances were higher during the wet season. Immigrations(plus births) rather than emigrations were thought to be the main factor for abundancevariations. Moreover, road kill by cars and fights with dogs were the main causes for deathof opossums. Based on time of captures of recently weaned opossums and females with lit-ters, the breeding season showed mainly two periods of births (between August and Novem-ber), though two litters were seen in April of two years. The mean litter size was 9.0 young(N = 14 litters). The white-eared opossum revealed a great potential to colonize urban envi-ronments where there are forest fragments, with its higher numbers of juveniles compen-sating adult deaths mainly by anthropical factors.

A variação sazonal na abundância de indi-víduos, as causas de morte, estação reproduti-va e tamanho de ninhada do gambá-de-ore-lha-branca, Didelphis albiventris, foram es-tudados em dois fragmentos florestais urba-nOS do sul do Brasil entre 1995 e 1997. Para

as capturas dos animais, 30 armadilhas fo-ram utilizadas semanalmente em uma dasáreas (5 ha) e 10 armadilhas duas vezes pormês na outra área (2,5 ha). A isca utilizada

foi uma mistura de banana com óleo de fí-gado de bacalhau comercial. Após um es-

forço de 1770 armadilhas armadas, 37 ani-mais foram capturados (20 fêmeas e 17 ma-chos) com a proporção sexual sendo signifi-cativamente igual (p > 0,40). Indivíduos jo-vens foram capturados mais sazonalmente,principalmente durante a estação úmida (pri-

Correspondence to: Nilton Carlos Cáceres, Departamento de Zoologia, Universidade Federal do Paraná, Caixa Postal 19020, Curitiba, PR81531-990, BrasilE-mail: ncaceres@garoupa.bio.ufpr.br

Volume 52(3)· May/June 2000 Ciência e Cultura Journal of the Brazilian Association for the Advancement of Science· 171

........................... Reports .

mavera e verão). Como jovens foram captu-rados proporcionalmente em maior nLÍme-ro, as maiores abundâncias foram observa-das durante a estação úmida. As imigrações,junto aos nascimentos, foram os principaisfatores a influenciar na variação das abun-dâncias. Atropelamentos por carros e brigas

com cães foram as causas de morte observa-das para a espécie. Baseando-se nos perío-dos de capturas de jovens recém-indepen-dentes das mães e de fêmeas com ninhadas,a estação reprodutiva de D. albiventris apre-sentou dois períodos principais de nascimen-tos (entre agosto e novembro). O tamanho

médio de ninhada foi de 9,0 filhotes (N = 14ninhadas). O gombá-de-orelha-branca apre-sentou grande potencial para colonizar am-bientes urbanos onde existem fragmentos flo-restais, com elevados números de jovens com-pensando as mortes de adultos principalmen-te por fatores antrôpicos.

A lthough the white-earedopossum, Didelphis albi-ventris Luod, 1841 (Fig. 1),shows a wide distribution in

South America (1), only a few studiesprovide ecological data (2,3,4) such asthe populatioo dynamics of this mar-supial (5,6), but several were conducted00 reproduction (6-12). The presentstudy presents data on some aspectsof the population dynamics, breedingseason and litter size of D. albiventrisin two small urban areas of southernBrazil.

Study areas are in Curitiba, state ofParaná (25°25'S; 49°18'W), at 940 mabove sea leveI. The first of them,namely COPEL, is 5 ha in size. The otherarea, namely UFPR, is 2.5 ha. They are 8km apart and are composed of disturbedprimary mixed ombrophyllous forestsand surrounding open forests (about10% sampled by traps in COPEL and20% in UFPR; see below). Thereis a wetseason (September to March) and a drierseason (April to August) in the region,but oot well-defined. The mean annualtemperature is 16.5°C and the mean an-nual rainfall ranges from 1100 to 1600mm(13).

Opossums were captured using 30live traps (40x20x20 em) in COPEL and10 traps in UFPR which were uniform1yplaced in these areas at fixed points (fol-lowing a grid pattern), These live trapswere baited during the aftemoon andchecked for opossum captures the fel-lowing morning once a week in COPEL(February 1995 - January 1996) and twicea month in UFPR (November 1996 - Feb-ruary 1997). The bait used was ripe ba-nana and codfish liver oil. After capture,animaIs were sexed, aged (7,8), markedby cornbinations of holes on each ear(4) and released. Females with litters hadtheir pouch young counted. Additionalreproductive data were obtained throughopossums randomly collected at othersites of Curitiba, usually near the studyareas, until1999. In COPEL, immigration

data were obtained by dividing the num-ber of new individuaIs in the area everytwo months by the number of "old" in-dividuaIs captured there during the pre-vious two months (14). The emigrationdata were calculated ioversely, takinginto accouot the exit of individuals out-side the area. The cause for death ofopossums was verified occasionallythrough direct observation in the studyarea and surroundings, besides otheranimals collected at other sites inCuritiba. The breeding season of opos-sums was determioed by the time whenpouch young aod recent pouch releasedopossums were present in the popula-tion as well by the degree of their devel-opment (15).

During the field work, 1770 livetraps were used which resulted in 37 in-dividuaIs of D. albiventris captured (20females and 17 males) and a total of 82.captures and reeaptures. The sex ratioobserved (56% females versus 44%males) did not differ signifieantly fromthe expeeted rate (p > 0.40, df = 1, Chi-Square test).

Abundances of D. albiveruris werehigh during the spring and summer andlow during winter months (Table 1).Simi-larly, in the COPEL area, the immigrationof new individuals was more pro-nounced between August-September1995 and December-January 1996 (from75 to 230% of new individuals, respee-tively, in relation to the 'resident' indi-viduaIs). During April to July, the immi-gration was only 20%. The emigrationwas high during Oetober to Januarywhen lOü% of individuals left the COPELarea. This exit rate was lower betweenApril and September when emigrationaveraged 50%.

Regarding age strueture, infant in-dividuals appeared seasonally in the ar-eas during the spring and summermonths. Subadults appeared later, mainlyin the autumn. Adult individuals weredominant during the winrer months. Inboth study areas pouch young occurred

172 • Ciência e Cultura Journal of the Brazilian Association for the Advancement of Science

at low rates in all seasons (Fig. 2) butadditional data from other sites ofCuritiba revealed eight litters duringAugust-September (57%), four littersduring November-January (29%) andonly two litters in April (14%) out of atotal of 14 litters seen.

The main causes for death of opos-sums were road kill (N = 10) and, see-ondarily, fights with dogs (N = 3). Adultmales of age c1ass 5 were mainly foundkilled on roads while adult females ingeneral were killed by dogs or cars. Someof these had pouch young.

The mean litter size was found tobe 9.0± 2.2 (range: 5 to 12;modalc1asses:8 and 12; N = 14liúers). The mean num-ber ofteats was 11.3 ( = 10 females).Two younger females (age c1ass 3)weighing 550 and 575 g were seen withlitters in January 1996 and February1999. Two females with well-developedteats but without litters (young left inthe den?) were trapped in September andNovember 1995.

Figure 1. Individual of lhe white-eared opos-sum, Didelphis albiv entris, in a tree in. theCOPEL area, Curitiba, southern Brazil.

Volume 52(3) • May/June 2000

........................... Reports .

Table 1, - Reíaríve abundance (number of individuals/number of traps used) of the white-earedopossum, Didelphis albiventris, in an ürbaa region of southern Brazi1.

Spritig Summer Autumn Wm/er

Number of .individuals trapped 18 6 5Total abundance 0.040 0.014 0.012,Adult abundance 0.002- 0,007 0.012Number of traps used 45Q 450 420

The sex ratio seen here for D.albiventris is similar to those found byother investigators that have workedwith Didelphis in South America (4,5,14,16,17 ,18), The population levels ob-served here are also similar to othersfound elsewhere (4,5) but they werelower th311those sympatric marsupial,D. aurita, in the COPEL area (14). Theabundance was high in the wet seasonbecause it is the time of the appearanceof many recentIy weaned opossums inthe population (15). This is confirmedby the fact that the number of poucbyoung in D. albiventris was higher th311that of adults and subadults (14,16). Thissituation is caused by the high mortal-ity of young opossums killed by preda-tors (19).

Immigration plus births contributegreatly to the population fluctuations ofD. albiventris, as was reported for D.aurita in southem Brazil as well (14). Thehigh rate of immigration seen here prob-ably occurs due to the changes caused

by the onset of the breeding season (20).Besides the higher number of litters dur-ing August 311dSeptember, the onset ofbreeding might be determined by theoccurrence of released young duringwetter periods of year, which is wellknown in Didelphis (6,7,14,2i). Thus,the onset of breec1ing for D. albiventrishere is thought to be during the start ofthe winter (July)(ll), a period that pre-cedes the wetter months. Breeding indi-viduals arise from two rather than threebirth peaks that usually occur in thisgenus (6,14,15,16,22,23), The breedingseason of D. albiventris in southernBrazil appears to end in the autumn as isusually the case for Didelphis (10,11,12,14,23). This time is indicated by the domi-nant numbers of subadults and aduItsin the population (Fig. 2).

As mentioned previously (14),anthropical factors are the main causesfor opossum mortality in urban environ-ments (2). Taking into account opossumdeaths on roads, the low time of perma-

DSubadult • Female with litter

100%

80%

60%

40%

20%

0%Spring Summer

Season

DAdult

Autumn Winter

Figure 2. Relative abundances of the white-eared opossum, Didelphis albiventris, captured intwo small [orest fragments of Curitiba, southern Braril, between 1995 and 1997, All blackcolumns a/ the top are adult females except for a black column at 'the bottom that belongs to ayoung female with litter. Adults are presented excluding pouch young females.

Volume 52(3)· May/June 2000

nence of individuals in the fragment (24),and the high emigration/immigrationrates (mainly in the wet season), 311al-most complete population turnover ofD. albiventris is expected in about 1yearin mixed forest fragments of southemBrazil (14). In addition,juveniles growrapidly (25) 311dreach high numbers inthe wet season and are therefore prob-ably strong competitors for resourcesin small forest fragments. Thus thewhite-eared opossum revealed a greatpotential to colonize urban environ-ments where there are forest fragments,with its higher numbers of juvenilescompensating adults killed on roads 311dby dogs,

The mean litter size of D. albi-ventris is about 77% of its me311num-ber of teats. This rate is higher than the50% reported for several marsupiaIs inVenezuela (26). In another study in theCOPEL area, a similar rate was found forD. aurita (76%) (23). In southeasternBrazil, a rate of 93% was seen for D.albiventris, and the teat number limitedthe litter size there (4). However, latitu-dinal and altitudinal variations may bethe most important factors to control lit-ter sizes in Didelphis (8,15,21). lu thisview, the mean litter size observed hereis relatively high when compared withother studies at lower latitudes and alti-tudes (4,8,9,10,27) except, however, forthe case reported in Argentina (12). •

References and notes

1. Emmons LH, F Feer 1990 Neotropicalrainforest mammals: A field guide. TheUniversity of Chicago Press, Chicago

2. Fonseca GAB, KH Redford, LA Pereira1982 Notes on Didelphis albiventris(Lund, 1841) of Central Brazil. Ci cuu34: 1359-1362

3. Streilen KE 1982 Behavior, ecology,and distribution of South American mar-supials, p 231-250, 111 Mammalian biol-ogy in South America. MA Mares, HHGenoways, eds, Univ Pittsburgh, Pitts-burgh

4, Monteiro-Filho ELA 1987 Biologiareprodutiva e espaço domiciliar de Di-delphis albiventris em uma áreaperturbada na região de Campinas,Estado de São Paulo tMammatia.Marsupialia). MSc Thesis, UniversidadeEstadual de Campinas, Campinas,' SP

5, CajaJ JL 1976 Una población de Didel-phis albiventris segun el método deBugoslavski. Physis C 35: 243-249

6, Mares MA, KA Ernest 1995 Population

Ciência e Cultura Journal ot the Brazilian Association for the Advancement ot Science • 173

· Reports .and community ecology of small mam-mais in a gallery forest of central Brazil.J Mammal 76: 750-768

7. Gilmore RM 1943 Mammalogy in anepidemiological study of jungle yellowfever in Brazil. J Mammal 24: 144-162

8. Tyndale-Biscoe CH, RB Mackenzie 1976Reproduction in Didelphis marsupialisand D. albiventris in Colombia. J Mam-mal 57: 249-265

9. Cerqueira R 1984 Reproduction de Di-delphis albiventris dans le nord-est duBrésil (Polyprodontia, Didelphidae).Mammalia 48: 95-104

10. Rigueira SE, CMC Valle, JBN Varejão, PVAlbuquerque, JC Nogueira 1987 Algumasobservações sobre o ciclo reprodutivoanual de fêmeas do gambá Didelphisalbiventris (Lund, 1841) (Marsupialia,Didelphidae) em populações naturais noEstado de Minas Gerais, Brasil. Rev Bras20014: 129-137

11. Nogueira JC 1988 Anatomical aspectsand biometry of the male genital systemof tbe white-bel ly opossurn Didelphisalbiventris Lund, 1841 during the annualreproductive cycle. Mammalia 52: 233-242

12. Regidor HA, M Gorostiague 1996 Re-production in lhe white-eared opossum(Didelphis albiventris) under tempera teconditions in Argentina. Stud NeotropFauna Environ 31: 133-136

13. Maack R. Geografia física do Estado doParanâ. José Olympio, Rio de Janeiro,RJ

14. Cáceres NC, ELA Monteiro-Filho 1998Population dynamics of the common

opossum, Didelphis marsupialis (Marn-mal ia, Marsupialia), in southern Brazil. 2Sdugetierk 63: 169-172, and Cáceres NC,ELA Monteiro-Filho 1996 Análisecomparativa populacional de dois gambásem mata urbana no sul do Brasil. Resumos,[U Congr Ecol Bras, Brasilia

15. O'Connell MA 1979 The ecology ofdidelphid marsupiais from nortbern Ven-ezuela, p 73-87. In Vertebrate ecology inlhe northern neotropics. JF Eisenberg ed,Smithsonian, Washington

16. Atramentowicz M 1986 Dynamique depopulation chez trois marsupiauxdidelphidés de Guyane. Biotropica 18: 136-149

17. O'Connell MA 1989 Population dy-namics of neotropical small mammals inseasonal habits. J Mammal 70: 532-548

18. Cherern JJ, ME Graipel, ME Menezes, MSoldateli 1996 Observações sobre a bio-logia do gambá (Didelphis marsupialis)na Ilha de Ratones Grande, Estado de SantaCatarina, Brasil. Biotemas 9: 47-56

19. Bergallo HG, NC Cáceres, ELA Monteiro-Filho 2000 Dinâmica de populações demarsupiais brasileiros. In Os marsupiaisdo Brasil: Evolução, biologia e ecologia.NC Cáceres, ELA Monteiro-Filho eds ~Submitted for publication

20. Ryser J 1992 The mating system andmale mating success of the Virginia opos-sum (Didelphis virginianay in Florida. J2001 Lond 228: 127-139

21. Fleming TH 1973 The reproductivecycles of three species of opossums andother mammals in the Panama CanalZone. J Mammal 54: 439-455

22. Sunquist ME, JF Eisenberg 1993 Repro-ductive strategies of female Didelphis. BuliFI Mus Nat Hist Biol Sci 36: 109-140

23. Cáceres NC, ELA Monteiro-Filho 1997Reproductive biology of the commonopossum, Didelphis marsupialis (Mam-malia: Marsupialia), in southern Brazil.Brenesia 47-48: 117-122

24. Cáceres NC, ELA Monteiro-Filho 2000Home range of the white-eared opossumDidelphis albiventris (Marnrnalia, Masur-pialia) in a forest fragment of southernBrazil ~ ln preparation

25. Cáceres NC, ELA Monteiro-Filho 1999Tamanho corporal em populações natu-rais de Didelphis (Mammalia: Marsupialia)do Sul do Brasil. Rev Bras Biol 59: 461-469

26. Eisenberg JF, DE Wilson 1981 Relativebrain size and dernographic strategies indidelphid marsupiais. Am Nat 118: 1-15

27. Catzeflis F, C Richard-Hansen, C Fournier-Chambrillon, A Lavergne, J Vié 1997Biométrie, reproduction et sympatriechez Didelphis marsupialis et D. albi-ventris en Guyane française (Didelphidae:Marsupialia). Mammalia 61: 231-243

28. Acknowledgements: 1 would like to thankValeriano M Cáceres-Júnior for help inthe field, Emygdio ELA Monteiro-Filho,Juliana Quadros and Susi M Pacheco andtwo anonyrnous reviewers for manuscriptcomments and Vernon E Thatcher andAdilson M Brito-Filho for the Englishtranslation. The "Curso de Pós-Gradua-ção em Zoologia - UFPR" and the Brazil-ian Councils for Research (CAPES andCNPq) gave general support for this study.