Experimentally assessing the effect of search effort on snare detectability

Reducing threats to biodiversity is the key objective of ranger patrols in protected areas. However, efforts can be hampered by rangers' inability to detect threats, and poor understanding of threat abundance and distribution in a landscape. Snares are particularly problematic due to their cryptic nature and limited selectivity with respect to captured animals' species, sex, or age. Using an experimental approach, we investigated the effect of search effort, habitat, season, and team on rangers' detection of snares in a tropical forest landscape. We provide an effort-detection curve, and use our findings to make preliminary predictions about snare detection under different scenarios of patrol effort. Results suggest that the overall probability of a searcher detecting any given snare in a 0.25/km(2) area, assuming 60 min (or approximately 2 km) of search effort is 20% (95% CI +/- 15-25%), with no significant effect of season, habitat or team. Our models suggested this would increase by approximately 10% with an additional 30mins/1 km of search effort. Our preliminary predictions of the effectiveness of different patrolling scenarios show that detection opportunities are maximised at low effort levels by deploying multiple teams to a single area, but at high effort levels deploying single teams becomes more efficient. Our results suggest that snare detectability in tropical forest landscapes is likely to be low, and may not improve dramatically with increased search effort. Given this, managers need to consider whether intensive snare-removal efforts are the best use of limited resources; the answer will depend on their underlying objectives.