Annual survival of Lesser Black-backed Gulls Larus fuscus and Herring Gulls L. argentatus in Schleswig-Holstein (Germany) and Denmark 2005–2014

Abstract

Survival of Lesser Black-backed Gulls and Herring Gulls was estimated by use of resighting data of individuals marked with colour-rings as chicks in breeding colonies on Langli (Denmark, North Sea), Amrum, Helgoland (Schleswig-Holstein, Germany, North Sea) and near Schleswig (Schleswig-Holstein, Germany, Baltic Sea) from 2005-2014 (Fig. 1). For Herring Gulls insufficient numbers of chicks had been colour-ringed on Langli and Helgoland to be included in the present analyses, and too few data were available for Lesser Black-backed Gulls from Schleswig. The seasons were defined as lasting from 1st August to 31st July the following year. For each bird a capture-observation history contained a ‘1’ in the year when it was ringed as well as for any other year in which it was observed at least once and a ‘0’ when the respective bird was not observed. The data were analysed using Cormack-Jolly-Seber models implemented in the program MARK that estimate apparent annual survival probability (φ) and resighting probability (ρ) independently from each other. A three step analysis was performed for each of the species: (1) we estimated the most appropriate model to estimate ρ, keeping φ constant. Then we (2) fitted a model set to estimate φ with ρ modelled according to step (1); and (3) we varied ρ in the most parsimonious model after step (2) to get the final model. Models were ranked according to the AICC.

In Lesser Black-backed Gulls, the most parsimonious model differentiated between three age classes (from ringing to fledging, first year of life, from start of second year onwards) and between the three breeding colonies (Langli, Amrum, Helgoland), but the model also showed interannual variation. For the first year of life, φ was between 0.48 and 0.89 (mostly 0.6-0.8, Fig. 7A). From the second year onwards, annual survival probability ranged from 0.82–0.97 with low standard errors and usually oscillating around 0.90 (Fig. 7B).

In Herring Gulls, the most parsimonious model again comprised three age classes and trends for each age class, but failed to show interannual variation and differences between the two breeding colonies Amrum and Schleswig. Compared to Lesser Black-backed Gulls, Herring Gulls had a lower, but nearly constant survival probability from the second year of life onwards (slightly increasing from 0.80 to 0.84; Fig. 8). For the first year of life, φ declined from 0.77 to 0.44 during the ten years of the study (Fig. 8). In connection with low breeding success observed recently in the Wadden Sea region, the low φ values in the first year may indicate unfavourable feeding conditions in the surroundings of the studied Herring Gull colonies. This is probably due to reduced discards from fisheries and the closing of landfills, the latter serving as winter feeding habitat especially for immature Herring Gulls, and may explain the decreasing size of the Herring Gull breeding population in northern Germany. It appears that Lesser Black-backed Gulls do not face such a problem, as they are mostly living far from the southern North Sea outside the breeding season, i.e. between SW Europe and W Africa.

In the future, an integrated monitoring comprising the measurement of breeding success as well as the determination of annual survival probabilities by the help of colour-ringing can provide better understanding of the observed growth rates in large gull opulations.