We placed four specimens of A. seemanni each in separate glass containers with the floor and two vertical walls covered with 20 microscope slides. Tarantulas remained in the container for 14 h, after which the slides were carefully examined. Seventy-two hours later, we sealed the spinnerets of the same individuals with paraffin and repeated the experiments. In the first trials with free spinnerets most slides (24 out of 40 horizontal slides and 16 out of 40 vertical slides) showed silk threads together with dislodged abdominal urticating hairs. None of the slides in the trials with sealed spinnerets showed silk threads; we found only urticating hairs. In both series of trials we observed the spiders walking on vertical surfaces. In addition, we made transverse cuts of A. seemanni tarsi and no structures interpretable as silk glands or silk conduits were observed.

Tarantulas entangle silk threads from the spinnerets with their tarsi (Fig. 1). They often use hind legs to entangle silk, but we also observed A. seemanni entangling with their other legs. This behaviour might explain the presence of the thread footprints photographed by Gorb et al.1. Tarantula silk released from spinnerets consists of multiple, roughly parallel silk threads which are better observed when magnified. This pattern could be maintained once the silk line is adhered to the substrate, when combed or stepped by tarsal scopulae. Tarsal structures interpreted as spigots by Gorb et al.1 are very different from the diverse kinds of spigot reported for spiders2,3,4. We found that the structures reported by Gorb et al.1 are very similar in morphology and size to fragments of tarsal thermosensory setae reported for other tarantulas5. In insects, a common origin or convergence was proposed for silk spigots and sensory setae6,7. The presence of conspicuous, dense scopulae with spatulated setae on the ventral surface of tarsi and metatarsi of all tarantulas seems incompatible with the presence of spinning spigots in a lower layer, because scopula setae might interfere with silk release. Furthermore, it has been explained that in A. seemanni the friction with scopula setae is enough to guarantee spider climbing8.

Figure 1: The zebra tarantula Aphonopelma seemanni entangling a silk thread released from the spinneret with its hind leg, while starting to climb a vertical glass wall.
figure 1

Arrows indicate the silk thread.

PowerPoint slide

During our observations, individuals of A. seemanni brushed spinnerets with their hind-leg tarsi while walking and releasing silk threads. The presence of abdominal urticating hairs with silk is indicative of this brushing. We did not find evidence of silk-like secretion through leg structures in tarantulas.