4.1 Agricultural intensification and contamination in Lake Naivasha catchment
Based on the secondary geospatial assessment, there has been an expansion of intensified agriculture in the L. Naivasha catchment–. Agricultural intensification between 1989 and 2019 in the L. Naivasha catchment suggests dynamic landscape changes, leading to contamination of surface waters. At the same time, lower river reaches have tended towards extensive agricultural production systems, while upper and middle reaches have tended towards intensification. The population changes in the upper reaches of L. Naivasha catchment , , and subsequent land subdivisions for small scale farming are possible explanations of intensification potential . Boserup (1965) postulated increased population density promotes agricultural intensification, with higher densities (above 600 persons per km2) resulting in unsustainable intensification . The implication of this is the degradation of the landscape in the catchment as a result of agricultural expansion , . Considering that increased forest and shrub cover enhances evapotranspiration, while increased bare land enhances surface water release, thus, increased bare land can have a negative impact on both water quality and quantity through sediment and pollutant emissions to surface waters . Further, intensification in the upper reaches suggest that agricultural activities would be the main source of agrochemicals , transported to L. Naivasha, highlighting the need to understand the spatial distribution of nutrients and pesticides in the catchment.
While the distribution of nutrient concentration in the catchment is likely to be a consequence of diffuse sources from land-use, the significantly higher nutrient concentrations at K1 (River Karati Highway Bridge) suggest higher point-source contribution from extensive agricultural farms around the site. The particular (K1) site has been documented by previous studies to show high concentrations of heavy metals , , and nutrients ,  which indicates potential negligent use of agrochemicals. Although Everard et al.  argue that the geological difference between the Karati River sub-catchment and the other river sub-catchments could explain the high chemical contamination at K1, the consistently higher concentrations at the site, indicates anthropogenic, point-source pollution. This study established that within the lake, the river mouth site (N1 – River Mouth) acts as major source of pollution into the lake, registering higher concentrations of nutrients compared with the other sites in the lake. Although similar trend of results within the lake have been reported, the current study recorded three times higher concentrations of total phosphorus compared with Kitaka et al.  and 10 % higher total nitrogenthan recorded by Ndungu et al. . The relatively lower concentrations of total nitrogen and phosphorus in lakes compared with rivers is consistent with both dilution and settling effects , , making L. Naivasha a sink for the nutrients accumulated from the rivers, a concept that was first documented by Vollenweider and Kerekes .
Although DDT is banned in Kenya, except for use by public health officials , its presence in high concentration in surface waters of the L. Naivasha catchment, and the subsequent high DDT:DDE ratios across the catchment indicate recent and continued use. The ∑DDT was highest in the lake samples, indicating that the lake is also a sink of DDT used in the catchment. The persistence of DDT in sediments could also indicate potential resuspension from the sediments , as well as the higher distribution potential of DDT compared with HCH or cyclodienes . ∑HCH was highest in the lake, which supports the findings by Onyango et al.,,  that there is continued application of technical lindane around the lake. While the concentrations reported in this study are within range of earlier studies by Onyango et al., , it is up to 700% higher than the values recorded in L. Naivasha by Njogu . The concentrations of ∑cyclodienes were highest in the lake sites, with the main river source of the cyclodienes being River Gilgil, and lower reaches of River Malewa. The distribution of the cyclodienes within the lake showed high concentrations within the deeper parts of the main lake (M3 – Hippo point) which has been documented by Ndungu et al. and Outa et al. as an area of lower disturbance. Further, the current study reports higher concentrations of cyclodienes compared with those reported by Gitahi et al. , suggesting continuous increase in use of cyclodienes based pesticides.
Despite that there was no significant relationship between the contamination by pesticides and enrichment of nutrients across the sites, it would be expected that agricultural intensification involved concurrent use of nutrients and pesticides , , and reflected in a correlation between occurrence and concentrations. However, differences in modes of action between nutrients and pesticides (Bainbridge et al., 2009; Damalas and Eleftherohorinos, 2011; EU, 2011) and the pathways of transfer to aquatic sources ,  would explain the absence of correlation between nutrients and pesticide concentrations in L. Naivasha catchment.
4.2 Implication of agricultural intensification in relation to aquatic ecology
Rattan & Chambers  argued that TN:TP stoichiometry was positively correlated with crop cover, associating agricultural land use and TN:TP ratios. On the other hand, the majority of the sites indicated a potential nitrogen limitation, but increased application of phosphorus-based fertilizers , , . There is continued use of DDT and technical HCH in the catchment, with consequent emissions to both the river sediments and lake water.
Evidence of increased pesticide residues in the forest areas are comparable with the higher concentration of DDT reported by Onyango et al.  within the upper reaches of the R. Malewa sub-catchment. Odongo et al.  and van Oel et al  argue that increasing forest cover is negatively related to proportion of agricultural land, leading to expected lower anthropogenic input of nutrients and pesticides into the aquatic ecosystem. However, forested areas in the L. Naivasha catchment are associated with fragmented, and small scale, intensified agricultural production, which could explain the increased pesticide concentrations, particularly through point sources in the forested areas. On the other hand, extensive systems can be associated with agricultural expansion without sustainability safeguards , especially in many of the catchments in sub-Saharan Africa with little to no strategic catchment management strategies. In the L. Naivasha catchment, increased and emission of nutrients and pesticides indicates, unsustainable intensification/expansion, with potential impact on biodiversity . An overview would be that it is a fragmented landscape, with trends of reduced forest cover and grassland, and increased crop production. The net result is increased diffuse fertiliser and pesticides pollution. This suggests the need to apply agricultural extension services to better manage sustainable intensification, with the goal to reduce overall agrochemical contamination , , coupled with improved monitoring of the use of inputs in catchments, and consideration of precision farming practices .
Integrated monitoring of nutrient and pesticide concentrations in surface waters and their association with LULC is a major need for development of policies and to support water and land management in many African catchments subject to intensification, or where plans are developing for intensification for food security and economic development . Achieving that without moving beyond environmentally safe limits with longer term impact on land productivity, ecosystem integrity and public health is a major challenge . When catchments include protected areas, or sites that have been designated under international agreement such as is the case for L. Naivasha as designated Ramsar site the balancing act of environmental protection can be even more difficult.
4.3 Management of agricultural intensification
The study identified excess recent and continued inputs of pollutants from agriculture to the surface waters of the L. Naivasha catchment as a management gap. This includes potential impact from banned pesticides used across the catchment; land use changes with complementary intensification practices; and high to very high risk of combined nutrients and pesticides chemical pollution in surface waters. These risks occur across many agricultural catchments in sub-Sahara Africa (SSA) , –. Standards for managing aquatic resources are highlighted in the Africa Water Vision 2025  advocating for a revision of water regulations and laws to give attention to water quality management. However, the blueprint ignores the potential of ecosystem-based management to allow land-use based water quality management and allocates only 3% of the annual investment (USD 0.6 billion) to institutional and policy reforms, and research and development. Based on the findings in this study, the revision of the Africa Water Vision 2025 needs to take into consideration water quality standards that incorporate combined nutrients and pesticide use and their emissions from land to water.
Kenya’s regulations on water quality standards  does not consider the potential of combined effects of nutrients and pesticide, neither does it consider monitoring, review, and policing framework for banned substances in surface waters. Moreover, the standards do not have provisions to monitor effluents associated with banned pesticides. This is a clear indication on mismatch between regulation, and enforcement, that requires the responsible agencies including the Water Resources Authority (WRA), National Environment Management Authority (NEMA), and the Kenya Bureau of Standards (KEBS) to actively revise the water quality standards, enforce the management of the standards, and invest in policy and institutional reforms to address management of agricultural intensification.
Policy and institutional reforms contributing to development of regulations that consider the increased risks to surface water and promotes enforcement for compliance to the regulations can provide considerable benefits for sub-Saharan Africa agricultural management. Examples for action include investment from nutrient credit trading, payment for environmental services, innovating financing, and financial instruments . Partnerships among national and subnational governments and the private sector can include more focus on the monitoring of aquatic resources through definition of indicators combining biological, physical, and chemical monitoring . Further, the partnerships can promote coherence of regulations across jurisdictions, and coordination among strategic water quality management and policy authorities. Catchment management partnerships – such as Imarisha in Naivasha , have an opportunity for inclusive and integrated management. At the same time, a monitoring regime which will require availability of consistent and continuous water quality data, to track and document progress in the development of water resources regulations and policy. The water quality guidelines in Kenya provides for frequency of monitoring, single chemical and biological standards, methodologies for collection and analysis of samples, and a reporting framework . However, the standards do not take into consideration combined contamination.
The study shows that intensification, accompanied by increased application of nutrients and pesticides , needs sustained advisory capacity in the use of agricultural inputs . However, the capacity and resources to manage the advisory services are still underdeveloped in Kenya, as in many sub-Saharan African countries. Reviewing the water quality standards taking into consideration combined contamination would need the integration of advisory platforms, including online advisory, that controls the emissions of pollutants to surface waters. Inadequate resources to manage advisory services, is exacerbated by inadequate standardized monitoring methodologies. The United Nations Environment Programme  in its Progress on Ambient Water Quality update, report findings from methodological considerations for monitoring water quality. However, the report recommends that the monitoring of ambient water quality should use national and/or subnational water quality standards. In many SSA countries, the standards are not comprehensive, a situation that reinforces the need for integrated review for water pollution management at national level.
An integrated review of standards would contribute to achieving and compliance with the Sustainable Development Goals (SDGs) including: SDG 2 through promotion of sustainable agriculture practices, SDG 14 through reduction of risks to aquatic biota, SDG 3 through reducing the potential of health complications from exposure to contaminants in the aquatic systems, SDG 6 through availing better water quality and promoting sanitation, and SDG 17 through promoting inclusive partnerships for monitoring and review. Further SDG indicator 6.3.2, provides a mechanism for determining whether, and to which extent, water quality management is successful, with a target to increase the proportion of water bodies with good water quality . The findings from this study, emphasizes the need for water quality managers to consider combined contamination from nutrients and pesticides, and how that is monitored.