Data from: Experimental evidence for a hidden network of higher-order interactions in a diverse arthropod community

In this study, we hypothesised that a complex network of higher-order interactions (HOIs), the modulation of a pairwise interaction by a third species, could be constantly modifying pairwise interactions and shaping ecological communities, and that consequently, the outcome of pairwise interactions would be a product of many influences from distinct sources. The study was conducted at Serra do Cipó (19°16'48"S - 43°35'20"W; 1170 m elevation), in southeast Brazil, in the Cerrado biome. The shrub B. dracunculifolia is a self-contained system with diverse arthropod fauna. We created four treatments using this system: (i) Ant Exclusion – all ants were excluded by applying a non-toxic resin to the basal stem of the plants; (ii) Live Gall Exclusion – Baccharopelma dracunculifoliae (Sternorrhyncha: Psyllidae), the commonest galling species found on B. dracunculifolia was removed from the plants by hand collecting the galls; (iii) Hatched Gall Exclusion - All hatched galls (a non-trophic engineering effect) of B. dracunculifoliae were removed by hand collecting the galls; and (iv) Control - no exclusion. Each treatment consisted of 16 replicated plants of B. dracunculifolia. Over two months, every week we quantified the changes in densities of several other species or guilds (hereafter groups; ants, free-feeding herbivores, predators, and aphids) as well as changes in the direct interactions of two galler species, such as gall induction (herbivory), parasitism by wasps, and inquilinism by aphids. Specifically, we combined direct observation and gall dissection data to quantify the effects of groups on each other’s population densities or traits (gall volume and plant size), including direct trophic and non-trophic interactions and the effect of the exclusion of a group on another group (hereafter, density effects). We also investigated how the direct interactions changed in different contexts defined by the density or exclusion of a third group. This allowed the construction of a unique “effect network” based on multiple manipulations performed simultaneously on the same system. Links were categorised into two types: node modulation (node-to-node effects), which are pairwise trophic and non-trophic interactions or density effects; and link modulation (HOIs; node-to-link effects), which are three-way interactions (interaction modification), or four-way interactions (modification of an interaction modification). Most pairwise interactions were affected indirectly by the manipulation of non-interacting taxonomic groups.