Plot placement when using a passive tracking index to simultaneously monitor multiple species of animals

1. What they did

Tracking plots are a cheap, simple way to follow changes in animal numbers, and earlier work had used a passive tracking index (PTI), with no bait or lure, on low-use dirt roads to monitor coyotes, bobcats, deer and dingoes. But roads are not a fair sample of the habitat they pass through, so the researchers asked whether plots placed off the road, in natural habitat, would work better, and how widely the index could be used for several mammal species at once.

On a ranch in south Texas in February and March 1999, they placed 40 plots at random on lightly used dirt roads, at least 0.8 km apart. Each plot was 1.5 m long and spanned the road (about 3 m), raked and smoothed to take clear tracks. Beside each road plot they placed a matching plot at least 30 m away on natural bare ground, on a randomly chosen side of the road. Every 24 hours they counted the sets of tracks of each species crossing each plot, then erased the tracks and smoothed the soil again.

The timing matched a separate study on the same ranch that tested devices for capturing coyotes. The plots were read for 4 days before those trials and 3 days after (rain ended the fourth day). Between the two survey periods, 90 coyotes were captured and removed, along with a few raccoons, javelina, deer, swine and rabbits caught by accident.

2. What they found

• The tracks of six mammals were easy to tell apart and count: coyotes, white-tailed deer, feral swine, javelina, raccoons and rabbits. Superimposed tracks caused no problems. Rodent tracks were too dense to count individual crossings, and bird tracks (roadrunners and quail) were not separated by species, so no index was calculated for them.
• Fewer than 1% of plot readings were lost to vehicles or livestock.
• Road versus off-road: road plots gave much higher index values than off-road plots for coyotes, feral swine, raccoons and rabbits. For deer and javelina there was no clear difference. Before the trials, for example, the coyote index was 0.790 on the road and 0.069 off it, while deer gave 0.755 and 0.819.
• Before versus after the coyote removals: index values fell sharply for every species, on and off the road. The one exception was coyotes on off-road plots, where the index was already low and rose slightly afterwards.
• After the trials, road plots still gave higher values than off-road plots for every species except deer, for which the off-road index was now higher than the road index.
• No raccoon tracks were found on off-road plots after the trials.
• Using only whether a plot was tracked or not (rather than counting track sets) gave a similar picture in some cases but lost sensitivity: after the trials, the on-road and off-road results for coyotes, swine and rabbits could no longer be told apart.

3. What they concluded

• The PTI picks up changes in population, or changes in activity after an event, for a range of species, and works best on lightly used dirt roads.
• In this dense, thorny brush most larger mammals seemed to use the roads as travel ways. An index best reflects the animals in an area when plots are placed on their predictable travel routes, because of, not despite, their preference for those routes.
• For coyotes and raccoons the drop in the index mainly reflected animals removed. For the species hunted for sport (deer, swine and javelina), few were removed, and the drop more likely reflected a learned response to the daily vehicles and shooting during the trials. Deer are hunted from the roads on this ranch, and may have learned to avoid roads when traffic comes with shooting. The small rise in coyotes off the road may show surviving coyotes avoiding the removal activity along the roads.
• The index should not be used to compare species, or the same species in very different habitats or seasons, because road use may differ.
• Indices show changes, not numbers of animals. Where actual numbers are essential, a proper density estimate is needed. Monitoring an untreated nearby site can help interpret results.
• The index could be useful for deer, feral swine and raccoons. The authors also saw human tracks on the plots overnight and suggest the method could index illegal border crossings from Mexico.

4. How tracking was or was not used

Tracking was the whole method. Plots were raked and smoothed so they recorded clear footprints, and each day the researchers identified the species and counted the number of separate track sets (crossings) on each plot. The paper describes the tracks of the six species as readily distinguished with careful inspection. No trailing was done.

5. Tracker or observer certification

Not reported. The paper does not say who read the plots or how their identification skill was checked; J. Guthrie is thanked for help in the field. The authors report that identification and counting were straightforward for the six species indexed, but not for rodents or birds.

6. Other technologies, skills or innovations

• Plot locations recorded with a Global Positioning System (GPS) unit and mapped with ArcView and Atlas Geographic Information System (GIS) software.
• A variance estimate for the index that makes no assumption that plots or days are independent and allows unequal numbers of days and plots, so lost plot readings and an unequal number of survey days caused no problem.

7. Implications for conservation (from the paper)

• Landowners could use the index to follow deer numbers when deciding how many hunters to allow each season.
• The authors were testing the index in a tropical habitat to track feral swine and the success of removal programs, and on beaches to track raccoons, which prey on sea turtle nests.
• Raccoon numbers could be monitored to plan the delivery of oral rabies vaccine baits.
• When a management action may affect animals, both their numbers and their behaviour can change, so index results need careful interpretation.