Predicting decomposition patterns of tree biomass in tropical highland microregions of Kenya
Abstract
Decomposition- and nitrogen-release patterns of biomass from three agroforestry multipurpose trees (Calliandra calothyrsus, Cordia africana and Grevillea robusta) were investigated in four contrasting environments (microregions) in the Kenyan tropical highlands during two cropping seasons. Dried leafy biomass was placed in 2-mm litter bags, buried at 15-cm depth and recovered after 2, 4, 7, 10, 15 and 20 weeks. Decomposition patterns were best described by first-order exponential decline curves. The decomposition rate constants ranged from 2.1 to 8.2 yr−1, and the rates of decomposition among the species were in the order: calliandra ≥ cordia > grevillea. There was a species-by-environment interaction during both seasons, but the nitrogen released did not follow such a pattern. Among the three tree species, calliandra released the highest amount of cumulative N, followed by cordia and grevillea. Using multiple regression techniques, decomposition pattern was described as a function of three groups of factors: biomass quality (N, C, lignin and polyphenol), climate (soil temperature and rainfall), and soil conditions (pH, soil organic C, total N and P). For all the species and factors combined, the adjusted R 2 values were 0.88 and 0.91 for seasons 1 and 2, respectively. Among the three groups of factors, climate and biomass quality had the most influence on decomposition rates. Climatic factors accounted for 75% of the total rate of decomposition in season 1 (‘irregular’ season with less rainfall and more soil temperature fluctuations), whereas biomass quality factors were more influential in season 2 (‘regular’ season), accounting for 65% of the total variability.
URI
https://www.researchgate.net/publication/226528442_Predicting_the_decomposition_patterns_of _tree_biomass_in_tropical_highland_microregions_of_Kenyahttp://hdl.handle.net/123456789/279