A nutrient and energy subsidy derived from foraging in agricultural croplands in several seasons and an expanded migration and winter range have been the major influences enabling geese to increase in numbers in recent decades. Climate warming on breeding areas and expanded breeding range are likely secondary causes. Reduced harvest rate appears to be an effect rather than a cause, even if harvest rate was limiting population size before the 1970s. While there are population density-dependent effects, such as decreases in body size in LSGO (Cooch et al. 1991, Cooch et al. 1991) and GSGO (Reed and Plante 1997) and poorer body condition/higher gosling mortality in LSGO (Cooch et al. 1993, Williams et al. 1993), these adverse effects are more than offset, at the population level, by increased adult survival (Francis et al. 1992) and by "cheating" (see below).
Once the nutrient-energy subsidy was established on migration and wintering grounds, the overall landscape use by geese became inherently unstable. The geese are recipients of an increasing nutrient and energy subsidy and as such they represent an output of the agro-ecosystem at the landscape level. Expected density-dependent effects such as declining natality and increasing mortality fail to operate because of this subsidy.
Geese also "cheat" density-dependent regulation by their dispersal behavior on the breeding grounds both within seasons and between years. Increased nesting at the edges of existing colonies leads to colony expansion (MacInnes and Kerbes 1987, Reed and Chagnon 1987, Alisauskas and Boyd 1994, Kerbes 1994, Cooke et al. 1995). Dispersal of family groups after hatch to areas distant from nesting sites ensures that the birds do not forage in the most severely degraded areas (Cooch et al. 1991, Hudson Bay Project, unpublished data, R. Alisauskas and S. Slattery, unpublished data). In addition, new nesting colonies establish away from traditional sites that have been degraded (Alisauskas and Boyd 1994, Kerbes 1994). The apparent decline of the McConnell River and west Hudson Bay nesting complex can be interpreted in this context. It may be an example of how local carrying capacity was exceeded as the population grew and occupied new areas, but that at some point further dispersal took the form of emigration to a distant habitat (e.g., to the Rasmussen Basin lowlands, McLaren and McLaren 1982 and perhaps to Queen Maud Gulf, cf. Kerbes 1994). It may appear that, if the birds can disperse, the problems of habitat destruction are less urgent. However, as we discuss below, under the continued pressure of expanding populations of geese, the rate of destruction is accelerating, the total area affected is large and significant, and the habitats remaining undamaged are non-preferred and even marginal and ultimately, finite.
The Wrangel Island LSGO population decline is real and appears to be related to density- independent factors, including weather conditions on the breeding grounds and the length of their migration route. A series of late summers in the early 1970s virtually eliminated recruitment of new breeders. A long-term cooling trend is also evident for the high latitudes of the Russian Far East (Cohen et al. 1994), unlike most other LSGO breeding areas. Harvest rates have also been higher for Wrangel Island birds than others until very recently (S. Boyd, pers. comm.). Historically, harvest on the breeding grounds was also very high.
Currently, both winter subpopulations have access to extensive agricultural lands (i.e., they should both benefit from the agricultural subsidy effect). Spring migration routes differ, however, with the California-Oregon group following an interior route coincident with western Arctic LSGO and central Arctic ROGO through the grain producing areas of Saskatchewan, Alberta and western Montana. At least part of the Washington-British Columbia wintering group migrates in steps from one natural river estuary/coastal marsh to another (e.g., Stikine River) where they feed principally on Carex lyngbyei (S. Boyd, pers. comm.). Thus, they differ in spring diets and may not benefit from the spring energy subsidy.