Monitoring changes in feral swine abundance and spatial distribution

1. What they did

Feral swine (also called feral hogs or feral pigs) were brought to North America and now damage habitats, prey on and compete with native species, harm agriculture and can carry diseases that spread to livestock or people. Managers need a simple way to tell whether swine numbers are going up or down. Counting animals exactly is costly and often unreliable, so the authors tested simpler indices based on tracks and signs.

The study took place in Jonathan Dickinson State Park, Florida, United States, where swine are a long-standing threat to rare habitats and species. The team:

• Set up tracking plots. They placed 22 plots at random on the park’s quiet, single-lane sandy dirt roads. Each plot was 1.5 m long and spanned the road (about 3 m). They raked and smoothed bare sand and kept markings to a minimum so the swine could not tell the plots from the rest of the road. Each plot’s location was recorded with the Global Positioning System (GPS).
• Counted track sets. On two consecutive days, they counted the number of swine track sets crossing each plot (each crossing counted as one visit), then erased the tracks and smoothed the surface again.
• Repeated this three times. In January 2000 before a control effort, in May 2000 after 25 swine had been removed, and in August 2000 to look for animals moving back in.
• Calculated three indices. A passive tracking index (PTI), the average number of crossings per plot per day; a new “index of pervasiveness”, which uses distances between plots with tracks to describe how swine activity is spread over the area; and a fresh damage index, the number of fresh rooting sites per km along their route between plots.

2. What they found

Swine numbers

• Between January and May, 25 swine were removed: 22 under a removal contract in April and three by park staff.
• The PTI fell by 81%, from 0.581 in January to 0.115 in May, a clear statistical difference.
• Fresh rooting along the route fell by 89%, from 1.75 to 0.19 sites per km.
• In August, after park staff removed three more swine, the PTI fell further to 0.070, and no fresh rooting was seen. There was no sign of the population building up again.

Where the swine were active

• Before control, swine signs along the route seemed concentrated in several areas; after control, activity was limited to one corner. The index of pervasiveness rose from 3.1 in January to 7.3 in May.
• In August, the index was 3.6, a pattern closer to January’s than May’s. In their summary, the authors say this may show pressure from swine moving back in.

Other animals

• Tracks of raccoons, white-tailed deer, armadillos and canids were also seen on the plots. Only raccoons were common enough to index. Their PTI did not change between January and May and rose in August, suggesting that the control effort did not harm other wildlife and that the drop for swine was due to the removals.
• Coyote and dog tracks could not be told apart with confidence, and dogs often walked the roads with hikers.

The plots themselves

• No trail of swine tracks turned towards or away from the plots, so the plots did not seem to attract or repel the swine. Tracks crossed them as part of the animals’ normal travel.
• Plot-to-plot differences were by far the largest source of variation, much larger than day-to-day differences.

3. What they concluded

• The passive tracking plots were a simple, easy way to measure changes in swine abundance and, from the same data, how their activity was spread across the area.
• Indices are often more practical than absolute population estimates. With so few swine (about 25 before control), capture-recapture methods would not have had enough data, and catch-per-effort or bait-take measures can change animal behaviour, especially in pigs, which learn quickly.
• Because the thick saw palmetto cover pushes swine to travel on the roads all year, plots on roads reflect their abundance well. Comparing index values from very different habitats could be misleading if swine use roads differently there.
• The PTI depends on finding and correctly identifying tracks and telling apart separate crossings. Tracks can be missed if plots are poorly prepared or observers are not trained, and rain, wind and traffic can erase them. Large groups crossing at once could cause problems in some places.
• With steady good weather, it is better to add more plots; if the weather may change, add more days or wait.
• The fresh damage index is a useful complement but should not be the main index, because judging “fresh” is somewhat subjective and swine may change their rooting habits with the seasons.
• The park would keep using the same plots every three months to watch for swine returning and to judge whether more control is needed.

4. How tracking was or was not used

Tracking was the core method. The team counted swine track sets on raked sand plots, identifying the species from their footprints and counting separate crossings of each plot. They also read the trails of tracks to check whether swine changed direction near the plots (none did). Along the route between plots, they looked for swine signs and judged the age of rooting: fresh rooting had moist overturned soil and living roots, while old rooting had dry soil and withered roots. Tracks of other species (raccoons, deer, armadillos and canids) were also identified.

5. Tracker or observer certification

Not reported. The paper does not say who read the tracking plots or what their experience was. The authors note that tracks may be missed if observers are not trained, and that coyote and dog tracks could not be told apart with confidence.

6. Other technologies, skills or innovations

• A new index of pervasiveness, adapted from Hopkins’ index of aggregation, that turns the GPS positions of plots with and without tracks into a single number describing how activity is spread over an area.
• A variance formula for the PTI that does not assume plots or days are independent.
• GPS to record plot locations, and Geographic Information System (GIS) software (ArcView and Atlas GIS) to calculate distances between plots.
• A simple rule for telling fresh from old rooting by the moisture of the overturned soil and the state of the roots.

7. Implications for conservation (from the paper)

• A practical way to index feral swine numbers is one of the greatest needs in feral swine management. Knowing their relative abundance and distribution helps with timing control, choosing where to control them, delivering oral vaccine baits and judging whether control has worked.
• Taking the index of pervasiveness together with the PTI can be a useful early sign of swine moving back into an area after control.
• Indices from passive tracking plots support more effective and economical management of swine damage in protected areas such as Jonathan Dickinson State Park.