Spaans A.L. (1971) On the feeding ecology of the Herring Gull Larus argentatus Pont. in the northern part of the Netherlands. ARDEA 59 (3-4): 73-188

This paper deals with the feeding ecology of the Herring Gull in the northern part of The Netherlands (Fig. 1). Most of the data were collected in 1966-1969. The aim of the study was to obtain quantitative data on the share of garbage and other offal, supplied indirectly by man, in the diet of this species, as a contribution toward the discussion of the causes of the population increase shown by the species during the present century. Chapter 2 gives a short review of the annual cycle of the species in the Wadden Sea area. The first gulls usually arrive at the breeding places in February, but are sometimes already present at the end of January. The first eggs are laid at the end of April or the beginning of May. The first young are ready to fly by the middle of July. At the end of August, most of the gulls have left the breeding colonies. An analysis of recoveries of Herring Gulls ringed as chicks on the Dutch Frisian Islands (Chapter 3) indicates that like most north-western European populations the gulls in the Wadden Sea area do not show a seasonal migration but disperse non-directionally over a limited area (Figs. 2-6, Tables 1 and 2). The majority of the population remains within 100 km of the ringing locality. During the first half of the present century more Herring Gulls travelled further from the ringing locality than after 1949, but the difference was significant only for the first-year gulls (Table 3). These differences are attributed to the intensive control of the Herring Gull in The Netherlands in the 1950s and early 1960s, resulting in a high mortality of the adult gulls during the breeding season and a low production of young. This in turn led to a smaller population and a relatively more favourable food situation at the beginning of the dispersal period than would normally have been the case. It is suggested that the migration tendency of part of the population was weakened as result of this situation. Most of the Herring Gulls of the Dutch Frisian Islands do not start to breed before the end of their fourth or fifth year. There was a tendency, although not a very pronounced one, to return to the colony of origin to breed; quite a number of gulls, however, settled in nearby colonies (Table 4). This dispersion is attributed to the disturbances at the nesting sites resulting from the intensive control measures. Most of the mortality among first-year gulls occurs in September and October, i.e. at the end of the dispersal period (Fig. 7). Second-year and third-year gulls show a more or less constant mortality throughout the year. Fourth-year and older birds die mainly during the breeding season (May-July). The latter pattern does not seem to be a result of the control program. Since most of the population spends the entire year within a limited area, it was possible to determine the distribution of the gulls outside the breeding season by counts (Chapter 4). Counts made regularly around high tide on Terschelling and Vlieland in 1966-1968 (Fig. 8) show that on the Frisian Islands the numbers drop rapidly after the breeding season. On Vlieland, the decrease was followed by a temporary increase, probably due to the dispersing gulls of the Terschelling colony, which accounts for about half of the breeding population of the Frisian Islands. On Terschelling the numbers were lowest during November and December, on Vlieland in January (1966-1967) and during November-February (1967-1968), after which the numbers increased gradually. The numbers on both islands showed appreciable fluctuations during the winter, probably under the influence of weather conditions. More or less simultaneous counts made in the Dutch Wadden Sea area give total numbers of 33,000-39,000 Herring Gulls (Table 5). The estimated minimum and maximum numbers in the winter of 1966-1967 were 21,000 and 37,000, respectively, and in the winter of 1967-1968 14,000 and at least 28,000. In 1967-1969, counts were made regularly at 13 refuse dumps on the mainland of the province of FryslGn (Figs. 9 and 10, Table 6). There, the lowest values were obtained during the breeding season. The numbers increased from the end of the breeding season onward; the winter level was reached as early as October, and the numbers remained high until April. During the winter months the numbers at the dumps showed marked fluctuations. There appeared to be a negative correlation between the feeding conditions on the mudflats of the Wadden Sea and the numbers of gulls at the dumps (Fig. 11, Table 8). In December 1967, a census was taken of the gulls at almost all of the 104 refuse dumps in the three northern provinces (Fig. 12), i.e. covering the main dispersal area on the mainland. The total shown by this census was 26,097 Herring Gulls, but as a result of the varying feeding conditions on the Wadden Sea this value must have fluctuated during the winter of 1967-1968 between 20,000 and at least 39,000 gulls (Table 9). There proved to be a positive correlation between the number of inhabitants of the relevant municipalities and the number of gulls at the dumps (Fig. 13), which indicates that the distribution of the gulls over the refuse dumps is closely related to the quantity of food locally available. It is suggested that this distribution is brought about by intraspecific competition during the dispersal period. Few gulls of other species were observed at the refuse dumps during the census. This is partially related to differences in the feeding ecology of the various species. But in parts of The Netherlands where Herring Gulls were scarce, Black-headed Gulls did occur in large numbers at refuse dumps during winter. The same held for some of the dumps in the northern provinces during the breeding season. It is therefore suggested that in the northern part of the country in winter a high proportion of the Black-headed Gulls retreats from the dumps as result of interspecific competition. The total number of Herring Gulls on the mainland of the provinces of FryslGn, Groningen, and Drenthe must have fluctuated during the winter of 1967-1968 between 24,000 and at least 47,000 (Table 11). Results of food analyses, based on examination of pellets, faeces, regurgitated food, and stomach contents, are given in Chapter 5. The food of full-grown Herring Gulls visiting the breeding colonies on Terschelling and Vlieland consisted mainly of marine invertebrates from the littoral and sublitoral zones (Tables 12-14). During the breeding season no difference in diet was observed between the sexes (Table 12). Examination of more than 27,000 pellets and faeces showed marked changes in the composition of pellets and faeces in the course of the season (Fig. 14). These seasonal variations are attributed mainly to changes in abundance, accessibility, and attractiveness of the various prey species. The food of chicks consisted mainly of fish (Tables 16-22), with a preponderance of Clupeidae, Gadidae, and flat-fish (Tables 23 and 24). Most or all of the Gadidae and flat-fish derived from discards of commercial fishery. This source was estimated to provide about one-fifth to one-third of the total food-supply (Table 26). Garbage provided about one-fourth of the total food-supply. Thus, on the average about 50% of the food was obtained through man. A positive correlation was found between brood-size and the percentage of gulls feeding garbage to their chicks in addition to natural food (including fish from commercial fishery) (Fig. 16). It is suggested that this represents a reaction of the parents to the increasing food requirements of the larger broods. The chicks completely digest fish bones, but juveniles, like the adults, regurgitate them. Some possible explanations of the differences in diet between chicks and adults are discussed. On the basis of the fluctuations in the breeding population of the Eider, Shelduck, and Avocet on Vlieland during the last four decades it is concluded that the Herring Gull cannot be considered harmful to these species in spite of the heavy toll of eggs and young birds taken by the gulls annually. It seems likely that the results of predation are counterbalanced by immigration. Outside the breeding season, a large proportion of the Herring Gull population feeds at refuse dumps and other places with human waste on the northern mainland of The Netherlands. For the winter of 1967-1968 this part of the population is estimated, on the basis of various counts made in the Wadden Sea area and on the mainland, to amount to at least 32% of the total winter population when foraging conditions on the Wadden Sea were favourable. When these conditions were unfavourable, this percentage increased to at least 77%. Some data on the food intake of full-grown gulls (Tables 29 and 30, Figs. 17 and 18) and chicks (Table 32, Fig. 19) are given. The intake of wild gulls foraging on Cockles is estimated to average 294 Cockles over one year old or 2,165 Cockles a, few months old, i.e. 167 ml Cockle flesh, per low-water period. Full-grown gulls in captivity showed the following mean daily intakes (fresh weight): 363, and 368 g Horse-mackerel; 199, and 229 g Sprat and young Herring; and 48 to 101 g bread. There was a distinct relationship between the mean quantity consumed daily and the mean changes in body weight (Fig. 18). The total food (fish) intake of two groups of chicks in captivity during the first 6 weeks (chick stage) amounted to 9.2 kg fresh weight (2.5 kg dry weight) and 8.4 kg fresh weight (2.4 kg dry weight), respectively, per chick. Full-grown Herring Gulls in captivity proved to be able to fast for 8 consecutive days without untoward consequences (Fig. 21), and to recover rapidly from the effects of starvation (Fig. 20). It is suggested that the Herring Gull is adapted to an irregular food-supply. Chapter 6 deals with the growth of chicks. In 1966 and 1967, but not in 1968, the growth-rate decreased from early-hatched to late-hatched chicks (Fig. 23, Table 33). Since this holds both for broods with one young and for broods with 3 young, the differences were suggested to be caused by weather conditions. In several cases chicks of bl1 grew faster than chicks of b/3 (Fig. 23, Table 33). The weight increase of chicks in supernormal broods was irregular (Fig. 24). Several chicks grew very slowly and reached a much lower maximum weight than did chicks of normal broods. Chicks that did not reach a body weight of 600 g all died during or shortly after the chick stage. According to the present findings, in the Wadden Sea area the Herring Gull does not seem to be able to raise more than three young, which is the maximum natural brood-size found. Chicks fed on garbage in addition to natural food generally grew faster than otherwise comparable birds in broods fed only on natural food (Table 34). Undernourished chicks show compensatory growth only under improved food conditions. Presumably in wild gulls the variations in growth-rate result in differences in fledging weights. The possible influence of fledging weight on post-fledging survival is discussed. The possible role of the extra food supplied indirectly by man on the regulation of the number of gulls is discussed in Chapter 7. On the basis of the data obtained during the present study, it seems probable that the northern part of The Netherlands has a greater winter population than it would have without garbage supplied by man. It seems likely that this extra food determined the level to which the population could increase.