Discussion
Predation of vertebrates by invertebrates might be more common than typically assumed, and spiders are the most cited invertebrate group showing such behavior (Barej et al. 2009). However, previous reports of spiders preying on amphibians point to an opportunistic behavior and provide no evidence of specialization. Here, we report systematic predation of amphibians by spiders using targeted traps. The observed traps were always built using two leaves that seemed to have the sole purpose for amphibian trapping. Vertebrates may hence not be only an opportunistic, indiscriminate or accidental prey, but rather a targeted systematically exploited food source of Olios sp. spiders. The behavior was observed independently in four spider individuals at four different sites suggesting that the trapping behavior is frequently performed by Olios sp. in north-eastern Madagascar. To our knowledge, this is the first observation of targeted vertebrate trapping by spiders.
The genus of Olios spiders is distributed globally with 246 species (Jäger 2012; Rayor 2018) and is included in the Sparassidae family, a group that is called ‘huntsman spiders’ (Rayor 2018). Most huntsman spiders do not build webs to capture their prey but actively stalk and run-down their prey with stealth and speed (Rayor 2018). However, some Sparassidae have been found in their own silk nest which is fastened with debris, living leaves or stems are completely surrounded by silk (Jackson 1987). Furthermore, most species in the family Sparassidae are nocturnal (Henon 2015; Rayor 2018).
The genus of Heterixalus frogs is arboreal and typically occurs in open areas such as clear-cut forest or rice fields (Blommers-Schlösser 1982; Raharivololoniaina et al. 2003). The species of H. andrakata is distributed in northern and north-eastern of Madagascar (Glaw and Vences 2007). During our ecological survey, we found H. andrakata to be mostly active at night, but recorded some daytime activity in agroforests. However, the species is typically hiding away during daytime between leaves.
A key factor facilitating the trapping behavior of Oliossp. may thus be that Heterixalus andrakata and possibly also other arboreal frogs try to hide from sunlight during the day in order to avoid dehydration (Rodel and Braun 1999). When temperatures rise, the frogs look for shade and cover away from the ground, which the spiders provide in form of their leaf trap. Additionally, the frogs might favor the seemingly protected traps in an attempt to hide from other predators such as birds that scan the vegetation for prey. The trap thus seems to be a necessity for the spider when attempting to catch frogs. Alternatively, the trap could also aim for catching prey other than frogs, such as insects. However, during our observation, no other prey was found in the traps and seeking shade seems of less importance for insects.
Interestingly, the majority of reports of amphibian predation by invertebrates stems from the Neotropics. Few predation events on Afrotropical anurans by invertebrates have been published (Barej et al. 2009). The only reports from Africa are from Tanzania and Uganda, where fishing spiders prey on tadpoles (Vonesh 2005), from South Africa where crabs predate on amphibians, and from Cameroon where wandering spiders prey on tree frogs (Barej et al. 2009). One report from Madagascar describe predation event on tree frogs (Boophis rufioculis ) by freshwater crab (Hydrothelphusa sp.) (Rosa et al. 2014).
Our observation is, to our knowledge, the first report of spider predation on vertebrates from Madagascar. Whether this geographic bias concerning amphibian predation by invertebrates is indeed reflecting a difference in the frequency of such behavior or whether the bias is due to more research being conducted in the Neotropics (Meyer et al. 2015; Martin et al. 2015) remains, however, unclear.