Jaguars (Panthera onca) remain virtually unstudied in the desert environments at the northern extent of their range. Historic sightings from the United States indicate a declining population of resident jaguars from the late 1800s into the 1940s, after which only occasional jaguars were reported until the present. After 2 sightings of jaguars in 1996, we established a camera monitoring program in southeastern Arizona. From March 2001 to July 2007, we maintained 9–44 trail cameras and conducted opportunistic track surveys. We documented 2 adult males and a possible 3rd unidentified jaguar with 69 photographs and 28 sets of tracks. One jaguar, originally photographed in 1996, was resighted 64 times during 2004–2007. This ≥ 13-year-old male used habitats from the Sonoran lowland desert at 877 m above sea level to pine–oak woodlands at 1,577 m, and covered 1,359 km2 in 2 mountain complexes. Despite speculation that recent sightings of jaguars in the United States represented dispersing transients on sporadic forays from Mexico, we documented jaguars in Arizona frequently, continuously, and year-round, and videotaped several scent-marking behaviors, indicating the residency of adult jaguars within Arizona. We outline the importance of maintaining cross-border connectivity for long-term survival of the wide-ranging and thinly distributed binational population of jaguars. We recommend further research and we stress the fragmentation consequences of the proposed United States–Mexico border fence to the northernmost jaguar population, and particularly to jaguars in the United States.
Jaguars (Panthera onca) are typically associated with the rain forests of Central and South America; however, the species historically ranged into the arid southwestern United States (Rabinowitz 1999; Sanderson et al. 2002; Seymour 1989). At risk throughout their range because of habitat loss and overhunting (Swank and Teer 1989), jaguars currently occupy only 46% of their former (pre-1900) range (Sanderson et al. 2002). A general shortage of information on the species in the American Southwest has led to the widely accepted assumption that jaguars observed in Arizona and New Mexico were not residents of the United States, but rather young, dispersing transients (Boydston and López González 2005; Rabinowitz 1999) on sporadic forays from Mexico (United States Fish and Wildlife Service 2006). The common assumption has been that the northernmost breeding population is concentrated at the junction of the Aros, Bavispe, and Yaqui rivers ∼200 km south of the United States–Mexico border (López González and Brown 2002; Martínez-Mendoza 2000; Rosas-Rosas 2006), and dispersing offspring occasionally stray north of the border (Brown and López González 2001; United States Fish and Wildlife Service 2006; Hatten et al. 2005; O'Neill and Van Pelt 2004). However, evidence of this type of wandering dispersal has not been documented.
Conversely, the confirmed historical records of jaguars from Arizona and New Mexico suggest a declining population of resident jaguars until the mid-1900s (Fig. 1), with reproduction occurring until 1910 and females documented as late as 1963 (Brown 1983; Brown and López González 2001; Hatten et al. 2005; Hoffmeister 1986; Rabinowitz 1999). Of the 61 historic records of jaguars in the United States from 1880 to 1995, sex was determined for only 25. Of those, 7 (28%) were females, and 3 (43%) of those were with young when killed. Thus, 12% of all known jaguars in Arizona were females raising young, clearly representing a breeding population north of the border.
Three policy changes in the late 1960s and early 1970s led to a sudden decrease in reported records, giving the impression that jaguars had disappeared from the Southwest, even if some persisted in the region. Bounties on predators were removed in the late 1960s (Brown and López González 2001; Hansen 1994, 2006), and it became illegal to kill jaguars under Arizona State Commission Order 14 in 1969 (Johnson and Van Pelt 1997). Establishment of the Endangered Species Act in 1972 resulted in local fear of federal restrictions on private property and general animosity toward endangered or threatened species (Brook et al. 2003; Haines et al. 2006; Treves and Karanth 2003). Not only did direct persecution of jaguars decrease, but the probability of any jaguars being reported was greatly reduced. This resulted in the appearance that jaguars had been extirpated from the southwestern United States, and there was no way of truly knowing their actual status in the region after 1970. The United States Fish and Wildlife Service, apparently considering the species extirpated, overlooked the jaguar for protection in the United States under the Endangered Species Act of 1973 (Rabinowitz 1999). A jaguar was subsequently killed in the Dos Cabeza Mountains of Arizona in 1986 (Brown and López González 2001), indicating the potential for persistence of jaguars in the region.
On 7 March 1996, mountain lion hunting guide and rancher, Warner Glenn, photographed a live, wild jaguar in the Peloncillo Mountains of southeastern Arizona (Glenn 1996). Six months later, a different jaguar was photographed in the Baboquivari Mountains of south-central Arizona (Childs 1998). Based on body size and conformation, it was determined that the Peloncillo jaguar was an adult male (≥3–5 years old), and the Baboquivari jaguar was a younger male between 2 and 3 years of age (O'Neill and Van Pelt 2004). In February 2006, W. Glenn photographed another male jaguar (different from the Peloncillo jaguar and the Baboquivari jaguar) in the Animas Mountains of southwestern New Mexico (W. Glenn, Malpai Borderlands Group, pers. comm.). Several photographs of the large (approximately 90 kg), snarling jaguar show rounded and yellow teeth, indicating the animal was an adult.
The 2 sightings in 1996 sparked interest in determining the current status of jaguars at the northern extent of their historic range, and reinforced the urgent need to identify and conserve jaguar habitat in the southwestern United States and northern Mexico (Johnson and Van Pelt 1997; O'Neill and Van Pelt 2004; Rabinowitz 1999). State and federal wildlife and land management agencies of the region formed the Arizona–New Mexico Jaguar Conservation Team (JAG Team) to protect and manage the jaguar in the Southwest (Johnson and Van Pelt 1997). The United States Fish and Wildlife Service formally listed the jaguar as an endangered species in 1997 (United States Fish and Wildlife Service 2006).
The conservation of populations at the periphery of a species range is now considered extremely important to the long-term survival of endangered species (Abbitt et al. 2000; Channell and Lomolino 2000; Nielsen et al. 2001); therefore, conservation of jaguars in the northernmost portion of their range (i.e., the borderlands population) deserves attention equal to or greater than that of core populations (Johnson et al. 2007). Furthermore, as global climate trends change toward hotter, drier environments (Karl et al. 1996; Walther et al. 2002), Jaguars living in the borderlands may become even more important to the survival of the species. Effective conservation of jaguars will require maintaining sufficient core and connective habitats to avoid population fragmentation and thus reduce the probability of extinction (Quigley and Crawshaw 1992; Sanderson et al. 2002; Weber and Rabinowitz 1996). The Secure Fence Act of 2006 (Secure Fence Act 2006) mandated the United States Department of Homeland Security to physically separate Mexico from the southwestern United States with steel fences 3–4 m high across 1,280 km of the United States–Mexico border, including ∼70% of the Arizona border (from Calexico, California, to Douglas, Arizona). Plans to complete the fence before 2009 have been expedited with no apparent regard to wildlife occupying the borderlands region (Segee and Neeley 2006). The border fence may effectively partition the already small, northernmost population of jaguars and isolate jaguars in the United States from the larger source population in northwestern Mexico (Brown 1983; Brown and López González 2001; Rabinowitz 1999; Spangle 2007).
The status, distribution, and basic ecology of jaguars living in the borderlands remain virtually unknown. The only existing data are from historical records of jaguars killed by hunters, trappers, and predator control agents over the past century (Brown 1983), which are considered valid only if accompanied by physical evidence (Tewes and Everett 1986). Several authors have used these records to develop maps of potential habitat for jaguars in the region (Boydston and López González 2005; Hatten et al. 2005; Menke and Hayes 2003). Records are scarce and difficult to verify, however, and this information is of unknown value for delineating the historic or current range and habitat use by jaguars in the region (Hatten et al. 2005; McCain et al. 2006; Sanderson et al. 2002).
Rabinowitz (1999) stressed the need to gather basic information on jaguars in the southwestern United States. In 2001, we initiated efforts to obtain data on jaguars in southern Arizona by implementing presence–absence surveys with trail cameras and track surveys in areas considered potentially suitable for the species (Brown and López González 2001; Hatten et al. 2005). The objectives of our study were to examine the current status and distribution of jaguars in southeastern Arizona, and to obtain baseline information on general habitat associations and travel corridors to guide meaningful conservation of the jaguar in the borderland region.
Materials and Methods
Study site.—The study area included mountain ranges in southern Arizona with potential jaguar habitat as identified by the JAG Team's Habitat Subcommittee and the Arizona Game and Fish Department, and in proximity to verified historic sighting locations (Brown and López González 2001; Hatten et al. 2005). The study extended from the crest of the Baboquivari Mountains east to the San Rafael Valley and approximately 80 km north of the international border, encompassing portions of Coronado National Forest, Buenos Aires National Wildlife Refuge, lands managed by the Bureau of Land Management and the state of Arizona, and several private ranches (Fig. 2). Exact locations and location names are withheld for the protection of the jaguars and their habitat, as requested by the Arizona Game and Fish Department and the JAG Team (T. Johnson, Arizona Game and Fish Department, pers. comm.; W. Van Pelt, Arizona Game and Fish Department, pers. comm.).
The borderlands region encompasses biotic communities of Madrean evergreen woodland and semidesert scrub grassland (Brown 1994). The oak woodland–oak grassland community is the dominant vegetation type between 1,100 m and 2,000 m. Below 1,100 m, mesquite–Sonoran desert scrub predominates (Brown 1994; Hatten et al. 2005; Toolin et al. 1979). The climate is semiarid with approximately 400 mm annual rainfall, half of which occurs between July and September (Hass 2002).
Trail camera surveys.—Based on population sampling techniques for tigers (Panthera tigris—Karanth 1995; Karanth and Nichols 1998; Karanth et al. 2004), we implemented a series of systematic surveys with automatic trail cameras arranged in grids. We placed at least 1 trail camera along the most probable travel route within each 5-km2 grid and maintained >1 km between cameras (Karanth et al. 2004; Silver et al. 2004). We recorded only the 1st detection when > 1 photograph was taken of the same animal at a given camera site within 60 min (Bridges et al. 2004).
We focused our surveys on major travel routes and natural funnels through core and connective habitats in the mountain ranges within our study area. Movement patterns and habitat use by jaguars were unknown in this region, but we assumed jaguars would use the landscape and travel routes similarly to pumas (Puma concolor), the other large felid in the area. Cameras were placed on likely travel routes along washes, trails, dirt roads, ridges, and canyon bottoms (Karanth et al. 2004; Silver et al. 2004), and in areas where wildlife travel was naturally directed by landscape features.
We employed an “adaptive cluster sampling” scheme, which builds on a simple or stratified sampling design by sampling all cells bordering those where presence had previously been recorded until the cluster is surrounded by cells that fail to detect presence (Karanth and Nichols 1998; Karanth et al. 2004). Where jaguars were photographed, we placed a 2nd camera on the opposite side of the trail to record the patterns of spots on both sides of the animals, following methodologies for unambiguous identification of individuals (Karanth et al. 2004; Silver et al. 2004; Wallace et al. 2003).
We used trail camera systems designed for continuous surveillance of medium to large mammals. We used 2 types of cameras: digital (Cuddeback Digital; Non Typical, Inc., Park Falls, Wisconsin) and film (CamTrakker; Camtrak South, Inc., Watkinsville, Georgia). All cameras were fully automatic and used passive infrared motion sensors that detect heat-in-motion. We attached camera units to trees, aimed approximately 0.3 m above the level of the trail or travel route. Cameras monitored activity 24 h per day with a 5-min delay between photographs to avoid multiple photographs of the same individual (Main and Richardson 2002). We serviced cameras approximately every 6 weeks to replace batteries and film or memory cards.
Activity patterns.—We examined daily activity patterns of jaguars using the time stamp on each photograph (Bridges et al. 2004; Pierce et al. 2000). We divided the 24-h day into eight 3-h time categories: 0131–0430 h = LateNight; 0431–0730 h = Dawn; 0731–1031 h = EarlyDay; 1031–1330 h = MidDay; 1331–1630 h = LateDay; 1631–1930 h = Dusk; 1931–2230 h = EarlyNight; and 2231–0130 h = MidNight.
Track surveys.—We conducted track surveys opportunistically, especially while traveling to service cameras. We searched for tracks in promising areas where substrates were favorable for track detection, such as along trails, dirt roads, sandy washes, canyon bottoms, and around ponds and water holes. Jaguar tracks were differentiated from those of pumas in the following ways: plantar pad is proportionally larger, wider, and more rounded without pronounced lobes at the base; plantar pad extends forward to the base of the toes, giving the overall track a compact or filled-in appearance, with little negative space between the toes or between the plantar pad and toes; toes are proportionally larger and more rounded; and front feet exhibit broad straddle, whereas hind feet have narrower straddle and fall inside front feet on overstep (Aranda 2000; Childs 1998).
Mapping.—Locations of photographs and tracks were recorded on handheld global positioning system units and imported into ARC-GIS 9.0 (ESRI, Redlands, California) for mapping and analyses of areas of use based on minimum convex polygons.
Animal care and use guidelines.—All field methodologies were approved by Humboldt State University Institutional Animal Care and Use Committee (IACUC nos. 04/05.W.43.A and 06/07.W.133.E) and followed the guidelines approved by the American Society of Mammalogists (Gannon et al. 2007). All field work was permitted by the United States Fish and Wildlife Service, Arizona Game and Fish Department, United States Department of Agriculture Forest Service, and Buenos Aires National Wildlife Refuge.
We operated between 9 and 44 trail cameras in southeastern Arizona from March 2001 to July 2007 for a total of 30,260 camera “trap” nights (Table 1). We obtained 69 photographs of jaguars (including 5 video clips) and documented 28 sets of jaguar tracks and 1 calf (Bos taurus) killed by a jaguar (Fig. 3). We photographed 2 adult male jaguars, dubbed Macho A and Macho B, and a possible 3rd individual that we were unable to identify or determine as to sex. In 2001–2004 we obtained 3 right-side profile photographs of Macho A; however, we were unable to photograph the spot pattern on his left side. We photographed Macho B 65 times during 2004–2007, recording his spot patterns on both sides with multiple simultaneous images from paired cameras. Shortly before and after we last photographed Macho A, we obtained 2 photographs of the left profile of a jaguar. These pictures may have been from the left side of Macho A (not observed in previous photos) or of a different, unknown jaguar. The 2 jaguars photographed by Glenn (1996; W. Glenn, Malpai Borderlands Group, pers. comm.) were different from the individuals photographed in this study.
|Year||Camera sites||Camera trap-nights||Jaguar photographs||Jaguar detection rates (photos/100 “trap” nights)|
|Year||Camera sites||Camera trap-nights||Jaguar photographs||Jaguar detection rates (photos/100 “trap” nights)|
Photographs and detection rates of jaguars (jaguars/100 trap nights) increased over the course of the study as monitoring intensified (Table 1). Detections were temporally clumped, with several photographs or tracks or both occurring over several months, followed by periods with no detections (Fig. 3), suggesting that jaguars were moving outside the area we surveyed. Nevertheless, photographs and tracks of jaguars were detected during every month of the year, indicating that movements outside of the study area were not seasonal. Jaguars were almost exclusively nocturnal, with most photographs (86.6%) taken between Dusk (1631–1930 h) and LateNight (0131–0430 h; x2 = 39.63, d.f. = 7, P < 0.005).
Based on body size and conformation, the JAG Team Scientific Advisory Group estimated ages of the 2 known jaguars. In 2001, Macho A was fully grown and estimated to be at least 3–5 years old (H. Quigley, Hornocker Wildlife Institute, pers. comm.), which meant he was 6–8 years old when photographed in September 2004. The pattern of spots on Macho B matched pelage features of the jaguar that was photographed in 1996 in the Baboquivari Mountains (Childs 1998). In 1996, this jaguar from the Baboquivari Mountains was fully grown and at least 2 years old, indicating that Macho B was 13–14 years old when photographed in July 2007 (Fig. 4).
Macho B was resighted 89 times between August 2004 and July 2007 including 60 photographs, 5 video segments, and 25 sets of tracks (which were attributed to Macho B based on track measurements [Childs 1998] and spatial and temporal association with photographic records). Also, in 2 of the 5 video segments Macho B exhibited territorial scent-marking behavior with backward urine-spraying, cheek-rubbing, and claw-raking.
Although our study design was insufficient to estimate true home-range or habitat selection, knowledge of the wide-ranging movements by Macho B through diverse habitat types is important for conservation of the species. The most continuous 12 months of detections for Macho B were between May 2006 and April 2007; 33 separate sightings were recorded for this jaguar. A minimum convex polygon assessment of these locations yielded an estimated area of use for Macho B to be 1,359 km2 (Fig. 2). The most continuous sequence of locations occurred from 4 October to 5 October 2005, in which Macho B crossed the border into the United States and traversed a 20-km straight-line distance northward along the Atascosa Mountain complex in 25.5 h. Jaguars used areas from 1,577 m elevation in the Baboquivari and Atascosa mountain complexes to flat lowland desert floor of the Altar Valley at 877 m (Fig. 5). The average elevation for the 89 different locations where jaguars were detected during the study was 1,159 m.
Our data on the presence-absence of jaguars provides valuable new information on the current status and distribution, basic habitat associations, travel patterns, longevity, and activity patterns of jaguars in the borderlands region. Although the types of data we generated likely underestimated actual animal movements (Soisalo and Cavalcanti 2006), the information does begin to describe core and connective habitats in the border region between the United States and Mexico. Examination of these data identifies high-priority areas for conservation of jaguars in the transborder region between these 2 countries (Fig. 2).
Of the 4 recent jaguars confirmed in the United States since 1996, none were dispersing or “transient” juveniles. Macho A and Macho B and the 2 different jaguars photographed by Glenn (in 1996 and 2006) near the Arizona–New Mexico border were all adult-aged animals well beyond the age of dispersal (15 months to 2 years—Nowak 2005; Seymour 1989; Sunquist and Sunquist 2002). Macho B was photographed north of the border over 11 total years, and he evidently resided in Arizona continuously and year-round from 2004 to 2007. These data clearly demonstrates the presence of resident, adult jaguars with at least some portion of their home ranges or territories within the continental United States.
Further evidence of residency within the United States was demonstrated in 2 video recordings of Macho B in which he scent-marked in the form of backward urine-spraying, cheek-rubbing, and claw-raking. Scent-marking in free-ranging felids serves as important chemical communication among conspecifics to display territoriality and to defend an established home range (Hornocker 1970; Rabinowitz 1986; Schaller 1967; Seidensticker et al. 1973; Smith et al. 1989).
Scent-marking behavior has not been well studied in jaguars (Rabinowitz 1986); however, knowledge of this behavior in other felids (Hornocker 1970; Schaller 1967; Seidensticker et al. 1973; Smith et al. 1989) suggests that Macho B was a territorial resident attempting to communicate with 1 or more jaguars in the area. Studies of this behavior in tigers in Nepal, for example, demonstrated that transient individuals did not begin to scent-mark until after they had established a territory (Smith et al. 1989). Also, both male and female tigers scent-marked, a behavior that increased for both sexes when females were in estrus (Smith et al. 1989).
Movements and habitat associations.—The arid environment in of the borderlands region of Arizona and Sonora, Mexico, contains resources and environmental conditions that are more variable than those in tropical habitats (Hass 2002). Because of this, we would expect jaguars to have larger home ranges and lower densities in the borderlands than in tropical rain-forest habitat (Rosas-Rosas 2006). Between 2004 and 2007, Macho B used an area of 1,359 km2 from the Baboquivari Mountain complex in the western part of our study area across the Altar Valley to the Atascosa Mountain complex in the east (Fig. 2).
In general, most locations of jaguars from our study were concentrated in the mountain ranges, which was likely an artifact of our sampling scheme. The majority of monitoring was focused in mountainous areas where landscape features helped direct animal movements and travel (Karanth et al. 2004). Passive monitoring was not done in the relatively featureless open desert flatlands because there were simply too many probable travel routes to monitor. Although Macho B would have crossed the Altar Valley between the Baboquivari and Atascosa mountains, we have no information on the extent of use of desert flatland areas by jaguars in the study area. Notably, however, Hibben (1948) reported following behind a jaguar on horseback more than 2 days as the animal crossed an open desert valley with little or no cover in southern Arizona.
Although the types of data we collected from camera traps and track surveys cannot be used for a quantitative assessment of habitat use by jaguars in the borderlands study area, we do know that Macho B used an extensive area, including habitats of the Sonoran lowland desert, Sonoran desert scrub, mesquite grassland, Madrean oak woodland, and pine-oak woodland. The wide-ranging movements of jaguars through varied habitats and topography evident from our study emphasize the importance of maintaining adequate expanses of intact core and connective habitats for the jaguar in this region.
On 5 occasions we photographed and tracked jaguars as they moved back and forth across the United States–Mexico border. Movements of jaguars across the international border appeared focused in key connective habitats in several mountain ranges and canyon bottoms that span the international border (Fig. 2).
Importance of cross-border connectivity.—Jaguars in Arizona and New Mexico, currently and perhaps historically, are likely neither dispersing transients from Mexico nor members of a distinct United States population. We suggest that jaguars in the United States and those recently documented in northeastern Sonora, Mexico (Brown and López González 2001; Rosas-Rosas 2006), represent small segments of a large but widely distributed, low-density population at the northern extreme of the species range. Although there has been speculation on the existence of a concentration of breeding jaguars in northeastern Sonora, Mexico (Brown and López González 2001; United States Fish and Wildlife Service 2006), Rosas-Rosas (2006) photographed only 5 jaguars over a 400-km2 area in this region during 5 years of research (estimated density = 1 jaguar/100 km2). López González (Universidad Autónoma de Queretaro, pers. comm.) estimated the same density and modeled that each jaguar would require approximately 1,800 km2. Examination of these data suggests that jaguars in northeastern Sonora, Mexico, are actually more sparsely distributed than any other documented population (Rosas-Rosas 2006). With these distributions and movement potentials, we suspect that a single, thinly distributed population likely inhabits the large area from southern Arizona and New Mexico, south through the mountains of eastern Sonora, Mexico.
The interdependence of the American and Mexican portions of the population of jaguars in the borderlands region has important implications for conservation. In Sonora, Mexico, jaguars are seriously threatened by loss of habitat, reduced prey populations, and hunting (Brown and López González 2001; Rosas-Rosas 2006; Valdez 1999). In contrast, jaguars in the United States occupy large expanses of public lands where federal protection for jaguars is enforced (United States Fish and Wildlife Service 2006), native prey are managed at healthy numbers (Arizona Game and Fish Department 2006), and a program to compensate producers for losses of livestock to jaguar depredation has alleviated concerns of local stakeholders (O'Neill and Van Pelt 2004; Sayre 2005). According to Spangle (2007), the overall area of potential habitat for jaguars in Arizona and New Mexico is equal to or greater than the area of suitable habitat for jaguars in Sonora, Mexico. Thus, the availability of suitable habitat for jaguars in the southwestern United States will be increasingly important for the long-term survival of the species in the borderlands region. Furthermore, with no known breeding north of the border since 1910, jaguars in the United States also are dependent on reproduction in Mexico.
Current threats.—The most critical and imminent threat to jaguars in the United States is the proposed border fence. The border fence will separate the small segment of the borderlands population in Arizona from those in northeastern Sonora, Mexico, thereby eliminating dispersal and preventing recovery in jaguar numbers or range north of the border. In August 2007, the Department of Homeland Security initiated construction of the border fence within the study area. Among the 1st areas to be fenced were 11 km across the southern boundary of Buenos Aires National Wildlife Refuge in the Altar Valley and a 4-km extension of the existing fences in Nogales, Arizona, west to the Atascosa Mountain complex. The United States Fish and Wildlife Service Biological Opinion on the border fence detailed the combined direct and indirect impacts to jaguars, emphasizing large-scale risks associated with fragmentation of an already small population related to loss of gene flow (Spangle 2007). The biological opinion also warned that migrant traffic and associated law enforcement activities will likely shift from the desert floor into mountainous habitats where disturbance by humans and habitat degradation will have greater negative effects on jaguars than before (Spangle 2007). This is important for jaguars because the mountainous Baboquivari and Atascosa Mountain complexes appear critically important for maintaining connectivity between the Arizona and Sonora, Mexico, portions of the jaguar population in the borderlands (Fig. 2). An extensive fence along the United States-Mexico border would likely effectively fence jaguars out of the United States, preventing dispersal and gene flow from northern Mexico, and bring an end to naturally occurring jaguars in the United States.
Direct killing of jaguars in predator control efforts is another very serious threat, especially in Mexico (Brown and López González 2001; Rosas-Rosas 2006). Also in the United States, recent debates over designation of Critical Habitat and federal restrictions on local land-use practices have alienated important local conservation stakeholders (Brook et al. 2003; Curtin 2002; Rabinowitz 1999), causing greater animosity toward jaguars that further threatens the safety of jaguars (United States Fish and Wildlife Service 2006). On a broader scale, Arizona has the fastest-growing human population in the United States (Bernstein 2006), and housing developments are severing connective habitats (Abbitt et al. 2000; Curtin 2002; Spector 2002). Finally, large-scale, open-pit mines threaten known core habitat of jaguars in the Atascosa Mountain complex (Ahern 2007) and potential habitat in the Patagonia/ Santa Rita Mountain complex (Gunzel 2006a, 2006b; Hardy 2007a, 2007b).
Research and management recommendations.—There is an urgent need for greater understanding of jaguars in the northern extent of their range, especially regarding cross-border connectivity. In Arizona and New Mexico there are 7 different mountain ranges with potentially suitable habitat for jaguars (Hatten et al. 2005; Menke and Hayes 2003) that have not yet been surveyed for jaguars (McCain et al. 2006). Jaguars have been recorded in 3 of these mountain ranges in the last 2 decades (Brown and López González 2001; Glenn 1996; Rabinowitz 1999; W. Glenn, Malpai Borderlands Group, pers. comm.). The current population of jaguars in the borderlands region, and particularly the United States portion of that population, appears to be dependent on large expanses of core and connective habitats for dispersal and cross-border movements within the bioregion of southern Arizona and New Mexico and eastern Sonora, Mexico. If this cohesive habitat and gene flow throughout the region are disrupted by the proposed fence along the border, it is our belief that the population of jaguars in the borderlands will be at great risk and that jaguars will not persist in the United States.
We 1st thank the Wildlife Conservation Society for their confidence and participation, which gave us momentum to research an unknown species in an unknown environment. We also received support from the Arizona Game and Fish Department, Arizona Heritage Fund, Disney Wildlife Conservation Fund, Switzer Foundation, Woodland Park Zoo, Phoenix Zoological Society, Milwaukee Zoo, Bergin County Zoo, Coronado National Forest, Redwood Sciences Laboratory, and several individual donors. We thank the Arizona-New Mexico Jaguar Conservation Team, its chairmen (T. Johnson and B. Van Pelt), and agency signatories for their collaborative support. We thank the Bell family for field accommodations and continuous technical and logistical support. M. Abbott, S. Avila, S. Bell, S. Bless, J. Brun, A. Childs, L. and M. Colvin, M. Hollister, J. McCain, S. Morse, M. Pruss, S. Pavlik, G. Paz, K. Shallcross, T. Snow, S. Stone, Sundog, M. Terrio, S. and V. Walkosak, and T. Wright all assisted with fieldwork. We thank J. Conklin, T. L. George, C. Hass, E. Heske, Z. Jones, S. Morse, R. Sweitzer, and 2 anonymous reviewers for insightful comments on the manuscript.