4.1 | Insights into the function of convergent morphological features
This study is the first to quantify the strong correlation between behavioral niches and a suite of convergent morphological features within the Mygalomorphae. In particular, features of the spinnerets, leg chaetotaxy, and eye group, as well as the rastellum and serrula, exhibit strong patterns of correlation with behavior, and an examination of their likely function provides insights into the potential drivers of convergent evolution within the group.
Spinnerets. Elongate, widely spaced posterior lateral spinnerets are correlated with web-building (Table 2; Fig. 3). Their length presumably allows for the efficient application of wide swathes of silk during the construction and repair of capture webs, as has been observed in Linothele (Eberhard & Hazzi, 2013; Nicolás Paz, 1988). Their widely separated position likely also aids in the independent, unilateral or asymmetrical use of each spinneret during web-construction, for example during the attachment of individual silk sheets (as observed in Linothele macrothelifera (Eberhard & Hazzi, 2013). In contrast, very short apical segments of the PLS (and short spinnerets in general), are correlated with structural modification of the retreat entrance (Table 2; Fig. 3) and are probably better for the precise application of strong, thin bands of silk (as observed in Ummidia : Coyle, 1981). The precise application of silk may be important for the integrity of these entrance structures, for example in the construction of a trapdoor hinge, or in the substrate/silk-matrix of a trapdoor or turret (Coyle, 1981; Coyle et al., 1992). During burrow and burrow-entrance construction, these short spinnerets have been observed to work together synchronously and/or rhythmically, usually applying silk to the same area, explaining their position close together on the abdomen in these species (Coyle et al., 1992; Mayo, 1988).
Rastellum and serrula . The rastellum is strongly correlated with both burrowing and door construction (Table 2; Fig. 3). Observations of burrowing taxa indicate that it is used for compaction of the burrow shaft and entrance structures (Coyle, 1981; Coyle et al., 1992) plus excavation (Gertsch, 1949; Nascimento et al., 2021). However, both burrowing and entrance modification occur in taxa that don’t possess a rastellum (e.g., Theraphosidae and Migidae, respectively), suggesting that other factors may also influence whether the structure is necessary, for example the substrate in which the spider burrows. The function of the serrula in spiders is generally assumed to involve manipulation of prey items (Jocqué & Dippenaar-Schoeman, 2006). We found that it was positively correlated with opportunistic retreats, and negatively correlated with burrowing (Table 2; Fig. 3). The functional reasons for this are unclear, although a speculative explanation for the negative correlation of the serrula with burrowing could be a tendency for it to become clogged with substrate while burrowing, because substrate is carried using the chelicerae/pedipalps during burrow construction, and so would likely come into contact with the serrula (Coyle, 1974, 1981; Mayo, 1988).
Leg chaetotaxy. Surprisingly, the so-called ‘digging spines’ – strong lateral spines on the anterior legs and pedipalps, did not show a positive correlation with digging, but only with burrow-entrance modification (Table 2; Fig. 3). That digging is not the primary role of these spines is supported by behavioral studies of burrowing taxa that observed that the chelicerae and fangs are used for substrate excavation, not the legs (Coyle, 1981; Coyle et al., 1992; Mayo, 1988; Nascimento et al., 2021). Furthermore, some taxa that do not burrow (e.g., many Migidae) still possess these spines, although they have lost other features associated with burrowing (e.g., pro-dorsal spine patches on patella III). We suggest that these spines function primarily during prey capture in species with modified burrow entrances, which tend to have smaller foraging areas (Main, 1982) and hunt by lunging from the burrow entrance and restraining prey with the anterior legs and pedipalps (Coyle, 1981, 1986; Hils & Hembree, 2015). Although no correlation was found between scopulae and behavior, in taxa that modify the burrow entrance scopulae clearly replace the function of digging spines because the only entrance-modifying taxa without digging spines possess scopulae, adding to the well-supported hypothesis that a function of both structures is to restrain prey (e.g., see Eggs et al., 2015; Pekár et al., 2011; Wolff & Gorb, 2016).
Enlarged posterior legs, a dorsal bias in spine position on the posterior legs, and the presence of pro-dorsal thorn patches on patella III are all correlated with both burrowing and burrow entrance modification (Table 2; Fig. 3). Behavioral studies on several burrowing species indicate that the posterior legs are braced against the burrow wall to anchor the spider (Bond & Coyle, 1995; Coyle, 1981; Decae & Bosmans, 2014; Hils and Hembree, 2015). This is done during routine movement, but also serves a defensive function in species that hold their burrow entrance shut when disturbed. Larger, stronger posterior legs and dorsal spines likely enhance this bracing function.
Eye group. The eye tubercle was found to be positively correlated with opportunistic burrowing, and a standard, compact, rectangular eye group was found to be negatively correlated with burrow entrance modification (indicating that change from this state generally occurs in taxa with modified entrances) (Table 2; Fig. 3). It seems most probable that these changes in the eye group relate to the amount and direction of light exposure (and therefore visual information) in different retreat types, for example, almost all opportunist taxa have relatively open retreat entrances, and when foraging at the retreat entrance, would be exposed to light from all directions. In contrast, burrowing taxa with modified entrances would be exposed to light from only one direction (the entrance), and to far less light in general. This is, however, in contrast to several previous studies which indicate that vision is not important for foraging in a range of mygalomorph species (see Coyle, 1986 for a list of relevant literature).