Male greater sage-grouse detectability on leks

It is unlikely all male sage-grouse are detected during lek counts, which could complicate the use of lek counts as an index to population abundance. Understanding factors that influence detection probabilities will allow managers to more accurately estimate the number of males present on leks. We fitted 410 males with global positioning system and very high frequency transmitters, and uniquely identifiable leg- bands over 4 years in Carbon County, Wyoming. We counted male sage-grouse using commonly used lek- count protocols and evaluated variables associated with our ability to detect marked males using sightability surveys on 22 leks. We evaluated detection probabilities of male sage-grouse based on factors related to bird characteristics such as age or posture, lek and group size, lek characteristics such as vegetation cover or aspect, light conditions, weather, and observer. We then applied the detection probabilities to more accurately estimate male counts on leks. Detection probabilities were generally high (x̄ = 0.87) but varied among leks from 0.77 to 0.93. Male sage-grouse detection declined with increasing sagebrush height and bare ground and increased with more snow cover. Detection probabilities were also lower when observers counted from a higher elevation than the lek. Our sightability models predicted detection well and can be used to accurately estimate male abundance on leks from lek counts, which is especially useful where accurate abundance estimates are required or inference about population status is based on only 1 count. Further, it is important to consider lek attendance as a component of counts on leks because it affects availability of male sage-grouse for detection during lek counts. Detection can be maximized by conducting lek counts from 30 minutes before sunrise to 30 minutes after sunrise, although current protocols recommend lek counts can be performed up to 1 hour after sunrise. Detection can also be maximized by conducting lek counts ≥2 days after snowfall, which maximizes attendance and detection.