The research, conducted on the grassland western side of the University of the Free State campus in Bloemfontein, represents the first investigation in South Africa to characterise both spider and springtail assemblages inside and around the mounds of this widespread termite species. Termitaria are a familiar feature of African savannas and grasslands, where they act as nutrient islands and provide refuge, food and spatial resources for a remarkable range of animals, from termitophilous arthropods to mammals and predacious ants. Yet despite decades of research on the ecology of T. trinervoides itself, including its feeding behaviour, mound thermoregulation and the opportunistic occupation of abandoned mounds by vertebrates, relatively little has been known about the arthropods that cohabit the active nests and how their communities connect to the wider grassland matrix.

The team selected twelve active and twelve abandoned mounds, each well established with a height exceeding 25 centimetres and a basal diameter greater than 50 centimetres, to improve comparability. At every mound, four pitfall traps were deployed in distinct microhabitats: one inside the mound structure itself, one on the outer surface at the top, one at the base of the mound, and one three metres away in the natural Themeda triandra-dominated grassland. Because active and abandoned mounds present contrasting internal thermal and structural conditions, the four trap positions of each mound type were treated as separate microhabitats, giving eight in total. Traps were operated for 21 days in March 2019, filled with ethylene glycol preservative, and checked daily. Inserting traps into active mounds required considerable care: a square of the outer crust was cut away, a hole dug into the tunnel structure with an augur, and the crust returned and sealed with damp sand so that the termite workers could repair the breach.

The haul was extraordinary in scale. In total, 838 spiders representing 59 species from 22 families were collected, alongside a staggering 217,857 springtails from just 24 species in nine families. The most abundant and species-rich spider families were the ground-dwelling Gnaphosidae with 270 individuals and ten species, Zodariidae with 86 individuals and seven species, Lycosidae with 86 individuals and six species, and Salticidae with 77 individuals and five species. These four families rank among the six most species-rich spider families in South Africa, and their dominance reflects the pitfall-trapping method, which primarily captures active ground-dwelling predators. The absence of orb-weaving Araneidae and the scarcity of plant-living Thomisidae were expected, since those families live predominantly on vegetation rather than on the soil surface.

The springtail fauna told a different story of numerical dominance by a handful of species. Brachystomellidae led with 56,521 individuals of a single species, followed by Bourletiellidae with 49,573 individuals across seven species, Sminthurididae with 44,491 individuals in three species, Isotomidae with 32,288 individuals of one species, and Entomobryidae with 26,216 individuals in seven species. Among the Entomobryidae, an undescribed species of the genus Cyphoderus accounted for 18,065 individuals, and a remarkable 96.5 percent of these were collected from inside active mounds. This finding is significant because springtails of the tribe Cyphoderini are commonly associated with ants and termites, moving freely within their nests as presumed scavengers, and their abundance is thought to depend on the narrow humidity and temperature ranges that social insect colonies maintain.

Statistical analysis revealed clear patterns in how these communities were distributed. For both spiders and springtails, abundance, species richness and Shannon’s diversity were significantly higher in traps at the base of the mounds and three metres away than inside or on top of the mounds, with all Kruskal-Wallis tests returning P values below 0.001. Interestingly, abundance peaked at the mound base rather than in the open grassland, suggesting that the mounds may act as structural barriers that disrupt the foraging movements of ground-dwelling spiders and concentrate their activity along the mound edge. Non-metric multidimensional scaling analyses, based on Bray-Curtis dissimilarity, showed that the assemblages inside active mounds were distinctly different from those in all other microhabitats for both taxa, with stress values of 0.116 for spiders and 0.046 for springtails indicating good to strong representation of the data.

Indicator species analysis identified three spiders statistically associated with abandoned mounds and their immediate surroundings: the ground spider Zelotes sclateri, the jumping spider Heliocapensis termitophagus, and the spitting spider Scytodes elizabethae. Of these, only H. termitophagus had previously been recognised as specifically associated with abandoned T. trinervoides mounds. No springtail species emerged as a statistically significant indicator of any microhabitat, although similarity percentage analysis showed that the undescribed Cyphoderus species contributed substantially to the differences observed between active and abandoned mounds. The abundance ratio of Cyphoderus between active and abandoned mounds was a striking 147.7 to 1, indicating that these springtails either maintain a relationship with the termites themselves or exploit the stable living conditions created inside active colonies, and may even follow the chemical trails of their hosts when encountered away from nests.

Why were so few spiders found inside the active mounds? The researchers point to the chemical defences of Trinervitermes, whose nasutiform soldiers release repellent terpene compounds when disturbed. These chemicals evidently do not deter particular nest associates such as Cyphoderus springtails or specialist mammalian predators like the aardwolf and aardvark, but the results suggest they may effectively exclude spiders from the mound interior. This could also explain why the mound-specialist jumping spider H. termitophagus is found only in abandoned mounds, where the termites and their chemical arsenal are gone. The abandoned structures, meanwhile, serve as refuges for many arthropods including ants, which in turn attract ant-eating zodariid spiders and ant-mimicking Micaria species among the gnaphosids, making old mounds prey-rich hunting grounds for opportunistic predators.

The study also yielded unexpected taxonomic bonuses. Several new springtail genus records for the Free State Province and South Africa emerged from the sampling, including Orchesellidae genera Capbrya and Nothobrya, the latter recorded in South Africa for the first time, and Mackenziellidae, a family represented worldwide by only a single described species, Mackenziella psocoides. These additions underscore how poorly the springtail fauna of southern Africa remains documented, with only 133 species recorded nationally against roughly 9,500 described globally. The researchers note that undescribed species from their samples will be formally described by specialist taxonomists, contributing a valuable baseline for future work on termitophilous springtails in the region.

The overall conclusion is that both active and abandoned T. trinervoides mounds should be treated as ecological islands within the grassland matrix, since their spider and springtail assemblages differed from those of the surrounding area, even though formal analysis of similarities found no statistically significant differences between microhabitats when tested independently. Mound size, measured as circumference and estimated volume, showed no meaningful effect on abundance or species richness, with most regression R-squared values below 0.1. The authors also acknowledge a sampling limitation: most pitfalls inside active mounds were filled with termites by the end of the 21-day period, likely reducing captures of both target groups, and future studies may need to excavate mounds to sample the deepest nest chambers. Even so, the image that emerges is vivid: millions of springtails and hundreds of spiders distributed across a grassland mosaic, with termite mounds standing as small, chemically guarded islands whose living interiors shelter a specialised springtail fauna and whose crumbling ruins become havens for a very different cast of eight-legged characters.

Subject of Research: Spider and springtail assemblages in and around active and abandoned snouted harvester termite mounds in South African grassland

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

Article References: Badenhorst, H., Haddad, C. R., & Janion-Scheepers, C. (2024). Small-scale variations in spider and springtail assemblages between termite mounds and the surrounding grassland matrix. African Invertebrates, 65(2), 339-367. https://doi.org/10.3897/afrinvertebr.65.139404

Image Credits: AI Generated

DOI: 10.3897/afrinvertebr.65.139404

Keywords: termites, Trinervitermes trinervoides, spiders, springtails, Collembola, ecological islands, grassland, South Africa, pitfall traps, biodiversity, termitophiles, Cyphoderus