Geostatistical methods to study spatial variability of soil cations after a prescribed fire and rainfall

Prescribed fire is a technique sometimes used in forest management but the effects on soil are not fully understood. Soil is a complex system and the spatial variability of properties and processes may increase after a disturbance like fire. We modelled three soil cations Ca2+, Mg2+ and K+ with probabilistic methods and geostatistics in order to assess the spatial effects of a prescribed fire within a plot (108 m2). Soil samples were collected during 6 different sampling campaigns (pre fire, post fire, 2 months, 5 months, 1 year, and 3 years after the prescribed fire). The sampling interval was designed to detect the effect of rainfall (1315 mm for the 3 years with 72 mm/h peak intensity) on the spatial soil changes as consequence of processes such as: leaching, soil erosion, etc. We developed an index of response to compare each cation after the prescribed fire with its pre-fire levels. Results showed that calcium and magnesium recover with time to their original concentrations. Potassium increase slightly in half of the plot just after fire; however, three years after fire the concentration was two orders of magnitude lower than the pre-fire state. The divalent cations (Ca2+ and Mg2+) were more homogeneous but the monovalent cation (K+) was more heterogeneous after the fire. Furthermore the maximum rainfall period triggered different spatial processes and the type of process was a function of the valence of the cation.