Meta-analysis in agronomy

A meta-analysis includes typically the following steps: (i) definition of the objective of the meta-analysis and of the response variable to be estimated from the data, (ii) systematic review of the literature and/or of the dataset reporting values of the response variable, (iii) analysis of data quality (i.e., quality of experimental design and measurement techniques, precision of the response variable), (iv) assessment of between-study variability and heterogeneity, (v) assessment of publication bias, (vi) presentation of the results and of the level of uncertainty.

A considerable amount of experimental data is available from papers published in agronomic journals, and such data could be reviewed, combined and analyzed with statistical techniques to rank cropping systems according to their impact on crop production and on key environmental variables, such as water nitrate content, emission of greenhouse gases (e.g., N₂O) or presence/absence of species of ecological interest (e.g., earthworms, birds). Meta-analysis appears to be a promising approach for assessing the agronomic and environmental performances of cropping systems.

For any questions contact David Makowski:

David.Makowski@inrae.fr