Reigniting healthy resilience: a trait-based approach to overcoming negative resistance and resilience in stream restoration

In an ecological sense, ‘resistance’ and ‘resilience’ describe the capacity of ecosystems to withstand and recover from disturbance, respectively. These community characteristics are often associated with good ecological health, underpinning restoration goals. However, degraded ecosystems can also be resistant and resilient to disturbance (negative resistance and resilience) making communities highly stable and thus less affected by restoration. First, I investigated this post-restoration scenario (i.e., physical recovery without biological recovery) in freshwater systems using a mesocosm experiment, demonstrating that presence of a persistent degraded invertebrate community can hinder biological recovery. I observed decreases in abundance of sensitive, colonist taxa in the presence of a degraded community, potentially driven by increased competition associated with resource depletion. Reductions in abundances of colonist taxa likely occurred by drift, but also by changes in life history, with Deleatidium spp. mayflies accelerating development to reduce time to emergence, but at a cost of body size. These mechanisms likely perpetuate the negative feedback loop that underpins negative resistance and resilience of degraded communities. Therefore, to facilitate biotic recovery, I propose that degraded communities must first be destabilised. This requires an understanding of how degraded communities are assembled, underpinned by trait- environment relationships. Thus, secondly, I conducted a meta-analysis of New Zealand streams to investigate how environmental gradients influence community assembly, using data across both anthropogenic and natural disturbance gradients. Trait-based ordination of communities using non-metric multidimensional scaling indicated different disturbance gradients work to shape communities along orthogonal axes in trait space. This knowledge can inform restoration by identifying novel disturbances that could be applied to disrupt existing trait-environment relationships, thereby displacing less desired taxa. Finally, I tested this theory in a further channel experiment to investigate how different community types responded to additional disturbance. Response to disturbance depended on disturbance type and community type (defined by disturbance history): if more of the same disturbance was applied, community composition remained relatively stable; if novel environmental filters were applied, community composition changed. Thus, there is potential to predictably alter communities and overcome negative resistance and resilience in restoration by applying different types of disturbances as a tool to direct community recovery and overcome negative resistance and resilience