Olof Andrén, Job Kihara, André Bationo, Bernard Vanlauwe, Thomas Kätterer
AMBIO: A Journal of the Human Environment 36 (5), 379-386, (1 July 2007) https://doi.org/10.1579/0044-7447(2007)36[379:SCADAI]2.0.CO;2
Soil biological activity was calculated on a daily basis, using standard meteorological data from African weather stations, a simple soil water model, and commonly used assumptions regarding the relations between temperature, soil water content, and biological activity. The activity factor re_clim is calculated from daily soil moisture and temperature, thereby taking the daily interaction between temperature and moisture into account. Annual mean re_clim was normalized to 1 in Central Sweden (clay loam soil, no crop), where the original calibration took place. Since soils vary in water storage capacity and plant cover will affect transpiration, we used this soil under no crop for all sites, thereby only including climate differences. The Swedish re_clim value, 1, corresponds to ca. 50% annual mass loss of, e.g., cereal straw incorporated into the topsoil. African mean annual re_clim values varied between 1.1 at a hot and dry site (Faya, Chad) and 4.7 at a warm and moist site (Brazzaville, Congo). Sites in Kenya ranged between re_clim = 2.1 at high altitude (Matanya) and 4.1 in western Kenya (Ahero). This means that 4.1 times the Swedish C input to soil is necessary to maintain Swedish soil carbon levels in Ahero, if soil type and management are equal. Diagrams showing daily re_clim dynamics are presented for all sites, and differences in within-year dynamics are discussed. A model experiment indicated that a Swedish soil in balance with respect to soil carbon would lose 41% of its soil carbon during 30 y, if moved to Ahero, Kenya. If the soil was in balance in Ahero with respect to soil carbon, and then moved to Sweden, soil carbon mass would increase by 64% in 30 y. The validity of the methodology and results is discussed, and re_clim is compared with other climate indices. A simple method to produce a rough estimate of re_clim is suggested.