Valorization of Fish Waste via Anaerobic Digestion: A Systematic Literature Review and Future Research Agenda

Fish processing and aquaculture wastes are protein- and lipid-rich by-products that can be used to produce renewable energy. However, evidence on their anaerobic valorization is limited. This bibliometric and systematic review maps the field and synthesizes evidence on the anaerobic digestion and dark fermentation of fish-derived waste. Scopus records from 2000 to 2025 were screened according to the PRISMA 2020 guidelines. A total of 164 articles comprised the bibliometric corpus and 120 research articles informed the qualitative synthesis. The annual publication growth rate was 13.29%, with 65% of publications occurring between 2019 and 2025. Most described experiments employed laboratory-scale batch assays, mesophilic conditions and co-digestion. Methane yields from fish offal, silage and recirculating aquaculture system sludge ranged from 48 to 1174 mL CH4/g VS. These differences reflect variations in feedstock composition and preparation, proportion of fish waste, selection of co-substrates and operating conditions. The process performance was mainly constrained by ammonia, volatile fatty acids and long-chain fatty acids. The modified Gompertz model predominated, whereas multi-step dynamic modeling remained rare. Microbial studies, primarily 16S rRNA gene surveys conducted in a batch-based manner, linked fish waste digestion to bacteria that degrade proteins and lipids, as well as hydrogenotrophic methanogens. However, community responses depended on the composition of the feedstock, inoculum and loading rate. Only three dark fermentation studies were identified, two of which used fish-derived substrates. Overall, progress toward industrial implementation requires fraction-specific characterization, validation in continuous systems, integration of hydrogen and methane production, dynamic modeling, multi-omics, digestate-safety assessment and integrated techno-economic and life cycle assessment based on pilot- and industrial-scale data. To assess potential database coverage bias, the search was repeated in Scopus and an equivalent search was run in Web of Science. Five additional eligible studies were identified.