SMALL MAMMAL COMMUNITY COMPOSITION WITHIN THE MAJOR LANDSCAPE DIVISIONS OF PATAGONIA, SOUTHERN ARGENTINA

Transcription

1 Mastozoología Neotropical; 4(2): SAREM, 1997 ISSN SMALL MAMMAL COMMUNITY COMPOSITION WITHIN THE MAJOR LANDSCAPE DIVISIONS OF PATAGONIA, SOUTHERN ARGENTINA J. Adrian Monjeau 1,2, Robert S. Sikes 1, Elmer C. Birney 1, Nadia Guthmann 2, and Carleton J. Phillips 3 1 Bell Museum of Natural History and Department of Ecology, Evolution, and Behavior, University of Minnesota, 100 Ecology Building, St. Paul, Minnesota Departamento de Ecología, Universidad Nacional del Comahue, CC. 1336, Bariloche, Río Negro, Argentina. 3 Department of Biological Sciences, Illinois State University, Normal, IL ABSTRACT: The relationships between small mammal community composition and the major landscape divisions in Patagonia are poorly understood. To address this issue we trapped rodents and marsupials at 14 localities across the breadth of southern Argentina. This transect included four of the five major biozones defined for Patagonia. Cluster analysis and non-metric multidimensional scaling were used to compare small mammal assemblages to two landscape classification schemes that have been proposed for this region. Localities formed four groups based on mammal species composition with each group matching closely one of the four megabiozones defined in the study area. Axes produced by the multidimensional scaling procedure were correlated with precipitation and elevation. Small mammal communities did not follow closely the finer subdivisions of vegetative types within the major biozones, such as the commonly recognized distinction between Monte and Patagonian vegetative regions. RESUMEN: Composición de comunidades de pequeños mamíferos dentro de las grandes divisiones del paisaje de la Patagonia, sur de Argentina. Las relaciones entre la composición específica de las comunidades de pequeños mamíferos y las grandes divisiones del paisaje no han sido todavía bien estudiadas. Con el objeto de aportar a dicho problema, realizamos capturas de roedores y marsupiales en 14 sitios en el sur Argentino, cubriendo el ancho del territorio desde la frontera con Chile hasta el Atlántico. La transecta incluyó cuatro de las cinco megabiozonas recientemente definidas para la Patagonia. Utilizamos un análisis de agrupamientos como método de clasificación jerárquica y non-metric multidimensional scaling como método de ordenación para comparar los ensambles de pequeños mamíferos con los dos esquemas clasificatorios de grandes biomas que se han propuesto para la Patagonia. Las localidades de muestreo definieron cuatro grupos sobre la base de su composición de pequeños mamíferos, coincidiendo con las cuatro megabiozonas propuestas para la región. Los ejes de ordenación correlacionaron con los niveles de precipitación y elevación de cada sitio. Las comunidades de pequeños mamíferos no responden claramente a las subdivisiones internas de las biozonas recientemente propuestas, basadas en cambios en el tipo de vegetación, como en el caso de la tradicional división entre Monte y Patagonia. Key words: small mammals, community ecology, landscape, Patagonia. Palabras clave: pequeños mamíferos, ecología de comunidades, paisaje, Patagonia. Recibido 24 octubre Aceptado 10 noviembre 1997.

2 114 J.A. Monjeau et al. INTRODUCTION Biomes and finer divisions of ecosystems are typically defined by vegetational composition and viewed as biological indicators of abiotic environmental conditions (Box, 1981; Grabherr and Kojima, 1993). A correspondence between vegetational composition and community composition at higher trophic levels is expected (Clements and Shelford, 1939; Holdridge, 1947, 1964). General correspondence between small mammal community composition and vegetation has been demonstrated in deserts, grasslands, and shrubbysteppes of North America (MacMahon, 1976; Grant and Birney, 1979; Parmenter and MacMahon, 1983, respectively), and elsewhere (Morton et al., 1994). Terborgh (1971) concluded that correspondence between avian communities and vegetation was more pronounced when examined across steep environmental gradients, and Pearson and Pearson (1982) showed that small mammal species composition also can change substantially over even short geographic distances in association with pronounced environmental gradients. with the exception of Pearson and Pearson (1982), Monjeau (1989), and Kelt (1996), no attempts have been made to characterize corresponding patterns east of the Andes in Patagonia where gradients and extremes of many environmental variables, such as precipitation, are at least as pronounced as those to the west. These gradients produce some of the sharpest biotic transitions known (Quintanilla, 1983; Veblen and Lorenz, 1988) and are also associated with a marked elevational component. Those studies that have examined mammal communities in eastern Patagonia (summarized by Monjeau, 1989) have been limited in geographic scope to the first 150 km east of the Andes. We recently had the opportunity to examine small mammal assemblages across the entire breadth of southern Argentina (Fig. 1), a region characterized by areas of both sharp and gradual environmental gradients. We used data on small mammal species captured to test the hypothesis that the community composition of southern Argentine small mammals reflects the major landscape divisions. Because two systems of landscape divisions have been Several recent studies in southern South America have been conducted along transects encompassing environmental gradients. For example, Kelt et al. (1994) investigated the guild composition of small mammal assemblages in the temperate rain forest in southern Chile, and subsequently extended the area of coverage to include a more arid region of Chile (Kelt, 1996). Patterson et al. (1989) and Pearson and Ralph (1978) examined community composition of small mammals along the Pacific slope of the Andes, and Meserve and Glanz (1978) and Meserve et al. (1991) assessed community composition and structure across latitudinal gradients. These investigators all documented changes in the make-up of small mammal communities that corresponded to changes in habitat. Furthermore, several of them showed that changes in mammalian community structure were correlated with abiotic parameters such as precipitation and temperature (Meserve et al., 1991) and elevation (Patterson et al., 1990). However, Fig. 1. Location of the general study area and approximate position of localities where small mammals were trapped in Argentina. Numbers correspond to the numbered list of localities in Table 1.

3 SMALL MAMMAL COMMUNITY COMPOSITION IN SOUTHERN ARGENTINA 115 proposed for Patagonia, our data presented us with the opportunity to compare the small mammal assemblages with landscape divisions (biozones) based solely on vegetational composition (summarized by Soriano, 1983 and Soriano and Paruelo, 1992) and also with a more recently proposed system that defines megabiozones on the basis of vegetational structure coupled with geomorphology and meteorological parameters (del Valle et al., 1995). MATERIALS AND METHODS Study Area and Environmental Classification The Patagonian region of Argentina is characterized by a pronounced gradient in annual precipitation that ranges from ca mm in the Andean Valdivian Forest to <200 mm on the eastern coast (Prohaska, 1976). Two systems of classifying Patagonian landscapes have been proposed and are summarized in Fig. 2. The traditional scheme, followed by Soriano (1983), Soriano and Paruelo (1992), and investigators referenced therein, characterized biozones based entirely on vegetation. Soriano (1983) modified this system slightly from earlier versions employed by Soriano (1949, 1956), Cabrera (1953, 1971, 1978), and Cabrera and Willink (1980), to recognize three major botanical provinces: 1) the Patagonian Botanical Province ( Patagonian semi-desert ); 2) the Monte Botanical Province ( Monte desert ); and 3) the Subantartic Botanical Province ( Subantartic forest ). The boundary between the Forest and Patagonian biozones results primarily from a precipitation gradient, whereas the boundary between Patagonian and Monte vegetation types results primarily from a temperature gradient (Soriano, 1983). Recently, del Valle et al. (1995) described five major megabiozones in Patagonia by integrating climatic, vegetational, and geomorphological features rather than relying solely on vegetational data. Relative to our study, the most important differences between the del Valle et al. (1995) classification and that of Soriano (1983) are: 1) the decreased emphasis on the classic boundary between the Monte and Patagonian phytogeographic regions; and 2) the eastward extension of the Occidental biozone, which includes Meseta de Somuncura and surrounding areas of higher elevation. Data Collection and Documentation We collected small mammals in March and April, 1992, along a transect in northern Patagonia between 40 and 44 S that extended from the temperate rainforest in westernmost Argentina eastward to the more arid Atlantic coast (Fig. 1). Our 14 trapping localities, which represent all of the major landscape divisions recognized by both classification schemes in Patagonia north of Santa Cruz Province, are characterized in Table 1 with respect to geographic coordinates, elevation, annual precipitation, landscape features, and correspondence with the classification systems. We collected data on small mammal species based on captures from ~200 traps per night at each locality (~1/2 Sherman live-traps and 1/2 Museum Special snap-traps). Voucher specimens were prepared for each species and deposited in the Bell Museum of Natural History at the University of Minnesota (skins and skeletal materials) and Illinois State University (tissues). Geographical coordinates at collecting localities were determined to the nearest 1/100 of a minute with a Magellan Geographic Position System instrument Model NAV 5000D based on a minimum of three satellite readings. We recorded approximate elevation at each trapping locality using topographic maps from the Argentine Instituto Geográfico Militar (IGM), 1:250,000 (IGM, Hojas 4169-I and IV, and 4369-III). Because no data were available from meteorological stations near our trapping localities, we estimated annual precipitation from data presented by Soriano (1983), del Valle et al. (1995), Instituto Nacional de Tecnología Agropecuaria-Patagonia (INTA, in litt.), and available climatological summaries (Prohaska, 1976). Species names employed throughout follow Monjeau et al. (1994). All specimens except those of the genus Eligmodontia were identified to species based on diagnostic morphological characters (Pearson, 1995). We assigned specimens of Eligmodontia to species based on mitochondrial DNA data (Hillyard et al., 1997) matched to specimens with diploid chromosome numbers characteristic of Eligmodontia typus and E. morgani (see Ortells et al., 1989; Kelt et al., 1991; Zambelli et al., 1992), or, in the absence of these data, we used discriminant function analyses to compare specimens of unknown identity with those independently identified by one of the other two methods (Sikes et al., 1997). We consider our data based on a single night s trapping effort to be inadequate for a reliable quantitative estimate of species abundance, and thus we used only qualitative data in our treatment of small mammal communities. Because our data concerning

4 116 J.A. Monjeau et al. Fig. 2. Schematic representation of the two current landscape classifications in Patagonia. a) Major vegetational types in Patagonia summarized in Soriano (1983). b) Megabiozones and internal divisions recognized by del Valle et al. (1995). Legends include only the districts and biozones discussed in the text.

5 SMALL MAMMAL COMMUNITY COMPOSITION IN SOUTHERN ARGENTINA 117 species presence are corroborated by larger, more quantitative investigations in at least some of these regions (Pearson and Pearson, 1982; Christie, 1987; Monjeau, 1989) and by unpublished distributional information obtained from museum records, we consider them to constitute an accurate representation of species assemblages for the purposes of this study. We collected no data within the Extra-Andean Austral Megabiozone and will not consider it further. Small mammal data for our locality within the Sub-Andean Sub-humid Megabiozone are taken from a capture-recapture study on small mammals near the airport at San Carlos de Bariloche, Río Negro, and are comparable in terms of season and total trapping effort to collecting efforts at other localities except that only Sherman live-traps were used at this locality. Identification of Eligmodontia at this locality was problematic as voucher specimens from the capture-recapture study were not available. However, three specimens of Eligmodontia captured ~10 km E of this site in 1995 were all identified as E. morgani based on mtdna data, and a sample of five skulls from specimens captured at the Airport locality in 1992 and 1993 were identified as E. morgani based on discriminant function scores as described by Sikes et al. (1997). Consequently, we have reported only E. morgani from this site in the data discussed beyond. Data Analyses Species presence or absence at each locality was scored and used in cluster analysis with the simple matching option in NT-SYS (Rohlf et al., 1982). We next used non-metric multidimensional scaling (NMMDA) in PC-ORD (McCune and Mefford, 1995) to gain insight about how species assemblages influence relationships between localities. This technique is the non-parametric analog of Principal Components Analysis (PCA) that is appropriate for categorical data (Jongman et al., 1987), and it has the advantage of representing interpoint distances more faithfully than does PCA (Rolhf et al., 1982). For a better understanding of the placement of individual localities in the ordination method, we combined the results of NMMDA with the groupings made by the cluster analysis into a single figure following Gauch (1982). The ordination axes or latent variables produced by NMMDA can be considered as hypothetical environmental variables constructed without reference to actual environmental measurements (Ter Braak and Prentice, 1987). Consequently, to assess the correlation between these axes and actual environmental parameters at each of our collecting localities we used Pearson product-moment correlation analysis in SAS (SAS Institute, 1990) to compare loadings of localities on the first two axes with measurements of elevation and annual precipitation. These environmental variables were selected because correlations between them and small mammal composition has been demonstrated previously (Pearson and Pearson, 1982; Patterson et al., 1990; Meserve et al., 1991) and because reliable estimates for these variables were readily available for our collecting localities. RESULTS We captured 430 specimens representing 12 species of sigmodontine rodents and 4 species of marsupials at the original 13 localities studied, and 37 individuals representing 5 species at the Airport site (Table 1). Cluster analysis showed that localities formed discrete groups based on small mammal composition (Fig. 3), and the groups identified agreed more closely with the major landscape divisions recognized by del Valle et al. (1995) than with those of the traditional classification scheme (Fig. 2a, b; Table 1). The two localities characterized by dense forest (Los Alerces and Puerto Blest) formed a cluster distinct from all other localities. The Airport locality did not show close affinity with any other locality, but joined the cluster composed of Mencué, Pampa de Agnia, and Meseta de Somuncura before that cluster joined any other. The fourth cluster included all localities on Península Valdés (Istmo Ameghino, Puerto Pirámide, and Caleta Valdés), Estancia María Sofía, and the three sites in the transect west of Dolavon in Chubut Province. The hierarchy of our cluster results corresponds closely to the Andean, Sub-Andean, Extra-Andean Occidental, and Extra- Andean Oriental megabiozones described by del Valle et al. (1995). The locality at Mencué is the only exception to a perfect match of all localities to their correct biozone as determined by cartographic representation. Mencué is located geographically in the Extra- Andean Oriental megabiozone, but clustered with localities of the Extra-Andean Occidental megabiozone based on small mammal composition. The results of our cluster analysis did not show separation of small mammal as-

6 Table 1. Geographic position, major landscape features, correspondence to the classification systems, and small mammal composition of the study sites. 118 Annual Geographic Elevation rainfall Correspondence to Small mammals Study sites position (m) (mm) Key characteristics previous classifications collected 1) Puerto Blest, S Southern beech (Nothofagus) del Valle et al. (1995) = Rhyncholestes raphanurus Nahuel Huapi National W temperate forest. Valdivian Andean megabiozone Dromiciops australis, Park, Río Negro forest in the study site Cabrera (1978) = Abrothrix longipilis, A. Subantartic forest olivaceus, Irenomys tarsalis, Oligoryzomys longicaudatus, Loxondontomys micropus 2) Futalaufquen Lake, S Southern beech (Nothofagus del Valle et al. (1995) = Abrothrix longilipis, A. Los Alerces National W temperate forest. Coihue (N. Andean megabiozone. olivaceus, Irenomys tarsalis, Park, Chubut dombeyi) forest with bamboo Cabrera (1978) = Oligoryzomys longicaudatus understory Subantartic forest 3) Estancia El Cóndor, S Grassy steppe with patches of del Valle et al. (1995) = Abrothrix longipilis, A. near Bariloche Airport W dense, tall and spiny shrubs Sub-Andean megabiozone xanthorhinus, Eligmodontia (10 km E Bariloche), Soriano and Paruelo (1992) = morgani, Oligoryzomys Río Negro Sub-Andean District longicaudatus, Reithrodon auritus 4) Estancia María S Shrubby open steppe with del Valle et al. (1995) = Thylamys pusillus, Akodon Sofía, Río Negro W presence of dispersed creosote Border of Extra-Andean iniscatus, Eligmodontia bush and scattered grasses. Oriental megabiozone typus, Oligoryzomys Rocky outcrops present Soriano (1983) = longicaudatus, Phyllotis Occidental District xanthopygus 5) 27 km N of Pampa S Bunchgrass with dispersed del Valle et al. (1995) = Abrothrix xanthorhinus, de Agnia, Chubut W shrubs isolated or in patches. Extra-Andean Occidental Eligmodontia morgani, Rocky outcrops present megabiozone Phyllotis xanthopygus Soriano (1983) = Occidental District J.A. Monjeau et al.

7 (Cont. Table 1) 6) 30 km N of Pampa S Bunchgrass with dispersed del Valle et al.(1995) = Lestodelphys halli, de Agnia, Chubut W bushes and patches. Small Extra-Andean Occidental Abrothrix xanthorhinus, rocky outcrops present megabiozone Eligmodontia typus, E. Soriano (1983) = Occidental District morgani, Phyllotis xanthopygus, Reithrodon auritus 7) 15 km NE Mencué, S Overgrassed bunchgrass with del Valle et al. (1995) = Abrothrix xanthorhinus, Río Negro W dispersed neneos (Mulinum Extra-Andean Oriental Eligmodontia morgani, spinosum). Meadows and megabiozone Phyllotis xanthopygus rocky outcrops are present Soriano (1983) = Central District 8) 280 km W of S Dense shrubby steppe with del Valle et al. (1995) = Thylamys pusillus, Dolavon, Chubut W dispersed bushes. Rocky Extra-Andean Oriental Eligmodontia typus, outcrops present megabiozone Phyllotis xanthopygus Soriano (1983) = Central District 9) 200 km W of S Shrubby steppe with dispersed del Valle et al. (1995) = Thylamys pusillus, Dolavon, Chubut W bushes. Rocky outcrops Extra-Andean Oriental Eligmodontia typus, present megabiozone Graomys griseoflavus, Cabrera (1978) = Phylotis xanthopygus Monte Botanical Province 10) Prahua Niyeu, S Bunchgrass with patches of del Valle et al. (1995) = Akodon iniscatus, Abrothrix Meseta de Somuncura, W shrubs. Huge rocky outcrops Extra-Andean Occidental xanthorhinus, Eligmodontia Río Negro are present megabiozone typus, E. morgani, Phyllotis Soriano (1983) = xanthopygus Central District SMALL MAMMAL COMMUNITY COMPOSITION IN SOUTHERN ARGENTINA 119

8 (Cont. Table 1) ) 100 km W of S Open shrubby steppe with del Valle et al. (1995) = Eligmodontia typus, Dolavon, Chubut W dispersed bushes. Rocky Extra-Andean Oriental Graomys griseoflavus, outcrops present megabiozone Phyllotis xanthopygus Cabrera (1978) = Monte Botanical Province 12) Istmo Ameghino, S Dense shrubby steppe with del Valle et al. (1995) = Thylamys pusillus, Akodon Chubut W dispersed bushes Extra-Andean Oriental iniscatus, Eligmodontia megabiozone typus, Graomys griseoflavus Cabrera (1978), Soriano (1983), and Bertiller and Beeskow (1981) = Central District 13) Puerto Pirámide, S Shrubby steppe and dispersed del Valle et al. (1995) = Thylamys pusillus, Akodon Chubut W grasses Extra-Andean Oriental iniscatus, Eligmodontia megabiozone typus Cabrera (1978), Soriano (1983), and Bertiller and Beeskow (1981) = Central District 14) Caleta Valdés, S Overgrassed shrubby steppe del Valle et al. (1995) = Eligmodontia typus, Chubut W with dispersed bushes Extra-Andean Oriental Graomys griseoflavus megabiozone Cabrera (1978), Soriano (1983), and Bertiller and Beeskow (1981) = Central District J.A. Monjeau et al.

9 SMALL MAMMAL COMMUNITY COMPOSITION IN SOUTHERN ARGENTINA 121 semblages consistent with the division between Monte and Patagonian vegetation types recognized by Ragonese and Piccinini (1969), Bertiller and Beeskow (1979), and Soriano (1983). Furthermore, our cluster analysis placed the locality at Meseta de Somuncura within the Occidental Patagonian region consistent with the classification scheme of del Valle et al. (1995) rather than within the Central Patagonian region recognized by Soriano (1983) (Table 1). The NMMDA placed localities in multidimensional space on the basis of their small mammal composition (Fig. 4). Localities identified by cluster membership formed four groups. The proximity of these groups to one another corresponded to the hierarchichal divisions obtained from the cluster analysis and with the classification scheme of del Valle et al. (1995). The Airport locality, representing the Sub-Andean megabiozone, was positioned between the forest and the more arid localities representing the Occidental megabiozone to the east. Meseta de Somuncura was positioned between the localities of the Occidental megabiozone and those of the Oriental megabiozone. The intermediate positioning of the Airport locality and Meseta de Somuncura indicates ecotonal small mammal community composition. The species with the highest loading on the X axis of the ordination diagram (Fig. 4) were Abrothrix olivaceus, Irenomys tarsalis, Rhyncholestes raphanurus, Dromiciops australis, Loxodontomys micropus, and Abrothrix longipilis (positive) and Phyllotis xanthopygus, Eligmodontia typus, Graomys griseoflavus, Thylamys pusillus, Akodon iniscatus, Lestodelphys halli, Abrothrix xanthorhinus, and Eligmodontia morgani Fig. 3. Hierarchical classification of the localities based on small mammal composition using cluster analysis. Cluster criterion was simple matching, type = UPGMA, and cophenetic correlation coefficient was Letters indicate clusters matching the landscape classification of del Valle et al. (1995) A = Andean megabiozone, B = Sub-Andean megabiozone, C = Extra-Andean Occidental megabiozone, D = Extra-Andean Oriental Megabiozone. Scale units depict similarity.

10 122 J.A. Monjeau et al. (negative) (Table 2). This axis separated forest from semi-desert localities. Localities positioned toward the left side of Fig. 4 were characterized by species that loaded negatively on this axis whereas species that loaded positively on this axis were more commonly found at those localities positioned toward the right side. On the Y axis Abrothrix xanthorhinus and Eligmodontia morgani Fig. 4. Ordination of the localities based on small mammal composition produced by non-metric multidimensional scaling. Groupings made by the cluster analysis shown in Fig. 3 are encircled. showed the highest positive loadings, whereas Graomys griseoflavus had the highest negative loading (Table 2). In general this axis separated localities in the Oriental and Occidental biomes and species that loaded positively were more commonly found in those localities positioned toward the top of Fig. 4. The correlation analyses between the loadings of localities (n = 14) on the X axis with precipitation and elevation both were statistically significant (r = , P = 0.016; and r = 0.703, P = 0.005, respectively). In contrast, the Y axis showed only marginal correlation with precipitation (r = , P = 0.068), and a slightly stronger, but still not statistically significant, correlation with elevation (r = 0.529, P = 0.052). Precipitation and elevation were not correlated (r = 0.17, P = 0.55). DISCUSSION The hierarchical classification resulting from the cluster analysis (Fig. 3) placed the localities into discrete groups that corresponded closely to the four major landscape divisions defined by del Valle et al. (1995). The results of our ordination support these groupings as they separated the forest from semi-desert localities on the X axis and localities in the Occidental biozone from those in the Oriental biozone on the Y axis (Fig. 4). Although the match of localities in the ordination diagram to biozones indicated by geographic coordinates was very close, there was ambiguity in placement of some localities. For example, the ordination procedure placed the collecting locality at Mencué with localities from the Occidental biozone even though Mencué is geographically located in the Oriental biozone (Figs. 1 and 2b). Some localities, such as the Airport and Meseta de Somuncura, were placed intermediate to major groups and appeared to represent ecotonal transitions between biozone types. One possible explanation for the intermediate positioning of these localities is the effect of habitat patches within the landscape matrix of a transition zone (Forman and Godron, 1986). Because the ecotones between two landscape types are far from discrete (Aguiar et al., 1988; Holland et al., 1991; Hansen and Di Castri, 1992), a gradual intermingling of patches from both landscapes is expected (the Gruyere effect, Rapoport, 1982). Although this intermingling of habitat types results in ambiguous classification of ecotonal localities based on cartographic representations where boundaries are represented as a line, the appearance of ambiguity is largely cartographic. Our collecting localities at Meseta de Somuncura and Mencué are placed by the ordination either in incorrect groups as interpreted from geographic position or are intermediate between biozones. However, both of these localities are higher in elevation than any of our other sites in this region. We believe that the placement of these localities by the ordination is in fact an entirely accurate

11 SMALL MAMMAL COMMUNITY COMPOSITION IN SOUTHERN ARGENTINA 123 Table 2. Loadings of the small mammal species on the axes produced by non-metric multidimensional scaling. Species X axis score Y axis score Thylamys pusillus Rhyncholestes raphanurus Lestodelphys halli Dromiciops australis Akodon iniscatus Abrothrix longipilis Abrothrix olivaceus Abrothrix xanthorhinus Eligmodontia typus Eligmodontia morgani Graomys griseoflavus Irenomys tarsalis Loxodontomys micropus Oligoryzomys longicaudatus Phyllotis xanthopygus Reithrodon auritus depiction of their biozone affiliations. Elevation is one of the major differences between Occidental and Oriental Patagonian biozones in the classification system of del Valle et al. (1995). Our data indicating that Somuncura represents the Occidental biozone match the classification of this area by del Valle et al. (1995). Furthermore, although del Valle et al. did not identify the mesa around Mencué as belonging to the Occidental biozone, the plant assemblages present there, such as meadows with Festuca pallescens and a shrubby steppe with Mulinum spinosum, Stipa humilis, and Senecio filaginoides, match the description of del Valle et al. (1995) for other areas they did include in this biozone. Our data suggest that ca. 900 m represents the threshold above which the Oriental biozone gives way to the Occidental at this latitude. Although we found a clear distinction between mammal assemblages belonging to Occidental and Oriental Patagonia, our results show no such discrimination between sites

12 124 J.A. Monjeau et al. located in Monte and Patagonian vegetation types. These two vegetation types traditionally were considered distinct, major divisions as reflected by their classification as botanical provinces (Cabrera, 1953; Cabrera and Willink, 1980; Soriano, 1983). However, this distinction was not mirrored by differences in small mammal composition. One potential explanation for small mammal composition not following the Monte-Patagonia division is that mammals are less closely tied to specific floristic composition than to the more general environmental features that determine landscape structure. Metaphorically speaking, small mammal assemblages are landscape ecologists rather than systematic botanists. Our results corroborate the finding that precipitation and elevation gradients are important features in the determination of small mammal composition in Patagonia (Pearson and Pearson, 1982; Patterson et al., 1990; Meserve et al., 1991; Kelt, 1994, 1996), and indicate that their influence extends over a greater geographic area and over less extreme gradients than previous studies in southern South America have addressed. The ordination of localities on the X axis produced by NMMDA (Fig. 4) and the correlation of loadings of localities on this axis with precipitation and elevation suggest that the primary influence of precipitation on small mammal communities occurs in the transition from the forest to the semi-desert. However, the relatively minor variation in precipitation among the semi-desert localities, where mean annual rainfall is ca. 200 mm, is not as highly correlated with small mammal composition. Therefore, elevation appears to be a better predictor of small mammal composition than is precipitation at localities in the semi-desert regions of Patagonia. Where water availability is rarely limiting, such as in our western localities, the influence of elevation on available moisture is not critical and serves only as a coarsefocus for species assemblages. However, in semi-desert localities to the east, where water availability is severely limited, the influence of elevation on available moisture becomes increasingly critical in determining which species can or cannot exist. Precipitation and elevation were not correlated among our study sites (P = 0.55). Nevertheless, the interaction between these two variables affects water availability substantially and thereby affects biotic responses in two ways. First, the decreased water holding capacity of the atmosphere at higher elevations increases water availability at a local scale through increased precipitation and condensation (Woodward, 1987). Secondly, the lower temperature at higher elevations decreases the rate of evapotranspiration (Solomon and Shugart, 1993). If the water balance in the general area is near the threshold for various plant and animal species, the subtle influence of elevation appears sufficient to shift community composition from one type to another. This interpretation fits with our identification of the mammal assemblage at Mencué as being biologically consistent with the Occidental biozone despite the fact that Mencué is geographically located in the Oriental biozone. Small mammal community composition shows structure that is more consistent with the integrative view of landscape divisions combining vegetation, meteorological, and geomorphological features proposed by del Valle et al. (1995) than with the traditional divisions based solely on vegetation (Cabrera, 1953; Soriano, 1983; and others). The data presented here provide additional insights to the hierarchical classification proposed by del Valle et al. and the relationship of specific abiotic variables to small mammal community composition. Additional studies considering more localities over larger geographic scales will help to characterize more fully the relationship between small mammal communities and landscape divisions in Patagonia. ACKNOWLEDGMENTS Funding for our field work was primarily from a grant to ECB from the Graduate School, University of Minnesota. CONICET and the Universidad Nacional del Comahue provided salaries for the Argentine in-

13 SMALL MAMMAL COMMUNITY COMPOSITION IN SOUTHERN ARGENTINA 125 vestigators, and provided local assistance. Travel expenses for CJP were provided by HCLAS grants from Hofstra University and from NIH grant GM42563 to CJP and D.E. Pumo. Centro Nacional Patagónico provided housing for our stay in Puerto Madryn. Collecting permits were issued to JAM and ECB by Administración de Parques Nacionales, Argentina, Secretaría de Turismo, Provincia de Chubut, Argentina, and Dirección Nacional de Fauna, Argentina. We thank E.J. Cushing, D.E. Kelt, and L. Frelich for the many important insights they shared with us regarding the multivariate statistical procedures. We acknowledge also the anonymous reviewers. Special thanks are extended to our field colleagues, botanists Luciana Ghermandi and Laura Margutti, mammalogist Sofía Heinonen Fortabat, and our guide and driver, Sr. Pablo Hube. Sergio Saba s insights regarding mammalian ecology and landscape divisions in Patagonia were especially helpful. Catedral Turismo made our trip to Puerto Blest possible by providing passage on their boat and hotel facilities. Héctor Oso Ferioli assisted our field work at Nahuel Huapi National Park. Los Alerces National Park provided housing for our stay in Futalaufquen. Finally, as always, thanks are extended to our spouses, all of whom did double duty while we enjoyed rural Patagonia. LITERATURE CITED AGUIAR, M.R.; J.M. PARUELO, R.A. GOLLUSCIO, R.J.C. LEON, S.E. BURKART, and G. PUJOL The heterogeneity of the vegetation in arid and semi-arid Patagonia: an analysis using AVHRR/NOA satellite imagery. Annali di Botanica, 46: BERTILLER, M.B. and A.M. BEESKOW Las unidades florístico-fisonómicas de la vegetación de la Península Valdés y sus alrededores. VII Reunión Argentina de Ecología, Resúmenes, pp BOX, E.O Predicting physiognomic vegetation types with climate variables. Vegetatio, 45: BRAAK, C.J.F. Ter and I.C. PRENTICE A theory of gradient analysis. Advances in Ecological Research 18: CABRERA, A.L Esquema fitogeográfico de la República Argentina. Revista del Museo de La Plata, 8: CABRERA, A.L Fitogeografía de la República Argentina. Boletín de la Sociedad Argentina de Botánica, 14:42. CABRERA, A.L La vegetación de Patagonia y sus relaciones con la vegetación altoandina y puneña. Pp In: Geological relations between the southern temperate zone with the tropical mountains (Troll, C. and W. Lauer, eds.). Erdwissenschafliche Forschung, Franz Seiner Verlag, Wiesbaden, Germany, XI: CABRERA, A.L. and A. WILLINK Biogeografía de América Latina. Organización de los Estados Americanos, Programa Regional de Desarrollo Científico y Tecnológico, OEA. Monografía Nro. 13. Serie Biológica, Washington, D. C., 2nd Edition, 122 pp. CHRISTIE, M.I. (ed.) Informe preliminar del relevamiento de mamíferos de los Parques Nacionales Lanín y Nahuel Huapi. Administración de Parques Nacionales, Vol. III. 87 pp. CLEMENTS, F.E. and V.E. SHELFORD Bio- Ecology. John Wiley and Sons, New York, 425 pp. FORMAN, R.T.T. and M. GODRON Landscape Ecology. John Wiley and Sons, New York, 619 pp. GAUCH, H.G Multivariate analysis in community ecology. Cambridge University Press, Cambridge, 298 pp. GRABHERR, G. and S. KOJIMA Vegetation diversity and classification systems. Pp In: Vegetation Dynamics & Global Change (Solomon, A.M. and H. Shugart, eds.). Chapman and Hall, New York, 338 pp. GRANT, W.E. and E.C. BIRNEY Small mammal community structure in North American grasslands. Journal of Mammalogy, 60: HANSEN, A.S. and F. DI CASTRI (eds) Landscape boundaries: consequences for biotic diversity and ecological flow, Springer-Verlag, New York, 452 pp. HILLYARD, J.R.; C.J. PHILLIPS, E.C. BIRNEY, J.A. MONJEAU, and R.S. SIKES Mitochondrial DNA analysis of the silky desert mouse Eligmodontia in northern Patagonia. Zeitschrift für Säugetierkunde, 62: HOLLAND, M.M.; P.G. RISSER, and R.J. NAIMAN (eds.) Ecotones: the role of landscape boundaries in the management and restoration of changing environment. Washington, D.C., USA, Chapman and Hall, 142 pp. HOLDRIDGE, L.R Determination of world plant formations from simple climatic data. Science, 105: HOLDRIDGE, L.R Life Zone Ecology. San José de Costa Rica, Tropical Science Center, 124 pp.

14 126 J.A. Monjeau et al. KELT, D.A The natural history of small mammals from Aisén Region, southern Chile. Revista Chilena de Historia Natural, 67: KELT, D.A Ecology of small mammals across a strong environmental gradient in southern South America. Journal of Mammalogy, 77: KELT, D.A.; P.L. MESERVE, and B.K. LANG Quantitative habitat associations of small mammals in a temperate rainforest in southern Chile: empirical patterns and the importance of ecological scale. Journal of Mammalogy, 75: KELT, D.A.; R.E. PALMA, M.H. GALLARDO, and J.A. COOK Chromosomal multiformity in Eligmodontia (Muridae, Sigmodontina), and verification of the status of E. morgani. Zeitschrift für Säugetierkunde, 56: JONGMAN, R.H.G., C.J.F. Ter BRAAK and O.F.R. van TONGEREN (eds.) Data analysis in community and landscape ecology. Pudoc Wagening, 299 pp. MACMAHON, J.A Species and guild similarity of North American desert mammal faunas: a functional analysis of communities. Pp In: Evolution of desert biota (Goodall, D.W., ed.). University of Texas Press, Austin, 250 pp. MESERVE, P.L. and W.E. GLANZ Geographical ecology of small mammals in the northern Chilean arid zone. Journal of Biogeography, 5: MESERVE, P.L.; D.A. KELT, and D.R. MARTÍNEZ Geographical ecology of small mammals in continental Chile Chico, South America. Journal of Biogeography, 18: MC CUNE B. and M.J. MEFFORD PC-ORD. Multivariate Analysis of Ecological Data, Version MjM Software, Gleneden Beach, Oregon. MONJEAU, J.A Ecología y distribución geográfica de los pequeños mamíferos del Parque Nacional Nahuel Huapi y áreas adyacentes. Tesis Doctoral, Universidad Nacional de La Plata, 253 pp. MONJEAU, J.A.; N. BONINO, and S. SABA Annotated checklist of the living land mammals in Patagonia, Argentina. Mastozoología Neotropical, 1: MORTON, S.R.; J.H. BROWN, D.A. KELT, and J.R. REID Comparison of community structure among small mammals of North American and Australian deserts. Australian Journal of Zoology, 42: ORTELLS, M.O.; O.A. REIG, R.A. WAINBERG, G.E. HURTADO DE CATALFO, and T.M.L. GENTILE DE FRONZA Cytogenetics and karyosystem-atics of phyllotine rodents (Cricetidae, Sigmodontinae). II. Chromosome multiformity and autosomal polymorphism in Eligmodontia. Zeitschrift für Säugetierkunde, 54: PARMENTER, R.R. and J.A. MACMAHON Factors determining the abundance and distribution of rodents in a shrub-steppe ecosystem: the role of shrubs. Oecologia, 59: PATTERSON, B.D.; P.L. MESERVE, and B.K. LANG Distribution and abundance of small mammals along an elevational transect in temperate rainforest of Chile. Journal of Mammalogy, 70: PATTERSON, B.D.; P.L. MESERVE, and B.K. LANG Quantitative habitat associations of small mammals along an elevational transect in temperate rainforests of Chile. Journal of Mammalogy, 71: PEARSON, O.P Annotated keys for identifying small mammals living in or near Nahuel Huapi National Park or Lanín National Park, southern Argentina. Mastozoología Neotropical, 2: PEARSON, O.P. and A.K. PEARSON Ecology and biogeography of the southern rainforest of Argentina. Pp In: Mammalian Biology in South America (Mares, M.A. and H.H. Genoways, eds.). Special Publication Series, Pymatuning Laboratory of Ecology, University of Pittsburgh, 6: PEARSON, O.P. and C.P. RALPH The diversity and abundance of vertebrates along an altitudinal gradient in Perú. Memorias del Museo de Historia Natural Javier Prado, 18:1-97. PROHASKA, F The climate of Argentina, Paraguay and Uruguay. Pp In: Climates of central and South America (Schwerdtfeger, W., ed.). World Survey of Climatology, Vol 12. Elsevier, New York, 532 pp. QUINTANILLA PÉREZ, V Geografía de Chile. Tomo III. Biogeografía. Instituto Geográfico Militar. 230 pp. RAGONESE, A.E. and B.C. PICCININI Límite entre el Monte y el Semi-desierto Patagónico en las Provincias de Río Negro y Neuquén. Boletín de la Sociedad Argentina de Botánica, 11: RAPOPORT, E.H Areography. Geographic strategies of the species. Pergamon Press, New York, 269 pp.

15 SMALL MAMMAL COMMUNITY COMPOSITION IN SOUTHERN ARGENTINA 127 ROHLF, F.J.; J. KISHPAUGH, and D. KIRK NT-SYS. Numerical taxonomy system of multivariate statistical program. State University. New York, Stony Brook. SAS Institute Inc SAS user s guide: statistics, version 6. Cary, North Carolina. SIKES, R.S.; J.A. MONJEAU, E.C. BIRNEY, C.J. PHILLIPS, and J.R. HILLYARD Morphological vs. chromosomal and molecular divergence in two species of Eligmodontia. Zeitschrift für Säugetierkunde, 62: SOLOMON, A.M. and H. SHUGART Vegetation Dynamics & Global Change. Chapman and Hall, New York, 338 pp. SORIANO, A El límite entre las provincias botánicas Patagónica y Central en el territorio del Chubut. Lilloa, 20: SORIANO, A Los distritos florísticos de la Provincia Patagónica. Revista de Investigaciones Agrícolas, 10: SORIANO, A Deserts and semi-deserts of Patagonia. Pp In: Temperate deserts and semi-deserts (West, N.E., ed.). Elsevier Scientific Publishing Company, Amsterdam, 522 pp. TERBORGH, J Distribution on environmental gradients: theory and a preliminary interpretation of distributional patterns in the avifauna of the Cordillera Vilcabamba. Ecology, 52: VALLE, H.F. del; J.C. LABRAGA, and J. GOERGEN Biozonas de la Región Patagónica. Pp In: Evaluación del Estado Actual de la desertificación en áreas representativas de la Patagonia: Informe Final de la etapa I. Proyecto INTA-GTZ (Eds.). Río Gallegos-Trelew-Puerto Madryn-Bariloche, 182 pp. VEBLEN. T.T. and D.C. LORENZ Recent vegetation changes along the forest/steppe ecotone of northern Patagonia. Annals of the Association of American Geographers, 78: WOODWARD, F.I Climate & Plant Distribution. Cambridge University Press, Cambridge, 173 pp. ZAMBELLI, A.; F. DYZENCHAUZ, A. RAMOS, N. de ROSA, R. WAINBERG, and O.A. REIG Cytogenetics and karyosystematics of phyllotine rodents (Cricetidae, Sigmodontinae) III. New data on the distribution and variability of karyomorphs of the genus Eligmodontia. Zeitschrift für Säugetierkunde, 57: SORIANO, A. and J.M. PARUELO Biozones: vegetation units defined by functional characteres identifiable with the aid of satellite sensor images. Global Ecology and Biogeography Letters, 2:82-89.