Hi, Sorry for the lack of details of my last e-mail. I'm working on feeding habitat selection of Ring-Billed Gull during the nidification period. During this period, gulls travel from the colony to feeding sites, as far as 65 km, and come back, 1 to 5 times a day. Gulls can stay only few minutes at a feeding site. The aim of our study is to link habitat selection and optimal foraging: we want to determine the energetic benefits linked to each habitat category and to determine the energetic costs linked to feeding sites in the gulls' home range during the whole nidification period. Our hypothesis is that gulls select feeding habitats so that their energetic benefits are optimized. So, we predict, for example, that feeding sites used far from the colony must present higher energetic benefits.
We are working with good resolution maps of agricultural boards, urban and suburban habitats. We have added to these maps pluviometry data, energetic benefits data for each habitat category, etc. First, we want to determine the scale(s) at wich occurs habitat selection, with first passage time (FPT). Second, we are thinking to build Cox proportional hazard models (CPH models) with residence time as response variable vs. distance from the colony, habitat category, energetic benefits, etc., as predictable variables. I wonder about processing FPT with "adehabitat" and my time lag: should I work with 4min time lag? We can continue this discussion in January! Thank you for your help and happy holidays, Martin Patenaude-Monette M.Sc. Candidate Université du Québec à Montréal Montréal, Québec
_______________________________________________ AniMov mailing list [email protected] http://lists.faunalia.it/cgi-bin/mailman/listinfo/animov
