Features Justification
Spinnerets: Elongate posterior lateral spinnerets (C11) Widely-separated spinnerets (C2) Pseudo-segmented apical segment of posterior lateral spinnerets (C10) Short apical segment of posterior lateral spinnerets (C9)
An association between “Dipluridae type” posterior lateral spinnerets, which are elongate and widely separated, and the construction of webs (sheet, funnel or curtain) has been proposed previously (Chamberlin & Ivie, 1945; Coyle, 1971; Eskov & Zonshtein, 1990). In some taxa with this spinneret type (and none without it) the spinnerets are pseudo-segmented, so this is also presumably associated with the same behavioral niche. At the opposite end of the spectrum, spinnerets with very short apical segments (traditionally called ‘domed’ or ‘triangular’ apical segments) show a clear pattern of association with burrowing spiders, many of which modify their burrow entrance.
Chelicerae and mouthparts: Presence of a rastellum (C51) Presence of a serrula (C43)
Observations of burrowing behavior indicate that the rastellum is used during burrow excavation and/or for modifying the burrow entrance (Coyle, 1971, 1981; Nascimento et al., 2021). Although the function of the serrula in Mygalomorphae is not well-established, we observed a potential association with spiders that construct opportunistic retreats and/or that do not construct a burrow. This is perhaps most evident in the Atypoidea, where the serrula is present in all species that show opportunistic retreat-construction habits (Hexurella, Mecicobothrium, Megahexura, and Hexura), and is absent in all genera that burrow (all Atypidae, Aliatypus, Atypoides, and Antrodiaetus).
Chaetotaxy of the anterior legs: Digging spines on legs I-II (C18) Presence of scopulae on the anterior tarsi/metatarsi (C20)
Strong lateral spines on at least metatarsi I-II, but usually also the tarsi and tibiae (previously called ‘digging spines’) have previously been associated with burrowing and/or trapdoor construction, and potentially prey capture (Raven, 1985). However, we observed that even in burrowing spiders, species with scopulae rarely possess these spines. We therefore hypothesized a positive correlation between digging spines and burrowing behaviors, but only when scopulae were not present. Scopulae have been studied extensively (Pérez-Miles et al., 2017; Wolff et al., 2013; Wolff & Gorb, 2012), with their major functions proposed as prey capture and locomotion. Pérez-Miles et al., (2017) identified an association between scopulae and particular burrowing behaviors, so we also tested this feature for correlation here as well. Characters of the tarsal extremities were not analyzed, as most showed no obvious association with behavioral niche (e.g., claw tufts and biserially dentate paired claws appear to have few or single origins and have rarely been lost despite the groups in which they are found inhabiting a range of behavioral niches) and we believe more subtle characters of claw dentition deserve more detailed attention prior to tests of association with behavior.
Chaetotaxy of the posterior legs: Leg III thicker and at least as long as leg II (C13) Spines of leg III mostly dorsal (C14) Patella III with pro-dorsal patch of >3 thorn-like setae (C15)
Behavioral observations have shown that in burrowing spiders leg III, and the posterior legs more generally, are used to anchor the spider in place in the burrow and for propulsion (presumably during prey capture) (Bond & Coyle, 1995; Coyle, 1981; Decae & Bosmans, 2014). We have observed that in burrowing spiders the posterior legs are generally larger relative to the anterior legs, have spines positioned mostly dorsally, and may be modified in other ways, either possessing a tibial saddle (a concave, asetose section of cuticle) or a patch of thorn-like spines on pro-dorsal patella III (and sometimes also on patella IV). We hypothesized that these characters are probably correlated with burrowing or entrance modification of some kind, and tested all of them except the tibial saddle, because this character is rare and restricted to relatively closely related taxa.
Eye group: Presence of a common tubercle C25) A compact, rectangular eye group (C22–23) A wide eye group (C22) Anterior lateral eyes in an advanced position relative to anterior median eyes (C23)
If we consider the ‘standard’ eye group to be a compact rectangle on a common tubercle, then this is modified in several ways within the Mygalomorphae. Firstly, the tubercle may be absent. Secondly, the formation of the eyes may be modified, with two common modifications being a widening of the eye group (e.g., in Actinopodidae and Migidae) or the anterior lateral eyes being positioned far advanced of the others (e.g., in Barychelidae and some Idiopidae). We observed that all modifications mentioned above are more common in spiders that modify the burrow entrance, and virtually never occur in non-burrowers, and therefore tested these characters for correlation with behavior.