Landscape use by gulls (Larus spp.)

Abstract

In contrast to the negative impacts of landscape change on many aspects of biodiversity, scavenging bird species, like gulls (Larus spp.), have reacted positively and expanded their ranges. This in tum has brought a number of problems, including; damage to town centres, an increased risk of disease transmission to humans, domestic animals and livestock, an increased risk of collision with aircraft and a threat to vulnerable seabird populations. A great deal of money has been invested in measures to mitigate these problems. However, these have often been hampered by a failure to understand the ecology of the system concerned. This study employs a variety of statistical techniques to investigate factors related to the spatial and temporal distribution of gulls, the possible problems they may cause and the efficacy of measures to reduce the impact of these problems. Using structural equation modelling (SEM) it was possible to show that in contrast to other groups, like corvids which use landfill sites close to their roosts throughout the year, gulls rely most heavily on landfill sites as a source of food during the winter. However, analysis of the spatial distribution of winter gull roosts using negative binomial generalised linear models (GLMs) showed that only roosts of the black-headed gull (Larus 'ridibundasv were positively influenced by proximity to landfill sites. In contrast to the winter, when roosts were widely distributed, during the summer roosts had a coastal distribution. The problems posed by gulls to air safety and human health were investigated by analysing the spatial patterns of accidents and the incidence of salmonella carriage by 3 wild birds. In the first, bivariate k-means clustering revealed that strikes on Royal Air Force (RAF) aircraft by gulls were clustered within the 6 km surrounding major (>1000 individuals) gull roosts and landfill sites. These results were used to identify additional areas capable of supporting large numbers of gulls, and hence likely to represent a threat to aircraft in the future. I used survival analysis to investigate temporal and spatial patterns in wild bird salmonella. Passerines were more likely to be infected with salmonella than non-passerines and further analysis was indicative of gulls having a higher rate of salmonella infection than other non-passerines. Salmonella prevalence in wild birds was greatest during the winter and spring, and in areas with large populations of cattle. Having investigated the factors determining where gulls were and the risks they posed to aircraft safety and human health, I analyse the efficacy of a range of management techniques to control problem gull populations, using linear mixed effects models (LMEs). This revealed that techniques with occasional lethal events were the most effective. By using a range of statistical techniques, it was possible to disentangle a series of complex and often interacting relationships between gulls, the landscape and humans.