Abstract
Bird counting inevitably suffers from imperfect detection, which varies across species, habitats, period of the day, and seasons. Although various modeling techniques have recently been developed to account for this phenomenon, the biological basis of natural variation in detection remains insufficiently known.
This study examined the bird species' detection rate throughout the day, considering their body mass and diet type, concerning the environment and weather characteristics. Species detection rates were significantly affected by the number of individuals of that species but were unrelated to body mass.
Overall, species with the highest detection rate were Corn bunting, Blackbird, European robin, House sparrow and Common chiffchaff. Granivores-insectivores and insectivores showed significant differences in detection rates throughout the day among habitats, with higher detection rates in grasslands during the afternoon.
Insectivores showed higher detection rates in farmland during midday (warmest time of the day). Granivores, omnivores and scavengers did not show changes in detection rates in different day periods.
Such patterns in daily detection rates were significant even when considering abundance and total species richness in each community. Finally, cloudiness was unrelated to the overall detection rate of birds, while temperature and wind affected detection rates in some guilds.
Our findings provide some advice for choosing a suitable ornithological sampling method by considering the avian communities composition in combination with the type of environment, the diet of bird species, and the period of the day.