Northern Prairie Wildlife Research Center
Ultimate factors are those that affect long-term survival and reproductive success. Optimal habitat for survival of the adult may differ from the optimum for reproduction; requirements for reproduction tend to be more stringent. Not only made by M. G. Anderson, F. C. Bellrose, must the bird be safe from predation, but so must the eggs and ducklings, life stages that are highly vulnerable. Food supplies must include the minimum for adult survival, as well as adequate calcium, protein, and energy for egg production (Holm and Scott 1954, Krapu 1979), and nutrients for the growth and development of young.
Proximate cues used by ducks to identify suitable breeding areas are unknown, but are thought to be indirect and related strongly to wetland features. Empirical evidence comes from studies in which wetlands were created or markedly improved, often leading to large increases in the number of breeding ducks (e.g., Hochbaum 1946, Hochbaum and Bossenmaier 1972, McKnight 1974). Although ducks typically use a variety of wetlands during the breeding period (Dzubin 1969, Dwyer et al. 1979), many of which do not persist for the entire breeding season, the conditions of wetlands in spring offer good cues to pond conditions in the subsequent summer (Leitch and Kaminski 1985). The customary measure of wetland quantity in extensive waterfowl studies is the number of wetland basins that contain water during some survey period, typically May (Crissey 1969).
We hypothesize 3 patterns of settling behavior and the ecological and species characteristics associated with each pattern: (1) homing, (2) opportunistic settling, and (3) flexible settling. It could be advantageous, however, for different age or sex classes within a species to follow different patterns.
One pattern is always to home, that is, for adults to return to the breeding area used the previous year and for yearlings to return to their natal area. Such philopatry has been termed "migrational homing" (Sowls 1955). Homing can be predicted if the temporal variability of resources is low, but would not be advantageous otherwise (Greenwood and Harvey 1982). Simply, if the birds' essentials are provided each year, there is no advantage to switching breeding areas.
The second pattern is for a bird to settle in the first site encountered along its migration path that appears to offer the requisites for survival and breeding. This pattern is akin to the "facultative migration" of Pulliam and Parker (1979) and would be appropriate if habitat conditions are unpredictable from year to year. Moreover, opportunistic settling minimizes any migration costs. Dispersal, such as caused by opportunistic settling, is an important means of adjusting to major changes in birds' habitats and is especially important in spatially heterogeneous habitats (Gauthreaux 1980,1982). One manifestation of opportunism is frequent pioneering, the movement of birds into habitats that were unavailable or not occupied during the previous year. Pioneering has been documented by anecdotes attributing increases in duck numbers in a local area as a rapid response to improved habitat and by large-scale surveys that show a redistribution of ducks from 1 year to the next when wetland conditions in part of the area change markedly.
A mixed pattern is for a bird to home to the area used the previous year, but move on (i.e., become opportunistic) if that habitat is not suitable. Drought displacement (the overflight of normal breeding areas) is included here. The pattern is similar to the "flexible homing" to wintering areas discussed by Bellrose and Crompton (1970) and can be viewed as a compromise between homing and opportunistic settling. The habitat chosen after moving on may enhance survival of the adult but reduce reproductive performance. Drought displaced birds are reported to have low reproductive success (Hansen and McKnight 1964, Murdy 1966, Calverley and Boag 1977, Derksen and Eldridge 1980), and wetlands in areas frequented by drought-displaced nonbreeding birds are typically less productive than those used by breeding birds (Murdy 1966).
In North America, most drought-displaced ducks are thought to move northward or northwestward (Hansen and McKnight 1964, Crissey 1969, Smith 1970), although Stoudt (1971) suggested that blue-winged teal may move southward. Evidence for drought displacement arises from the following 2 sources: (1) local studies in northern areas that show greater densities of ducks, and sometimes the presence of species not usually found there, in years of drought in the prairie and parkland (Hansen 1960, Hansen and McKnight 1964, Derksen and Eldridge 1980); and (2) extensive surveys that show greater proportions of ducks in northern areas than in prairie and parkland during years of severe prairie drought (Hansen and McKnight 1964, Crissey 1969, Smith 1970).
Ecologists long have recognized that many life history characteristics are associated with one another, e.g., longevity and fecundity. Efforts have been made to relate suites of characteristics to features of the habitat used by the animals, specifically contrasting highly variable (and unpredictable) environments with those more stable (and predictable) (e.g., Pianka 1974). Comparisons often have been cast in terms of r- and K-selection (MacArthur and Wilson 1967), commemorating the parameters of the logistic equation of the growth of populations subjected to high levels of density-independent (r - selected populations) or density-dependent (K-selected populations) mortality (Stearns 1977, Boyce 1984). The introduction of these parameters has brought much controversy to the discussion of covarying life history characteristics. Pending a resolution of this controversy, we will use the expressions, reselection and K-selection, to refer to populations that are influenced relatively more by factors that operate in a density-independent manner and populations more influenced by density-dependent factors, respectively.
The settling pattern of a species may be related to the stability of the habitat it selects. The theory of settling presented here suggests that opportunistic responses to habitat change will be associated with the use of variable habitats and with relatively r - selected species, and that homing will be associated with the use of stable habitats (e.g., Wiens 1976) and with relatively K-selected species. We hypothesize that a given settling pattern or selected breeding habitat is within a suite of life history characteristics that can be regarded as relatively r - selected or K-selected, while recognizing that other kinds of selection could have caused these suites to covary (Stearns 1977). If we are correct, then species of ducks could be arrayed along an axis that reflects settling pattern and habitat selection; this axis would correspond to an axis of r- and K-selection. Several authors (Patterson 1979, Bailey 1981, Vickery and Nudds 1984) have done related comparisons, mostly attempting to place duck species along a continuum of r- and K -selection.