A predator-habitat assessment for felids in the inland Atlantic Forest of eastern Paraguay: a preliminary analysis. (Habitat issues).

Abstract

Jaguar (Panthera onca), puma (Puma concolor), and possibly six species of small cats (ocelot, Leopardus pardalis; margay, Leopardus wiedii; oncilla, Leopardus tigrinus; Geoffroy's cat, Oncifelis geoffroyi, pampas cat, Oncifelis colocolo, jaguarundi, Herpailurus yagouaroundi) co-exist within Mbaracayu Forest Nature Reserve in eastern Paraguay. At the landscape level this 64,000-hectare island of Inland Atlantic Forest, surrounded by agricultural land, is a mosaic of forest habitats and interspersed grasslands. Habitats on the reserve include low, medium, and high forest, as well as dry and wet grasslands. The adaptive nature of most predators led us to predict that felids would occur uniformly across habitats. Tests of independence between species and habitats, however, suggest distinct associations between felids and habitats. Explanations for habitat affinities include interference competition or simply following prey to their preferred habitats. Thus, while felids in the inland Atlantic forest may be habitat generalists across their entire range, they exhibit some habitat preferences within Mbaracayu reserve, possibly as an adaptation to interspecific competition and/or prey availability. Further research is needed to determine whether these patterns continue long-term or are an artifact of the timing of our current data collection efforts.

Reserva Natural del Bosque Mbaracayu

The Mbaracayu reserve is the largest tract of undisturbed forest in eastern Paraguay covering approximately 64,000 hectares of the quarter million hectare upper Jejui watershed. Only the combined reserve areas in Brazil and Argentina around Yguazu Falls contains more of the Alto Parana formation of the Atlantic Rainforest. Located at approximately 55 [degrees] west and 24 [degrees] south, the reserve is mostly between 150 to 300m in elevation and is drained to the west by the Paraguay River. The area averages about 1800 mm of rainfall per year but is characterized by extreme unpredictability in monthly patterns from year to year. The typical dry season lasts from May to September (Sanchez 1973; FMB unpublished data). Seasons are associated with marked temperature fluctuations, with average daily high-low temperatures of 14 to 25 [degrees] C in July and 22 to 34 [degrees] C in January. The reserve contains about 90% of all species classified as rare or endangered within Paraguay (FMB 1992) and was chosen as the top priority conservation site in eastern Paraguay (Keel et al. 1993). The Mbaracayu reserve is located in the traditional homeland of Ache hunter-gatherers. The Ache have exceptional knowledge of the Paraguayan forest because they lived off wild resources until recently and most adults have spent most of their lives in the forest (Hill and Hurtado 1989). Further, the Ache are allowed exclusive rights to hunt and forage within the reserve.

Felid species within Mbaracayu Reserve

The diversity of mammalian carnivores (Mammalia, Carnivora) existing within the bounds of the reserve is high with published lists of species ranging between 13 to 17 (Hill et al. 1997; FMB 1997; MNHNP 1996). There is disagreement about the occurrence of some species, with the following listed: three or four canids [bush dog (Speothos venaticus), maned wolf (Chrysocyon brachyurus), crab-eating fox (Cerdocyon thous), pampas fox (Pseudalopex gymnocercus)], two procyonids [South American coati (Nasua nasua), crab-eating raccoon (Procyon cancrivorus)], four or five mustelids [tayra (Eira barbara), lesser grison (Galictis cuja) long-tailed otter (Lontra longicaudus), giant otter (Pteronura brasiliensis), hog-nosed skunk (Conepatus chinga)], and, five to eight felids [jaguar (Panthera onca), puma (Puma concolor), ocelot (Leopardus pardalis), margay (L. wiedii), oncilla (L. tigrina), jaguarundi (Herpailurus yagouaroundi), Geoffroy's cat (Oncifelis geoffroyi), pampas cat (O. colocolo)]. Many of these carnivore species, including all canids and felids, are listed by the Convention on International Trade in Endangered Species of Wild Fauna and Flora (CITES), either on Appendix I or Appendix II (Table 1). We chose to focus on felids as a first analysis of habitat associations by predators in Mbaracayu reserve.

The jaguar is found in many habitats throughout its range (southwestern United States through northern Argentina). This is the largest native cat in the western hemisphere with males weighing as much as 136 kg. Jaguars are a top predator, and, in Mbaracayu Reserve, evidence of predation by a jaguar on a puma has been documented (K. Hill, personal observation). Jaguars are thought to be dietary generalist and take advantage of any prey available. Pumas currently exist over a larger area (northern Canada through southern Chile and Argentina) than jaguars and function as top predators in much of their range. Pumas are also large cats (males weighing as much as 103 kg); they are considered deer specialists, although they feed on a diverse prey spectrum.

Several species of small cats potentially co-exist within Mbaracayu Reserve. The presence of the ocelot, oncilla, jaguarundi, and Geoffroy's cat at this site is confirmed by voucher specimens at the Natural History Museum of Paraguay (MNHNP 1996). Margays are considered residents (Hill et al. 1997) and pampas cat is rumored to exist in the cerrado portion on the western edge of the reserve (E. Esquivel, personal communication). This discrepancy of confirmed felid species, coupled with the limited number of local names for small cats in the area (Table 2), led us to consider all small cats as a group.

Major habitat types within Mbaracayu Reserve

Until recently, categorizing the forest mosaic of Mbaracayu was difficult. Sources for identifying individual plants and some plant communities have been available for reference for several years (Lopez et al. 1987) but no consistent description of key habitat types was available. The publication of "Plantas Comunes de Mbaracayu" provides a scientifically based habitat classification system for studies within Mbaracayu (Marin et al. 1998). Habitat descriptions provided by the Ache were synthesized into the categories outlined in Marin et al. (1998). In our treatment of habitats within Mbaracayu we recognize three (low, medium, and high) forest types, grasslands, and cerrado.

Low forest

Located on dry land above stream drainages, low forests are characterized by trees <15 m high with a typical diameter at breast height (dbh) <10 cm and none with a dbh >25 cm. The understory is dominated by bromeliads (Bromeliaceae) and ferns (Blechnaceae, Cyatheaceae). Major fruiting plants within low forest are Faramea porophylla and Coussarea contracta within the Rubiaceae. Low forest constitutes approximately 30% of the area of the Reserve.

Medium forest

Medium forest is characterized by variability. Some portions of medium forest are dominated by bamboo, either the large Guadua angustifolia bamboo or the small diameter Merostachys clausseni bamboo. There are few trees associated with the large bamboo stands and the tree height associated with small bamboo understory ranges from 15 to 25 m. Copaifera langsdorfii or Hexachlamys edulis can be the dominant woody tree species. Other portions of medium forest are dominated by lianas. These areas are typically far from a permanent water source and ground cover is sparse. Although riparian forests are distinguished by the Ache as well as in several treatments of the reserve, here it is included in medium forest for the purpose of analyses. These areas are in the downslope near streams and rivers and are generally richer in fruit species. As with low forest, medium forest covers approximately 30% of the Reserve area.

High forest

High forest, the most common plant community within Mbaracayu reserve, is described in detail in Hill et al. (1996). The canopy of this forest is relatively tall (typically 12 to 20 m with some trees exceeding 30 m) and trees commonly have epiphytic orchids (Orchidaceae) and philodendrons (Araceae). It is found at intermediate elevations above the water table on gently sloping ground. The basal area of tree species >10cm dbh is 39 [m.sup.2] per ha in high forest (Keel 1987). Ground cover in high forest is generally sparse including ferns, heliconias (Heliconia psiattacorum; Musaceae), and bromeliads. Approximately 30% of the Reserve area is comprised of high forest.

Grasslands

Grasslands are areas of open graminoid (i.e., Loudetia flammida, Axonopus suffultus, Andropogon bicornis) meadow with patches of cerrado-like vegetation (i.e., palms, Butia paraguayensis; Palmae). They are usually wet during part of the year and are characterized by poor drainage and a layer of black organic matter overlaying white sandy soils. Several leguminous species (Mimosa spp., Desmodium spp., and Chamaecrista spp.) also characterize this habitat. Wet grasslands are dominated by several grasses (Setaria paucifolia and Paspalum plicatum; Gramineae and Rhynchospora glabosa; Cyperaceae), some of which can reach two meters in height. These areas are wet throughout the year with standing water up to half meter deep during some parts of the year.

Cerrado

A large area of true cerrado habitat (~7,000 hectares) now exists within the reserve. Cerrado is a dry grassland/forest interface dominated by mixed-height grasses (Annona coriacea, Dugetia furfuracea; Annonaceae, and Eragrostis polytricha; Gramineae) and palms (Butia paraguayensis, Palmae). Because this area was not a part of the reserve when data collection was initiated, however, it has been excluded from surveys since its purchase.

Research question

Given the information at hand, we sought to determine if there are patterns of association between specific felids and major habitat types. We hypothesize that: felids in Reserva Natural del Bosque Mbaracayu exhibit no detectable habitat associations. Our rationale for this are (1) predators are generally highly adaptive, and (2) the habitats in Mbaracayu exist as a mosaic with no distinct patches available for exclusive residence.

Data collection and analysis

A stratified random sample of walking transects through the Mbaracayu reserve was employed as the field method of game censusing. Data for this analysis come from four census periods: June 1994 to April 1995, July 1995 to June 1996, July 1996 to May 1997, and July 1997 to May 1998. Five Ache assistants, walking in parallel, and a data recorder, walking 5 m directly behind the middle researcher, censused the same 91 transects each year of the study. One individual walked directly on the transect line while two assistants were located on each side, separated by 25 and 50 m. Each transect began at a specified GPS location and proceeded along a specified compass bearing throughout the day for approximately 5 km. Transects passed through all habitat types and were only suspended when water exceeded half meter in depth. Transects were continued on the far side of water bodies (swamp, streams, river, etc.) that had to be crossed. A GPS location and habitat description were noted for all types of animal encounters (for all animals at least 0.5 kg in size). The type of sighting was noted: (1) animal was seen, heard, or found in a burrow, or (2) identified through fresh sign or feces. All locations during the study were determined using a Trimble Pathfinder Pro Global Positioning System (GPS) receiver. Data were then categorized according to species (jaguar, puma, small cat) and habitat type prior to statistical analysis. In order to detect deviations from random habitat associations, a Chi-square test for independence between species occurrence and habitat was performed using SAS (SAS Institute Inc., Cary, NC 27513). Data were pooled across survey years to ensure a robust analysis.

Results and discussion

The Chi-square test resulted in a significant deviation from random ([chi square] = 22.29, P = 0.0012) indicating that some species were non-randomly associated with habitats. The relationship between species occurrence within each habitat is displayed in Figure 1. Jaguars occur within high forest habitat greater than expected and in low forest and grassland habitats less than expected. This apparent preference for high forest by jaguars may be related to greater availability of prey species; an issue that re quires further investigation. Jaguars appear to be the exclusive top predator in Mbaracayu reserve; their diet overlaps and includes all other cat species.

[FIGURE 1 OMITTED]

Pumas occur within all habitat types as expected. The lack of a pattern by pumas may reflect an overall high abundance of brocket deer, Mazama spp., within Mbaracayu reserve (Hill unpublished data).

Small spotted cats occur within low forest and grassland habitats greater than expected and in high forest habitat less than expected. The apparent preference for low forest and grasslands by small spotted cats may reflect an avoidance of jaguars, and other large predators, or a response to more abundant prey. Data for small mammal and small bird population densities are not available.

Thus, while felids in the inland Atlantic forest may be habitat generalists across their entire range, they exhibit some habitat preferences within Mbaracayu reserve, possibly as an adaptation to prey availability and interspecific competition. Research is needed to test our hypothesis that small cats appear to be associated with distinct habitats, in response to potential predation by jaguars and pumas.

Table 1. CITES listings for mammalian carnivores in Paraguay (MAG-
CITES 1999).

Carnivore Species     Scientific Name            CITES Listing

Canidae
  bush dog            Speothos venaticus         Appendix I
  maned wolf          Chrysocyon brachyurus      Appendix II
  pampas fox          Pseudalopex gymnocercus    Appendix II
Mustelidae
  long-tailed otter   Lontra longicaudus         Appendix I
  giant otter         Pteronura brasiliensis     Appendix I
Felidae
  jaguar              Panthera onca              Appendix I
  puma                Puma concolor              Appendix II
  ocelot              Leopardus pardalis         Appendix I
  margay              Leopardus wiedii           Appendix I
  oncilla             Leopardus tigrina          Appendix I
  jaguarundi          Herpailurus vagouaroundi   Appendix II
  Geoffroy's cat      Oncifelis geoffroyi        Appendix I
  pampas cat          Oncifelis colocolo         Appendix II
Table 2. Indigenous and local common names for possible small cat
species in Reserva Natural del Bosque Mbaracayu. No Ache names
exist to distinguish Margay, Geoffroy's cat, or Pampas cat.
Jaguarete'i, and Tirika are each used for multiple cat species
(Villalba and Yanosky 2000, MAG-CITES 1999, FMB 1997).

English Common Name   Ache Name(s)   Paraguayan Common Name(s)

ocelot                    Kaja       Jaguarete'i, Jaguatirika
margay                     --           Jaguarete'i, Tirika
oncilla                 Kajamini              Tirika
Geoffroy's cat             --            Tirika, Tirika'i
pampas cat                 --                  Osio
jaguarundi             Mberembo,         Jaguarundi, Eira
                       Mbekrymba

Literature cited

FMB (Fundacion Moises Bertoni). 1992. Reserva Natural Mbaracayu. Plan operative. Fundacion Moises Bertoni, Asuncion, Paraguay.

FMB (Fundacion Moises Bertoni) 1997. Reserva Natural del Bosque Mbaracayu. Plan de manejo 1997-2001. Fundacion Moises Bertoni, Asuncion, Paraguay.

Hill, K. and A. Hurtado. 1989. Huntergatherers of the New World. American Scientist 77:436-443.

Hill, K., J. Padwe, C. Bejyvagi, A. Bepurangi, F. Jakugi, R. Tykuarangi, and T. Tykuarangi. 2000. Sustainability of Ache hunting in the Mbaracayu Reserve, Paraguay. Pp 79-105 in J. Robinson and E. Bennett, eds. Hunting for sustainability in tropical forests. Columbia University Press, New York.

Hill, K., J. Padwe, C. Bejyvagi, A. Bepurangi, F. Jakugi, R. Tykuarangi, and T. Tykuarangi. 1997. Impact of hunting on large vertebrates in the Mbaracayu Reserve, Paraguay. Conservation Biology 11:1339-1353.

Keel, S. 1987. Informe de viaje al terrano del Banco Mundial. Centro de Datos para la Conservacion, Asuncion, Paraguay.

Keel, S., A. Gentry, and L. Spinzi. 1993. Using vegetation analysis to facilitate the selection of conservation sites in eastern Paraguay. Conservation Biology 7:66-75.

Lopez, J.A., E.L. Little, Jr., G.F. Ritz., J.S. Rombold, W.J. Hahn, W.R. Sterrett, and M.J. McDonald. 1987. Arboles Comunes del Paraguay. Cuerpo de Paz, Coleccion e Intercambio de Informacion, Asuncion, Paraguay.

MAG-CITES (Ministerio de Agricultura y Ganaderia -- Oficina CITES-Paraguay). 1999. Especies Nativas de Paraguay. MAG-CITES, Asuncion, Paraguay

Marin, G., B. Jimenez, M. Pena-Chocarro, and S. Knapp. 1998. Plantas Comunes de Mbaracayu. The Natural History Museum, London.

MNHNP (Museo Nacional de Historia Natural del Paraguay). 1996. Colecciones de Flora y Fauna del Museo de Historia Natural del Paraguay. Direccion de parques Nacionales y Vida Silvestre, Ministerio de Agricultura y Ganaderia, Asuncion, Paraguay.

Sanchez, T.F. 1973. The climate of Paraguay. Pp 33-38 in J.R. Gorham, ed. Paraguay: ecological essays. Academy of the Arts and Sciences of the Americas, Miami.

Villalba, R. and A. Yanosky. 2000. Guia de Huellas y Senales: Fauna Paraguaya. Fundacion Moises Bertoni and U.S. Agency for International Development, Asuncion, Paraguay.

Gerald L. Zuercher
Ph.D. Candidate, Kansas Cooperative Fish and Wildlife Research Unit,
Kansas State University, 205 Leasure Hall, Manhattan, KS 66506; (785)
532-6336; (fax) [phone]; Research Associate, Sunset Zoological
park; [e-mail]

Philip S. Gipson
Unit Leader, Kansas Cooperative Fish and Wildlife Research Unit, Kansas
State University, 205 Leasure Hall, Manhattan, KS 66506; (785)
532-6070; (fax) [phone]; [e-mail]

Kim Hill
Associate Professor, Department of Anthropology, University of New
Mexico, Albuquerque, NM 87131; [phone]; (fax) [phone];
[e-mail]