Photo Credit: Neil Palmer / CIAT

Original work authored by Nathaniel Robinson, Cindy Cox, Jawoo Koo, and Zhe Guo. Funding provided by PIM / IFPRI.

To help meet Sustainable Development Goals there is a need to develop informative indices that capture the level of sustainable intensification in agricultural systems. Through various indicators that measure the relationship between farming systems and natural resources, we can compare sustainability of agroecosystems geospatially and over time, estimate potential productivity of a place given the availability of natural resources, and identify geographies and farming systems that overly-extract (and thus require high inputs) or under-assimilate available natural resources and ecosystem services (such as water, nutrients, and sunlight). Measuring the impact of agricultural activity on ecosystems and the ability to capture ecosystem services are perhaps the most important and challenging components in the development of such indices (other components include productivity and socioeconomics). Net Primary Productivity (NPP) is one possible co-indicator that, when in combination with other indicators, may provide insight on the status of sustainable intensification in global agriculture.

Values of NPP provide estimates of the net amount of biomass (vegetation) produced by plants in a particular geography. Biomass is the ultimate energy resource for all species on Earth. Tracking the trend of biomass production is important for understanding energy flow in ecosystems. Remote sensing data from satellites are increasingly available, including imagery data for global estimates of NPP at high spatial (about 1 km) and temporal (weekly) resolutions. NPP can be combined with other geospatial information such as agricultural production and technologies that provide more context. For example, trends in spatially disaggregated crop production data and fertilizer use can be compared against human appropriated NPP (HANPP) to provide a proxy indicator for agricultural intensification and sustainability. As a pilot study, we at HarvestChoice estimated NPP and changes in NPP between 2000 and 2005 in Kenya. Our primary objective was to gain insight on the use of NPP as part of a suite of ecosystem-driven indicators that attempt to measure intensification and sustainability in agricultural systems.


Net Primary Productivity

NPP in Kenya is highest in the central highlands on either side of the Rift Valley, near Lake Victoria, and along the coast. Low levels of NPP are shown across all of Northern Kenya into South Eastern Kenya and through the Rift Valley. Between 1999 and 2009 NPP increased overall by approximately .002 kg of carbon per square meter per year.

Spatial Production Allocation Model

SPAM data showed a slight decrease in agricultural production between 2000 and 2005 from approximately 7.88 million tons of carbon to 7.65 million tons of carbon, a decrease of .23 million tons. However, for two of the most common crops in Kenya, maize and beans, production increased between 2000 and 2005. Maize production increased by 247 thousand tons of carbon from 2.82 million tons of carbon to 3.07 million tons of carbon. While beans increased by 20 thousand tons of carbon from 132 thousand tons of carbon to 152 thousand tons of carbon.


The proportion of total NPP removed from the system (NPPh) to the total NPP (NPPact) increased from 2000 to 2005 from 2.9 percent to 4.5 percent (SPAM 2005 data). This is consistent with the HANPP literature for Kenya.

In particular locations in Kenya the ratio of HANPP to actual NPP approaches 1. This is potentially an indication of areas where agricultural production is unsustainable. Either substantial external inputs are necessary to increase agricultural production or the agricultural production in the area is efficiently allocating available resources and managing soil resources.

Production Diversity

We calculated a crude estimate of diversity that only considers the number of different crops per pixel, where NPPi is the total production for the ith crop, and NPPt is the total production for all crops. This value ranges from 0 to 1, 0 being the least diverse and 1 being the most diverse. Diversity can be used as an indicator for sustainability as more diverse systems are generally better able to take advantage of available natural resources by increasing the length of the cropping season, and therefore the green area under the curve, and increasing ecosystem function performed by an assemblage of different crop types. Of particular concern are areas in Western Kenya where it appears the diversity of agricultural production is decreasing.

The results are preliminary but this exploratory exercise at least shows the potential to overlay NPP data with global agricultural data sets, such as SPAM. Our next steps are to include seasonal parameters (rainfall, temperature, solar radiation) so we can overlay these with NPP, determine potential NPP in cropping areas, and scale up our methods to include larger regions in sub-Saharan Africa. We would also like to map belowground NPP and are exploring this possibility.


HarvestChoice, 2014. "Net Primary Productivity and Sustainable Intensification: An Exploratory Exercise." International Food Policy Research Institute, Washington, DC., and University of Minnesota, St. Paul, MN. Available online at

Oct 1, 2014