Climate change is predicted to impact tropical rainforests, with droughts becoming more frequent and more severe in some regions. We currently have a poor understanding of how increased drought will change the functioning of tropical rainforest. In particular, tropical rainforest invertebrates, which are numerous and biologically important, may respond to drought in different ways across trophic levels. Ants are a diverse group that carry out important ecosystem processes, shaping ecosystem structure and function through predation and competition, which can influence multiple trophic levels. Hemiptera are a mega-diverse order, abundant in tropical rainforests and are ecologically important. To understand the roles of ants in exerting predation and competition pressure on invertebrates in tropical rainforests during drought and a post-drought period, we established a large-scale ecosystem manipulation experiment in Maliau Basin Conservation Area in Malaysian Borneo, suppressing the activity of ants on four 0.25 ha plots over a two-year period. We sampled hemipterans found in the leaf litter during a drought (July 2015) and a post-drought period (September 2016) period. We found significant shifts in the assemblage of hemipterans sampled from the leaf litter following ant suppression. Specifically, for ant suppression plots, the species richness and abundance of herbivorous hemipterans increased only during the post-drought period. For predatory hemipterans, abundance increased with ant-suppression regardless of drought conditions, and we found marginal evidence for a species richness increase during the post-drought period with little or no change in the drought period. These results illustrate how ants in tropical forests structure invertebrate communities and how these effects may vary with climatic variation.  

We used a set of experimental ant-suppression plots in Maliau Basin Conservation Area in Malaysian Borneo (4.8531° N, 116.8439° E), established by Griffiths et al. (2018) as part of the Biodiversity and Land-Use Impacts (BALI) project. The conservation area is 588 km² of primary rainforest, ranging from 300 m above sea level (asl) to 1675 m asl. The mean annual rainfall for the Basin is 2,838 ± 93 mm (from climate-data.org).

     The experimental design consisted of eight plots located within a 42-ha area. Each plot was 50 m x 50 m with treatment plots having an additional a 15 m buffer zone around each treatment plot. Four plots were treated to suppress ants and four plots served as controls with no treatment (Figure 1). Epigaeic ants were suppressed, starting in October 2014, primarily with Synergy Pro^TM (active ingredients: hydramethylon and pyriproxyfen). These baits are specific: theoily nature of the baits attracts ants, while previous baiting trials have shown no discernible environmental effects (reviewed by Vander Meer et al. 2007). The ants take the bait granule back to their nests, which limits exposure to non-target organisms. Initial poisoning of the ant suppression plots also included cat food baits soaked in imidacloprid, however, the most effective treatment was Synergy Pro, which was used subsequently throughout the experiment. Our suppression method specifically targets ground ants, as baits are scavenged and the active ingredient is consumed within ant nests. We tested for pesticide residue and found no evidence of contamination in seedlings or the soil. Other invertebrate groups were not adversely affected by the application of ant baits (Griffiths et al. 2018). Our suppression treatments reduced ant populations by as much as 93% in treated plots (Griffiths et al. 2018). 

We sampled hemipterans in the leaf-litter using Winkler extractions in July 2015, during a drought and again in September 2016 during a post-drought period. Two, 50 m transects were placed perpendicularly crossing across the middle of each ant-suppression and control plot. Every 7m along each transect, we collected and sieved 1 m² of leaf litter, which was hung in Winkler bags for three days, collecting a total of 15 leaf litter samples per plot. There are several groups of tropical hemipterans that are found in the leaf litter (e.g. Schizopteridae), here we sampled the subset of the total hemipteran community that are readily sampled using Winkler extractions. Although this sampling method may include species that also occur in other layers of the rainforest, we expected that the invertebrates sampled in leaf litter would be the most sensitive to the suppression of ground-layer ants.

We sorted adult hemipterans from samples to morphospecies and classified most morphospecies to either genera, family or suborder, following Capinera (2008). All specimens were identified as either predators or herbivores based on front leg modifications (for grasping and snatching), stylet shape and size (Capinera 2008). 

Related data sets on the effects of ant suppression can be found at: https://doi.org/10.5285/5321bc6e-be35-4ed3-9b56-25598d61ac8f