Open AccessGround surface vs. black globe temperature in northern bobwhite resource selection
Author(s) -
Olsen Benjamin R. L.,
Fulbright Timothy E.,
Hernández Fidel,
Grahmann Eric D.,
Wester David B.,
Hehman Michael W.
Publication year - 2018
Publication title -
ecosphere
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.255
H-Index - 57
ISSN - 2150-8925
DOI - 10.1002/ecs2.2441
Subject(s) - wet bulb globe temperature , colinus , microclimate , akaike information criterion , environmental science , operative temperature , atmospheric sciences , selection (genetic algorithm) , thermal , geography , ecology , meteorology , statistics , mathematics , air temperature , biology , physics , computer science , quail , artificial intelligence
Abstract Understanding thermal ecology of endotherms is becoming increasingly important with predicted increases in temperature associated with climate change. Patterns of resource selection may depend in part on temperature thresholds outside which animals avoid warmer or cooler temperatures. During summer, northern bobwhites ( Colinus virginianus; hereafter bobwhite) in semiarid areas often exist at the limits of their thermal tolerance. We sought to determine (1) selection bounds for black globe and ground surface temperatures, (2) whether ground surface temperature is a better predictor of the thermal environment used by bobwhites than black globe temperature, and (3) how time of the day influences thermal resource use. We radio‐marked 40 bobwhites and located them two to three times a week from April to September in 2014–2016. At each location, we measured ground surface and black globe temperatures at the bird location and at a location 20 m away in a random direction. We calculated continuous selection functions to estimate bounds of thermal characteristics suitable for bobwhites. We used generalized linear models to determine times of day bobwhites select cooler microclimates. Finally, we used case‐controlled logistic regression models to generate thermal resource selection functions. We compared models that included ground surface temperature, black globe temperature, and an index created from both variables with principal component analysis and selected the best model based on Akaike's information criterion adjusted for sample size ( AIC c ) and log‐evidence ratios. Upper and lower selection bounds for black globe temperature and ground surface temperature were 24.5–42.5°C and 23.0–39.5°C, respectively. A principal component of black globe combined with ground surface temperature was a better predictor of bobwhite resource selection than either black globe temperature or ground surface temperature alone. In early afternoon (13:01–17:30 hours) during August, bobwhite locations had 13°C cooler black globe temperatures and 18°C cooler ground surface temperatures than random locations. Availability of thermally usable coverts (woody plant thickets) may be a limiting factor for bobwhites in the study area based on the findings that mean ground surface and black globe temperatures at used locations were close to the upper bounds for use in late summer during the middle of the day.