The Efficacy of Video Cameras to Account for Northern Bobwhites Flushed, but Undetected During Aerial Surveys

Abstract

Over the past 20 years, conventional distance sampling from a helicopter platform has been used to estimate northern bobwhite ( Colinus virginianus ; hereafter, bobwhite) density over large areas of rangeland vegetation. However, it has been speculated that aerial surveys can complicate the ability to meet the distance sampling assumption of detecting 100% of the target objects on the transect line due to the restricted observer view from the helicopter. We attempted to use video cameras to determine whether missed detections occurred and whether digital methods could improve the precision of bobwhite density estimates. Our objectives were to 1) determine whether video cameras are a viable option to detect if coveys are flushing behind the helicopter and missed by observers, 2) determine whether coveys are flushing underneath the helicopter and missed by observers, and 3) explore the use of video cameras in a mark-recapture distance sampling (MRDS) framework. We recorded video while traversing line-transects with a helicopter during 4 distance-sampling surveys across 2 ranches in South Texas, USA. For objective 1, we reviewed footage from cameras with a backward-facing view and detected only 1 pair of bobwhites (0.001% of 889 coveys detected) that flushed on video footage recorded during the surveys but were unnoticed by observers. These results indicated that when coveys flushed, they rarely flushed behind the helicopter, and the helicopter flew at what seemed to be the proper speed and altitude to detect late flushes. For objective 2, we reviewed footage from a helicopter-mounted camera that was recorded within a swath underneath the helicopter’s center. We recorded 22 flushes within the swath, none of which was missed by the observers in the helicopter; as a result, we could not complete an MRDS analysis in Program Distance. This study improved confidence in fulfilling the assumptions of distance sampling and resulting density estimates but was limited to flushing birds only.