Ecological Vulnerability and Biodiversity Conservation within Agroecosystems: An Integrated Bibliometric Mapping and Systematic Literature Review

<ns5:p>Background Accelerating climate change, habitat fragmentation, and intensive conventional agriculture have elevated ecological vulnerability and biodiversity conservation within agroecosystems into critical global frontiers. Despite expanding literature, a comprehensive synthesis of the intellectual structure and empirical transitions within this domain remains fragmented. This study maps the historical development, global scientific contributions, and thematic trajectories over an 11-year period to bridge the science-policy implementation gap. Methods Combining bibliometric mapping and systematic literature review (SLR) methods, we evaluated publications indexed in the Scopus database between 2015 and 2025. Following a rigorous PRISMA-compliant framework, 96 highly relevant articles were analyzed. Quantitative synthesis and network visualization were performed using the bibliometrix R-package and VOSviewer software to evaluate annual scientific outputs, country contributions, multi-country collaboration networks, and conceptual frameworks. Results The findings reveal an exponential growth in scientific production post-2018, peaking towards 2025, heavily driven by global policies such as the Kunming-Montreal Global Biodiversity Framework (CBD 2022) and UN SDG 15. China (via its Ecological Conservation Redline strategy), the United States, and Brazil emerged as the dominant global contributors. Multi-country publications (MCP) rose sharply from <15% to >40%, led by key China-USA and Europe-Brazil collaborative hubs. Keyword co-occurrence network analysis identified a paradigm shift toward advanced standardization, featuring geodetectors and spatiotemporal frameworks alongside structural indicators like functional traits and functional redundancy. Crucially, the data uncovers persistent limitations, including geographic data asymmetry, a reliance on macro-scale satellite modeling, and a distinct lack of localized plot-level field validation. Conclusions Global literature demonstrates an accelerating shift toward integrated landscape-driven adaptive management. To advance the discipline, future research must transition from broad regional assessments to transnational, multi-taxa ground-truthing field trials that synchronize ecological safety indices with localized stakeholder agricultural practices.</ns5:p>