Comparing and combining use of autonomous recording units and traditional counts to monitor Northern Bobwhite

Abstract

We examined the use of autonomous recording units for monitoring Northern Bobwhite (Colinus virginianus) in South Carolina and compared results with those of traditional point count surveys conducted simultaneously at overlapping points. We assessed seasonal patterns and quail encounter rates for traditional and recorded surveys and used random forest modeling to determine which location and survey-based variables are most important. We found both survey methods have similar encounter rates, but seasonal occupancy rates are significantly higher when the more extensive automated recording data are used. Both survey methods indicate that location-based variables are most important to encounter rate, and both adequately account for survey detectability, but the use of recordings reduces survey bias. The autonomous recording method also permits an increase in survey and season length without increasing the labor necessary for monitoring. Both survey methods indicate a peak encounter rate in June consistent with the current protocols and with similar studies in nearby regions. The use of recordings, however, allows for extensive tracking of seasonal patterns and would be beneficial for long-term monitoring. Overall, traditional methods are more conducive to abundance surveys, whereas recordings are more appropriate for occupancy or encounter rate studies. We suggest a combination of both point count and autonomous recording methods as a feasible way to expand and improve monitoring of bobwhite populations.