We need a modern soil health management and monitoring framework

Over 95 percent of smallholder farms in Kenya show severe depletion of essential soil nutrients – nitrogen, phosphorous and potassium. Moreover, Kenya’s agricultural soils have dangerously low soil organic matter and exhibit worrying trends of acidification. This is according to a report by the National Accelerated Agricultural Input Access Programme (NAAIAP) released February 18, 2014.

The report reveals that all the farms surveyed require an average of 8 tons of manure per hectare, in addition to about 200 kilograms per hectare of different compound inorganic fertilizer. The NAAIP soil testing report is based on traditional labor-intensive soil survey methods and conventional soil laboratory techniques. 9600 soil samples were collected from 4800 farms spread over 42 out of 47 counties.

The soil testing report serves the purpose of directing public and policy focus to the magnitude of Kenya’s soil fertility crisis. However, the 464-page NAAIAP report, dense with text and tables is an unequal response to Kenya’s soil fertility and agricultural productivity crisis. The report is gravely limited in its capacity to provide a robust and reliable basis for fertilizer use recommendation at the farm level. Moreover, 9600 samples are too few as a basis for extrapolation over 42 counties, given the high spatial variability of soil properties. Relying on this report to fix our soils and boost land productivity is like using horses and bayonets to wage war in the 21^st century.

Invariably, the soil sampling and analytical methods would yield unhelpful blanket recommendations. For example, it is curious that farms in the humid Trans Nzoia East sub-county and farms in the semi-arid Ijara sub-county in Wajir have the same fertilizer recommendations – 7-8 tons of manure per hectare; 250 kg per hectare of NPK (23:23:0) at planting; 125-150 kg per hectare of CAN.

We need techniques, which can be deployed at a national scale for rapid assessment, testing, generation of spatially explicit soil management recommendations and routine monitoring of soil quality. Management recommendations must be buttressed by farm level agronomic trials, over multiple seasons, to ascertain fertilizer response. The NAAIAP report does not mention any farm level trials.

Advances in infrared spectroscopy, GIS and Remote Sensing now permit rapid diagnostic assessment of soils at higher sampling intensities. These methods were developed and tested 10 years ago at the World Agroforestry Centre here in Nairobi by a team of landscape ecologists, systems agronomists, including myself.

Ethiopia’s Agricultural Transformation Agency is using these techniques to establish the Ethiopian Soil Information System (EthioSIS). The final product will be interactive high resolution (0.1 - 1 ha) grid maps of key soil properties such as pH, texture, mineralogy, organic matter, nutrient content, erosion and other soil degradation prevalence estimates across all of Ethiopia. This system will enable Ethiopia to provide spatially explicit, evidence-based and targeted recommendations for: fertilizer applications, and water management practices.

The hundreds of thousands of archived topsoil held by Kenya Agricultural Research Institute, including the 9600 samples analyzed for NAAIP as well as samples held by universities could be a great starting point for building reflectance spectral libraries for rapid characterization of soil quality at the farm level. Working with farmers and using the Land Degradation Assessment Framework (LDSF), I would recommend additional soil samples from 75 percent of all farms.  

A large national spectral library of soils can be used to set up distributed soil testing and farm advisory centres at the sub-county or ward level. All it would take is an infrared spectrometer, an electric socket and a technician. With such as set up one can establish a farm level soil report card, where farmers, with basic training in sample collection and pre-processing, could turn in soils from their field for testing. The report would contain critical soil quality parameters such as organic carbon, nitrogen, soil acidity, soil bulk density, phosphorous, potassium and other essential elements.

Moreover, the farm level soil report card could also be used to monitor landscape level processes, facilitating identification of areas at risk of soil degradation and designing locally appropriate recommendation domains for integrated management of land health, beyond fertilizer application.

Civilizations have fallen because they failed to manage their soils. We have an opportunity to harness the capability of infrared spectroscopy, remote sensing, spatial statistics, the cellphone and Ushahidi’s crowd sourcing applications to change the way soil information is collected analyzed and used to guide soil management decisions at the farm and landscape level.