Supplementary food is less variable than natural food. While feeding from constant food sources tends to be preferred by animals they must sometimes forage from more variable resources. However, the ways in which animals deal with the temporal and spatial variability of more natural food is not entirely understood. I investigated the decisions free-living rufous hummingbirds made when foraging from variable resources, where variability was encountered over time or within a bout via four field experiments. In addition, I investigated their use of wild flowers and differences in the use of supplementary food by these birds in different regions using surveys and by manipulating the distribution of feeders at feeding sites. I investigated the possibility of a genetic explanation for any differences in feeder use between regions using analysis of microsatellite DNA and banding data. Hummingbirds seemed to prefer to make foraging decisions based on past behaviour or post-ingestive feedback rather than on sensory information such as taste, which may be harder to assess accurately. Birds choosing between constant and variable rewards with equal means preferred the constant rewards when variability was high but tended to prefer the variable reward when variability was low. This seems to be a result of hidden time and other costs associated with foraging on highly variable resource but not on less variable ones, combined with potential benefits of information seeking from less variable resources. In addition, these preferences between constant and variable resources were affected by preceding foraging conditions. The number of birds using feeders was affected by population density and air temperature. Microsatellite data showed the rufous hummingbird population to have a fairly panmictic population structure. Investigating influences on foraging decisions at a large scale (population density) and small scale (resource variability) has provided a much wider understanding of their foraging behaviour than either could alone.

<p>Modern economic activities, like industry and agriculture, as well as household activities, generate an important amount of refuse. The way we collect, transport, and dispose it will determine the level of environmental contamination. May animals exploit refuse as a food source (i.e., anthropogenic food subsidy) and gulls are the most important group. Refuse subsidizes energetically gull populations, which impacts on their acquisition and allocation of resources, as well as on the environment, with ecological and evolutionary consequences are not well understood. In this dissertation we evaluated potential impacts of refuse on gulls by doing a literature review as well as empirical research on the Kelp Gull (<i>Larus dominicanus</i>) in the Rio de la Plata Estuary in South America. Direct and indirect impacts of refuse on gull species and the environment have been observed during the review process. We have detected positive impacts of refuse on body size, chick growth, fecundity, reproductive success, and population dynamics. However, negative impacts were also found focusing on fecundity, reproductive success, and population dynamics. Indirect negative impacts on other species, water bodies, and airport security were also found. Refuse produces numerous impacts on gulls at the individual, population, and species levels, with indirect negative consequences on ecosystems. There is a need to reduce the access of gulls to sources of refuse to mitigate the existing and potential conflicts with human activities and other species, especially those that are threatened and endangered. During our empirical research we found that refuse was ingested and assimilated by Kelp Gull chicks during the chick rearing period and that the ecological niche width increased with the age of the chick. We propose that parents incorporate isotopically unique food sources to nestling’s diet during their growth, increasing isotopic diversity of nestlings. Additionally, we found that refuse could affect foraging decisions of females during the pre-incubation period, which could positively affect future fecundity and negatively impact reproductive success. We found also that refuse consumption on fecundity and reproductive success of gulls is generally studied at the colony level, using conventional diet techniques, but not much has been done using stable isotopes at the individual level, making comparisons among studies and conclusions difficult to address. We encourage other researchers to continue incorporating the isotopic ecology perspective to study the effect of food subsidies on gulls. Additionally, we found that Kelp Gull on the coast of the Rio de la Plata Estuary ingest plastic debris. We conclude that plastic bags and plastic films might be the most important source of contaminants for the Kelp Gull on the coast of the estuary. Main findings of this dissertation suggests the need for an improvement of waste management practices and a regulation of plastic production and use in Uruguay to reduce plastic ingestion by gulls. Finally, next steps for research are provided in this important area of environmental science and natural resource management.</p>