WHAT CONTROLS THE NUMBERS OF SEA-BIRDS? 121 



Wherever the ecology of these species has been intensively studied, 

 real differences in food patterns are apparent. Thus, of the three 

 skuas, the pomarine skua remains addicted throughout the breeding- 

 season primarily to the "orthodox" skua life of parasitising and robbing 

 other sea-birds for their food. The arctic skua has the most generalised 

 feeding habits: in certain parts of its range where there are no rodents 

 it lives rather like the pomarine skua, provided the pomarine does not 

 overlap with it: in areas where there are voles, mice or lemmings, 

 it eats quantities of these in the breeding season. The long-tailed skua, 

 while it does parasitise sea-birds to a slight extent in its winter range, 

 seems to have become primarily a rodent-feeder and its breeding range 

 is practically limited to those parts of the world in which rodent food 

 is abundant. 



Many overlaps occur in the gulls of Britain and north-western 

 Europe where, at the two ends of a long chain of forms which embraces 

 the whole of the northern hemisphere (see Fig. 4^, p. 41) the herring- 

 gull and lesser blackback share a breeding-range, and, at least through 

 most of the year, a feeding-range. What their original food spectra 

 were and that of the other close relation which overlaps with them — 

 the great blackback — might be discovered by the study of populations 

 which have no connection with fishing-ports. Where such independent 

 populations exist, if they do, they have certainly not yet been critically 

 compared. In north-western North America the short-billed gull, 

 which is an eastward extension across the Bering Straits of the common- 

 gull or mew-gull, overlaps with the ring-billed gull. The two birds 

 are fortunately sufficiently different in the zone of overlap to behave 

 as distinct species. As far as we know there has been no precise investi- 

 gation of their feeding habits. The information collected by Bent 

 (1921) gives us no real clue as the observations were not made in the 

 area of overlap. Both appear to have similar nesting habits, both 

 breeding on the ground and in trees. (Map, fig. 5, p. 44). 



From Salomonsen's account of the glaucous and Iceland gull in 

 Greenland it is evident that there is a fairly clear difference in their 

 nest-site selection. The nests of the Iceland gull are placed on narrower 

 ledges, and at the mixed colonies they, "keep at a respectful distance 

 below the glaucous gulls . . . being separated from them by a belt 

 of no-man's-land." The Iceland gull is a fish — and crustacean-eater 

 primarily, and also a scavenger, but it does not take any living bird, 

 and in Salomonsen's opinion the food-competition between the two 

 species is "slight or non-existent." The glaucous gull is a predator 



