Distance sampling survey effort to improve density estimates of northern bobwhite

Abstract

AbstractDistance sampling from aerial platforms can provide researchers with precise and efficient density estimates for wildlife populations, particularly over large areas. The intensity of the distance sampling survey depends on the survey effort. Effort can be referred to as coverage, where transect spacing is determined by the observer's sightability distance of the object or animal. Managers and researchers in Texas often use helicopters to survey and collect data for northern bobwhite (Colinus virginianus) density estimates. For bobwhites, 100% coverage represents transect spacing of 200 m, assuming observers can cover out to 100 m on either side of the transects. Often surveys are conducted at 50% coverage or 400 m spacing to reduce cost. Still, the implications of lowering survey effort on the precision of density estimates are unknown, particularly in coverage prescriptions. We flew 2,641 km of line transects each December from 2014 to 2017 and detected 2,333 bobwhite coveys across 7,648 ha of rangeland on the San Antonio Viejo Ranch in Jim Hogg County, TX, USA. We conducted line‐transect distance sampling from a helicopter platform at 50 and 100% coverage and simulated results at coverage <50% using empirical data. Based on the results of simulated surveys at <100% coverage, surveying for bobwhites using helicopters with less than 50% coverage results in variable density estimates (wide 95% confidence intervals) and low precision (coefficient of variation >20%). Empirical surveys conducted at 50 and 100% coverage show little variation in density estimates between the 2 levels of coverage. However, when broken into substrata (areas less than ~1,000 ha), conducting surveys at 50% coverage resulted in <60 detections, which led to low precision in density estimates (coefficient of variation >20%). Our results are based on surveys conducted at the juncture of the Coastal Sand Sheet and Tamaulipan Thorn Scrub ecoregions in south Texas and need to be evaluated in other arid and semi‐arid rangeland systems.