Using occupancy analysis to validate the use of footprint tunnels as a method for monitoring the hedgehog Erinaceus europaeus

1. What they did

Many mammals are hard to see, so they are often surveyed by their signs instead. But a sign survey can miss an animal that is really there, and that can lead to wrong conclusions. Hedgehogs are a species of conservation concern in the United Kingdom (UK), and their numbers in Britain may be falling fast. The authors tested whether footprint tunnels, run by volunteers, could reliably show whether hedgehogs were present at a site.

• The tunnels. Each tunnel was a corrugated plastic tunnel, 1200 mm long, with sausage bait in the middle and an ink of vegetable oil and carbon powder that left footprints on paper.
• The surveys. Members of local mammal groups and university students surveyed 111 sites between April 2011 and September 2013, mostly in southern England. Sites (such as farms and golf courses) were chosen without knowing whether hedgehogs lived there. At each site, 10 tunnels were placed along hedgerows and fences, which hedgehogs often follow, at least 100 m apart. Tunnels were checked on five mornings in a row, and all footprint papers were sent to the authors to check.
• The analysis. Each night counted as a repeat survey. Occupancy analysis used these repeats to estimate how often hedgehogs were missed where present, and how many nights are needed to be confident they are truly absent. The authors also tested whether season, the habitat around the site, or the presence of badgers affected hedgehog presence, and used simulations to estimate how many sites a national survey would need to detect a change.

2. What they found

• The chance of detecting hedgehogs stayed the same from night to night, so there was no sign that hedgehogs got used to (or avoided) the tunnels over time.
• Hedgehog footprints were found at 38.7% of sites. The estimated true occupancy was almost the same (39.2%), so very few occupied sites were missed.
• On any one night, the chance of detecting hedgehogs at an occupied site was 0.593 (all years together).
• Across years, up to 3.6 nights were needed to be 95% confident that no footprints meant no hedgehogs. Four nights were enough.
• Season (spring, summer, autumn) and habitat did not improve the models.
• Hedgehogs were about twice as likely to be present at sites without badgers (occupancy 0.506) as at sites with badgers (0.247), although the badger model was only slightly better supported than a model with no covariates.
• With fewer than 1000 sites (typical for UK volunteer surveys), the method could detect a 25% change in occupancy. For example, with four nights per site, 641 sites would detect a 25% change with a power of 0.95. Detecting a 10% change would need thousands of sites.

3. What they concluded

• The footprint tunnel method reliably detects hedgehogs where they are present, and is suitable for monitoring changes in their distribution.
• Tunnels work in many habitats and at the scale of single farms. They suit volunteers: they are placed in known spots, so nobody has to search for scarce field signs; they are checked by day; and the footprints are kept for checking.
• Detection did not vary between April and September, and survey time could be cut from five nights to four. Using volunteers is likely to cut costs and increase statistical power.
• Occupancy was similar to or lower than in other studies, suggesting hedgehogs are now more patchily distributed than they were in the past.
• More broadly, occupancy analysis is a powerful way to validate indirect survey methods for mammals.

4. How tracking was or was not used

Tracking was used as a sign survey: hedgehog presence was recorded from footprints left on inked paper in baited tunnels. Volunteers collected the papers and the authors checked the identifications. Footprints of other animals were also recorded, but the paper does not say which species, how hedgehog prints were told apart from them, or whether any papers were reclassified. Badger presence was judged partly from field signs, along with spotlight counts and conversations with landowners.

5. Tracker or observer certification

Not reported. Surveys were carried out by members of local mammal groups and university students supervised by the authors. No training or skill test is described; all footprint papers were returned to the authors for verification, but the paper does not describe how this was done.

6. Other technologies, skills or innovations

• Footprint tunnels designed by Ian Stephens: corrugated plastic, 1200 mm long, baited with hotdog sausages, with an ink of vegetable oil and carbon powder.
• Occupancy modelling in PRESENCE software and simulation-based power analysis in R.
• Land cover mapping (Land Cover 2007) to describe habitat within 500 m of each site.

7. Implications for conservation (from the paper)

• Hedgehog populations in Britain may be declining rapidly, so a reliable monitoring method is urgently needed. This method avoids some problems of other methods: low detection rates, non-random choice of sites, and bias because hedgehogs avoid roads.
• With volunteer sample sizes, a national survey could detect changes in hedgehog occupancy of about 25%.
• The method can show the effects of management at the scale of single farms.