Ein Vergleich zwischen revier- und individuenbasierten Verfahren zum Abschätzen von Brutbestandsindizes im Monitoring häufiger Brutvögel

In the German monitoring of common breeding birds (Monitoring häufiger Brutvögel, MhB), population trends are currently calculated with the rtrim software based on territory numbers that are estimated from simplified territory mapping data. In the present study, we compare population trend estimates resulting from the current standard approach with those resulting from different statistical methods or from key figures which differ from estimated territory numbers. We base our analysis on mappings made in Baden­Württemberg, collected within the German monitoring of common breeding birds. Volunteer observers collect data on four seasonal visits per sampling site. Currently, the subsequent data processing consists of two steps. Based on the data collected per plot, the observers estimate the number of territories per species, in accordance with predefined rules. Then, species­wise population trends are calculated with the widely used rtrim software, based on the estimated number of territories per site and year. Estimating the number of territories is time­consuming and subject to individual variation. The motivation for the present study is to evaluate alternative approaches to reduce the desk work of the volunteers and to simplify data processing. For a set of 51 common species, we calculated the following alternative target variables from the observation data: the number of observed individuals per visit, the number of individuals per visit from visits considered for subsequent territory estimation only (differs between species according to breeding season), and the number of individuals of the visit with the highest number of individuals. For all 51 species, we analysed these different target variables using rtrim, a Generalized Linear Mixed Model (GLMM), and a Binomial N­Mixture Model. We then compared the resulting index values and trends. For many species, similar trends resulted from either estimated territory numbers or alternative key figures. This is especially true for species where the number of detected individuals and the number of estimated territories is highly correlated among sites. The alternative metrics “number of individuals per visit from visits considered for subsequent territory estimation only” or “number of individuals of the visit with the highest number of individuals” often provide trends that are very similar to those of the standard approach. This is true for the calculation by rtrim as well as by GLMM. Together with other methods still under development, the presented results should help to streamline the procedure of trend analyses, and to relieve the volunteer observers.

Keywords: monitoring, breeding birds, territory mapping, trend estimate