Spiders are known to commonly use aerial dispersal, so-called ballooning, especially at juvenile stages. They produce a silk thread that allows them to rise up in the air to disperse, which serves as inbreeding avoidance or to find an optimal over-winter habitat. Studies of phenology, species and meteorological factors associated with aerial dispersal have been limited to laboratory settings, with few data obtained under natural settings and no studies to date executed in France. To understand aerial dispersal, we conducted daily sampling between 2000 and 2002 at a height of 12m. For adults, high proportions of "ballooners" were observed during four seasonal peaks, with dispersal most prevalent during summer, while for juveniles dispersal was protracted across summer and fall. Linyphiidae is the most abundant family among the 10,879 individuals caught. We show a significant and negative influence of high wind speeds on ballooning, an effect that increased even under low temperatures (<19°C). At wind speeds greater than 4m·s(-1) dispersal becomes difficult, and is almost impossible beyond 5.5m·s(-1). Ballooning ability is reported for the first time for several species. This study increases our knowledge on aerial dispersal in spiders in an agricultural context. Such behaviour can be seen as a survival strategy to escape from a disturbed and unstable landscape.

We describe a method for tracking the path of animals in the field, based on stereo videography and aiming-angle measurements, combined in a single, rotational device. In open environments, this technique has the potential to extract multiple 3D positions per second, with a spatial uncertainty of <1 m (rms) within 300 m of the observer, and <0.1 m (rms) within 100 m of the observer, in all directions. The tracking device is transportable and operated by a single observer, and does not involve any animal tagging. As a video of the moving animal is recorded, track data can easily be completed with behavioural data. We present a prototype device based on accessible components that achieves about 70% of the theoretical maximal range. We show examples of bird ground and flight tracks, and discuss the strengths and limits of the method, compared with existing fine-scale (e.g. fixed-camera stereo videography) and largescale tracking methods (e.g. GPS tracking).