Spatial proximity and prey vibratory cues influence collective hunting in social spiders

Abstract

Social spiders are thought to predominantly receive information about their environment through vibrational cues. Thus, group living introduces the challenge of distinguishing useful vibrational information from the background noise of nestmates. Here we investigate whether spatial proximity between colony-mates may allow social spiders (Stegodyphus dumicola) to reduce background noise that might obstruct vibrational information from prey. To do so, we constructed experimental colonies and measured whether the number of spiders in proximity to one another whilst resting could predict the number of spiders that participated in prey capture. Additionally, we exposed spider colonies to five different simulated vibrational cues mimicking prey to determine which cue types spiders were most responsive to. We found that the number of spiders huddled together prior to foraging trials was positively correlated with the number of spiders participating in collective foraging. Furthermore, colonies responded more quickly to pulsed vibrational cues over other types of vibrational patterns. Together these data reveal that both social interactions and prey cues shape how social sit-and-wait predators experience and respond to their environment.

Keywords: collective action; communication; foraging; seismic cues; sociality; spiders.

Figure 1.

Relationship between the number of spiders resting in contact prior to foraging and the average number of spiders attacking during foraging. The line represents the best fit regression and the shaded area is the 95% confidence intervals.

Figure 2.

The effect of vibratory cue on collective foraging measured as latency to attack in seconds (A) and total number of attackers (B). Error bars depict standard error. Purple bars refer to cues delivered by the Magic Purple Bullet vibratory device, and grey bars refer to cues given by Arduino devices. Bars that do not share a letter reflect statistically significant differences using a post hoc Tukey test.