Quantifying ecosystem resilience: catastrophic collapse and recovery of a large river food web

In early July 2013 the impacts of one of the UK's largest pesticide pollution events were first detected by citizen scientists involved in biomonitoring of riverfly populations along the River Kennet, as local extinctions of invertebrate life were reported along 15km of the river's length. The Environment Agency were notified and the source of the spill has so far been tracked back to a water treatment works in Marlborough, where it appears the widely-used organophosphate pesticide Chlorpyrifos was accidentally discharged into the river. Further discharges from the treatment works were immediately suspended: thus we are dealing with a single point source event that has affected over 1/3 of the entire river.

This Urgency Grant is designed to sample the river's biota as comprehensively as possible over multiple scales in space (from local patches to 100m reaches, over 6km of the river's length) and time (from days to one year after the spill) and across multiple levels of biological organization (from genes-to-ecosystems). The samples we will collect will enable us to assess both the impact and initial recovery following the spill. We also have access to extensive citizen scientist biomonitoring data collect by ARK volunteers on a monthly basis from dozens of sites along the river both before and after the impact., which will complement our new intensive sampling. Our sample collection will be spaced through time at approximately logarithmic intervals from the date of the spill, so we can front-load it towards the initial acute and direct phase, while also being able to track chronic and indirect effects as the perturbation ripples through the food web. This unique case study will give us invaluable new insights into of the fragility and resilience of natural systems, and especially of how pesticide effects are felt far beyond those of their usual target organisms (invertebrates in this case). We have selected 6 study reaches (3 upstream and 3 downstream of the spill, separated by 1km intervals), and profound effects are already evident beyond those of invertebrate extinctions: for instance, fish are being left without food and microbes without consumers to keep them in check. The food web is therefore already undergoing profound restructuring, which we predict will continue to intensify in the immediate aftermath of the spill before entering the recovery phase - the pesticide has a very short residence time, so effects being manifested now are the consequences of its extirpation of the invertebrate fauna.

As Urgency grants are limited to 1 year of funding and £65k we are focusing on urgent sample collection and preliminary laboratory processing, with minimal analytical work targeting a subset of samples, dates and sites. We will apply to process the remainder via other funding schemes (e.g. NERC Standard Grants). Our interdisciplinary team will collect an unprecedented breadth and depth of samples: from molecular metasystematics, to quantitative food web characterization and multiple measures of ecosystem functioning. The data will allow us to ask numerous novel questions in both pure and applied ecology, and to investigate the still largely unknown links across organizational levels. For instance, do the microbial biofilms switch on specific genes to process the pesticide and do their blooms when released from control by invertebrates threaten the local fish populations by pushing the system towards anoxia at night as respiration strips the oxygen from the water? Such complex, indirect effects cannot be detected in traditional laboratory ecotoxicological studies - they can only be seen in large field perturbations, but such a comprehensive and co-ordinated coverage of responses to a pesticide spill have never been attempted before. This study will deliver uniquely valuable samples - and ultimately data - to address this glaring gap in our current knowledge of stressor impacts in fresh waters.

The proposed project has the potential to have huge impacts on both the academic (see Academic Beneficiaries) and non-academic communities (see Pathways to Impact attachment and our 5 Letters of Support), as it addresses a fundamental ecological question of growing significance: how do pesticides affect natural systems, both directly (e.g. via toxic effects) and indirectly (e.g. food web effects of invertebrate loss, via starvation of predatory fish and release of microbial resources from top-down control)? At an international level it has clear resonance for bodies involved with both scientific research and its implementation into resource management and policy, and at a national level, where the role of stressors in complex multispecies systems is becoming increasingly evident. For instance, as the "ecosystem approach" becomes applied increasingly widely as a means to link the underlying science more explicitly to future regulatory and management decisions, the need for basic and applied research in this area will continue to grow rapidly.

The main national beneficiaries of this research among the end-user and stakeholder communities include the major UK environmental/conservation agencies (Department of Environment Food & Rural Affairs (DEFRA), Environment Agency (EA), Natural England (NE), Centre for Ecology & Hydrology (CEH), Centre for Environment, Fisheries and Aquaculture Science (CEFAS)), for whom understanding, predicting and mitigating the impacts of stressors in natural ecosystems is an essential remit falling under the 10 year Living With Environmental Change (LWEC) programme (www.nerc.ac.uk/research/programmes/lwec/). In particular, the outputs from our sampling and bioinformatics databases can ultimately be used to inform policy decisions regarding perturbations to natural ecosystems.

The research outputs will also aid other regulatory and legislative end-users in identifying systems that may be particularly sensitive to perturbations (e.g. ARK - see Letter of Support - have highlighted the cocktail of stressors that lowland rivers such as the Kennet are exposed to). These include environmental consultancies (e.g. ENSIS - see Letter of Support), conservation bodies (e.g. Association of Rivers Trusts; Angling Trust; The Wild Trout Trust - see Letters of Support), trust funds and commercial companies (e.g. John Spedan Lewis Foundation - see Letter of Support), as well as industries that emit discharges to aquatic systems, and which are subject to biomonitoring and legislative regulation (e.g. Wessex Water; Thames Water).

Finally, the theroretical developments and new data gathered will be of particular interest to members of the scientific community, including those working in general environmental stressor research outside the current Research Team's remit, as well as those involved in more closely-aligned research networks (e.g. EU RIVFUNCTION, EUROLIMPACS, REFRESH and REBECCA projects; NERC BESS Thematic Programme) - this proposal will open up important and exciting new avenues of both pure and applied research by using the Kennet as an invaluable case study of the devasting impacts of large-scale accidental pesticide spills.

Note: please see also the separate Pathways to Impact attachment for further details on impact and the means of dissemination of our results to both specialist and general audiences.