Competitive release of invasive frugivorous mammals leads to perverse outcomes for native seed abundance

Abstract The control or eradication of invasive mammals is crucial for conserving biodiversity in vulnerable island ecosystems, but often targets only a subset of the wider guild of interacting mammals. This selective removal can disrupt trophic dynamics, leading to unintended consequences. In particular, the effects of ‘competitor release’, where one competitor is removed and numbers of the residual (i.e. remaining) competitor species increase, are likely to be substantial but are still little understood compared to the more studied paradigm of ‘mesopredator release’. Opportunistic observations of competitor release following management interventions are increasing, but the outcomes for indigenous biodiversity are still unclear. We selectively removed one or both of two competing invasive mammal species (ship rats [ Rattus rattus ] and brushtail possums [ Trichosurus vulpecula ]) and recorded the effects on the residual competitor and a shared resource (community‐level seed availability) in the cool temperate rainforests of New Zealand. We used a replicated before–after‐control‐impact (BACI) design to measure baseline seed availability and relative abundances of ship rats and possums for 19 months across 8 experimental blocks and then implemented either rat‐only control, possum‐only control, rat and possum control or no control (no treatment) for a further 20 months. Control of either species was associated with increased detections of the untargeted species, providing evidence of competitor release. Combined possum and rat control increased overall seed availability, as did possum‐only control. Rat‐only control had a negative impact on seed availability, likely because of increasing possums. When we analysed each plant taxon separately, either combined control or possum‐only control increased seed availability in five of ten taxa: Beilschmiedia tawa , Fuscospora fusca , Leucopogon fasciculatus , Pectinopitys ferruginea , Pseudopanax spp. Possums appeared to have the greatest impact on seed availability, but rats clearly had additional impacts. Synthesis and applications . Our study provides replicated, experimental evidence that competitor release can result in an unintended and negative conservation outcome: seedfall from indigenous trees decreased when ship rats alone were controlled. Overall, our results support others' recent conclusions that comprehensive pest management strategies are key to maximise the benefits of control.