Digital mapping of soil nutrients for the Republics of
Burundi and Rwanda
María Ruipérez González1 , Bas Kempen1, Prem S. Bindraban2, Sandra Wolters1, Cyrille Hicintuka3,
Marcien Nibasumba4, Zacharie Nzohabonayo5, John Wendt6, Oscar Nduwimana5, John Veerkamp7
1
4
ISRIC – World Soil Information, Wageningen, The Netherlands, 2 VFRC, Washington D.C., USA, 3ISABU – Institute des Sciences Agronomiques du Burundi, Bujumbura, Burundi,
Mininistère de l’Agriculture et de l’Elevage, Bujumbura, Burundi, 5IFDC, Bujumbura, Burundi, 6 IFDC – East and Southern Africa Division, Nairobi, Kenya7IFDC, Kigali, Rwanda
Abstract
Results
Lack of awareness about soil fertility constraints is a major
limitation to developing sound liming and fertilizer
recommendations in sub-Saharan Africa. Detailed maps of
soil nutrient concentrations and soil acidity can help to
identify areas with soil fertility constraints. For this
purpose, maps of primary (P, K), secondary (Ca, Mg, S)
and micronutrients (Cu, Zn, B), as well as pH, soil acidity
(Al+H), effective CEC and organic matter were generated
for the 0-20 cm soil layer by means of digital soil mapping
using random forest models at a 250 m spatial resolution
for Burundi and Rwanda. The models explained between
20 and 45% of the variation in the data.
Twelve soil nutrient maps were generated for each country. Fig. 2
shows examples for Ca, S, Cu and B. Covariates related to climate
and terrain proved to be the best predictors.
A continuous growth of the VFRC network with
partners from the agricultural and environmental
fields is foreseen, while participation of partners from
the health sector, the food and chemical industries,
as well as national and international development
organizations and policy will be essential for a
balanced representation of required expertise. We
are, therefore, looking for partners and endeavor to
catalyze a global dialogue on R&D of, and the need
for, innovative fertilizers. We therefore invite you to
join VFRC in its effort and look forward to partnering
with you.
Fig. 2. Predicted secondary (Ca, S) and micronutrients (Cu, B) for the 0-20 cm layer.
Objective
To update the currently available soil nutrient maps of Burundi
and Rwanda in terms of resolution and soil variables, including
micronutrients, to support soil fertility management programs.
Methodology
Around 1000 field observations of soil fertility parameters for
each country were used to calibrate random forests models
(Fig. 1).
Over 100 environmental GIS data layers representing land
cover, soil, terrain and climate were used as covariates in the
models (Fig. 1).
Model calibration and mapping was done for each country
separately.
A kriging step was added if the model residuals showed