Dynamics of land-to-lake transfers in the Lake Victoria Basin

They presented their findings at a stakeholder workshop in Kisumu at the Kenya Marine Fisheries Research Institute (KMFRI). Stakeholders were gathered from academia, research institutes and government, along with community representatives and they were all invited to share their experiences in land and lake management.  

Changes to Kenyan land-management practices are urgently needed for sustainable agricultural productivity and to reduce the growing problem of soil erosion and transfers into Lake Victoria, which compromise the growing economic and food security dependence on fisheries aquaculture. The goal of the workshop was to find a way to better coordinate disparate research and create partnerships to improve communications that will better inform land-management decision makers. 

The workshop 

The workshop covered three exercises across three groups, with the first two set up as ice breakers and to help participants consider how they could translate research findings into impact. These exercises directed discussion towards the preparation of policy briefs: what they are, who reads them and how they can be effective, alongside examples of community engagement to change behaviour or practice.  

The third exercise generated the most discussion on how data and research should be coordinated and shared, with examples of good practice being quite rare owing to a lack of resources and expertise. Additionally, the exercise discussed whether there could be a process to enforce the delivery of research harmonisation, improved reliability of data quality and the ability to consider multiple research outcomes from numerous projects. Again, examples of good practice were limited, although less direct means of communication to policy decision makers were discussed; for example, via media or a community bottom-up approach. 

Overall, the workshop demonstrated that the key to enabling a positive change in behaviour and practice for land and lake management is ensuring community engagement from the outset of a research project, such as engaging with focus groups, having community representatives or enabling citizen science participation. The workshop participants agreed that a committee was essential to share research outputs with relevant stakeholders in the Lake Victoria basin. A virtual platform is also essential to a functional framework, so that research outcomes are better shared, data is used in multiple ways to realise efficiency gains and long-lasting impact is created from research. However, such a platform would require adequate resourcing and continued support from stakeholders, alongside engagement from policy decision makers.  

Joint research with the University of Eldoret and KMFRI 

Research shared with this group was funded via previous Royal Society and NERC grants. It initially involved mapping the geochemistry across the Winam Gulf catchment of the Lake Victoria basin, to model the areas at greatest risk of soil erosion and identify more precisely the locations within a river catchment suitable for targeting limited resources to train farmers and test intervention methods to reduce soil erosion.  

Secondly, sediments and sediment cores were surveyed across the Winam Gulf catchment, to generate a chronology of sediment run-off over the past 100 years, as well as the extent of metal and nutrient land-to-lake transfers. This will help us to better understand the effects of poor land management on the lake environment.  

Field collections, measurements, and digitisation and modelling of data were described in previous blogs, with partnerships cemented by the exchange of technicians, early career researchers and principal researchers. The opportunities created by this collaborative project collectively and individually demonstrate the potential for scientific research to address environmental issues whilst developing scientific capacity in Kenya and the UK. The two-way exchange of staff and paired UK/Kenya PhD students provided an enriching experience for all involved. 

Continuing partnerships  

Ongoing efforts will see improvements to modelling the risk associated with soil erosion, including translating a machine-learning model for predicting risk at the sub-catchment scale to other similar land-lake environments, to determine changes in soil losses/sediment transfer at historical scales (100 years) and through dynamic modelling (within few years). Funding to support an advisory forum for Lake Victoria has the potential to set a template for the region on how to better coordinate research data. This will be of particular value to developments in machine learning, which can analyse vast amounts of data quickly. Machine learning could provide a broader, more effective perspective beyond the typical scope of a research project but is dependent on harmonised approaches to data capture and quality. 

Data and information are very expensive due to their endless nature and the value attached to them; thus, quality collaboration between KMFRI, BGS and UoE has created a unique platform to provide baseline and useful data and information on lake-land interphase, which will form the foundation of lake-basin management, planning and conservation in the region. 

Data and information are very expensive due to their endless nature and the value attached to them; thus, quality collaboration between KMFRI, BGS and UoE has created a unique platform to provide baseline and useful data and information on lake-land interphase, which will form the foundation of lake-basin management, planning and conservation in the region. 

Acknowledgements 

We would like to thank: 

• Olivier Humphrey and Andy Marriot, who provided expertise in machine learning, and sampling strategy and fisheries, respectively (BGS) 

• Job Isaboke and Sophia Dowell, joint UK/Kenya PhD students 

• Staff from all three institutes that supported laboratory and field work, logistical arrangements and community engagement 

• Collins Ongore, Job Mwamburi and George Basweti (KMFRI)  

• Elliott Hamilton and Amanda Gardner (BGS) 

• Prof William Blake for guidance on soil erosion sampling strategy and translation of data outcomes into useful data tools to advise on land management (University of Plymouth). 

Funding 

This work was financially supported by:  

• Natural Environment Research Council (grant numbers NE/R000069/1, NE/X006255/1, NE/S007334/1 and GA/19S/017) 

• Royal Society (grant number ICA/R1/191077]) 

• British Academy (grant number WW21100104] ) 

• Commonwealth Scholarship Council UK for professional fellowships  

About the authors 

Dr Christopher M Aura: Director of Freshwater Research at the Kenya Marine Fisheries Research Institute. Chris was a co-PI on the joint research, with oversight on the lake management, sampling and community engagement. 

Prof. Odipo Osano: Professor of Environmental Sciences at the University of Eldoret.  Odipo was a co-PI on the joint research, with oversight on the land sampling and community engagement, overall coordination of Kenyan activities. 

Dr Michael Watts: Head of Inorganic Geochemistry and Lead for International Geoscience R&D at BGS.  Michael was the PI for UK funded grants and overall coordinator for the project.