The scientific basis of climate-smart agriculture: a systematic review protocol.

Abstract

Background: 'Climate-smart agriculture' (CSA)-agriculture and food systems that sustainably increase food production, improve resilience (or adaptive capacity) of farming systems, and mitigate climate change when possible-has quickly been integrated into the global development agenda. However, the empirical evidence base for CSA has not been assembled, complicating the transition from CSA concept to concrete actions, and contributing to ideological disagreement among development practitioners. Thus, there is an urgent need to evaluate current knowledge on the effectiveness of CSA to achieve its intended benefits and inform discourse on food, agriculture, and climate change. This systematic review intends to establish the scientific evidence base of CSA practices to inform the next steps in development of agricultural programming and policy. We will evaluate the impact of 73 promising farm-level management practices across five categories (agronomy, agroforestry, livestock, postharvest management, and energy systems) to assess their contributions to the three CSA pillars: (1) agronomic and economic productivity, (2) resilience and adaptive capacity, and (3) climate change mitigation in the developing world. The resulting data will be compiled into a searchable Web-based database and analytical engine that can be used to assess the relative effectiveness and strength of evidence for CSA, as well as identify best-fit practices for specific farming and development contexts. This represents the largest meta-analysis of agricultural practices to date. Methods/Design: This protocol sets out the approach for investigating the question: How do farm-level CSA management practices and technologies affect food production and/or farmers' incomes, resilience/adaptive capacity, and climate change mitigation in farming systems of developing countries? The objective of this ongoing systematic review is to provide a first appraisal of the evidence for CSA practices in order to inform subsequent programming. The review is based on data found in English-language peer-reviewed journals with searches using terms relevant to CSA practices and CSA outcomes. Searches were conducted via Web of Science (WoS) and Scopus. Articles located were screened first by abstract and then full text according to predefined eligibility criteria for inclusion in the review. Data capturing the context of the study (e.g., geographic location, environmental context), management practices, and impacts (e.g., indicators of CSA outcomes) will be compiled from those studies that meet the predetermined criteria. Statistical relationships between practices and impacts will be evaluated via meta-analytical approaches including response ratios and effect sizes. Mechanisms to identify bias and maintain consistency continue to be applied throughout the review process. These analyses will be complemented 5 with an analysis of determinants of/barriers to adoption of promising CSA practices covered in the meta-analysis. Results of the review will be incorporated into a publicly available Web-based database. Data will be publicly available under Creative Commons License in 2016.

Key words