Soil C and N dynamics in primary and secondary seasonally dry tropical forests in Mexico

The nature and size of soil carbon (C) and nitrogen (N) pools and turnover were compared in secondary and primary forests in a seasonally dry tropical region of Mexico. Total soil C and N, microbial biomass C and N, mineral (ammonium and nitrate) N pools and potential mineralization and nitrification were measured in samples collected during the dry and rainy seasons in early-, mid-, late-successional and primary forests. We hypothesized that the previous agricultural land use of secondary forests would result in lower soil C and N stocks than in primary forest soils, as well as in the seasonal dynamics changes of these two elements. The expected pattern of decreasing soil C and N after a previous agricultural land use did not occur. Soil C was unaffected by the successional stage of the forest. In addition, early- and mid-successional forests registered the highest total and mineral N pools and potential N transformations, whereas primary forests had the lowest N pools and potential cycling. The total soil C and N pools did not change with the sampling season. However, the nitrate pool decreased at the beginning of the rainy season in all forest soils, as did the ammonium pool in primary forests. A striking contrast of the effects of the rainfall (i.e., dry season versus rainy season) seasonality on the microbial biomass and its C:N ratio was observed among forests; late-successional and primary forests recorded the lowest values of both parameters at the beginning of the rainy season, whereas early- and mid-successional forests showed the highest values at this sampling date. Therefore, potential N transformations in all forests were the highest during the rainy season. Our study on the consequences of the land cover change on soils, following the discontinuation of agricultural practices, allows us to conclude that the nutrient dynamics in this ecosystem will vary depending on the successional stage of the forests. This work suggests that the full restoration of soil C and N dynamics will take ca. 60 years of secondary succession. (C) 2005 Elsevier B.V All rights reserved.