Are northern bobwhites an umbrella species for open‐land birds in Ohio?

Abstract

Birds that inhabit open lands such as grasslands and shrublands are rapidly declining across North America. A common practice for multi-species management is to focus on umbrella species whose habitat requirements overlap with several other species.We evaluated whether the northern bobwhite (Colinus virginianus; bobwhite) could serve as an umbrella species for openland birds in Ohio, USA. We related landscape metrics to abundance patterns and assessed whether bobwhite occupancy positively predicts presence of open-land birds. We combined bird survey data from the second Ohio Breeding Bird Atlas (2006-2011) with land cover data from the 2011 National Land Cover Database (Homer et al. 2015) to construct single-season N-mixture models to identify landscape metrics that influence bobwhite abundance. Bobwhite abundance was positively predicted by forest cohesion, percent agriculture, percent barren, and percent grassland. Of the 34 focal species, bobwhites were a significant positive predictor for 12, and a significant negative predictor for 10. The model with only bobwhite occupancy probability as a predictor was the best supported model for only willow flycatcher (Empidonax traillii).These results suggest that bobwhite land cover type requirements are too specialized to meet the needs of broader species guilds, instead affording protection for a narrower range of individual species that share specific habitat requirements with bobwhites. Management for bobwhites may still be able to promote co-occurrence for declining species across multiple guilds by identifying locations where focused management can

Integrating multiple data sources improves prediction and inference for upland game bird occupancy models

“…The occupancy model results generally match previous literature on bobwhite and pheasant occupancy, although many previous studies measured covariates at different scales (Duren et al 2011, Jorgensen et al 2014, Rosenblatt et al 2022). Notably among detection covariates, pheasant detection probability decreased with date for upland surveys but increased with date for BBS surveys.…”

“…Bobwhite occupancy was associated especially with forest-agriculture edge density, proportion of agriculture, and first-order stream density. The first is known to be associated with bobwhite occupancy (Twedt et al 2007, Duren et al 2011, Rosenblatt et al 2022), as bobwhite are edge specialists. First-order stream density is not generally considered an indicator of suitable bobwhite habitat.…”

“…We used the R package landscapemetrics (Hesselbarth et al 2019) to calculate 4 metrics of landscape configuration: forest and early successional patch cohesion (hereafter cohesion), forest-agriculture edge density (where agriculture means specifically row crops), and forest-early-successional edge density. All 4 of these landscape configuration metrics and the land cover composition covariates were chosen because they were meaningful predictors of bobwhite occupancy in previous models of occupancy and abundance (Duren et al 2011, Rosenblatt et al 2022). We used locations of first-order streams from the Illinois Statewide Streams Application layer (Illinois Department of Natural Resources 2014) to calculate the density of first-order streams in each grid cell, as the habitat around these streams is often the only grassland available in parts of Illinois and may not be classified as grassland in the NLCD.…”

“…Bobwhite are native and pheasants are non-native to the USA, but both are economically important game species. Bobwhites are associated with heterogeneous landscapes and edges, and tend to use woody cover more than pheasants do (Twedt et, Wilson, and Keister 2007, Duren et al 2011, Rosenblatt et al 2022). Pheasants are primarily associated with open grassland habitats (Clark, Schmitz, and Bogenschutz 1999, Jorgensen et al 2014, Kauth 2020), although pheasants may use wetlands and wooded areas for shelter from severe winter weather (Gabbert et al 1999).…”

“…Specifically, we fit models using 6 combinations of datasets (all 3 [integrated], upland, BBS, BBS and upland, BBS and eBird, upland and eBird) and compared these combinations of datasets in terms of predictive ability, parameter inference, multimodel inference, and predicted occupancy. We expected that bobwhites would be most associated with heterogeneous landscapes, which in Illinois would likely be reflected by a mixed landscape of forest and agriculture (Veech 2006, Duren et al 2011, Rosenblatt et al 2022). We expected that pheasants would be most associated with grasslands, row-crop agriculture, and first-order streams, which are often surrounded by grasslands (Jorgensen et al 2014, Kauth 2020).…”