Soil dynamics in Pinus halepensis reforestation: Effect of microenvironments and previous land use

Reforestation with Pinus halepensis has been the most frequently used method for restoring supposedly degraded lands in Mediterranean semiarid areas, a key objective of which is to improve soil conditions to trigger succession. In this paper we study the effect of a 30 year-old reforestation on lands formerly dedicated to two different uses: shrubland and barley fields. The study area is located in SE Spain in a semiarid area with an annual rainfall of 298 mm. We hypothesize that the previous land use will influence present day soil characteristics. In addition, reforestations have led to the formation of microenvironments associated to: (i) tree development, which generates a radial gradient of influence; (ii) disturbance by mechanical works with resulting microenvironments such as escarpments, showing a worse initial state and, therefore, slower dynamics. Ten plots were set up on former shrubland and ten on a former barley field. Each plot represented a set of related microenvironments that followed a gradient of tree influence/mechanical disturbance. The microenvironments on former shrubland were: under escarpment (ES), plantation line adjacent to P. halepensis stem (PL), beneath A halepensis crown (BC), the upper bank where part of the material of the terrace was dumped (UB) and original unaltered hillslope taken as pseudocontrol of the original soil (PC). In the former barley field, only PI, BC and PC microenvironments were present as no terracing was carried out. Pine growth created a strong gradient of litter input in the sequence PL>BC>PC which, however, had no significant effect on soil organic carbon (SOC) or other soil variables (except light changes in extractable K and pH). The incorporation of organic matter into the soil was probably delayed by the quality of recalcitrant pine litter and the unfavorable semiarid climate. In addition, disturbance created by reforestation works seems to have been buffered by redistribution of fine particles through short-range erosion and sedimentation. In the external part of the terrace, the values of SOC and extractable Na were higher, accompanied by increased electrical conductivity and decreased pH. The most striking differences found were related to former land use. SOC was three times higher in former shrubland than in former barley field. Likewise, available P was higher in the former barley field and extractable Na was higher in the former shrubland, both apparently related to the pre-reforestation state. The results indicate that soil dynamics thirty years after the reforestation has hardly erased the differences attributable to former land use. Soil dynamics induced by the input of organic matter from trees is slower that expected when the reforestation was planned. We conclude that any future application of reforestation should take into account the slow soil dynamics observed in this study in order to be more effective. (C) 2009 Elsevier B.V. All rights reserved.