Northern bobwhite brood ecology and population dynamics in southwest Missouri

Abstract

Northern bobwhite (Colinus virginianus) have experienced widespread, long-term population declines. Bobwhite populations in Missouri have declined 2.7% annually between 1966 and 2015, and 3.5% annually between 2005 and 2015, for a cumulative loss of 80% since 1966. Maintaining healthy populations requires a better understanding of how population growth and vital rates across seasons are affected by habitat and land management practices. We examined bobwhite brood ecology and population dynamics in southwest Missouri. My dissertation objectives were to: (1) Examine the influence of cover type, management, and weather on juvenile bobwhite body condition; (2) Estimate bobwhite juvenile survival from hatch as a function of herbaceous cover, habitat management practices, woody vegetation structure, and landscape patterns; (3) Determine population-level bobwhite brood resource selection patterns for managed grasslands and agricultural habitat; and (4) Estimate site-specific bobwhite fecundity, seasonal age- and sex-specific survival, and the relative contributions of these vital rates to trends in abundance. Among our five study sites, Wade and June Shelton Memorial Conservation Area, Stony Point Prairie Conservation Area, and Wah'Kon-Tah Prairie are native grasslands ranging in size from 320 to 3030 acres. These sites were considered extensively managed, with fire, grazing, mowing and haying implemented to maintain continuous tracts of native grasslands. By contrast, Robert E. Talbot and Shawnee Trail conservation areas (3635–4361 acres) are intensively managed sites. In addition to grassland habitats, both of these areas incorporated small units of agriculture, food plots, and woody vegetation strips for wildlife use. We used radio telemetry and brood capture data to evaluate the influence of habitat management on juvenile body condition, survival, and brood habitat selection patterns. Brood attending adults were tracked daily from nest hatch to brood capture at approximately 3-weeks old. On capture, juveniles were fitted with transmitters and observed through the life of the bird or the life of the transmitter. Habitat was characterized using maps of herbaceous cover (native- mixed- and cool-season grasslands, and agricultural cover), management activities (prescribed burning, conservation grazing, mowing/haying), and woody vegetation structure (shrub and tree cover). To evaluate environmental influencers on bobwhite juvenile body condition, we captured 216 individuals from 33 broods >16-days old across our study sites in 2017 and 2018. We used the residuals from a linear regression of tarsus length and body mass as an index of body condition in a generalized linear mixed model evaluating effects of season, weather, and habitat. We found some support for improved body condition early in the breeding season, under warmer average temperatures, and in native grasslands that were burned and grazed in the previous two years. We estimated bobwhite juvenile survival based on observations of 705 individuals from 75 broods for a total of 14,904 exposure days. This included 493 individuals observed from hatch to capture and 212 radio-tagged individuals. We estimated 28.6% 114-day period survival for juvenile bobwhite 2016–2018. Survival was highest on native grasslands burned and grazed at least once in the previous two years and survival increased with local shrub cover. Local agricultural cover also improved survival, however landscape interactions suggest survival is low in small units of cultivated crops surrounded by grasslands or woody vegetation. We evaluated patterns in brood resource selection and the effects of those choices on brood survival using integrated step selection analysis. We quantified resource selection behavior of 101 bobwhite broods by comparing each of 2,788 chosen daily steps to 10 random available steps not taken. Consistent with our body condition and survival results, native grasslands that were burned and grazed at least once in the previous two years had the highest relative probability of use. Selection for idle native grasslands increased with age. We also found selection influenced brood success; broods that succeeded were more likely to choose available habitats with more shrub cover, while failed broods avoided available habitats with more shrub cover. Successful broods also selected areas farther from trees than failed broods. To quantify full annual cycle population dynamics in an integrated population model, we tracked 766 juveniles and 618 adults during the breeding season, we tracked 772 juveniles and 349 adults during the non-breeding season, we monitored success of 276 nests incubated among 576 adults, and conducted whistle counts in May 2016–2018. Talbot Conservation Area had the lowest population growth rate ([lambda]TAL = 0.31, 95% CRI: 0.03, 0.65), due to low fecundity and summer adult survival. Shelton Memorial Conservation Area had low estimated breeding season juvenile survival, non-breeding season survival, and the second lowest population growth rate ([lambda]SLT = 0.37, 95% CRI: 0.00, 0.74), possibly due to its small area (320 acres) and low bobwhite density. Stony Point Prairie had high fecundity and adult breeding season survival probabilities, but the lowest non-breeding season survival among all sites, which depressed population growth ([lambda]STP = 0.51 95% CRI: 0.08, 0.85). Extensive native grasslands had consistently higher fecundity and adult summer survival probabilities than intensively managed conservation areas. Wah'Kon-Tah Prairie, our largest extensively managed site, had the highest population growth rate ([lambda]WKT = 0.55, 95% CRI: 0.13, 0.94). We found declining population trends on all conservation areas. Non-breeding season survival from 1 November to 30 April was most strongly correlated with abundance, followed by juvenile breeding season survival. Greater, less variable non-breeding season survival is required to achieve population stability. Further evaluation of intra-annual environmental influencers and the relationship between habitat area and population viability would better inform bobwhite management approaches. Important next steps in this research include examining finer temporal scales that separately evaluate breeding, winter, and transitional periods in the annual cycle. Additionally, effects of weather and climate on population viability will be important considerations as Missouri experiences warmer temperatures and potentially extreme changes in precipitation patterns. Finally, this study generated valuable data that can be used to simulate regional population trends and projections across the surrounding landscape.