Data collection and preparation
We used a sample of individuals from 11 species to investigate shape complexity in the amblypygid pedipalp. This incorporated 82 individuals in the tibia analysis, and 77 in the femur analysis. Disparity in the number of specimens between analyses resulted from the exclusion of individuals with damaged pedipalp segments. Species were chosen to represent a diversity of pedipalp lengths and morphotypes, and cover a large taxonomic range, spanning two of the five amblypygid families and encompassing six out of seventeen amblypygid genera. Specimens were placed in to species by taxonomists based on morphology. The species chosen also span a wide geographic range, originating from South and Central America, Africa and Asia. Multiple replicates of males and females were collected for each species to facilitate comparisons between the sexes (see Supplementary Material S1). All specimens were held in spirit at the Royal Central African Museum (Tervuren, Belgium), the Natural History Museum (London, England), The American Museum of Natural History (New York, USA) or the Natural History Museum in Vienna, Austria.
Photographs were taken with a Canon EOS D750 attached to a copy stand to ensure the camera remained perpendicular to the specimens. Photographs of smaller specimens were taken with a 60 mm macro lens. Pedipalps were not detached from specimens, but pedipalps were orientated in a standard position parallel to the camera lens. The pedipalps occupied as much of the camera field of view as possible in order to minimise differences in relative image resolution between species. Binarised outlines were subsequently obtained from photographs by manually tracing the contour of the dorsal surface of the femur and tibia segments in Inkscape. Outlines were then converted to JPEG images in preparation for morphometric analysis.
Linear metrics of pedipalp length and overall body length (taken as a proxy for overall body size) were taken using digital callipers with a measurement precision of 0.01 mm. Relative pedipalp length was defined as pedipalp tibia length divided by body length. Blind repeated measures were taken on a set of Damon variegatus of varying size, the mean measurement error was 2.80% (SD–1.46%, 10 specimens). All specimens were sexed by lifting the genital operculum in order to determine the presence/absence of the spermatophore organ. Any individual listed as juvenile or immature was excluded from the analysis. Wherever possible, linear measurements and photographs were collected from the right pedipalp. However, in instances where the right pedipalp was damaged, the left pedipalp was mirrored and used instead.