Suppression of Australian savanna ants shows “the little rulers” do not rule over spider or beetle assemblages

Ants are often referred to as “the little things that rule the world” because of the critical roles they play as ecosystem engineers and through trophic and non-trophic interactions. We describe an experimental test of the influence of ants on spiders and beetles in an Australian tropical savanna. We experimentally suppressed ant abundance through baiting in six plots, each matched with adjacent reference plots, across two sites (Territory Wildlife Park [TWP] and Tropical Ecosystem Research Centre [TERC]), and documented the impact on ground and arboreal spider and beetle communities. At the TWP site, ants were highly abundant and diverse, dominated by species of aggressive Iridomyrmex. Ant diversity was lower and Iridomyrmex low in abundance at the TERC site. At the TWP site, suppression treatment predominantly affected species of Iridomyrmex, reducing overall ant abundance on the ground by 52%–77%, but because few Iridomyrmex forgage arboreally, there was no change on vegetation. Suppression treatment had little effect on ant abundance on both the ground and vegetation at the TERC site. Despite marked ant suppression on the ground at TWP, we observed little response from either spiders or beetles after two years. Among spiders, the only response was a decrease (by 50%) in abundance of the ant-associated family Prodidomidae. One beetle family (Scarabaeidae) also declined after ant suppression but was a suspected nontarget casualty of baiting. No spider or beetle family showed the increase in abundance that we predicted. Curiously, an increase was shown by the spider family Clubionidae at TERC despite minimal ant suppression. At both sites, experimental treatment had little influence on spiders and beetles on vegetation, consistent with the lack of suppression of ants on plants. The general lack of effect of ant suppression implies that there is surprisingly limited regulation of arthropod communities by ants in our study system. This contrasts with findings from other ant manipulation experiments and is particularly surprising as ant abundance is so high in Australian savannas. While our experiment involved ant suppression rather than elimination, our results suggest that Australian savannas are resilient to large reductions of a ubiquitous faunal group.