Research Article

 

Current and future estimates for the fire frequency and the fire rotation period in the main woodland types of peninsular Spain: a case-study approach

 

Antonio Vázquez

Centro de Investigación Forestal CIFOR-INIA. Ctra. A Coruña, km 7.5 28040-Madrid, Spain.

José M. Climent

Centro de Investigación Forestal CIFOR-INIA. Ctra. A Coruña, km 7.5 28040-Madrid, Spain.

Luis Casais

Universidad Complutense de Madrid UCM, Ftad. Farmacia, Ciudad Univ. s/n 28040-Madrid, Spain

José R. Quintana

Universidad Complutense de Madrid UCM, Ftad. Farmacia, Ciudad Univ. s/n 28040-Madrid, Spain

 

Abstract

Aim of study: Fire regimes are frequently dynamic and change as a function of the interactions between the three main fire drivers: fuels, ignitions and climatic conditions. We characterized the recent period (1974-2005) and performed estimates for the future fire regime.

Area of study: We have considered five pine and another four woodland types by means of the analyses of 100 reference areas in peninsular Spain.

Material and Methods: The estimates of the expected alterations in fire frequency and the fire rotation period were based on models previously developed for the climatic scenarios SRES A2 and B2.

Main results: The results point to the large variability in fire frequency and rotation periods between the woodland types as defined, and also among the reference areas delimited for each of them. Fire frequencies will increase for all woodland types while very relevant shortenings of the fire rotation periods are expected. For the 32 yr period analysed, rotation periods longer than 500 yr were obtained in 54% of the reference areas while this percentage would decrease to 31% in the B2 and to 29% in the A2 climatic scenario. In the most affected woodland type, P. pinaster, from a median rotation period of 83 yr it would decrease to 26 yr in the B2 and to 20 yr in the A2 climatic scenario.

Research highlights: We conclude that the predicted increases in fire activity will have adverse effects on some of the main Spanish woodland types due to the expected future disruptions in the fire regime.

Keywords: Forest fires; fire regime; fire frequency; fire rotation period; climatic change.

Abbreviations: SRES: Special Report on Emissions Scenarios; IPCC: Intergovernmental Panel on Climate Change; RA: Reference Areas.

Citation: Vázquez de la Cueva, A., Climent, J.M., Casais Calo, L., Quintana, J.R. (2015). Current and future estimates for the fire frequency and the fire rotation period in the main woodland types of peninsular Spain: a case-study approach. Forest Systems, Volume 24, Issue 2, e031, 13 pages. http://dx.doi.org/10.5424/fs/2015242-06454.

Received: 26 Jun 14. Accepted: 05 May 2015

Copyright © 2015 INIA. This is an open access article distributed under the Creative Commons Attribution License (CC by 3.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Funding: The authors do not received specific funding for this work.

Competing interests: The authors have declared that no competing interests exist.

Correspondence should be addressed to Antonio Vázquez de la Cueva: vazquez@inia.es


 

CONTENTS

Abstract

Introduction

Materials and methods

Results

Discussion

Acknowledgements

References

IntroductionTop

Wildland fires are a global ecological disturbance process with complex spatial and temporal distributions. In recent years we are acquiring a greater knowledge about its global distribution (Bond & Keeley, 2005; Krawchuk et al., 2009; Archibald et al., 2013) thanks, to a greater extent, to global remote sensing products (e.g. Riaño et al., 2007) and we are now able to propose a hypothesis about global activity patterns (Krawchuk & Moritz, 2011; Pausas & Ribeiro, 2013).

Wildland fire activity can be characterized by it fire regime, that is the type of fire history registered in a defined space and time. Fire regimes result from the interactions between climate, fuels and ignitions as their three main drivers (e.g. Davis & Burrows, 1994; Moreno et al., 1998). Fire regime is determined by a large number of factors such as the presence of ignition sources, the topography, prevention and extinction efficiency, existence of climatic conditions which favour the spread of fire and the amount, type and arrangement of fuels at several spatial scales (Rothermel, 1983). The fire regime represents a particular combination of fire characteristics such as frequency, intensity, size, season, type and extent (e.g. Trabaud, 1981; Bond & van Wilgen, 1996; Archibald et al., 2013).

Fire regimes frequently change in time and space, not only because of the fortuitousness of fire events (Moritz et al., 2012; San-Miguel-Ayanz et al., 2013) at short spatial and temporal scales, but also as a function of the interactions between the main fire drivers: fuels, ignitions and climatic conditions. This variability in drivers is exacerbated in the context of global change in which we are now (IPCC, 2007; 2014). This has been the case in southern European countries in which all the three fire drivers have been changing quickly since the middle of the past century leading to strong changes in fire regime (Le Houerou, 1987; Moreno et al., 1998; Pausas & Fernandez-Muñoz, 2012; San-Miguel-Ayanz et al., 2013; Moreno & Chuvieco, 2013).

The woodland area burned accounted for 37% (see results) of the total forest area affected by fire in peninsular Spain in the period analysed (1974-2005) while the rest were scrublands, pastures and cultivated areas. Nevertheless, the ecological relevance of ecosystems dominated by trees, its economic importance as ecosystem services suppliers and the possibility of identification of the affected woody species in the available fire reports support the analyses of the fire regime in different woodland types in this paper. Additionally, fire regimes for woodland types with a large distribution area could be very variable depending on the climatic, landscape and cultural differences within that area (e.g. Trabaud, 1994; Moreno et al., 1998; Díaz-Delgado et al., 2004; Moreira et al., 2011, Vila-Cabrera et al., 2012). In general, the spatial distribution of different types of forest over large areas is mainly conditioned by climatic gradients which control primary productivity and the structure of vegetation, but the spatial distribution of woodland types and structures in anthropogenic, or man-made, landscapes as those found in the Mediterranean basin is largely dependent on human activities and can determine also the behaviour of fire (e.g. Ruiz-Benito et al., 2012; Alvarez et al., 2012).

In addition to the availability of fuel to carry the fire and the ignition sources requirements, the climatic/weather conditions determine to a large extent fire activity. Several studies have indicated that large percentages of variance in fire activity can be explained as a function of simple climatic variables (e.g. Piñol et al., 1998; Pausas, 2004; Flannigan et al., 2009; Koutsias et al., 2013). According to the IPCC (2007; 2014) reports, the risk of fire, length of the fire season, fire frequency and severity are very likely to increase in Mediterranean type ecosystems. Climatic conditions which favour the spread of fire will become more frequent in the Mediterranean basin (Moriondo et al., 2006; Lung et al., 2013) while several studies concerning the Iberian Peninsula, point to an increment in fire risk under future climate scenarios (e.g. Moreno, 2005; Carvalho et al., 2010, Vázquez et al., 2012). This possible increase in fire activity could have important effects on forest landscapes (e.g. Schumacher & Bugmann, 2006). However, in spite of the key role of fire in forest distribution and structure there is still a lack of knowledge of current fire activity in specific woodland types and also a lack of estimates for future climatic scenarios.

In order to improve this knowledge, the strategy followed in this work was based on the delimitation of several Reference Areas (RA) or case-studies for each of the nine woodland types considered. In this way we can evaluate the current fire regime registered, and perform future estimates, for each of the species considered. With this approach we hope to provide a picture of the broad range of variability in current and also in future estimates. To quantify the expected alterations in future fire regime parameters we have used statistical models previously published (Vázquez et al., 2012). The models used were based in the relationships between fire regime parameters and monthly meteorological variables for several ecozones in which peninsular Spain was divided up for the climatic scenarios (Nakicenovic et al., 2000) SRES A2 and B2 (2071-2100).

We have three main objectives in this work. The first one is to provide a characterization of the fire frequency and the fire rotation period registered in the main woodland types defined and its spatial variability by means of the reference areas (RA) considered. The second is to explore the differences in these parameters between the main pine species and the third to perform estimates of future fire regime parameters such as the fire frequency and the fire rotation period in the RA defined for the nine woodland types. Our main hypothesis is that fire regime parameters will be very different between the woodland types considered and also among the reference areas considered for each of the woodland types. Additionally we expect a higher fire activity in some of the woodland types more used in afforestation and reforestation programs and large differences between the different pine woodlands that constitute five of the nine woodland types considered.

Material and methodsTop

Forest fire data and woodland types

The forest fire data were provided by the Spanish Forest Administration, covering the 32 yr period 1974-2005. Each fire record was referenced to the 10x10 km grid-unit based on the UTM projection in which the fire started. Reliable data were not available for the whole period for two of the 47 provinces (Álava and Navarra) in peninsular Spain, so they were excluded in the analyses. Insular provinces were also excluded. Raw (1974-1990) and elaborated (1991-2005) data were screened and checked against the published official fire statistics. Only fires with a reported size ≥ 0.1 ha of burned area are considered. When a fire burns woodlands or dense forest areas, each of the combinations of the tree species affected, the stand development or the type of property (e.g. public or private) of the area affected was recorded individually along with area burned. Based on this information we have defined the “number of cases” as the number of times each of the nine woodland types considered has been affected by fire. We have used the number of cases for each woodland type as a surrogate for computing the fire frequency in each woodland type.

The woodland (or dense forest) types considered are made up by the five main pine species (Pinus sylvestris L., P. pinea L, P. halepensis Mill., P. nigra Arn. and P. pinaster Ait.), the genera Eucalyptus spp. and Quercus spp. and two miscellaneous types grouped into “other conifers” and “other broadleaves”. In this way we have grouped the trees affected by fire into nine types with an emphasis on the main Pinus spp. growing in peninsular Spain. These will be named as woodland types, in spite of the fact that they could be dense forests, as opposed to scrublands (Table 1).

Table 1. Wooded forest area (based on the MFE50 Forest Map of Spain) and number of grid-units with ≥ 100 ha of the woodland type defined. The proportion (%) of grid-units affected by fire is calculated from grid-units with ≥ 100 ha of forest area and ≥ 100 ha of area burned for each of the woodland types. Number of cases and area burned by the fires reported for the nine woodland types considered in peninsular Spain for the period 1974-2005. The proportions (%) for the number of cases and the woodland area burned make 100%. Only fires with a reported size ≥ 0.1 are considered. The two provinces (Álava and Navarra) with incomplete data are excluded giving a total of 4864 grid-units


Reference areas (RA) and current fire regime

We have defined a total of 100 reference areas (RA) or case-studies to get an idea of the variability in the fire activity for each woodland type. The selected reference areas and the distribution of the woodland types considered are displayed in Fig. 1. The area covered by each woodland type, represented in grey tones in Fig. 1, has been derived from the Forest Map of Spain (MFE) at a 1:50.000 scale. In order to compare this information with the fire data, we have calculated for each of the 5007 10x10 grid-units used as reference for fire location in peninsular Spain, the area covered by each of the woodland types. The total size of each RA is very similar and close to 300,000 ha. The shape is elliptical and they are rotated to encompass in each of them a large area of each woodland type to which the RA is dedicated, that is each RA is built for a specific woodland type. It is important to note that in the process of selection and location of the RA we considered only the species distribution, disregarding forest fire activity or area burned. The RA are then built as case-studies of the main distribution areas of each woodland type and are located in regions with a good representation of the species of interest. More details of the RA are shown in Table 2 and 3 and Fig. 2 and shown later.

Figure 1. Maps displaying the location of the RA defined for the nine woodland types and the fire rotation period calculated for the current period (1974-2005). All the reference areas are of similar size (close to 300,000 ha) and shape but they can change in alignment. Grey tones indicate the proportion of area covered by the woodland type of interest in each 10 x10 km grid-unit: light grey between 100 and 1000 ha and dark grey more than 1000 ha.

Figure 2. Box-plots displaying the range of values in the total forest area (woodland and shrublands), the total woodland area (of the nine woodland types) and the area of the woodland type selected for which the RA is dedicated. The number of RA (reference areas or case-studies) is also displayed. The data are derived from the Forest Map of Spain.

Table 2. Characterization of the Reference Areas (RA) selected. Number of RA, wooded forest area (based on the MFE50 Forest Map of Spain), number of cases and area burned for the RA selected for each woodland type. The proportions (%) for the wooded forest area, the number of cases and the woodland area burned are calculated between the values reported in the table and the totals for peninsular Spain displayed in Table 1. Only fires with a reported size ≥ 0.1 are considered


Table 3. Mean and standard error for the elevations and a climatic characterization of the RA selected for each woodland type. The observed (Ninyerola et al., 2005) precipitation and temperature and the modelled temperatures for the control period (1961-1990) and the climatic scenarios B2 and A2 (2071-2100) derived from the PROMES Regional Climate Model (AEMET 2009) are provided


We used two fire regime parameters to assess the variability in the current fire regime in the nine types of woodlands and to estimate its future variability. The first one is fire frequency defined as the number of fire cases in 10,000 ha of wooded forest surface and year. As explained before, we used the number of cases in each woodland type as a surrogate for fire incidence. The number of years with information was assumed to be 32 in all the grid-units in spite of some missing information. The second fire regime parameter is the fire rotation period defined as the number of years needed to burn an area equivalent to the one used as reference (Heinselman, 1973). In the same way as for fire frequency the surfaces involved were only those of the woodland type for which we are calculating the fire rotation period. The fire rotation period calculated in this not spatially explicit way, did not take into account the recurrence of fires in the same places.

Future fire regime parameters estimates

We estimated the future fire incidence in each RA by applying to each one a “ratio value” according to the ecozone in which it is located (Vázquez et al., 2012). Ecozones are based on the concept of potential natural vegetation. The map of Vegetation Series for Spain (Rivas-Martínez 1987) was used to classify each 10x10 km grid-unit according to the dominant vegetation type that was them assigned to the ecozones considered. Ecozones 1 to 6 are dominated by mountain conifers (Pinus sylvestris, P. uncinata, Abies alba, Juniperus thurifera) and deciduous trees (Fagus sylvatica, Quercus robur, Q. pyrenaica, Q. faginea). Ecozones 7 to 13 are dominated by sclerophyllous species, among which Q. rotundifolia is prevalent. Q. suber dominates type 14 areas and type 15 areas are dominated by Q. coccifera. Although conifers (mainly pine species) represent a high proportion of the current forested area of Spain, on this map they are mostly considered as subdominant in other vegetation units, forming secondary forests or increasing in extension as a result of afforestation.

Ratio values are thought to represent the proportion between current (1974-2005) and future (2071-2100) fire incidence (Vázquez et al., 2012), and were based on linear regressions between monthly temperatures and fire activity variables. The ratio applied to each RA for obtaining future estimates was based on the ratio value of the ecozone in which the RA is located (Fig. 3). From the 100 RA selected we got estimates for 97 since no valid models were obtained for ecozones 1 and 2. The models and limitations were more deeply explained in Vázquez et al. (2012). We assumed that the ratios obtained for the different ecozones for the number of fires and for the wooded area burned can be applied equally to the number of cases, that is the number of times each of the woodland types has been affected by fire and to the area burned by each of them. Future scenarios were based on the anthropogenic emission estimates of the IPCC SRES (Special Report on Emissions Scenarios) A2 and B2 (Nakicenovic et al., 2000). In both SRES A2 (high emission) and B2 (low emission) scenarios Europe would undergo a warming in all seasons.

Figure 3. Ratios values estimated between current fire incidence (number of fires and wooded area burned) and the estimates for the SRES climatic scenarios B2 and A2. No valid models were obtained for ecozones 1 and 2. Data based on Vázquez et al. 2012.

ResultsTop

Forest fires in the period 1974-2005 in peninsular Spain

In the period 1974-2005 wildland fires affected to close to 2.2 x 106 ha of woodland forest area, which represented just 37% of the total area burned in this period that can be quantified in more than 5.9 x 106 ha in 343,723 fires with a reported size ≥ 0.1 ha. The five species of Pinus accounted for 34% of the wooded forest area in peninsular Spain but were implicated in 52% of the number of cases and 71% of the wooded forest area burned. The remaining tree species integrated in the other four wildland types accounted for more a greater wooded forest area (66%) and were implicated in less cases (48%) and in much less area burned (29%) (Table 1). A large proportion of the number of cases is due to P. pinaster (33.9%) followed by Eucalyptus spp. (19.4%) and Quercus spp. (14%). In relation to the woodland area burned the largest proportion is due also to P. pinaster (31.6%) but P. halepensis (23%) is in the second position. The Pinus species, except P. pinaster, have greater proportions in the area burned than in the number of cases. The opposite is observed in the remaining woodland types indicating smaller fires. The proportion of grid-units affected by fire, taking into account only the grid-units with more ≥ 100 ha of wooded forest area as displayed in Fig. 1 and also ≥ 100 ha of wooded area burned, is in general greater in Pinus. The largest value is obtained in Eucalyptus spp. (53%) followed by P. pinaster (52.2%), P. sylvestris (37%) and P. halepensis (29.7%) (Table 1).

Reference areas (RA)

The number of RA selected ranged from 7 for P. nigra to 17 for Quercus spp. These 100 reference areas (RA) encompassed 32.4% of the area covered by wooded forest in peninsular Spain ranging from 202,009 ha in P. pinea to 1,749,047 ha in Quercus spp. The RA selected accounted for 33% of the number of cases registered and 40.1% of the wooded area burned in peninsular Spain in the 32 yr period considered. These percentages were quite variable in the nine woodland types (Table 2).

The total forest area proportion (shrublands and woodlands) was high in all the RA with median values up to 200,000 ha. The median woodland area was close to 150,000 and that of the woodland type for which the reference area was defined had median values close to 50,000 ha (Fig. 2). Mean elevation of the RA ranged from less than 400 m in Eucalyptus spp. to more than 1,000 m in Pinus sylvestris. Mean precipitation ranged from close to 500 mm in P. halepensis to more than 1,100 mm in Eucalyptus spp. Mean observed temperatures ranged from 10.4 ºC in P. sylvestris to 14.4 ºC in P. pinea. Differences in temperature were small between the observed data and the data estimated from climate models but the expected increase in mean yearly temperature in the climatic scenarios B2 and A2 was clear in all woodland types (Table 3).

Current fire regime (1974-2005) in the RA

Based on the fire regime parameters calculated it was possible to distinguish three groups of woodland types (Fig. 1 and grey bars in Fig. 4a and b). The first group would be integrated by the two pines with lower fire incidence (Pinus sylvestris and P. nigra) and the three miscellaneous woodland types (“Other conifers”, Quercus spp. and “Other broadleaves”). Fire frequency was low (median values < 1 cases 10,000 ha-1 yr-1) and the calculated fire rotation period was high (median values > 478 yr). The second group would be integrated by P. pinea and P. halepensis with median fire frequencies between 1.9 and 2 and median fire rotation periods between 272 and 178 yr respectively. Finally, the third group would be integrated by P. pinaster and Eucalyptus spp. the two woodland types with the higher fire incidence. Median fire frequency was 3.1 in P. pinaster and 10.2 in the Eucalyptus spp. Median calculated fire rotation period was 83 yr in P. pinaster and 103 yr in Eucalyptus spp. The spatial variability in the values registered for the fire rotation period within the RA defined for each woodland type was high for all of them. For some woodland types (e.g. P. sylvestris in the North West, P. pinaster in the West and P. halepensis in the East) the values of the rotation period (less than 50 yr) were extremely short (Fig. 1).

Figure 4. Box-plots showing the range in the calculated values of (a) the fire frequency and (b) the fire rotation period of the wooded area in the current (1974-2005) and estimates in the future climatic scenarios (SRES A2 and B2) for the RA defined for each of the nine woodland types. The number of reference areas selected for each woodland type is displayed along with some outliers (circles) and extreme (stars) values.

Future fire regime estimates in the A2 and B2 climatic scenarios

Future fire frequency estimates indicated an increase in fire frequency in all the RA defined for the nine woodland types. As an example, in the pine species with higher frequencies (P. pinaster) median values for the RA would increase from 3.1 cases 10,000 ha-1 yr-1 to 7.3 in the B2 and 8.7 in the A2 scenario (Fig. 4a).

Values of the fire rotation period higher than 500 yr were obtained in 52 of the 97 RA. This figure would decrease to 30 in the B2 and to 28 in the A2 climatic scenarios. The future estimates indicated shortenings of the rotation periods from median values of 572 yr to 267 yr in B2 and 222 yr in A2 for P. nigra, the pine species with a larger rotation period. The shorter rotation period was obtained for P. pinaster (median value of 83 yr), and this value would still decrease to 36 yr in the B2 and 20 yr in the A2 climatic scenarios. The longer rotation periods (much longer than 500 yr) were obtained for the “other conifers”, Quercus spp. and “other broadleaves”. In these three woodland types values shorter than 500 yr are expected in the A2 scenario for Quercus spp. (median value of 421 yr) and in “other broadleaves” in the B2 and A2 with median values of 346 yr and 276 yr respectively (Fig. 4b).

In Pinus sylvestris woodlands, fire rotation periods shorter than 50 yr were registered in two RA located in north-western Spain (Fig. 1). Under the A2 scenario, one more RA located in the western Central Range was predicted to experience this short rotation period (Fig. 5). In P. pinea none of the RA displayed rotation periods shorter than 50 yr in the 32 yr considered but this fire rotation period value was estimated to occur in three out of the eight RA under the A2 scenario. In P. halepensis, the number of RA with shortest fire rotation periods (< 50 yr) would increase from three to six under the A2 scenario. In P. nigra, while none of the seven RA registered the shortest rotation period, one RA located in north-eastern Spain would experience lower than 50 yr rotation periods under the A2 scenario. In P. pinaster, values of the rotation period shorter than 50 yr were obtained in six of the 15 RA defined (Fig. 1). In the A2 scenario these values were estimated in 11 of the 15 RA. In the Eucalyptus spp. only in one of the eight RA the calculated value of the rotation period was shorter than 50 yr (Fig. 1) but this figure would increase to five RA in the A2 scenario (Fig. 5). In the “Other conifers” the calculated rotation periods were very long (longer than 1000 yr) in all the RA except in one located in north-western Spain in which the estimated rotation period would be shorter than 50 yr in the A2 scenario (Fig. 5). In all the RA defined (17) for the Quercus spp. the rotation periods obtained were longer than 250 yr (Fig. 1) while it was estimated than in seven of them the rotation period would be shorter than 250 yr in the A2 scenario. Finally, in the “Other broadleaves” type only in one of the 11 RA defined the rotation period was shorter than 250 yr while these figure were estimated in four RA in the A2 scenario (Fig. 5).

Figure 5. Maps displaying the location of the RA defined for the nine woodland types and the fire rotation period estimates for the A2 climatic scenario (2071-2100). In three of the 100 RA there are no estimates of the future fire rotation period (displayed in white). Grey tones indicate the proportion of area covered by the woodland type of concern in each 10 x 10 km grid-unit: light grey between 100 and 1,000 ha and dark grey more than 1,000 ha.

DiscussionTop

Recent fire regime in the woodland types

This work provides an evaluation of regional fire regime parameters focusing on individual forest species and it also provides estimates for future climatic scenarios. Our results confirm a generally greater fire activity in pine woodlands, large differences between the different pine species and a large variability in the fire regime among RA belonging to the same woodland type. The results confirm also a higher fire activity in some of the species more used in afforestation and reforestation programs, such as P. pinaster, P. halepensis and Eucalyptus spp. These results agree with our starting hypothesis and point to a large relevance of the region in which the RA is located in addition to the woodland type. These patterns could be due to several factors such as the strong effect of human activities (Moreno et al., 1998; Pausas et al., 2008; Moreira et al., 2011; Brotons et al., 2013) and also to gradients of forest productivity which are very relevant in fuel accumulation and therefore in fire activity (Vázquez et al., 2006) that agree also with the global fire-productivity relationship suggested recently by Pausas & Ribeiro (2013).

Our results on the current fire regime in pine woodlands show large differences between species. Pinus sylvestris and P. nigra registered a low fire incidence in most of the RA followed by P. pinea and P. halepensis with a generally medium incidence and finally a large incidence in the RA selected for P. pinaster. These three groups defined according to the fire regime registered partially agree with the strategies (or syndromes) proposed to understand the role of fire in the patterns of diversification of pines (Keeley, 2012). According to this work, Pinus sylvestris, P. pinea and P. pinaster should been considered as belonging to the fire-tolerating syndrome (able to survive low intensity surface fires due to several traits such as height coupled with the self-pruning of dead branches and a thick bark). On the other side P. halepensis would belong to the fire-embracer pine syndrome that enhance the spread of wildfires by limited self-pruning of dead branches that increase the probability of lethal crown fires (Keeley, 2012). However, other authors have already pointed out that the late flowering and absence of serotinous cones in P. sylvestris, P. nigra and P. pinea indicate that their natural forest did not evolve under conditions involving frequent crown fires while, by contrast, serotiny and early flowering in P. halepensis and P. pinaster make them able to regenerate successfully after the fire (Tapias et al., 2004). Meaningfully, these last two species -the most widely used in afforestation in Spain- are those more affected by fires (33% and 24% respectively of the wooded area burned).

Although our calculations of the fire rotation period are based only in the wooded forest area, which accounts for just 37% of the total area burned in the 32 years considered, it is worthwhile to compare some fire rotation periods obtained in other places as a reference. In Portugal, the mean fire rotation period is estimated to be 82 years (1975-2005) while calculations in 21 forest regions ranged from 31 to 762 yr. In that period a 60% of the area affected burned only once, a 23% burned twice and a 17% three or more times (Oliveira et al., 2012). Moreno et al. (1998) estimated that a fire rotation period shorter than 100 yr occurred in 22% of the grid-units in which peninsular Spain was divided during the period 1974–1994. A fire rotation period of 51 yr was reported for a study in a P. pinaster dominated area in central Spain while a spatially explicit calculation (taking into account the areas burned more than once) of this metric yielded 64 years (Vázquez & Moreno, 2001). Díaz-Delgado et al., (2004) estimated a fire rotation period of 133 yr in Catalonia, NE Spain, for the period 1975-1998.

The variability detected in the RA belonging to the same woodland type has been large. For example, in P. halepensis the median value (in the 11 RA) of the fire rotation period is 182 yr but three of them have registered values shorter than 50 yr. Some of the largest fires registered in Spain have occurred on the Mediterranean coast and many of these were caused by lightning, as happened in 1994 (Moreno et al., 1998). These fires affected to a large extent afforested or previously cultivated areas colonised by P. halepensis (e.g. Moya et al., 2007). P. halepensis forests have been the most affected by high intensity crown fires (Eugenio et al., 2006; Alvarez et al., 2012). In a comparison between natural and planted pine forest in Spain, planted pine forests were shown to occur on poorer soils, to experience higher human-caused disturbance rates (fire frequency and human-caused mortality) and to have structural characteristics different from natural pine forests (Ruiz-Benito et al., 2012). Species composition has also been demonstrated to affect deeply the probability of fire in NE Spain (González et al., 2006). Focusing on other pines (e.g. P. sylvestris and P. pinea) it is clear that there are some RA with low fire incidence and long fire rotation periods and other RA with short rotation periods. In the case of P. sylvestris these last are located in north-western and central-western Spain, out of the current main distribution area of the species while in P. pinea, short rotation periods occur in South-Central Spain. In both cases, there has been an important spread through afforestation of the surfaces covered by these species.

At a regional scale is relevant to note also the strong relationship between fire activity and the productivity of the area. This pattern has been suggested for peninsular Spain (e.g. Vazquez et al., 2006) and one possible explanation is that higher productivity is linked to more favourable site conditions for vegetation growth that could give rise to larger fuel loads and faster biomass accumulation. These patterns could partially explain the high incidence and recurrence of fires in north-western of the Iberian peninsula: the Spanish region of Galicia (e.g. Vélez, 1990; Martínez-Fernández et al., 2013) and also north of Portugal (e.g. Rego, 1991; Oliveira et al., 2012). In any case, land-uses and human activities are also determinants in fire regimes in humanized landscapes.

Fire regime estimates and forest changes

Our analyses assume a 3.7 times increase in the forest area burned in the RA defined in the A2 scenario. This figure is close to the 3.9 mean ratio provided by Vázquez et al. (2012) for the wooded area burned in the work used as a base for the analyses performed here. This 3.7 times increase is translated into shorter fire rotation periods which is in agreement with the results obtained in other approaches carried out in the Iberian Peninsula. In Portugal, Carvalho et al. (2010) reported even greater values than the ones used in this work with an increment of 478% (> 5.7 times the current level) for the SRES A2 scenario. A simulation analysis performed in three areas of NE Spain show that currently fuel-humidity limited regions could suffer a drastic shift of fire regime with an up to 8 times increase of annual burned area while the impact is predicted to be less pronounced in drier areas (Loepfe et al., 2012).

Fire regimes are dynamic and vary in response to changes in drivers such as ignition frequency, vegetation patterns, climate and human activities (e.g. Bond & van Wilgen, 1996; Moritz et al., 2012; Pausas & Fernández-Muñoz, 2012). In spite of the uncertainties associated with future climatic scenarios and the simplicity of the models, the results point to a significant increase in fire activity in all the RA defined. This would lead to a shortening of the fire rotation period that could affect negatively key processes such as regeneration (e.g. Carnicer et al., 2014). The expected effects of the increasing fire activity could lead to changes in the composition of species and also to changes in forest structure, as already postulated by other authors (Mouillot et al., 2002; Vilà-Cabrera et al., 2012). P. sylvestris, P. nigra and P. pinea frequently exhibit problems for regeneration after the fire, probably due to the lack of an aerial seed bank (Tapias et al., 2004) but other factors, including different eco-physiological needs should not be dismissed (Rodrigo et al., 2004; 2007). Thus, our results support and run in parallel with those that have predicted vegetation changes as a consequence of changing fire regimes (e.g. Rodrigo et al., 2004; Schumacher & Bugmann, 2006).

Even for the pine species with easier regeneration after fire the short recurrence intervals are a problem for the sustainability of woodland. The recruitment of P. halepensis is limited by too short fire recurrence intervals (Eugenio et al., 2006) and can be replaced by grasses or shrubs under recurrent fires (e.g. Arnan et al., 2013). Oliveira et al. (2012) indicated that the recent fire regime observed in Portugal has already led to a substantial decrease in the area of P. pinaster. The different minimum maturation time (or age needed for the first effective reproduction) between species, but also between populations within species can be a key factor when dealing with a shorter fire recurrence period (Tapias et al., 2004; Climent et al., 2008). Also, genetic and plastic differences in other traits, like cone serotiny (in the fire-embracing strategy) or bark thickness (in the fire-avoiding strategy) can be critical in order to understand future individual survival and recruitment under altered fire regimes (e.g. Rigolot, 2004)

The fire regime estimates performed are only based on the effect that temperature increase will have on the empirical relationships established between fire regime and climate. That is, we have only considered the direct effects of climatic change on fire activity. However, the overall picture is more complex and includes the effects of climatic change on vegetation composition and structure and also the effects of vegetation changes on fire regimes (Mouillot et al., 2002; Moritz et al., 2012; Pausas & Ribeiro, 2013). The combination of both effects could make fire an even more important factor in vegetation dynamics, especially in fire prone areas. In the Mediterranean basin, predicted climate change is likely to exacerbate fire prone weather conditions, but the mid- to long-term impact of climate change on fire regime is not easily predictable (Loepfe et al., 2012; Brotons et al., 2013). These authors point to several key processes such as the possible reduction of biomass accumulation rates due to climate change and the role of humans on fuel management, fire ignitions and fire extinction.

The predicted increases in fire activity, illustrated by higher fire frequencies and the shortenings of the fire rotation periods shown in this work highlight the need to place more emphasis on fuel management and species use, the only direct aspects in which man plays a role. To minimize the adverse effects of fire on forest ecosystems they should be less susceptible to fire and more resilient to the expected future disruptions in fire regimes.


AcknowledgementsTop

We thank the Spanish forest administration for allowing us access to the fire data and also the Agencia Española de Meteorología (AEMET) for the use of the PROMES Regional Climate Model. We thanks also Christopher Twigg for the English corrections.

ReferencesTop

AEMET, 2009. Generación de escenarios regionalizados de cambio climático para España. Report with contributions of: M Brunet, MJ Casado, de Castro M, Galán P, López JA, Martín JM, Pastor A, Petisco E, Ramos P, Ribalaygua J, et al. Gobierno de España. Agencia Estatal de Meteorología, Ministerio de Medio Ambiente y Medio Rural y Marino. Available at www.aemet.es.
Álvarez A, Gracia M, Vayreda J, Retana J, 2012. Patterns of fuel types and crown fire potential in Pinus halepensis forests in the Western Mediterranean Basin. For Ecol Manage 270: 282-290.
Archibald S, Lehmann CER, Gómez-Dans JL, Bradstock RA, 2013. Defining pyromes and global syndromes of fire regimes. PNAS 110(16): 6442-6447. http://dx.doi.org/10.1073/pnas.1211466110
Arnan X, Quevedo L, Rodrigo A, 2013. Forest fire occurrence increases the distribution of a scarce forest type in the Mediterranean Basin. Acta Oecologia-Int J Ecology 46: 39-47. http://dx.doi.org/10.1016/j.actao.2012.10.005
Bond WJ, Keeley JE, 2005. Fire as a global «herbivore»: the ecology and evolution of flammable ecosystems. Trends Ecol Evol 20(7): 387-394. http://dx.doi.org/10.1016/j.tree.2005.04.025
Bond WJ, van Wilgen BW, 1996. Fire and Plants. Population and Community Biology Series 14, Chapman & Hall, 263 pp.
Brotons L, Aquilue N, de Cáceres M, Fortin MJ, Fall A, 2013. How fire history, fire suppression practices and climate change affect wildfire regimes in Mediterranean Landscapes. Plos ONE 8(5): e62392. http://dx.doi.org/10.1371/journal.pone.0062392
Carnicer J, Coll C, Pons X, Ninyerola M, Vayreda J, Peñuelas J, 2014. Large-scale recruitment limitation in Mediterranean pines: the role of Quercus ilex and forest successional advance as key regional drivers. Glob Ecol Biogeog 23(3): 371–384. http://dx.doi.org/10.1111/geb.12111
Carvalho A, Flannigan M, Logan K, Gowman L, Miranda AI, Borrego C, 2010. The impact of spatial resolution on area burned and fire occurrence projections in Portugal under climate change. Clim Change 98: 177-197. http://dx.doi.org/10.1007/s10584-009-9667-2
Climent J, Prada MA, Calama R, Chambel MR, de Ron DS, Alía R, 2008. To grow or to seed: ecotypic variation in reproductive allocation and cone production by young female Aleppo pine (Pinus halepensis, Pinaceae). Am J Bot 95: 833-842. http://dx.doi.org/10.3732/ajb.2007354
Davis FW, Burrows DA, 1994. Spatial simulation of fire regime in Mediterranean-Climate landscapes. In: The role of fire in Mediterranean-type ecosystems (Moreno J.M., Oechel, W.C., eds). Ecological Studies 107, Springer- Verlag, NY. pp. 117-139.
Díaz-Delgado R, Lloret F, Pons X, 2004. Statistical analysis of fire frequency models for Catalonia (NE Spain, 1975-1998) based on fire scar maps from Landsat MSS data. Int J Wildland Fire 13: 89-99. http://dx.doi.org/10.1071/WF02051
Eugenio M, Verkaik I, Lloret F, Espelta JM, 2006. Recruitment and growth decline in Pinus halepensis populations after recurrent wildfires in Catalonia (NE Iberian Peninsula). For Ecol Manag 231(1-3): 47-54.
Flannigan MD, Krawchuk MA, De Groot WJ, Wotton BM, Gowman LM, 2009. Implications of changing climate for global wildland fire. Int J Wildland Fire 18: 483-507. http://dx.doi.org/10.1071/WF08187
González JR, Palahí M, Trasobares A, Pukkala T, 2006. A fire probability model for forest stands in Catalonia (north-east Spain). Ann For Sci 63: 169–176. http://dx.doi.org/10.1051/forest:2005109
Heinselman ML, 1973. Fire in the virgin forest of the Boundary Waters Canoe area, Minnesota. Quaternary Research 3: 329-382. http://dx.doi.org/10.1016/0033-5894(73)90003-3
IPCC, 2007. Alcamo J, Moreno JM, Nováky B, Bindi M, Corobov R, Devoy RJN, Giannakopoulos C, Martin E, Olesen JE, Shvidenko A, 2007:. Europe. Climate Change 2007: Impacts, Adaptation and Vulnerability. Contribution of Working Group II to the Fourth Assessment Report of the IPCC (Eds ML Parry, OF Canziani, JP Palutikof, PJ van der Linden, CE Hanson) pp. 541-580. Cambridge University Press: Cambridge, UK.
IPCC, 2014. Climate Change 2014: Impacts, Adaptation, and Vulnerability. Working Group II Contribution to the IPCC 5th Assessment Report. Chapter 23 Europe.
Keeley JE, 2012. Ecology and evolution of pine life histories. Ann For Sci 69:445-453. http://dx.doi.org/10.1007/s13595-012-0201-8
Koutsias N, Xanthopoulos G, Founda D, Xystrakis F, Nioti F, Pleniou M, Mallinis G, Arianoutsou M, 2013. On the relationships between forest fires and weather conditions in Greece from long-term national observations (1894–2010). Int J Wildland Fire 22: 493–507. http://dx.doi.org/10.1071/WF12003
Krawchuk MA, Moritz MA, 2011. Constraints on global fire activity vary across a resource gradient. Ecology 92(1): 121-132. http://dx.doi.org/10.1890/09-1843.1
Krawchuk ME, Moritz MA, Parisien MA, Van Dorn J, Hayhoe K, 2009. Global Pyrogeography: the Current and Future Distribution of Wildfire. Plos ONE 4(4): e5102. http://dx.doi.org/10.1371/journal.pone.0005102
Le Houerou HN, 1987. Vegetation wildfires in the mediterranean basin: evolution and trends. Ecologia Mediterranea 13(4): 13-23.
Loepfe L, Martinez-Vilalta J, Piñol J, 2012. Management alternatives to offset climate change effects on Mediterranean fire regimes in NE Spain. Clim Change 115(3-4): 693-707. http://dx.doi.org/10.1007/s10584-012-0488-3
Lung T, Lavalle C, Hiederer R, Dosio A, Bouwer LM, 2013. A multi-hazard regional level impact assessment for Europe combining indicators of climatic and non-climatic change. Glob Environ Change 23: 522–536. http://dx.doi.org/10.1016/j.gloenvcha.2012.11.009
Martínez-Fernández J, Chuvieco E, Koutsias N, 2013. Modelling long-term fire occurrence factors in Spain by accounting for local variations with geographically weighted regression. Nat Hazards Earth Syst Sci 13: 311–327. http://dx.doi.org/10.5194/nhess-13-311-2013
Moreira F, Viedma O, Arianoutsou M, Curt T, Koutsias N, Rigolot E, Barbati A, Corona P, Vaza P, Xanthopoulos G, Mouillot F, Bilgili E, 2011. Landscape–wildfire interactions in Southern Europe: implications for landscape management. J Environ Manage 92: 2389-2402. http://dx.doi.org/10.1016/j.jenvman.2011.06.028
Moreno JM (Ed.), 2005. Evaluación preliminar de los impactos en España por efecto del cambio climático, ECCE Proyect. Ministerio de Medio Ambiente, Final Report. (Madrid, Spain).
Moreno JM, Vázquez A, Vélez R, 1998. Recent history of forest fires in Spain. In: “Large Forest Fires” Moreno, JM (ed.) Backhuys Publishers, Leiden, The Netherlands, 159-185.
Moreno MV, Chuvieco E, 2013. Characterising fire regimes in Spain from fire statistics. Int J Wildland Fire 22: 296–305. http://dx.doi.org/10.1071/WF12061
Moriondo M, Good P, Durao R, Bindi M, Giannakopoulos C, Corte-Real J, 2006. Potential impact of climate change on fire risk in the Mediterranean area. Clim Resarch 31: 85-95. http://dx.doi.org/10.3354/cr031085
Moritz MA, Parisien MA, Batllori E, Krawchuk MA, Van Dorn J, Ganz DJ, Hayhoe K, 2012. Climate change and disruptions to global fire activity. Ecosphere 3(6): UNSP 49.
Mouillot F, Rambal S, Joffre R, 2002. Simulating climate change impacts on fire frequency and vegetation dynamics in a Mediterranean-type ecosystem. Glob. Change Biol. 8: 423–437. http://dx.doi.org/10.1046/j.1365-2486.2002.00494.x
Moya D, Espelta JM, Verkaik I, López-Serrano FL, de las Heras J, 2007. Tree density and site quality influence on Pinus halepensis Mill. reproductive characteristics after large fires. Ann For Sci 64: 649–656. http://dx.doi.org/10.1051/forest:2007043
Nakicenovic N, Alcamo J, Davis G, de Vries B, Fenhann J, Gaffin S, Gregory K, Grübler A, Yong Jung T, et al., 2000. Emissions Scenarios. A Special Report of Working Group III of the Intergovernmental Panel on Climate Change, Cambridge University Press, 599 pp.
Ninyerola M, Pons X, Roure JM, 2005. Atlas Climático Digital de la Península Ibérica. Metodología y aplicaciones en bioclimatología y geobotánica. Universidad Autónoma de Barcelona, Bellaterra, Spain.
Oliveira SLJ, Pereira JMC, Carreiras JMB, 2012. Fire frequency analysis in Portugal (1975-2005), using Landsat-based burnt area maps. Int J Wildland Fire 21(1): 48-60. http://dx.doi.org/10.1071/WF10131
Pausas JG, 2004. Changes in fire and climate in the eastern Iberian peninsula (Mediterranean basin). Clim Change 63: 337-350. http://dx.doi.org/10.1023/B:CLIM.0000018508.94901.9c
Pausas JG, Fernández-Muñoz S, 2012. Fire regime changes in the western Mediterranean Basin: from fuel-limited to drought-driven fire regime. Clim Change 110: 215–226. http://dx.doi.org/10.1007/s10584-011-0060-6
Pausas JG, Llovet J, Rodrigo A, Vallejo R, 2008. Are wildfires a disaster in the Mediterranean basin? A review. Int J Wildland Fire 17: 713–723. http://dx.doi.org/10.1071/WF07151
Pausas JG, Ribeiro E, 2013. The global fire–productivity relationship. Glob Ecol Biogeog 22: 728–736. http://dx.doi.org/10.1111/geb.12043
Piñol J, Terradas J, Lloret F, 1998. Climate warning, wildfire hazard and wildfire occurrence in coastal eastern Spain. Clim Change 38: 345-357. http://dx.doi.org/10.1023/A:1005316632105
Rego FC, 1991. Land use changes and wildfires. In ‘Responses of forest ecosystems to environmental changes’ (Eds. A Teller, P Mathy, JNR Jeffers). Elsevier Applied Science: London. pp. 367–373.
Riaño D, Moreno-Ruíz JA, Isidoro D, Ustin L, 2007. Global spatial patterns and temporal trends of burned area between 1981 and 2000 using NOAA-NASA Pathfinder. Glob. Change Biol 13:40–50. http://dx.doi.org/10.1111/j.1365-2486.2006.01268.x
Rigolot E, 2004. Predicting postfire mortality of Pinus halepensis Mill. and Pinus pinea L. Plant Ecology 171(1-2): 139–151. http://dx.doi.org/10.1023/B:VEGE.0000029382.59284.71
Rivas-Martínez S, 1987. Mapa de las series de vegetación de España (1:400000). Publ. Ministerio de Agricultura, Pesca y Alimentación, Instituto Nacional para la Conservación de la Naturaleza: Madrid.
Rodrigo A, Quintana V, Retana J, 2007. Fire reduces Pinus pinea distribution in the Northeastern Iberian Peninsula. Ecoscience 14(1):23-30 http://dx.doi.org/10.2980/1195-6860(2007)14[23:FRPPDI]2.0.CO;2
Rodrigo A, Retana J. Pico FX, 2004. Direct regeneration is not the only response of Mediterranean forest to large fires. Ecology 85(3): 716-729. http://dx.doi.org/10.1890/02-0492
Rothermel RC, 1983. How to predict the spread and intensity of fires? USDA, Forest Service Gen. Tech. Rep. INT-143: Ogden, UT.
Ruiz-Benito P, Gomez-Aparicio L, Zavala MA, 2012. Large-scale assessment of regeneration and diversity in Mediterranean planted pine forests along ecological gradients. Diversity and Distributions 18(11): 1092-1106. http://dx.doi.org/10.1111/j.1472-4642.2012.00901.x
San-Miguel-Ayanz J, Moreno JM, Camia A, 2013. Analysis of large fires in European Mediterranean landscapes: Lessons learned and perspectives. For Ecol Manage 294: 11–22.
Schumacher S, Bugmann H, 2006. The relative importance of climatic effects, wildfires and management for future forest landscape dynamics in the Swiss Alps. Glob Change Biol 12: 1435–1450. http://dx.doi.org/10.1111/j.1365-2486.2006.01188.x
Tapias R, Climent J, Pardos JA, Gil L, 2004. Life histories of Mediterranean pines. Plant Ecology 171:53-68. http://dx.doi.org/10.1023/B:VEGE.0000029383.72609.f0
Trabaud L, 1981. Man and tire: Impact on Mediterranean vegetation. In: ´Mediterranean-Type Shrublands´. (Eds. F. di Castri, D.W. Goodall, RL.Specht). Ecosystems of the World 11, Elsevier, Amsterdam, pp. 523-537.
Trabaud L, 1994. Wildland fire cycles and history in central southern France. In: Proceedings of the Second International Conference on Forest Fires Research, Coimbra (Portugal), November 1994, pp. 545-556.
Vázquez A, Moreno JM, 2001. Spatial distribution of forest fires in Sierra de Gredos (Central Spain). For Ecol Manage 147: 55-65.
Vázquez de la Cueva A, García del Barrio JM, Ortega M, Sánchez Palomares O, 2006. Recent fire regime in peninsular Spain in relation to forest potential productivity and population density. Int J Wildland Fire 15: 397-405. http://dx.doi.org/10.1071/WF05071
Vázquez de la Cueva A, Quintana JR, Cañellas I, 2012. Fire activity projections in the SRES A2 and B2 climatic scenarios in peninsular Spain. Int J Wildland Fire 21: 653-665. http://dx.doi.org/10.1071/WF11013
Vélez R, 1990. Algunas observaciones para una selvicultura preventiva de incendios forestales. Ecología, Fuera de Serie 1: 561–571.
Vilà-Cabrera A, Rodrigo A, Martinez-Vilalta J, Retana J, 2012. Lack of regeneration and climatic vulnerability to fire of Scots pine may induce vegetation shifts at the southern edge of its distribution. J Biogeog 39(3): 488-496. http://dx.doi.org/10.1111/j.1365-2699.2011.02615.x



Webpage: www.inia.es/Forestsystems