Small-scale variations in spider and springtail assemblages between termite mounds and the surrounding grassland matrix

The snouted harvester termite ( Trinervitermes trinervoides (Sjöstedt, 1911)) is a widespread grass-eating termite species that constructs thermoregulated dome-shaped mounds. However, little is known about the influence of these mounds on the arthropod assemblage structure in the surrounding grassland matrix, and whether the mounds represent ecological islands. Spiders and springtails are two ecologically important arthropod groups often associated with termites or their mounds. We investigated their assemblage composition inside and around active and abandoned T. trinervoides mounds in a central South African grassland. In total, 838 spiders (59 spp., 22 families) and 217 857 springtails (24 spp., 9 families) were collected from 96 pitfall traps, placed at four microhabitats in and around each of 12 active and 12 abandoned mounds during March 2019. The most abundant and species-rich spider families include the Gnaphosidae (n = 270, 10 spp.), Zodariidae (n = 86, 7 spp.), Lycosidae (n = 86, 6 spp.) and Salticidae (n = 77, 5 spp.), whereas the springtail fauna was dominated by Brachystomellidae (n = 56 521, 1 species), Bourletiellidae (n = 49 573, 7 species), Sminthurididae (n = 44 491, 3 species), Isotomidae (n = 32 288, 1 species) and Entomobryidae (n = 26 216, 7 species). Indicator analysis showed that the spiders Zelotes sclateri Tucker, 1923, Heliocapensis termitophagus (Wesołowska & Haddad, 2002) and Scytodes elizabethae Purcell, 1904 are associated with abandoned mounds, but no springtails showed an association based on the IndVal analysis of the eight microhabitats (lumped data), even though the undescribed Cyphoderus sp. were mostly collected inside active mounds. The mounds thus had a negligible influence on the spatial distribution of springtails in the surrounding grassland. The different spider and springtail assemblages sampled indicate that both active and abandoned mounds function as ecological islands in grasslands, but that mound size does not affect their abundance or species richness in the different microhabitats sampled.