More data, fewer shifts: Molecular insights into the evolution of the spinning apparatus in non-orb-weaving spiders

Abstract

All spiders produce silk and use it for various functions throughout their lives, but not all spiders produce the same silks, or use them for the same functions. These functions may include building shelters, protecting eggs, and trapping prey. The “RTA clade” of spiders (grass spiders, jumping-spiders, wolf spiders, hackled-band weavers, etc.) is an extremely diverse group (∼18,000 species, representing nearly half of all described species), with great variation in ecology and morphology, including variation in the cribellum, a specialized silk-producing organ. The loss of the cribellum, a structure that produces fibers contributing stickiness to prey snares and which is invariably associated with a set of accessory structures, has been studied in orb-web-weavers and shown to have been lost once during the evolutionary history of the group, but never regained. Relative to the orb-weavers, evolution of the structure remains less-thoroughly studied in the RTA clade. As the cribellum is one member of a suite of traits, the combined action of which is essential in prey-capture, its loss should have ecological correlates or physiological trade-offs of evolutionary interest. Using molecular data from nuclear genes (ribosomal DNAs 18S and 28S, and protein-coding Histone H3), as well as mitochondrial data (Cytochrome oxidase I) totaling ∼3400 base pairs, we developed a phylogenetic hypothesis for three-clawed lineages in this group, focusing on families where taxonomy and previous cladistic analyses suggest multiple losses, or possibly loss and secondary gain, of the cribellum. Results of Bayesian and direct-optimization (POY) analyses agree on a well-resolved and robust agelenid clade that includes the putative subfamilies Ageleninae, Tegenariinae, Textricinae and Coelotinae, but excludes the cribellate New Zealand genus Neoramia. Optimizing the pattern of cribellum evolution onto these trees shows that the cribellate state is conserved in large clades and has undergone fewer shifts than current taxonomy implies. The dominant pattern is one of repeated loss of the cribellum, though loss and regain remains a possibility in some groups.