Every year approx. half a million hectares of land are burned by wildfires in southern Europe, causing large ecological and socio-economic impacts. Climate and land use changes in the last decades have increased fire risk and danger. In this paper we review the available scientific knowledge on the relationships between landscape and wildfires in the Mediterranean region, with a focus on its application for defining landscape management guidelines and policies that could be adopted in order to promote landscapes with lower fire hazard. The main findings are that (1) socio-economic drivers have favoured land cover changes contributing to increasing fire hazard in the last decades, (2) large wildfires are becoming more frequent, (3) increased fire frequency is promoting homogeneous landscapes covered by fire-prone shrublands; (4) landscape planning to reduce fuel loads may be successful only if fire weather conditions are not extreme. The challenges to address these problems and the policy and landscape management responses that should be adopted are discussed, along with the major knowledge gaps.

T he aim of this work consi.sts of monitoring the recovcry process a&r jiw by mean.s of satellite imagery. The objective,s are to a.s.sess the regrowth pathways followed by diflerent species populations after a disturbance, to analyze the speed of recovery in the years JXlowing fire, and, finally, to estimate rate.s of regrowth. The test area is located in the north of the province of Alicante, on the Mediterranean coast (If Spain. This area, e%specially prone to forest j&s, show.s a remarkable land-use histoq and human pressure. The test areas belong to dicfcrent microclimatic xnes, shozc; diverse tiegetation communities, and have diflerent degrees of .stoniness; so we attempted to discover their post&-e behaviors according to their biogeographical conditions. To accomplish these objectives, we used nine Landsat 5 thematic mapper images from 1984 to IYY4 to which geometric and radiometric corrections were applied. Once the comparability betu>ecn images was guaranteed, we generated a normalized diflercncc vegetation index (NDVL) for each date. First, ux rlelrlonstr~~tclcl that the t1iflerence.s between ND\? itt1agcJ.s were snitablc for mappin, 0 hwned areas. Second, we nndertook a nonlinear regression analysis between NDVI values and the time elapsed .since the jire to a.s.se,s.s the recocety processes. The exponential adjustment between NDVI and time was in accord with the asyn@otic bellacior observed ~chen the rNmX?t7J process is complete. The paramcter,s supplied by the proposed tncthod arc' helpjid in yuanti$jng the ejt2cct.s of jire on dijj%rerrt eco- .

In fire-prone areas, like the Mediterranean, land abandonment and forestation may interact with fire to alter landscape properties and eventually fire hazard and occurrence. However, the spatial interactions among the two processes (land-use/land cover change [LULC] and fire) are poorly known. Here, we analysed the relative effect of LULC change and fire on the landscape structure of an area of Central Spain frequently affected by fire. A series of Landsat MSS images from 1975 to 1990 was analysed to quantify annual changes in LULC, map fire perimeters and evaluate the changes in landscape properties. The temporal dynamics were analysed by annually computing the fraction occupied by each LULC type and landscape structural properties (number, size, shape and arrangement of patches) that might play a role in fire propagation. All of these were calculated separately for the unburned or the burned areas during the study period, as well as for the entire area. At the whole landscape level, or in the unburned area, LULC changes were small, yet the two more flammable LULC types tended to increase, and the landscape tended to become more homogeneous. In the burned area, the area covered by pine woodlands tended to decrease, and that covered by shrublands to increase. Burned areas turned into shrublands only five years after fire. Landscape indices indicative of reduced fragmentation were also found. Both LULC change and fire altered landscape patterns in the whole area to create a less fragmented and more contiguous landscape than in 1975. The changes induced in the whole landscape by fire, in spite of the overall low disturbance rate, were sufficient to closely determine the changes in landscape composition (LULC types) and patterns.

No abstract

No abstract