Strong negative effect of diurnal rainfall on nocturnal activity of a wandering spider in Central Amazonia
Rainfall is a common phenomenon in tropical forests influencing the behavior of many animals, however, little is known about its post-occurrence effect on behavior. We investigated the effect of diurnal rainfall on the nocturnal activity of the wandering spider species Phoneutria reidyi in nights without rainfall. Our study included two different areas and periods: a coconut plantation, located 108 km from the city of Manaus, containing an area of 80 m x 80 m with 105 palms, from July 2014 to July 2015; and an area of 80 m x 80 m in a rainforest fragment in Manaus, from December 2015 to March 2016. Each night, we counted active spiders (found outside refugees) searching carefully on the vegetation and on the ground using a headlamp between 19:00-00:00. Spiders were not captured to avoid the effect of disturbance, and were measured by approaching a caliper rule. We used an index to measure the diurnal rainfall effect (DRE) which was the percentage of change in the number of active spiders at night after a diurnal rainfall, considering 100 % the number of spiders active at the previous or following night, without previous rainfall during the day. This pairwise approach was used to avoid seasonal bias and included 15 pairs of nights in the plantation, and 15 pairs in the forest. A total of 2 243 active spiders were counted. The number of active spiders was always smaller in nights after diurnal rainfall, with a mean reduction of 53.4 %. The abundance of active spiders reduced significantly in both areas after a diurnal rainfall, and the effect was not different between areas. Larger spiders (mostly adults) reduced their activity (-62.8 %) more than smaller spiders (juveniles, -48.5 %). The amount of rainfall during the day had no effect on the nocturnal activity, i.e., the effect of strong diurnal rainfall is similar to the effect of a weak rainfall. The air temperature did not change significantly at night after diurnal rainfalls. The seasonality of rainfall apparently has a weak or absent effect on the abundance of P. reidyi, which was approximately constant through one year. We believe that the moisture, which may affect the chemical cues of prey, is the major cause to reduce the active spiders after rainfall, but we discuss other potential causes. Smaller individuals are probably more active under less favorable conditions due to the stronger need of food for growing. If this effect of rainfall on the behavior is common for wandering spiders in general, the rainfall events may have important consequences for the entire community of arthropods and small vertebrates. We suggest that studies based on relative abundance of spiders should take in account this potential effect in collection and analysis of data.