|
Résumé :
|
Eleven years of monitoring at Zackenberg Research Station in high-arctic Northeast Greenland documented a relatively stable population of about 21-27 pairs of Long-tailed Skuas in a 17.1 km(2) study area. The average density, 1.2-1.6 pairs per km(2), was among the highest known. Territories were well spaced out with mean distances between nests/territory centres of 450-600 m, and situated very much in the same areas between years. The extent to which the skuas laid eggs was significantly correlated with lemming abundance at snowmelt, but at least some pairs initiated breeding even in lemming-poor years. Timing of egg-laying within each year was very synchronous, but the median date of egg-laying varied about 10 days between years. This variation was correlated neither with abiotic parameters like snowmelt nor with lemming abundance, but there was a weak correlation with early spring abundance of wolf spiders. Excluding supposed re-laid clutches, the annual mean clutch size ranged between 1.0 and 1.9 and was highly correlated with lemming abundance. Hatching success varied between 0 and 94%, but was neither correlated with fox nor lemming abundance. Likewise, the number of fledged young was not correlated with any abiotic or biotic parameter tested. By quantifying lost breeding potential during the different stages of reproduction, we identified the period between territory establishment and egg-laying as the most important for variation in reproductive success across years. Breeding potential lost during this stage was closely, but non-linearly related to lemming density at snowmelt. Juvenile production could be termed 'eruptive', in that only one year produced many juveniles (1.05 per pair)-a year both with many lemmings and many foxes. Based on estimates of mortality we suggest that reproduction during the study years in this population was not sufficient to maintain a stable population size.
|