Dispersion and dispersal of the Dominican gull in Wellington, New Zealand

The basic population structure of the Dominican gull, Larus dominicanus, in Wellington, New Zealand, was examined by total censuses, fixed transects, and by banding. A census during the period of peak autumn population in 1965 revealed 12,236 in the Wellington area, and over 84% within II miles of the harbour. There was an average of 48.5 birds/mile of coast; 48.8% of all gulls were at or near refuse tips or meatworks. On mainly uninhabited coast there was an average of only 2.4 birds/ mile, showing the distribution of the species is now greatly influenced by human occupation. Flocks are not discrete but comprise some birds that attend more or less regularly, and varying numbers of itinerants (from several roosting sites) that tend to 'knit' the population together. The population is at a minimum from November to January during breeding, and maximum from March to May. Itinerants comprised an average 9.4% of the population.
The number and percentage of adults in flocks increases rapidly from February until April or May and then declines. First-year birds begin to appear in flocks in January and increase in number and percentage during the year. This increase is affected by the very sedentary behaviour of local young gulls, the immigration of young from flocks adjacent to the city and. after June, to the departure of adults to breeding colonies. The percentage of young was apparently related to the proximity of feeding sites; flocks feeding usually had relatively more young than those resting, thus suggesting the young spend more time feeding than adults. There was no simple correlation between percentage of young and flock size. During summer breeding birds tend to visit flock sites near their colonies rather than ones further away. In periods of minimum attendance at colonies the flocks may be supplemented by up to 37% of the breeding adults. Attendance at flocks was also affected by the location of roosting sites and (for feeding flocks) by the type of food available.
Dispersal patterns were obtained from 275 dead recoveries as well as from sight records of 7,096 birds banded as fledglings in five colonies, and 337 gulls banded while roosting. The low proportion of recoveries in mountains or uninhabited country, and the high proportion in towns or populated areas does not distort the picture of gull dispersal but simply reflects true habitat preference and distribution as shown by aerial surveys. Mean recovery distance in the first year was 16.8 miles, after the first year (second to sixth years only) 11.6 miles and all ages combined 15.2 miles. If birds recovered in their natal colonies are ignored, the mean recovery distance after the first year was 14.4 miles. Dispersal was restricted and randomly directed. 64% came from Wellington and Lower Hutt cities. Dispersal to the South Island was very limited and was apparently related to the food supply and number of feeding sites near the colonies. There was no movement north in winter.
The dispersal pattern of gulls banded while roosting on Somes Island was apparently not different from that of birds born on the island. A survey of colour-banded birds showed again that gull distribution was affected by feeding sites near the colonies. A small number of recoveries in Wellington (which, however, represents half the total from gulls banded in colonies at the north end of the South Island) indicates some dispersal across Cook Strait to Wellington. In comparison with other species Dominican gulls in Wellington are remarkably sedentary.