Landscape fragmentation and connectivity for carnivores in the Upper San Pedro Basin

1. What they did

Between the Huachuca Mountains and the San Pedro River in southeastern Arizona, fast housing growth around Sierra Vista threatens to cut off the routes animals may use to move between the mountains and the river. The area is home to around 15 carnivore species, several at the northern edge of their range. Following animals with radio collars is expensive, so the author used track and sign counts, which cost much less, to answer two questions: which carnivores are present and what parts of the landscape they use for travel, and where corridors between the mountains and the river might still be possible.

• The tracking study. Twenty routes along dirt roads, firebreaks, trails and washes, ten on Fort Huachuca and ten in the San Pedro Riparian National Conservation Area (NCA), were each walked at least three times between October 1998 and January 2000. Routes were chosen for good tracking surfaces, not at random. Surveys took place at least two weeks after rain, except three after fresh snow. Every sign (tracks, scat, dens, carcasses, scrapes, diggings) was identified to species, with field cards built from track guides and from feet of live-trapped and road-killed animals, and its position taken with a Global Positioning System (GPS) receiver. A new record was made only if the sign was over 1 km from the last one of that species, or in a different habitat. The author noted topography, vegetation, distance to water and washes, canopy cover or path width, where on the path the sign lay, and whether the animal was crossing or following the path.
• Telling individuals apart. Puma and black bear tracks were measured (rear plantar pad width for pumas, front pad width for bears) to estimate how many individuals were present.
• The corridor model. Using a Geographic Information System (GIS), the author scored twelve washes that run from Huachuca woodland to within 1 km of the river. Each wash got an “impedance” score: its length plus weighted counts of roads and housing near it. The focus was on pumas and black bears (large areas needed) and coatis (woodland animals with limited dispersal).

2. What they found

What was detected

• 65 surveys covered 386 km and recorded 601 sign units of 13 species (hooded and striped skunks counted together). No sign of jaguar, ocelot or long-tailed weasel was found.
• Routes averaged 7.9 species, up to 11. Mephitis skunks were found on all 20 routes, bobcats on 19, and coyotes, gray foxes and hog-nosed skunks on 18 each. Black bears were the only species found at just one site (Fort Huachuca).
• Longer routes and more surveys gave more species. New species added per survey fell to almost zero by the fourth survey.
• Carnivore diversity was highest on woodland routes (0.88 species per km), lowest on scrub and grassland routes (0.46) and in between on riparian routes (0.52). It did not differ with housing nearby or distance to the Mexican border.

Habitat and travel

• Badgers were found more in grassland, coyotes in grassland and scrub, gray foxes and hog-nosed skunks in scrub and woodland, coatis in woodland and raccoons in riparian habitat than expected. Bobcats, kit foxes, pumas and Mephitis skunks used vegetation types in proportion to what was available.
• Pumas and raccoons left more sign in washes; coatis and gray foxes more on ridges and side hills.
• For every species with enough data, about half of the sign was within 10 m of a wash, sign was found mostly under less than 25% canopy cover, and almost half was on paths 3 m wide or narrower.
• Sign tended to be at the side of wider paths, and on wider paths animals were more likely to be crossing than following.
• Most species left more sign on washes and trails than on roads, compared with what was available. Coyote, coati, gray fox and kit fox sign appeared in proportion. Hog-nosed skunks used surfaces in proportion once busy roads were removed, which may mean they avoid traffic, or that their sign is harder to see there.

Species notes

• Pumas were common (62 track sets, 5 scats, 1 scrape) on 14 routes, but scarce on the southernmost river routes. Pad measurements suggested at least four individuals, likely including an adult male and an adult female using the river regularly and at least one subadult passing through. One puma had a deformed toe and could be recognized from a single track.
• Black bears were rare (6 track sets, 1 scat) and only in Fort Huachuca woodland; measurements suggested at least four bears.
• Ringtails and spotted skunks left very little sign; the author believes this reflects poor detectability rather than rarity.
• Coati groups seldom follow roads, while adult males often do, so lone males were more likely to be detected.
• The Palominas route, near heavy night-time foot traffic by people crossing the border, lacked bear, coati and puma sign, while the busy daytime birding area near San Pedro House had among the most species.

The corridor model

• Memorial Wash, draining the east slope of Coronado National Memorial, scored best (impedance 22.6): it is the shortest route and crosses only a dirt road and one 2-lane highway. Hunter, Carr, Miller and Ramsey washes came next, but cross State Route 92 through box culverts and run through housing.
• Woodcutter’s Wash scored worst (530.4), running through some of the densest parts of Sierra Vista and under two 4-lane highways.
• The five best washes lie mostly on private and state trust land, open to development.

3. What they concluded

• Track surveys were a very cost-effective way to monitor carnivore distribution and habitat use here, except for ringtails, spotted skunks and possibly coatis, where other methods (live-trapping, camera bait stations) may work better.
• If sign reflects habitat use, all species used roads, trails and washes, but preferred trails and washes to roads.
• The tracking could not show whether carnivores actually move between the mountains and the river. The river appears to give year-round habitat to most carnivores, and may act as a corridor or oasis for black bears and coatis.
• The model ranks washes objectively, but the weights were arbitrary and need field testing, with cover added. The author suggests impedance scores might one day be linked to species thresholds (for example, up to 300 for pumas, 200 for black bears).
• The author proposes three ecological linkages, Memorial Wash (Palominas), Hunter Wash and the Babocomari River, each with a 200 m protected core along the wash and a 1000 m buffer, with Memorial Wash as the top priority.

4. How tracking was or was not used

Tracking was the main field method. Along fixed routes, the author identified tracks, scats, dens, diggings, scrapes, nose circles and carcasses to species, recorded where each was found and in what habitat, and noted whether tracks crossed or followed the path. Where possible, several tracks in a track set were measured and the most common value used, to tell individual pumas and bears apart (for pumas following Fjelline and Mansfield 1989). No trailing was reported. Three surveys were done in fresh snow.

5. Tracker or observer certification

Not reported. The author did the surveys, with field assistance from M. Tuegel; no tracking qualification or test of identification skill is described. The author mentions a 4-year study of coatis in the Huachuca Mountains. Identification relied on field cards built from published track guides and from measured feet of trapped and road-killed animals.

6. Other technologies, skills or innovations

• Garmin 45 GPS receiver with United States Geological Survey (USGS) topographic maps.
• Manifold GIS for habitat availability along routes and for the wash corridor model, with housing density mapped from county 911 addresses and aerial photos.
• A simple impedance score built in Microsoft Excel combining wash length, roads and housing.

7. Implications for conservation (from the paper)

• Rapid subdivision of ranches threatens to sever wildlife links between the San Pedro River and the Huachuca Mountains.
• The proposed linkages would keep the washes free of new housing, industry and roads, limit recreation to daytime, keep cover along the washes, and replace box culverts with overpasses where possible.
• Washes that end short of the river among houses could trap bears, pumas, bobcats or coatis in neighbourhoods, raising human-wildlife conflict.
• The area south of the border towards the San Pedro should also be protected, including reducing disturbance from border enforcement.
• More is needed on the very rare long-tailed weasel in southeastern Arizona.