Publications

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
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.