Northern Prairie Wildlife Research Center
From the present analysis, the relative paucity of negative correlations between residuals from the best-fitting models and counts of other species does not support a strong role for interspecific competition among ducks at Woodworth. The larger number of positive correlation coefficients and the general symmetry between species pairs are consistent either with an active association between those species or, more plausibly, with the hypothesis that the species were responding similarly to changes in environmental conditions that the measured variables did not adequately capture.
Wiens (1989b) observed that only rarely are community patterns the result of a single process, and this contention is consistent with results presented here. Ducks are widely recognized as responding immediately to dynamic wetland conditions by settling in areas where wetlands are favorable and deserting areas with dry basins, although philopatry also can influence where ducks settle (Johnson and Grier, 1988). The proximate habitat features that most influence the number of ducks using a particular breeding area are the number, type, and condition of wetlands in the area. This is especially true in an area such as Woodworth, where upland cover suitable for nesting dabbling ducks is abundant. The present analysis clearly shows opportunistic responses by waterfowl to wetland conditions, with such responses most pronounced in blue-winged teal, northern shoveler, northern pintail, and ruddy duck. The three dabbling ducks, but not the ruddy duck, make extensive use of seasonal wetlands (Stewart and Kantrud, 1973), i.e., those wetlands that are more variable from year to year than are semipermanent wetlands.
The lack of correlation between counts of a species in successive years is consistent with opportunistic settling in dynamically varying habitat, and it is prevalent among the dabbling ducks. In contrast, there are strong correlations in numbers between adjacent years in the diving ducks and ruddy duck, and these are species that make more use of semipermanent wetlands (Stewart and Kantrud, 1973) which are more persistent from one year to the next. In sum, the variability of a species' density clearly reflects the variability in its preferred habitat.
Philopatry also can contribute to a correlation between counts in successive years. Philopatry is an important characteristic of many species, particularly those that breed in stable habitats (e.g., Wiens, 1976). Among the waterfowl, diving ducks and possibly the ruddy duck tend to be most philopatric (Johnson and Grier, 1988). The present study found positive correlations between counts of all four of these species in successive years. Leitch and Kaminski (1985) reported similar findings for lesser scaup and canvasbacks but not for redheads and ruddy ducks. Although reproductively successful females of several dabbling-duck species will return to their previous breeding areas (Johnson and Grier, 1988; Lokemoen et al., 1990), the present study found few positive correlations between counts in successive years; these were for American wigeon and green-winged teal, and were not significant. Thus dabbling ducks seem to closely follow the habitat resources they select; there is little tracking inertia (sensu Hilden, 1964). Leitch and Kaminski (1985), in contrast, reported significant correlations between counts in successive years for five of seven dabbling ducks, including American wigeon but not greenwinged teal.
Dabbling ducks as a group tended to respond to changes in habitat conditions in a similar fashion, but not identically. American wigeon and green-winged teal covaried most closely, as reflected by the correlation between numbers of the two species, their similar declines in the Woodworth Study Area, and changes in neither species relating to measured habitat conditions on the study area. The other five dabbling-duck species were positively correlated to one another and to wetland numbers, but differed in other ways. For example, the mallard and northern shoveler did not decline during the study period, as did the other species.
Numbers of the three diving ducks, and to a lesser degree the ruddy duck, were correlated with one another. All increased during the study period, regardless of continent-wide changes in populations. Increases at Woodworth were due possibly to increased emergent cover in wetlands, which produced more favorable conditions for nesting. Only the ruddy duck and possibly the redhead seemed to respond to local wetland conditions, as indexed by the May , I pond count.
The importance of the overall, continent-wide populations in influencing numbers at Woodworth varied by species. Counts in the Woodworth Study Area most closely tracked continental populations of the ruddy duck, northern pintail, and blue-winged teal. The ruddy duck and blue-winged teal do have somewhat more southerly breeding ranges, so that Woodworth is nearer the core of their ranges. For species that settle opportunistically, sites that the birds encounter early in their northward migration will, if habitat conditions are suitable, be occupied first. Accordingly, those sites should be occupied to their i capacity and most likely will exhibit stronger relations between bird numbers and habitat conditions (Wiens, 1977).