Factors influencing detection in occupancy surveys of a threatened lagomorph

Abstract

Successful recovery of populations of rare and cryptic species requires accurate monitoring of changes in their distribution and densities, which in turn necessitates considering detection rates. Development of population monitoring protocols is needed to aid recovery of the New England cottontail (Sylvilagus transitionalis; currently the top-priority Species of Greatest Conservation Need in the northeastern United States), which lives in dense shrubby habitat and is difficult to detect. To address this need, we conducted repeated, systematic, presence-absence surveys to determine patch-specific detection probabilities and factors influencing detection of the New England cottontail. We surveyed cottontails during 2-6 visits on 30 sites with known occupancy in the northeastern United States during the winters of 2010 and 2011. For each survey visit, we determined whether cottontails were detected by the presence of fecal pellets on fresh fallen snow and subsequent species identification by genetic analysis. Detection probabilities were modeled in Program PRESENCE to explore the influence of snow condition and depth, temperature, wind, number of pellet deposition days, woody stem density, patch size, and search effort. The overall probability of detecting a New England cottontail during a single survey visit was 0.73. The most influential factor in detection was prior knowledge of site-specific cottontail activity. Snow depth <30.5 cm and the number of days without high winds following a snowfall had a positive influence on detection. Patch size had a negative effect on detection when surveys were restricted to 20 minutes. In the absence of prior knowledge, 2-3 surveys conducted with snowpack <30.5 cm and 2-4 days after a snowfall without high wind should yield reliable occupancy status with 95% confidence in detection. Incorporating our recommendations into monitoring programs will improve the accuracy of patch-specific occupancy data for New England cottontail. Our approach and findings may be applicable to monitoring other rare, cryptic, or threatened species that occupy dense habitats, especially where patch-level occupancy knowledge is required. Ó 2014 The Wildlife Society.

Spatially explicit abundance estimation of a rare habitat specialist: implications for SECR study design

“…, Brubaker et al. ). Yet, occupancy is typically an inadequate surrogate for abundance (Johnston et al.…”

“…In the case of the New England cottontail, implementing winter fecal pellet surveys is extremely labor intensive (Brubaker et al. ). Knowledge of the intensity of sampling and the number of repeat surveys required for robust estimation ensure that limited resources are put to effective use.…”

“…Sampling under ideal survey conditions (Brubaker et al. ) will therefore increase detectability, potentially enabling desired levels of precision and bias to be obtained from a less‐intensive sampling regime (e.g., single survey visit with intensive sampling).…”

“…The 50‐m buffer comprised an area beyond which a cottontail would not venture into non‐habitat (Brubaker et al. , Cheeseman ). Sites were comprised of early‐successional habitat, where dominant vegetation included juniper ( Juniperus spp.…”

Monitoring a New England Cottontail Reintroduction with Noninvasive Genetic Sampling

“…We recorded spatial information for search transects and pellets (Garmin GPSMAP 64 s, Olathe, KS, USA). Pellet surveys consisted of 1 survey during December–March in winter 2013/2014 (hereafter winter 2014) and 2 independent surveys annually in 2014/2015 through 2017/2018 (hereafter winter 2015 through 2018), under optimal survey conditions to detect New England cottontails (2–5 days after snowfall, with snow depth <30.5 cm when possible and wind speed <40 km/hour; Brubaker et al 2014). We collected fecal pellets with gloves to prevent contamination in the field and stored them in 15‐mL conical tubes at −20° C until DNA extraction.…”