National Quail Symposium Proceedings


Developing an effective monitoring program for Montezuma quail (Cyrtonyx montezumae) is challenging because the technique must be practical for surveying vast, remote landscapes while accounting for the species’ low detectability. We used call-back surveys within a presence–absence framework to estimate occupancy and detection probability of Montezuma quail and used this information in conjunction with habitat data to develop an estimated probability of occurrence map for the species. We established survey points at 4 sites in western Texas (n = 20–30 points/site) and conducted 5 repeat surveys/season during June–August 2007 and 2008. We documented abiotic conditions (temperature, time of day, survey number, and year) during surveys and quantified microhabitat (% bare ground, food-plant density, vegetation height, and visual obstruction) and macrohabitat (vegetation type, elevation, aspect, and slope) at survey points. We then used an information-theoretic approach to evaluate the influence of micro- and macro-habitat on detection probability and occupancy at a local and regional scale, respectively. At a microhabitat scale, the most parsimonious model (ΔAICc R2 = 0.46) suggested detection probability was influenced primarily by year (βYear = 0.91, 95% CI = 0.24–1.57), with occupancy being influenced primarily (but minimally) by year (βYear = –59.7, 95% CI = –179.0–59.6) and vegetation-height (βVH = 67.7, 95% CI = –71.9–207.4). This model indicated that detection probability decreased from 2007 (0.40; 95% CI = 0.31–0.49) to 2008 (0.21; 95% CI = 0.14–0.32), as did occupancy (1.00 vs. 0.72, respectively), which corresponded to a transition from a relatively wet to dry year. At a macrohabitat scale, the most parsimonious model (ΔAICc R2 = 0.20) suggested occupancy was influenced by elevation (βElevation = 1.11 ± 0.56) and vegetation type (βVegetation type 2 = –3.17 ± 1.26; βVegetation type 3 = –1.20 ± 1.18), and we used these variables to construct a first-approximation, probability of occupancy map. Given our findings, presence–absence surveys may be a viable approach for monitoring Montezuma quail populations through time, and use of a probability of occupancy map can help with efficient allocation of survey points and effort. However, the viability of using a presence–absence approach to monitor Montezuma quail populations will depend on whether sampling effort can be increased sufficiently to obtain more precise estimates of occupancy. In addition, our probability of occupancy map should be regarded as a first approximation and further research should be conducted to refine the relationships.