Small Mammal Community Structure and Composition in the Cerrado Province of Central Brazil

Home , Caviidae, Thalpomys

CORE Metadata, citation and similar papers at core.ac.uk

Provided by UNL | Libraries

University of Nebraska - Lincoln DigitalCommons@University of Nebraska - Lincoln

Faculty Publications from the Harold W. Manter Laboratory of Parasitology Parasitology, Harold W. Manter Laboratory of

1986

Small mammal community structure and composition in the Cerrado Province of central Brazil

Michael A. Mares University of Oklahoma Norman Campus

Kristina A. Ernest University of Arizona

Donald D. Gettinger University of Oklahoma Norman, [email protected]

Follow this and additional works at: https://digitalcommons.unl.edu/parasitologyfacpubs

Part of the Parasitology Commons

Mares, Michael A.; Ernest, Kristina A.; and Gettinger, Donald D., "Small mammal community structure and composition in the Cerrado Province of central Brazil" (1986). Faculty Publications from the Harold W. Manter Laboratory of Parasitology. 679. https://digitalcommons.unl.edu/parasitologyfacpubs/679

This Article is brought to you for free and open access by the Parasitology, Harold W. Manter Laboratory of at DigitalCommons@University of Nebraska - Lincoln. It has been accepted for inclusion in Faculty Publications from the Harold W. Manter Laboratory of Parasitology by an authorized administrator of DigitalCommons@University of Nebraska - Lincoln. Mares, Ernest & Gettinger in Journal of Tropical Ecology (1986) 2. Copyright 1986, Cambridge University Press. Used by permission.

Journalof TropicalEcology (1986) 2:289-300. With5 figures

Small mammal community structureand composition in the Cerrado Province of central Brazil

MICHAEL A. MARES, KRISTINA A. ERNEST* and DONALD D. GETTINGER

StovallMuseum and Departmentof Zoology, Universityof Oklahoma, Norman,Oklahoma 73019 USA

ABSTRACT. Communitystructure, macrohabitat selection, and patternsof species co-occur- rence were examined duringa 14-monthstudy of small mammalsin the Cerrado Provinceof central Brazil. Data were collected frommark-recapture grids in brejo and galleryforest, and fromlive-trapping and specimen collection in all habitat types withincerrado (campo limpo, campo sujo, cerrado [s.s.], cerradao,brejo, valley-sidewet campo, and galleryforest). Gallery forest supported the highestspecies richness,most complex verticaldistribution of species, highestlevel of trophicdiversity, and highestmacroniche diversity. Degree of habitatselection varied widely. All habitat types supported both rodentsand marsupials,although marsupials tended to be much less common in the grasslands(campos) than in woodlands (cerrado) and forests(cerradao, gallery forest). Some species, such as Didelphis albiventris,occurred in all habitat types, while others were much more restricted.Oryzomys bicolor, for example, ocur- red only in galleryforest. No habitat type had a completelydistinct fauna: overlap in species composition always occurredwith at least one other habitat type. Because of the greatvaria- bility of habitats, and the fact that subsets of the mammal fauna were frequentlyhabitat specific,the overall species richnessof any portion of mixed cerrado vegetationis remarkably high.

KEY WORDS: Brazil, communitystructure, gallery forest, habitat selection,marsupials, neo- tropics,rodents, tropical savanna.

INTRODUCTION South Americasupports many habitats,but two major phytogeographicpro- vinces,both tropical,predominate. The Amazon rainforest, a mosaicof forest communities,encompasses about 5-6 million kmi2,while the cerrado,a complex of savanna grasslandsand forestedareas, extends over 1.5 millionkm2 (e.g. Eiten 1974, Fearnside1982). Together,these areas comprise42% of the South Americancontinent. The flora of both areas has been examinedcur- sorily,while their faunas remain largely unknown (e.g. Alho 1982a, Mares1982, 1986). The taxonomyof the mammalsin these habitatsis poorlyknown, and little ecological researchhas been conductedon thisgroup (Mares 1982, Pine 1982). Both the Amazon Basin and cerrado are experiencingrapid develop- ment (Goodland & Irwin 1977, Myers1980) and theirfaunas are considered

* Present address: Departmentof Ecology and EvolutionaryBiology, Universityof Arizona, Tucson, Arizona85721, USA.

290 MICHAEL A. MARES, KRISTINA A. ERNEST AND DONALD D. GETTINGER threatened.In this paper, we examine habitat selectionby small mammals, especiallyrodents, in the CerradoProvince of centralBrazil. Most of our data weregathered in one of the Cerrado'smost unusual habitats, the galleryforest. Most descriptionsof tropicalforest in South Americaconsider the lowland Amazon forest,the montanetropical forest rimming the Amazon Basin,or the montanerain forestof the east coast of Brazil. However,there is an extensive evergreentropical forest that occurs along many thousands of kilometersof the cerradowatercourses and extendsthroughout the regionin a serpentinefashion. Ultimately,this galleryforest joins the Amazon or Atlantic coastal forest (Eiten 1972, 1974), and so it is not surprisingthat many floral and faunalele- mentsare shared among these sylvanhabitats (Alho 1982b, Cerqueira1982). Little ecological work has been done on the mammalsof the galleryforest (Paula 1983), althoughsome cerrado mammals have been examinedecologically (Alho 1981a, 1981b, 1981c, Dietz 1983, Souza & Alho 1980). The habitat selection of tropical mammals,forest or otherwiseis largely unknown,but it has been suggestedthat habitatspecificity in tropicalregions is more pronouncedthan in temperateareas (MacArthuret al. 1966). This has been especiallytrue forresearch done on birds (MacArthuret al. 1972). Here we presentpreliminary data on habitatselection, species richnessand macroniche compositionof the small mammalfauna of fourhabitats of the cerrado of Brazil. We comparethese data to otherstudies conducted in the cerradoand point out the importanceof havingdetailed information on speciescoexistence beforegeneralizations on diversitypatterns can be made. Our resultsapply to both ecologicaland biogeographicstudies of thisregion.

STUDY AREA Cerrado(sensu lato) is a majorphytogeographic province located in the central plateau region of Brazil (Figure 1; Cabrera & Willink1973). The climate is tropical,with distinct wet (October-April)and dry (May-September)seasons. Annualprecipitation ranges from 750-2000 mm (Eiten 1974). Cerrado(s. 1.), the dominantflora of the province,is a mosaic of xeromorphicupland vegetation typesranging from open grasslandsto closed woodlands.Trees and shrubs withcontorted trunks, thick bark, twisted limbs, and broad,thick leaves give a characteristicappearance to these upland formations.Edaphic factors,such as soil depth,drainage, and nutrientquality, influence the structureof the flora, and often produce physiognomicgradients (Eiten 1972). The upland florais commonlydivided into fivephysiognomic types (Eiten 1972, Goodland 1971): (1) cerradao,a forestof largeand moderatelytall trees,usually with a closed canopy and few shrubs;(2) cerrado(sensu stricto), a shortertree/shrub wood- land withan open canopy and sparsegrass understory; (3) campo cerrado,an open tree/shrubwoodland with a groundcover of grasses;(4) campo sujo, a grasslandwith scattered trees and shrubs;(5) campo limpo,an open grassland withfairly low grasses. Mares, Ernest & Gettinger in Journal of Tropical Ecology (1986) 2. Copyright 1986, Cambridge University Press. Used by permission.

Mammalecology in theBrazilian cerrado 291

-20

-10

-0

-10

-20

-30

-40

-50

Figure1. Map showingthe limitsof the Cerradophytogeographic Province in South America.

The flora changesdrastically along the drainagelowlands, where the water table is close to the surface.The dominantcerrado vegetation is replacedby mesophyticevergreen forest, which formscorridors along streamsand rivers. These galleryforests often ascend the slopes of the valleyand are usually50- 200 m wide (Eiten 1974). Two typesof lowlandmarshes are frequentlyfound along the edge of gallery forest: brejo, a sedge-dominated,permanently- inundatedmarsh; and valley-sidewet campo, a grass-dominated,seasonally- inundatedmarsh that grades into upland cerrado vegetation. In thispaper, we group the vegetationaltypes of the CerradoProvince into fourhabitats (Figure 2) for purposes of statisticalanalysis. We will consider cerradao,cerrado, and campo cerradoas a singlehabitat, cerrado. Further, we combine campo limpo,campo sujo, and valley-sidewet campo into the single habitat,campo. Membersof these habitat sets grade imperceptiblyinto one another,and our collectingeffort was not sufficientlyfine-grained to distin- guishbetween the relatedhabitat pairs. Brejo and galleryforest are considered separatehabitats. Data presentedhere were collected near Brasilia, Districto Federal ( 150 56' S, Mares, Ernest & Gettinger in Journal of Tropical Ecology (1986) 2. Copyright 1986, Cambridge University Press. Used by permission.

292

Mammalecology in theBrazilian cerrado 293

470 53' W). Althoughcerradao is poorly representedin this area, all other vegetationtypes were common. Most field work was done on the Ecological Reserveof the InstitutoBrasilieiro de Geografiae Estatisticaand the Fazenda Agua Limparesearch station of the Universidadede Brasilia.

METHODS BetweenAugust 1983 and November1984, smallmammals were sampled from all vegetationtypes: cerradao,cerrado (s.s.), and campo cerrado(4801 trap- nights);campo sujo, campo limpo, valley-sidewet campo (4096 trap-nights); galleryforest (70 465 trap-nights);and brejo (5739 trap-nights).Sherman (23 X 8 X 9 cm) and Tomahawk (48 X 15 X 15 cm) live-trapswere used to sample both the lower (<100 g) and upper (100 g - 2 kg) size classes of small mammals. Traps were placed at 10 or 15 m intervalsalong transect lines or in grids. Captured animals were anaesthetizedwith ether, and examined,measured and identified.Standard mark-recapture data were recorded,the animal was markedand, afterrecovery from anaesthesia, released at the point of capture. Althoughthe numberof trap-nightsin galleryforest is fargreater than that of the brejo and uplandhabitats, we feelthat the latterare sufficientfor the com- parisonsmade in thispaper. Voucher specimensfor all species (except Chiro- nectes and Coendou) were deposited in the Stovall Museum of Science and History,The Universityof Oklahoma, and the Departamentode Biologia Animal,Universidade de Brasilia. Statisticalanalyses. A speciesX habitatmatrix was constructedto analyse similarityin rodent species compositionbetween habitats.Zero (0) signified eitherthe absence of a particularspecies from a habitat,or a habitatin whicha species was only rarelycaught, suggesting that the habitatwas not preferred by that species. One (1) indicatedpresence of the speciesin preferredhabitat. The unweightedpair-group method of clusteringwith arithmeticaveraging (Sneath & Sokal 1973, UPGMA) was employed to summarizesimilarities betweenhabitats based on the simplematching coefficients, and to summarize similaritiesbetween species. A similaranalysis was performedon a macronicheX habitatmatrix. Macro- niches (Eisenberg1981) were definedas the locomotoradaptation-diet com- binationin each habitat.Zero (0) signifiedabsence of a macronichein a given habitat;one (1) indicatedthat at least one speciesoccupied that macroniche in a given habitat. Again, UPGMA was used to summarizesimilarity between habitatsbased on simplematching coefficients. Taxonomic determination.Whenever possible, series of study specimens (skin and skeleton) were made and comparedwith specimens in museumsas well as with publisheddescriptions of species. The taxonomicidentifications are as preciseas we can make themat the moment.However, two species,both membersof the Oryzomysnigripes complex, are listedas 0. nigripesA and 0. nigripesB. Also, we have been unable as yet to determinewith certainty the properbinomial for one specieseach of Marmosa,Proechimys and Dasyprocta. Mares, Ernest & Gettinger in Journal of Tropical Ecology (1986) 2. Copyright 1986, Cambridge University Press. Used by permission.

294 MICHAEL A. MARES, KRISTINA A. ERNEST AND DONALD D. GETTINGER

RESULTS Seven species of marsupials(family Didelphidae) and 18 species of rodents (familiesMuridae, Caviidae, Dasyproctidae, Echimyidae; Figure 3) were captured. Two additional rodent species (familiesHydrochaeridae, Erethizonti- dae) were seen but not captured(the presenceof Hydrochaerisalso was detec- ted by the unique footprintsof this species). These mammalsdisplay a wide rangeof body masses,food habits,and locomotoradaptations (Table 1). Table 1. Body mass, diet, locomotor adaptation and preferredhabitat of marsupialsand rodentscap- tured or seen in the cerrado. Diets are: C, carnivore;F, frugivore;G, granivore;H, herbivore;I, insecti- vore; 0, omnivore.Locomotor adaptationsare: A, arboreal; Aq, aquatic; C, cursorial;S, scansorial;Saq, semiaquatic;T, terrestrial.Habitats are: B, brejo; C, cerrado; F, forest;M, campo.

Body mass Locomotor Species (g) Diet adaptation Habitat

Didelphidae Chironectesminimus 600-790[11] C[11] Aq[11] F Didelphis albiventris 700-2000* 0[3] T/A[3] F, B, M, C Marmosaagilis 18-45* 0/I A/T* F Marmosa sp. 36* 0/I T F Monodelphisamericana 35-70* O/C T* F Monodelphisdomestica 50-155[6] O/C T C Monodelphiskunsi 9-15* I[1] T C Muridae Akodon cursor 30-55* 0[3] T[3]* F Bolomys lasiurus 30-75* O[10] T[10]* M, C Calomyscallosus 20-60* G/I[14] T* M, C C. tener 9-15* G/O T* M, C Nectomyssquamipes 200-450* 0[3] Saq* F Oryzomysbicolor 20-40* 0 A* F 0. capito 40-75* 0[7] T* F 0. concolor 45-95* O/F A/T* F t0. nigripesA 16-35* 0[3] T/A* F tO. nigripesB 12-25* 0 T/A* B, M O. subflavus 60-115* F/O T* M, C Oxymycterusroberti 50-95* I[12, 13] T* B, M Rhipidomysmastacalis 55-95* 0[13] A[13]* F Thalpomyslasiotus 17-30* O T* M, C Erethizontidae Coendou prehensilis 1-5 kg[12] H/F[8, 12] A[12] F Caviidae Cavia aperea 400-1000[6] H[12] C M Hydrochaeridae H. hydrochaeris 27-80 kg[12] H[12] Aq F, B Dasyproctidae Dasyprocta sp. 1200-3000[12] H/F[12] C F Echimyidae Proechimyssp. 225-425* H/F[5, 12] T[13] * F Thrichomysapereoides 175-395* H/F T/S[14] C

Citations: * this study; 1 = Anderson 1982; 2 = Charles-Dominiqueet aL 1981; 3 = Crespo 1982; 4 = Davis 1947; 5=Emmons 1982; 6=Fadem et al. 1982; 7=Fleming 1970; 8=Mares & Ojeda 1982;9=Mareset al. 1981a; 10=Mares et al. 1981b; ll=Marshall 1978; 12=Nowak & Paradiso 1983; 13=O'Connell 1982; 14 = Streilein1982. t Added in proof. Identificationsnow known to be: 0. nigripesA= 0. nigripes(Olfers); 0. nigripesB= 0. fornesiMassoia. Mares, Ernest & Gettinger in Journal of Tropical Ecology (1986) 2. Copyright 1986, Cambridge University Press. Used by permission.

Mammalecology in theBrazilian cerrado 295

'I',

I-FOREST- I BREJO- I+CAMPO - CERRADO-i

Akodon cursor Bolomys lasiurus Calomys callosus C. tener Nectomys squamipes Oryzomysbicolor 0. capito 0. concolor 0. nigripes A O. nigripes B 0. subtlavus Oxymycterusroberti Rhipidomys mastacal is rhalpomys lasiotus Coendou prehensil is Cavia aperea H. hydrochaeris Dasyprocta sp. Proechimys sp. Thrichomysapereoides

Figure3. Preferredhabitats of rodent species in the cerrado.Bars indicatepresence or absence of rodents in the fourmajor habitatsconsidered.

Galleryforest showed the highestspecies richnessof both marsupials(five species) and rodents(11 species). The latterwere representedby sevenmurids and one species each of Erethizontidae,Hydrochaeridae, Dasyproctidae and Echimyidae.Brejo had the lowest rodent diversity:two muridspecies and Hydrochaeris.Didelphis was the only marsupialcaptured in brejo. Campo and cerradowere intermediatein diversity.Campo was a preferredhabitat for six muridspecies and one caviid.Again, Didelphis was the onlymarsupial. Cerrado had fivemurids, one echimyidand threemarsupials. Rodent species clusteredinto two main groups(Figure 4A). The firstgroup contained species that occurredin the forest.Campo, campo/brejo,campo/ cerrado,and cerradospecies were subsets that comprised the second group,the non-forestrodents. Campo and cerradowere the mostsimilar habitats in termsof rodentspecies composition(Figure 5). Sixty-twopercent of the campo species also occurred Mares, Ernest & Gettinger in Journal of Tropical Ecology (1986) 2. Copyright 1986, Cambridge University Press. Used by permission.

296 MICHAEL A. MARES, KRISTINA A. ERNEST AND DONALD D. GETTINGER

FOREST BREJO CAMPO CERRADO

00 02 04 06 08 Simplemntching

FOREST

BREJO

CAMPO

CERRADO

0.2 0.4 0.6 0.8 10

Simple matching

Figure 4. Dendrogramof habitat relationships: (A) based on presence or absence of rodent species, cophenetic correlationcoefficient= 0.914; (B) when clusteredby macronichespresent or absent in each macrohabitat,cophenetic correlation coefficient = 0.966. in cerrado,while 83% of cerradospecies occurredin campo. Some overlapin rodent species was found betweencampo and brejo (two species), and forest and brejo (one species). Forest was the most distinctivehabitat: only one of its 11 species(9%) was sharedwith another habitat. Galleryforest supported the highestnumber of macroniches(Table 2), and was quite distinctfrom the otherhabitats in termsof macronichesfilled (Figure 4B). Campo and cerradosupported intermediate numbers of macroniches,and werethe mostsimilar habitats. Brejo supportedthe fewestmacroniches.

DISCUSSION The CerradoProvince supports a verydiverse small mammal fauna. Indeed, we collected 27 species of smallmammals in our immediatearea, and severalother species have been reportedfrom the cerradoof centralBrazil (e.g. Didelphidae: Philanderopossum; Muridae: Akodon nigrita,Holochilus brasiliensis,Jusceli- nomys candango,Kunsia fronto,Plectomys paludicola; Caviidae: Galea spixii; Agoutidae:Agouti paca; Echimyidae:Carterodon sulcidens, Clyomys laticeps, Euryzygomatomysspinosus; Alho 1982a). If these species are included,the cerradosupports an exceedinglyrich small mammalfauna. We mustpoint out Mares, Ernest & Gettinger in Journal of Tropical Ecology (1986) 2. Copyright 1986, Cambridge University Press. Used by permission.

Mammalecology in theBrazilian cerrado 297

.~~~~~~~~~~~~~~ I Akodon cursor Nectomyssquamipes Oryzomysbicolor 0. capito 0. concolor 0. nigripes A Rhipidomysmastacalis Coendou piehensilis Dasyprocta sp. Proechimyssp. H. hydrochaeris Bolomys lasiurus Calomys callosus C. tener Oryzomys subflavus Thalpomys lasiotus Thrichomysaperoides Oryzomys nigripes B Oxymycterusroberti Cavia aperea

0.25 0.50 0.75 1.00 Figure5. Dendrogramof rodent species relationshipswhen clusteredby presence or absence in the four major macrohabitats,cophenetic correlation coefficient = 0.952. that such high richnessis not foundwithin any one habitator even withina fairlyextensive, but localized, area. Thus, while the cerrado does supporta complexsmall mammalcommunity, faunal lists reported in the literature(e.g. Alho 1982a, Redford& Fonseca,in press)are somewhatmisleading. Very likely such lists accentuate diversitybecause they include species fromthroughout the extensivecerrado region, where numerous other microhabitats are found and where species fromother phytogeographic provinces extend into the cerrado. In some cases, however,elevated species richnessis assumedbecause of lack of sufficientfield researchto determineexactly which species occur in which habitats. Thus, we doubt that any other species occur in the gallery forestswhere our high-intensitytrap gridswere in operation.It is significant thatall butone ofthe species(Agouti paca) we consideras potentialinhabitants of the regionare non-forestspecies: our more limitedtrap effortoutside the forestmakes us less definiteabout speciesoccurrence in thosehabitats. There was verylittle overlap between forest rodent species and thoseof the grasslandhabitats (Figures 3-5). The brejo had no species that was limitedto thathabitat, while both the cerradoand campo had one speciesthat occurred only in that particularhabitat. Campo and cerradowere more similarin their species compositionthan any otherhabitats (Figures 4 and 5), and the brejo Mares, Ernest & Gettinger in Journal of Tropical Ecology (1986) 2. Copyright 1986, Cambridge University Press. Used by permission.

298 MICHAEL A. MARES, KRISTINA A. ERNEST AND DONALD D. GETTINGER

Table 2. Macroniches(locomotor adaptation-dietcombinations) present in each habitat. Codes are combinationsof the codes fordiet and locomotoradaptations (see Table 1)

Habitat

Macroniche Forest Brejo Campo Cerrado

Aq-C + - - Aq-H + + - SAq-O + C-H/F + _ _ C-H _ + T-O/C + - - + T-O + - + + T-O/I + - - T-I + + + T-G/I - _ + + T-G/O - + + T-F/O - + + T-H/F + - _ T/S-H/F - - - + T/A-O + + + + A/T-O/I + - - A/T-O/F + - - A-O + - - A-H/F + - _ Numberof macroniches 13 3 7 8 sharedtwo species withthe campo and only one, the capybara(Hydrochaeris hydrochaeris),with the forest.Thus when habitatsand specieswere clustered, two maingroups developed: the forestand the cerrado(s. 1.). Our analysesshow that the galleryforest is quite distinctfrom all othercer- rado habitatsfrom the standpointof the small mammalfauna. A macroniche analysisagreed almost perfectly with the clusteranalysis based on speciescom- position.This indicatesthat not only is the forestfauna more diverse in terms of numberof species, but also in broad niche types. This can be viewed as supportingthe idea that the galleryforest of centralBrazil is an extensionof both the Amazonian and Atlantic tropical forestsinto the uplands of the country(cf. Cerqueira1982, Lovejoyet al. 1984, Redford& Fonseca,in press). This resultsin a situationwhere cerrado endemics, primarily grassland species such as Plectomysor Juscelinomys,live alongsiderain forestelements such as Rhipidomys,Oryzomys bicolor, or Coendou. The interdigitationof such a rich communitywith complex grass/scrub habitats helps explain the elevatedspecies richnessof mammals,and presumablyof other taxa as well, in the Cerrado Province. The habitatswe have consideredin thisreport are macrohabitatsin thateach is made up of distinctivemicrohabitats. In this paper, we presentthe broad view, but preliminaryanalysis of our extensivedata set forthe galleryforest mammalcommunity shows that microhabitatselection can be quite specific among the species consideredin this report. Unpublishedstudies by L. Z. Nitikmanand M. A. Mareshave shownthat the rodentspecies of galleryforest areas subdividethe forestinto at least fiveprincipal microhabitats, and reveal Mares, Ernest & Gettinger in Journal of Tropical Ecology (1986) 2. Copyright 1986, Cambridge University Press. Used by permission.

Mammalecology in theBrazilian cerrado 299 thatspecies occurrencemay differgreatly from one galleryforest patch to the next. Thus, the galleryforest fauna of the CerradoProvince allows forgreat diversificationin nichetypes. Among the smallmammals are foundspecies that are arborealor terrestrial,aquatic or scansorial,diurnal or nocturnal,insecti- vorous,herbivorous, frugivorous or omnivorous. Our findingscan be interpretedas supportingconservation efforts of gallery foresthabitat in Brazil. The fauna of the cerradois morethan doubled by the presenceof a forestmammal community. Any damageto the galleryforests, such as loggingor burning,will greatlyreduce diversityin the area. Otherdata we have gatheredshow that populationsizes of forestmammals are frequently quitelow, makingthem even more sensitive to habitatdisruption.

ACKNOWLEDGEMENTS We thankthe manyBrazilians who helpedin waystoo numerousto enumerate, althoughwe mustsingle out a few.This workwould have been impossiblewith- out the cooperationof Dr CleberAlho of the Universidadede Brasilia.Domi- ciano Dias of the InstitutoBrasileiro de Geografiae Estatistica(IBGE) was extremelyhelpful during our stay in Brazil. Dr Robert Cavalcantimade the Fazenda Agua Limpa available to us for research.Dr Tom Lacher, Leslie Nitikmanand JanetBraun were of assistancethroughout the project.Drs C. 0. Handley and Al Gardnerassisted with specimenidentification. This research was supportedby National Science Foundation grantsDEB 82-13675 and INT 82-12576 to MAM. Dr Alho's work on thisproject was supportedby a grantfrom the Conselho Nacional de DesenvolvimentoCientifico e Tecnolo- gico (CNPq) of Brazil.

LITERATURE CITED

ALHO, C. J. R. 1981a. Roedores silvestresdo cerrado: habitos e habitats.Pp. 155-156 in Alho, C. J. R. (ed.). Resumos das Comunica,oes Cientificasdo VIII CongressoBrasileiro de Zoologia. Universidade de Brasilia,D.F., Brasil. ALHO, C. J. R. 1981b. Mata ciliar como refilgioda fauna do cerrado em caso de fogo? Pp. 173-174 in Alho, C. J. R. (ed.). Resumos das ComunicagoesCientificas do VIII CongressoBrasileiro de Zoologia. Universidadede Brasilia,D.F., Brasil. ALHO, C. J. R. 1981c. Small mammalpopulations of Braziliancerrado: the dependence of abundance and diversityon habitatcomplexity. Revista Brasileira de Biologia 41:223-230. ALHO, C. J. R. 1982a. Brazilianrodents: their habitats and habits.Pp. 143-166 in Mares,M. A. & Geno- ways, H. H. (eds). Mammalianbiology in South America. PymatuningLaboratory of Ecology Special PublicationsNo. 6, Linesville,Pennsylvania. ALHO, C. J. R. 1982b. Quantitativecomponents of threecerrado landscape habitatsin Brazil. Tropical Ecology 23:125-133. ANDERSON, S. 1982. Monodelphiskunsi. Mammalian Species 190:1-3. CABRERA, A. L. & WILLINK, A. 1973. Biogeografia de America Latina. Washington,D.C., OEA, MonografiasCientl'ficas, Serie Biologica, 13. CERQUEIRA, R. 1982. South American landscapes and their mammals. Pp. 53-75 in Mares, M. A. & Genoways,H. H. (eds). MammalianBiology in South America. PymatuningLaboratory of Ecology Special PublicationsNo. 6, Linesville,Pennsylvania. CHARLES-DOMINIQUE, P., ATRAMENTOWICZ, M., CHARLES-DOMINIQUE, M., GERARD, H., HLADIK, A., HLADIK, C. M. & PREVOST, M. F. 1981. Les mammiferesfrugivores arboricoles noc- turnesd'une foretGuyanaise: inter-relationsplantes-animaux. Terre et Vie 35:341-435. CRESPO, J. A. 1982. Ecologi'a de la comunidad de mamiferosdel Parque Nacional Iguaz'u,Misiones. Revistadel Museu Argentinode CienciasNaturales "BernardinoRivadavia" 3:45-162. Mares, Ernest & Gettinger in Journal of Tropical Ecology (1986) 2. Copyright 1986, Cambridge University Press. Used by permission.

300 MICHAEL A. MARES, KRISTINA A. ERNEST AND DONALD D. GETTINGER

DAVIS, D. E. 1947. Notes on the life historiesof some Brazilianmammals. Boletim do Museu Naciona4 Rio de Janeiro76:1-8. DIETZ, J. M. 1983. Notes on the naturalhistory of some small mammalsin CentralBraziL Journalof Mammalogy64:521-523. EISENBERG, J. F. 1981. The mammalianradiations. University of Chicago Press,Chicago. EITEN, G. 1972. The cerradovegetation of BraziL Botanical Review 38:205-341. EITEN, G. 1974. An outline of the vegetationof South America. Pp. 529-545 in Proceedingsfrom The Symposiaof the FifthCongress of the InternationalPrimatological Society. Nagoya,Japan. EMMONS, L H. 1982. Ecology of Proechimys(Rodentia, Echimyidae) in southeasternPeru. Tropical Ecology 23:280-290. FADEM, B. H., TRUPIN, G. L., MALINIAK, E., VAN DE BERG, J. L. & HAYSSEN, V. 1982. Care and breeding of the gray, short-tailedopossum (Monodelphis domestica). LaboratoryAnimal Science 32:405-409. FEARNSIDE, P. M. 1982. Deforestationin the BrazilianAmazon: how fast is it occurring?Interciencia 7:82-88. FLEMING, T. H. 1970. Comparativebiology of two temperate-tropicalrodent counterparts.American MidlandNaturalist 83:462-47 1. GOODLAND, R. J. 1971. A physiognomicanalysis of the 'Cerrado' vegetationof CentralBrasiL Journal of Ecology 59:411-419. GOODLAND, R. J. & IRWIN, H. S. 1977. Amazonian forestand cerrado: developmentand environ- mental conservation.Pp. 214-233 in Prance, G. T. & Elias, T. S. (eds). Extinctionis forever.New York Botanical Garden,Millbrook, New York. LOVEJOY, T. E., RANKIN, J. M., BIERREGARD, Jr,R. O., BROWN,Jr, K. S., EMMONS, L H. & VAN DER VOORT, M. E. 1984. Ecosystem decay of Amazon forestremnants. Pp. 295-325 in Nitecki, M. H. (ed.). Extinctions.University of Chicago Press,Chicago. MACARTHUR, R. H., DIAMOND, J. M., & KARR, J. R. 1972. Density compensationin island faunas. Ecology 53:330-342. MACARTHUR, R. H., RECHER, H. & CODY, M. 1966. On the relationbetween habitat selectionand species diversity.American Naturalist 100:319-332. MARES, M. A. 1982. The scope of South Americanmammalian biology: perspectiveson a decade of research. Pp. 1-26 in Mares, M. A. & Genoways, H. H. (eds). Mammalianbiology in South America. PymatuningLaboratory of Ecology Special PublicationsNo. 6, Linesville,Pennsylvania. MARES, M. A. 1986. Conservationin South America: problems,consequences and solutions.Science 233: 734-739. MARES, M. A. & OJEDA, R. A. 1982. Patternsof diversityand adaptationin South Americanhystricog- nath rodents. Pp. 53-75 in Mares, M. A. & Genoways, H. H. (eds). Mammalian biology in South America. PymatuningLaboratory of Ecology Special PublicationsNo. 6, Linesville,Pennsylvania. MARES, M. A., OJEDA, R. A. & KOSCO, M. P. 1981a. Observationson the distributionand ecology of the mammalsof Salta Province,Argentina. Annals of the CarnegieMuseum 50:151-206. MARES, M. A., WILLIG, M. R., STREILEIN, K. E. & LACHER, Jr,T. E. 198 lb. The mammalsof north- easternBrazil: a preliminaryassessment. Annals of the CarnegieMuseum 50:81-137. MARSHALL, L G. 1978. Chironectesminimus. Mammalian Species 109:1-6. MYERS, N. 1980. Conversionof tropicalmoist forests. National Academy of Sciences, Washington,DC. NOWAK, R. M. & PARADISO, J. L. 1983. Walker'sMammals of the World,4th Edition. The John Hop- kinsUniversity Press, Baltimore, Maryland. O'CONNELL, M. A. 1982. Population biology of North and South Americangrassland rodents: a com- parativereview. Pp. 167-185 in Mares, M. A. & Genoways,H. H. (eds). Mammalianbiology in South America. PymatuningLaboratory of Ecology,Special PublicationsNo. 6, Linesville,Pennsylvania. PAULA, A. C. DE. 1983. Relac3es espaciais de pequenos mamiferosem uma comunidade de mata de galeria do Parque Nacional de Brasilia. Master's Thesis, Universidadede Brasilia, Departmento de Biologia Vegetal. PINE, R. H. 1982. Currentstatus of South Americanmammalogy. Pp. 27-37 in Mares,M. A. & Geno- ways, H. H. (eds). Mammalianbiology in South America. PymatuningLaboratory of Ecology Special PublicationsNo. 6, Linesville,Pennsylvania. REDFORD, K. H. & DE FONESCA, G. A. B. A zoogeographicanalysis of the non-volantmammalian fauna of the Cerrado.Neotropica (in press). SNEATH, P. H. A. & SOKAL, R. R. 1973. Numericaltaxonomy. W. H. Freeman,San Francisco. SOUZA, M. J. & ALHO, C. J. R. 1980. Distribui'ao espacial do roedor silvestreZygodontomys lasiurus em habitatnatural do cerrado.Brasil Florestal 44:31-74. STREILEIN, K. E. 1982. Ecology of small mammalsin the semiaridBrazilian Caatinga. I. Climate and faunalcomposition. Annals of the CarnegieMuseum 51:79-107.

Accepted17 June 1986