© Marcel Burkhardt
Different types of data contain different fragmented information about where birds go and in what numbers they go there. Ring re-encounter data give precise information about the finding location, but the absolute numbers are not proportional to the absolute numbers of birds since the probability of finding and recording a ringed bird tremendously differs among countries and seasons. Geolocator data provide detailed information about the timing of migration as well as the stopover sites and the residence areas during the non-breeding season of a few birds that returned to and were recaptured in the breeding area. Stable isotope data give some indication about the area where the bird has moulted. Parasites can also contain vague information about the whereabouts of a bird.
We develop statistical methods that allow us to put together the information from the different types of data to quantify how many birds of which populations are going to which locations.
The aim of the R-package ‘birdring’ is to provide a collection of R-functions to help with the analysis of ring re-encounter data. At present, it contains functions to read EURING data into R, to draw maps for visualizing recovery data, to re-code EURING code into interpretable names, and to calculate the loxodromic and orthodromic distances between two encounter locations. The package also allows spatially different re-encounter probabilities to be estimated using the division-coefficient method. A function to obtain the proportional overlap between prior and posterior distributions facilitates parameter estimability, which is often an issue in mark–recapture models. The package is a work-in-progress and we welcome additional contributions of functions for the analysis of ring re-encounter data
Migration and distribution models for ring re-encounter and other data
For several bird species, we analyzed ring re-encounter data to infer non-breeding distribution or flyway use while taking into account heterogeneity in ring re-encounter probability. To do so, we developed stochastic models in the framework of state-space or multi-state models. Geolocator and stable isotope data were integrated into these analyses.
Kasper Thorup, Center for Macroecology, Evolution and Climate
Natural History Museum of Denmark; University of Copenhagen
Universitetsparken 15; DK-2100 Copenhagen; Denmark
Petr Procházka, Institute of Vertebrate Biology, Academy of Sciences of the Czech Republic, Brno, Czech Republic