Artificial Intelligence and Peer Review Capacity in Forestry: A Decade of Evolution and an Introduction to Proposed MCFNS Guidelines for AI-Assisted Review

Abstract

Ten years after an opinionated discussion of traditional peer-review processes and their alternatives was published in this journal, the landscape of scientific publishing in forestry has changed considerably. The fundamental supply-demand imbalance identified in 2015 has worsened: reviewer non-response rates already averaged 47.8% in 2015, and projection models now indicate a demand-to-supply ratio rising from roughly 1.0:1 in 2002 to 4.5:1 in 2024 and a projected 9.4:1 by 2032. Against this backdrop, artificial intelligence (AI) tools have moved from speculative alternatives to active components of many researchers' review workflows, with a 2025 survey reporting that 53% of researchers used AI assistance in peer review (Frontiers Media, 2025). Yet journal-level guidance has lagged adoption: most major forestry journal publishers (Elsevier, Springer Nature, Taylor & Francis) now bar reviewers from uploading manuscripts to AI tools, while Frontiers uses an AI assistant for research integrity assessments, leaving the active middle ground of AI-assisted review largely ungoverned. This paper revisits the four structural premises and four alternative review models discussed in 2015, assesses how each has evolved over the intervening decade, and introduces a new MCFNS section on AI applications in forestry. It also provides context for the accompanying AI-Assisted Review Guidelines, proposed for discussion and voluntary use, summarizes empirical findings on AI versus human review performance (including recent biotechnology-sector evaluation of GPT-5, Qwen-Plus, and Gemini 2.5 Pro across 398 open peer-reviewed preprints), situates the MCFNS approach within the shadow AI versus governed AI framework recently formalized in the operations literature, and presents an anonymized scored example applying the structured framework to a recently declined MCFNS submission. As a demonstration of the framework's applicability, this manuscript was self-assessed using the same B1–B12 rubric prior to publication; the evaluation, including pre- and post-revision scores and a full issue disposition table, is reported in Appendix A. MCFNS, which pioneered open and hybrid peer-review in forestry, is a natural venue to lead a second wave of reform with a structured, evidence-based framework for responsible AI integration.