Dromiciops Gliroides Thomas, 1894, Last of the Microbiotheria (Marsupialia), with a Review of the Family Microbiotheriidae

UNIVERSITY OF ILLINOIS LIBRARY AT URBAiMA-CHAMPAIGN BIOLOGY MP Z 2 ""J"*

£ o

Dromiciops gliroides Thomas, 1894, Last of the Microbiotheria (Marsupialia), with a Review of the Family Microbiotheriidae

Philip Hershkovitzt

4ay 28, 1999 •ublication 1502

PUBLISHED BY FIELD MUSEUM OF NATURAL HISTORY to Fieldiana , vrmation for Contributors

MultiMate

References: i should follow B

n the Bi Information Sei I 1983) published by

Refe «d in die fi Press, Stanfi ontane and lowland n , j R L] 51: 567-601. rructvue, ph) -gy, Cultural E J M 1979 »ng the Siona: patterns Publishers. Th , and Stars. Mouton R A St 1 Spirits, Shamans, J. H.. ed.. Handbook of S tribes of pp. 785-821. In Steward. rican Indiai e Andean Ci an Institution, Wa DC. rns and fern allies of Guatemala. Part II. P

'-•-- text <» ot as Illustrations: 111 "plates' ' itements in 8, are detad

.. numbei top. All illustrations should be gure inch uld, whenever practica by ll in uld be mounted on boards in tto obtained Pen and in! ied drawii limitation;

Lack and white prints Red. author upon publication ui Audi

litor.

Page Proofs: I

of Paper). This paper meets the requirements ot ANSI/NISO Z39.48-1992 (Permanence FIELDIANA

Zoology NEW SERIES, NO. 93

Dromiciops gliroides Thomas, 1894, Last of the Microbiotheria (Marsupialia), with a Review of the Family Microbiotheriidae

Philip Hershkovitzt

Curator Emeritus Division of Mammals Department of Zoology- Field Museum of Natural History- Roosevelt Road at Lake Shore Drive Chicago, Illinois 60605-2496 USA

Accepted June 14, 1996 Published May 28, 1999 Publication 1502

PUBLISHED BY FIELD MUSEUM OF NATURAL HISTORY © 1999 Field Museum of Natural History ISSN 0015-0754 PRINTED IN THE UNITED STATES OF AMERICA Dromiciops gliroides Thomas, 1894, Last of the Microbiotheria (Marsupialia), with a Review of the Family Microbiotheriidae Dromiciops gliroides Thomas (Isla Chiloe, Chile). Photograph courtesy of Sr. Felipe Montiel Vera. m/^

Table of Contents 9. Testis of Dromiciops 18 10. Rete testis sections 19

1 1 . Rete testis of various marsupials 20 Abstract 1 12. Cloaca in Dromiciops and Thylamys 21 Introduction 1 13. External ear of Dromiciops and Order Microbiotheria Ameghino, 1887, Marmosa 22 Cohort Microbiotheriomorphia 1 4. Mammary field of Dromiciops 23 2 Ameghino 15. Tarsal bones 24 Biological Origin 5 16. Upper dental system of Dromiciops 25 Geographic Origin 7 17. Fragment of left ramus of Adinodon Weddellian Biotic Province of pattersoni 26 Antarctica 7 1 8. Caecum of Dromiciops 27 Weddellian Land Habitats 7 19. Geographic distribution of Dromiciops Weddellian Land Mammals 8 gliroides 29 Recorded Microbiotheres 9 20. Caudal fields of Dromiciops and Biogeography: The Nothofagus- Metachirus compared 30 C«t/S0l/£4-MlCROBIOTHERE ASSOCIATION ... 10 2 1 . Cheiridia of Dromiciops and Small Rodent Associates 11 Metachirus nudicaudatus 31 Family Microbiotheriidae Ameghino 12 22. Four types of spermatozoa 36 Genus Dromiciops Thomas, 1894 12 23. in female and Microbiotheriid Characters Distinctive Reproductive system male Philander and female Dromiciops .. 38 Among Marsupials 13 24. Skull of Dromiciops 40 Subsidiary Characters 18

25 . of skull 41 Epitome 25 Diagrams Dromiciops 26. bullae 44 Dromiciops gliroides Thomas 26 Auditory Acknowledgments 56 27. Transitional vertebrae 45 Addendum 56 28. Thoracolumbar flexure 46 Gazetteer 56 29. Left upper and left lower tooth rows of Literature Cited 57 Dromiciops, Philander, Caenolestes 47 30. 2 3 1 in p , dp , m Dromiciops; p 2 , dp,, m, in occlusal of Dromiciops; aspect p 2 , dp,, m, in Dromiciops 50 in List of Illustrations 31. p,_ 2 , p, erupting, dp, functioning, m, Philander 51

32. 2 3 3 1 2 p , p erupting, dp , m , m , 3 Frontispiece: Dromiciops gliroides Thomas pseudoquadritubercular m 52

Plate 1 . Thomas 2 Dromiciops gliroides 33. Karyotype of Dromiciops 53 Plate 2. Nest and nestlings of Dromiciops gliroides Thomas 3 1. Geographic distribution of living Dromiciops 4 2. Odyssey of the Microbiotheria- List of Tables Nothofagus-Chusquea association 5 3. Bamboo in upper and lower stories 6 4. Leaves of the bamboo Chusquea 1. and dental culeou 8 External, cranial, measurements 32 5. Mandibular symphyses of Dromiciops and Marmosa 14 2. Measurements of seminiferous tubules and rete testis 34 6. Mandibular symphyses of various 3. Postcranial skeletal measurements and marsupials 15 42 7. Auditory bulla of Dromiciops, auditory proportions bulla of a Microbiotherium 16 4. Dental symbols 48 8. Posteroventral portion of Dromiciops 5. Sequence of dental eruption 52 cranium 17 6. Dromiciops nests and occupants 54

Dromiciops gliroides Thomas, 1894, Last of the Microbiotheria (Marsupialia), with a Review of the Family Microbiotheriidae

Philip Hershkovitzf

Abstract

The tiny Chilean and bordering Argentinian endemic Dromiciops gliroides Thomas, lone survivor of the South American Cohort Microbiotheriomorphia, is basal to all known marsupials. Indications are that the ancestral microbiothere may have originated in South America during the Late Mesozoic in a cool, humid, low-latitude biome dominated by a beech-bamboo (Nothofagus-Chusquea) plant association, the bamboo being its primary nesting material. A postulated worldwide climatic warming initiated a southward migration of cool-loving elements of the Southern Hemisphere. The philopatric microbiotheres evidently clung to their Nothofa- gus-Chusquea niche and nesting material as the association shifted from subequatorial into Patagonian and Antarctic latitudes. A climatic reversal during the Tertiary returned the Noth- ofagus-Chusquea-vmcvobioihcvc association to Patagonian latitudes. An expanding arid scrub savanna farther north, however, halted the Nothofagus community shift in that direction. At the same time, habitat was being lost to the increasingly colder climate advancing from the south. Extinction followed. The most northern progression of the association was in the west, where rains intercepted by the rising Andes provided a favorable environment. The Nothofagus- Chusquea-microbiothere association now survives in the cool, humid Valdivian region of Chile and its narrow Argentine extension as a relictual enclave sharply delimited by the warmer arid environment to the north and the colder, drier environment to the south. The present ecogeographically restricted Cohort Microbiotheriomorphia is described and compared with its sister Cohort Didelphimorphia, which, by virtue of its adaptability, fecundity, and diversity, had dispersed into all continents. The narrow climatically controlled distribution and what little is known of the life history and anatomy of the single surviving microbiothere, Dromiciops gliroides, are reviewed.

Introduction recognize the relationship between the living Dromiciops and the extinct microbiotheriid mar- then known from the Miocene. The monito del monte, Dromiciops gliroides supial only The with its Thomas (frontispiece; Plates 1, 2), a mouse opos- Microbiotheriomorphia, single sum with a simian vernacular name, is the lone known family Microbiotheriidae, was reviewed Marshall with attention to cra- survivor of the South American marsupial family by (1982), given Microbiotheriidae Ameghino, 1887. The monito nial, dental, and external characters. The clade as a its dis- del monte and a dozen or so known extinct spe- was treated family of Didelphoidea, cies of the same family compose the monophy- tinctive characters unperceived except for men- letic Cohort Microbiotheriomorphia Ameghino. tion of small canine teeth. It has since been de- Together with Cohort Didelphimorphia Gill, 1872, termined (Hershkovitz, 1992a) that microbiothe- it constitutes the Infraclass Marsupialia, also riids are monophyletic and must have arisen at known as Metatheria. Reig (1955) was the first to some time before the divergence of didelphoids,

FIELDIANA: ZOOLOGY, N.S., NO. 93, MAY 28, 1999, PP. 1-60 Plate 1 . Dromiciops gliroides, natural size: top, note incrassate prehensile tail displayed by adult; bottom, same individual cupped in hand. Animal captured in Chile by Dr. Bruce Patterson and donated to the Chicago Zoological Society, Brookfield, Illinois. Photograph by Mike Greer; original photograph courtesy of the Zoological Society (reproduced from Hershkovitz, 1992a).

which are characterized by their distinctive stag- of south central Chile, from the latitude of Con- gered lower third incisor (Hershkovitz, 1995). ception (ca. 37°00', 72°30') to the southern border of Isla Chiloe (ca. 44°00', 72°00'), thence east- ward into of in western Order Microbiotheria Ameghino, adjacent parts Argentina and southwestern Rio PAST: 1887, Cohort Microbiotheriomorphia Neuquen Negro. El Molino-Santa Lucia For- Ameghino Successively: Upper mation, Cochabamba, south central Bolivia (Early Distribution (Figs. 1-4)—PRESENT: Cool, Paleocene; see Marshall et ah, 1995, for details); humid, dominantly Nothofagus-Chusquea forests Itaboraian Formation in Rio de Janeiro, south-

Plate 2. A, Arboreal nest of Dromiciops gliroides constructed of moss and the leaves of bamboo (Chusquea culeou). B, Nestling. C, Young taken from back of mother. (Copied from photographs by Mann, 1958.)

FIELDIANA: ZOOLOGY HERSHKOVITZ: LAST OF THE MICROBIOTHERIA c c 55 65 e 60

"35° 35°

55 c

— 40°

40° 60°

. Caradon Vaca

f I fCanodon Hordo 45°

Puerto Deseado MICRDBIOTHERIIDAE +

50° Monte Leon La Cueva Oueqiia Quanada 50° Santa Cruz Monte Observacion Killik Aike Norte (Fel ton's Estarcia Rfo Gal legos

C5^

Fig. 1. Distributions of living Dromiciops gliroides in Chile and bordering Argentina (see Fig. 19, p. 29, for locality records, and p. 9 for names and ages of extinct Oligocene-Miocene microbiotheriids in southern Argentina).

FIELDIANA: ZOOLOGY Fig. 2. Geoclimatic stages in the odyssey of the Nothofagus-Chusquea-m'icrob'wtheriid association in southern

South America and Antarctica. Stage 1 : South central Bolivia (Early Paleocene). Stage 2: Sao Jose de Itaboraf, Rio de Janeiro (Middle Paleocene). Stage 3: La Meseta Formation, Seymour Island. Antarctic Peninsula (Eocene). Stage 4: Reversion to southern Patagonia (Argentina) (Oligocene-Miocene), continuing into (Stage 5) central Chile (Ter- tiary-Recent). Arrows point to direction of migration with changing climates.

eastern Brazil (Middle Paleocene); Colhuehuapian eutherian-like grade characters may have ap- and Santacrucian formations from Lago Colhue- peared in latest Jurassic or earliest Cretaceous. Its Huapi (46°30'S) to Rio Gallegos (51°36'S), Chu- origin may have been from the same metatherian but and Santa Cruz provinces, Argentina (Oligo- stock that later gave rise to Cohort Didelphimor- cene, Miocene); La Meseta Formation, Seymour phia, distinguished by the staggered i„ precaudal Island, Antarctic Peninsula, Weddellian Province, cloaca, and other derived characters fully evolved Antarctica (Eocene). by late Early Cretaceous (Hershkovitz, 1982, 1992a, 1995) but inferentially not before the appearance of the microbiotheriids, which lack these Biological Origin derived characters. The tree showing phylogenetic Monophyletic Microbiotheriomorphia with its position and relationship of didelphoids (Hersh- combination of prototherian-, metatherian-, and kovitz, 1992a, p. 210) is subject to modification

HERSHKOVITZ: LAST OF THE MICROBIOTHERIA FIELDIANA: ZOOLOGY within a Metatheria perceived as independently South America, New Zealand, and encompassed differentiated from a therian stock. islands. The southern provincial border was determined as the Transantarctic Mountain front; the northern limit was said to coincide with the edge of the continental shelf, including southern Ar-

Geographic Origin gentina and central Chile (map. Fig. 2, p. 5).

The oldest known didelphoids are North Amer- ican Albian age (late Early Cretaceous: Adinodon [Hershkovitz, 1995], Kokopellia [Cifelli, 1993], and Holoclemensia [Slaughter, 1968a,b]; see also Weddellian Land Habitats Turnbull, 1995). Both the phylogenetically older microbiotheres and the geologically younger di- Forest-covered habitats within a Ant- delphoids are known as fossils from the Bolivian temperate arctic zone such as the Meseta Formation of the El Molino Formation of earliest Paleocene (Mar- Island were known shall et al., 1985, 1995; Marshall & de Muizon, peninsular Seymour (Fig. 2) to exist Permian 1988). It has been argued that the earlier Albian during through Early Tertiary pe- age of some didelphoids presupposes a North riods (Seward, 1914). Woodburne and Zinsmeister American origin of marsupials, microbiotheres (1984, p. 935) note that the "presence of large not having been distinguished at the time from the logs (up to a meter in diameter), together with derived didelphoids. Microbiotheres, however, are other plant debris indicates that the Antarctic Pen- unknown in North America. All locality records, insula was still heavily vegetated during the Eo- whether of extant or known extinct microbio- cene [and that] preliminary analysis of the abun- theres, are from well south of the equator in South dant palynomorph floras of the La Meseta For- Pen- America, including one from the Antarctic mation [of Seymour Island] indicates that condi- insula et Goin (Carlini al., 1991; & Carlini, 1995). tions on the northern part of the Antarctic It that the continent of mar- appears, therefore, Peninsula were similar to present-day humid-tem- is based on the supial origin properly phyloge- perature climates in Tasmania, New Zealand, and netically basal South American or Weddellian mi- southern South America." Similar conclusions crobiotheres and not the (tentatively) geologically were reached by Doktor et al. (1996) from the older but phylogenetically younger North Ameri- plant and fish assemblages from the Eocene La can didelphimorphs. Meseta Formation. In my discussion of marsupial ankle bones and Dettman (1989, p. 89), in her paper "Antarc- phylogeny (Hershkovitz, 1992a, p. 206), the hab- tica: Cretaceous cradle of austral temperate rain- itat of the Microbiotheriomorphia was inadver- forests?" observed that "fossil evidence from tently given as North American. It should have Antarctica and associated in the been South American. closely regions Cretaceous southern Gondwana assembly con- firms that Antarctica was a Cretaceous originator and dispersal region of certain elements of today's Weddellian Biotic Province of southern hemispheric humid and prehumid for- Antarctica ests. Antarctica origins are indicated for the fern Lophosoria, podocarp gymnosperms, and several of the to their The Weddellian Paleogene-Cretaceous Biotic lineages Proteaciae; migration present of distribution was Province was distinguished by Zinsmeister (1976) regions probably step-wise. on the basis of similarities between the late Cre- Antarctica also served as a Cretaceous dispersal taceous and Early Paleocene shallow-water mol- corridor for other angiosperms represented today luscan faunas of Antarctica, Australia, southern in mid to low latitude austral regions."

Fig. 3. Top. bamboo (Chusquea sp.) in second growth forest of La Petard State Park. Iporanga, Sao Paulo. Bottom. Chusquea in lower story clearing of La Petard State Park. Sao Paulo. Microbiothcriids are not known to have occurred in Sao Paulo (see map. Fig. 1, p. 4). Photographs by Barbara Brown, December 1989.

HERSHKOVITZ: LAST OF THE MICROBIOTHERIA 2mm

Chusquea culeou

Fig. 4. Leaves of the bamboo Chusquea culeou, used by Dromiciops for weaving its waterproof nest.

Weddellian Land Mammals burne and Zinsmeister (1984, p. 933), "the Sey- mour Island polydolopids represent a sample of a The first land mammal known from Antarctica, mammal population that likely had been endemic described by Woodburne and Zinsmeister (1984), to the region for about 10 Ma, and had actually is the polydolopoid marsupial Antarctodolops dai- dispersed from South America at about 50 Ma, lyi from the Late Eocene of Seymour Island, when the group was most abundant and diverse northern Antarctic Peninsula. The degree of bio- there. . . . We also conclude from the number of logical differentiation from its nearest South specimens represented, and geological evidence

American relative suggested "a late Cretaceous . . . that the entry of polydolopids into Antarctica presence of marsupials in Greater Antarctica and reflects waif dispersal via an archipelago between their isolation in Australia by whatever means by southern South America and the Antarctic Pen- the close of that period." In the opinion of Wood- insula rather than a continuous land connection."

FIELDIANA: ZOOLOGY The overall evidence suggests, continue Wood- crept in. Compare, for example, lower canines of burne and Zinsmeister (1984, p. 942), that "mar- numbers 2 and 6 in Marshall (1982, p. 47). Iden- supials must have been present on Antarctica tification of number 2 may be equivocal. It carries sometime from the Late Cretaceous to the Eocene. the same name and registry number, PU 15038, The Oligocene and younger marsupials on Aus- of the figured auditory region of Microbiotherium tralia . . . require an earlier—probably a much ear- tehuelchum (Marshall, 1982, p. 30), which is a lier (because of their diversity in the Oligocene true microbiothere. Neither it nor the aforemen- and Miocene)—ancestor there or in a closely tioned figured canine belong to the same taxon or neighboring place. That place was Antarctica to the neotype of M. tehuelchum Ameghino, 1 887 from at least 80 Ma and, diminishingly, to about (MLP36). The identification of Mirandatherium, 38 Ma ago." although not queried, requires reexamination. The The recovery of a ground sloth "Megatherioi- Pediomyinae, generally classified as microbio- dea" reported by Carlini et al. (1991, p. 15) added therioid (cf. Marshall et al., 1990), are not so re- another order of South American mammals to the garded here. The named forms follow. known Late Eocene fauna of the La Meseta For- mation. Other remains in the same Seymour Is- Microbiotherium acicula Ameghino, 1891 land Formation and underlying Cross Valley (Marshall, 1982, p. 12)—Santacrucian members of the Sobral Formation (Paleocene) re- (Monte Observacion), Santa Cruz Prov- ported by Carlini et al. (1991) are a "probable ince, Argentina (Miocene). microbiotheriid, polydolops, birds and plants." Microbiotherium patagonicum Ameghino, Isolated teeth of the unique South American deer- 1887 (Marshall, 1982, p. 14)—Santacru- like "ungulates" representing the families Theri- cian, Santa Cruz Province, Argentina dodontidae and Trigonostyliodae, recovered by (Miocene). Bond et al. (1990, p. 3) from the Seymour Island Microbiotherium divisum Ameghino, 1902 Formation, are the first eutherians known for the (Marshall, 1982, p. 22)—Colhuehuapian region. beds, at the Barranca south of Lago Col- Discovery by Pascual et al. (1992a,b) of a hue-Huapi, Chubut, Argentina (Miocene). platypus (Monotremata) tooth in sediments of the Microbiotherium tehuelchum Ameghino, 1887 Banco Negro Inferior (45°30'S, 67°1 l'W), Chubut (Marshall, 1982, p. 26)—Santa Cruz For- Province, southern Patagonia or within the Pata- mation (Corriquem-Kaik; Quequa-Quema- gonian Terrane, a subset of the Weddellian Prov- da; La Cueva; Monte Observacion; Killik ince, added another biogeographic dimension to Aike) Santa Cruz, Argentina (Miocene). the region. Microbiotherium praecursor Ameghino, 1898 (Marshall, 1982, p. 19)— "Cretaceo inferior." Age and locality unknown. Microbiotherium gallegosense Sinclair, 1906

Recorded Microbiotheres (Marshall, 1982, p. 35)—Santa Cruz For- mation, N bank Rio Gallegos, Santa Cruz, The 13 recorded extinct species, including 2 Argentina (Miocene). undescribed (types, tentatively and dubiously as- IPachybiotherium acclinum Ameghino, 1902 signed individuals), and 1 living species, distrib- (Marshall, 1982, p. 61)—Colhuehuapian uted among the 7 genera, are listed below (and in beds, probably Barranca, Chubut, Argen- Fig. 2). References to original descriptions, tax- tina (Miocene)—microbiotheriid identity

onomies, synonymies, geographic data, and fig- questioned by Marshall (1982, p. 63). ures of teeth and bone fragments of the first 8 ?Eomicrobiotherium gaudryi Simpson, 1964 species are from Marshall (1982). The determi- (Marshall, 1982, p. 58)—Barranca, S Lago nations of the fossils as microbiotheres are ac- Colhue Huapf, Chubut, Argentina (Mio- cepted as tentative from Marshall (1982, 1987), cene)—The genus reviewed by Marshall

Marshall and de Muizon (1988), and Marshall et (1982, p. 57) includes a second, unrevised al. (1990). Description of Pucadelphys andinus is species, E. gutierrezi del Corro, but the mi- from Marshall and de Muizon (1988). Discussion crobiotheriid affinities of the genus are of the genus as a microbiothere is based on data problematic (Marshall, 1982, p. 59). from Marshall et al. (1995). Pucadelphys andinus Marshall and de Muizon, Some discrepancies in earlier reports may have 1988—Tiupampa local fauna. El Molino-

HERSHKOVITZ: LAST OF THE MICROBIOTHERIA Santa Lucia Formation, Cochabamba, Bo- Biogeography: The Nothofagus- livia (Early Paleocene). CTiMSrfjrwtfa-Microbiothere Association Khasia cordillerensis Marshall and de Muizon, 1988—Tiupampa local fauna, El Molino Knowledge of living microbiotheres and nest- Formation, Cochabamba, Bolivia (Early ing preferences indicates that the habitat of Drom- Paleocene). iciops is restricted to the Nothofagus-Chusquea Mirandatherium alipioi Paula-Couto, 1962— association of central Chile and bordering Argen- Sao Jose de Itaborai, Rio de Janeiro, Brazil tina. Fossil evidence suggests that habitat prefer- (Middle Paleocene); Marshall (1987, pp. ence of Dromiciops may have changed little in 114, 137, 148). time. The fossils also record the latitudinal trans-

Microbiotherium sp., Antarctica, La Meseta lations experienced by microbiotheres concomi- Formation, Seymour Island, Antarctic Pen- tant with the climatic shiftings of their nearly ob- insula, Late Eocene. Carlini et al. (1991); ligate Nothofagus-Chusquea-dominated habitat Goin and Carlini (1995). (Figs. 2-4). Dromiciops gliroides Thomas, 1894—Isla de The oldest known microbiofherians are from Chiloe; Los Lagos, Central Chile (Recent). the El Molino Formation in Tiupampa, Cocha- bamba, Bolivia (Marshall & de Muizon, 1988, p. 23; Marshall et al., 1995), now regarded as Early Among 169 newly recovered dental specimens Paleocene. The next chronological record of mi- of Colhuehuapian-Santacrucian age, Bown and crobiotherian occurrence is the mid-Paleocene It- Fleagle (1994) distinguished the teeth of three aborafan of Rio de Janeiro, southeastern Brazil. species of the genus Microbiotherium from Oli- This suggests a climatic change that shifted the gocene and Miocene Patagonian rocks. The ma- Bolivian Nothofagus-Chusquea biome with its terial included a new species of Colhuehuapian microbiotheres to the more southern or possibly Microbiotherium, closest in size to M. praecursor, then cooler latitudes of coastal Rio de Janeiro. a new Santacrucian Microbiotherium from the Continuation of the warming trend southward led Pinturas Formation, intermediate in size between the Nothofagus-Chusquea-microbiothere associ- M. acicula and M. praecursor, and the common ation into Antarctica during the Eocene (Carlini Pinturas and Santacrucian M. tehuelchum. et al., 1991; Goin & Carlini, 1995) and led to the Four partial skeletons and an extra skull of Pu- extinction of the Itaboraian species of the lower cadelphys andinus Marshall and de Muizon latitudes. (1988) recovered from the Santa Lucia Formation A climatic reversal brought freezing tempera- of Tiupampa, about 95 km SE of Cochabamba tures into Antarctica, ending the southward ad- (65°35'W, 18°02'S), Cochabamba Department, vance of the Nothofagus-Chusquea-microbio- south central Bolivia, were studied by Marshall et there association and shunting it back into the al. (1995). The geochronological position had pre- warmer Patagonia of the then Oligocene-Mio- viously been, and continues to be, referred to as cene. The northward trend, however, was re- the El Molino Formation. strained by the more rapidly advancing arid scrub Skeletal measurements provided by Marshall et savanna arriving from the north, and habitat loss al. (1995, p. 156) reveal Pucadelphys andinus to to the falling temperatures moving in from the be relatively small, comparable in size to living south. Rains intercepted by the western versant of caenolestids, marmosids, and the microbiothere the rising Andes provided scope for continuation Dromiciops gliroides. Cranially, Pucadelphys, of the northward trend of the Nothofagus-Chus- with its more or less evenly proportioned skull, gwea-microbiothere association into the Valdivian resembles Dromiciops as contrasted with the long region of Chile and bordering parts of western rostrate caenolestid and narrowly triangulate mar- Argentina during the Neogene. Continued favor- mosid skulls. Notwithstanding, caenolestids, like able conditions at least locally into the Quaternary Pucadelphys, but unlike the others, are terrestrial, allowed the association to persist as a relict of the non-prehensile-tailed, and locomotorily similar. once cool, humid Tertiary Antarctic climate (Fig. Nevertheless, phylogenetic, morphological, and 2). The Tertiary microbiotheres east of the Andes behavioral comparisons and alignments of Puca- in southern Argentina disappeared as their special delphys made by Marshall et al. (1995) were with habitat gave way before the advancing aridity of the least likely, most derived, and unrelated, stag- the north and the cold climate of the south. gered-toothed didelphoid Didelphis. The Chilean-Argentine habitat of surviving

10 FIELDIANA: ZOOLOGY Dromiciops, shown juxtaposed against that of the Nahuel Huapf Park's glacier-carved valleys are extinct Tertiary Argentine microbiotheres (map, covered by

Fig. 1 ), documents the geographic positions of the latest climatic shift of microbiothere habitat. great native forests dominated by beech trees of the genus with bamboo as a Mammalian survivors of what had been a rich Nothofagus, major understory component. Although man has had relatively little im- Late Cretaceous-Tertiary Weddellian mammalian pact on these forests, glaciers, avalanches, mudslides, fauna are monotremes, the diversified American volcanic ashfalls and fires have all influenced the distri- didelphoids, their Australian descendants, and the bution of the beech trees and the bamboo. Precipitation also has had an effect: bamboo is more abundant in the microbiothere line that led to Dromiciops gli- wet, mild climate near the Chilean border (over 4000 roides. The latter is now confined to its shrinking mm annual precipitation), and disappears toward the dri- habitat in the cool, humid Val- man-endangered er, colder eastern edge of the park 60 km away (600 mm forests of divian Nothofagus-Chusquea Chile and annual precipitation). Chusquea grows and thrives near Nahuel 760 adjacent Argentina. Lago Huapf (elevation m), where most pre- falls as rain, but is also found to 1450 m. The climatic changes described above have cipitation up ... It occurs both in pure stands in the open as well as been inferred from the stepwise shifts of habitat beneath the dense canopy of Nothofagus forests. and fauna that might have begun north of and earlier than the time of the El Molino Formation in The foregoing is, in effect, a description of the Bolivia. Fossil records the reconstruc- support ecological niche of the philopatric Dromiciops. tions. The climatic shifts appear to be in rough agreement with latitudinal paleotemperature re- constructions by Frakes et al. (1994). The philopatric microbiotheres, linked to their Nothofagus- Small Rodent Associates Chusquea association, moved with it first southward, then back northward. Small mammals observed by A. K. Pearson et The ecologically based reconstruction by Pas- al. (1994, p. 118) frequenting the Chusquea cu- cual and Juaraguizar (1990, pp. 23-60) of chang- leou thickets in the Argentine Parque Nacional ing climates and evolving vertebrate faunas of the Nahuel Huapf include the mouse Irenomys tar- southern half of South America and northwestern salis (Sigmodontinae), one of which they wit- Antarctica throughout the Cenozoic is panoramic nessed "shinny up bamboo culms and climb in scope. Didelphimorphs are accorded prime about in the foliage. Dromiciops australis [they treatment but microbiotheres are not mentioned. observed] . . . uses bamboo leaves to construct its A report by A. K. Pearson et al. (1994) on the spherical nest. A variety of small rodents eat the biology of the bamboo Chusquea culeou of the underground rhizome buds and emerging shoots Dromiciops-Nothofagus-Chusquea association is of C. culeou, among them Abrothrix longipilis, relevant to the subject. The pertinent parts are Aconaemys fuscus, Irenomys tarsalis, Auliscomys summarized below. [= Loxodontomys] micropus and Ctenomys men- ,, The groves of Patagonian bamboo, Chusquea docinus. A. K. Pearson et al. (1994) cited O. P. culeou (Fig. 3), used by Dromiciops gliroides for Pearson (1983) for the rodents. The long vegeta- habitat, and its bamboo leaves (Fig. 4) for nest tive phase of the Chusquea culeou, more than 54 construction, were studied for over 7 years by A. years before seeds are produced, eliminates the K. Pearson et al. (1994), mostly in the Parque bamboo as a regular source of food for Dromi- Nacional Nahuel Huapf in southwestern Argenti- ciops. The delayed production of bamboo seeds na. The park and the region westward into the when it comes culminates in an explosive increase same latitudes of Chile to the coast and Isla Chi- of mice, for which I had proposed the Spanish loe include many known Dromiciops habitats. ratada (Hershkovitz, 1962, pp. 42, 277). According to A. K. Pearson et al., three species Microbiotheres may never have been abundant of bamboo occur in the Argentine national park. or widely dispersed from their Nothofagus-Chus- Only Chusquea culeou is widely distributed. It oc- quea association in the Southern Hemisphere. In- curs between 35°S and 47°S. In Argentina, it cov- sofar as known, differentiation at any time was ers a narrow band along the eastern Andean restricted to no more than one to a few species, slopes of the Provinces of Neuquen, Rio Negro, the genera questionable. In sharp contrast, the har- and northern Chubut, or coincident with the Ar- dy, adaptable, prolific, speciose didelphimorphs gentine portion of the Dromiciops range. Accord- spread in all directions, populated three conti- ing to the authors (Pearson et al., 1994, p. 94) the nents, and experimented in the other three.

HERSHKOVITZ: LAST OF THE MICROBIOTHERIA 11 Remarks—Resemblances between Dromiciops ning microbiotheres in Australia are based pri- and living Australian marsupials noted by Dyzen- marily on the erratic hypothesis of ankle bone chauz et al. (1993, p. 82) must be attributed to morphology (Szalay, 1982). Nevertheless the retention of shared characters such as a i primitive staggered 3 as a significant phylogenetic marker karyotype of 14 chromosomes or presumed sin- separating didelphimorphs and microbiotheriids gle-headed sperm. There is no positive evidence appears to have been recognized by Woodburne that microbiotheres originated or ever lived in and Case (1996). " much less that Microbiotherium dis- Australia, Woodburne and Case's (1996, p. 142) exposi- persal would likely have been from rather than to tion of their stand is summarized in their words: Australia" (Kirsch et al., 1991, p. 10465). Mar- supials and placentals were unknown in Australia before the Eocene (Archer, 1993, p. 8). Long be- Hershkovitz (1982, 1992[a], 1995) posits that Drom- fore that, in the latest Cretaceous or earliest Pa- iciops, and by inference all microbiotheres, are more microbiotheres were in Boliv- leocene, disporting primitive than all other marsupials based on the lack of ia de (Marshall & Muizon, 1988). Furthermore, a staggered i3 [= i3]. Hershkovitz (1995) proclaims that are much older than consid- Australian marsupials are derived from staggered didelphoids (a) commonly ered as based on other such as a cen- lower third incisor didelphoids (Hershkovitz, features, V-shaped trocrista (upon which he does not comment), but also 1982, 1995), not from microbiotheriids. Szalay's that the members of this group are plesiomorphic in hav- (1982a,b) postulate that microbiotheriids together ing this trait (p. 160, no. 5) [see my reply, pp. 14, 16, all Australian form a with marsupials monophy- to their egregious perversion of my statements]. Other letic than not a i3 i there is clade based primarily on ankle bone mor- having staggered [= 3 ], nothing in Hershkovitz (1995) to that microbiotheres are phology does not agree with the data (cf. Hersh- suggest more primitive than all other marsupials. Pending further kovitz, 1992a). evidence, the [DNA] results of Springer et al. (1994, 1996 [in press]) and Retief et al. (1995) are followed here, i.e., that Dromiciops is rooted within the Austral- idelphia, rather than being the sister taxon of all mar- Family Microbiotheriidae Ameghino supials.

Genus Dromiciops Thomas, 1894

Type Species—Dromiciops gliroides Thomas, The DNA-DNA hybridization studies of Wester- 1894. man and Edwards (1991), showing a trichotomy The scenario of land mammal dispersal among of divergence between microbiotheres, didelphids, South America, Antarctica, and Australia in the and dasyurids, were not considered but are no less Late Cretaceous to Early Tertiary epochs, de- speculative. scribed by Woodburne and Case (1996), is based Although Woodburne and Case cited Hershkov- on a review of paleontological, phylogenetic, geo- itz, 1992a, they failed to note that lower incisor physical, and climatic evidence. They regard the orientation is but one of a number of distinctive Microbiotheriidae as australidelphian and dating features described. Those features show that mi- from 63 Ma. They ask (1996, p. 138), without crobiotheres must have arisen at a time before offering a solution, "to what order of Australian those archaic or unique external, cranial, and den- marsupials do microbiotheriids exhibit a special tal characters seen only in Dromiciops had dis- relationship?" The answer is, none. Microbiothe- were or most were riids are not Australian. There is absolutely no appeared, disappearing, likely evidence that microbiotheriids occur on that con- never present in other fully developed marsupials. are the basic therian characters that tinent, or ever did. Their appearance in South They among and America predates the arrival of any marsupials in absolutely decisively separate Dromiciops Australia (Archer, 1993). Recorded similarities (the only known complete microbiotheriid) from all other one of those between Dromiciops gliroides and some species marsupials. Any cranial, of Australian marsupials and eutherians must be dental, or external characters mentioned and num- attributed to parallelisms, convergences, or com- bered 1 to 7 below, invalidates the postulate that mon retentions of primitive marsupial or therian Dromiciops or a Dromiciops-like morphotype is characters. or was near kin to Australian didelphimorph mar- The Woodburne and Case arguments for pin- supials and their American originators.

12 FIELDIANA: ZOOLOGY Microbiotheriid Characters Distinctive monplace among reptiles, particularly therapsids Among Marsupials (cf. Romer, 1966, p. 182), where it is often correlated with rounded incisive arch and the basi-

1. Four lower incisors evenly spaced in Dromi- cranial sagittal presphenoidal crest. All appear to ciops (Figs. 5, 6) be plesiomorphic features.

The maximum j incisor formula of marsupials 3. Entotympanic bone component in auditory is believed to have been derived from a 5 formula. bullae of Dromiciops and Microbiotherium Loss of the first lower incisor is attributed to con- (Fig. 7) traction of the lower jaw at the birth of the earliest metatherian or ancestral form (Hershkovitz, 1982, The earliest known marsupial auditory bulla is that microbiothere p. 195) or may have been inherited directly from of the Tiupampa Pucadelphys the ancestral therapsid. The remaining four inci- andinus (Marshall & de Muizon, 1995). The bulla sors are not crowded in microbiotheriids. is of the tripartite type (Hershkovitz, 1992b, p. 25, The evolutionary sequence of mandibular con- Fig. 10), with the hiatus between alisphenoid and traction in marsupials, Dromiciops excepted, petrous wings of the bulla. The Dromiciops bulla caused further crowding and elimination of addi- is a large, highly derived, inflated globe composed tional lower incisors (cf. Marshall et al., 1990, p. of the tympanic wing of the alisphenoid bone, the 466). The contraction in didelphoids, dasyurids, joined tympanic wing of the petrous bone, a great- and others forced the root of the numerical second ly pneumatized mastoid bone with the mastoidal lower incisor, or phylogenetic third, out of line process absorbed, a narrow lamina of the basi- with adjacent incisors (Figs. 5, 6). An upgrowth sphenoid bone, and a sutured ventromedial bone of the alveolus on the buccal side of the staggered between alisphenoid and petrous wings identified tooth appears as a buttress. Attenuation or elon- as "entotympanic," an element not present in any gation of the mandibular body reduced or elimi- other marsupial. The bone is not homologous with nated the staggered condition in many Australian the so-named entotympanic bones of eutherians. marsupials. The condition is discussed and figured That of Dromiciops is interpreted as an adventi- by Hershkovitz (1992a, p. 200; 1995). tious element that, during the course of auditory Uncrowded four lower incisors are qualities of bullar evolution, filled the midventral and medial Dromiciops and other known microbiotheriids. gap before the approximating alisphenoid and pet- The condition may be directly derived from the- rous wings might have closed it. Closure by the rapsids with upper incisive formulae of the pre- latter element is almost complete in Caluromys maxilla ranging from 8 to 4; other shared therap- (Caluromyidae) and some others. sid traits are rounded incisive arch, short mandib- Intervention of the entotympanic bone for com- ular symphysis, and the presphenoidal sagittal plete bullar gap closure in Dromiciops is a unique crest. microbiotheriine trait fully accomplished in the Miocene Microbiotherium sp. (Fig. 7B; see Se- 2. Symphysis menti and lower incisive arch of gall, 1969). The bulla as described above is fig- Dromiciops and Marmosa (Figs. 5, 6) ured in a skull of Dromiciops by Hershkovitz

(1992a, p. 197; 1992b, p. 24, Fig. 10). "The incisive arch of Dromiciops is rounded, the mandibular symphysis shallow and extends 4. Presphenoidal sagittal crest of the basicranium to line i back a between 4_5 , sometimes between (Fig. 8, labeled s) i -c. all 5 In other marsupials examined, the arch and symphysis are angular the latter terminating A sagittally keeled presphenoid of the meso- behind at a line between lower canine and pre- pterygoid fossa is present among living marsupi- molars. ... In eutherians, the symphysis also ex- als only in Dromiciops (not seen in others). The tends back to a line between lower canine and midventral crest or keel extends back from the premolars but .... dental formulae and diaste- vomer for the length of the basicranium to the mata of eutherians and marsupials are different basioccipital. The crest is widely distributed with dental points of reference not strictly com- among living and extinct eutherians (Hershkovitz parable. . . . 1992a, p. 199). It is figured in therapsids by Rom- "The short Dromiciops symphysis menti [isj er (1966, pp. 182, 185, labeled ps), and in Dromi- not matched in any other living mammal" (Hersh- ciops by Hershkovitz (1992a, p. 200, PI. IX, la- kovitz, 1992a, pp. 201, 202, PI. X) but is com- beled s).

HERSHKOVITZ: LAST OF THE MICROBIOTHERIA 13 B

Fig. 5. Ventral aspect of mandibles showing short rounded symphysis (A) of Dromiciops (fmnh 127454), compared with long angular symphysis (B) of a didelphoid (Marmosa, fmnh 69321).

5. Rete testis (Figs. 9-11) a perineum that eliminates the cloaca. Exceptions or intergrades occur, however. The Dromiciops rete testis, according to Wool- It is interesting to note that a precaudal cloaca ley (1987), differs from that of all other marsu- persists in the eutherians Ochotona (pikas) and pials in the structure of the rete, greater number many Insectivora, particularly tenrecids. of tubules, and encasement in a mediastinum. The The evolutionary stages of the cloaca in Amer- character is also mentioned in Hershkovitz ican and Australian marsupials and in mono- (1992a, p. 203). tremes are outlined by Hershkovitz (1992a, p. 203). 6. Cloacas: basicaudal and precaudal (Fig. 12) 7. Sexual dimorphism The cloaca in both sexes of Dromiciops is basicaudal, a feature shared only with monotremes Sexual dimorphism with respect to overall size, among mammals, and with reptiles. The derived canine length, coloration, and perhaps other dis- cloaca is precaudal in all other marsupials except play characters is absent in Dromiciops. In all oth- the Didelphidae (as restricted by Hershkovitz, er marsupials males are consistently larger, the ca- 1997), Macropodidae, and Phalangeridae. In these nine tooth usually larger, and the coloration of the three families the mouth of the rectal and urogen- venter, particularly the mammary field of females, ital ducts of both sexes is completely separated by usually different in the sexes (Hershkovitz, 1997).

14 FIELDIANA: ZOOLOGY Fig. 6. The staggered or crowded lower third incisor of the Didelphimorphia. From top: Dasyurus (Dasyuridae, Australian); Marmosa (Marmosidae, American); Echymipera (Peramelidae, Australian); hoodon (Peramelidae, Aus- tralian); Sipalocyon (Borhyaenidae, Miocene American); Sarcophilus (Dasyuridae, Australian), dorsal and ventral aspects. Arrows point to staggered lower third (second in line) incisor.

Comment to character 1 above. In two previous ment, without acknowledgment of source, to wit, publications as well as here, Hershkovitz (1982; that "the staggered i3 is here regarded as a de- 1995) stated that the even spacing of the lower rived state which appears to be synapomorphic for incisors of the Microbiotheriomorphia is the ple- methatherians except Pucadelphys and Micro- siomorphic arrangement and that the staggered biotheriidae." Notwithstanding, and contrary to third lower incisor of all Didelphimorphia is the their conclusions, Marshall et al. (1995, p. 17) er- derived condition. In a third work (1997), Hersh- roneously referred Pucadelphys to the family Di- kovitz referred the metatherian Pucadelphys an- delphidae. Their explanation (Marshall et al., dinus Marshall and de Muizon to the Microbio- 1995, p. 84): "because its molar structure (which theriidae on the basis of its nonstaggered lower is currently the foundation of metatherian system- incisors. In a disturbing perversion of the facts, atics) is indistinguishable from that family. We Marshall et al. (1995, p. 68) declared that "Given thus give preference to molar structure in classi- the nearly universal occurrence of this staggered fying Pucadelphys within Metatheria." The mo- i3 in metatherians, Hershkovitz concluded that lars are indeed metatherian characters but the phy- this state i dental features [the staggered 3] was plesiomorphic[!J logenetically most significant of for this group [Metatheria]." Marshall et al. marsupials are in the antemolar fields, most no- (1995, p. 68) did at least arrive at my real assess- tably the lower incisive.

HERSHKOVITZ: LAST OF THE MICROBIOTHERIA 15 18 28 f

OVAL SHAPED AREA (MIDDLE COMPARTMENT) - POSTERIOR = entotympanic COMPARTMENT = tympanic process ofperiotic B

16 FIELDIANA: ZOOLOGY Fig. 8. Dromiciops gliroides. Two views of basicranium showing basisagittal crest (s): Bones are a, alisphenoid: b, basioccipital: e. "entotympanic"; p, petrous; mt. mastoid; s. basisagittal crest of presphenoid. Figure from Hersh- kovitz (1992a).

Fig. 7. A, Auditory bullar portion of Dromiciops skull; bone 24 is the so-called "entotympanic." which in microbiotheriids fills the gap between the alisphenoid tympanic process (17) and the petrous tympanic process (19). For names of other cranial elements, see key (p. 39). B, Crushed, disassociated auditory bulla of Microbiotherium sp. (Miocene. Patagonia) was probably indistinguishable in undamaged state from a Recent Dromiciops gliroides bulla. Figure copied from Segall (1969) with original labels capitalized; italicized labels added.

HERSHKOVITZ: LAST OF THE MICROBIOTHERIA 17 Fig. 9. Testis, efferent duct system, and epididymis of Dromiciops gliroides: (a) dorsal aspect and (b) ventral = = = = aspect. T testis; E epididymis; S 1 first segment of efferent duct system; S2 second segment of efferent = = duct system; RD extratesticular rete duct; VD vas deferens. Figures copied from Woolley (1987, p. 219); reproduced courtesy of Surrey Beatty & Sons Pty. Ltd.

Subsidiary Characters absence of incisor crowding in microbiotheriids. The Dromiciops small canine may be 1. Ectotympanic bone fully enclosed within an nearer primitive proportions, rather than sec- inflated auditory bulla. Character state is de- ondarily reduced. Upper and lower canines of rived (Fig. 7). In the didelphoids Caluromys Pucadelphys are large. and Caluromysiops (Caluromyidae) and in Gli- ronia the same bone is to (Glironiidae) nearly 4. Pouch with 4 nipples, 2 on each side, none but the bulla is not inflated. entirely enclosed, centered (Fig. 14). The same pattern occurs in "There is no evidence of an ossified ectotym- the caenolestids Caenolestes and Lestoros

. . . nor are there facets panic [in Pucadelphys] (Kirsch & Waller, 1979), but not in Rhyncho- in the ear which marked the site of at- region lestes, with 7 (3-1-3) (Patterson & Gallardo, tachment of this bone" et (Marshall al., 1995, 1987). Nipples in Australasian marsupials p. 64). range from 2 to 12, the number even in each species (Collins, 1973).

2. Ears are relatively small, tapered, hairy in cor- relation with the cool habitat (Fig. 13). Incidental Comments

1. Tarsal bones in marsupial phylogeny: A bio- 3. Small, nearly incisiform lower canines in logical fallacy (Fig. 15) Dromiciops contrast with the large canines of other marsupials except diprotodonts (Figs. 5, As cited from Hershkovitz (1992a, p. 181, ab- 6) and could be but need not be correlated with stract): The "arrangement of the Marsupialia [by

18 FIELDIANA: ZOOLOGY Fig. 10. Dromiciops gliroides: A, Left top and bottom, section through (a) main channels and (b) the two branches = = = = of the rete. (Level of sections shown in Fig. 1 1.) R rete; M mediastinum; TR tubuli recti. Scale line 0.1 mm. B, Right top and bottom, section through the rete and tubuli recti of (a) Didelphis albiventris and (b) Philander = = = opossum. RD rete duct; TR tubuli recti. Scale line (in right lower quadrant) 0.2 mm. Copied from Woolley (1987, pp. 223, 224). Reproduced courtesy of Surrey Beatty & Sons Pty. Ltd.

Szalay, 1982b] into cohort Ameridelphia encom- continents. Surprisingly, the confession is fol- passing all New World marsupials except Micro- lowed by a non sequitur (p. 143), which declares biotheriidae, and cohort Australidelphia contain- that "Hershkovitz' (1982, 1992[a], 1995) propos- ing all Australian [marsupials] and the American als seem unsupportable in the face of a large num- Microbiotheriidae, based primarily on the pattern ber of character conflicts." They fail to mention of articulation between the foot bones astragalus a conflict but conclude, despite their awareness of and calcaneus, has no leg to stand on." It is the inconsistencies (p. 142), that "ankle joint mor- shown (Hershkovitz, 1992a) that the joint patterns phology remains the best character supporting are variable and intergrading and that the "con- monophyly of the Australidelphia"; the statement tinuous" pattern said to be exclusively Australi- is iterated on page 155. delphian evolved independently more than once Tarsal bones are mammalian features, and nei- from the "separate" joint pattern said to be an ther of the two derived patterns is restricted to exclusive feature of Ameridelphia. The two pat- either ameridelphian or australidelphian. The geo- terns occur in both hemispheres. The morphology graphic terms are, nevertheless, useful in a ver- of astragalus and calcaneus of the metatherian nacular sense. Dromiciops, treated by Szalay as the australidel- varies between the two phian "morphotype," pat- 2. Ankle bones terns but is essentially ameridelphian or didelphoid, and little if at all different from the ankle Woodburne and Case object to my failure bones of some didelphoid mouse opossums (cf. (Hershkovitz, 1992a) to "address the other ankle Gracilinanus marica [Hershkovitz, 1992b, PI. joint morphologies utilized by Szalay (1982a, b)." HI]). To this may be added Szalay 's (1994) treatise, Woodburne and Case (1996, p. 142) acknowl- which is replete with descriptions of ankle bones, edge the variability of the morphology and dis- many of them fragmentary, none diagnostic. peral of both types of ankle bones on the two Careful study of the data reveals that ankle bone

HERSHKOVITZ: LAST OF THE MICROBIOTHERIA 19 1 mm

Marmosa impavida

1 mm

Caenolestes obscurus

Didelphis albiventris

Dromiciops gliroides

20 FIELDIANA: ZOOLOGY Fig. 12. A, Cloaca precaudal in Thylamys, bifurcated glans penis showing at opening, scrotum pendulous. B, = Cloaca basicaudal in Dromiciops, scrotum sessile. Bar 1 mm.

patterns with their consistencies and inconsisten- suggested. Although primitive in design no fea- cies as tabulated by Hershkovitz (1992a, p. 186) ture of the Dromiciops molars is peculiar to the provide more than adequate justification for rejec- genus. Molars of caenolestids and the didelphoid tion of Szalay's hypothesis of phylogenesis based Caluromysiops are also euthemorphic but more on them. molarized. Molar crown patterns of all other marsupials including Caluromys are dilambdomorph- ic with the eocrista 3. Molars (Fig. 16) W-shaped [or centrocrista] secondarily derived from the euthemorphic pat- The incomplete quotation from Woodburne and tern (cf. Hershkovitz, 1977, p. 279)." The buccal Case (1996, p. 143) says that Hershkovitz (1992a, " shelf is usually variably developed, the stylocone 208) holds that 'retention' in of p. Dromiciops (cusp B) often poorly developed or absent. a narrow buccal shelf in the while upper molars, The "current definition" of Marsupialia men- a would virtually 'lacking' stylar cusp B, place tioned by Woodburne and Case appears defective outside of all under cur- Dromiciops Marsupialia in light of the variability of width of stylar shelf rent definitions." and degree of development of stylar cusp B in Hershkovitz said Actually, (1992a, p. 208) microbiotheriids and marsupials generally (cf. molars retain the "Dromiciops early mammalian Hershkovitz, 1972, Fig. 19). My description of high cusped tritubercular euthemorphic crown Dromiciops dentition agrees with that of Reig pattern with buccal shelf narrow, stylocone (cusp (1955), who was the first to recognize its micro- B or j of authors) diminutive or hardly more than biotheriid characters.

Fig. 1 1. Diagrams of testicular rete of Marmosa, Caenolestes, Dromiciops, and Didelphis; redrawn from Woolley (1987, p. 220); reproduced courtesy of Surrey Beatty & Sons Pty. Ltd.

3 HERSHKOVITZ: LAST OF THE MICROBIOTHERIA 21 Dromiciops

Mar"mo s a

Fig. 13. Hirsute ear of Dromiciops gliroides contrasted with comparatively naked ear of a mouse opossun (Marmosa).

22 FIELDIANA: ZOOLOGY ifl HERSHKOVITZ: LAST OF THE MICROBIOTHERIA 23

Fig. 16. Upper dental system of Dromiciops gliroides, from Marshall (1982, fig. 16); a, labial; b, occlusal; c, = 1 lingual. Bar 6 mm. Stylar cusps of m labeled.

6. Caecum, declared loss (Fig. 18) Epitome

and Archer state that Aplin (1987, p. xxxvii) The five most important events in the history with all of the Das- Dromiciops shares members of marsupial classification are the following. yuromorphia and Notoryctemorphia, and with (alone within the Order Tarsipes Diprotodontia), 1. Determination of the South American the derived loss of an intestinal caecum. Hume Dromiciops as a genus of microbiotheriid (1982) is cited as source. (Reig, 1955). to Hume Australian mem- According (1982), 2. Systematic separation of microbiotheriids bers of the families noted and Archer by Aplin with 4 evenly spaced lower incisors from all lack a but nowhere in his text is Drom- caecum, other marsupials or didelphimorphs, all with iciops mentioned. This taxon, like all other Amer- derived staggered lower third incisors ican marsupials described or figured by Hume (Hershkovitz, 1982, 1995). (Marmosa [sensu lato], Philander, Chironectes, 3. Discovery of Adinodon (Didelphoidea), old- Didelphis, Caenolestes), has an intestinal caecum. est known marsupial with staggered i„ re- That of Dromiciops (fmnh 129815) is present and covered from the Albian (Lower Creta-

Texas 1 95 1 Hershkov- figured. ceous) of (Patterson, ; itz, 1982, 1995). The possibility that Holoclemensia 7. Seeming similarities Slaughter (1968b) is the same as or nearly related to Adinodon has been suggested by Some seeming dental and cranial similarities Turnbull (1995), but its lower incisors are between microbiotheres and Recent Glironia, unknown. Caluromys, and Caluromysiops prompted Reig 4. Persistence of the derived character of a to those to the Microbiothe- (1955) assign genera staggered i, suggests a Late Jurassic origin riinae, perceived by him as a subfamily of Didel- of didelphimorphs and an earlier origin of phidae. the phylogenetically antecedent microbio-

HERSHKOVITZ: LAST OF THE MICROBIOTHERIA 25 lmm

buttress

1/2 mm ADINODON PALTERSONI

Fig. 17. Adinodon pattersoni (holotype); fragment of left ramus cleared with oil of anise for revealing alveolar outlines. buccal surface of i Patterson same Top, fragment 2 _ 5 , c, p, ,, m b redrawn from (1956); bottom, specimen

full i . with i2_5, c only, seen from labial surface tilted slightly inward to reveal depth of staggered alveolus of 3 Copied from Hershkovitz (1982, 1995).

theriids now known from the Early Paleo- CHILE: Valdivia (near Union, type locality cene of Bolivia. 40°17'S, 73°05'W); holotype a mounted 5. Marsupialia (Metatheria) evolved indepen- skin, skull inside, Museo Nacional, Santi- dently of Eutheria, and each arose from a ago de Chile; name preoccupied by Didel- different stem of the Therapsida. phys australis Goldfuss (1809:219) (Pen- nants New Holland opossum, a phalanger- id). F. 1893b: 33, (animal)— Dromiciops gliroides Thomas Philippi, pi. CHILE: Valdivia (near Union). R. A.

Synonymy Philippi, 1894:33, pi. 4, fig. 2 (animal)— CHILE: Valdivia; Llanquihue; Araucania. Didelphys elegans Cunningham (not Thylamys Wolffsohn and Porter, 1908:69—CHILE: elegans Waterhouse, a marmosid), 1871: Curacautin; local names: monito del mon- 362. te, kumiuma, wenukiki, kongoi-kongoi; gli- Didelphys australis F. Philippi, 1893a:318— roides Thomas, a synonym.

26 FIELDIANA: ZOOLOGY Fig. 18. Caecum of Dromiciops gliroides (fmnh 129815). Alleged to be absent in microbiotheriids by Aplin and Archer (1987).

Dromiciops australis Thomas, 1919:212—AR- Meserve et al., 1988:721—CHILE: Osorno GENTINA: Neuquen (Beatriz, Nahuel (La Picada); diet. Patterson et al., 1989: Huapi, 800 m); CHILE: Cautin (Temuco). 67—CHILE: Osorno (La Picada); distri-

Cabrera, 1919:30, pi. 1, fig. 1 (head). bution, abundance, capture (snap trap pos-

Krieg, 1925:173, fig. 2 (head, tail), fig. 3 itive, live trap negative). Patterson et al., (pouch), characters, comparisons. (Reig, 1990:620—CHILE: Osorno (La Picada);

1955, p. 127)—Dromiciops a microbiothe- distribution, habitat.

riid. Mann, 1955:159, figs. 1, 2, 4, 7 (tail, Dromiciops australis australis, Osgood, 1943: hand, foot, marsupium), figs. 3, 5 (skull), — 48, fig. 2 (skull) characters; distribution; fig. 6 (liver), fig. 8 (female reproductive CHILE: Llanquihue (Cayetue, Todos tract)—CHILE: Valdivia (between Valdi- Lago Santos; Peulla, Todos Santos); Con- via and Puerto Montt); characters, habits, Lago ception (Lota, southwest of Conception); thermoregulation. Mann, 1958:209, figs. 1, Malleco (Rio Colorado; Victoria; Sierra 2(9 genitalia), fig. 3 (marsupium), fig. 4 Nahuelbuta). Santos, 1946:191, pi. 1 (rhi- (teats), fig. 5 (newborn), fig. 6 (juvenile), 2 — — narium, foot, hand, tail), pi. (animals) fig. 7 (nest) reproductive organs, habits ARGENTINA: (Huemul, Nahuel of newborn. Lyne, 1959:84—distribution Neuquen Huapi; Villa Angostura, Nahuel Huapf); of vibrissae. Segall, 1969:489, figs. 1-4 characters; habits; local names, llaca, mon- (skull; middle ear, auditory ossicles)—mid- ito del monte. Greer, 1965:103—CHILE: dle ear region, comparisons, classification Malleco (Cordillera de Nahuelbuta; Cor- (Microbiotheriinae). Reig et al., 1972:39— = dillera de los Andes); habits, measure- karyotype, 2n 14. Reig et al., 1977:199, = ments. and Frassinetti, 1980: 202—karyotype, 2n 14. Marshall, 1978: Tamayo 327—CHILE: Nuble (Rio Itata). Pearson, 1, fig. 1 (skull), figs. 2, 3 (dentition), fig. 4 1983:483—ARGENTINA: characters (sen- (map distribution); characters, habits. Mar- sorial, nocturnal, insectivore, omnivore). shall, 1982:35, fig. 12 (map distribution), 13 (skull), 14-19 fig. figs. (dentition), fig. D[romiciops] australis australis, Oliver Schnei- 21 (size relationships); characters, compar- der, 1946:68—CHILE: Conception (Parque isons. Pearson and Pearson, 1982:129— Pedro del Rio; Huepil; N Rio Laja). ARGENTINA: Rio Negro (Parque Nacion- al Nahuel Huapf); habitat, rodent censuses Dromiciops gliroides Thomas, 1894:187—type in three Nothofagus-Chusquea forests. description. Thomas, 1919:212—CHILE:

HERSHKOVITZ: LAST OF THE MICROBIOTHERIA 27 •j Cautin (Temuco). Krumbiegel, 1941:11 — Abrothrix longipilis, Akodon olivaceus, Chele- CHILE: Valdivia. Hershkovitz, 1992a: 181, mys macronyx, Geoxus valdivianus, Loxodon-

pis. Ill, IV (tarsal bones), pi. VII (ani- tomys micropus, Irenomys tarsalis, and Octo- mal)—ankle bones, phylogeny, classifica- don bridgesi. Chelemys, Geoxus, or the Octodon tion. could have excavated the tunnel described below by Greer. australis 1943: Dromiciops gliroides, Osgood, Pearson & Pearson (1982) censused small 50—CHILE: Chiloe (Quellon). Cabrera, mammals in three study sites in the Parque Na- 1958:5—classification. and Fras- Tamayo cional Nahuel Huapi, Rio Negro Province, in sinetti, 1980:377—CHILE: Isla Chiloe (0- the southern rain forests of Argentina. All sites 300 m). were dominated by the southern beech Notho- fagus and bamboo Chusquea. The Puerto Blest [Didelphys (Peramys)] australis, Trouessart, site of 1.01 ha at 770 m and 1898:1244—listed. study elevation, annual rainfall of about 3,000 mm yielded four Holotype—Male, skin and skull, British sigmodontine genera {Irenomys tarsalis, Geox- Museum (Natural History) M.92.9.5.3; collect- us valdivianus, Abrothrix longipilis, Oligory- ed 20 May 1868 by Robert O. Cunningham on zomys longicaudatus), Rattus sp., and the mar- the voyage of H.M.S. Nassau. supial Dromiciops. Excluding Rattus, the single Type Locality—Huite, near Ancud Dromiciops and Geoxus were least trapped. Ir- (41°52'S, 73°50'W), northeastern Chiloe Island, enomys (5.1 per ha) and Akodon (3.9 per ha) Chile. were most abundant. Trapping was during May Distribution (Fig. 19)—The rain forests of and November 1978 and April 1980. central Chile and bordering parts of Argentina The second site (Rio Castano) was about 20

from about 36°S to 43°S on Chiloe but to 41°S km S of the first, at 950 m elevation, 2,000 mm on the mainland, or from the latitude of Con- annual precipitation, with freezing temperatures ception to the southern border of Isla Chiloe, registered during every month of the year. east to the lake district of Nahuel Huapi in Ne- Poorest catches for the site were Dromiciops uquen, and Rio Negro, Argentina. The eleva- (0.5 per ha) in November 1978 and Oligory- tional range extends from sea level in Isla Chi- zomys longicaudatus (0.3 per ha) in May 1978. loe to 1450 m in Malleco Province. Highest yield was Chelemys macronyx (14.7 Habitat and Associates—Dromiciops lives per ha) in May 1978. Other captures were in the humid Nothofagus and Aracauria forests Geoxus valdivianus, Akodon olivaceus, Abro- intermixed with bamboo (Chusquea) thickets. thrix longipilis, and Loxodontomys micropus. "Dromiciops was taken mostly in short-statu- The third site was at Nire, about 43 km SSW red forests with low density of shrubs and high of Bariloche, at about 1030 m. Dromiciops was herbaceous groundcover, and these predominate not found and may not occur there. A species at higher elevations" (Patterson et al., 1990, p. each of Geoxus, Chelemys, Akodon, Abrothrix, 630). In a test of elevational preferences of 10 and Loxodontomys were found, the same cap- species of small mammals (8 sigmodontines, 2 tured at the second site. marsupials), Patterson et al. (1989) found that An unreported collection from Isla Chiloe between elevations of 425 and 1135 m, Drom- and Osorno Province made by Patterson anc iciops was captured most frequently at 715 m. Gallardo (1984, field catalog at fmnh) includec The animal was taken in flat, open snap traps the same Dromiciops associates noted, in ad- only. The closed metal Sherman live traps were dition to Abrothrix sanborni and the caenolestic ignored. Rhyncholestes raphanurus. The mean annual temperature of the Dromi- Reported from La Picada, Osorno, by Mes ciops habitat is 10°-11°C, the annual rainfall erve et al. (1988) were Dromiciops gliroides about 3,000-4,000 mm (Mann, 1978; Pearson Rhyncholestes raphanurus, and Abrothrix san & Pearson, 1982). Dromiciops was absent in borni, in addition to those species already men the dry matorral country in the vicinity of San- tioned. tiago. The poor showing of Dromiciops in the Ar Sigmodontines captured together with Dro- gentine localities may not reflect the true pop miciops in Greer's (1965) Malleco Province ulation density throughout the year and in othej traplines were Oligoryzomys longicaudatus, parts of its range. Dromiciops is morphologi

28 FIELDIANA: ZOOLOGY T72 36

38 PACIFIC OCEAN

C H I

25 SO 100 ISO KILOMETERS

42 ARGENTINA ISLA CHILOE

76 74 72 70

Fig. 19. Distribution of Dromiciops gliroides in Chile and bordering Argentina. See gazetteer (pp. 56-57) for geographic and political data.

HERSHKOVITZ: LAST OF THE MICROBIOTHERIA 29 B B

Dromiciops gliroides Metachirus nudicaudatus

Fig. 20. Tails (sections enlarged, not to scale). Left: Prehensile tail of Dromiciops gliroides. A, Dorsal surface at base. B, Terminal portion of same. C, Ventral surface of terminal portion enlarged showing dermal plica permitting folding. Right: Nonprehensile tail of Metachirus nudicaudatus. A, Base of dorsal surface, brush short. B, Terminal portion of tail. C, Ventral surface of terminal portion, dermal plica absent.

cally, physiologically, ecologically, and behav- legs like underparts, outer sides brown like sides iorally different from the rodents where both of trunk; face (Fig. 13) buffy to ochraceous, muz- were trapped. zle darker, eye ring blackish, the color often extending as a dark, contrasting band across cheek to tip of nose; pale patch present above each eye; Characters fine mystacial vibrissae extending to ears when — laid back (Fig. 13); ears comparatively short, base Outer Parts and Coloration (Frontispiece) rounded tapering to tip, thinly hirsute on inner and Tail prehensile, the dermal folds well marked outer surfaces, bushy at base; hands and feet (Fig. (Fig. 20); pollex and hallux opposable (Fig. 21); 21) above whitish, the digits fringed; digital tufts pelage dense, color pattern agouti; dorsum domi- of feet usually thick, often hiding claws; tufts of nantly chocolate brown, sides of trunk with a con- manual digits short, thin, hardly visible, claws not trasting pale patch on each shoulder, postscapula, protruding beyond pads; pedal claws slightly or hip and rump; underparts from chin to cloaca pale not protruding; tail (Fig. 20) slightly longer than ochraceous to buffy or nearly white, the gray hair head and body combined, prehensile, thickly pi- bases showing through; inner sides of arms and lose, proximal 4th or 5 th incrassate, ventral sur-

30 FIELDIANA: ZOOLOGY s: S. Table 1. Dromiciops gliroides: means and extremes (mm) of external, cranial, and dental measurements of 25- 40 individuals.

Sex Character Mean (extremes)

66 Total length 208 (192-222) 9? Total length 212 (202-229)2 12 66 99 Compared with the fmnh material (Table 1 ), the Santos series has a larger body and a longer tail, but significantly smaller hind feet and shorter ears. There is no accounting for the discrepancies even if the smaller foot measurement given was without claw, and the shorter ear measured from crown, not notch. Local or geographic size differences are expected. Greer (1965, p. 104) recorded the following measurements (mm) from two localities in Malleco Province, sexual size differences not ev- ident: the seminiferous tubules in sexually mature didelphoids and caenolestids examined ranged from 0.17 to 0.27 mm. The tubule of Dromiciops, however, was too young to measure. Woolley (1987) 3 "2 z 3 could confirm that in the didelphoids and caenolestids observed by Rodger (1982), the rete is a simple duct system lying within the testicular tis- 28~ sue in the anterior pole of the testis. A single duct — f", o o — o o o was found in all of those mentioned except one 4> SS J E Didelphis albiventris (Fig. 11), in which the main duct bifurcated. No distinct mediastinum was c found in examined any species by Rodger (1982). 2 E ' 5 vd (N © — —'< © O "In the microbiotheriid, the rete o»"5 E- Dromiciops, v x OS J Q S [Fig. 10] ... consisted of a network of intercon- nected channels lying among the seminiferous tubules in the anterior pole of the testis. It was en- cased in a thin but distinct mediastinum formed as an of the fibrous tissue of the tunica inturning * After the testicular tissue E 3U333330C — albuginea. penetrating ** o-g-a-o-a-o-o.^ o /") in Tt ©

The morphology of the testis, as adduced by S. sj .b! a t-S-l offers evidence of a a* -« "a -S g Woolley, convincing signifi- CO dj cant distinction between Dromiciops and didel- 3-2 I g •S 15 £ J2 phoids and caenolestoids on the one hand, and an Q

34 FIELDIANA: ZOOLOGY uration, is restricted to the American fauna even greater separation from Australian marsupi- marsupial and has been used to separate them from their Australian als. counterparts. The evidence provided suggests sperm Spermatozoa (Fig. 22)—Four morphological pairing will occur in all those American species not yet types of spermatozoa have been distinguished in examined, such as Lestodelphys, Lutreolina, Caluromy- American marsupials, three of them by Biggers et ciops [sic], Glironia, Lestoros and Rhyncholestes, with the exception perhaps of Dromiciops, which has al. (1963) and the fourth by Temple-Smith (1987) sperm which more closely resemble those of Australian mar- (Fig. 22). Descriptions of the first three are sum- supials. marized from Biggers (1966, p. 269), that of Drom- iciops from Temple-Smith (1987, pp. 180-182). Temple-Smith (1987, p. 180), added that Didelphis Type—Head flattened, hook-shaped, with two directed limbs, one backwardly recent observations on sperm structure, and in particular ac- thick, the other thinner, longer, tapered; sperm head morphology have cast more light on the evo- affinities of rosome on anterior part of thick limb; mid- lutionary this species [Dromiciops gliroides]. . . . Although these ultrastructural observations . . . piece attached by fine filament to base of have been made on inappropriately fixed, field material, cleft separating limbs of head: Didelphidae sufficient detail is present in spermatozoa from the cauda Mar- {Didelphis, Philander, Chironectes); epididymidis to indicate a remarkable similarity in

mosidae {Marmosa); Monodelphidae sperm head structure between Dromiciops . . . and many Australian ... In with material {Monodelphis); Metachiridae {Metachi- species. fact, appropri- ately prepared for electron microscopy, it would proba- rus); spermatozoa paired. bly be difficult to readily distinguish the sperm head of —Head acro- Caluromys Type saucer-shaped, Dromiciops from such species as Trichosurus vulpecula. some on concave side; midpiece inserted Cercartetus nanus and some other related possum spe-

in center of convex side of head: Caluro- cies. . . . The structural features which characterize these species ... are all features shared by Dromiciops sper- myidae {Caluromys); spermatozoa paired. matozoa . . . but not by those of any other extant Amer- Caenolestes Type—Head cleft on one side into ican species. which midpiece is inserted; acrosome not studied: Caenolestidae {Caenolestes); sper- In Didelphis, according to Rodger (1982, p. matozoa paired. 269) "paired ejaculated spermatozoa are trans- Dromiciops Type—Head cuneiform, acrosome ported to the oviducal isthmus where they sepa- restricted to anterior third of dorsal surface rate, and fertilization is monospermic." of nucleus, midpiece not preserved; sper- Didelphoid sperm pairing is clearly a derived matozoan pairing not determined; the sin- condition, unique among mammals but unproven gle head said to be similar to that of Aus- in Dromiciops. The condition evolved after sep- tralian Trichosurus vulpecula (Phalangeri- aration of Australian marsupials from their Amer- dae). ican progenitor(s), which, at the time, still pos- Pairing—The swimming of spermatozoa in sessed the simple ancestral type of epididymal closely coupled pairs as a regular phenomenon sperm, at least in the Antarctic Sounders. Any has been observed in the epididymides and in similarity between sperm morphology of living mated females of all the genera mentioned above Dromiciops and living Australian marsupials may except Dromiciops. The pairing involves juxta- be attributed to retention of the relevant primitive position at the acrosome level of the heads of two character state in both taxa. spermatozoa. According to Biggers (1966), the Marsupia and Mammary Field (Fig. 14)— principal source of this information, spermatozoan Marsupials were so named under the mistaken be- pairing has not been observed in Australian mar- lief that all were characterized by an abdominal supials. This may well be one of the most signif- or abdomino-inguinal pocket containing the mam- icant differences separating all living Australian mary glands suckled by sheltered young. Austra- marsupials from American didelphoid opossums lian marsupials, with few exceptions, have pouch- and paucituberculates. es. The majority of New World marsupials lack Relationships between American and Austra- them. An abdomino-inguinal cutaneous pocket lian marsupials based on sperm morphology, mat- housing milk glands is possessed by females of uration, and pair formation was discussed by the marsupial families Didelphidae (Didelphinae Temple-Smith (1987). Biggers suggested that [Philander, 2 species; Didelphis, 3 species]; Chi- ronectinae [Chironectes, 1 species]; Lutreolininae sperm pairing, also regarded as a feature of sperm mat- [Lutreolina, 1 species]), Caluromyidae {Caluro-

HERSHKOVITZ: LAST OF THE MICROBIOTHERIA 35 nucleus plasma membrane

head- nuclear membrane

implantation fossa acrosome

connecting piece

midpiece

Didelphis type

J/ww

Caluromys type

Caenolestes type

Dromiciops type

36 FIELDIANA: ZOOLOG mys, 3 species; Caluromysiops, 1 species), Gli- pouch for incubation of usually 1, occasionally 2 roniidae (Glironia, 1 species, not a caluromyid, eggs. The platypus (Ornithorhynchus) has no has no pouch), and Microbiotheriidae (Dromi- pouch. An abdominal pouch is no more plesio- ciops, 1 species with pouch). Marsupium and morphic in Prototheria than in Metatheria. scrotum are not homologous. Six types of mammary fields defined by the The pouch or pocket is composed of a pair of nipple pattern, all even-numbered, are described recessed longitudinal skin folds, one on each side and illustrated by Tyndale-Biscoe and Renfree of the midabdomino-inguinal region. In didel- (1986, p. 38). Number 1 is ovate with limits of phoids they are united caudally but separate prox- the "pouch area," or mammary field, outlined di- imally. In Didelphis, the lip of each lateral skin agrammatically. The field "has no covering fold fold is known to appose the other to form a com- of skin" and the (10) nipples are fully exposed. pletely but flexible shuttered waterproof pocket This type is said to be representative of "Didel- for shielding the suckling young (cf. Enders, phidae, Caenolestidae, Dasyuridae." No Ameri- 1937, p. 25; 1966, p. 199). The folds of Chiro- can marsupial mammary field conforms to it or to nectes are similar. any of the other types outlined. The types (1-4) The pouch of Dromiciops gliroides consists of shown by Tyndale-Biscoe and Renfree (1986) an ovate midabdominal orange or reddish mam- were copied from Woolley (1974, p. 12), who had mary field roughly 10 X 15 to 13 X 20 mm in constructed them from dasyurid models. Ameri- mature females. The field is partially shielded by can types were not mentioned. There is, however, a pair of lateral cutaneous folds, one on each side, a superficial resemblance of type 1 to the 4-nipple joined caudally in a recessed pocket but separated mammary field of Dromiciops (Fig. 14). laterally, leaving exposed the pair of abdominal Reproductive System (Fig. 23)—The female nipples on each side. The lateral cutaneous folds reproductive system of Dromiciops is the normal are recessed a millimeter or two on each side, marsupial type characterized by medial ureters, about 1 mm deeper caudally. The central mam- double vaginas, and a middle or pseudovagina for mary field is swollen when filled with milk. delivery of young. The system in Dromiciops as The maximum 4 young of Dromiciops may not seen and illustrated by Mann (1958, p. 209; 1978, need the protection provided by apposing lateral p. 28) is reproduced here (Fig. 23). The male re- folds such as those of Didelphis. The Chusquea productive system of the species has not been nest built for them by the pregnant mother is well seen by me. Available (emnh) entire specimens hidden and waterproof when covered with moss preserved in alcohol were either too young or had or bryophytes. The survival rate of young may be been deprived of their genitalia by an unknown high and compare favorably with that of didel- pursuer of knowledge. phoids with much larger litters. Remarks—According to Mann (1958, p. 209), Didelphoids other than those mentioned above the scrotal color changes from whitish in young are pouchless. The mammae are arranged bilat- to bright dark red in adults. In 22 males at hand, erally, in some extending from inguinal to pec- of various ages and from several localities, neither toral regions, in others not beyond the abdomen. scrotum nor other abdominal part or area is red- In most species there may be 1, rarely a few, ad- dish. In 16 females the mammary field is orange ditional nipples in the medial abdominal region. or yellowish to reddish in 7 full adults, including The number of nipples in didelphoids ranges from all parous individuals. 5 to 7 in Marmosops impavidus to 27 in Mono- R. A. Philippi (1894, p. 35) had been informed delphis henseli, the total in each species being odd by the person who brought him the holotype of D. (Hershkovitz, 1992b, Table 1, pp. 14-15). australis that up to 5 young had been seen in a nest. The marsupium must have arisen independently This has never been verified. A 5th young could not in each group. Among Monotremata, the spiny survive for lack of a nipple, but it is not uncommon anteaters {Tachyglossus) develop a temporary for newborns of a litter to outnumber the nipples.

Fig. 22. Diagrams of 4 types of American marsupial spermatozoa: From top to bottom: Didelphis (sperm pair. and unpaired at right), Caluromys (sperm pair, and unpaired), Caenolestes (sperm pair, and unpaired), and Dromiciops (unpaired) side and front. From Temple-Smith (1987. p. 177); labels added. Reproduced courtesy of Surrey Beatty & Sons Pty. Ltd.

tr HERSHKOVITZ: LAST OF THE MICROBIOTHERIA 37 fallopian tube

kidney

epididymis

urogenital sinus or duct

glare penis prong

urethral opening

lateral vagina

pseudovagina

urogenital sinus or duct

Fig. 23. Metatherian reproductive systems. A, Female (9). Each independent uterus connects to the vagina o the same side; each vagina empties into the urogenital sinus or duct through the pseudovagina. which serves as th birth canal for both vaginas. The pseudovagina usually disappears after parturition and re-forms for the next birth B, Male (6). The penis is everted independently; the paired testes are external and contained in a prepucial pendulou scrotum (not shown). Diagrams based on the reproductive system of Philander opossum. C, System of 9 Dromiciop

38 FIELDIANA: ZOOLOGY Breeding—The breeding season occurs during 12. the southern spring. Suckling young have been recorded in November (Osgood, 1943), December (Pearson, 1983; Greer, 1965), and January (Mann, 1978). Hunsaker (1977, p. 143) states that Drom- iciops "breeds October through May, pouch young in November and December" but fails to give the source of data. None of the many adult females collected by Patterson, Meserve, and Lang at La Picada during February and March 1984 harbored suckling young, although some were still lactating (Patterson, pers. comm.).

The male, according to Mann (1978, p. 35), stays with the female at least during the breeding season. This has not been verified, nor is there evidence that the marsupial male helps care for the young. Caecum (Fig. 18)—Aplin and Archer (1987, p. xxxvii) affirm that "Dromiciops australis shares with all of the members of two Australian orders (Dasyuromorphia and Notoryctemorphia) and with Tarsipes (alone within the Order Diprotodontia) the derived loss of an intestinal caecum (Hume 1982)." According to Hume (1982), Australian members of the families noted by Aplin and Archer lack a caecum, but nowhere in his text is Drom- iciops mentioned. This genus, like all other Amer- ican marsupials described or figured by Hume

{Marmosa [$./.], Philander, Chironectes, Didel- phis, Caenolestes), has an intestinal caecum. That of Dromiciops is similar (Fig. 18).

Explanation of symbols 1-28 and a-y in Fig- ures 7, 23, 25, and 26

Bones

1. nasal

2. frontal

3. parietal 4. interparietal 5. premaxillary 6. maxillary 7. lacrymal 8. jugal (zygomatic) 9. squamosal (temporal) 10. sphenoid (includes orbitosphenoid, basisphe- noid) 11. palatine Fig. 24. Skull of Dromiciops gliroides: dorsal, ventral, and left lateral aspects; mandible below.

Skull (Figs. 8, 24-26)—Skull broad, uncrest- border of premaxillary bones; inferior postorbitai ed; braincase vaulted, temporal ridges absent; process sharply pointed; premaxillary symphysis fronto-parietal suture more or less transverse; in- comparatively shallow, more nearly rounded thar terparietal bone large; nasals flared at nasomaxil- angular; incisive foramina short, not extending lary suture; nasal tips not projected beyond distal behind plane of canines; palate comparative!)

Fig. 25. Diagrams of Dromiciops gliroides skull: 3 sides of cranium and 3 of mandible showing all elements See page 39 for key to symbols.

40 FIELDIANA: ZOOLOGY HERSHKOVITZ: LAST OF THE MICROBIOTHERIA 41 Table 3. Skeletal dimensions (mm) and proportions of Dromiciops, compared with three caenolestoids and one didelphoid.

Name Table 3. Extended. B Ja

o i- ^ u

•— '-1 ~ —

'on 5

= S ~ 3 _g Son S

i_ cd -a c o _•- a. 2 r- M

u O rf-O fill •5 £ 1 >. u S fl ^ is 5 °<

~ O iu u

. PS a. oo nO ... p >>

d 2 c o.

44 FIELDIANA: ZOOLOGY in 60

Hxg ~

zz'i 1

E= Fig. 28. Thoracolumbar flexure in Dromiciops gliroides (fmnh 127460) and a large Marmosa (fmnh 124611). The spinal curvature is normal in marsupials, rodents, and other "hunching" or "curling up" mammals, particularly when torpid.

1 3 Incisors simple, spatulate, diastema between the of p , fully erupted p about one-third again larger 2 first to either size first and 2 upper incisors about equal width of (taller) than p ; of lower premolar p 2 tooth, first upper incisor about one-third again about equal, shorter than canine, cingula well de- 2 than i the incisors well than higher , remaining sequentially fined, distostyles developed; p, larger p 2 3 1 smaller, i about half the size of i . All lower in- with proportionately longer distostyle. cisors about same size, none crowded but crowns Third upper deciduous premolar about half bulk touching or slightly overlapping, none staggered. or less of displacing premolar, its paracone, me- Canines stout, little curved, upper conical seen tacone, and protocone well defined, cingula and from inner side, nearly twice as high as first in- terminal styles present; third lower deciduous pre- cisor and first premolar, diastema about as wide molar greatly reduced but trigonid and talonid 5 as height of i or slightly wider; lower canine low, well defined, the latter shorter than the former but

incisiform, touching i 5 . slightly wider. First upper premolar low with buccal cingulum, First molar, with all primary cusps, buccal shelf mesostyle and distostyle (heel) present, the worn with mesostyle A, ectostyle B (= stylocone), ec- 2 2 tooth followed by diastema; p about twice bulk tostyle C, ectostyle D and terminal style E; m like

Fig. 29. Upper dental arcade of (A) Dromiciops gliroides (Microbiotheriidae), (B) Philander opossum (Didel- 1-2 phidae), (C) Caenolestes fuliginosus (Caenolestidae), and (D) Caenolestes fuliginosus (occlusal aspect); m quad- 3-4 ritubercular, with hypocone (5); m tritubercular. Mandibular dental arcade of (E) Dromiciops gliroides (Microbiothe- riidae), (F) Philander opossum (Didelphidae), and (G) Caenolestes fuliginosus (Caenolestidae). For names of dental elements see page 48.

46 FIELDIANA: ZOOLOGY A DROMICIOPS

12 3 4 5 B PHILANDER

12 3 4 5

C CAENOLESTES

CAENOLESTES

E DROMICIOPS

PHILANDER C

2 3 4 5

CAENOLESTES

\ HERSHKOVITZ: LAST OF THE MICROBIOTHERIA 47

I Table 4. Explanation of symbols used for elements of marsupial teeth adapted from quadritubercular molars. For names and locations of all dental elements see Hershkovitz (1977, pp. 299-301); the same symbols are used for homologues in upper and lower elements of tritubercular and quadritubercular molars.

Upper teeth Lower teeth

Cones 1 Conids 1

eocone (paracone) 1 eoconid (protoconid) protocone 2 metaconid 3 paraconid metacone 4 hypoconid hypocone present in 1st and 2nd upper molars of Caenolestidae

2 2 Conules and styles Conulids and stylids

A mesiostyle-a (parastyle) A mesiostylid-a (parastylid) E distostyle-fr (metastyle; hypoconule) E distostylid-b (hypoconulid) d metaconule (hypertrophied to cusp proportions) rare B ectostyle-y (stylocone) ectostyle-A: C ectostyle-l (mesostyle; stylar cusp C of authors) D ectostyle-m Cristae 3 Cristids 3

(Cristae extend from/to or between (Cristids extend from/to or between cusps shown in brackets; reference cuspids shown in brackets; reference cusps in parentheses are not cuspids in parentheses are not necessarily elements of the necessarily elements of the indicated crista; primes (') have indicated cristid; primes (') have mostly been deleted in this study) been mostly deleted in this study) / eocrista [(A)-(B)-C-(E)] / eocristid [{A)-3-l-4-(E)]

I' paracrista [/-(A)] or alternate routes or portions Vx [l-(B)-(l)]

I" postmetacrista [1 -(b) in absence of 4 or 4-(E)] I" postmetacristid [4-(E)]

T" centrocrista [1-4] T" centrocristid [1-4] II epicrista [/-(2)] may be incomplete II epicristid [l-(2)] may be incomplete

/// protoloph (protocrista) [2 crested /// protolophid (protocristid) [2- portion of cingulum B] (3) or 2 crested portion of cingulid B]

IV plagiocrista (metaloph) [2-d-(4)] IV plagiocristid [incomplete or absent] Basins or fossae Basins or fossids

pretrigon basin or fossa "trigonid" basin or fossids trigon basin or fossa anterior talonid basin or fossid talon basin or fossa posterior talonid basin or fossid

post-talon basin or fossa post-talonid basin or fossid pretalon basin or fossa intercingular basin A Upper and lower teeth Cingula and cingulids Main enamel folds Primary ex ectoflexus (between / and 4) A buccal or external (buccal shelf)

B anterolingual or anterior (primary lingual shelf)

C posterolingual cingulum

1 Most cones (conids) are numbered in the order of their origin and development. 2 Most conules (conulids) and styles (stylids) are listed in the order of their position from buccal to lingual and anterior to posterior. Rare or infrequent elements of the Dromiciops trigon such as conule d are included. Supernu- meraries or gemini of established cusps (cuspids) are not identified individually. 3 Most cristae (cristids) are numbered in the order of their appearance or development in phylogeny, others are numbered opportunistically; all cristae (cristids), except /-/"' inclusive, are modified parts of cingula (cingulids): homologies of the talonid cristids are not certain in every case. 48 FIELDIANA: ZOOLOGY 3 2 4 1 like re- ual or modifications m or slightly larger; m m ; m greatly specific by members of most duced, simplified, its metacone reduced or absent, marsupial families (Reig et al., 1972). stylar shelf and cusps reduced or absent; m, 3 or Sharman (1973, p. 492, Fig. 4) declared that talonid than alone has 4 times larger than dp 3 , larger trigon- "Dromiciops twenty-two autosomal than about all but its id; m 2 one-third or more larger m,; m, arms one pair of autosomes being meta- talonid centric or same size; m4 about two-thirds bulk m 3 , submetacentric." The authority and greatly reduced but evidently functional, in source of figure given are Reig et al. (1972), 4 marked contrast to nearly obsolete talon of m . whose depicted karyotype of female Dromiciops, Diphyodonty (Figs. 30-32)—Functional di- as well as Sharman's copy, shows the expected 24 phyodonty in marsupials is limited to the 3rd pre- autosomal arms. Later, Sharman (1982) gave a re- molar field. The old, worn, fully molarized but constructed account of the female Dromiciops minute dp3, fitted to the small jaw of juvenals, is karyotype, relating it to Australian types with 14 loosened by its growing alveolus and displaced by chromosomes. In this the investigator evidently the much larger erupting 3rd premolar. Shown conformed to the dubious Szalay (1982a,b) con- (Figs. 30-32) is the deciduous tritubercular dp3 cept of the American and Australian marsupial di- of Dromiciops in place or being displaced by the chotomy based on unrealistic tarsal bone patterns. single-cusped third premolar. In eutherians the Sex chromosome mosaicism in Dromiciops so- first functional teeth are the incisors (t) and pre- matic tissue from Valdivia and Concepcion, Chile, all set. molars (ij), replaced by the second Molars was reported by Gallardo and Patterson (1987, p.

(I) are deciduous, erupt last, and are not replaced. 113) in bone marrow cells of 5 adult males with Sequence of Dental Eruption (Table 5)—All 13 chromosomes, the Y being the chromosome incisors and canines and the first 2 (of 3) pre- believed absent. The complement of 2 females molars are fully erupted in 29 of the 38 available consisted of the normal 14. The investigators were skulls. One or more teeth are unerupted or par- aware that a small chromosome like the Y might tially erupted in the remaining 9 skulls. be overlooked. The "universal absence in all In two specimens of British Museum Dromi- counted plates [291 from males] makes this alter- 3 ciops, Tribe (1990, p. 568) found p erupted be- native highly unlikely. Available data favor a so- 4 fore m . In the Dromiciops at hand (Table 5), matic elimination of the Y chromosome." The eruption of the 3rd upper premolar precedes that normal complement of 14 in spermatocytes has of the upper 4th molar in 4 specimens, follows in been reported by Fernandez et al. (1979). 1, and erupts simultaneously in 2. The lower 3rd Brown and White Adipose Tissue—Absence premolar erupts after m4 in 5 specimens, precedes of brown adipose tissue (BAT) or brown fat in in 1, and appears at about the same time in 3. Dromiciops was reported by Hayward and Lisson One or more of the 4 lower incisors of Dromi- (1992). The examination of representatives of all ciops may overlap slightly but none are staggered families of Australian marsupials, the 2 families as in didelphoids. First and last functional molars of monotremes, and 3 orders of American mar- are particularly variable. supials also showed absence of BAT. At the same PSEUDOQUADRITUBERCULARITY (Fig. 32) Left time, the authors confirmed the presence of BAT 3 m (fmnh 134624) not completely erupted, ap- in all eutherians and disproved its occurrence in pears to be quadri tubercular. The "4th" cusp, other vertebrates believed to store brown fat. however, is not a hypocone (5). It is a hypertro- The phylogenetic significance of brown fat as phied metaconule (conulid d), an element of crista a unique character of eutherians rates with the pat- IV that extends from protocone (cusp 2) to me- tern of the eutherian reproductive system as evi- tacone (cusp 4). The metaconule is normally ab- dence of the independent origins of Eutheria and sent in Dromiciops and most didelphoids. The Metatheria. true hypocone (cusp 5) rises from cingulum C or Torpor and Hibernation—Dromiciops be- the posterolingual shelf of the quadritubercular comes torpid for the 2 coldest months of the year. molar. It is absent in American marsupials except Santos (1946) noted that during that period they

1 - 2 in the first 2 upper quadritubercular molars (m ) are often found in hollows of old or dry trees of caenolestids. where firewood gatherers come upon them. Live Karyotype (Fig. 33)—The chromosome com- animals Greer (1965, p. 105) captured in Malleco plement of Dromiciops gliroides is the primitive during winter assumed a flexed position in the = 2n 14 (Spotorno and Fernandez, 1971). The cold of night. When early morning temperatures karyotypic formula is shared with slight individ- dropped to 4.5°C or lower, bodies of the animals

HERSHKOVITZ: LAST OF THE MICROBIOTHERIA 49 dp 3

P2 P3 ™i

P2 P3 mi

50 FIELDIANA: ZOOLOGY m. Pi P2 dp 3

Fig. 31. Philander opossum (juvenal). Right lower cheek teeth shown are first 2 premolars erupted, third premolar and deciduous and first molar of (p,) erupting displacing dp,, (m,); compare morphology dp, with p ; and m,. For names of dental elements see page 48.

became stiff and cold to the touch. One, after be- Dromiciops in Patagonia, Argentina, found it tor- ing warmed in Greer's shirt pocket, revived, pid in the traps during winter. He caught two on crawled out of the pocket, and resumed activity 18 April "when patches of snow lay in the forest, as usual. and the overnight temperature was 0°C. ... A A male and four females were observed during captive that we awakened abruptly during the day periods of torpor by Grant and Temple-Smith on two occasions had a deep rectal temperature (1987, p. 273): of only 33° and 33.5°C, but nevertheless was alert and capable of quick and agile movements." Ol- All individuals were curled on their sides with the tail iver Schneider (1946) described the animal enter- passed over the head and fore feet and nose applied ing torpor as lying down on one side, doubling closely to the curled ventral side of the body. The rear and cold and stiff. The animal is feet were often exposed and trembled during arousal. At up, turning by some time during arousal all animals shifted their posi- nature, Oliver Schneider declared, "timid, unso- tions from this curled to a lying, posture sitting position. ciable and treacherous [!]." It is normally noctur- In all instances of the were closed and the torpor eyes nal. ears folded until surface down T s [body temperature] The heartbeat before torpor, determined by was close to that of an active animal. . . . Mann (1978), is 230 per minute, and drops to less than 30 minute when the animal is chilled to Torpidity lasted over 36 hours in two animals. Ap- per The shallow, bouts of which nea during torpidity was observed lasting 2 to 30 torpor. daily torpor, last minutes to less than 20 character- minutes, followed by several deep breaths or a may hours, ize all but one of the cases of described series of up to 30 rapid respirations. Respiration torpor rates during arousal increased from less than 100 above, including those for which time values were iRPM to over 200 RPM. Weight of the animals not available. The minimum body temperatures of the ranged from about 18 to 30 g; body surface tem- torpid examples of American marsupials perature of the active animal ranged from 28° to should range from 16° to 28°C, according to Geis- 4°C. er (1994). The same authority described hiberna-

O. P. Pearson (1983, p. 483), who trapped tion in marsupials as a deep, prolonged torpor.

Fig. 30. A, Dromiciops gliroides (juvenal): 3 upper left cheek teeth shown are, respectively, developing second

: 3 1 premolar (p ). fully functioning deciduous third premolar (dp ), and fully erupted first molar (m ); compare mor- : phology of p and dp\ B, Dromiciops gliroides (juvenal): 3 left lower cheek teeth shown are, respectively, premolar

I 3 1 of (p : ). premolar (p,) erupting and displacing dp,, and fully erupted lower molar (m,); compare morphology

and . in Ip, p : C, Dromiciops gliroides (juvenal): occlusal view of same lower teeth shown B; compare morphology

)f and . For dental dp, p 2 names of elements see page 48.

iERSHKOVITZ: LAST OF THE MICROBIOTHERIA 51 2 3 Fig. 32. Dromiciops gliroides p -m (fmnh 134624) with "fourth" cusp, a hypertrophied metaconule (d) or 2 1 functional deciduous third 3 followed formed "pseudohypocone." p fully erupted, p- displacing premolar dp , by fully

1 4 3 last or lost. For names of dental elements see 48. m ; diminutive molar (m ) unerupted page

Table 5. Sequence of dental eruption in Dromiciops: of 38 available skulls, 9 with 1 or more unerupted or partially erupted teeth are analyzed; remaining skulls not listed. ro

CO <

I CD Table 6. Revised summary of nests and occupants of Dromiciops gliroides found and/or reported by Jimenez and Rageot (1979, Table 1, p. 84).

Date to a nest in the wild caused the mother to flee, gentina, recorded its diet as insects, larvae, nest- abandoning her 4 young, 2 males and 2 females ing birds, and small mice. Mann (1955) confirmed (Table 6). that Dromiciops is carnivorous and insectivorous. A female captured in the bamboo thicket Captives held by O. P. Pearson (1983) ate apples, proved to be one of the most gentle encountered. grubs, flies, and small lizards, but not frogs. From From the beginning she allowed her head to be all accounts, it appears that captive Dromiciops stroked. Let out of the cage at night to roam free- prefers anything off the dining table to food

ly, she always returned to her nest between 7 and caught or picked outdoors. 8:30 in the morning. Stomach contents of the Dromiciops captured Trapability—Patterson et al. (1990, p. 630) by Meserve et al. (1988) in La Picada contained noted a peculiarity of Dromiciops gliroides not 71.8% invertebrates, of which 58.6% were mature exhibited by any of the other 9 small mammals arthropods. Annelids and invertebrates other than they trapped at the same time. The "highly scan- arthropods were not eaten, but some seeds and sorial habits of this species," they affirmed, "are vegetative matter, including bryophytes, were underscored by prehensile tail, opposable toes, consumed. Fungi, so much a dietary component and by their unqualified aversion to entering en- for co-occurring Rhyncholestes, Geoxus, and closed live traps; not one of 61 Dromiciops cap- Abrothrix, were hardly touched. tured at La Picada was taken in a Sherman live Greer (1965) returned to Michigan from Chile trap." with a live Dromiciops that weighed 26.5 g on Diet—Mother and young ate fruit without pref- arrival. After an average daily consumption of 4. 1 erence for any one kind. Meat, insects, or other g ham and 6 cc water, the weight of the animal proteins as part of the diet were not mentioned by increased 3 g in 5 days. the captors, Jimenez and Rageot (1979). An adult According to Mann (1978), the tongue is coated pair captured later rejected the proffered insect with a mucilagenous saliva that helps Dromiciops larvae and bananas but took apples, and bread capture and hold slow-moving prey, a question- with or without butter. Other captured Dromiciops able interpretation. accepted a variety of fruit. The tame and docile Nests (Plate 2, Table 6)—Greer (1965) found captive female that roamed the house never ac- nests in thickets of Chusquea bent to the ground cepted insects. Her preferred foods were bananas, by a fallen tree, but none in trees. He mentioned grapes, cherries, pears, honey, and marmalade. On trapping an animal in front of a burrow, implying occasion she ate a piece of cooked meat or fish. it might have been used by the monitos, but he The small birds offered were first ignored but later saw no nest. Mann (1978) saw nests in moss and eaten piecemeal. under rocks, fallen logs, and heaps of branches. A male held captive for nearly a year by Ji- During his exploration of Volcan Osorno, Mann menez and Rageot (1979) remained timid and (1978, p. 35) discovered 5 Dromiciops nests in never really tamed. His diet was 80% fruit, the hollow trees, fallen trunks, and branches of bam- remainder insects. A mouse placed in his cage boo {Chusquea) situated 1 or 2 m above ground. elicited no attention. Raw or cooked meat, eggs, The nests were made of interlaced leaves of bam- and honey failed to attract him. He did like pears, boo with openings on the side 4 or 5 cm wide. bananas, apples, tomatoes, cooked potatoes, rice, The entrances to 2 nests were each provided with and milk dessert. a short hallway fashioned of bamboo leaves. Four A captured female and her brood killed a young of the nests were roofed with moss. sparrow. She skinned the head first, then ate the Jimenez and Rageot (1979) found nests in hol- brain, following with the remainder of the bird, low trees or in tangled branches about 1 or 2 m the feathers excepted. Her young helped consume above ground. The nests described by Jimenez the bird. and Rageot were round or ovoid, usually made of The male found torpid in a bamboo nest (Table leaves of bamboo interwoven with the smallest Jimenez and ate fruit and leaves on the or with bamboo „ 6) by Rageot 80% 20% inside, made dry live insects. At one time he ate an earthworm leaves and concealed the bamboo in- „ only among in his Another a lizard Found in some nests were p placed cage. time, placed tertwinings. ectopara- J in the cage was killed and its entrails were eaten. sites, insect chitin, and eggshell. Nests seen by On the other a live introduced into Pearson in were also ^ hand, mouse (1983) Patagonia spherical, the cage lived congenially with the opossum. made of bamboo leaves, and located in shrubs or Krieg (1925), who observed the species in Ar- low trees. Most, perhaps all, other American mar-

HERSHKOVITZ: LAST OF THE MICROBIOTHERIA 55 supials build nests, many of leaves, none, to my hanced its quality. Financial support for the sur- knowledge, of bamboo. All, like Dromiciops, vey of neotropical mammals was received from avail themselves opportunistically of any suitable the Barbara E. Brown Mammal Research Fund. natural refuge, such as forest debris, tree hollows, depressions under logs, tunnels, and the like. Nothing is known of the nesting habits of canopy- inhabiting opossums. Rodents nesting in bamboo- Addendum covered tall trees, however, are known. The Rat- tus-siztd echimyid Olallamys (= Thrinacodus) I The important contribution by Szalay (1994) on captured in the Cordillera Oriental near Bogota, the evolution of marsupials came to my attention Colombia, lived in Chusquea-shrouded trees, in- too late for critical examination here. It seems also cluding tall pines, ate its fruit and shoots, and fab- that Szalay became aware of my (Hershkovitz, ricated small ovoid nests with the leaves. The re- 1992a) paper on marsupial tarsal bones and rejec- lated Kannabateomys of southeastern Brazil, Par- tion of his (1982b) classification based on them aguay and northern Argentina also lives in Chus- too late for consideration in his book beyond a quea-covered trees. perfunctory reference thereto (p. 63). Although of Enemies—Particular enemies or predators of doubtful value here, Szalay's work may be useful Dromiciops have not been mentioned by any ob- for my monograph (in preparation) on New World server to my knowledge. Mann (1978, p. 29) sug- marsupials. gested that ill-tasting secretions of dermal glands repel potential predators. More likely, the secretions are markers and sexual attractants. The Chil- ean environment harbors a few species of carni- Gazetteer of Selected Collecting vores that must prey on small mammals, probably Localities from Maps (Fig. 19)* including Dromiciops, the marmosid Thylamys elegans, and small game. Pearson and Pearson Antillanca, Puyehue, 4040/7209, 970 m, Osorno. (1982, p. 136) observed that the "dense bamboo Beatriz, Lago Nahuel Huapi, 4058/7130, 800 m, provides an effective screen against raptors and Neuquen, Argentina. that small carnivores such as skunks, weasels and Carmen, Puerto, 4308/7346, sea level, Isla Chiloe. wildcats are scarce or absent." Cayutue, 4114/7217, 250 m, Llanquihue. Specimens Examined (Skin, Skeletal, Dental, Chiloe, 4230/7355, 700 m, Isla Chiloe. Entire)—Total 59. CHILE: Isla Chiloe (Coc- Colorado, Rio, Llanquihue. auque, 1; locality not recorded, 1); Llanquihue Colorado, Rio, ca. 3825/7130, 900 m, Malleco. (Peulla, 2; Rio Colorado, 3; Rio Negro, 5; Contao, Concepcion, 3650/7303, 9 m, Concepcion. Palana, 15); Malleco (La Auracania, 3; Sierra Na- Contao, Palena, Llanquihue. huelbuta, 2; Victoria, 1); Osorno (Valle de la Pi- Huemul, Nahuel Huapi, ca. 4055/7130, 767 m, cada, 35; Osorno, 3). Neuquen, Argentina. Huite, Puerto, 4207/7326, Isla Chiloe. Itata, Rio, 3623/7252. Nuble or Biobio. La Picada, Forest Preserve, 41/7230, 450 m, Acknowledgments Osorno. Lota, 3705/7310, sea level, Concepcion. I am beholden to Kathy Kazol Telfer, Cameron Locar, Lago, 4011/7130, Neuquen, Argentina. Pfiffner, and Maggie Robertson for the illustra- Maicolpue, 4033/7346, 50-110 m, Osorno. tions, and to technical assistants Eunice Hoshizaki Madera, Puerto (Isla Victoria), ca. 4056/7133, ca. and Barbara Brown for their help in manuscript 800 m, Neuquen, Argentina. production. I am particularly grateful to my col- Nahuelbuta, Sierra, 3748/7304, 1440 m, Malleco. league Bruce Patterson for use of his photographs Nahuel Huapi National Park, 4100/7130, Neu- of bamboo, the female Dromiciops with exposed quen, Argentina. mammary field and nipples, and use of data from Negro, Rio, Llanquihue. Patterson's and Milton Gallardo's field book of Nielol, Cerro, 3840/7238, Cautin. Chilean mammals. Bruce Patterson's critical read- * ing added polish and clarity to the completed Unless indicated otherwise, all localities are Chil- manuscript, and three anonymous reviewers en- ean.

56 FIELDIANA: ZOOLOGY and S. Marenssi. 1990. Los Osorno, Volcan, 4106/7230, Osorno. primeros "ungulados" Sudamericanos de la Antarctica. Ame- Osorno, 4031/7309, Osorno. extinguidos ghiniana. 26: 3-4. Peulla, Lago Todos Los Santos, 4106/7202, ca. Bown, T. M., and J. G. Fleagle. 1994. New Colhue- 200 m, Llanquihue. huapian and Santa Crucian Microbiotheriidae and Puerto Montt, 4128/7257, Osorno. Caenolestidae from Patagonian Argentina. Journal of Puyehue National Park, Lago Paraisa, 4048/7217, Vertebrate Paleontology, 14(3): 18a (Abstract). 950 m. Osorno. Cabrera, A. 1919. Genera Mammalium. Monotremata. Museo Nacional de Ciencias Naturalcs. Quellon-Quellon (Puerto Quellon?), 4307/7337. Marsupialia. Madrid, 177 pp. sea level, Isla Chiloe. . 1958 (1957-1961 ). Catalogo de los mamfferos Quellon, Rio Yaldad, 4307/7344, Chiloe. de America del Sur. Revista del Museo Argentino de Seno Reloncavi, 4138/7245, Llanquihue. Ciencias Naturales "Bernardino Rivadavia." 4: iv + Temuco, 3844/7236, 113 m, Cautin. 732 pp. Todos Los Santos, Lago, 4106/7215, ca. 200 m, Carlini, A. A.. R. Pascual. M. A. Reguero. C. J. Scil- lato-Yarre, E. P. Tonni. and S. F. Vizcaino. 1991. Llanquihue. The first land mammal from Ant- Union, 4017/7305, 29 m, Valdivia. Paleogene placental arctica; its paleoclimatic and biogeographical bear- Valdivia, 3948/7314, ca. sea level, Valdivia. ings. VIII Jornados Argentinas de Paleontologfa de Victoria, 3813/7220, 351 m, Malleco. Vertebrados. Resumenes, 7-9 Mayo, 1991. La Rioja. Victoria, Isla, 4056/7133, ca. 800 m, Neuquen, Cifelli, R. L. 1993. Early Cretaceous mammals from North America and the evolution of dental Argentina. marsupial characters. Proceedings of the National Academy of Villa Angostura, Neuquen. Argentina. Science, USA. 90: 9413-9416. Yaldad, Rio, 4307/7344, Isla Chiloe. Collins, L. R. 1973. Monotremes and Marsupials: A Reference for Zoological Institutions. Smithsonian In- stitution Press, Washington, DC. 323 pp. Literature Cited Cunningham, R. O. 1871. Notes on the Natural History of the Straits of Magellan and West Coast of Patagon- Ameghino. F. 1887. Enumeracion sistematica de las es- ia. Edinburgh, 517 pp. pecies de mamfferos fosiles coleccionados por Carlos Dettman, M. E. 1989. Antarctica: Cretaceous cradle of Ameghino en los terrenos eocenos de la Patagonia austral temperate rainforests, pp. 89-105. In Crane. J. austral depositados en el Museo de la Plata. Boletfn A., ed.. and Evolution of the Antarctic Biota. Museo de la Plata. 1: 1-26. Origins The Geological Society Special Publications, no. 47. . 1891. Nuevos restos de mamfferos fosiles des- Geological Society, London. cubiertos por Carlos Ameghino en el Eoceno inferior Doktor. M.. A. Gazdzicki, A. Jerzmanska. S. J. Poreb- de la Patagonia austral. Especies nuevas. adiciones y ski, and E. Zastawniak. 1996. A plant-and-fish as- correcciones. Revista Argentina Historia Natural, 1: from the Eocene La Meseta formation of 289-328. semblage Seymour Island (Antarctic Peninsula) and its environ- 1898. Sinopsis geologico-paleontologica. Se- mental implications. //; Gazdzicki, A., ed., Palaeon- gundo censo de la Repiiblica Argentina. Buenos Aires, tological results of the Polish Antarctic expeditions. 1: 112-225. Part II. Palaeontologia Polonica. 55: 127-146. 1902 (1902-1904). Premiere contribution a la Dyzenchauz, F J.. M. A. Barros. J. A. W. Kirsch, and connaissance de la faune des couches mammalogique 0. A. Reig. 1993. Dromiciops australis: G-banding a Colpodon. Boletfn Academia Ciencias. Cordoba. 17: analysis reinforces its close association to Australian 5-70. marsupials, p. 82. Abstracts. Sixth International Ther- 1987. Recent advances Aplin, K. P., and M. Archer. iological Congress. University of New South Wales. in marsupial systematics with a new syncretic classi- Sydney. Australia. xv-lxxii. In Possums and fication, pp. Archer, M.. ed.. Enders, R. K. 1937. Panniculus carnosus and forma- vol. 1. & Sons, Opossums, Surrey Beatty Chipping tion of the pouch in didelphids. Journal of Morphol- Norton, New South Wales. ogy. 61(1): 1-26. Archer, M. 1993. Evolution of Australian mammal . 1966. Attachment, nursing and survival of fauna, 8-9. Abstracts, Sixth Con- pp. Theriological young in some didelphids. pp. 195-203. In Rowlands. gress. of New South Wales, Aus- University Sydney, 1. W., ed.. Comparative Biology of Reproduction in tralia. Mammals. Symposium of the Zoological Society of Biggers, J. D. 1966. Reproduction in male marsupials, London, no. 15. 251-277. In I. pp. Rowlands, W, ed.. Comparative Fernandez. R., S. Berrios. and J. Pincheira. 1979. Po- of in Mammals. Biology Reproduction Symposium. sition of the nucleolus within the nuclei of pachytene of London, vol. 15. Zoological Society spermatocytes of Dromiciops australis and Marmosa Biggers, J. D., R. F. S. Creed, and E. D. DeLamater. elegans (Didelphoidea-Marsupialia). Experientia. 35: 1963. Conjugated spermatozoa in American marsu- 1021-1023. Journal of 6: 324. pials. Reproduction and Fertility, Frakes. L. A., J. L. Probst, and W. Ludwig. 1994. Bond, M., R. Pascual, M. Reguero. S. Santillana, Latitudinal distribution of paleotemperature on land

HERSHKOVITZ: LAST OF THE MICROBIOTHERIA 57 and sea from Early Cretaceous to Middle Miocene. Jimenez, J., and R. Rageot. 1979. Notas sobre la biolo- Comptes Rendus de l'Academie des Sciences. Earth gia del monito del monte. Anales del Museo de His- and Planetary Sciences, II, 318(9): 1209-1218. toria Natural. Valparaiso, Chile, no. 12: 83-88. Gallardo, M. H., and B. D. Patterson. 1987. An ad- Kirsch, J. A. W, A. W Dickerman, O. A. Reig, and ditional 14-chromosome karyotype and sex chromo- M. S. Springer. 1991. DNA hybridization evidence some mosaicism in South American marsupials. In for the Australian affinity of the American marsupial Patterson, B. D., and R. M. Timm, eds., Essays in Dromiciops australis. Proceedings of the National Honor of Philip Hershkovitz. Fieldiana: Zoology, n.s., Academy of Science, USA, 88: 10465-10469. no. 111-115. 39: Kirsch, J. A. W., and P. F. Waller. 1979. Notes on the Geiser, F. 1994. Hibernation and daily torpor in mar- trapping and behavior of the Caenolestidae. Arkiv supials: A review. Australian Journal of Zoology, 42: Zoologie, 14: 1-104. 1-16. Krieg, H. 1925. Biologische Reisestudien in Sudamer- Goin, F. J., and A. A. Carlini. 1995. An Early Tertiary ika. Chilenische Beutelratten. Zeitschrift fur Mor- microbiotheriid marsupial from Antarctica. Journal of phologic und Okologie, 3: 169-176. Vertebrate 205-207. Paleontology, 15(1): Krumbiegel, I. 1941. Die Saugetiere der Sudamerika- Goldfuss, G. A. 1809 (1809, 1812). Vergleichende Na- Expeditionen Prof. Dr. Kriegs. Zoologischer Anzeiger, turbeschreibung der Saugethiere. Waltherschen Kunst- 135(1/2): 1-12. und Buchhandlung. Erlangen, Germany. Lyne, A. G. 1959. The systematic and adaptive signif- Grant, T. R., and P. D. Temple-Smith. 1987. Obser- icance of the vibrissae in the Marsupialia. Proceedings vations on torpor in the small marsupial Dromiciops of the Zoological Society of London, 133(1): 79-132. australis (Marsupialia: Microbiotheriidae) from south- Mann, G. 1955. Monito del monte Dromiciops aus- ern 273-291. In Possums Chile, pp. Archer, M., ed., tralis Philippi. Investigaciones Zoologicas Chilenas, and Opossums: Studies in Evolution. Surrey, Beatty 2(9/10): 159-166. and Sons, Chipping Norton, New South Wales. . 1958. Reproducion de Dromiciops australis J. K. 1965. of Malleco Province Greer, Mammals (Marsupialia, Didelphydae). Investigaciones Zoologi- Chile. Publications of the Museum, Michigan State cas Chilenas, 4: 209-213. University, Biological Series, 3(2): 53-151. -. 1978. Los pequenos mamiferos de Chile. Gay- J. P. A. Lisson. 1992. Evolution of Hayward, S., and ana Zoologfa, 40: 1-342. brown fat: Its absence in and monotremes. marsupials Marshall, L. G. 1978. Dromiciops australis. Mam- Canadian Journal of Zoology, 70: 171-179. malian Species, 99: 1-5. Hershkovitz, P. 1962. Evolution of Neotropical crice- . 1982. Systematics of the South American mar- tine rodents (Muridae), with special reference to the supial family Microbiotheriidae. Fieldiana: Geology, phyllotine group. Fieldiana: 46: 1-524. Zoology, n.s., 10: vii + 1-75.

. 1977. Living New World (Platyrrhi- Monkeys 1987. Systematics of Itaboraian (Middle Pa- ni) with an Introduction to Primates. University of leocene) age "opossum-like" marsupials from the Chicago Press, Chicago, xiv + 1 1 17 pp. limestone quarry at Sao Jose de Itaborai, Brazil, pp. 1982. The staggered marsupial lower third in- 91-160. In Archer, M., ed., Possums and Opossums:

cisor i . 6: 191-200. 3 Geobios, Memoire Special, Studies in Evolution. Surrey, Beatty and Sons, Chip- 1992a. Ankle bones: The Chilean opossum ping Norton, New South Wales. Dromiciops gliroides Thomas, and marsupial phylog- Marshall, L. G., J. A. Case, and M. O. Woodburne. eny. Bonner Zoologische Beitrage, 43(2): 181-213. 1990. Phylogenetic relationships of the families of 433-505. In H. 1992b. The South American gracile opossums, marsupials, pp. Genoways, H., ed., Current 2. Plenum New York. genus Gracilinanus Gardner and Creighton, 1989 Mammalogy Press, (Marmosidae, Marsupialia): A taxonomic review with Marshall, L. G., and C. de Muizon. 1988. The dawn notes on general morphology and relationships. Field- of the age of mammals in South America. National iana: Zoology, n.s., 70: 1-56. Geographic Research, 4(1): 23-55. 1995. The staggered marsupial third lower in- Marshall, L. G., and C. de Muizon. 1995. Part II: cisor: Hallmark of Cohort Didelphimorphia, and de- The skull. In Muizon, C. de, ed., Pucadelphys andinus scription of a new genus and species with staggered (Marsupialia, Mammalia) from the early Paleocene of

i3 from the Albian (Lower Cretaceous) of Texas. Bon- Bolivia. Memoires du Museum National d'Histoire ner Zoological Beitrage, 45(3/4): 153-169. Naturelle, Paris, 165: 21-90.

-. 1997. Composition of the family Didelphidae Marshall, L. G., C. de Muizon, M. Gayet, A. Lavenu, Gray, 1821 (Didelphoidea, Marsupialia) with a review and B. Sige. 1985. The "Rosetta Stone" for mam- of the morphology and behavior of the included genus malian evolution in South America. National Geo- Philander Tiedemann, 1808. Fieldiana: Zoology, n.s. graphic Research, 1: 274-288. 86: 1-103. Marshall, L. G., C. de Muizon, and D. Sigogneau- Hume, I. D. 1982. Digestive Physiology and Nutrition Russell. 1995. Pucadelphys andinus (Marsupialia, of Marsupials. Cambridge University Press, Cam- Mammalia) from the early Paleocene of Bolivia. Me- bridge. moires du Museum National d'Histoire Naturelle, 165: 1-164. Hunsaker, D., II. 1977. Ecology of New World marsupials, pp. 95-155. In Hunsaker, D., II, ed., The Bi- Meserve, P. L., B. K. Lang, and B. D. Patterson. ology of Marsupials. Academic Press, New York. 1988. Trophic relationships of small mammals in a

58 FIELDIANA: ZOOLOGY Chilean temperate rain forest. Journal of Mammalogy, . 1893b. Un nuevo marsupial chileno. Anales de 69(4): 721-730. la Universidad de Chile, 85: 31-34. Norris, C. A. 1993. Changes in the composition of the Philippi, R. A. 1894. Ein neues Beuteltier Chiles. Ar- auditory bulla in southern Solomon Islands popula- chiv fiir Naturgeschichte, 1(1): 33-35. tions of the cuscus, orientalis brevi- gray Phalanger Pridmore, P. A. 1994. Locomotion in Dromiciops aus- Journal ceps (Marsupialia, Phalangeridae). Zoological tralis (Marsupialia: Microbiotheriidae). Australian of the Linnaean 107(2): 93-106. Society, Journal of Zoology, 42: 679-699. Oliver Schneider, C. 1946. Catalogo de los mamiferos Reig, O. A. 1955. Noticia preliminar sobre la presencia de la de Boletin de la Sociedad provincia Concepcion. de microbiotherinos en la fauna Sudamericana. Inves- de de Chile, 21: 67-83. Biologi'a Concepcion. tigaciones Zoologicas Chilenas, 2(8): 121-129. Osgood, W. H. 1943. The mammals of Chile. Field Reig, O. A., R. Fernandez, and O. A. Spotorno. 1972. Museum of Natural History, Zoological Series, 30: 1- = Further occurrence of a karyotype of 2n 14 chro- 319. mosomes in two species of Chilean didelphoid mar- E. J. L. Pascual, R., M. Archer, Ortiz Juareguizar, supials. Zeitschrift fur Siiugetierkunde, 37(1): 37-42. Prado, H. Godthelp, and S. J. Hand. 1992a. First Reig, O. A., A. L. Gardner, N. O. Bianchi, and J. L. discovery of monotremes in South America. Nature, Patton. 1977. The chromosomes of the Didelphidae 356: 704-705. (Marsupialia) and their evolutionary significance. Bi- E. Ortiz J. L. Pascual, R., M. Archer, Juareguizar, ological Journal of the Linnaean Society, 9: 191-216. Prado, H. Godthelp, and S. J. Hand. 1992b. The Retief, J. K., C. Krajewski, M. Westerman, R. H. first non-Australian monotreme: An early Paleocene Winkfein, and G. H. Dixon. 1995. Molecular phy- South American Platypus (Monotremata, Ornitho- logeny and evolution of marsupial protamine PI genes. rhynchidae). In Augee, M. L., ed.. Platypus and Proceedings of the Royal Society of London (B). 259: Echidnas. of South Royal Zoological Society New 7-14. Wales, Sydney, Australia. Rodger, J. C. 1982. The testis and its excurrent ducts Pascual, R., and E. O. Juareguizar. 1990. Evolving in American caenolestid and didelphid marsupials. climates and mammal faunas in Cenozoic South American Journal of Anatomy, 163: 269-282. America. Journal of Human Evolution, 19(1/2): 23- 60. Romer, A. F. 1966. Vertebrate Paleontology, 3rd edi- tion. University of Chicago Press, Chicago. Patterson, B. 1956. Early Cretaceous mammals and R. the evolution of mammalian molar teeth. Fieldiana: Rosenmann, M., and Ampuero. 1981. Hibernacion en australis. Archivos de Med- Geology, 13(1): 1-105. Dromiciops Biologi'a y icina Experimentales (Santiago de Chile), 14: 294 Patterson, B. D., and M. H. Gallardo. 1987. Rhyn- (Abstract). cholestes raphanurus. Mammalian Species, 286: 1-5. Santos, J. G. 1946. La comadrejita enana Dromiciops Patterson, B. D., P. L. Meserve, and B. K. Lang. australis australis 4: 191- 1989. Distribution and abundance of small mammals (F. Philippi). Holmbergia, 195. along an elevational transect in temperate rain forests W. 1969. The middle ear of of Chile. Journal of Mammalogy, 70(1): 67-78. Segall, region Dromiciops. Acta Anatomica, 72: 489-501. Patterson, B. D., P. L. Meserve, and B. K. Lang. 1990. Quantitative habitat associations of small mam- Seward, A. C. 1914. Antarctic fossil plants. British Antarctica 1910. Natural mals along an elevational transect in temperate rain- ("Terra Nova") Expedition forests of Chile. Journal of Mammalogy, 71(4): 620- History Reports, Geology, London, 1(1): 1-49. 633. Sharman, G. B. 1973. The chromosomes of non-eu- Paula-Couto, C. de. 1962. Didelfideos fosiles del Pa- therian mammals, pp. 485-530. In Chiarelli, A. B., leoceno de Brazil. Revista Museu Argentino Ciencias and E. Capanua, eds., Cytotaxonomy and Vertebrate Evolution. Academic London. Naturales "Bernardino Rivadavia," Zoologi'a, 8(12): Press, 135-166. . 1982. Karyotypic similarities between Dromi- Pearson. A. K., O. P. Pearson, and I. A. Gomez. 1994. ciops australis (Microbiotheriidae, Marsupialia) and Biology of the bamboo Chusquea culeou (Poaceae: some Australian marsupials, pp. 71 1-721. In Archer, Bambusoideae) in southern Argentina. Vegetatio, 3: M., ed., Carnivorous Marsupials, vol. 2. Royal Zoo- 93-126. logical Society of New South Wales, Sydney, Austra- lia. Pearson, O. P. 1983. Characteristics of a mammalian fauna from forests in Patagonia, in southern Argenti- Simpson, G. G. 1964. Los mamfferos Casamayorenses na. Journal of Mammalogy, 64(3): 476-492. de la coleccion Tounnouer. Revista Museu Argentino Ciencias Naturales "Bernardino Rivadavia." Paleon- Pearson, O. P., and A. K. Pearson. 1982. Ecology and 1: 1-21. biogeography of the southern rainforests of Argentina, tologia, pp. 129-142. In Mares, M. A., and H. H. Genoways, Sinclair, W. J. 1906. Marsupialia, pp. 333-460. In Re- eds., Mammalian Biology in South America. Special ports of the Princeton University Expeditions to Pa- Publication Series, Pymatuning Laboratory of Ecolo- tagonia 1896-1899. Vol. IV, Paleontology 1, Part III: gy, University of Pittsburgh. Mammalia of the Santa Cruz Beds, pis. XL-LXV. Princeton N.J. Philippi, F. 1893a. Ein neues Beutelthier Chiles. Ver- University, Princeton, handlungen Deutsches Wissenschaft. Verein, Santiago Slaughter, B. H. 1968a. Earliest known marsupials. de Chile, 5: 318-319. Science, 162: 254-255.

HERSHKOVITZ: 59 ~.\ LAST OF THE MICROBIOTHERIA . 1968b. Holoclemensia instead of Clemensia. viventium quam fossilium fasc. 5. Friedlander and Science, 162: 1306. Sohn, Berlin.

Spotorno, O. A., and D. R. Fernandez. 1971. The Turnbull, W D. 1995. Trinity mammal jaws from the chromosomes of the "monito del monte," Dromiciops late Early Cretaceous of North Texas, pp. 266-269. In R. and H. australis Philippi. Mammalian Chromosome Newslet- Linsky, J., Renz, eds., Proceedings, 10th International ter, 12(2): 40-41. Symposium of Dental Morphology, Ber- lin. Springer, M. S., J. A. W. Kirsch, and J. A. Case. 1996. and M. Renfree. 1986. The chronicle of marsupial evolution. In Givnish, T. Tyndale-Biscoe, H., Repro- ductive of Univer- J., and K. J. Sytsma, eds., Molecular Evolution and physiology marsupials. Cambridge Adaptive Radiation. Cambridge University Press, sity Press, Cambridge. Cambridge (in press). Westerman, M., and D. Edwards. The relationship of australis to other Data from Springer, M. S., M. Westerman, and J. A. W. Kirsch. Dromiciops marsupials: DNA-DNA hybridisation studies. Australian Journal 1994. Relationships among orders and families of of Zoology, 39: 123-130. marsupials based on 12S ribosomal DNA sequences and the timing of the marsupial radiation. Journal of Wolffsohn, J. A., and C. E. Porter. 1908. Catalogo Mammalian Evolution, 2: 85-1 15. metodico de los mamfferos Chilenos existentes en el Museo de Valparaiso en Diciembre de 1905. Revista Szalay, F. S. 1982a. Phylogenetic relationships of the Chilena de Historia Natural, 12(1): 66-85. marsupials. Geobios Memoire Special, 6: 177-190. Woodburne, M. O., and J. A. Case. 1996. Dispersal, . 1982b. A new appraisal of marsupial phylog- vicariance, and the Late Cretaceous to Early Tertiary eny and classification, pp. 621-640. In Archer, M., land mammal biogeography from South America to ed., Carnivorous Marsupials, vol. 2. Royal Zoological Australia. Journal of Mammalian Evolution, 3(2):, of New South Australia. Society Wales, Sydney, 121-161. 1994. of the Evolutionary History Marsupials Woodburne, M. O., and W. J. Zinsmeister. 1984. The and an of Characters. Cam- Analysis Osteological first land mammal from Antarctica and its biogeo- bridge University Press, Cambridge. graphic implications. Journal of Paleontology, 58(4): Tamayo, H. M., and C. Frassinetti. 1980. Catalogo 913-948. de los mammfferos fosiles vivientes de Chile. Bol- y Woolley, P. A. 1974. The pouch of Planigale subtil- etin Museo Nacional de 37: 323-399. Chile, issima and other dasyurid marsupials. Journal of the of Temple-Smith, P. 1987. Sperm structure and marsupial Royal Society Western Australia, 57(1): 11-15. In Possums phylogeny, pp. 171-193. Archer, M., ed., . 1987. The seminiferous tubules, rete testis and and Opossums: Studies in Evolution, vol. 1. Surrey efferent ducts in didelphid, caenolestid and micro- Beatty & Sons, Chipping Norton, New South Wales. biotheriid marsupials, pp. 217-227. In Archer, M., ed. Possums and vol. 1. & Sons Thomas, O. 1894. On Micoureus griseus, Desm. with Opossums, Surrey Beatty and of New South Wales the description of a new genus and species of Didel- Royal Zoological Society Norton, New South Wales. phidae. Annals and Magazine of Natural History, se- Chipping ries 6, 14: 184-188. Zinsmeister, W J. 1976. Biogeographic significance ol the Late Mesozoic and Early Tertiary molluscan fau- . 1919. On small mammals collected by Sr. E. nas of Island (Antarctica Peninsula) to tht Budin in northwestern Patagonia. Annals and Maga- Seymour final breakup of Gondwanaland, pp. 349-355. // zine of Natural History, series 9, 3: 199-212. Gray, J., and A. Boucot, eds.. Historical Biogeogra Tribe, C. J. 1990. Dental age classes in Marmoset in- phy, Plate Tectonics and the Changing Environment cana and other Journal of didelphoids. Mammalogy, Proceedings, 37th Annual Biological Colloquium anc 71(4): 566-569. Selected Papers. Oregon State University Press, Cor Trouessart, E.-L. 1898. Catalogus mammalium tarn vallis.

60 FIELDIANA: ZOOLOGY

A Selected Listing of Other Fieldimm

iblication

allicebu H ! Philip Publication 141

m Graci!' mosidae, Marsupialia): A Taxonomic R By Philip H iblication 1441, $17.00

e Distribution ai Mamma! , Biliran,

C. B. I > kart, Lawrence R. I man, and Ruth

Publication 1449, $18.00

le Description of a New Spe^ >uth American Hocicud* lodontinae, Mm Critical R

l )4. 43 paj Publication 1460, $13.00

stematic Review of Som I ongtail Macaques, Macaca fasci

i Publication 1470, $35.00

ler by publication number and/or ask for a free copy of our prkv residents add current destination tax. All foreign ord< U.S. bank or the U.S. subsidiary of any foreign bank. Prices and terms subject to change withou

• to:

MUSEUM OF NATURAL HISTORY Library—Publications Divii

: Roosevelt Road at Lake Shore I