TOWARD A SUSTAINABLE AND SUSTAINING LAKE ECONOMY: AN EXPLORATORY FRAMING OF KENYA’S ENABLING ENVIRONMENT FOR AQUACULTURE IN LAKE VICTORIA A Capstone Project Paper Presented to the Faculty of the Graduate School of Cornell University in Partial Fulfillment of the Requirements for the Degree of Master of Professional Studies by Eli Wind Newell August 2025 © 2025 Eli Wind Newell ABSTRACT Cage culture, or fish farming in floating cages, is now an established practice in Lake Victoria, Kenya, and has helped offset declining yields from the lake’s capture fishery. However, cage culture is still a relatively recent phenomenon compared to fishing and is not yet well regulated. Most of the growth in recent years appears to have gone to a few commercial farms while the outcomes for small and medium farmers are more varied. Smallholder aquaculture has significant potential to enhance regional nutrition security and economic growth, yet the sector remains underdeveloped due to numerous persistent challenges. In response, the Lake Victoria Aquaculture Association is leading a concept to facilitate better access to inputs, services, and markets through a coordinated approach. The present study evaluates farmer interest in this concept to inform implementation and contributes to understanding of the systemwide enabling factors supporting or hindering farmers, including policy, cooperation, and trust. 2 BIOGRAPHICAL SKETCH Eli Wind Newell has been drawn to agriculture and food systems for his whole life. Having spent his childhood in his vegetable garden, running his informal neighborhood business Eli’s Eggery, and volunteering at every farm he could, Eli is still driven today by the same purpose and utility of feeding people while stewarding our natural resources. After working for four seasons as an assistant farmer on a farm in Massachusetts and wanting to learn more from agricultural systems abroad, Eli went to Cornell University to study International Agriculture and Rural Development. While at Cornell, Eli joined the lab of Dr. Rebecca Nelson in the School of Integrative Plant Science and the lab of Dr. Eugene Won in the Department of Animal Science. He took a strong interest in circular economy solutions for producing agricultural inputs, where otherwise burdensome wastes are transformed into valuable products. With these mentors, Eli proposed and contributed to research and development of fertilizers, soil amendments, and tilapia feeds derived from waste streams such as urine and seafood processing waste. Eli also completed a fellowship as a Lund Fellow for Regenerative Agriculture at Cornell’s student farm and an internship with Dr. Charles Midega of Poverty & Health Integrated Solutions in Kenya as a Laidlaw Scholar, focused on circular economy opportunities in Western Kenya. Eli has presented his work at multiple international conferences, including his research on urine-derived fertilizers, soilless horticultural media made from crop residues, and aquaculture wastewater as a water and nutrient source in soil-based cropping systems. In 2024, Eli was awarded the Richard A. Church Senior Service Award for outstanding leadership and service to the College of Agriculture and Life Sciences. 3 Dedicated to the future generations who will inherit the consequences of our choices. 4 ACKNOWLEDGMENTS I am deeply grateful to my advisor, Dr. Ndunge Kiiti, for her extraordinary commitment and support, exemplified in the late nights and early mornings on Zoom, the boats and bumpy roads to different fish farms, her thoughtful encouragement, and her generosity with her networks. My world has expanded under your mentorship. Thank you to Mr. Pete Ondeng and Mr. Victor Didi of the Lake Victoria Aquaculture Association for so warmly including me in your work and partnering in this journey. I also thank Dr. Ed Mabaya, Dr. Lori Leonard, and my wonderful mentor Dr. Rebecca Nelson for making it possible for me to join the MPS Program as a Teaching Assistant for Dr. Nelson’s class, FoodCycle: Systems Thinking for Organic Resources. Thank you to my mentors Dr. Eugene Won and Dr. Charles Midega; to my ever- encouraging friends Jefferson Kangacha, Logan Bonn, Trisha Bhujle, and Adam Sharifi; to my family; and to my MPS cohort for your unwavering support. Special thanks to Dr. Eric Teplitz and Dr. Katie Fiorella for sharing your survey data and your insightful guidance. I am grateful to the Toward Sustainability Foundation, the Polson Institute for Global Development, the Mario Einaudi Center for International Studies, and the Department of Global Development for generously funding this work. Most importantly, thank you to all the farmers and other industry experts who gave their time, expertise, and trust. I hope we will continue our conversations. 5 TABLE OF CONTENTS ABSTRACT ……………………………………………………………………. 1 BIOGRAPHICAL SKETCH …………………………………………………… 2 ACKNOWLEDGEMENTS …………………………………………………….. 4 TABLE OF CONTENTS ………………………………………………………. 5 LIST OF FIGURES …………………………………………………………….. 7 LIST OF TABLES ……………………………………………………………… 8 LIST OF ABBREVIATIONS …………………………………………………... 9 CHAPTER 1 – INTRODUCTION ……………………………………………... 11 1.1 Background ……………………………………………………………………… 11 1.2 The Lake Victoria Aquaculture (LVA) Association ………………………… 14 CHAPTER 2 – LITERATURE REVIEW ……………………………………… 16 2.1 Introduction ……………………………………………………………………... 16 2.2 History of Aquaculture in Kenya and Lake Victoria ………………………. 16 2.3 Landscape of Literature ……………………………………………………….. 17 2.4 An Improved Conceptual Framework ……………………………………….. 22 2.5 Blue Transformation Core Concepts ………………………………………… 25 2.5.1 Cooperation, Planning, and Governance ……………………………... 25 2.5.2 Innovative Technology and Management ……………………………... 26 2.5.3 Access to Resources and Services ……………………………………… 26 2.5.4 Resource Use Efficiency and Impact to Environment ……………….. 27 2.5.5 Monitoring and Reporting ………………………………………………. 31 2.5.6 Value Chains ……………………………………………………………… 34 CHAPTER 3 – METHODS …………………………………………………….. 35 3.1 Research Design ………………………………………………………………... 35 3.2 Study Timeframe ………………………………………………………………... 37 3.3 Research Relationship …………………………………………………………. 37 3.4 Site Selection ……………………………………………………………………. 38 3.5 Participant Selection …………………………………………………………… 41 3.6 Data Collection …………………………………………………………………. 41 6 3.7 Supplemental Dataset ………………………………………………………….. 43 3.8 Analytical Methods …………………………………………………………….. 44 3.9 Validation ……………………………………………………………………….. 44 CHAPTER 4 – FINDINGS …………………………………………………….. 45 4.1 Sample Description …………………………………………………………….. 45 4.2 Cooperation, Planning, and Governance …………………………………… 51 4.3 Innovative Technology and Management …………………………………… 54 4.4 Access to Resources and Services ……………………………………………. 56 4.4.1 Financial Resources and Services ……………………………………… 57 4.4.2 Extension Resources and Services ……………………………………... 58 4.5 Resource Use Efficiency and Impact to Environment ……………………... 61 4.6 Monitoring and Reporting …………………………………………………….. 63 4.7 Value Chains ……………………………………………………………………. 66 4.7.1 Upstream …………………………………………………………………... 66 4.7.2 Downstream ……………………………………………………………….. 70 CHAPTER 5 – DISCUSSION ………………………………………………….. 73 5.1 Introduction …………………………………………………………………….. 73 5.2 Supportive Policy ………………………………………………………………. 74 5.2.1 Nutrient Management ……………………………………………………. 75 5.2.1.1 Nutrient Management in Aquaculture ……………………….. 75 5.2.1.2 Nutrient Management Outside of Aquaculture ……………... 76 5.3 Cooperative Governance and Facilitation of Services ……………………. 79 5.4 Trust in Peers and Institutions ……………………………………………….. 84 CHAPTER 6 – CONCLUSION AND NEXT STEPS …………………………. 86 APPENDIX A …………………………………………………………………... 87 APPENDIX B …………………………………………………………………... 88 REFERENCES …………………………………………………………………. 90 7 LIST OF FIGURES Figure 1. Tilapia Cage Culture in Lake Victoria ……………………………….. 13 Figure 2. Output of Scholarly Publications on Lake Victoria Aquaculture by Country ……………………………………………… 17 Figure 3. Institutional Collaboration Network for 30 Most- Published Institutions ………………………………………………… 18 Figure 4. Thematic Map of Lake Victoria Aquaculture Scholarly Research …………………………………………………... 20 Figure 5. Concept Map of Blue Transformation Roadmap …………………….. 21 Figure 6. Improved Conceptual Framework ……………………………………. 23 Figure 7. Visual Concept of Farm-Level Aquaculture Nutrient Mass Balance ………………………………………………………… 33 Figure 8. Research Design Map ………………………………………………… 36 Figure 9. Map of Kenya, Lake Victoria, and Study Area ………………………. 39 Figure 10. Cage Distribution and Site Suitability ………………………………. 40 Figure 11. Interview With Participant in Kisumu ……………………………… 43 Figure 12. Distribution of BMP Information Sources ………………………….. 60 Figure 13. The Varied Fates of Fish Viscera ………………………………….. 63 Figure 14. Distribution of Farm Records Kept …………………………………. 65 Figure 15. Density Plot of Reported Likelihood of Using a Hub Offering Selected Services ………………………………………….. 81 Figure 16. Frequency of Perception About Which Technical Support Areas Would Benefit Farmer ………………………………. 82 8 LIST OF TABLES Table 1. Categorical Descriptive Statistics of Study Participants ……………… 46 Table 2. Continuous Descriptive Statistics of Study Participants ……………… 46 Table 3. Labor Sources …………………………………………………………. 47 Table 4. Ranked Challenges to Farm Operation and Succes …………………… 49 Table 5. Ranked Success Factors and Importance to Operation ……………….. 51 Table 6. Collaborative Counts by Area ………………………………………… 52 Table 7. Rating Collaboration and Cooperation ………………………………... 53 Table 8. Rating Aquaculture Development Policy in Lake Victoria, Kenya …... 54 Table 9. Frequency of Seeking Technical Support or Training ………………… 59 Table 10. BMP Information Sources …………………………………………… 60 Table 11. Experience of Algal Blooms and Upwellings in Past Year ………….. 62 Table 12. Farm Recordkeeping Frequencies …………………………………… 65 Table 13. Feed Sourcing Descriptive Statistics ………………………………… 67 Table 14. Factors Considered in Feed Purchasing Decisions …………………... 69 Table 15. Post-Harvest Storage and Processing ………………………………... 71 Table 16. Reported Likelihood of Using Resource Hubs with Selected Services ……………………………………………….. 80 9 LIST OF ABBREVIATIONS ABD Aquaculture Business Development ADP Adenosine Diphosphate ATP Adenosine Triphosphate BMP Best Management Practice BMU Beach Management Unit CAFO Concentrated Animal Feeding Operation DNA Deoxyribonucleic Acid DO Dissolved Oxygen FAO The Food and Agriculture Organization of the United Nations FCR Feed Conversion Ratio FRRI Fisheries Resources Research Institute (Uganda) HABs Harmful Algal Blooms HDPE High-Density Polyethylene IRB Institutional Review Board KeFS Kenya Fisheries Service KMFRI Kenya Marine and Fisheries Research Institute LVA Lake Victoria Aquaculture MPS Master of Professional Studies NDPES National Pollutant Discharge Elimination System NGO Non-Governmental Organization 10 NMB Nutrient Mass Balance NRCS Natural Resources Conservation Service OCHA United Nations Office for Coordination of Humanitarian Affairs RAS Recirculating Aquaculture System RNA Ribonucleic Acid SAWA Sustainable Activities in Water Areas SFD Shit-Flow Diagram TFRI Tanzania Fisheries Research Institute TN Total Nitrogen U.S. United States UV Ultraviolet WHO World Health Organization 11 CHAPTER 1 – INTRODUCTION 1.1 Background The world’s wild fisheries are fully exploited, with no lasting increase in annual harvest over the last 20 years. As a result, all new seafood demand has been met by expanding aquaculture, or the cultivation of farmed fish, primarily in freshwater lakes and ponds (FAO, 2024). This rapid development of aquaculture, at 6.6% growth per year, or roughly an additional 3 million tons of production worldwide per year, employs close to 20 million people in primary production alone, supporting countless livelihoods and now making up 8-10% of the world’s animal protein supply (FAO, 2024). At the same time, the expansion of freshwater aquaculture has tested these fragile aquatic environments in new ways that can imperil the wild capture fisheries and even compromise the aquaculture system itself (Fiorella, 2023). East Africa’s Lake Victoria is the largest freshwater fishery in the world, producing one million metric tons of catch annually and employing 800,000 people directly and indirectly, with another 3 million dependents relying on their income (Nyamweya et al., 2023). Kenya, which controls 6% of the lake (Tanzania and Uganda control 51% and 43%, respectively) and accounts for about 10% of the annual wild catch, has seen declining yields from the capture fishery since 2006, following a rapid increase through the preceding decade (Njiru et al., 2018; Shitote et al., 2022). While research suggests this decline can be offset by further expanding smallholder cage fish farming (Awuor et al., 2019), a form of aquaculture where fish are farmed in floating cages, 12 there is conflicting information about how that expansion has played out in recent years and who has benefited. While much literature on the development of cage culture (Figure 1) on the Kenyan section of Lake Victoria refers to explosive growth of the sector (Teplitz et al., 2025; Aura & Ntiba, 2024), this growth is not well documented and appears to not have occurred across the sector, which ranges from very small smallholders to very large commercial farms. According to the most recent official State of Aquaculture Report in Kenya published by the Kenya Marine and Fisheries Research Institute (KMFRI) in 2021, aquaculture production rose sharply after cage culture was first introduced in Lake Victoria in 2009. It peaked in 2014 at 24,096 metric tons and, as of 2019, had dropped to 18,542 metric tons (Munguti et al., 2021). Over the course of the present study, several stakeholders, with credible industry knowledge, independently estimated that today’s production has risen to 35,000 metric tons per year, but there does not appear to be an official statistic supporting that figure.1 More recent data on the number of cages, instead of output, show that the number of cages in the Kenyan part of Lake Victoria rose from 1,663 cages across 39 establishments in 2016 to 5,357 cages across 71 establishments in 2020. However, the number has stagnated since then with a drop to 5,242 cages across 127 establishments in 2022 (Aura et al., 2024). The stagnation of cage proliferation coupled with the increase in establishments suggests a high rate of cage abandonment or sale to new entrants. Meanwhile, one very large farm, Victory Farms, is reported to have reached 18,000 metric tons of annual tilapia 1Aura et al. (2024) estimate 21,000 mt total annual tilapia production in 2022, with 17,280 mt coming from large firms. 13 production in 2024 after securing $35 million in Series B venture capital investment (The Fish Site, 2025). This suggests that aquaculture’s growth could be limited to a few successful commercial operations and may not be captured in official statistics even a few years old. Figure 1 Tilapia Cage Culture in Lake Victoria, June 2025. (Author’s photo). Note. Cage frames can be made from wood, metal, or HDPE plastic and are anchored in place with concrete weights. Nets are affixed to the floating frame and extend downward, where they contain the cultivated fish. 14 In recent years, Nile tilapia (Oreochromis niloticus) has become a slim minority of the wild catch from Lake Victoria, making up less than 5% of the total catch lake-wide and around 5,000 metric tons of annual catch in Kenyan waters (Njiru et al., 2018). However, it makes up the lion’s share of aquaculture production. While demand for tilapia has soared in the Kenyan market, farmers have not universally been able to exploit the opportunity even though Lake Victoria stands as a ready host for cage culture enterprises to capitalize on that demand. The challenges to development faced by Lake Victoria’s aquaculture and fisheries sector are many, spanning underdeveloped value chains for inputs like feeds and fingerlings (Obiero et al., 2022), lack of technology and expertise (Munguti et al., 2021), market access challenges (Gichuru et al., 2019), environmental risks (Njiru et al., 2019), and fragmented stakeholder groups (Munguti et al., 2021). 1.2 The Lake Victoria Aquaculture (LVA) Association The Lake Victoria Aquaculture (LVA) Association was established in 2023 to bring the sector together for advocacy and programming that supports the sector’s continued and equitable development. Their mission is quite clear: “to lead private sector coordination, policy advocacy, and reform in Kenya’s aquaculture industry” (Ohala Africa Foundation, 2025). In December 2024, the LVA Association achieved a notable lobbying success with the suspension of a major aquaculture policy (The Fisheries Management and Development Act, 2024) mainly for failure to solicit public participation in the regulation’s development. The present study was developed in this context, under the partnership of the LVA Association and its leadership, to 15 understand the challenges and opportunities small and medium-scale aquaculture farmers experience and how the LVA Association, as a recently formed industry group, could support their commercial success. Thus, the purpose of this research is to support the LVA Association’s effort with the practical goals to (1) inform the LVA Association’s strategic planning, (2) inform the design and evaluate feasibility of county-level resource hubs as a potential program, including documenting farmer interest and concerns, and (3) characterize the spread of farm management approaches and value chain positions. These goals were approached with mixed methods, driven by qualitative analysis of interviews and supported by quantitative description, ultimately contributing to the underlying intellectual goals to (1) understand what factors farmers consider to be important to their success, and what they consider to be holding them back, and (2) understand Kenya’s Lake Victoria aquaculture context, including the state and nature of the enabling environment in which aquaculture enterprises develop and operate. 16 CHAPTER 2 – LITERATURE REVIEW 2.1 Introduction Literature review supports this study by helping to articulate existing theory of aquaculture development in the Lake Victoria region while identifying gaps in research and theory. 2.2 History of Aquaculture in Kenya and Lake Victoria Lake-based aquaculture in Kenya traces to the decline of the capture fishery. This decline was caused by overfishing and the introduction of invasive Nile perch (Lates niloticus), which outcompeted many of the native tilapia and other cichlids (McGee et al., 2015). Smallholder and commercial cage farming of Nile tilapia (Oreochromis niloticus) took off in the early 2010s but has suffered setbacks including fish kills in recent years that have been linked to eutrophication and upwelling (Aura & Ntiba, 2024). The growth of cage culture from 2009 has occurred at the same time as its governance was reshaped by Kenya’s 2010 Constitution and subsequent laws that devolved many agriculture and fisheries functions from national to county-level jurisdiction. Still, national agencies, like Kenya Marine and Fisheries Research Institute (KMFRI) and Kenya Fisheries Service (KeFS), support the sector. 17 2.3 Landscape of Literature: Study Themes and Collaborators Driving Lake Victoria Aquaculture Research A bibliometric analysis of 218 publications, identified on August 1, 2025 in the Scopus database using the advanced query below, shows that while Kenya controls only 6% of Lake Victoria, the country dominates the scientific discourse on aquaculture in this region. Kenya's single-country publications (publications for which all author affiliations are in the same country) alone exceed the total output of scientific publications from any other country (Figure 2). TITLE-ABS-KEY ( "Lake Victoria" ) AND (TITLE-ABS-KEY ( "aquaculture" OR "cage culture" OR "fish farming" OR "fish farms" OR "fish farm" OR "fish farmer" OR "fish farmers" OR "tilapia" )) Figure 2 Output of Scholarly Publications on Lake Victoria Aquaculture by Country Note: Only includes publications with available country information. Figure processed by author in R Studio with the Bibliometrix package (Aria & Cuccurullo, 2017). 18 For publications with authors from multiple institutions, a network analysis of the 30 most prolific institutions (Figure 3) shows that the public fisheries research institutes from Kenya (KMFRI), Uganda (Fisheries Resources Research Institute-FRRI), and Tanzania (Tanzania Fisheries Research Institute-TFRI) form the center of the network, publishing more than most other institutions and collaborating frequently with each other. Still, KMFRI stands out as the most prolific and collaborative while the large contributions of prominent universities in the Lake Victoria Basin are also notable (Makerere University, Uganda; Moi University, Kenya; University of Eldoret, Kenya; Egerton University, Kenya; University of Nairobi, Kenya; Maseno University, Kenya; and Sokoine University, Tanzania). Figure 3 Institutional Collaboration Network for 30 Most-Published Institutions 19 Note: Processed by author in R Studio with the Bibliometrix package (Aria & Cuccurullo, 2017). Affiliations from the same institution (e.g., different departments from the same university) were grouped together, as were institutions with multiple spellings (e.g., “KMFRI,” “Kenya Marine and Fisheries Research Institute,” and “Kenya Marine Fisheries Research Institute”). Using author-provided keywords and Fuzzy Sets theory, a map is developed (Figure 4) to show the thematic structure and evolution of the scholarly literature on Lake Victoria aquaculture’s subthemes, based on the frequency with which similar keywords appear together (Cobo et al., 2011). Haplochromines (non-tilapia cichlids), Nile tilapia, Lake Victoria, gene flow, and economics are determined to be “motor themes,” meaning they are highly developed as their own research areas and are also highly connected to other research areas. FliC (a protein associated with a specific pathogen), cholesterol, feeding patterns, fish production, and biomass are determined to be “niche themes,” meaning they are highly developed as their own research areas but are not researched in association with other themes. Archaea, cage fish farming, freshwater, cichlids, membrane bioreactor, and tilapia lake virus are determined to be “emerging or declining themes,” meaning they appear frequently enough to emerge as recurring themes but are neither well developed as their own areas of research nor well connected to other areas. Lastly, cichlidae, bacillariophyta (diatoms), fish farming, and blue economy are determined to be “basic themes” meaning they appear frequently in connection to other themes in the literature but are not well developed as their own themes (Cobo et al., 2011). 20 Figure 4 Thematic Map of Field of Lake Victoria Aquaculture Scholarly Research Note: Processed by author in R Studio with the Bibliometrix package (Aria & Cuccurullo, 2017). Density refers to the frequency with which keywords within a given subtheme appear together and centrality refers to the frequency with which keywords with a given subtheme appear with keywords in other subthemes. Circle size represents number of publications. In addition to scientific research on aquaculture in Lake Victoria, this literature review also draws from the planning literature. The Blue Transformation Roadmap 2022- 2030 produced by the Food and Agriculture Organization (FAO) of the United Nations presents five targets for aquaculture development to produce aquatic food in a 21 way that contributes to food supply, is environmentally sustainable, and equitably benefits stakeholders. The roadmap lays out additional targets for fisheries and value chains and describes priority actions for each. The areas of the roadmap relevant to this study are reconfigured into a concept map (Figure 5), where core concepts are adapted from the Blue Transformation targets and contributing concepts and intermediate outcomes are adapted from the priority actions. Figure 5 Concept Map of Blue Transformation Roadmap Note: Author interpretation of adapted targets and actions from Blue Transformation Roadmap 2022-2030 (FAO, 2022). 22 2.4 An Improved Conceptual Framework The challenges of expensive inputs, lack of technology and expertise, market access, environmental risks, and fragmentation are known independently. This study contributes to existing knowledge by evaluating these challenges in relation to each other, and in relation to a larger picture of the aquaculture value chain. For example, Lake Victoria aquaculture has a major constraint not captured in the FAO Blue Transformation Roadmap which is that the FAO framework only considers the post- harvest value chain. This study contributes to the FAO framework by incorporating the pre-production value chain (feeds and fingerlings) as well, in addition to several important factors relating to the broader systemwide enabling environment (Figure 6). 23 Figure 6 Improved Conceptual Framework Management of environmental and biosecurity risks to aquaculture is a key part of farm management, but these risks are also influenced by factors outside of a farmer’s control. Existing literature suggests nutrient pollution, harmful algal blooms, water column inversions, and pathogen transmission are major risks that need to be managed. Interview data illuminate additional risks such as wind which can move cages and overturn brooder nets, rising water levels which can erode lakeside hatchery ponds, water hyacinth rafts which can smother cages, heavy rains which can make roads impassable for receipt of feeds and sale of fish, and theft and fisherman interference as additional environmental and security risks. Several of these risks to 24 aquaculture, including nutrient pollution and pathogen transmission, are also exacerbated by aquaculture itself and can thus be minimized by farm management informed by those risks. Access to high-quality, affordable inputs contributes to farm management and is a key part of the enabling environment for individual farms and the sector as a whole to succeed. Access to such inputs requires that those inputs exist in the first place, which requires a functioning supply chain (or hatchery capacity in the case of fingerlings), uninterrupted manufacturing or import schedules, and testing to guarantee quality. However, it is not enough for the inputs to just exist; farmers need to be able to access the inputs. This access is often determined by ability to secure credit or other financial services. Access to financial services is also a key component when it comes to investment in farm infrastructure for production, processing, or transport. As with any technical industry, access to technology and technical expertise contributes to the success of a farm or the industry. Again, this depends on the existence of such resources, which requires investment in developing technology and technical knowledge. To access these resources, financial services need to be in place for farmers or groups of farmers to pay for them, or they need to be provided by public-private partnerships or public entities with the institutional infrastructure and social capital to deliver. This core concept also informs farm management and contributes to the enabling environment for a reliable feed supply. 25 Market access is important for farmers so they can sell their fish when they want to, without risking losses. Again, just for markets or demand to exist is not enough— farmers need to be able to access the market to avoid losses and justify growth. Lastly, the development and implementation of record keeping instruments and data- driven farm management tools is an intermediate outcome that supports profitable, environmentally responsible farm management. Adoption of such tools requires trust in the implementing institutions, and the benefits of their use are greatest if farms cooperate in data-aggregation. For this intermediate outcome in particular, and for the broader model in general, supportive policy, cooperative governance and facilitation of services, and trust in peers and institutions are the underlying criteria that enable the arrows to move forward. Further, supportive policy emerges from industry-derived advocacy and lobbying as has been shown by LVA’s December 2024 action. 2.5 Blue Transformation Core Concepts While many of the concepts and relationships between them in the FAO’s Blue Transformation Roadmap are underdeveloped, under-connected, or largely unrepresented in the Lake Victoria aquaculture literature, they are still explored in this section. 2.5.1 Cooperation, Planning, and Governance The absence of organized producer groups has been noted as a barrier to achieving economies of scale and accessing credit or inputs and the FAO identifies regional cooperation as a key target area for a successful blue transformation. Munguti et al. 26 (2021) similarly identify greater cooperation as an important factor for the success of aquaculture development in Kenya. Supporting this position, Gichuki et al. (2025) find that tilapia farmers, in this case practicing pond culture instead of cage culture, benefited from joining cooperative or common interest groups. These farmers saw their yields increase by more than 6% last year, which corresponded to a more than 30% increase in sales. 2.5.2 Innovative Technology and Management Many farmers or investors aspiring to become farmers lack technical skills for effective aquaculture management. A 2021 report by KMFRI identifies low expertise and technical know-how as primary impediments to aquaculture’s success, especially in the areas of feed formulation and fish disease management (Munguti et al., 2021). 2.5.3 Access to Resources and Services Access to high quality, affordable, and sustainable inputs is a persistent challenge for monogastric livestock because they have a high dietary protein requirement. In the case of the tilapia aquaculture that is most prevalent in Lake Victoria, fish feed alone accounts for over 60% of production costs with fingerlings accounting for another 33% in some cases (Obiero et al., 2022). Farmers in the region have expressed frustration due to the high cost of feeds. This has led them to advocate for the establishment of a local industry to produce feeds less expensively (Miima et al., 2023). Often, these input costs and other barriers to resources and services are prohibitive for smallholders, especially women. In a study of cage culture adoption in 27 Siaya County, Orinda et al. (2021) found that just over three percent of cage culture adopters were women. 2.5.4 Resource Use Efficiency and Impact to Environment The cautionary tales of the Central American great lakes show how large cage-culture developments, which in their case produce much of the fresh tilapia eaten in the Western Hemisphere, can utterly change a lake’s water quality and ecology, and undermine other uses of the lake (Fadum et al., 2025; Fadum & Hall, 2022; McCrary et al., 2007). Unlike these Central American lakes where the aquaculture presence is consolidated to a few large companies, the aquaculture presence in Lake Victoria is attributable to hundreds or thousands of largely uncoordinated smallholders resulting in a tragedy of the commons. Fish excreta and uneaten feed contribute to Lake Victoria’s precariously high nutrient load, and in turn, farmers are subject to algal blooms that trigger devastating fish kills in addition to the water column inversions that occasionally trigger the same (Olokotum et al., 2020; Orinda et al., 2021). As the limiting nutrient in most lakes (Howarth et al., 2021), phosphorus requires special attention. Phosphorus is a critical element to some of life’s most essential structures and compounds including cell membranes, RNA, DNA, and ADP/ATP. Since every cell is necessarily surrounded by a phosphorus-laden membrane, an organism that cannot get more phosphorus cannot sustainably make more cells. It is easy to imagine why crops can be stunted by inadequate phosphorus availability, and even easier to imagine why unicellular organisms like algae simply cannot proliferate without it. 28 While any algal bloom and decomposition can create risks for those who depend on the lake, cyanobacteria are uniquely threatening. Many cyanobacteria, but most importantly those of the genera Anabaena (also known as Dolichospermum) and Microcystis, produce a class of secondary metabolites called microcystin (Olokutom et al., 2020; Sitoki et al., 2012). Microcystin is a hepatotoxin that can cause illness by inhibiting phosphatase (Svirčev et al. 2010) and is itself resistant to being broken down enzymatically (Somdee et al., 2013). The proliferation of these toxins in the Lake Victoria basin and specifically around the populated gulfs affects people as well as aquatic life. Microcystin can accumulate in fish such as the silver cyprinid Rastrineobola argentea, also known as “omena” or “dagaa,” which are eaten whole, including the liver where the microcystin concentrates. Microcystin levels in these fish, often consumed by children, have been recorded at five to ten times World Health Organization (WHO)-acceptable levels (Roegner et al., 2023). Microcystin is an alarming problem and is also a relatively recent problem. 2 Krienitz et al. (2002) 2 Until the 1980s, diatoms of several genera, primarily Aulacoseira, Cyclostephanos, and Nitzchia, dominated Lake Victoria’s algal community (Njagi et al., 2022). Since then, diatoms, which do not produce microcystin, have been outcompeted by cyanobacteria for three reasons, two of which are attributable to soaring phosphorus loading to the lake. First, diatoms have cell walls formed from silica (𝑆𝑖𝑂2 ∙ 𝑛𝐻2𝑂), meaning they need abundant dissolved silicon available to them in their aquatic environment, or at least relative to their other nutrient requirements. As phosphorus enrichment from external sources accelerated, silicon became the limiting factor for diatoms. Diatom communities fertilized by newly abundant phosphorus grew quickly, immobilizing available silicon, which would be remobilized after the diatoms died and fell to the lake floor (Hecky et al., 2010). Over five decades, silicon in Lake Victoria became both limiting and non-renewable. Pacini et al. (2018) measured a stark decline in soluble reactive silica in Kenya’s Lake Naivasha from 14-15.7 mg l-1 in the 1960s to a mere 1-2 mg l-1 in 2015, and they note that Lake Victoria has seen similar trends. Hecky et al. (2010) corroborate this claim, not only showing that dissolved silicon declined at inshore measurement stations in Lake Victoria from the 1960s to the 1990s but also that this trend was reflected by lake sediments, showing an accumulation of biogenic silicon during the same period. As phosphorus enrichment and resulting silicon depletion disadvantaged the diatoms, the cyanobacteria had yet another advantage over the diatoms—nitrogen fixation. 29 were among the first to report cyanotoxic algal blooms in Lake Victoria just two decades ago (Sitoki et al., 2012). With rapid growth of Lake Victoria’s cyanobacteria population relative to its diatom population and abundant phosphorus, all algae began to compete for nitrogen. Cyanobacteria, many of which can fix their own nitrogen, are not limited by nitrogen availability. Of particular importance in Lake Victoria is the Anabaena genus of nitrogen-fixing cyanobacteria, contributing to a two- to ten-fold increase in primary production since the 20th century (Guya, 2020). The extent to which cyanobacteria overcome nitrogen limitation is strikingly illustrated by the assessment by Hecky et al. (2010) that as much as two-thirds of the total nitrogen (TN) in the system is fixed from the atmosphere by cyanobacteria. For nitrogen-fixing cyanobacteria, phosphorus is the limiting nutrient, which has become abundant in Lake Victoria (Hecky et al., 2010). Lake Victoria’s exceptionally low total nitrogen to total phosphorus ratio relative to other lakes, combined with an exceptionally high nitrogen fixation rate, greatly advantage nitrogen-fixing cyanobacteria and disadvantage non-fixing diatoms. Consequently, Lake Victoria diatoms are now mostly silicon- and nitrogen-limited while the cyanobacteria are phosphorus-limited (Guya, 2020). Overwhelmingly, the shift from diatoms to cyanobacterial dominance is attributed to nitrogen fixation and silicon depletion, both driven by phosphorus enrichment (Hecky et al., 2010). Total phosphorus in Lake Victoria’s Nyanza Gulf doubled from 2000 to 2006 (Sitoki et al., 2012) and soluble phosphorus increased more than ten-fold from 4 µg l-1 in 1990 to 57 µg l-1 in 2008 (Juma et al., 2014), a shift driven by fluvial (river- 30 carried) nutrient loading from agricultural and urban areas (Njagi et al., 2022; Sitoki et al., 2012). The result has been larger, more frequent, and more toxic algal blooms. Phosphorus’s importance to life and its significant flow into Lake Victoria have transformed the algal communities and now imperil the region’s fisheries and livelihoods. As a result, Lake Victoria’s carrying capacity for aquaculture activities should account for the phosphorus burden they pose in the form of fish excreta and sinking uneaten feed. Aura et al. (2024) calculate the carrying capacity for the most suitable part of Kenya’s section of Lake Victoria, an area of just 190km2, or 4.6% of Kenya’s Lake Victoria Territory. Using this zoning, they estimate the carrying capacity at 109,000 metric tons of tilapia production annually. Considering this number from a phosphorus loading standpoint, producing that much tilapia would lead to between 882 and 2,147 metric tons of external phosphorus loading to Lake Victoria annually. This estimate assumes tilapia feeds are 1.2-1.7% phosphorus by dry weight (Chatvijitkul et al, 2018), that Nile tilapia are 0.75% P on a fresh weight basis (Boyd et al., 2007), and Nile tilapia farms in Lake Victoria achieve a Feed Conversion Ratio (FCR) between 1.3-1.6 (Musa et al., 2022a). The best-case scenario would be equivalent to Kenya’s 2002 levels of phosphorus pollution from all urban wastewater and runoff, and the worst-case scenario would be equivalent to more than all 1,925 mt of P estimated to have reached Lake Victoria from all of Kenya’s Lake Victoria-bound river basins in 2002 (Kayambo & Jorgensen, 2005). Using Kenya-specific values for P loading from urban wastewater and runoff, industrial loading and river flows, and estimating Kenya’s portion of runoff and atmospheric deposition at 6% of the lake’s total, 2002 total phosphorus loading to the Kenyan part of Lake Victoria was 4,657 mt 31 (Kayombo & Jorgensen, 2005). The worst-case scenario under 109,000 mt annual production, therefore, represents a nearly 50% increase in phosphorus loading compared to total annual loading from all sources before aquaculture was established in Lake Victoria. Considering this scenario and the widespread establishment of aquaculture operation in zones determined to be unsuitable for aquaculture (Aura et al., 2021; SAWA, n.d.), the true carrying capacity is probably lower than the 109,000 mt estimate. Already, the majority of farmers experience algal blooms; in a 2023 survey by Teplitz et al. (2024) of 172 farmers, 140 (81%) said they had experienced algal blooms at their cages, of which 61 said their fish were affected. Though not always associated with algal blooms, mass fish kills within the prior year also affected 40 percent of the farmers Teplitz et al. (2024) surveyed. 2.5.5 Monitoring and Reporting Lakes are liable to severe ecological change from aquaculture-derived nutrient loading, ultimately threatening both capture and culture fisheries. Lake Victoria specifically has suffered from harmful algal blooms (HABs) and fish kills exacerbated by excessive nutrient loading. While Lake Victoria’s nutrient loading comes from multiple sources in addition to aquaculture, aquaculture’s contribution should be monitored and should not be overlooked. In Honduras’s Lake Yojoa, for example, cage culture has been shown to contribute 85% of the total N loading to the lake and 95% of the total P loading (Fadum et al., 2025), even as the company responsible for Lake Yojoa’s aquaculture has made major commitments and earned multiple certifications for sustainability (Fadum et al., 2025). 32 Unlike terrestrial farming, where farmers operate on owned or leased land, cage farmers operate in public waters. The activity of one affects a common resource, affecting all. Each farm poses a risk to the environment and to each other, which must be managed. Concentrated animal feeding operations (CAFOs) in the United States (U.S.) are similarly recognized for their nutrient pollution risk, which farms mitigate with nutrient management planning informed by whole-farm nutrient mass balance (NMB) calculations (Soberon et al., 2013). Additionally, NMB assessment informs better and often more profitable management. This assessment offers insight into how farms can more efficiently use their inputs by identifying where nutrient imports are excessive or nutrient exports are insufficient. New York dairy farms participating in NMB assessment have been found to meaningfully improve their nutrient balances over the course of several years beyond their first assessment year (Cela et al., 2015; Soberon et al., 2015). Cage culture nutrient mass balances have been advocated and modeled for several species and geographies (Song et al., 2023; Suresh et al., 2023; Chary et al. 2022; Cai et al., 2016; Bureau & Hua, 2010), including tilapia (Oreochromis niloticus) in Lake Victoria (Musa et al., 2023b), but no assessment has been developed and studied as a practical tool for farmers. NMB assessment is especially important for aquaculture because the unaccounted losses are unseen and expensive. Feed inputs alone account for the majority of farm expenses, so any feed that sinks uneaten is both wasteful and polluting (Yuan et al., 2024). Simply transitioning from pelleted feed to extruded feed can greatly improve 33 feed utilization, reducing economic and environmental costs (Musa et al., 2023a). However, this benefit, for example, might not be apparent to a farmer who doesn’t track nutrient balances. More drastic management changes like shifting cages between production cycles can also confer benefits (Musa et al., 2022b), but whether it is worth doing that might depend on the farm’s nutrient balance. A visual summary of the parameters an aquaculture-specific NMB assessment might account for is shown in Figure 7. Figure 7 Visual Concept of Farm-Level Aquaculture Nutrient Mass Balance Note: Author’s interpretation of calculator described by Soberon et al. (2013), adapted to a cage culture application. Nutrient mass balances refer to a calculation of 34 “nutrients in” minus “nutrients out” and can be normalized by the farm’s total productive output. This way, peer farms can be compared to each other. Excessively large balances indicate low efficiency and high unaccounted losses. The example of nutrient pollution management from New York dairy farms, where farmers track their nutrient balances and are able to compare their farm’s performance with their anonymized peers, is a model for data-driven management that relies on sound record-keeping to improve sustainability and profitability (Van Almelo et al., 2016). 2.5.6 Value Chains Farmers face challenges in selling their fish profitably because fish are perishable and cold-chain storage and transportation infrastructure remains insufficient. A 2019 study by KMFRI researchers found that 20-40% of Lake Victoria’s harvest was lost post- harvest because of an absence of preserving technologies like those that make up a cold chain (Gichuru et al., 2019). 35 CHAPTER 3 – METHODS 3.1 Research Design This study employed numerous methods informed by the project’s goals, research questions, and conceptual framework, which are all reciprocally influenced by each other. Additionally, they take into account any anticipated validity threats that could undermine the credibility of the data and analysis. The underlying research paradigm, or worldview about “what we can know about our world (ontology) [and] how we can know it (epistemology)” that guides these methodological decisions (Mayan, 2023), borrows from two established paradigms. First, the pragmatic paradigm informs this practical mix of methods and the deliberate selection of participants covering a variety of perspectives (Mayan, 2023). Second, this research borrows from the postpositivist or neopositivist paradigm; the interpretative description of the phenomena studied is not explicitly focused on causal relationships as in a positivist paradigm and instead recognizes the need to for qualitative data to explain processes that are not captured in numbers alone (Mayan, 2023). Following this paradigm, the qualitative analysis draws primarily from participant’s explicit invocation and discussion of key phenomena instead of relying on easily biased inductive approaches (Roulston, 2014). The overall study design is represented in Figure 8, using a graphical layout and study design exercises adapted from Maxwell (2013). The research design was responsive to early interviews, where the concept of “enabling environments” emerged as an essential theme to explore, leading to the newly focused research questions on trust, 36 policy, and cooperation. This approach, where the research focus is allowed to respond to an evolving conceptual framework and learnings, is represented by the dashed arrows in Figure 8. Figure 8 Research Design Map Note: Graphical layout adapted from Maxwell (2013). 37 3.2 Study Timeframe The bulk of the data collection process took place in January 2025. Selected stakeholders were engaged through a combination of methods. Structured survey questions were designed to produce quantitative data, while observation and semi- structured interview questions accompanying the survey, as part of the same Cornell IRB-exempted questionnaire, produced qualitative data. A preliminary literature review informed the structured survey questions, but the literature review was further enriched and developed in the process of analyzing, interpreting, and contextualizing the data obtained. Bibliometric analysis was also used to quantitatively assess the field of academic literature on the subject of aquaculture in Lake Victoria published since 1951, though 88% of publications are since 2000. Lastly, the data generated through each of these processes were analyzed with coding, categorization, and descriptive statistics, as relevant. 3.3 Research Relationship The present study was guided by the strategic planning and program design needs of the LVA Association. In Fall 2024, LVA Board Secretary Pete Ondeng solicited the assistance of a Master of Professional Studies (MPS) student from Cornell’s Department of Global Development via Prof. Ndunge Kiiti, the author’s advisor. The author, Mr. Ondeng, and Dr. Kiiti met regularly on Zoom from October to December, 2024 to build rapport and plan this study. 38 While no relationship was built over time between the author and the participants prior to the administration of the surveys, the participants or their Beach Management Unit (BMU; Obiero, 2015) chairmen, where relevant, were known either to Mr. Ondeng or to LVA associate Victor Didi, who is also a fish farmer and a member of the BMU at Dunga Beach in Kisumu. Each interview was conducted either by the author or by Mr. Didi; Dr. Kiiti assisted once when it was necessary because of limited time. Mr. Didi provided additional translation support where needed on all interviews, for any questions that required elaboration or explanation in Luo (local language) or Kiswahili (national language). 3.4 Site Selection Kenya’s primary administrative units are counties, of which five border Lake Victoria. Of these five counties, Siaya, Kisumu, and Homa Bay were selected for the study (Figure 9) because most of the cage culture occurs in these counties (Aura et al., 2024) and because of proximity to the research team. The largest population center on the Kenyan part of Lake Victoria is the city of Kisumu, which appears to be the primary market or at least initial destination for tilapia produced across the five counties. A more detailed map of cage culture density in the lake and human population density on land is included in Appendix A. 39 Figure 9 Map of Kenya, Lake Victoria, and Study Area Within Kenya’s Five Riparian Counties Note: Rendered in QGIS by author. Basemap from ESRI (Canvas/World_Light_Gray_Base (MapServer), n.d.); Kenya administrative boundaries from OCHA Regional Office for Southern and Eastern Africa (ROSEA) (2023); and Lake Victoria shapefile from Africa Geoportal (Frank, 2024). The density and distribution of tilapia cages in Lake Victoria are shown in Figure 10 along with a site suitability map for aquaculture development produced by the Sustainable Activities in Water Areas (SAWA) initiative in partnership between the Government of Kenya, Gatsby Africa, Longline Environment, and ThinkAqua, a British NGO focused on smallholder aquaculture development. 40 Figure 10 Cage Distribution and Site Suitability A B Note: A. Cage density and distribution as of 2022 (Aura et al., 2024). B Aquaculture development site suitability map accounting for shipping lanes, protected conservation areas, and other parameters (SAWA, n.d.). By spanning the three most invested of the five riparian counties, this study covers a wide range of farm scales and proximity to markets while still serving the stakeholders best positioned to viably implement a strategic hub for aggregating and accessing resources as proposed by the LVA Association. 41 3.5 Participant Selection Participants for this study were chosen using purposeful selection (Maxwell, 2013) in accordance with Mayan’s (2023) sample size guidelines for an interpretive description study. The goals of purposeful selection, in this case, are to (1) solicit perspectives from specific stakeholder groups with which the LVA Association would prioritize engagement in future programming, and more importantly, (2) to capture the remarkable heterogeneity of the sector. In a 2023 survey of 172 cage farmers across the five riparian counties, Teplitz et al. (2024) collected extensive data on farm characteristics and management practices. They applied K-means and latent class analysis to group farms but found that the resulting clusters did not meaningfully explain variation in production metrics. This suggests that farms are heterogenous in both characteristics and performance, and that these differences are not well captured by cluster structures. In other words, the sector is highly heterogeneous. Survey participants and farm tours for this study were therefore selected purposefully to cover some of the sector’s heterogeneity. 3.6 Data Collection Overall, the study participants are stakeholders known to the LVA Association and are actively involved in the aquaculture sector. For smallholder and commercial aquaculture actors, this means rearing fish. For suppliers, this means either selling aquaculture-specific products or having aquaculture-specific clients. LVA Association determined eligible participants prior to active recruitment. Once potential 42 interviewees were identified, each participant was given an overview of the study objectives, what data would be collected, and how those data will be used. Any questions were then taken, and potential participants were given the chance to consider whether they wanted to participate or not. Then, participants were explicitly asked if they agreed to participate in the research according to Cornell’s IRB protocols for verbal consent. Interviews focused on (1) the quantity, cost, and sources of inputs (feed, fingerlings) they use; (2) experience with technology and accessing technical expertise (e.g. about feed formulation and fish disease management); (3) experience accessing markets; (4) perceptions of and experience with environmental risks, including harmful algal blooms and water column inversions; and (5) perceptions of and attitudes toward stakeholder fragmentation and cooperation, though other topics also emerged during interviews. Research data were recorded on a paper survey form through interviews (Figure 11) and were subsequently recorded in a digital Microsoft Excel spreadsheet for analysis. The research interview was guided by a prepared, printed questionnaire and participants were informed that their responses would be recorded by taking notes on the paper questionnaire. Participants were also informed that some of the interview questions were open-ended, meaning their answers may not be easily captured by taking notes alone, and all or part of the interview would therefore be recorded with audio recording to capture complete responses. Additionally, audio recordings and 43 interview notes were subsequently replaced with transcripts as text files and audio files were permanently deleted after being transcribed. Figure 11 Interview With Participant in Kisumu 3.7 Supplemental Dataset While the small sample size is appropriate for a pilot study and for describing the study’s core concepts and a range of stakeholder experiences with those concepts, a much larger sample is required to make generalizable claims about specific farm practices across the aquaculture sector. Using data shared from the Teplitz et al. 44 (2024) survey, the present study includes descriptive analysis of farm record-keeping and sources of information about best management practices (BMPs). 3.8 Analytical Methods All quantitative data evaluated in this study were processed in R Studio (Version 2024.09.1+394) and analyzed with univariate techniques including frequency tables and density plots. Where intersections between groups were of interest, Complex Upset plots were used to visualize these intersections, using the UpSetR package in R (Lex et al., 2014). Qualitative interview data analysis is guided by Maxwell (2013), such that transcripts were evaluated first by coding responses by topic and then categorizing both by content or thematic issues (what core concept was discussed) and theory (what systemwide enabling factors were implicated). 3.9 Validation Validity threats were addressed using a combination of countermeasures including triangulation between methods, using multiple interviewers, describing contradictory evidence, and considering frequency of theme occurrences. Respondent validation (Maxwell, 2013) was informally used in conversation with participants but is formally planned for a future workshop. 45 CHAPTER 4 – FINDINGS 4.1 Sample Description Eight smallholder farmers were surveyed using the semi-structured questionnaire to generate quantitative and qualitative data (Table 1). These 8 smallholder farmers span two Beach Management Units (BMUs) in Kisumu and run operations ranging from one cage to 15 cages, and in one case managing an additional 30+ cages belonging to other people. The participants included one woman and seven men. Six of the 8 farmers considered themselves to be small farmers, with explanations including “the type of cage being used,” “the number of cages I have is less,” “I am still owning small size cages,” and “the money to expand is not there.” One farmer considered his operation to be a large farm despite owning only 5 cages, explaining that he has “grown faster than the others.” The average number of cages owned by each farmer interviewed was 4.5 (Table 2). The eighth farmer self-classified as medium and did not elaborate. One participant was only several months into the first production cycle, while others had over a decade of experience. 46 Table 1 Categorical Descriptive Statistics of Study Participants. Characteristic Count % of valid responses Gender Male 7 87.5 Female 1 12.5 Education Primary 1 12.5 Secondary 5 62.5 University 2 25.0 Prior occupation Fishing 4 50.0 Fabrication 1 12.5 Ecotourism 1 12.5 Vegetable farming 1 12.5 Government services employee 1 12.5 Year established current business 2024 1 12.5 2023 1 12.5 2022 2 25.0 2021 1 12.5 2020 1 12.5 2019 1 12.5 2014 1 12.5 Source of initial capital Savings 6 75.0 Loan 1 12.5 Table banking 1 12.5 Has other income source Yes 7 87.5 No 1 12.5 Other income source (not mutually exclusive) Ecotourism 2 28.6 Shopkeeping 2 28.6 Restaurant 1 14.3 Fabrication 1 14.3 Fishing 1 14.3 Table 2 Continuous Descriptive Statistics of Study Participants Variable n Mean SD Min Max Years of individual experience 8 6.13 5.19 2 16 Years in business 8 4.00 3.32 1 11 Number of cages 8 4.50 5.47 1 15 47 All but one participant had attended at least secondary school, and half of participants had transitioned to aquaculture from fishing. All but one participant maintained other sources of income, with the most common being ecotourism and shopkeeping. While three of the farmers who had previously been fishermen had left fishing behind, with explaining how with his new venture into aquaculture "I don't toil the whole night looking for fish that aren't there," One of the farmers does continue fishing. This farmer explained that fishing is a crucial source of income to finance the aquaculture enterprise, since fish farming requires months of spending money on labor and feed before there is any return on investment, while a daily wild catch is easily converted to daily income. Most farmers hired labor to help operate their farms, while some relied on family members and one, belonging to a women’s group with shared ownership, engaged in a rotating group labor scheme (Table 3). Table 3 Labor Sources # of people employed from [labor category] Labor sources Count % of valid responses Mean SD Min Max Hired labor 1.57 1.72 0 5 Yes 6 75.0 No 2 0.0 Family 0.50 0.76 0 2 Yes 3 62.5 No 5 37.5 Rotating group labor scheme 1.88 5.30 0 15 Yes 1 12.5 No 7 87.5 48 To test assumptions about challenges holding farmers back, the farmers were asked to rank how severely each of 7 assumed challenges affected their operation. Then, every pairwise combination of challenges was presented in competition, for example “between input cost and accessing financial services, which is a bigger challenge to your operation?” to reflect how any organization with scarce resources for programs addressing these challenges might have to triage what challenges to engage with (Table 4). One farmer opted out of these questions, so the descriptive statistics consider only 7 participants. On average, “Input cost” was considered a bigger challenge than 5.2 of the other 6 challenges and was ranked as “severely or always” affecting the farmers. “Accessing financial services,” “access to technical support,” and “environmental risks” were also ranked highly, whereas “fragmentation among stakeholders,” “market access,” and “post-harvest loss and cold-chain access” were ranked more moderately. 49 Table 4 Ranked Challenges to Farm Operation and Success # of other challenges [challenge] beats or is equal to in pairwise comparisons Challenge and strength of effect on operation Count % of valid responses Mean SD Min Max Input cost 5.2 1.3 3 6 Severely or always 6 85.7 Moderately or frequently 0 0.0 Minimally or occasionally 1 14.3 No effect 0 0.0 Accessing financial services 4.2 1.5 2 6 Severely or always 5 71.4 Moderately or frequently 2 28.6 Minimally or occasionally 0 0.0 No effect 0 0.0 Access to technical support 4.0 1.0 3 5 Severely or always 6 85.7 Moderately or frequently 1 14.3 Minimally or occasionally 0 0.0 No effect 0 0.0 Environmental risks 4.0 1.0 3 5 Severely or always 5 71.4 Moderately or frequently 2 28.6 Minimally or occasionally 0 0.0 No effect 0 0.0 Fragmentation among stakeholders 2.0 1.9 0 4 Severely or always 1 14.3 Moderately or frequently 3 42.9 Minimally or occasionally 2 28.6 No effect 1 14.3 Market access 2.0 1.6 0 4 Severely or always 0 0.0 Moderately or frequently 4 57.1 Minimally or occasionally 0 0.0 No effect 3 42.9 Post-harvest loss & cold-chain access 1.6 1.8 0 4 Severely or always 3 42.9 Moderately or frequently 0 0.0 Minimally or occasionally 2 28.6 No effect 2 28.6 50 In addition to being asked about these aspects of aquaculture business in terms of challenges, the farmers interviewed were also asked about them as “success factors” and asked to rank the importance to the farm’s operation and success for each one (Table 5). “Affordable inputs” and “financial services” both emerged as the most important success factors, with 6 farmers classifying them as "extremely or always” important to farm success and 1 farmer classifying them as moderately or frequently important. Notably, both of these variables ranked as both highly challenging and highly important to success. “Market access,” on the other hand was ranked as highly important but not as highly challenging. “Technology and technical support,” “environmental risk management,” “cooperation among stakeholders,” and “post- harvest infrastructure” generated more varied views about their importance to success. 51 Table 5 Ranked success factors and importance to operation. Success factor and importance to operation Count % of valid responses Affordable inputs Extremely or always 6 85.7 Moderately or frequently 1 14.3 Minimally or occasionally 0 0.0 Not important 0 0.0 Financial services Extremely or always 6 85.7 Moderately or frequently 1 14.3 Minimally or occasionally 0 0.0 Not important 0 0.0 Technology & technical support Extremely or always 4 57.1 Moderately or frequently 2 28.6 Minimally or occasionally 0 0.0 Not important 1 14.3 Environmental risk management Extremely or always 3 42.9 Moderately or frequently 3 42.9 Minimally or occasionally 0 0.0 Not important 1 14.3 Cooperation among stakeholders Extremely or always 4 57.1 Moderately or frequently 1 14.3 Minimally or occasionally 2 28.6 Not important 0 0.0 Market access Extremely or always 6 85.7 Moderately or frequently 1 14.3 Minimally or occasionally 0 0.0 Not important 0 0.0 Post-harvest infrastructure Extremely or always 2 28.6 Moderately or frequently 2 28.6 Minimally or occasionally 1 14.3 Not important 2 28.6 4.2 Cooperation, Planning, and Governance To quantify the current extent of cooperation or collaboration between and among the smallholder fish farmers interviewed, each was asked whether they collaborate with any organization in the areas of cage maintenance, marketing products, fish 52 production, credit sourcing, disease prevention, feed management, financial literacy, or environmental risk mitigation (Table 6). Three farmers reported collaboration with organizations for cage maintenance, which they all clarified to mean working with a local craftsman for cage repairs. Two farmers reported collaborating with an organization for marketing products and one farmer reported collaborating with an organization for fish production, credit sourcing, disease prevention, and feed management. For all of these categories where only one farmer reported collaboration, the collaborating organization is a women’s group in which its members share responsibility for the cage in its control. No farmers reported collaboration around financial literacy or environmental risk mitigation. Table 6 Collaborative Counts by Area Collaborates with any organization in [area] Count % of valid responses Cage maintenance Yes 3 37.5 No 5 62.5 Marketing products Yes 2 25.0 No 6 75.0 Fish production Yes 1 12.5 No 7 87.5 Credit sourcing Yes 1 12.5 No 7 87.5 Disease prevention Yes 1 12.5 No 7 87.5 Feed management Yes 1 12.5 No 7 87.5 Financial literacy Yes 0 0.0 No 8 100.0 Environmental risk mitigation Yes 0 0.0 No 8 100.0 53 In addition to participation in collaborative groups, awareness of such groups was also of interest, to understand if farmers who were not collaborating with others had the option to. When asked how well aquaculture stakeholders collaborate or cooperate in the region, the responses evenly spanned the full range from “very well” to “not at all” (Table 7). All participants were also asked if they were aware of any organized farmer groups for accessing technical support. Three farmers said that the BMUs had helped them access technical support for feed formulation and financial management. Interestingly these three farmers did not fall all in one BMU but rather were split between the two BMUs engaged in the survey. Still, every farmer said they were not aware of any organized farmer groups for accessing formal credit sources and only one farmer said they were aware of a group for accessing inputs, which they said coordinates its members in the sourcing of fingerlings. Table 7 Rating Aquaculture Collaboration and Cooperation in Lake Victoria, Kenya Rating how well aquaculture stakeholders collaborate or cooperate in region Count % of valid responses Very well 2 28.6 Adequately 2 28.6 Poorly 2 28.6 Not at all 1 14.3 While the farmers’ scoring of cooperation takes a wide spread, there was much more consensus about how well current policies support aquaculture development in the region (Table 8). Asked how well current policies support aquaculture development in the region, most said “poorly,” none said “very well,” and only one said “adequately.” The one farmer who said current policies are adequate clarified that her women’s 54 group has been a beneficiary of a policy that grants financial support for women in aquaculture from the government in partnership with a non-governmental organization (NGO). When asked what policies supporting aquaculture development they were aware of, four of the six farmers who ranked policy support at poor or worse said they were not aware of any policy. The other two said, “there are no policies supporting cage culture. There is only regulation of fishing nets, but it is not enforced,” and “there is a county policy governing cleanliness of the lake, though there is no help from the County when loss of fish is experienced.” With the exception of one policy for targeted support to women, broadly supportive policies seem to be mostly unknown and poorly implemented when known. Table 8 Rating Aquaculture Development Policy in Lake Victoria, Kenya Rating how well current policies support aquaculture development in the region Count % of valid responses Very well 0 0.0 Adequately 1 14.3 Poorly 5 71.4 Not at all 1 14.3 4.3 Innovative Technology and Management While the FAO’s Blue Transformation Roadmap (2022) emphasizes genetic improvement of reared tilapia stock, the technology and management challenges demanding innovation in Lake Victoria are more existential than that. Two of the farms visited in this study reported operating below capacity because of large system breakdowns that are frequent and unpredictable. In both cases, the farms were 55 operating at about half of their reported capacity despite facing demand significantly greater than what they can supply. One of the farmers explained that they had suffered high losses in their hatchery and brood ponds in the preceding months from rain, flooding, and wind that overturned the nets containing their broodstock. Despite these challenges, when asked about what is going well, the same farmer identified fingerling production as an area of success. The main challenge, however, that was common to both farms had to do aquaculture systems requiring constant electricity while grid power could not be counted upon or was absent entirely. Recirculating aquaculture systems (RAS) have many advantages for incubating tilapia eggs and larvae and for rearing fry into fingerlings; water quality, pathogen exposure, and feeding are tightly controlled, meaning the fry and fingerlings grow faster and healthier. The downside of RAS, however, is its constant power requirement to (1) pump the water back to the top of the system after solids have been filtered mechanically and nitrogenous wastes have been filtered biologically, (2) keep the water oxygenated with aerators, and (3) treat pathogens and algae in the water with UV light. This is a problem, explained one of the visited farmers, because “RAS is constrained by electricity, the power is unreliable, and we have had no power for the last several months.” While this farm does have a backup generator, the farmer explained that it is too expensive to run as a long-term solution. As a result, this farm’s incubation facility and nursery were both non-operational at the time of the visit. The same was true of the other farm that had suffered similar system breakdown in a different county. When asked if they would do anything differently, if they were to go 56 back, this farmer explained that no, “RAS is essential so has to be there” and that they plan to install a solar and battery system in the future. Even if these farms are successful in installing solar and battery systems, the required generation capacity for RAS is still significant and the demand is still constant. One farm, facing similar constraints, appears to have developed an innovation that addresses many of the risks associated with RAS while maintaining many of the benefits (Alando, 2024). This farm has set up a modified flow-through system that requires running a pump only a couple times per day to draw water from the lake to an elevated tank. Gravity then takes that water through a solar water heater for disinfection, then continues through the system and back to the environment, without being recirculated or requiring all the many parts to go right that a RAS requires. Innovations on appropriate technologies like this, instead of just transplanting systems that have worked in places with more reliable power, are promising and reward the farms that adopt them. 4.4 Access to Resources and Services The priority actions described in the FAO’s Blue Transformation Roadmap (2022) under their “access to resources and services” target emphasize access for women and youth. The scarcity of women and absence of youth participating in this study, corroborated by Orinda et al. (2021) documenting the uniquely high barrier to entry these groups face, suggests that even as the study participants are found to have challenges accessing resources and services, their level of access is still greater than more marginalized groups. 57 4.4.1 Financial Resources and Services Every smallholder fish farmer interviewed either had another income stream, investment from family members, or had invested from their savings into their aquaculture enterprise. Without these other sources of capital, which many said were still insufficient, many of the farmers questioned how else one would get started. Several farmers noted their inability to get enough credit to buy feed, which is a challenge because fish require daily feeding for at least eight months before the farmer receives a return on investment at harvest. As one farmer explained “banks give a loan today and then want repayment next month. But harvest doesn’t come until eight months later.” While only one farmer successfully secured a loan to help with the initial capital investment, a few others have been able to borrow for more recent investments. Three farmers reported having borrowed money to operate or invest in their business within the prior 12 months while five had not. Those who had borrowed secured their loans through either M-Shwari (a loan service for M-Pesa mobile money users), a commercial bank, or a combination of banks and friends. These farmers explained that their reasons to borrow included “to boost business” and “for fish stocking.” Additionally, one farmer explained that he has a close relationship with his feed supplier, which sometimes enables him to get credit from his feed supplier to collect feeds even when he cannot pay. Those who had not borrowed in the past 12 months explained that they either were not able to obtain a loan because “there is no entity to 58 give a loan” (n = 1) or they did not need to because they had enough savings (n = 3), or they were able to invest with money from another business (n = 1). 4.4.2 Extension Resources and Services The survey asked about extension services, trainings, and technical support in two ways and the discrepancy in answers was not reconciled or clarified, so both are reported here. First every farmer but one reported accessing technical support or training, though the frequency with which farmers accessed such support varied and some gaps were still identified. Two farmers reported accessing technical support or training monthly or more, 2-3 times per year for three farmers, and less than two times per year for one (Table 9). When asked about extension services and training specifically, two farmers reported that they do not receive extension services or trainings, while the other six highlighted receiving trainings on disease management (n = 2), water quality (n = 1), financial management (n = 1), feed formulation and quality (n = 2), general fish farming practices (n = 2), and fingerling handling (n = 1). All of these trainings were reported to have been one-time occurrences, except for one farmer’s quarterly participation in the water quality training, another farmer’s 2-3 times per month participation in the general farming practices training, and another farmer’s quarterly participation in the feed formulation training. Training and extension sources included KMFRI; ABD; Kisumu County; Ministry of Mining, Blue Economy, and Maritime Affairs; FUGO Feeds company, and directly from the BMU in partnership with private companies. Two farmers remarked that they wished trainings were more consistent and two other farmers praised the trainings they had 59 received, saying they “acted as an eye opener to sustainable aquaculture practices,” “improved feeding knowledge,” and “influenced cage model.” Table 9 Frequency of Seeking Technical Support or Training. Frequency of seeking technical support or training Count % of valid responses Regularly (monthly or more) 2 25.0 Occasionally (2-3x/year) 3 37.5 Rarely (less than 2x/year) 2 25.0 Never 1 12.5 Regarding gaps in training resources, one visited farm that makes its own feeds noted that “we need better expertise and maybe trainings too for staff,” explaining that some of the process is done by feel, such as the water content before the feed is heated and extruded into pellets. This part is important, he explained, because the water content at the time of heating is what determines whether the feed floats for 10 minutes at worst or 2.5 hours at best. When feed floats for longer, it is more likely to be eaten and converted into tilapia than to sink and be wasted. In the 2023 survey conducted by Teplitz et al. (2024), participants were asked where they get information on “best management practices,” or BMPs. Participants were given the option to select all that apply between Kenya Fisheries Service (KeFS), other farmers, online groups, online websites, WhatsApp groups, KMFRI, an NGO, or any other source. Using data shared from this survey, Table 10 shows that “other farmers” was the most common source of information about BMPs, with 52% of farmers getting information that way either in combination with other sources or alone. As Figure 12 shows, other farmers are often the only source of information 60 about BMPs that participants reported. KeFS, KMFRI, and NGOs are also commonly reported sources of information but are much more likely to be reported in combination with each other or other sources. Table 10 BMP Information Sources BMP information source Yes No n % Yes Other farmers 90 82 172 52.33 KeFS 85 87 172 49.42 KMFRI 84 88 172 48.84 NGO 47 125 172 27.33 Other source 16 156 172 9.30 Website 10 162 172 5.81 Online group 9 163 172 5.23 WhatsApp group 3 169 172 1.74 Note: Data from survey described by Teplitz et al. (2024). Figure 12 Distribution of BMP Information Sources, Alone and in Combination 61 Note: Horizontal bar plot represents total number of farmers that reported each BMP information source. Vertical bars represent the number of farmers who reported the combination of sources specified in the plot with connected dots. 4.5 Resource Use Efficiency and Impact to Environment The large number of largely uncoordinated aquaculture smallholders, combined with the smaller number of larger commercial farms, leads to a tragedy of the commons. Fish excreta and uneaten feed, which represents a resource inefficiency, contribute to Lake Victoria’s high nutrient load, and in turn, farmers face algal blooms that contribute to devastating fish kills as well as other adverse, but less devastating, consequences. Seven out of eight participants reported having experienced an algal bloom in the past year (Table 11). While aquaculture exacerbates nutrient pollution and its consequences, there are other environmental challenges to which farms do not contribute but are still subject to and need to be managed when relevant. Six out of the eight participants reported having experienced upwellings, when the deep low-oxygen water rises into the surface waters where the fish are held, in the last year, with effects including “some fingerlings died,” “lost a whole cage investment,” “killed fish in entire cage,” and as described most succinctly by 3 of the participants, “fish kills.” Another farmer, described a recent fish kill that he attributed to water pollution, noting that this instance killed “just a few fish and not the whole cage.” Farmers also noted other environmental risks, including water level rise, wind, waves, and rain. One visited farm had lost several hatchery ponds to Lake Victoria’s rising water level, while also noting how a period of high winds had overturned some nets, disrupting the 62 hatchery operation. Another farmer explained that “waves move cages to vulnerable places,” and yet another described flooding and non-navigable roads in the rainy season as a major environmental threat, preventing buyers from getting to the beach to purchase fish and preventing farmers from transporting feed to the beach to take to the cages. Table 11 Experience of Algal Blooms and Upwellings in Past Year. Variable Count % of valid responses Experienced algal blooms in past year Yes 7 87.5 No 1 12.5 Experienced upwelling in past year Yes 6 75.0 No 2 25.0 The concept of resource use efficiency and the impact on the environment naturally invokes the concept of a circular economy, in which wastes that might otherwise burden the environment are put to productive use. One such waste stream in aquaculture is the viscera, or the guts, that are taken out of the fish when the fish are cleaned. The fates of these viscera are varied (Figure 13); in some cases they are pollutants and in others they are resources. For example, one large farm visited in this study extracts fish oil from guts, then composts what remains and gives the compost away to nearby communities. The viscera are also put to use at one BMU visited in this study, where a biodigester is installed, taking fish viscera as a feedstock to make biogas. At the other BMU visited in this study, however, the viscera were observed to be discarded directly into the lake. 63 Figure 13 The Varied Fates of Fish Viscera. A. B. C. Note. A. Processing at a large farm for value recovery. Fish oil is extracted from guts by massaging the guts to release fat. The fat is then skimmed off, congealed in an ice bath, and then rendered. B. Biogas digester at a Kisumu landing site. C. discarded into Lake Victoria at another Kisumu landing site. 4.6 Monitoring and Reporting While monitoring and reporting applies at both the farm and regional or wider scales, this section focuses on the farm or business. Regional or national monitoring and reporting is considered later in this paper. Two areas where monitoring came up at the farm or business level are water quality and post-harvest distribution. Two farms visited during this study explained that they measure dissolved oxygen (DO) every morning and adjust their feeding accordingly, not feeding when DO is less than 4mg O2 per liter.3 One farm noted that DO gets low in their hatchery ponds, which is a 3One important reason not to feed when DO is low is because a fish’s oxygen demand while digesting is huge compared to its baseline state, and a high enough density of fish feeding and digesting in low-DO conditions will quickly deoxygenate the water. 64 challenge. They said they would like to add paddle wheels for aeration, but they would need to expand their solar power generation capacity. The other example of data-driven management comes at the distribution stage. One farm visited for this study manages its own distribution network and explained that its stores that sell to retailers are supplied by refrigerated hubs that each service many stores, but the stores themselves do not have refrigeration. This system is possible because they intensively keep data on purchasing patterns in each store to inform inventory distribution, ultimately saving on cold-chain investment while keeping spoilage at less than 1%. In order to gain useful management insights from farm records, those records have to be there. As the data from the survey described by Teplitz et al. (2024) show, the vast majority (99%) of farmers surveyed reported keeping records (Figure 14). Recordkeeping in each category of feeding, mortalities, and production, was reported by over 70% of farmers interviewed (Table 12). The most common combination of recordkeeping categories reported was feeding, mortalities, production, and fish growth, with 27% of participants reporting the collection of all four. Water quality was the least recorded parameter, with only 5% of participants reporting that they keep water quality records. 65 Figure 14 Distribution of Farm Records Kept, Alone and in Combination Note: Horizontal bar plot represents total number of farmers that reported keeping each category of records. Vertical bars represent the number of farmers who reported the combination of recordkeeping categories specified in the plot with connected dots. Table 12 Farm Recordkeeping Frequencies Recordkeeping category Yes No n % Yes Feeding 148 23 171 85.55 Mortalities 138 33 171 80.71 Production 121 50 171 70.76 Growth 83 88 171 48.54 Water quality 9 162 171 5.26 Note: Data from survey described by Teplitz et al. (2024). 66 4.7 Value Chains While the FAO Blue Transformation Roadmap (2022) focuses on value chains as one of the three objectives for equitable and sustainable aquatic food system development, the focus only includes downstream value chains. As Orina et al. (2019) map out, though, cage farmers are just as exposed to value chain conditions upstream from them (i.e. for inputs) as they are to the value chain downstream that ultimately returns them revenue. The present study considers both, finding that the inputs value chain fails to deliver in a way that the participating farmers are satisfied with. 4.7.1 Upstream Feed is the biggest and most important cost for fish farms. As one farmer explained, feeds account for 75% of his costs, but that “any operation should want feed to account for a very high percentage of production costs” because that means overhead is low. The other side of feeds being the dominant cost, though, is that farms are strongly affected by any changes in price or supply chain. As another farmer said, “Feeds are the biggest headache—they are expensive to buy and expensive to transport.” What feeds farmers choose to use is an important choice, partially determining how much it costs to grow the fish, and how much fish there will be to sell. As still another farmer explained: Lack of high-quality feed is a big challenge. Many people using sinking mash feed, which leads to a 16-month production period with high waste and pollution, and low feed conversion efficiency and harvest weight. Floating feed gives us an 8-month production period. 67 Four smallholder farmers interviewed use only commercial feeds, while two use only non-commercial feeds and two use a combination (Table 13). Many commercial feeds are imported, including from Egypt, India, and Mauritius, according to one feed distributor. All four of the interviewed smallholders using non-commercial feeds include ochong’a (Caridina nilotica; a small freshwater shrimp), in their feeds while two also use both bread crumbs and omena (Rastrineobola argentea), and one uses maize, sunflower, cotton seed, and snail shell in their reported mix along with the ochong’a. Five farmers get their feed or feed ingredients from an agro-dealer while three get theirs directly from a mill and one from a commercial feed manufacturer’s depot. Table 13 Feed Sourcing Descriptive Statistics. Variable Count n % of valid responses Use commercial feed 8 Yes 6 75.0 No 2 25.0 Use non-commercial feed 8 Yes 4 50.0 No 4 50.0 Non-commercial feeds used (not mutually exclusive) 4 Ochong’a 4 100.0 Bread crumbs 2 50.0 Omena 2 50.0 Maize 1 25.0 Sunflower 1 25.0 Cotton seed 1 25.0 Snail shell 1 25.0 Feed source (not mutually exclusive) 8 Agrodealer 5 62.5 Directly from mill 3 37.5 Manufacturer depot 1 12.5 68 When asked about what factors were important in purchasing decisions about feed, all eight smallholders said that they considered quality (Table 14). In addition to quality, six farmers said they considered consistency in supply and six said they considered price, while two did not. Five farmers said they considered proximity to source, five said they considered certification and traceability, and five said they considered their relationship with their supplier. When asked which factor was more important, comparing every combination of these factors, quality beat 4.17 of the others on average, consistency in supply beat 3.17 of the others on average, and price and proximity to source both only beat 2 of the other factors on average. Certification and traceability and relationship with supplier were the least important factors, each only on average beating 1.50 and 1.17 of the other factors, respectively, which were mostly each other. When asked to describe the feeds they favored, the participants’ descriptions included “expensive,” “good protein,” “good quality with high protein,” and “has enough crude protein.” 69 Table 14 Factors Considered in Feed Purchasing Decisions # of other factors [factor] beats or is equal to in pairwise comparisons Factors considered in feed purchasing decisions Count % of valid responses Mean SD Min Max Quality 4.17 0.98 3 5 Yes 8 100.0 No 0 0.0 Consistency in supply 3.17 1.60 1 5 Yes 6 75.0 No 2 25.0 Price 2.00 1.41 0 4 Yes 6 75.0 No 2 25.0 Proximity to source 2.00 1.79 0 5 Yes 5 62.5 No 3 37.5 Certification and traceability 1.50 1.97 0 5 Yes 5 62.5 No 3 37.5 Relationship with supplier 1.17 1.17 0 3 Yes 5 62.5 No 3 37.5 One farm visited during this study had determined that commercial feeds, aside from being used to maintain breeding stock, were not worth the money. Even after turning to making their own feed, though, they still claimed that “the biggest challenge, by far, is feeds.” Feeds are still a major challenge, the farmer explained, because the supply chain for feed components is unreliable and the wider enabling environment is lacking. Additionally, there have been equipment issues. When the extruder, which is the machine used to make feed pellets, is out of commission, it takes more than a week to get a repair done because there is only one person in the county who can work on motors like the extruder’s motor. 70 4.7.2 Downstream The participants in this survey reported selling their fish primarily to restaurants and local retailers. All but one participant reported selling to customers on credit. Additionally, the use of contracts varied by BMU. At one BMU far from Kisumu City’s population center, all farmers had a contract with the BMU to market fish through the BMU, which facilitates contracts with hotels, or restaurants, and schools. No such arrangement was reported at the other BMU, which is much closer to Kisumu city and where only one participant reported having advance contracts directly with customers. The others explained that contracts were not necessary, that “there is no need, buyers are around,” “the focus is on selling locally to retailers who come to the beach,” or “I am still a smallholder farmer.” One assumption that was tested in this survey was the importance of a functioning cold chain. While this study did not follow any fish beyond the farm gate where the cold chain is likely more consequential, the importance of cold chain infrastructure within the farm or BMU seems minimal or at least case specific. One farm visited for this survey, for example, explained that they had built a cold room after a customer complained about spoiled fish, but they generally harvest only what has been ordered so are rarely holding inventory and therefore do not experience spoilage. At one point, farmers from one BMU explained that they had received coolers from the government 5 years ago, but “nobody used them.” Two years later they received another cold- chain facility for the same purpose, which again has not been used as intended. “It is being used to charge electric motorbikes,” said one farmer. “Cold storage is not a priority because the market for fresh fish is not satisfied,” he continued. In all, most 71 smallholders surveyed did not seem concerned about the cold chain. Of the seven farmers who opted in to the survey questions on post-harvest processes, five (71.4%) indicated that they did not store harvested fish (Table 15), while one stored the harvested fish on ice and the last one stored them in a cold room owned by their BMU (not the BMU with the repurposed facility). All but one farmer graded their fish by size, and four farmers reported cleaning the fish they harvested, compared to three who said they did not clean their fish. The one farmer who did not grade the harvested fish explained that he only would harvest the desired size of fish. Table 15 Post-Harvest Storage and Processing Variable Count % of valid responses Storage of harvested fish None 5 71.4 Refrigeration – Ice 1 14.3 Refrigeration – Cold room owned by BMU 1 14.3 Fish processing Grading Yes 6 85.7 No 1 14.3 Cleaning Yes 4 57.1 No 3 42.9 When asked if they were able to satisfy peak demand, only one participant claimed to be able to. For those unable to satisfy peak demand, coping strategies included “harvesting based on demand”, “Other farmers complement my supply”, “I cooperate with other farmers to fulfill”, and “referrals to other farmers.” To emphasize the large demand for fish, one farmer said, “sometimes I have to hide from my customers.” While keeping fish alive is an effective way to keep them fresh, a distinct 72 disadvantage of harvesting on demand is that one’s real-time harvesting and processing capacity limits the amount of fish that can be prepared for sale and ultimately sold. Though this model was widespread among smallholders and no farmers reported wasted or spoiled fish between harvest and the farm gate, the medium and large farms visited during this survey had all made substantial cold chain infrastructure or distribution investments. When asked about marketing constraints, the participants mentioned “congestion in some seasons”, “overproduction from other farmers in some seasons”, “none”, “competition, product quality, and clients’ varying choices and preferences”, “fluctuating markets in various seasons”, and lastly, “no constraints, I harvest on demand and satisfy demand in general, but holidays are difficult—some customers get fish the next day” With marketing constraints spanning both excessive demand and excess supply, the participating farmers were asked about excess supply as well. Most farmers claimed they have excess supply at least occasionally, two of whom said they cope with this excess supply by giving fish on credit, thereby expanding their customer pool and customers’ purchasing capacity, and one of whom said they sell into their networks in Eldoret and Nairobi when their supply is greater than local demand. Nobody indicated that they adjust their prices. 73 CHAPTER 5 – DISCUSSION 5.1 Introduction Smallholder cage culture appears to be a risky venture in which farmers worry they don’t have the money, the technical know-how, nor the political support to thrive in the long term. When asked what would help them grow and make more money, the participating farmers generally responded with something that fits in these categories. Three farmers commented that “material costs are very high,” “cost of production is too high,” or “the cost of fingerlings and feed needs to be addressed,” with a fourth remarking how “loans and financing to buy cages” would help them. A future inquiry should investigate whether production costs are too high to be profitable (i.e. total production costs are greater than the value of fish sold) or if they are still profitable over a full harvest cycle, just too expensive to afford in the short term. This distinction has important implications for what a supportive program should look like. Answering the same question on what it takes to grow and make money, three farmers suggested better guidance on feed recommendations, BMPs, and technology to support implementation of BMPs. “We are gambling on feeds,” said one. “We don’t know which feeds are best and we want guidance on this.” Another farmer worried that his peers are overstocking their cages and need more authoritative guidance on how densely to stock fingerlings. A third said he knew he should withhold feed sometimes because of water quality, but he doesn’t have the instrumentation or capacity to test water quality. A fourth noted the environmental threats to his operation such as upwelling and asked what could be done to mitigate the risk while recognizing 74 that his cages are in a zone considered unsuitable for cage farming and it is not feasible to move to a suitable zone. In addition to financial and information or technical services, a third group of answers emerged from the farmers’ responses, focused on addressing conflict and social impediments. One farmer identified a need for “regulation of fishermen who impose on aquaculture,” while another elaborated on a hidden cost of theft, explaining that “we do not know how much we lost to theft, and then we are overfeeding costly feeds because we are feeding for a higher stocking density than reality.” Between farmers, another farmer remarked, “people don’t cooperate because they want to grow on their own. But there are benefits to cooperating with others.” Each of these identified areas is evidence of an operating environment that has made it difficult for small farmers to grow into thriving commercial entities while larger companies have captured most of the sector’s growth. This section elaborates on three key aspects of that operating or enabling environment. 5.2 Supportive Policy Even though The Fisheries Management and Development (Aquaculture) Regulations (2024) included provisions to address water quality and other areas of concern, it was not seen as an answer to one stakeholder’s claim that “there needs to be political will to enforce good practices, and we need good education on best practices too.” These regulations were instead seen as undermining the sector with onerous fees, without tools to help farmers comply, and as having been developed without public 75 participation. The LVA Association demonstrated in 2024 the power of an industry association in lobbying against policy seen as unsupportive; 2025 or 2026 perhaps will show whether generative or constructive advocacy can lead to the creation and implementation of a policy that is seen to be supportive. One stakeholder pointed out that policy that is good for aquaculture development does not have to be tailored to aquaculture alone; “from the government we are needing reliable power and needing solar or generators, and infrastructure including roads.” Meanwhile, relating specifically to fisheries, Aura et al (2021) call for development and enforcement of regulations to govern siting of cages to minimize conflicts and optimize for both capture and culture fisheries uses. 5.2.1 Nutrient Management Supportive policy extends beyond direct regulation of aquaculture to include regulation of other sectors that affect aquaculture. Nutrient pollution, for example, is not caused only by aquaculture. Nutrient pollution from other sources also threaten aquaculture and should also be regulated. 5.2.1.1 Nutrient Management in Aquaculture As the U.S. has regulated CAFO permitting and incentivized nutrient management through the Clean Water Act, the NRCS 590 Standard, and NDPES Permits, Kenya has recently attempted to impose licensing regulations on the aquaculture sector with the introduction of The Fisheries Management and Development (Aquaculture) Regulations (2024). Though this law called for water quality monitoring and 76 management of pollution, it was not accompanied by a mechanism for farmers to comply. Instead, it was accompanied by onerous licensing fees that would have put many farmers out of business. The law was suspended by the courts in December 2024 (Matete, 2024) and is now in the process of being repealed (Oirere, 2025). As Kenya’s aquaculture regulations are being rewritten, the timing is critical to introduce nutrient mass balances as an accessible and effective indicator to help farms prosper while complying with best environmental practices. Aquaculture farmers and regulators would benefit from a user-friendly NMB tool to estimate farm nutrient mass balances, informing better management for lower investment risk, improved sustainability, and greater profitability. 5.2.1.2 Nutrient Management Outside Aquaculture Nutrient pollution from sources other than aquaculture, including urban sewage, fertilizes toxic algal blooms that have produced microcystin levels that vastly exceed WHO-accepted limits and have led to anoxia and devastating fish kills (Olokotum et al., 2020; Mchau et al., 2019; Njiru et al., 2018; Sitoki et al., 2012; Stager et al., 2009; Ochumba, 1990). Today, excess nutrient loading poses an existential threat to Lake Victoria, its fisheries, and the tens of millions of livelihoods supported by the lake (Roegner et al., 2020; Njiru et al., 2019; Hecky et al., 2010). In November 2022, for example, news outlets reported fish kills bearing death tolls in the millions of fish in a single anoxia event, blaming the casualties on raw sewage discharge, among other causes (Bizot, 2022). In the lakeside city of Kisumu, half of all human waste is emptied from latrines but never delivered to treatment; ultimately, these wastes are 77 either dumped to the lake or somewhere where they are liable to leach downstream to the lake (Midega, 2022; Peal et al., 2020; Furlong & SFD Promotion Initiative, 2016). Just as urban population growth and land use change have diminished wetland services and exacerbated phosphorus loading, population-driven rural land use change has also contributed enormously (Njagi et al., 2022). Agricultural runoff, specifically of eroding soils, accounts for a major portion of phosphorus loading. Fluvial erosion is largely attributable to bare soils and reduced vegetation cover (Odada et al., 2004), and the highest deforestation rates are in highlands already vulnerable to erosion (Juma et al., 2014). Atmospheric deposition, including rain and particulate dryfall of ash from biomass burning and soil and dust from wind erosion (Njagi et al., 2022), also accounts for a major source of phosphorus to Lake Victoria; estimates include 36% of total phosphorus (Scheren et al., 2000), 44% (Agwanda & Iqbal, 2019), and 55%, or 1.80- 2.70 kg ha-2 year-1 (Tamatamah et al., 2005). Burning of agricultural residues and wild vegetation are both widespread during the June-October dry season; correspondingly, dry phosphorus deposition to Lake Victoria was measured and recorded to be nearly three times higher during those months than November-May when burning is not prevalent (Tamatamah et al., 2005). Ash from these fires is carried to the lake on southerly and southeasterly prevailing winds from the Mara region’s savannahs to the southeast and agricultural areas to the south (Tamatamah et al., 2005). Water-borne and air-borne erosion combined, the need to keep soil in place and minimize biomass burning is clear; the fluvial and aerial flow of phosphorus from rural soils to Lake 78 Victoria continues to deplete agricultural areas of nutrients that are already yield- limiting and continues to fertilize algal blooms downstream that threaten the survival of the fishery and the livelihoods that depend on it. To manage anoxic fish kills triggered by dying algal blooms and toxic microcystin, algal abundance and community composition both must be addressed. Overall algal abundance is limited by phosphorus in Lake Victoria, and while the non-toxic diatoms are silicon- and nitrogen-limited, their toxic cyanobacterial counterparts are phosphorus-limited. In the Nyanza Gulf of Lake Victoria, where the Kenyan cities of Kisumu, Kisii, and Homa Bay are situated, these threats to ecosystem health and integrity are significantly driven by municipal and industrial effluent discharge, including sewage (Guya, 2020). Further, Njagi et al. (2022) show that along with a growing population, land use changes around Lake Victoria have transformed the watershed’s nutrient flows. Between 1985 and 2014, the lakefront cities expanded rapidly and enormously, some as much as 1022% (Njagi et al., 2022). Guya (2020) argues that combined, the private development of wetlands that once helped detain nutrient flows to the lake, and the increased nutrient flow from increasing anthropogenic activity, together present a critical need for conservation of wetlands that can intercept nutrient inflow from land. In all, while nutrient management in aquaculture needs to be a priority for aquaculture developers and policymakers, it also needs to be a priority across the lake, which will require regulation of all contributing sectors. In this way, a supportive policy environment includes regulatory de-risking of external threats to aquaculture. 79 5.3 Cooperative Governance and Facilitation of Services Without strong organization, smallholders fail to achieve economies of scale and the benefits that can be gained only at scale. Such benefits are exemplified by the successful valorization of viscera wastes at one very large farm and one BMU while another BMU appears to burden the environment with these wastes, at least some of the time. While it may not pay off for any individual or small producer to invest time and resources in processing wastes to recover some value from them and keep them from polluting, as it does pay off for a large, well-capitalized operation, it can pay off for a group of individuals or small producers cooperating with each other. Beyond managing wastes, one farmer surveyed for this study pointed out how aggregating groups can also improve access to inputs and training, mentioning that there is already an association that provides trainings, fingerlings, and feeds to pond aquaculture farmers only, and could potentially be a model for cage culture as well. Another farmer pointed out that cooperating more openly with other producers “could give a lot of information and knowledge.” “Cooperation is very important for people to learn from each other. Without that we cannot move anywhere,” said another. Already, some farmers within the same BMU are supporting each other with their labor; two farmers described helping each other with fish harvests and in emergencies like clearing floating water hyacinth rafts that interfere with the cages. In addition to informal local collaboration, more organized regional cooperation can be pursued. Every farmer interviewed was presented with a concept for county-level hubs that could potentially aggregate demand for inputs with bulk procurement, organize the provision of financial services, develop and share new market linkages, enlist 80 technical support for farmers, conduct environmental risk monitoring, or provide cold- chain storage and distribution infrastructure. When asked how likely they would be to use a hub offering each service, inputs and financial services emerged as the clear winners with five out seven farmers saying they would be extremely likely to use such a hub, one farmer saying they would be highly likely, and one saying they would be moderately likely (Table 16). Market access facilitation and technical support followed, with much less interest expressed in environmental risk monitoring and cold chain services, also shown in Figure 15. Table 16 Reported Likelihood of Using Resource Hub with Selected Services Likelihood of using a hub offering [service category] (n (%)) Service category None Minimally Moderately Highly Extremely Affordable input services 0 (0.0) 0 (0.0) 1 (14.3) 1 (14.3) 5 (71.4) Financial services 0 (0.0) 0 (0.0) 1 (14.3) 1 (14.3) 5 (71.4) Market access facilitation 0 (0.0) 1 (14.3) 1 (14.3) 1 (14.3) 4 (57.1) Technical support 0 (0.0) 0 (0.0) 3 (42.9) 0 (0.0) 4 (57.1) Environmental risk monitoring 0 (0.0) 1 (14.3) 1 (14.3) 2 (28.6) 3 (42.9) Cold-chain services 2 (28.6) 2 (28.6) 0 (0.0) 1 (14.3) 2 (28.6) 81 Figure 15 Density Plot of Reported Likelihood of Using a Hub Offering Selected Services Of the seven farmers who participated in the follow-up question of what kind of technical support would benefit them most, all seven identified “feed management.” Additionally, “feed management” and “disease prevention and biosecurity” was the most common combination (Figure 16). For “marketing and financial literacy,” “cage maintenance,” and “disease prevention and biosecurity,” these areas were each identified by four farmers as areas where technical support would benefit them. Environmental risk mitigation was the least commonly selected, picked by only two farmers. One farmer noted that coordination between farmers is a prerequisite for progress in several of these challenge areas, especially biosecurity. 82 Figure 16 Frequency of Perception About Which Technical Support Areas Would Benefit Farmer Most. Note. Horizontal bar plot represents total number of farmers that selected each support area. Vertical bars represent number of farmers who selected the specific combination of support areas specified in the plot with connected dots. In addition to the structured questions on the envisioned hub services, the participants were asked how else they could imagine the hubs supporting them. Most responses were practical and directly linked to the business of buying inputs or selling products, possibly suggesting why environmental monitoring and cold-chain services ranked the lowest when farmers were asked which services they would be likely to use. Responses ranged from being broad or repeating the proposed structure, such as “improvement from a smallholder farmer to a medium farmer,” “promoting mine and 83 others’ businesses,” “bridging gaps that lead to higher investment,” and “providing financial services, trainings, market accessibility, and affordable inputs,” to more specific, such as “cutting transport costs on inputs, and also providing trainings and technical support,” “feeds can be bought because it would save time and transport, making feeds affordable,” “we would be financially managed and supported,” “if I have excess fish this would guide me on where to sell or market them,” and “we could coordinate higher prices if acting in cooperation.” Two farmers also noted security as an issue they could imagine the hubs addressing. After discussing these possible services and programs that a hub could deliver, the farmers were asked for their concerns about the hub concept. Two farmers said they had “no concern,” to which one added, “with such a group we would go farther.” Three other farmers were slightly more cautious; one wanted to know how a hub would balance so many priorities, suggesting that it “should prioritize trainings, then the other services can come later,” another pointing out that “it will support aquaculture business, but it will require proper management,” and the third asking “whether local residents will be employed in the hub and if it will develop community.” Lastly, two farmers were more distrusting, asking “how the hub will benefit from the farmers.” One of these two farmers asked again, “how farmers’ information will be protected? How will the hubs benefit from farmers—is it a money minting idea? And can it handle all the proposed services?” This concept of trust and whether programs have ulterior motives would continue to resurface. 84 5.4 Trust in Peers and Institutions Greater cooperation has been identified as an important factor for the success of aquaculture development in Kenya, with one study finding that pond tilapia farmers who joined cooperative or common interest groups saw 6% yield increases and a 30% increase in sales (Gichuki et al., 2025). Asked about barriers to effective cooperation they perceive, two farmers in the present survey identified “individualism,” while another said “greediness,” another said “nobody has the time,” another said “mistrust and individualism,” and one said there were no barriers. In fact, 63% of participants independently raised individualism, greed, and distrust as barriers to success and cooperation, and most of these participants reported having previous disputes with other smallholders or fishermen. Asked about disputes with other producers, four said they had experienced conflict; two reported “fish theft,” another reported “struggling for cage spaces,” and the fourth clarified that his conflict is “with fishermen coming too close and stealing, not with other aquaculture farmers.” All four were interviewed at the same BMU. No farmers at the other BMU reported such disputes. Aside from harboring some distrust toward each other or toward fishermen sharing the lake, some farmers expressed distrust in the institutions meant to serve them. When asked if they share production data with any agency, one farmer said, “no, we only share our hatchery and harvest protocols. Those guys give us problems. What do they give us in return—no training, no energy subsidy, no feed subsidy?” This farmer went on to share their concern about the suspended Fisheries Management and Development Regulations (The Fisheries Management and Development Act, 2024), 85 saying they worried that implementation of the licensing provision would bring predatory audits, adding “they just want a cut.” The consequences of distrust at all scales, shown by this example, by one farmer’s concern about the resource hub concept and the safety of their information, and by the frequently reported security conflicts including one farmer even describing a family member being caught stealing fish from his pens, are easy to imagine. In such an environment, the FAO Blue Transformation (2022) targets for monitoring and reporting, data collection, and governance, among others, seem very difficult to achieve. 86 CHAPTER 6 – CONCLUSION AND NEXT STEPS In general, the interviewed farmers seem to perceive their enabling environment as holding them back. Farmers who pursued aquaculture as an investment or as an alternative to fishing have experienced conflict and distrust while navigating challenging upstream value chains and a regulatory environment they see as unsupportive. Inadequate financial services prevent farmers from making the investments required to succeed, including investments in feeds. When presented with a concept for county-level resource hubs, financial services and inputs procurement emerged as the clear favorites that farmers could see themselves engaging with. The core concepts of access to high-quality, affordable inputs; access to financial services; access to technology and technical expertise and extension; access to markets, farm management, and management of environmental and biosecurity risks to aquaculture were shown to present varied opportunities and challenges to farmers. Critically, these core concepts were also shown to be exposed to the underlying systemwide enabling environment which at its best is defined by supportive policy, cooperative governance and facilitation of services, and trust in peers and institutions. A workshop is planned to assess these preliminary conclusions, and a draft workshop plan is included in the Appendix. 87 APPENDIX A: MAP OF CAGE LOCATIONS RELATIVE TO POPULATION CENTERS Source: (Irungu & FAO, 2023) 88 APPENDIX B: LVA WORKSHOP FALL 2025 DRAFT PLAN Workshop Title Designing Resource Hubs for a Thriving Aquaculture Sector Purpose Bringing together aquaculture stakeholders to co-develop a shared vision and actionable plan for Aquaculture Resource Centers (ARCs). Objectives 1. Build relationships and shared understanding among farmers, feed and fingerling suppliers, technical service providers, researchers, and policymakers. 2. Identify common challenges and opportunities where ARCs could provide coordinated support. Present findings from January pilot survey and invite response. 3. Co-develop a vision for ARCs that are responsive to local needs and scalable across the region. 4. Define clear goals for ARC development and establish actionable steps, roles, and responsibilities to move forward. Expected Outcomes 1. Strengthened networks and new partnerships across aquaculture production and value chain actors. 2. An articulated set of prioritized functions and services that ARCs should provide, grounded in stakeholder needs. 3. A shared vision for how ARCs can support sector prosperity, growth, resilience, and sustainability. 4. Concrete action points, timelines, and task plans to guide next steps toward ARC development. 5. A workshop report summarizing key insights, commitments and a roadmap for implementation. This report should inform ongoing planning, resource mobilization, and coordination among partners. Target Participants (TOTAL ~60)  Smallholder and commercial fish farmers  Feed, fingerling, and equipment suppliers  Existing aquaculture cooperatives/associations, including BMUs  County and national government representatives, possibly  Extension officers and technical support providers  Researchers/scientists and research/academic institutions  NGOs and development organizations  Financial service providers 89 Draft Program Morning session 1 – target Objective 1  Who is here? (Borrowed from Johannes Lehmann’s 3-round charette, in rotating groups of four): o Charette #1 ▪ How did you get here? ▪ What got you initially excited to engage with LVA or the ARC concept, and how? o Charette #2 ▪ How is it (your business, area of work, etc.) going now? ▪ What gets you excited at the moment? o Charette #3 ▪ What gets you excited about the future? ▪ Where do you want to be in 3-5 years?  Stakeholder identification and mapping exercise Tea Break Morning session 2 – target Objective 2  SWOT Analysis: present findings from pilot surveys, open discussion: do participants agree? Disagree?  What are the strategic issues?  Formulate strategies to manage the issues. Lunch Afternoon session 1 – target Objective 3  What does ARC success look like? What does it take? Tea Break Afternoon session 2 – target Objective 4  Final reflections and next steps  Discussion on sustaining momentum  Establish key goals and action points + timelines to achieve those goals for ARCs to be realized and deliver on their on purpose. Closing remarks and appreciation 90 REFERENCES Agwanda, P. O., & Iqbal, M. M. (2019). Engineering control of eutrophication: Potential impact assessment of wastewater treatment plants around Winam Gulf of Lake Victoria in Kenya. Journal of Coastal Research, 91(sp1), 221. https://doi.org/10.2112/SI91-045.1 Alando, P. (2024, September 7). African Blue banks on solar heating to produce fish hatchlings. Daily Nation. https://www.gatsbyafrica.org.uk/app/uploads/2024/09/aquaculture-focus-seeds-of- gold-7th-sept-24.pdf Aria, M., & Cuccurullo, C. (2017). bibliometrix: An R-tool for comprehensive science mapping analysis. Journal of Informetrics, 11(4), 959–975. https://doi.org/10.1016/j.joi.2017.08.007 Aura, C., Lewo Mwarabu, R., Sangara Nyamweya, C., Onyango Ongore, C., Musa, S., Last Keyombe, J., Guya, F., Fonda Awuor, J., Owili, M., & Muriithi Njiru, J. (2024). Unbundling sustainable community-based cage aquaculture in an afrotropical lake for blue growth. Journal of Great Lakes Research, 50(5), 102410. https://doi.org/10.1016/j.jglr.2024.102410 Aura, C. M., Musa, S., Nyamweya, C. S., Ogari, Z., Njiru, J. M., Hamilton, S. E., & May, L. (2021). A GIS‐based approach for delineating suitable areas for cage fish culture in a lake. Lakes & Reservoirs: Science, Policy and Management for Sustainable Use, 26(2), e12357. https://doi.org/10.1111/lre.12357 https://doi.org/10.2112/SI91-045.1 https://www.gatsbyafrica.org.uk/app/uploads/2024/09/aquaculture-focus-seeds-of-gold-7th-sept-24.pdf https://www.gatsbyafrica.org.uk/app/uploads/2024/09/aquaculture-focus-seeds-of-gold-7th-sept-24.pdf https://doi.org/10.1016/j.joi.2017.08.007 https://doi.org/10.1016/j.jglr.2024.102410 https://doi.org/10.1111/lre.12357 91 Aura, C. M., & Ntiba, M. J. (2024). Possible lessons on rapid assessment of fish kills in cages in Lake Victoria, Kenya for informed decision making. Aquatic Ecosystem Health & Management, 27(3), 65–73. https://doi.org/10.14321/aehm.027.03.65 Awuor, F. J., Obiero, K., Munguti, J., Oginga, J. O., Kyule, D., Opiyo, M. A., Oduor- Odote, P., Yongo, E., Owiti, H., & Ochiewo, J. (2019). Market linkages and distribution channels of cultured, captured and imported fish in Kenya. Aquaculture Studies, 19(1), 57–67. https://doi.org/10.4194/2618-6381-v19_1_06 Bizot, O. (2022, November 18). “Thousands of livelihoods destroyed” after masses of fish die in Kenya’s Lake Victoria. The Observers. https://observers.france24.com/en/africa/20221118-thousands-of-livelihoods- destroyed-after-masses-of-fish-die-in-lake-victoria Boyd, C. E., Tucker, C., Mcnevin, A., Bostick, K., & Clay, J. (2007). Indicators of resource use efficiency and environmental performance in fish and crustacean aquaculture. Reviews in Fisheries Science, 15(4), 327–360. https://doi.org/10.1080/10641260701624177 Bureau, D. P., & Hua, K. (2010). Towards effective nutritional management of waste outputs in aquaculture, with particular reference to salmonid aquaculture operations. Aquaculture Research, 41(5), 777–792. https://doi.org/10.1111/j.1365- 2109.2009.02431.x Cai, H., Ross, L. G., Telfer, T. C., Wu, C., Zhu, A., Zhao, S., & Xu, M. (2016). Modelling the nitrogen loadings from large yellow croaker (Larimichthys crocea) cage aquaculture. Environmental Science and Pollution Research, 23(8), 7529–7542. https://doi.org/10.1007/s11356-015-6015-0 https://doi.org/10.14321/aehm.027.03.65 https://doi.org/10.4194/2618-6381-v19_1_06 https://observers.france24.com/en/africa/20221118-thousands-of-livelihoods-destroyed-after-masses-of-fish-die-in-lake-victoria https://observers.france24.com/en/africa/20221118-thousands-of-livelihoods-destroyed-after-masses-of-fish-die-in-lake-victoria https://doi.org/10.1080/10641260701624177 https://doi.org/10.1111/j.1365-2109.2009.02431.x https://doi.org/10.1111/j.1365-2109.2009.02431.x https://doi.org/10.1007/s11356-015-6015-0 92 Canvas/World_Light_Gray_Base (MapServer). (n.d.). [Map]. Esri. https://server.arcgisonline.com/ArcGIS/rest/services/Canvas/World_Light_Gray_Base /MapServer/tile/{z}/{y}/{x} Cela, S., Ketterings, Q. M., Czymmek, K., Soberon, M., & Rasmussen, C. (2015). Long- term trends of nitrogen and phosphorus mass balances on New York State dairy farms. Journal of Dairy Science, 98(10), 7052–7070. https://doi.org/10.3168/jds.2015-9776 Chary, K., Brigolin, D., & Callier, M. D. (2022). Farm‐scale models in fish aquaculture – An overview of methods and applications. Reviews in Aquaculture, 14(4), 2122–2157. https://doi.org/10.1111/raq.12695 Chatvijitkul, S., Boyd, C. E., & Davis, D. A. (2018). Nitrogen, phosphorus, and carbon concentrations in some common aquaculture feeds. Journal of the World Aquaculture Society, 49(3), 477–483. https://doi.org/10.1111/jwas.12443 Cobo, M. J., López-Herrera, A. G., Herrera-Viedma, E., & Herrera, F. (2011). An approach for detecting, quantifying, and visualizing the evolution of a research field: A practical application to the Fuzzy Sets Theory field. Journal of Informetrics, 5(1), 146–166. https://doi.org/10.1016/j.joi.2010.10.002 Fadum, J. M., & Hall, E. K. (2022). The interaction of physical structure and nutrient loading drives ecosystem change in a large tropical lake over 40 years. Science of The Total Environment, 830, 154454. https://doi.org/10.1016/j.scitotenv.2022.154454 Fadum, J. M., Ross, M. R. V., Tenorio, E. A., Barby, C. A., & Hall, E. K. (2025). Nutrient loading from a sustainably certified aquaculture operation dwarfs annual nutrient inputs from a large multi‐use watershed, Lake Yojoa, Honduras. Earth’s Future, 13(3), e2024EF004807. https://doi.org/10.1029/2024EF004807 https://server.arcgisonline.com/ArcGIS/rest/services/Canvas/World_Light_Gray_Base/MapServer/tile/%7bz%7d/%7by%7d/%7bx%7d https://server.arcgisonline.com/ArcGIS/rest/services/Canvas/World_Light_Gray_Base/MapServer/tile/%7bz%7d/%7by%7d/%7bx%7d https://doi.org/10.3168/jds.2015-9776 https://doi.org/10.1111/raq.12695 https://doi.org/10.1111/jwas.12443 https://doi.org/10.1016/j.joi.2010.10.002 https://doi.org/10.1016/j.scitotenv.2022.154454 https://doi.org/10.1029/2024EF004807 93 FAO. (2022). Blue Transformation—Roadmap 2022–2030. FAO. https://doi.org/10.4060/cc0459en FAO. (2024). The State of World Fisheries and Aquaculture 2024. FAO. https://doi.org/10.4060/cd0683en Fiorella, K. J. (2023). Understanding interactions between wild fisheries and aquaculture is essential to sustainable and equitable aquaculture development. Fisheries Management and Ecology, 30(6), 573–577. https://doi.org/10.1111/fme.12576 Frank, S. (2024). Lake Victoria Shapefile [Dataset]. Africa Geoportal. https://www.africageoportal.com/datasets/1c2c6e1107e74e1daef7fb319f1e3553_0/abo ut Furlong, C. & SFD Promotion Initiative. (2016). SFD Report Kisumu, 2016. Sustainable Sanitation Alliance. www.sfd.susana.org Gichuki, C. N., Ndiritu, S. W., & Emodoi, A. B. (2025). Impact of common interest group participation and aquaculture development programs on fish productivity and net returns: Evidence from Nile tilapia farming. Aquaculture International, 33(1), 55. https://doi.org/10.1007/s10499-024-01707-w Gichuru, N., Nyamweya, C., Owili, M., Mboya, D., & Wanyama, R. (2019). Poor management of Lake Victoria fisheries (Kenya); a threat to sustainable fish supplies. In FAO, Africa’s inland aquatic ecosystems: How they can increase food security and nutrition.: Vol. 32(2), 38–43. Nature & Faune Journal. Guya, F. J. (2020). Biogeochemical characterization, phosphorus sources and intrinsic drivers to its speciation within the Nyanza Gulf of Lake Victoria. Lakes & Reservoirs: https://doi.org/10.4060/cc0459en https://doi.org/10.4060/cd0683en https://doi.org/10.1111/fme.12576 https://www.africageoportal.com/datasets/1c2c6e1107e74e1daef7fb319f1e3553_0/about https://www.africageoportal.com/datasets/1c2c6e1107e74e1daef7fb319f1e3553_0/about https://doi.org/www.sfd.susana.org https://doi.org/10.1007/s10499-024-01707-w 94 Science, Policy and Management for Sustainable Use, 25(1), 31–43. https://doi.org/10.1111/lre.12305 Hecky, R. E., Mugidde, R., Ramlal, P. S., Talbot, M. R., & Kling, G. W. (2010). Multiple stressors cause rapid ecosystem change in Lake Victoria. Freshwater Biology, 55, 19– 42. https://doi.org/10.1111/j.1365-2427.2009.02374.x Howarth, R. W., Chan, F., Swaney, D. P., Marino, R. M., & Hayn, M. (2021). Role of external inputs of nutrients to aquatic ecosystems in determining prevalence of nitrogen vs. Phosphorus limitation of net primary productivity. Biogeochemistry, 154(2), 293–306. https://doi.org/10.1007/s10533-021-00765-z Irungu, K. M., & FAO. (2023). Inventory of Cages in Lake Victoria—Kenya [Map]. Lake Victoria Fisheries Organization (LVFO). https://geodata.lvfo.org/geonetwork/srv/api/records/5c272897-0fb8-4f9c-bca8- 9e8d71409b94?language=all Juma, D. W., Wang, H., & Li, F. (2014). Impacts of population growth and economic development on water quality of a lake: Case study of Lake Victoria Kenya water. Environmental Science and Pollution Research, 21(8), 5737–5746. https://doi.org/10.1007/s11356-014-2524-5 Kayombo, S., & Jorgensen, S. E. (2005). Lake Victoria: Experience and Lessons Learned Brief. Lake Basin Management Initiative. https://api.semanticscholar.org/CorpusID:53662964 Krienitz, L., Ballot, A., Wiegand, C., Kotut, K., Codd, G., & Pflugmacher, S. (2002). Cyanotoxin-producing bloom of Anabaena flos-aquae, Anabaena discoidea and https://doi.org/10.1111/lre.12305 https://doi.org/10.1111/j.1365-2427.2009.02374.x https://doi.org/10.1007/s10533-021-00765-z https://geodata.lvfo.org/geonetwork/srv/api/records/5c272897-0fb8-4f9c-bca8-9e8d71409b94?language=all https://geodata.lvfo.org/geonetwork/srv/api/records/5c272897-0fb8-4f9c-bca8-9e8d71409b94?language=all https://doi.org/10.1007/s11356-014-2524-5 https://api.semanticscholar.org/CorpusID:53662964 95 Microcystis aeruginosa (Cyanobacteria) in Nyanza Gulf of Lake Victoria, Kenya. Journal of Applied Botany - Angewandte Botanik, 76(5–6), 179–183. Lex, A., Gehlenborg, N., Strobelt, H., Vuillemot, R., & Pfister, H. (2014). UpSet: Visualization of Intersecting Sets. IEEE Transactions on Visualization and Computer Graphics, 20(12), 1983–1992. https://doi.org/10.1109/TVCG.2014.2346248 Matete, F. (2024, December 28). Court suspends implementation of new fisheries regulations. The Star. https://www.the-star.co.ke/news/realtime/2024-12-28-court- suspends-implementation-of-new-fisheries-regulations Maxwell, J. A. (2013). Qualitative research design: An interactive approach (3rd edition). Sage. Mayan, M. J. (2023). Essentials of Qualitative Inquiry (2nd ed.). Routledge. https://doi.org/10.4324/b23331 McCrary, J. K., Murphy, B. R., Stauffer, J. R., & Hendrix, S. S. (2007). Tilapia (Teleostei: Cichlidae) status in Nicaraguan natural waters. Environmental Biology of Fishes, 78(2), 107–114. https://doi.org/10.1007/s10641-006-9080-x McGee, M. D., Borstein, S. R., Neches, R. Y., Buescher, H. H., Seehausen, O., & Wainwright, P. C. (2015). A pharyngeal jaw evolutionary innovation facilitated extinction in Lake Victoria cichlids. Science, 350(6264), 1077–1079. https://doi.org/10.1126/science.aab0800 Mchau, G. J., Makule, E., Machunda, R., Gong, Y. Y., & Kimanya, M. (2019). Harmful algal bloom and associated health risks among users of Lake Victoria freshwater: Ukerewe Island, Tanzania. Journal of Water and Health, 17(5), 826–836. https://doi.org/10.2166/wh.2019.083 https://doi.org/10.1109/TVCG.2014.2346248 https://www.the-star.co.ke/news/realtime/2024-12-28-court-suspends-implementation-of-new-fisheries-regulations https://www.the-star.co.ke/news/realtime/2024-12-28-court-suspends-implementation-of-new-fisheries-regulations https://doi.org/10.4324/b23331 https://doi.org/10.1007/s10641-006-9080-x https://doi.org/10.1126/science.aab0800 https://doi.org/10.2166/wh.2019.083 96 Midega, C. A. O. (2022). Opportunities for circular bionutrient economy in Kenya: Sanitation and waste stream characterization. Urban Agriculture & Regional Food Systems, 7(1), e20034. https://doi.org/10.1002/uar2.20034 Miima, D. A., Mugalavai, E. M., & Wakhungu, J. W. (2023). Freshwater aquaculture and household performance in Busia County, Kenya. African Journal of Empirical Research, 4(2), 1071–1081. Munguti, J., Obiero, K., Orina, P., Mirera, D., Kyule, D., Mwaluma, J., Opiyo, M., Musa, S., Ochiewo, J., Njiru, J., Ogello, E. O., & Hagiwara, A. (Eds.). (2021). State of Aquaculture Report in Kenya 2021: Towards nutrition sensitive fish food production systems. Techplus Media House, Nairobi, Kenya. Musa, S., Aura, C. M., & Okechi, J. K. (2022a). Economic analysis of tilapia cage culture in Lake Victoria using different cage volumes. Journal of Applied Aquaculture, 34(3), 674–692. https://doi.org/10.1080/10454438.2021.1884632 Musa, S., Aura, C. M., Tomasson, T., Sigurgeirsson, Ó., & Thorarensen, H. (2022b). Impacts of Nile tilapia cage culture on water and bottom sediment quality: The ability of an eutrophic lake to absorb and dilute perturbations. Lakes & Reservoirs: Science, Policy and Management for Sustainable Use, 27(4), e12413. https://doi.org/10.1111/lre.12413 Musa, S., Aura, C. M., Tomasson, T., Sigurgeirsson, Ó., & Thorarensen, H. (2023a). A comparative study of the effects of pelleted and extruded feed on growth, financial revenue and nutrient loading of Nile tilapia ( Oreochromis niloticus L.) cage culture in a lacustrine environment. Journal of Applied Aquaculture, 35(3), 633–655. https://doi.org/10.1080/10454438.2021.2011528 https://doi.org/10.1002/uar2.20034 https://doi.org/10.1080/10454438.2021.1884632 https://doi.org/10.1111/lre.12413 https://doi.org/10.1080/10454438.2021.2011528 97 Musa, S., Aura, C. M., Tomasson, T., Sigurgeirsson, Ó., & Thorarensen, H. (2023b). Nutrient budget of cage fish culture in a lacustrine environment: Towards model development for the sustainable development of Nile tilapia (Oreochromis niloticus) culture. In N. N. Gabriel, E. Omoregie, & K. P. Abasubong (Eds.), Emerging Sustainable Aquaculture Innovations in Africa (pp. 365–381). Springer Nature Singapore. https://doi.org/10.1007/978-981-19-7451-9_15 Njagi, D. M., Routh, J., Odhiambo, M., Luo, C., Basapuram, L. G., Olago, D., Klump, V., & Stager, C. (2022). A century of human-induced environmental changes and the combined roles of nutrients and land use in Lake Victoria catchment on eutrophication. Science of The Total Environment, 835, 155425. https://doi.org/10.1016/j.scitotenv.2022.155425 Njiru, J. M., Aura, C. M., & Okechi, J. K. (2019). Cage fish culture in Lake Victoria: A boon or a disaster in waiting? Fisheries Management and Ecology, 26(5), 426–434. https://doi.org/10.1111/fme.12283 Njiru, J., Van Der Knaap, M., Kundu, R., & Nyamweya, C. (2018). Lake Victoria fisheries: Outlook and management. Lakes & Reservoirs: Science, Policy and Management for Sustainable Use, 23(2), 152–162. https://doi.org/10.1111/lre.12220 Nyamweya, C., Lawrence, T. J., Ajode, M. Z., Smith, S., Achieng, A. O., Barasa, J. E., Masese, F. O., Taabu-Munyaho, A., Mahongo, S., Kayanda, R., Rukunya, E., Kisaka, L., Manyala, J., Medard, M., Otoung, S., Mrosso, H., Sekadende, B., Walakira, J., Mbabazi, S., … Nkalubo, W. (2023). Lake Victoria: Overview of research needs and the way forward. Journal of Great Lakes Research, 49(6), 102211. https://doi.org/10.1016/j.jglr.2023.06.009 https://doi.org/10.1007/978-981-19-7451-9_15 https://doi.org/10.1016/j.scitotenv.2022.155425 https://doi.org/10.1111/fme.12283 https://doi.org/10.1111/lre.12220 https://doi.org/10.1016/j.jglr.2023.06.009 98 Obiero, K., Brian Mboya, J., Okoth Ouko, K., & Okech, D. (2022). Economic feasibility of fish cage culture in Lake Victoria, Kenya. Aquaculture, Fish and Fisheries, 2(6), 484– 492. https://doi.org/10.1002/aff2.75 Obiero, K. O., Abila, R. O., Njiru, M. J., Raburu, P. O., Achieng, A. O., Kundu, R., Ogello, E. O., Munguti, J. M., & Lawrence, T. (2015). The challenges of management: Recent experiences in implementing fisheries co‐management in Lake Victoria, Kenya. Lakes & Reservoirs: Science, Policy and Management for Sustainable Use, 20(3), 139–154. https://doi.org/10.1111/lre.12095 OCHA Regional Office for Southern and Eastern Africa (ROSEA). (2023). Kenya— Subnational Administrative Boundaries [Shapefile]. Humanitarian Data Exchange. https://data.humdata.org/dataset/cod-ab-ken Ochumba, P. B. O. (1990). Massive fish kills within the Nyanza Gulf of Lake Victoria, Kenya. Hydrobiologia, 208(1–2), 93–99. https://doi.org/10.1007/BF00008448 Odada, E. O., Olago, D. O., Kulindwa, K., Ntiba, M., & Wandiga, S. (2004). Mitigation of environmental problems in Lake Victoria, East Africa: Causal chain and policy options analyses. AMBIO: A Journal of the Human Environment, 33(1), 13–23. https://doi.org/10.1579/0044-7447-33.1.13 Ohala Africa Foundation. (2025). The Ohala SAFI Project. https://ohalakenya.org/ohalasafi/ Oirere, S. (2025, March 21). Kenya embarks on process of repealing its fisheries act. SeafoodSource. https://www.seafoodsource.com/news/supply-trade/kenya-embarks- on-process-of-repealing-its-fisheries-act https://doi.org/10.1002/aff2.75 https://doi.org/10.1111/lre.12095 https://data.humdata.org/dataset/cod-ab-ken https://doi.org/10.1007/BF00008448 https://doi.org/10.1579/0044-7447-33.1.13 https://ohalakenya.org/ohalasafi/ https://www.seafoodsource.com/news/supply-trade/kenya-embarks-on-process-of-repealing-its-fisheries-act https://www.seafoodsource.com/news/supply-trade/kenya-embarks-on-process-of-repealing-its-fisheries-act 99 Olokotum, M., Mitroi, V., Troussellier, M., Semyalo, R., Bernard, C., Montuelle, B., Okello, W., Quiblier, C., & Humbert, J.-F. (2020). A review of the socioecological causes and consequences of cyanobacterial blooms in Lake Victoria. Harmful Algae, 96, 101829. https://doi.org/10.1016/j.hal.2020.101829 Orina, P., Ogello, E., Kembenya, E., Muthoni, C., Musa, S., Ombwa, V., Mwainge, V., Abwao, J., Ondiba, R., Kengere, J., & Karoza, S. (2021). The state of cage culture in Lake Victoria: A focus on sustainability, rural economic empowerment, and food security. Aquatic Ecosystem Health & Management, 24(1), 56–63. https://doi.org/10.14321/aehm.024.01.09 Orinda, M., Okuto, E., & Abwao, M. (2021). Cage fish culture in the Lake Victoria region: Adoption determinants, challenges and opportunities. International Journal of Fisheries and Aquaculture, 13(2), 45–55. https://doi.org/10.5897/IJFA2020.0798 Pacini, N., Hesslerová, P., Pokorný, J., Mwinami, T., Morrison, E. H. J., Cook, A. A., Zhang, S., & Harper, D. M. (2018). Papyrus as an ecohydrological tool for restoring ecosystem services in Afrotropical wetlands. Ecohydrology & Hydrobiology, 18(2), 142–154. https://doi.org/10.1016/j.ecohyd.2018.02.001 Peal, A., Evans, B., Ahilan, S., Ban, R., Blackett, I., Hawkins, P., Schoebitz, L., Scott, R., Sleigh, A., Strande, L., & Veses, O. (2020). Estimating safely managed sanitation in urban areas; Lessons learned from a global implementation of Excreta-Flow Diagrams. Frontiers in Environmental Science, 8, 1. https://doi.org/10.3389/fenvs.2020.00001 Roegner, A., Corman, J., Sitoki, L., Kwena, Z., Ogari, Z., Miruka, J., Xiong, A., Weirich, C., Aura, C., & Miller, T. (2023). Impacts of algal blooms and microcystins in fish on https://doi.org/10.1016/j.hal.2020.101829 https://doi.org/10.14321/aehm.024.01.09 https://doi.org/10.5897/IJFA2020.0798 https://doi.org/10.1016/j.ecohyd.2018.02.001 https://doi.org/10.3389/fenvs.2020.00001 100 small-scale fishers in Winam Gulf, Lake Victoria: Implications for health and livelihood. Ecology and Society, 28(1), 49. https://doi.org/10.5751/ES-13860-280149 Roegner, A., Sitoki, L., Weirich, C., Corman, J., Owage, D., Umami, M., Odada, E., Miruka, J., Ogari, Z., Smith, W., Rejmankova, E., & Miller, T. R. (2020). Harmful algal blooms threaten the health of peri-urban fisher communities: A case study in Kisumu Bay, Lake Victoria, Kenya. Exposure and Health, 12(4), 835–848. https://doi.org/10.1007/s12403-019-00342-8 Roulston, K. (2014). Analysing Interviews. In U. Flick (Ed.), The SAGE handbook of qualitative data analysis (pp. 297–312). SAGE. SAWA. (n.d.). Spatial Plan [Map]. Sustainable Activities in Water Areas (SAWA). https://sawa.blue/explore Scheren, P. A. G. M., Zanting, H. A., & Lemmens, A. M. C. (2000). Estimation of water pollution sources in Lake Victoria, East Africa: Application and elaboration of the rapid assessment methodology. Journal of Environmental Management, 58(4), 235– 248. https://doi.org/10.1006/jema.2000.0322 Shitote, Z., Munala, N. O., & Maremwa, J. S. (2022). Feasibility for cage farming in Africa: The case of the Kenyan part of Lake Victoria. Aquatic Ecosystem Health & Management, 25(4), 22–27. https://doi.org/10.14321/aehm.025.04.22 Sitoki, L., Kurmayer, R., & Rott, E. (2012). Spatial variation of phytoplankton composition, biovolume, and resulting microcystin concentrations in the Nyanza Gulf (Lake Victoria, Kenya). Hydrobiologia, 691(1), 109–122. https://doi.org/10.1007/s10750-012-1062-8 https://doi.org/10.5751/ES-13860-280149 https://doi.org/10.1007/s12403-019-00342-8 https://sawa.blue/explore https://doi.org/10.1006/jema.2000.0322 https://doi.org/10.14321/aehm.025.04.22 https://doi.org/10.1007/s10750-012-1062-8 101 Soberon, M. A., Cela, S., Ketterings, Q. M., Rasmussen, C. N., & Czymmek, K. J. (2015). Changes in nutrient mass balances over time and related drivers for 54 New York State dairy farms. Journal of Dairy Science, 98(8), 5313–5329. https://doi.org/10.3168/jds.2014-9236 Soberon, M. A., Ketterings, Q. M., Rasmussen, C. N., & Czymmek, K. J. (2013). Whole Farm Nutrient Balance Calculator for New York Dairy Farms. Natural Sciences Education, 42(1), 57–67. https://doi.org/10.4195/nse.2012.0020 Somdee, T., Thunders, M., Ruck, J., Lys, I., Allison, M., & Page, R. (2013). Degradation of [ Dha 7 ]MC-LR by a microcystin degrading bacterium isolated from Lake Rotoiti, New Zealand. ISRN Microbiology, 2013, 1–8. https://doi.org/10.1155/2013/596429 Song, Y., Li, M., Fang, Y., Liu, X., Yao, H., Fan, C., Tan, Z., Liu, Y., & Chen, J. (2023). Effect of cage culture on sedimentary heavy metal and water nutrient pollution: Case study in Sansha Bay, China. Science of The Total Environment, 899, 165635. https://doi.org/10.1016/j.scitotenv.2023.165635 Stager, J. C., Hecky, R. E., Grzesik, D., Cumming, B. F., & Kling, H. (2009). Diatom evidence for the timing and causes of eutrophication in Lake Victoria, East Africa. Hydrobiologia, 636(1), 463–478. https://doi.org/10.1007/s10750-009-9974-7 Suresh, K., Tang, T., Van Vliet, M. T. H., Bierkens, M. F. P., Strokal, M., Sorger- Domenigg, F., & Wada, Y. (2023). Recent advancement in water quality indicators for eutrophication in global freshwater lakes. Environmental Research Letters, 18(6), 063004. https://doi.org/10.1088/1748-9326/acd071 Svirčev, Z., Baltić, V., Gantar, M., Juković, M., Stojanović, D., & Baltić, M. (2010). Molecular aspects of microcystin-induced hepatotoxicity and hepatocarcinogenesis. https://doi.org/10.3168/jds.2014-9236 https://doi.org/10.4195/nse.2012.0020 https://doi.org/10.1155/2013/596429 https://doi.org/10.1016/j.scitotenv.2023.165635 https://doi.org/10.1007/s10750-009-9974-7 https://doi.org/10.1088/1748-9326/acd071 102 Journal of Environmental Science and Health, Part C, 28(1), 39–59. https://doi.org/10.1080/10590500903585382 Tamatamah, R. A., Hecky, R. E., & Duthie, H. C. (2005). The atmospheric deposition of phosphorus in Lake Victoria (East Africa). Biogeochemistry, 73(2), 325–344. https://doi.org/10.1007/s10533-004-0196-9 Teplitz, E. M., Mwainge, V. M., Otuo, P. W., Ogwai, C., Awandu, H., Onsongo, K., Kemei, N., Gonzalez, G., McIntyre, P. B., Ivanek, R., Getchell, R., Mulanda, C. A., & Fiorella, K. J. (2024). Descriptive characterization of cage aquaculture production practices in Lake Victoria, Kenya to improve fish health management. https://doi.org/10.2139/ssrn.4950050 Teplitz, E. M., Mwainge, V. M., Wacira, T. N., Ogwai, C., Mayianda, M. J., Ochieng, L., Patel, E., Getchell, R. G., Aura, C. M., & Fiorella, K. J. (2025). Disease and antimicrobial resistance surveillance for Nile tilapia pathogens in Lake Victoria, Kenya. Journal of Fish Diseases, e70022. https://doi.org/10.1111/jfd.70022 The Fish Site. (2025, July 9). First tilapia farm in Africa joins ASC Improver Programme. The Fish Site. https://thefishsite.com/articles/first-tilapia-farm-in-africa-joins-asc- improver-programme The Fisheries Management and Development Act, Pub. L. No. 126, Cap. 378 2169 (2024). https://kenyalaw.org/kl/fileadmin/pdfdownloads/LegalNotices/2024/LN126_2024.pdf Van Almelo, J., Ketterings, Q. M., & Cela, S. (2016). Integrating record keeping with Whole Farm Nutrient Mass Balance: A case study. Journal of Agricultural Science, 8(6), 22. https://doi.org/10.5539/jas.v8n6p22 https://doi.org/10.1080/10590500903585382 https://doi.org/10.1007/s10533-004-0196-9 https://doi.org/10.2139/ssrn.4950050 https://doi.org/10.1111/jfd.70022 https://thefishsite.com/articles/first-tilapia-farm-in-africa-joins-asc-improver-programme https://thefishsite.com/articles/first-tilapia-farm-in-africa-joins-asc-improver-programme https://kenyalaw.org/kl/fileadmin/pdfdownloads/LegalNotices/2024/LN126_2024.pdf https://doi.org/10.5539/jas.v8n6p22 103 Yuan, X., Wang, C., Miao, Q., & Zou, C. (2024). Investigation on quantitative evaluation method for feed diffusion effect in deep-sea aquaculture cages. Ocean Engineering, 295, 116759. https://doi.org/10.1016/j.oceaneng.2024.116759 https://doi.org/10.1016/j.oceaneng.2024.116759