Résumé
Successful management practices for declining bird species depend often on long-term surveys acquired by point counts. Despite high standardization of field protocols, uncertain detection probability remains an important source of variability and bias in point-count data. This effect is of main importance in low-responsive species as the Red-legged partridge (
Alectoris rufa
), but it can be counterbalanced, increasing detection probability. In this 2-year study, we sampled using traditional point-count methods, followed by playback sessions for each repetition. We measured detection probability and the efficiency of playback for detectability in the context of a feasibility study on long-term point-count surveys for a harvested game bird, the Red-legged partridge. The results for both study years show a distinct increase in detection probability (23% and 45%, respectively) when using playback vs. the traditional point-count method. We also tested our results for heterogeneity, trap dependence, and time dependence, and no effect was detected. Thus, we suggest that the future design of long-term surveys on Red-legged partridges should consider abundance indices using playback sessions.