Modelling hydrological connectivity in burned areas. A case study from South of Spain

Abstract

Hydrological connectivity (HC) depends on the spatio‐temporal interactions of hydrological and geomorphic processes as well as on the human footprint on the landscape. This study deals with the modelling of hydrological connectivity in a burned area with different levels of fire severity. Namely, the objectives are to: i) characterize and ii) modelling the pre‐ (PreF) and post‐fire (PostF) scenarios, as well as iii) evaluate the effect of the vegetation changes due to the fire and the initial post‐fire management practices (construction of new skid trails and check‐dams) on the magnitude and spatial pattern of connectivity. Four post‐fire scenarios are simulated: immediately after the fire (PostF1), with new skids and without check‐dams (PostF2), with new skids and check‐dams and without vegetation recovery (PostF3), and with new skids and check‐dams and incipient vegetation cover (PostF4). The study area corresponds to eleven headwater subcatchments (total area of 330 ha) that cover the entire burned area of the mountain in the West and Southwest facing hillslopes. This site is located in the province of Malaga, South of Spain, and all sub‐catchments are disconnected between them. The fire started in 2014, 27 June and lasted two days affecting 222.9 ha. The landscape is mainly mountainous, with very steep slopes and marble rocks, Mediterranean climate, and land use of shrubs and pine forests (pre‐fire scenario). Settlements, olive and fruit groves appear at the bottom of the slopes. After the wildfire, land management were carried out in order to remove completely the burned trees and thus new skid trails were built. Then, eleven concrete check‐dams and twelve wooded check‐dams were built in the main gullies. The different scenarios of linear landscape elements, vegetation cover and modifications on the topography related to the construction of new trails and check‐dams were included in the simulations. The IC index of hydrological connectivity was chosen to perform this metric at a spatial resolution of 2.5 x 2.5 meters. The C‐RUSLE factor and the D‐Infinity algorithm were chosen to assess the downslope and upslope components of the simulation. The analysis of the different spatial patterns and temporal changes was done considering the different levels of fire severity and changes on hydrological connectivity were analysed at each sub‐catchment. Most of the burned area was affected by very low and low burned severity with 32.8 and 41.0% of the total area, respectively. The tracks and the different gully networks defined the main overland flow pathways. Values of HC changed in accordance with the changes in the vegetation cover, whereas the spatial patterns of HC changed following the different post‐fire management practices. The overall value of HC in the eleven sub‐catchments increased 31.4% after the forest fire, and slight increments of connectivity were simulated during the second (1.8%) and third (6.5%) post‐fire scenarios. However, the incipient recovery of the vegetation explained the small decrease of HC in the fourth post‐fire scenario (‐3.3%). We can conclude that the forest fire and the subsequent management practices triggered a clear increment of the HC in the study area. Further research should be focussed on the effect of silted check‐dams, the progressive vegetation recovery and other support practices that can help to reduce the observed and simulated high values of connectivity in the burned area.