Providing the evidence base for environmental risk assessments of novel farm management practices

An environmental risk assessment of a new agricultural management practice depends upon the provision of empirical evidence of cause and effect. This will invariably be derived from comparative experiments testing the null hypothesis that a change in management will have no effect on an assessment endpoint (the metric on which policy decisions will be based). Crucial to the design of these experiments is the answer to the question of 'what to measure?'. The selection of these measurement endpoints and the design of sampling protocols will be determined by the properties of the environmental stressors associated with the change in management practice and the taxa that are exposed to their effects, as well as logistic and financial considerations. The rationale for deciding what to measure in the context of these various criteria is reviewed. For a measurement endpoint to be a valid indicator of the risk of a negative impact of management on the assessment endpoint, a predictable and quantifiable link must be made between the two. It should also be recorded at the appropriate taxonomic resolution to safely assume that all the constituent parts will both respond in a similar way to the management stressor and have a similar effect on the assessment endpoint. Protocols must be designed with the spatial and temporal properties of the management stressor and the measurement endpoint in mind and a consideration of the statistical power of the experiment to detect changes. Where there is a lag in the response time of a measurement endpoint to a stressor due to inertia in the system, an accurate measurement of the effect of the novel management may require experiments running over several years. Throughout, care must be taken that the statistical and biological validity of a sampling regime is not compromised in the face of logistic and financial pressures. The Farm Scale Evaluations of the management of Genetically Modified Herbicide Tolerant crops are presented as a case study to illustrate the concepts discussed. (C) 2008 Elsevier Ltd. All rights reserved.