Shrubland responses to experimental burning, cutting and ploughing treatments were studied over 15 years in two shrubland communities dominated by Erica australis. The treatments represent the most frequent forms of anthropogenic disturbances experienced by these communities throughout their history. The response to burning and cutting treatments is similar, and the succession process is characterised as autosuccession. The highest values for herbaceous annuals and perennials were observed in the third and fourth years. Generally, herbaceous species remain present throughout the study period, while woody taxa ones increase their cover values over time. The quantity of herbaceous species present is in inverse proportion to the quantity of woody taxa. The woody species that appear immediately after treatments are sprouting species, namely Erica australis and Arctostaphylos uva-ursi. The response to ploughing is slower, reflecting the recovery mechanism (seedlings). However, after 15 years, there are no significant differences in regeneration between treatments. The first stages of this post-ploughing succession are dominated by annual species until the fourth or fifth years, after which woody species begin to dominate and herbaceous taxa decrease considerably. Woody species with high germination values are Halimium alyssoides and Halimium umbellatum. These shrubland communities have a very high resilience to such perturbations and start regenerating rapidly, reaching the original state in about 9 years. The appearance of the climax arboreal species of the area, Quercus pyrenaica, when it comes from germination, occurs 15 years after the perturbations.

The effects of wildfire on vegetation regeneration in communities dominated by Quercus pyrenaica and those dominated by Pinus pinaster in NW Spain were compared. In order to study changes in the composition and structure of both types of community, permanent plots were established in areas dominated by Q. pyrenaica and those dominated by P. pinaster. All were burned by wildfires at the end of summer. In each plot a permanent transect of 20 m  1 m was established. Basal cover of the plant species present was analysed. In both types of community the global cover values generally increased throughout the study period. In the oak communities the species that appear in the first years are those that will dominate in the mature stage, like Q. pyrenaica and Erica australis. Both species are typical resprouters: from shoots on the rhizome or the stem of subterranean roots in the case of Q. pyrenaica and from lignotubers in the case of E. australis. Among the other species, herbaceous perennials dominated during the first year after the fire, Luzula lactea being the most representative. The percentage of bare soil decreased very rapidly after the first year of regeneration. However, in the P. pinaster communities the species that appeared with higher cover values during the first and second year after fire were seeders, like P. pinaster and Halimium alyssoides. Other species that appeared in these communities were Chamaespartium tridentatum, and E. australis. Amongst the herbaceous perennials, the most representative species was the Liliacea Ornithogalum umbellatum, which appeared throughout the study period in all the burned plots. The percentages of bare soil were higher than in the oak communities. Structural parameters such as diversity and specific richness were much higher in the community dominated by oak than in the pine stand. In general, regeneration after wildfire in the Pinus community was slower than in oak communities.

The aim of this study is to compare the recovery dynamics in three shrub communities subjected to experimental burning and cutting, and situated on an altitudinal gradient. Climatic features are different in each area, but all had the common characteristic of very homogeneous vegetation cover before the disturbances, with only one shrub species clearly dominant, a different taxon in each area, and with different regeneration strategies. The first area was a heathland dominated by Calluna vulgaris, situated at an altitude of 1600 m, with a continental climate (mean annual precipitation 1320 mm). The second area was a heathland dominated by Erica australis, located at an altitude of 1000 m (mean annual precipitation 840 mm). The third area was a Cistus ladanifer shrubland, located at 900 m altitude, with a Mediterranean climate similar to that of the previous area, but with lower mean annual precipitation (470 mm). Erica australis recovers by vegetative resprouting, but Cistus ladanifer is an obligate seeder, as is Calluna vulgaris in these areas. Each experimental disturbance was carried out over 100 m 2 in each area. Post-fire recovery is faster in Cistus ladanifer: 2 years after burning there was 40% cover vs. less than 20% in the other two species. However, recovery after cutting was similar for Cistus ladanifer and Erica australis. Calluna vulgaris recovers very slowly, with cover values below 20% even 10 years after both disturbances. Cover of dominant shrub species is negatively correlated with cover of herbaceous species. So different recovery of dominant species lead a different community dynamic in each area.

A detailed understanding of the spatial patterns of burning is valuable for managing biodiversity and ecosystems. This research assesses the performance of several spectral indices derived from Landsat data when modelling fire occurrence probability by means of logistic regression. Normalized Difference Vegetation Index (NDVI), Normalized Difference Moisture Index (NDMI), Normalized Burn Ratio (NBR) and the greenness and wetness components of the Tasseled Cap Transformation were tested. Landscape variables (topography, accessibility and structural vegetation) were also included as predictors in models development. Although fire risk is closely related to weather and vegetation status at a given time, it is also strongly linked to fire history, and changes in predictor values in years previous to the fire were events also considered. The models generated correctly classified about 70% of the validation data set. The inclusion of pre-fire spectral indices improved models ability to predict fire occurrence. Although the NBR-based model was the most accurate, TCWetness and NDVI-based models showed similar results, while TCGreenness performed worst. Models with no spectral indices described the fire-proneness of the landscape structure, while the inclusion of spectral indices improved the recognition of particular spatial conditions. Slope and distance to the nearest path were also identified as valuable predictors. All the models identified the main fire risk zones in the study area. Their integration into a single, integrated model properly described fire-proneness and is suggested to be a valuable tool for the identification and management of fire risk. The method used is simple, describes the key variables and spatial pattern of the fire regime and is suited to operational use in Mediterranean ecosystems.

No abstract