Mapping linkages between ecosystem services and agricultural and landscape management using structural topic modeling of scientific literature

Abstract

This paper provides a systematic overview of existing literature on the interdependence between terrestrial Ecosystem Services (ES) and agricultural landscape management. To this end, we constructed a structured methodological approach that future researchers and other stakeholders may use to explore potential gaps for potential avenues for research. The framework starts with a constructed bibliometric dataset comprising 2478 paper abstracts (and associated metadata) published between 2010 and 2019 that contain references to specific ES classes and case study location references worldwide. Next, the dataset was used to construct a Structural Topic Model (STM), which features marginal effect estimation of metadata covariates on topic prevalence and topic prevalence correlation. The estimated topic graph structure is then used for community detection of interconnected topics via cluster identification, which allowed us to determine the current structure of the academic discourse on the interaction between terrestrial ES and agricultural landscape management. We expect this novel approach to enable identification of research priorities, prioritization of project portfolios for our field of interest, and reusable methodological guidelines that can be employed to target other fields of knowledge similarly.