Ecological enhancement measures, such as farmland margin restoration, are efficient in addressing biodiversity loss, yet traditional pairwise meta-analyses often fail to comprehensively assess the diverse range of margin interventions available. This study introduces the well-established network meta-analysis (NMA) framework to agroecology to systematically assess the effectiveness of different farmland margins in conserving natural enemies across main grain crops such as wheat, maize, soybean, and rice. Using Bayesian and Frequentist NMA frameworks, we analyzed 20 studies retrieved through systematic literature screening. Both methods consistently ranked treatments as follows: flower strips > grass strips > crops > spontaneous plants > bare fields. However, confidence intervals for most comparisons included zero, indicating the ranking should be viewed as an indicative trend, not a definitive hierarchy. Meanwhile, there are limitations of this model, including the sparse data network, potentially large sampling errors within studies, and limited data. This study demonstrates NMA's potential for agroecology and establishes a framework for restoration-focused comparisons, though it notes that robust application requires additional data and adaptations. • Introducing the NMA framework to agroecology. • The results is only for reference due to limited data. • The application of NMA methods in agroecology faces certain limitations. • More standardized primary data are essential to unlock the full potential of NMA.
Wang et al. (Sun,) studied this question.