Mimosa species (Leguminosae) and rainfall seasonality affect spatial dynamics of soil carbon and nitrogen in a semiarid ecosystem

Abstract Background and aims The Mimosa species (Leguminosae) promote soil fertility and restoration; however, their impact on soil C and N cycling have never been assessed, and even neither with respect to their topographic positions at semiarid ecosystems. Our objective was to examine seasonal variation of the effect of three Mimosa species on soil C and N dynamics, including M. luisana at two contrasting topographic positions. Methods We measured the total and available forms of soil C and N, microbial biomass, and net mineralisation and nitrification rates in soils collected under the canopies of M. texana var. filipes , M. polyantha , M. luisana , and in open areas at plain sites, as well as under the canopy of M. luisana and open areas at plain vs. hillslope sites, during the dry, early-rainy and rainy seasons. Results At plain sites, the three Mimosa species contributed with higher litter mass and quality, and 1.5 to 2 times more soil C and N concentrations under their canopies. The soil under Mimosa species had the double or more microbial biomass, labile C, inorganic N as well as faster C and N mineralisation and nitrification rates than open areas. Soil N mineralisation and nitrification rates varied with rainfall seasonality, having a peak during the rains that ranged from 54 to 202 µg N g −1 and 14 to 180 µg NO 3 g −1 , respectively among Mimosa species. Mimosa luisana affected soil C and N dynamics similarly at plain or hillslope sites. Conclusions The three Mimosa species not only form resource islands but also fertility islands, which couple soil C and N dynamics, acting as sources or sinks of these nutrients depending on rainfall seasonality. Moreover, M. luisana drives soil C and N cycling that may be magnified by the topography. In turn, Mimosa (legumes)-topography interaction should be included in spatial and temporal heterogeneity models of soil resources in semiarid ecosystems.