Conclusions and discussion

Investment in reproduction in a long-lived ani-mal represents a trade off between the availabil-ity of current resources, the cost of the reproduc-tion attempt and the probability of surviving to breed again in a future year. It seems reasonable to assume that female condition determines the level of effort invested in reproduction, up to the point where the effort threatens her own future survival. In terms of initial investment, it appears that given the relatively long period of pre-nest-ing feedpre-nest-ing in Greenland, clutch size decisions may be made by Whitefronts on the breeding

ar-eas (see Raveling 1978, Ganter & Cooke 1996) based on their own internal condition and the prevailing environmental conditions. This would involve assessment by the females of their ability to meet the nutritional demands of laying differ-ing numbers of eggs and incubatdiffer-ing the clutch given available stores and the supplement possi-ble from exogenous food sources. Evidence is ac-cumulating to suggest that exogenous sources supply much, or perhaps all of the fat needed for egg formation (Choinière & Gauthier 1995, Ganter

& Cooke 1996, Meijer & Drent 1999). Hence, ac-cess to adequate food resources prior to first egg date may have a considerable impact on the abil-ity of a bird to reproduce successfully.

We know little about female condition and its po-tential to affect reproductive success (before, dur-ing and after nestdur-ing), so the accumulation of knowledge relating to this parameter remains a priority. In particular, following body mass changes in particular individuals during the pe-riod from first arrival in west Greenland through to the end of incubation would be highly desir-able. Tracking changes in body mass by capture and the use of balances under nests offers the opportunity to assess and contrast the potential of different individuals to successfully invest stores and reserves in reproductive attempts.

Similarly, it is of great interest to understand more about how brood females recoup stores and re-serves exploited during the laying and incuba-tion period, a process about which we know noth-ing at present. Clearly behavioural adaptations (i.e. mechanisms resolving the conflict between self-maintenance and investment in brood pro-tection) and dietary selection are both potentially involved, but could differ between individuals.

Despite the recent expansion in total population size, the absolute numbers of successfully breed-ing pairs returnbreed-ing with young to the two major wintering sites combined have been more or less constant, suggesting some density-dependent mechanism is operating on the breeding grounds which restricts recruitment. Amongst known age marked individuals at Wexford, the probability of recruitment has declined over time and the mean age of first breeding has increased from c.3 prior to 1988 to c.5 years of age in subsequent years.

Although difficult to measure in an objective way, the extent of breeding habitat available

through-out the summer range does not appear limiting.

Nevertheless, the extent of habitat available in spring for pre-nesting feeding as well as later in the summer, are likely to vary with weather con-ditions, especially in the north. The Wexford geese breed mainly in the north of the breeding range, and the recent declines in fecundity of birds win-tering at that site seem likely to be the result of conditions these birds encounter on their pre-breeding and nesting areas. Their migration to west Greenland differs little in distance or route from the Scottish wintering element of the popu-lation that breed further south. They could expe-rience less access to energy-rich foods (such as barley and potatoes) in western than in southern Iceland, which could enable greater energy stores to be accumulated by predominantly Scottish birds staging in these Iceland lowlands (MS4, MS19). However, if it is exogenous energy derived on the breeding areas that represents a major de-terminant of clutch size or quality, early arrival to staging areas in southern Iceland (MS19) would give these birds an advantage over geese breed-ing further north. The latter would not only com-pete with local breeders in the staging areas of central west Greenland but also then migrate northwards within Greenland with a high prob-ability of encountering severe weather conditions on arrival at ultimate nesting grounds. As goose densities have increased in recent years, it may be that all potentially breeding White-fronted Geese are encountering more competition for lim-ited resources in spring, but the birds still need-ing to continue north face increased competition from non-breeders. As there is no further habitat to the north of the current range into which to expand, it might therefore be expected that geese breeding in the north of the range show greater density-dependent effects on the summering ar-eas than those nesting further south. There is some evidence that this is the case; on Islay, the production of young per potentially breeding fe-male has not declined significantly, the decline in successful breeding being compensated for to some extent by increases in mean brood size in very recent years. Increased mean brood size im-plies (i) adequate stores to lay large clutches, (ii) to incubate these successfully and (iii) to raise goslings to fledging. If the breeding range of the Islay-wintering birds has not changed, it is un-likely that their increased brood sizes have been brought about by change in habitat. Since their breeding area (mainly 66-69ºN) is the area with greatest density of colonising breeding Canada

Geese, it might be expected that inter-specific in-teractions in this area would reduce reproductive output.

It may have been the case in the 1970s (prior to the period of population expansion by restriction of winter hunting) the northern breeders had an advantage over southern breeders. Staging fur-ther south on the summer grounds gave the op-portunity to accumulate nutrient and energy stores remote from breeding areas, but still time arrival to nesting areas to optimise food availabil-ity there. This seems to be reflected in greater pro-ductivity amongst Wexford compared to Islay birds at that time (Figures 6.2 and 6.4). Subse-quently, conditions of greater population size have increased local feeding densities in spring and substantially increased moult migrant non-breeder numbers using northern areas to regrow flight feathers. These changes in local density at key stages of the life cycle could potentially have turned the strategic advantage into an increasing disadvantage, especially at a time when a series of late springs has constrained the overall avail-ability of early season food.

It seems likely that breeding habitat has not changed in extent, but that the quantity (and

pos-sibly the quality) of the resources available to fe-males arriving in west Greenland have increased.

Global climate models suggest there will be a short term and moderate warming of the central west Greenland coastal strip, whilst summer tem-peratures further north will be expected to fall.

While increases in total population size may re-duce overall access to finite food resources through competition, best quality individuals able to de-fend rich spring feeding areas could rapidly gain condition to invest in a clutch to be laid locally.

Birds breeding further north face increased feed-ing competition, and poorer summer conditions later in the season on their own pre-nesting and breeding areas further north. Consequently, Wex-ford-wintering geese may be undergoing the very declines in fecundity predicted on the basis of cli-mate change by Zöckler & Lysenko (2000), whilst Islay-wintering birds enjoy the positive benefits of this change. It would be interesting to analyse the historical meteorological archive to determine whether the difference in fecundity of these two elements of the population can be related to weather patterns in the north and south of the range. If not, there seem grounds for assessing in more detail the alternative explanations for these differences in breeding success in different ele-ments of the population.

[Blank page]