Habitat selection and social structure of mottled ducks in a Texas coastal marsh

Abstract

Breeding chronology and habitat selection by mottled ducks (Anas f. fulvigula) were studied at San Bernard National Wildlife Refuge (SBNWR), TX, 1985-87. The Habitat Suitability Index (HST) model for the mottled duck also was tested. Nesting chronology was influenced by habitat conditions. Mean nest initiation dates were 25 March 1985, 4 May 1986, and 4 March 1987. Mean nest initiation dates were inversely related to October - February rainfall totals. The chronology of reproduction by mottled ducks was determined using pair status, nesting chronology and brood observations. Peak rates of nest initiation were measured beginning 20 March 1985, 1 May 1986, and 20 February 1987. Peak hatch rates occurred beginning 18 April 1985, 30 May 1986, and 20 March 1987. Most broods had fledged by 5 September 1986. These dates provided the best indication of when most mottled ducks entered each respective phase of the annual cycle. The dates established in chronology studies were used to divide the annual cycle into seasons for the study of habitat selection. In each season, mottled ducks of all social classes selected habitats which were both low in salinity and nontidal. Intermittent stream and brackish marsh quadrats were selected by mottled ducks each season. Quadrats in impounded fresh marsh also were selected during each season except nesting 1987. Structural parameters affecting habitat use were determined by cannonical discriminant function analysis. Although habitat use could be classified based on structure, the important characteristics varied among seasons and years. This variation was due to fluctuations in habitat availability. The HSI scores for hen with brood cover were correlated with habitat use by mottled ducks. The characteristics of high quality brood habitat specified by the model provided high quality habitat for other social classes of mottled ducks during most seasons. Sample quadrats did not include nesting cover; therefore, HSI and HSI for nesting cover only were not correlated with habitat use.