How effective are interventions to reduce damage to agricultural crops from herbivorous wild birds and mammals? A systematic review protocol

Abstract Background An important conservation challenge is to mitigate negative impacts that wild birds and mammals can have on human practices and livelihoods, and not least on agricultural crops. Technical interventions to limit the number and severity of damages are available, but evaluations of intervention effectiveness are usually limited in scope, and meta-analyses are rare. This protocol describes a systematic review that seeks to answer the following question: How effective are evaluated interventions in reducing damage from herbivorous wild birds and mammals on agricultural crops? Methods The literature searches are made in the databases Scopus and Zoological Record. The search string is based on a Population-Intervention-Comparator-Outcome (PICO) formatted research question, and search terms fall within five categories: Wildlife type (Population), Damage object (Population), Counteraction (Intervention), Evaluation (Comparator), and Damage (Outcome). Initial scoping searches informed amendment of the search string. A set of 19 benchmark articles were used to estimate the ability of the scoping search to capture relevant literature. To be eligible for inclusion in the review, original articles should study cases where settings of exposure to interventions (measures implemented to reduce damages on agricultural crops caused by terrestrial birds and mammals) are compared to a control setting without exposure to interventions. Eligible studies will be subject to data extraction, systematically documented in an Excel spreadsheet. Associated risk of bias will be critically appraised for the included articles according to seven criteria: 1. risk of confounding biases, 2. risk of post-intervention selection biases, 3. risk of misclassified comparison biases (observational studies only), 4. risk of performance biases (experimental studies only), 5. risk of detection biases, 6. risk of outcome reporting biases, and 7. risk of outcome assessment biases. The results will be reported in narrative and, if possible, quantitative syntheses. The quantitative synthesis will include a summary statistic calculated based on the data of each study and illustrated graphically in a forest plot. If possible, meta-regression analyses will be conducted.