SUMMARY

Theraphosid tarantulas, like all other spiders, secrete silk from spigots on the abdominal spinnerets. A few years ago, it was proposed that the large tarantula Aphonopelma seemanni could extrude silk from specialized spigots on the tarsi to help adhesion to vertical surfaces. This suggestion was later questioned because silk was not observed after the spinnerets had been sealed. Recently, experiments with the tarantula Grammostola rosea again suggested tarsal silk secretion. All observations of the supposed tarsal silk were made in spiders with functional spinnerets, thus contamination with silk coming from the spinnerets could not be excluded. Recent morphological arguments also questioned putative tarsal spigots and proposed that they are actually contact chemoreceptors. We here test the supposed tarsal silk secretion in Aphonopelma seemanni, Avicularia avicularia, Brachypelma vagans and Grammostola mollicoma using similar experimental conditions as the previous authors, but with sealed spinnerets. Our results clearly demonstrate that when spinnerets are sealed, tarantulas do not show any tarsal silk secretion. We reinterpret those putative tarsal spigots and discuss possible evolutionary implications of these findings.