| Research from the last two decades has elucidated the importance of migration in the annual cycle of ducks, but many aspects of migration ecology remain poorly understood due to the difficulty of investigating movements that occur over large spatial scales, at substantial heights and at night. Weather surveillance radar (WSR) offers a unique tool for observing movements of birds aloft, but until now has been used primarily to address questions only relevant to broad taxonomic groups. Using thermal infrared imaging, portable radar, and natural history, I ground-truthed WSR echoes originating from a complex of wetlands in the central Illinois River valley to develop a technique for identifying and enumerating ducks as they emigrated from this important stopover area. With this technique, I quantified duck emigrations during 7 falls (1996, 1997, 2003, and 2005-2008). I used WSR-derived estimates of annual turnover in combination with aerial inventory estimates of duck use to estimate the average amount of time ducks spent at my study site during fall (stopover duration). The mean stopover duration estimate of 11 days (SD = 4 days) was much shorter than a historical estimate (28 days) that has been use for regional waterfowl conservation planning. I also regressed average annual stopover duration estimates against an index of annual foraging habitat quality and found a strong, positive relationship (r2 = 0.71), suggesting ducks assessed local habitat conditions and adjusted time spent at the site. Weather influences the timing of migration in many avian taxa, but this relationship is poorly understood for ducks. An evaluation of competing models including 15 years of data indicated following winds aloft, no precipitation, less cloud cover, decreasing temperatures, increasing barometric pressure and date best predicted emigration (R2 = 0.52). Based on this model, the odds of a duck emigration occurring when winds were following and precipitation was absent were 13.2 to 1.0 (95% CI 7.8--22.4). Finally, the notion that ducks rely on leading-lines of rivers for visual orientation during flight is a dominant paradigm in waterfowl science. I examined departure tracks of emigration events from my study site during 1995--2009 and found ducks had a significant SSE directional preference (152°; P < 0.05), which differed significantly in all years from the course of the Illinois River (220°; P ≤ 0.001). This pattern was markedly different than the river-oriented route described for ducks departing this site in the mid-20th century. Thus, leading lines appear to have been unimportant for orientation in the majority of duck emigration events from the major stopover area examined here. |
