Links between vegetation patterns, soil C and N pools and respiration rate under three different land uses in a dry Mediterranean ecosystem

Purpose Soil respiration (R-s) is controlled by abiotic soil parameters interacting with characteristics of the vegetation and the soil microbial community. Few studies have attempted a comprehensive approach that simultaneously addresses the roles of all the major factors known to influence R-s. Our goal was to explore the links between heterogeneity in R-s, above-ground plant biomass and belowground properties in three representative land-use types in a dry Mediterranean ecosystem: (1) a 150-year-old mixed Aleppo pine-kermes oak open forest, (2) an abandoned agricultural field, which was cultivated with cereal for several years until abandonment in 1980, when establishment of typical Mediterranean shrubland vegetation started and (3) a rain-fed olive grove, which has been cultivated for 100 years. Materials and methods We selected two distinctive sampling periods coinciding with annual minimum or near minimum (December) and maximum (April) rates of R-s in this dry Mediterranean ecosystem. In each sampling period, R-s, temperature and moisture, aboveground plant biomass, carbon (C) and nitrogen (N) contents in both light and heavy soil organic matter fractions, extractable dissolved organic C (EDOC), as well as microbial and fine root biomass were measured within each land-use type. Results and discussion Across sites, R-s rates were significantly higher in April (3.07 +/- 0.1 mu molm(-2) s(-1)) than in December (1.30 +/- 0.1 mu molm(-2)s(-1)). The labile soil organic matter fractions (light fraction C and N contents, microbial biomass C and EDOC) were consistently and strongly related to one another, and to a lesser extent, to the C and N contents in the heavy fraction across sites and seasons. Linear models adequately explained a large proportion of the within-site variability in R-s (R-2 values ranged from 41 to 91 % depending on land use and season) but major controls on R-s differed considerably between sites and seasons. Primary controls on spatial patterns in R-s were linked to recent plant-derived C inputs in both forest and olive grove sites. However, in the abandoned agricultural field site R-s appeared to be mainly driven by microbial activity, which could be sustained by intermediate or recalcitrant C and N pools derived from previous land use. Conclusions Conversion of native woodland to agricultural land and subsequent land abandonment leads to profound changes in the relationships between R-s, aboveground biomass and belowground properties in this dry Mediterranean ecosystem. While above-and belowground vegetation are the primary controls on spatial variability in labile soil C pools and R-s in the open forest and olive grove sites, a complete lack of influence of current vegetation patterns on soil C pools and respiration rates in the abandoned agricultural field was observed.