Meta-analysis (MA) is increasingly adopted in agricultural and ecological sciences, fields at the interface with plant pathology. Our review on the use of MA in the field, since 1999, identified 79 original research papers. Fifty percent of these works were published in two APS journals, emphasizing their central role in disseminating MA research. Analysis of the number of authors per study and the authorship network revealed a tightly knit field. The network shows a few "hub" authors or groups that significantly influence research directions, with a clear geographical clustering in North and South America. Most of the MA papers focused on the effects of disease control, where fungicide was the main product. The MA investigations often focus on response variables like yield and disease severity for which the analyzed effect sizes differ, but the log of the ratio of means and untransformed means are predominant. The study selection to be included in the MA often follows systematic review standards or when clear quantitative criteria are used for study inclusion. Yet, the inclusion of data from reports, newsletters, and research trial data adds breadth. Frequentist rather than Bayesian inference is most common and SAS and R are the preferred software. Our review underscores the escalating importance of MA in plant pathology, especially for determining the mean and variability of the effects of different control methods on disease and yield. Efforts should continue to improve the detailed reporting of MA methods used, data sharing and reporting of measures of heterogeneity among studies.
Read full abstract