Human activity influence on gull-vegetation interaction
GUILLEMETTE, M. 1997. Influence de l'activité humaine sur l'interaction goéland-végétation et sur le maintien des populations aviennes de la RPNAM . Rapport présenté au service de conservation des ressources naturelles, Parcs Canada, Région du Québec 181 pages + annexes.
© Parks Canada / É. Le Bel / L 30 04 95
This study was conducted during the summers of 1994, 1995 and 1996 in order (1) to investigate diet, productivity and the naturalness of the Herring Gull population nesting on île Nue de Mingan , (2) to measure the impact of Herring Gulls reproductive activities on the vegetation and floristic composition of île Nue , (3) to determine the influence of human disturbance on nesting success of Common eiders ( Somateria mollissima ) and (4) to measure predation rate of gulls on the ducklings of Common eiders and to quantify the impact of boat traffic on the gull-duckling interaction. All these objectives were aimed at monitoring stress points of the system and the ecological integrity of MANPR . For the first objective, nesting chronology, attendance pattern, nesting success and productivity of fledglings were estimated following marked nests and ringed chicks in the course of the season. Diet was identified by (a) catching live individuals from which their stomach contents were pumped, by (b) collecting non-digestible pellets regurgitated by adults on roosting areas and by (c) collecting regurgitations offered by adults to the chicks. Results showed that clutch size and productivity were very low for the three years of the study which indicate that this population will decrease if such a situation prevails in the future. Diet of chicks was essentially composed of capelin ( Mallotus villosus ) whereas diet of adults was much more diversified being composed of prey of terrestrial (small fruits, insects, eggs and chicks of birds), littoral (molluscs, echinoderms and crustaceans), pelagic (capelin) and artificial origin (garbage). The relative importance of each of these groups varied seasonally and it is concluded that fisheries discards and refuses are, at the moment, of little importance for this colony. To the contrary, results suggest that the low level of productivity observed in this study is linked with shortage of their main natural prey, the capelin.
The influence of reproductive activities of gulls on the vegetation were assessed using an experimental approach comparing, for four points in time, control and treatment plots (of 25 m 2 each). Frequency of occurrence and percentage cover were quantified for all plant species present while the density of individuals was evaluated only for a number of species. Soil samples were taken in control and treatment plots concomitantly with measurements of vegetation. Results of this experiment showed that growth and percentage cover of Ledum groenlandicum decreased while pH and phosphorus levels increased significantly. When compared with former studies, our results suggest that the main plant association of île Nue changed from Empetrum-Ledum-Epilobium to Deschampsia-Stellaria-Epilobium association. It also appears that Empetrum nigrum (crow berries) disappeared from the gull colony. Finally, our results indicate that roosting areas used by gulls were characterised by a nearly complete absence of plant cover while nutriment levels were high when compared to control and treatment plots.
For the third objective, we first described the nesting habitat used by Common eiders by measuring for each nest, vegetation height, distance from the coast and concealment index. Clutch size and laying chronology were also assessed by using a regression equation relating density index of eggs and the number of incubation days. The impact of human disturbance on nesting success was estimated experimentally by conducting repeated visits to each individual nests using three different treatments based on the frequency of visits and the occurrence of the first visit. Disturbance responses and hatching success varied little in relation to the frequency of visits whereas it was more severe when a sequence of visits started in the beginning of the incubation period, indicating that the period at which a disturbance occurs is a more important factor. When the first visit is applied to the first half of the incubation period, it was demonstrated that it decreased hatching success by 17 % (forested islands) and 37 % (open islands). For the three different treatments applied, the first visit is always more destructive on a relative scale with about 50 % of mortality (all nests that failed to hatch) while the response toward disturbance is also influenced by density of nesting gulls and nesting habitat.
For the fourth objective, predation rate of gulls on ducklings leaving the nest and feeding on the rearing areas was measured using focal-animal sampling. Results for the two periods differed strikingly being 6,2 attacks per hour per crèche when leaving the colony whereas it was 0,09 on the water. Although leaving the nest and walking toward the beach is a short phase (about 10-40 min) of the reproduction period compared to the time spent on the rearing areas (12-16 weeks), this indicates that it is the most sensitive period of the reproduction. Unplanned disturbances (caused by our research activities) of duckling leaving the nests showed that hunting success of gulls shift from 33 % for undisturbed broods to 100 % for disturbed one. Simulating disturbance now of crèches on rearing areas with a small boat, we showed that predation success of gulls during such interactions increased tenfold from 4 % for undisturbed crèches to 40 % for disturbed ones. This experiment also showed that distance between the boat and the crèche was more important (presence/absence of predation) than the speed at which the boat approached the crèche . Although the outcome of boat-crèche-gull interactions is striking, the occurrence of such interactions was negligible at île à la Chasse and was 10 to 15 less than the crèche-gull interactions recorded at Firmin and Marteau islands which are characterised by high boat traffic. This conclusion is based on the quantification of the disturbance "très proche" (corresponding to the one simulated in our experiment) and the recording of alert behaviour of crèches during our observations. A further analysis suggest that disturbance by boats in this high traffic sector may be as deadly as harassment by gulls acting on their own (without any human-driven stimuli). Finally, partial information gathered in 1994 indicates that eiders bagged at the opening of the hunting season were all ducklings unable to fly. The impact of such a phenomenon on the nesting population is unknown.
In a final chapter, I analysed the demographic trends of Herring Gulls and Common Eiders in the last seventy years in the MANPR and the North Shore of the Gulf of St. Lawrence as a whole. It appears that gull population of the North Shore increased steadily since the fifties to maintain itself at a high level since the sixties. Based on this, and indirect evidence reported in other studies, I suggest that the actual population level of Herring Gulls nesting on the North Shore is unnatural. For the Common Eider, the majority of the North Shore population is found in the MANPR where populations increased recently. Based on the absence of nesting eiders at Bouleaux islands bird sanctuary, which was an important colony from 1925 to 1960, and new colonies identified in 1988, I suggest that the population of eiders of MANPR could increase further in the future. Finally, using the specific example of interaction between vegetation, colonial birds, foxes and man, I discuss the applicability of the ecological integrity concept to the MANPR by showing how difficult it can be to state about the naturalness of an island and its bird population.
This report is available at the Regional Library of Parks Canada in Quebec (in French).