Integrated planning for landscape diversity enhancement, fire hazard mitigation and forest production regulation: A case study in central Portugal

Land Use Policy 61 (2017) 398–412

Contents lists available at ScienceDirect

Land Use Policy journal homepage: www.elsevier.com/locate/landusepol

Integrated planning for landscape diversity enhancement, fire hazard mitigation and forest production regulation: A case study in central Portugal Isabel Navalho a , Cristina Alegria b,c,∗ , Luís Quinta-Nova b,d , Paulo Fernandez b,d a MsC in GIS, Instituto Politécnico de Castelo Branco, Escola Superior Agrária, Quinta da Senhora de Mércules, Apartado 119 6001-909 Castelo Branco, Portugal b Instituto Politécnico de Castelo Branco, Escola Superior Agrária, Department of Natural Resources and Sustainable Development, Quinta da Senhora de Mércules, Apartado 119 6001-909, Castelo Branco, Portugal c CERNAS—Centro de Estudos de Recursos Naturais, Ambiente e Sociedade, Instituto Politécnico de Castelo Branco, Escola Superior Agrária, Castelo Branco, Portugal d GeoBioTec—Geobiociências, Geoengenharias e Geotecnologias, University of Aveiro, Aveiro, Portugal

a r t i c l e

i n f o

Article history: Received 11 May 2016 Received in revised form 22 November 2016 Accepted 23 November 2016 Keywords: Land cover change Species suitability Area control method Silvicultural prescription Landscape composition and structure

a b s t r a c t Forest fires and forest biodiversity are related issues of major concern in Mediterranean countries and require an integrated approach to landscape planning. The aim of this study was to develop a GIS approach for regulating forest production while promoting landscape diversity and mitigating fire hazard. A study area located in the centre of Portugal was chosen. The area was primarily occupied by maritime pine and had a high fire hazard, low tree species diversity and an extensive protection area. The classical area control method was used to assist in forest production regulation. Species suitability maps were produced for 21 recommended species for afforestation in the study area. Maritime pine management compartments were defined, and a 50-year harvesting plan was proposed. In each harvested compartment, protection areas were identified for species conversion (e.g., native oaks and/or broadleaves). Afforestation species were proposed according to the species suitability maps produced earlier. Low flammability species that produce high-quality wood, non-wood products and landscape enhancement were preferred. A comparison of the land cover in the study area in 2007 to that anticipated in 2064 via the proposed plan showed that a more fragmented landscape structure could be achieved by introducing 16 species of lower flammability than maritime pine into the study area. This study proved the usefulness of this methodological approach for guiding sustainable changes in homogeneous, unmanaged forest landscapes prone to fire. Further research is needed regarding integrated planning approaches that incorporate environmental, economic and social dimensions (e.g., human desertification of rural areas). © 2016 Elsevier Ltd. All rights reserved.

1. Introduction In the Mediterranean region, changes in land use patterns and the impacts of socioeconomic factors on land management practices have resulted in major modifications to forest ecosystems during the second half of the 20th century (Piussi and Farrell, 2000; Fernandes et al., 2014; Seijo et al., 2015). Many rural areas have experienced substantial population decreases, leading to the abandonment of agricultural land and a reduction in the consumption

∗ Corresponding author at: Instituto Politécnico de Castelo Branco, Escola Superior Agrária, Department of Natural Resources and Sustainable Development, Quinta da Senhora de Mércules, Apartado 119 6001-909, Castelo Branco, Portugal. E-mail addresses: [email protected] (I. Navalho), [email protected] (C. Alegria), [email protected] (L. Quinta-Nova), [email protected] (P. Fernandez). http://dx.doi.org/10.1016/j.landusepol.2016.11.035 0264-8377/© 2016 Elsevier Ltd. All rights reserved.

of forest fuels (Piussi and Farrell, 2000; Nunes et al., 2005; Jones et al., 2011). Additionally, a prominent feature of forest policies in the Mediterranean region is large-scale afforestation and subsequent reforestation with fast-growing pioneer conifers for wood production and land restoration (Jones et al., 2011; Fernandes et al., 2014). As biomass removal has decreased with afforestation, fire hazards have increased (Fernandes et al., 2014). Throughout this period, fires have destroyed many large forest areas (Piussi and Farrell, 2000; FAO, 2013), particularly in southern Mediterranean countries (e.g., Greece, Italy, Portugal and Spain). The relationships between landscape structures and fire regimes are interactive (Loepfe et al., 2010). Generally, when agricultural land is abandoned, vegetation growth inevitably occurs and leads to spontaneous forest regeneration, often transforming from a mixture of fragments of different ages and sizes to an integrated

I. Navalho et al. / Land Use Policy 61 (2017) 398–412

homogenization of the forest landscape (Geri et al., 2010). Homogenous landscapes with high fuel loads and high connectivity favour high fire intensity and considerable potential for spreading (Loepfe et al., 2010; Fernandes et al., 2014). However, fires also increase landscape homogeneity by reducing fuel loads and changing land use/cover types (e.g., forests into scrublands), thereby increasing fire propagation (Loepfe et al., 2010). Fires are selective, with small fires exhibiting stronger land cover preferences than large fires. In general, forests are usually more fire prone than agricultural areas, but less susceptible than scrubland (Nunes et al., 2005; Moreira et al., 2011; Oliveira et al., 2014). From a management perspective, land cover is the only landscape variable that influences fire behaviour and can be manipulated. In that context, the development of land use policies that conserve landscape diversity by promoting a fragmented landscape structure, namely, combining the effects of strategically placed area-wide fuel treatments, fire resistant forest types and agricultural mosaics, are recommended (Badia et al., 2002; Loepfe et al., 2010; Fernandes et al., 2014). The environmental consequences associated with land cover/land use changes due to the abandonment of agricultural areas and/or the afforestation intensification may occur as a result of gradual decreases in landscape diversity and complexity and an increase in vulnerability to certain hazards such as forest fires (Serra et al., 2008). Therefore, analyses of landscape patterns (e.g., composition and structure) are essential for forest conservation and should be a priority in forest management programmes (Teixido et al., 2010). Multiple contributions can be made to the diversity and complexity of the landscape matrix: (1) at the landscape level, by considering the spatial arrangement of different-aged plantation stands with respect to other landscape components, especially native forest remnants (Brockerhoff et al., 2008), and (2) at the stand level, by considering appropriate management choices regarding composition, structure (e.g., age structure, vertical structure, spatial heterogeneity and trees species), rotation lengths and harvesting approaches (Kerr, 1999; Brockerhoff et al., 2008). Introducing species and age diversity throughout a forest can increase its resilience to pests, diseases and fire and expand the associated economic opportunities. Replanting can also offer the opportunity to establish woodland replacement that responds to new markets and is potentially more resilient to climate change and fire. Moreover, promoting diversity in forests is essential to preserving biodiversity and expanding habitats, as well as contributing to enhanced landscape quality and recreational opportunities (Grant et al., 2012). One way of introducing age diversity into a homogeneous, unmanaged forest landscape is by regulating forest production. In that context, the area control method (Davis and Johnson, 1987) is one of the easiest approaches to ensure the progression of stand age classes. This approach requires that those areas in the oldest age class are harvested and regenerated each year, thereby becoming the areas in the youngest age class the following year. All other age classes are treated as a prescribed and annually increase one age class. The structure of the forest remains constant from year to year. The same number of hectares is cut each year, the same approximate harvest is produced each year, and harvest equals growth, which ensures the sustainability of the exploited resource (Davis and Johnson, 1987). Currently, the Portuguese landscape is predominately occupied by forests (3,154,800 ha; 34%) and scrubland and pastures (2,853,228 ha; 32%). Only 24% of the country is currently maintained as agricultural areas (2,114,278 ha). Portuguese forests are chiefly composed of planted eucalyptus forests (Eucalyptus globulus Labill.; 811,943 ha; 26%) and maritime pine forests (Pinus pinaster Aiton; 714,445 ha; 23%) (ICNF, 2013). Portuguese forest stands are differentially prone to fire, with mature forests of broadleaved deciduous and mixed forests having a lower fire hazard compared

399

to pure pine forests, eucalyptus plantations, or mixed pine and eucalyptus stands (Fernandes, 2009; Moreira et al., 2009; Silva et al., 2009; Fernandes et al., 2010). Accordingly, a more prominent role should be given to the expansion of deciduous broadleaved and mixed forests rather than pine stands or exotic species plantation forests in fire-prone landscapes. In addition, regarding afforestation policies, the creation of non-contiguous forest patches smaller than 30 ha is advisable (Fernandes et al., 2010; Moreira et al., 2011). In fact, the fragmentation of a fire-prone landscape with patches in different succession stages, the introduction of narrow corridors between wooded patches and the promotion of convoluted perimeters are effective measures to reduce the potential fire size (Moreira et al., 2011). Therefore, the definition of landscape-level management rules is important for promoting landscapes that are less fire prone where the species, stand variables (e.g., stand density and stand height) and topographic variables (e.g., slope and aspect) of contiguous patches can be used to predict fire spread and severity (Fernandes et al., 2010; Moreira et al., 2011). Although Portugal has a significant legal framework regarding forest policies (FAO, 2013), which was developed in consideration of complexly related issues such as climate change, forest health, fire and forest multifunctionality, obtaining concrete results is a challenge that requires an integrated landscape planning approach (Tortora et al., 2015). This paper presents a GIS methodology for an integrated landscape planning approach supported by existing methods to promote landscape diversity enhancement, fire hazard mitigation and forest production regulation in homogeneous unmanaged forest landscapes prone to fire. In that context, a study area that is mainly occupied by maritime pine and has a high fire hazard, low tree species diversity and an extensive protection area was chosen. The objectives of the study were as follows: (1) to investigate land cover change trends from 1990 to 2007; (2) to produce a suitability map for 21 recommended species for afforestation in the study area; (3) to determine the appropriate management compartments using the area control method to regulate maritime pine forest areas over a 50-year rotation period (both to assist in the regulation of this forest and to identify the compartments associated with the protection area); (4) to compare the landscape structure (patch number and size by species), composition (area by species) and flammability (fire hazard prediction) in 2007 and 2064; and (5) to discuss the application of these planning methods in terms of Portuguese forest policy.

2. Materials and methods 2.1. Study area The study area (3100 ha) is located in the centre of Portugal (Fig. 1a) and is primarily occupied by continuous areas of maritime pine forest. In 2003, this area was severely devastated by a wildfire (1516 ha; 49%; Fig. 1b). In 2007, the study area landscape was dominated by maritime pine forest (1817 ha; 59%; Fig. 1c), with mature stands in the south and young regenerated stands in the north. This study area has low species diversity, as 92% of the land cover is composed of maritime pine, eucalyptus and shrubs (Fig. 1c). Therefore, regarding both study area land cover and topography (based on a digital elevation model (DEM); Fig. 1d), it is not surprising that the vast majority of this area is classified as extreme to very extreme fire hazard (2868 ha; 95%; Fig. 1e) according to the official fire hazard map from 2011 (DGT, 2016). Furthermore, this area also comprises a vast protection area (1822 ha; 59%; Fig. 1f) (DR, 2015a) due to its topography and network of rivers and streams (Fig. 1g). Additionally, the mountain area to the west (Fig. 1d) is included in the Geopark of “Naturtejo da Meseta Meridional”, which has been considered a UNESCO Global Geopark since 2006.

400

I. Navalho et al. / Land Use Policy 61 (2017) 398–412

Fig. 1. Study area: a) maritime pine distribution in 2005 (Godinho-Ferreira et al., 2005); b) annual burnt areas (1990–2014); c) forest land cover in 2007 (COS07); d) topography − DEM; e) fire hazard in 2011; f) protection areas; g) rivers and streams; h) soil diagnosis characteristics; i) ecological zoning (SM − Sub-Mediterranean; SA.AM − Sub-Atlantic.Atlantic-Mediterranean; and SM − Sub-Atlantic); and j) roads.

2.2. Land cover change (1990–2007) and annual burnt areas (1990–2009) An analysis of land cover change in the study area from 1990 to 2007 was performed using land cover maps from 1990 (COS90) (DGT, 2015a) and 2007 (COS07) (DGT, 2015b). These land cover maps (shapefile format) are available at a 1:25,000 scale and have

a minimum mapping unit (MMU) of 1 ha and a five-level classification system of 238 land cover classes. Additionally, annual burnt area maps (shapefile format) from 1990 to 2009 (ICNF, 2015) were used to assess the impacts of forest fires in the area. To quantify the changes in forest landscape composition, the land cover maps COS90 and COS07 were overlaid to determine the transition matrix (i.e., a matrix that summarizes the class transitions from one land

I. Navalho et al. / Land Use Policy 61 (2017) 398–412

401

the study area (Fig. 1i): “Sub-Mediterranean” (SM − elevation less than 400 m), “Sub-Atlantic to Atlantic-Mediterranean” (SA·AM − elevation between 400–700 m) and “Sub-Atlantic” (SA − elevation between 700–1000 m). Soil diagnostic characteristics and ecological suitability maps (Fig. 2) were produced for each of the species included in the analysis considering three suitability classes: “Superior” (3), “Reference” (2) and “Inferior” (1) (Appendix A Tables A2 and A3). A reference class was defined by species monographic notes and characterized as having no limitations for species survival, growth and development. The superior class and the inferior class were considered as having fewer and more limitations, respectively, compared to the reference class (Dias et al., 2008). The final “Species suitability” map for each species was obtained by combining the maps of soil diagnostic characteristics and ecological suitability for each species. The resulting “Species suitability” map was re-classified as superior (3), reference (2) and inferior (1) by assigning the lowest classification observed either in the soil diagnostic characteristics map or ecological suitability map of that species (Fig. 2). Finally, the “Species suitability zones” map was produced by combining the 21 species suitability maps considering the superior (3) and reference (2) classes only. 2.4. Regulation of maritime pine forest production

Fig. 2. GIS workflow for species suitability maps production.

cover class to another over the 17-year analysis period). The purpose of performing the analysis of land cover change was to provide information to guide landscape planning for diversity enhancement and fire hazard mitigation. 2.3. Species suitability mapping Suitability maps were produced for 21 forest species, which were selected in consideration of the Regional Forest Management Plan guidelines (DR, 2006) and the potential vegetation progression in the study area. Species suitability areas were defined according to the methodology developed by Dias et al. (2008), which considers two parameters: i) the soil diagnostic characteristics that constrain forest species development and ii) the ecological distribution of species (Fig. 2). The soil map of the study area (shapefile format; scale 1:25,000) (DGADR, 2015) was interpreted to identify the diagnostic characteristics (Dias et al., 2008) and obtain the “Soil diagnostic characteristics” map (Fig. 1h). The characteristics considered for soil diagnosis were as follows: expandable depth, active limestone, textural discontinuity, vertical characteristics, salinity, external drainage, internal drainage, water storage, effective thickness, rocky outcrops and urban area (Appendix A Table A1). The ecological zoning map (shapefile format; scale 1:1,000.000) (Albuquerque, 1954; APA, 2007) defines the potential geographic distributions of species according to two main vectors: one related to the maritime influence (coastal to inland) versus the Mediterranean influence (north to south) and the other related to elevation (bottom land to mountain areas). This map allowed for the identification of the following three ecological zones in

Digital topographic maps at a scale of 1:25,000 (IGeoE, 1993) were used to support the vectorization of the “Rivers and streams” (Fig. 1g) and “Roads” (Fig. 1j) maps. These maps were used to support the delimitation of the “Units” map, which was produced for administrative management purposes only. Subsequently, two management compartment maps were produced: the “Compartments-South” map for the mature maritime pine forest in the southern zone and the “Compartments-North” map for the young regenerated maritime pine forest in the northern zone. The classical production regulation model of the area control method (Davis and Johnson, 1987) was used to assess the sizes of management compartments (e.g., the number of hectares harvested per year) and the number of compartments that should be established during a rotation (e.g., the number of years to final harvest). A 50-year rotation was considered, including a final clear-cut at 45 years and five more years to guarantee natural regeneration after harvesting (Alegria, 2011). To support the application of the area control method, a set of inventory data collected from previous studies was used to assess maritime pine forest productivity in the southern zone. As a result, the map “Compartments-South” was derived by dividing the mature maritime pine area in the southern zone into 50 compartments of equal productivity as follows: four compartments of 43 ha for low productivity areas, 18 compartments of 28 ha for intermediate productivity areas and 28 compartments of 19 ha for high productivity areas. The map “Compartments-North” for young maritime pine forest management in the northern zone was derived by creating 21 compartments of 27 ha (considering the entire area as an intermediate productivity class because no inventory data were available). The boundaries of these management compartments were drawn based on the road, river and stream networks; therefore, they are irregularly shaped. Moreover, the vast majority of these compartments were split into non-contiguous patches smaller than 30 ha (Fernandes et al., 2010; Moreira et al., 2011) to reduce both fuel continuity and erosion risk to some extent. Finally, compartments were organized into a temporal management schedule of 50 years to achieve a fully regulated forest within this rotation. The age of each stand within management compartments was obtained using data from the nearest inventory sample plot as a reference. However, it is important to acknowledge the need to harvest stands younger or older than 45 years until a fully regulated forest is attained.

402

I. Navalho et al. / Land Use Policy 61 (2017) 398–412

2.5. Integrated landscape planning Maritime pine areas were identified in protection areas (Fig. 1c, f) and compartments re-arranged as production areas (maritime pine) or protection areas (oak conversion). Finally, maritime pine forest areas and native oak conversion areas were evaluated over the 50-year rotation period. Afforestation areas for conversion were proposed according to the species suitability maps produced previously. Study area landscape patterns (composition and structure) were analysed by evaluating various landscape metrics. The landscape patterns (composition and structure) of the study area were analysed by evaluating various landscape metrics. These can be broadly classified into five components: i) number of patch types; ii) proportion of each patch type; iii) spatial arrangement of patches (namely, patch aggregation level); iv) patch shape; and v) contrast between neighbouring patches. Additionally, the former three are more important than the latter two in determining landscape patterns (Peng et al., 2010). Accordingly, the following three landscape metrics were used: area by land cover type, number of patches by land cover class and patch mean size by land cover class. Regarding the fire hazard, the official model (AFN, 2012; DGT, 2016) adapted from Chuvieco and Congalton (1989) can be expressed as a weight additive equation as follows: FH = 0.59LC + 0.21S + 0.09R + 0.06A + 0.05P where FH is the fire hazard, LC is the land cover, S is the slope (%), R is the road network (m and m ha−1 ), A is the aspect (◦ ) and P is the population density (inhabitants km−2 ) (Appendix A Table A4). The model was produced in 2010 and validated using the burnt areas (larger than 20 ha) observed during that summer. A good level of agreement was verified, as 87% of these burnt areas were classified as having “Extreme” or “Very extreme” fire hazards (DGT, 2016). In this model, land cover (LC) is the most important variable for assessing the fire hazard (a weight of 59%). Therefore, this study focused on assessing land cover susceptibility by analysing the species flammability classification (IGP, 2004), as proposed in the model methodology (Appendix A Table A4). Ultimately, the fire hazard was inferred based on the species flammability assuming that all the other variables (i.e., slope, road network, aspect and population density) remained constant during the analysis period (2007–2065). Additionally, the degree of fuel continuity due to the spatial arrangement of fuel across the landscape (different species compartments or same-species compartments of different ages) was explored using the evaluated landscape metrics (e.g., the number of patches by species and the sizes of patches by species). The type of fuel (e.g., fuel models adapted from Anderson (1982) for Portugal) and the degree of vertical fuel continuity were explored considering the progression of the age classes of a stand by species (e.g., compartment age and density) for insight regarding fire spread and severity. In general, tall, open stands have lower fire severity than short, close stands (Godinho-Ferreira et al., 2005; Fernandes, 2009; Fernandes et al., 2010). Finally, forest land cover in 2007 was compared to estimated forest cover in 2064 based on the landscape structure (number and size of patches by species), composition (area by species) and flammability (fire hazard prediction). 3. Results 3.1. Land cover change (1990–2007) and annual burnt areas (1990–2009) Several forest fires occurred from 1990 to 2007 (Fig. 1b). These fires had a major impact on the land cover changes observed in

2007 (Table 1). The northern zone was severely affected by wildfire in 2003, but the southern zone was still mainly occupied by mature maritime pine (i.e., maritime pine forest) (Fig. 1c). Most burnt areas of the maritime pine forest changed to maritime pine open forest (602.5 ha), indicating that the species had the ability to regenerate new maritime pine areas. However, a substantial portion of the burnt area was transformed to scrubland (494.0 ha), eucalyptus plantations (124.4 ha) or new plantations (73.9 ha). Scrubland was observed in areas that were burnt once or twice (either 1991 and 2003 or 1992 and 2003). These scrubland areas are mainly located in high elevation zones (e.g., 700–1000 m, outside the optimal zone for maritime pine) and are currently dominated by strawberry trees (Arbutus unedo L.), one of the less flammable species in Portuguese scrubland areas (flammability class 1). As a result, the forest composition (Table 1) did not significantly change and was still dominated by maritime pine forest (from 78% to 59%). However, substantial maritime pine forest fragmentation was observed (e.g., in the northern zone). A large maritime pine forest patch of 1060 ha still dominated the southern zone (Fig. 1c). Likewise, two large open maritime pine forest patches (249 ha and 120 ha) were identified in the northern zone (Fig. 1c). Eucalyptus areas and new plantation areas (e.g., eucalyptus plantations) increased significantly (from 0.9% to 6.8% and 2.7%), mainly in post-burnt areas. However, afforestation areas of eucalyptus and new plantations larger than 20 ha were not identified. Additionally, areas of native oak afforestation were practically non-existent. Therefore, the fire hazard is still a major concern in this study area due to both the forest landscape structure (e.g., spatial arrangement of fuel across the landscape, with large continuous patches of maritime pine) and composition (e.g., dominated by the maritime pine and eucalyptus, which are very flammable). 3.2. Species suitability assessment Suitability maps produced for the 21 recommended forest species were obtained from the “Species suitability zones” map which provides an integrated output by organizing species into six zones considering reference and superior suitability areas only (Fig. 3). It is clear that native oaks will thrive in both zones one and two (400–700 m). In zone three (less than 400 m), almost every oak will thrive except the Pyrenean oak. In zones four and five (700–1000 m), only the Pyrenean oak will thrive. Maritime pine will also thrive in both zones two and three. Therefore, the possibility exists of promoting pure stands of oaks, mixed stands of oaks or mixed stands of maritime pine and oaks. However, other broadleaved species are also suitable in both zones two and three (e.g., wild cherry, black walnut and sycamore maple in zone two and strawberry trees, narrow-leaved ash, common alder, black and white willows and black and white poplars in zone three). The choices for zones one, four and five are more limited. Nevertheless, numerous possibilities exist for landscape diversification and enhancement. The maps produced in this study are important tools for decision support and will subsequently be used to define species afforestation priorities when converting maritime pine areas to protection areas. 3.3. Regulation of maritime pine forest production First, the study area was organized into 54 administrative management units ranging from 24 to 91 ha (Fig. 4a) considering both the river and stream and road networks (Fig. 1g, j). Then, the mature maritime pine forest in the southern zone was organized into 50 compartments and split into 130 sub-compartments ranging from 1 to 21 ha (Fig. 4b). Similarly, the young maritime pine forest in the northern zone was organized into 21 compartments and split into 51 sub-compartments ranging from 1 to 31 ha (Fig. 4c).

I. Navalho et al. / Land Use Policy 61 (2017) 398–412

403

Fig. 3. Suitability maps for each of the 21 recommended species and species suitability zones map (Pp − Pinus pinaster Ait., Ppn − Pinus pinea L., Ec − Eucalyptus globulus Labill., Qp − Quercus pyrenaica Willd., Qs − Quercus suber L., Qfb − Quercus faginea Lam. subsp. broteroi (P. Cout.) A. Camus), Qr − Quercus rotundifolia Lam., Cs − Castanea sativa Mill., Pa − Prunus avium L., Pll − Prunus lusitanica L. subsp. lusitanica, Au − Arbutus unedo L., Fa − Fraxinus angustifolia Vahl, Ag − Alnus glutinosa (L.) Gaertn., Csp − Cupressus sempervirens L., Cl − Cupressus lusitanica Mill., Jn − Juglans nigra L., Ap − Acer pseudoplatanus L., Sa − Salix atrocinerea L., Ss– Salix salviifolia Brot., Pn − Populus nigra L., Pal − Populus alba L.).

11.6 0.4 1.1 0.0

0.4

1.1 38.0 1.2 1.2 0.0 5.3 0.2 2.4 1124.9 36.4 5.5 0.2 15.4 0.5

0.1 129.3 4.2

1.4 0.0

7.1 189.2 6.1

0.0

0.0 1.2 1.1 0.1

Legend: COS land cover classes − Artificial surfaces (1); Agricultural areas (2); Cork oak forest (31111), Oak forest (31113), Eucalyptus forest (31115), Broad-leaved forest (31117); Coniferous forest (312): Maritime pine forest (31211); Maritime pine and eucalyptus mixed forest (31321); Natural grasslands (321), Moors and heathland (322), Sclerophyllous vegetation (323), Cork oak open forest (32411), Eucalyptus open forest (32415), Maritime pine open forest (32431), Eucalyptus and coniferous open mixed forest (32455), Maritime pine and broad-leaved open mixed forest (32461), Other woodland (32471), Cuts (32481), New plantations (32482); Beaches, dunes, sands (331), Bare rocks (332), Sparsely vegetated areas (333), Burnt areas (334); Water bodies (512). Forests are classified as “Forests” when the ground cover (GC) is greater than 30% and as “Open forests” when the GC ranges from 10% to 30%.

3.4 0.1 8.1 0.3 24.4 0.8 5.6 692.4 22.3

8.3 0.3

11.1 15.7 0.5

0.2 1.2 84.3 2.7

1.7 0.5 5.5 0.2

0.5 1.8 0.0 0.4 1.1 0.1 4.5 2.9

73.9 0.0 3.3 3.0 8.1 3.5 14.9 0.9 26.2 2.2 0.6 7.7 868.7 35.5 105.3 42.9 1.0

17.7 124.4 2.9 16.0 15.9 14.5 0.4

0.1

4.5 1.4

4.3

1.2

494.0 0.8 72.1 49.1 1.0 1.8 2.9 54.7 735.2 23.7

8.1

0.2

11.2

1.6 602.5 7.9 14.9 34.6

0.1

8.3

0.0

2.7

1.1

0.2

2.7

17.2 284.5 0.0 27.0 2259.3 49.2 220.7 150.3 1.4 1.8 4.7 83.8 3100 100.0

Sum 32481 32455 32431 322

323

32411

32415 0.8 0.2 1.4 61.0

31211 31117 31115 31111

0.8 4.5 5.0 107.4

2 1

5.5 9.4

1 2 31113 31115 31211 31321 322 32431 331 332 333 334 Sum %

1990

Table 1 Land cover change transition matrix (1990–2007) in hectares.

31321

321

0.4 58.5 0.0

2007

2.5 22.8

32461

32471 0.9 1.2

32482

332

512

0.6 9.2 0.0 0.9 72.9 1.6 7.1 4.8 0.0 0.1 0.2 2.7 100.0

I. Navalho et al. / Land Use Policy 61 (2017) 398–412

%

404

A harvesting schedule was proposed for these compartments (Tables 2 and 3) to produce a fully regulated maritime pine forest over a 50-year period (e.g., sustainable maritime pine forest production). The first compartment scheduled to be clear-cut was compartment number 16 in the southern zone (Fig. 4b; Table 2). This compartment has an area of 20 ha, with 18 ha in the protection area and, thus, in need of conversion, including five ha in zone one and 13 ha in zone three. In the following year, compartment number 37 was clear-cut, and this process continued for all compartments during the period (Fig. 4d; Table 2). After 30 years, maritime pine harvesting is proposed in the southern zone (Fig. 4c; Table 3). After the 50-year rotation is completed (1st cycle: 2015–2064), the maritime pine production area will be reduced to 836 ha: 549 ha in the southern zone and 287 ha in the northern zone (Table 2 and 3). A total of 983 ha of maritime pine in the protection area, 670 ha in the southern zone and 313 ha in the northern zone, will be converted during this 1st cycle of maritime pine forest production regulation. The 2nd cycle (2065–2114) of maritime pine production regulation (836 ha) requires the application of an area control method to define new production compartments. Furthermore, the maritime pine in zones four and five should be excluded from the overall maritime pine production area. 3.4. Integrated landscape planning Considering both the “Species suitability zone” map and each “Species suitability” map (Fig. 3), it was proposed (Table 4) that the protection area of maritime pine should be converted as follows: 4% Pyrenean oak (35 ha), 24% cork oak (237 ha), 10% Portuguese oak (100 ha), 10% holm oak (100 ha), 24% chestnut (236 ha), 18% other broadleaved (181 ha), 8% umbrella pine (78 ha) and 2% other coniferous (2 ha). Thus, the goal is to alter the forest composition in the study area by 2065 so that the percentage of maritime pine forest decreases from 84% (1847 ha) to 42% (866 ha) and the areas of native oak, chestnut other broadleaves, umbrella pine and other coniferous increase to 23% (474 ha), 11% (236 ha), 9% (187 ha), 4% (78 ha) and 1% (17 ha), respectively (Table 1 and Table 4). The area of Pyrenean oak is small, as only zone four exhibits superior suitability, and it is shared with both sycamore maple and chestnut. However, chestnut was significantly promoted in zone two, as it is also of superior suitability for this species. Likewise, cork oak, Portuguese oak and holm oak were proposed for afforestation in zone three. The remaining area was divided among other species (e.g., based on reference suitability) to promote landscape diversification and enhancement. A greater emphasis was given to promoting species that produce high-quality wood (e.g., chestnut, wild cherry and black walnut) and non-wood products (e.g., cork oak, chestnut and umbrella pine). Others species were chosen for aesthetic purposes only (e.g., Portuguese laurel, sycamore maple, ash, alder, willows, poplars and cypress) (Fig. 5a). However, eucalyptus afforestation should potentially be restricted to 10% (209 ha; Table 1) and limited to suitable zones one, two and three. Current plantations in protection areas (more than 50%) should be converted in the future based on species suitability zones (Fig. 3). A comparison of landscape structure (Fig. 5b) between the maritime pine land cover in 2007 (before) and the proposed final land cover in 2064 (after) revealed an immense increase in the number of patches (61–340) and a decrease in the mean patch area (34 ha to 4 ha). The land cover flammability analysis (Table 4; Fig. 5c) clearly showed that an essential reduction in fire hazard should be expected because the land cover area by flammability class (previously in class two) decreases by half, and the remaining area is classified in lower flammability classes (e.g., five to seven). This was accomplished by landscape composition change as a result of

I. Navalho et al. / Land Use Policy 61 (2017) 398–412

405

Fig. 4. Maritime pine forest production regulation: a) administrative management units; b) management compartments for mature maritime pine forest (South); c) management compartments for young maritime pine forest (North); d) maritime pine management compartments over protection areas for species conversion; and e) compartments detail − production area (maritime pine) and protection area (species conversion) by species suitability zone.

the introduction of 16 tree species with lower flammability than maritime pine (Table 4; Fig. 5c). 4. Discussion 4.1. Trends in land cover change The analysis clearly identifies the following as the main drivers of land cover change in the study area from 1990 to 2007: (1) forest fires and (2) eucalyptus afforestation. These land cover changes were similar to those observed at the national level, as maritime

pine areas are decreasing in post-burnt areas and subsequently recovering as scrubland or being converted to eucalyptus plantations (Caetano et al., 2009; Jones et al., 2011; ICNF, 2013). This result indicates that fire hazards will continue to be of major concern in the study area due to both the forest landscape composition and structure (the spatial arrangement of fuel across the landscape made of continuous patches of very flammable species such as maritime pine and eucalyptus), as well as the vertical fuel continuity of a stand if no management strategy is applied. Therefore, methodologies that support integrated landscape planning on a medium- to long-term schedule, produce a frag-

406

I. Navalho et al. / Land Use Policy 61 (2017) 398–412

Fig. 5. Forest cover planning: a) land cover maps before and after (Pp − Pinus pinaster Ait., Ppn − Pinus pinea L., Qp − Quercus pyrenaica Willd., Qs − Quercus suber L., Qfb − Quercus faginea Lam. subsp. broteroi (P. Cout.) A. Camus), Qr − Quercus rotundifolia Lam., Cs − Castanea sativa Mill., Pa − Prunus avium L., Pll − Prunus lusitanica L. subsp. lusitanica, Au − Arbutus unedo L., Fa − Fraxinus angustifolia Vahl, Ag − Alnus glutinosa (L.) Gaertn., Csp − Cupressus sempervirens L., Cl − Cupressus lusitanica Mill., Jn − Juglans nigra L., Ap − Acer pseudoplatanus L., Sa − Salix atrocinerea L., Ss– Salix salviifolia Brot., Pn − Populus nigra L., Pal − Populus alba L.); b) land cover metrics before and after (patches mean area and number of patches); and c) land cover flammability classes (1–7) and species area before and after.

I. Navalho et al. / Land Use Policy 61 (2017) 398–412

407

Table 2 1st cycle (2015–2064): maritime pine harvesting schedule in the Southern zone and compartments re- arrangement as production area (maritime pine) and as protection area (species conversion by suitability zone). 1st cycle − harvesting schedule South

Species conversion schedule (2015–2064) by suitability zone

Order

Year

Comp S number

Area (ha) Prod Pp

Area (ha) Prod

Area (ha) Prot

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 Total

2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 2026 2027 2028 2029 2030 2031 2032 2033 2034 2035 2036 2037 2038 2039 2040 2041 2042 2043 2044 2045 2046 2047 2048 2049 2050 2051 2052 2053 2054 2055 2056 2057 2058 2059 2060 2061 2062 2063 2064

16 37 4 5 6 39 20 47 19 48 49 42 40 24 44 28 25 45 36 35 38 31 2 3 27 9 23 21 22 41 33 30 7 26 34 46 11 13 1 18 17 8 32 12 10 29 15 43 50 14

20 30 19 19 21 28 19 28 19 28 32 28 29 20 29 18 18 29 28 28 30 43 19 20 19 18 19 17 16 28 31 42 21 19 28 30 19 19 19 19 19 18 43 19 20 44 19 28 28 18 1220

2 10 18 11 13 14 6 15 12 22 20 16 10 10 17 7 14 10 9 1 16 18 15 16 10 1 2 14 2 20 1 17 6 2 1 12 8 13 1 8 15 10 15 9 11 21 10 9 21 8 549

18 20 1 8 8 14 13 13 7 6 12 12 17 10 12 11 4 19 19 27 14 24 4 4 9 17 17 3 14 8 30 25 15 17 27 18 11 6 18 11 4 8 28 10 9 23 9 19 7 10 670

Zone 1 Oaks all

Zone 2 Oaks all

Zone 3 Qp excluded

5 3 1 4 2 1 2 2

13 17

2 5 6 10 3 3 4 4 1 2 4 1

7 1

2 18 3 4 1 3 2 1 2 1 3 11 4 11 1 9 4 4

4 6 13 11 11 7 4 7 12 6

62

Zone 6 No suitability

5

2

15 18 27 10 2

1

8 14 15 2 12 27 13 11 6 26 9 7 2 2

4

10

Zone 5 Qp only

9 8

16

15

Zone 4 Qp only

4 8

3 9 10 4 1 179

7 4 4 1 10 9 8 9 3 9 416

4

2

13

3

Legend: Comp S number − compartments over the Southern zone; Prod Pp − maritime pine compartments; Prod − maritime pine production area; Prot − maritime pine area over protection area for species conversion; Zones 1–6–species suitability zones; Qp − Quercus pyrenaica Willd.

mented landscape structure and introduce fire resistant forest types (Godinho-Ferreira et al., 2005; Silva et al., 2009; Fernandes, 2009; Moreira et al., 2009; Moreira et al., 2011; Fernandes et al., 2010) are of utmost importance. 4.2. GIS methodology for integrated landscape planning The methodological approach used in this study successfully facilitated integrated landscape planning to enhance landscape diversity, mitigate the fire hazard and regulate forest production within one rotation of 50 years. The methodology used to define species suitability maps is important for decision support when selecting the best species for afforestation (e.g., fire resistant forest

types such as deciduous broadleaved) and identifying a wide range of possibilities for landscape diversification and fire mitigation. The area control method proved was easy to apply and allowed for the organization of a homogeneous, unmanaged forest area currently prone to fire into management compartments according to the progression of stand age classes (e.g., different succession stages) to reduce the associated fire hazard. The method also answers the management questions of “when”, “where” and “how much” in terms of the area of maritime pine forest that should be harvested each year. At the same time, this method considers species conversion for landscape diversification and fire hazard mitigation after each compartment is harvested. The spatial arrangement of different-aged compartments improves the landscape diversity

408

I. Navalho et al. / Land Use Policy 61 (2017) 398–412

Table 3 1st cycle (2015–2064): maritime pine harvesting schedule in the Northern zone and compartments re- arrangement as production area (maritime pine) and as protection area (species conversion by suitability zone). 1st cycle − harvesting schedule North

Species conversion schedule (2015–2064) by suitability zone

Order

Year

Comp N number

Area (ha) Prod Pp

Area (ha) Prod

Area (ha) Prot

30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 Total

2044 2045 2046 2047 2048 2049 2050 2051 2052 2053 2054 2055 2056 2057 2058 2059 2060 2061 2062 2063 2064

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21

28 29 29 27 28 29 30 29 33 28 25 31 28 28 27 27 27 29 27 33 28 600

5 12 12 4 5 16 10 15 20 16 11 13 13 28 17 8 5 7 24 28 18 287

23 17 17 23 23 13 20 14 13 12 14 18 15 0 10 19 22 22 3 5 10 313

Zone 1 Oaks all

Zone 2 Oaks all

Zone 3 Qp excluded

17 15 14 17 21 9 15 14 13 12 14 16 11

Zone 4 Qp only

Zone 5 Qp only

Zone 6 No suitability

6 2 3 6 2 4 5

2 4

10 15 18 22 3 5 10 271

4 4

13

29

Legend: Comp N number − compartments over the Northern zone; Prod Pp − maritime pine compartments; Prod − maritime pine production area; Prot − maritime pine area over protection area for species conversion; Qp − Quercus pyrenaica Willd. Table 4 1st cycle (2015–2064): maritime pine protection areas − species conversion areas (ha) by suitability zone and flammability class evaluation (2–high and 7–low). Species suitability zones Species

Flammability class

Pp Ppn Ec Qp Qs Qfb Qr Cs Pa Pll Au Fa Ag Sa/Ss Csp Cl Jn Ap Pn/Pal Total area (ha) Flammability class

2 3 4 6 6 6 6 7 7 7 7 7 7 7 3 3 7 7 7

Zone 1 Oaks all

Zone 2 Oaks all

Zone 3 Qp excluded

Zone 4 Qp only

Zone 5 Qp only

78 15

78 15

100

5

137 100 99

236 47 15

2 3 3 3

15 15 2 62 7

14 35 6 3 450 6

3 3 3 3

14 3 429 5

Total area (ha)

29 6

2 2 2 13 7

35 237 100 99 236 47 17 0 6 6 6 3 14 52 37 10 983

Legend: Pp − Pinus pinaster Ait., Ppn − Pinus pinea L., Ec − Eucalyptus globulus Labill., Qp − Quercus pyrenaica Willd., Qs − Quercus suber L., Qfb − Quercus faginea Lam. subsp. broteroi (P. Cout.) A. Camus), Qr − Quercus rotundifolia Lam., Cs − Castanea sativa Mill., Pa − Prunus avium L., Pll − Prunus lusitanica L. subsp. lusitanica, Au − Arbutus unedo L., Fa − Fraxinus angustifolia Vahl, Ag − Alnus glutinosa (L.) Gaertn., Csp − Cupressus sempervirens L., Cl − Cupressus lusitanica Mill., Jn − Juglans nigra L., Ap − Acer pseudoplatanus L., Sa − Salix atrocinerea L., Ss– Salix salviifolia Brot., Pn − Populus nigra L., Pal − Populus alba L.

and ensures sustainable wood production (Kerr, 1999; Brockerhoff et al., 2008; Grant et al., 2012). In fact, a comparison of the forest land cover in the study area in 2007 to that proposed in 2064 showed that a more fragmented landscape structure was achieved (e.g., non-contiguous forest patches with convoluted perimeters that were smaller than 30 ha and in different succession stages), where 16 species of lower flammability than maritime pine were introduced (e.g., deciduous broadleaved) (Godinho-Ferreira et al.,

2005; Silva et al., 2009; Fernandes, 2009; Moreira et al., 2009; Moreira et al., 2011; Fernandes et al., 2010). Fire hazard mitigation was achieved by a combination of effects associated with properly managing the forest type, forest age and spatial landscape patterns (i.e., patch size and forest fragmentation). Forest types such as native oaks and chestnut forests were introduced because they have much lower flammability levels than do maritime pine and eucalyptus (based on different fuel types/models and different fire behaviours) (Fernandes,

I. Navalho et al. / Land Use Policy 61 (2017) 398–412

2009; Moreira et al., 2009; Silva et al., 2009; Fernandes et al., 2010). The progression of stand age class (e.g., different succession stages) based on the area control method creates different fuel loads and vertical fuel continuities between compartments. For instance, after a compartment is harvested (e.g., clear-cutting maritime pine), a break in the fuel load is observed. Over time, due to stand establishment and growth, the vertical structure will change from short and open (in the establishment phase) to short and close and eventually to tall and close (in the harvest phase). The combination of cover type (the dominant overstorey species) and forest structure, which is defined as a combination of generic stand density (close or open) and height (low or tall), can be used in fuel models to assess the fire hazard (Fernandes, 2009). Four groups of fire hazards were identified by Fernandes (2009) for Portuguese forests based on increasing order of fire potential risk: (1) open and tall forest types, and closed and tall Quercus suber and diverse forests; (2) closed, low woodlands of deciduous oaks, Q. suber and diverse forests, closed and tall Pinus pinaster woodland and tall Eucalyptus globulus plantations; (3) open and low forest types; (4) and dense, low stands of P. pinaster, E. globulus and Acacia. Thus, the fire severity generally decreases as the species flammability decreases, the stand height increases (less fuel vertical continuity) and the stand density decreases (less fuel load) (Godinho-Ferreira et al., 2005; Silva et al., 2009; Fernandes, 2009). Hence, based on landscape planning, a variety of fuel types/models, fuel loads and vertical fuel continuities may be found in contiguous compartments across the landscape. Consequently, the fragmentation of a homogeneous, fire-prone landscape in compartments smaller than 20 ha, either for different ages of the same species or different species, will have a positive impact on fire hazard mitigation (Badia et al., 2002; Moreira et al., 2009; Loepfe et al., 2010; Fernandes et al., 2010; Moreira et al., 2011; Fernandes et al., 2014). Finally, although topographic variables (e.g., slope and aspect) have an important impact on fire spreading and severity (e.g., positively related to steep slopes and warm-facing slopes) (Fernandes et al., 2010), these variables were not explored; however, they should be considered when selecting the most appropriate species for compartmental afforestation. Both official models in use in Portugal to assess the fire hazard, i.e. the one used in this study adapted from Chuvieco and Congalton (1989) and an alternate model referenced in Parente and Pereira (2016), have an important limitation: the flammability assigned to each vegetation type considers the dominant overstorey species only. The first relies on COS land cover classes (a minimum mapping unit of 1 ha), and the latter uses the Corine Land Cover (CLC) map (a minimum mapping unit of 25 ha), which is too broad for local-scale landscape analyses, such as the one performed in this study. Therefore, the fuel complexity of a stand (i.e., vertical structures of both the overstorey and understorey species) should be considered to objectively quantify the fire hazard, as stand structure rather than cover type is the major determinant of fire vulnerability (Fernandes, 2009). Finally, to set guidelines for fire hazard mitigation at the landscape scale, it would be helpful to investigate fire behaviours for different forest types, forest structures and spatial patterns (Fernandes et al., 2010; Moreira et al., 2011).

409

5. Conclusions Increasing the species diversity is a useful planning mechanism that can achieve a range of environmental, social and economic objectives and increase the ecological resilience of forests (Kerr, 1999; Brockerhoff et al., 2008; Grant et al., 2012). This is particularly important in Portugal, where fire was one of the most important drivers of land cover change from 1995 to 2010. Specifically, fires decreased maritime pine areas, and those areas became scrubland, pastures and eucalyptus forest (Caetano et al., 2009; Jones et al., 2011; ICNF, 2013). Consequently, the Portuguese national strategy goals for forest composition by 2030 (DR, 2015b) will not be attained if efforts regarding integrated landscape planning are not made. The proposed goals are as follows: maritime pine +2–11%; umbrella pine +15–32%, other coniferous +10–56%, cork oak +1–13%, holm oak +0–5%, others oaks +10–40%, chestnut +17–42% and other broadleaves +11–22%. The eucalyptus area is proposed not to increase. The GIS methodology explored in this study proved its effectiveness and can be used to assist in decision making regarding integrated landscape planning at the regional, municipal and local scales. For example, in Portugal, it can be used to improve Regional Forest Management Plans, which are currently under revision regarding forest composition goals. Likewise, the forest composition goals at the regional scale are based on the Portuguese national strategy of increasing the area of broadleaved trees (e.g., native oaks), a goal that is not currently being met. Additionally, because these plans are developed at the regional scale, they provide little detail at the municipal scale. Therefore, it would be valuable to have a Municipal Forest Management Plan with more specific guidelines regarding the definitions of forest areas functionalities and land cover management rules for regulating sustainable changes in forest landscapes. However, serious constraints exist regarding the application of those guidelines, including the lack of rural registers in the vast majority of areas in the central and northern portions of the country and the need for managing forest abandonment areas using third party entities. Therefore, an integrated planning approach is necessary and should encompass not only the environmental and economic dimensions of the problem but also the associated social dimension (e.g., human desertification of rural areas). Acknowledgements The participation of both co-authors Luís Quinta-Nova and Paulo Fernandez was supported by FCT − Foundation for Science and Technology under Funding of RD Units Reform Plan − UID/GEO/04035/2013. The participation of the co-author Cristina Alegria was supported by CERNAS-IPCB [UID/AMB/00681/2013 funding by FCT]. The authors gratefully acknowledge the comments of the reviewers which have greatly improved the manuscript. Appendix A.

410

I. Navalho et al. / Land Use Policy 61 (2017) 398–412

Table A1 Study area soil diagnosis characteristics (Dias et al., 2008).

Legend: Portuguese soils classification by SROA available at http://www.dgadr.mamaot.pt/nota-explicativahttp://www.dgadr.mamaot.pt/nota-explicativa. Note: The correspondence between SROA soils classification and the World Reference Base (WRB), the international standard for soil classification system endorsed by the International Union of Soil Sciences, is available in E. C. Sousa, M. Madeira & F. G. Monteiro. 2004. A Base de Referência para os Solos do Mundo e a Classificac¸ão dos Solos de Portugal. Revista de Ciências Agrarias 27(1):13–23.

Table A2 Species soil diagnosis characteristics (Correia and Oliveira, 2002; Correia and Oliveira, 2003; Dias et al., 2008). Species (common name and scientific name)

Symbol

Superior (3)

Reference (2)

Inferior (1)

Maritime pine

Pinus pinaster Ait.

Pp

Expandable depth

Textural discontinuity

Umbrella pine

Pinus pinea L.

Ppn

Expandable depth

–

Eucalyptus

Eucalyptus globulus Labill.

Ec

Expandable depth Textural discontinuity

–

Rocky outcrops Unproductive areas Textural discontinuity Rocky outcrops Unproductive areas Rocky outcrops Unproductive areas

Pyrenean oak Cork oak

Quercus pyrenaica Willd. Quercus suber L.

Qp Qs

Expandable depth

Textural discontinuity

Rocky outcrops Unproductive areas

Portuguese oak

Qfb

Holm oak

Quercus faginea Lam. subsp. broteroi (P. Cout.) A. Camus Quercus rotundifolia Lam.

Qr

Expandable depth Textural discontinuity

–

Rocky outcrops Unproductive areas

Chestnut Wild cherry

Castanea sativa Mill. Prunus avium L.

Cs Pa

–

Expandable depth

Textural discontinuity Rocky outcrops Unproductive areas

Portugal laurel

Pll

Strawberry tree

Prunus lusitanica L. subsp. lusitanica Arbutus unedo L.

Au

Expandable depth

Textural discontinuity

Narrow leaved ash

Fraxinus angustifolia Vahl

Fa

–

–

Rocky outcrops Unproductive areas Expandable depth Textural discontinuity Rocky outcrops Unproductive areas

Common alder Black willow White willow Italian cypress

Alnus glutinosa (L.) Gaertn. Salix atrocinerea L. Salix salviifolia Brot. Cupressus sempervirens L.

Ag Sa Ss Csp

Expandable depth

Textural discontinuity

Cypress Black walnut Sycamore maple Black poplar White poplar

Cupressus lusitanica Mill. Juglans nigra L. Acer pseudoplatanus L. Populus nigra L. Populus alba L.

Cl Jn Ap Pn Pal

Rocky outcrops Unproductive areas

I. Navalho et al. / Land Use Policy 61 (2017) 398–412

411

Table A3 Species ecological suitability (Correia and Oliveira, 2002; Correia and Oliveira, 2003; Dias et al., 2008). Species

Symbol

Superior (3)

Reference (2)

Inferior (1)

Pinus pinaster Ait. Pinus pinea L. Eucalyptus globulus Labill. Quercus pyrenaica Willd. Quercus suber L. Quercus faginea Lam. subsp. broteroi (P. Cout.) A. Camus Quercus rotundifolia Lam. Castanea sativa Mill. Prunus avium L. Prunus lusitanica L. subsp. lusitanica Arbutus unedo L. Fraxinus angustifolia Vahl Alnus glutinosa (L.) Gaertn. Salix atrocinerea L. Salix salviifolia Brot. Cupressus sempervirens L. Cupressus lusitanica Mill. Juglans nigra L. Acer pseudoplatanus L. Populus nigra L. Populus alba L.

Pp Ppn Ec Qp Qs Qfb Qr Cs Pa Pll Au Fa Ag Sa Ss Csp Cl Jn Ap Pn Pal

SA·SM SM SM SA SM SM SM SA and SA·SM SA SA·SM SM and SA·SM SA·SM and SA SA·SM and SA SA·SM and SA SA·SM and SA – SA·SM – SA – –

SM SA·SM SA·SM SA·SM SA·SM SA·SM SA·SM – SA·SM SA SA SM SM SM SM SM SM and SA SA·SM and SA SA·SM SM, SA·SM and SA SM, SA·SM and SA

SA SA SA SM SA SA SA SM SM SM – – – – – SA·SM and SA – SM SM – –

Legend: SA − Sub-Atlantic; SA·SM − Sub-Atlantic.Atlantic-Mediterranean; SM − Sub-Mediterranean.

Table A4 Fire hazard model variables and weights and flammability classes (1–7) for forest species (COS land cover) (IGP, 2004; DGT, 2016). Variables and weights

Classes

Weight (%)

COS land cover classes

Land cover (0.59)

1 2

100 80

3

70

4

40

5 6

30 10

7

1.5

– Maritime pine forest (31211), Maritime pine open forest (32431) Umbrella pine forest (31212), Coniferous forest (31213), Umbrella pine open forest (32432), Coniferous open forest (32433) Eucalyptus forest (31115), Eucalyptus open forest (32415) – Cork oak forest (31111), Holm oak forest (31112), Oak forest (31113), Cork oak open forest (32411), Holm oak open forest (32412), Oak open forest (32413) Chesnut forest (31114), Broad-leaved forest (31117), Chesnut open forest (32414), Broad-leaved open forest (32417), Cuts (32481), Burnt areas (33411)

>40 30–40 20–30 10–20 0−10 <25 25–50 50–100 100–150 <5 5–12.5 12.5–20 20–30 30–40 40–65 65–85 >85 135◦ −225◦ 225◦ −315◦ 45◦ −135◦ 315◦ −45◦ - 1 Flat <250

100 66.67 22.38 11.43 3.81 100 46.32 20.58 9.55 50 23.52 10.29 5.14 5.14 10.29 23.52 50 100 57.45 21.28 6.38 0 100

250–1500 >1500

21.05 100

Slope (%) (0.21)

Roads network (0.09)

Distance to roads network (m)

Roads network density (m ha−1 )

Aspect (◦ ) (0.06)

Population density (inhabitants km−2 ) (0.05)

412

I. Navalho et al. / Land Use Policy 61 (2017) 398–412

References AFN, 2012. Plano Municipal de Defesa da Floresta Contra Incêndios. Guia Técnico Nacional. Autoridade Florestal Nacional (Accessed 03.03.20165) http://www. icnf.pt/portal/florestas/dfci/Resource/doc/guia-tec-pmdfci-abril12. APA, 2012. Atlas do Ambiente Digital. Agência Portuguesa do Ambiente. Ministério do Ambiente, do Ordenamento do Território e do Desenvolvimento Regional (Accessed 07.07.2015) http://www.apambiente.pt/index. php?ref=19&subref=174. Albuquerque, J., 1954. Carta ecológica de Portugal. Ministério da Economia, Direcc¸ão Geral dos Servic¸os Agrícolas, Repartic¸ão de Estudos, Informac¸ão e Propaganda, Lisboa. Alegria, C., 2011. Simulation of silvicultural scenarios and economic efficiency for maritime pine (Pinus pinaster Aiton) wood-oriented management in centre inland of Portugal. For. Syst. 20, 361–378, http://dx.doi.org/10.5424/fs/ 20112003-11070. Anderson, H.E., 1982. Aids to Determining Fuel Models for Estimating Fire Behaviour. U.S. Forest Service, Ogden UT. Badia, A., Saurí, D., Cerdan, R., Llurdés, J.-C., 2002. Causality and management of forest fires in Mediterranean environments: an example from Catalonia. Environ. Hazards 4, 23–32, http://dx.doi.org/10.3763/ehaz.2002.0403. Brockerhoff, E.G., Jactel, H., Parrotta, J. a., Quine, C.P., Sayer, J., 2008. Plantation forests and biodiversity: oxymoron or opportunity? Biodivers. Conserv. 17, 925–951, http://dx.doi.org/10.1007/s10531-008-9380-x. Caetano, M., Nunes, V., Nunes, A., 2009. CORINE Land Cover 2006 for Continental Portugal, Relatório técnico. Instituto Geográfico Português, Lisboa. Chuvieco, E., Congalton, R.G., 1989. Application of remote sensing and geographic information systems to forest fire hazard mapping. Remote Sens. Environ. 29, 147–159, http://dx.doi.org/10.1016/0034-4257(89)90023-0. Correia, A.V., Oliveira, A.C., 2002. Principal Espécies Florestais Com Interesse Para Portugal − Zonas De Influência Mediterrânica. Estudos E Informac¸ão n◦ 318. Direc¸ão geral de Florestas, Lisboa. Correia, A.V., Oliveira, A.V., 2003. Principal Espécies Florestais Com Interesse Para Portugal − Zonas De Influência Atlântica. Estudos E Informac¸ão n◦ 322. Direc¸ão Geral de Florestas, Lisboa. DGADR, 2015. Carta de solos de Portugal. Direc¸ão-Geral de Agricultura e Desenvolvimento Rural (Accessed 22.05.2015) http://www.dgadr.mamaot.pt/ cartografia/cartas-solos-cap-uso-analogico. DGT, 2015a. Carta de ocupac¸ão do solo 1990 para Portugal. http://www. dgterritorio.pt/cartografia e geodesia/cartografia/cartografia tematica/carta de ocupacao do solo cos /cos 90/, (Accessed 09.03.2015). DGT, 2015b. Carta de ocupac¸ão do solo 2007 nível 2 para Portugal. http://www. dgterritorio.pt/cartografia e geodesia/cartografia/cartografia tematica/carta de ocupacao do solo cos /cos 2007/, (Accessed 09.03.2015). DGT, 2016. Cartografia de risco de incêndio florestal. http://www.dgterritorio.pt/ cartografia e geodesia/cartografia/cartografia tematica/cartas de risco/, (Accessed 08.03.2016). DR, 2006. Plano regional de ordenamento florestal da Pinhal Interior Sul. Decreto Regulamentar n◦ 8/2006. Diário da República, I Série − n◦ 138 de 19 de julho. http://www.icnf.pt/portal/florestas/profs/prof-do-pinhal-interior-sul, (Accessed 07.07.2015). DR, 2015. Aprovac¸ão da 1. a revisão do Plano Diretor Municipal de Oleiros. Aviso n.◦ 11679/2015. Diário da República, II Série − n◦ 200 de 13 de outubro. http:// www.cm-oleiros.pt/conteudos/366/383/publicacao-pdm-oleiros-em-diarioda-republica, (Accessed 20.01.2016). DR, 2015b. Aprovac¸ão da atualizac¸ão da Estratégia Nacional para as Florestas (ENF). Resoluc¸ão do Conselho de Ministros n◦ 6-B/2015. Diário da República n◦ 24/2015, 1◦ Suplemento, I Série de 2015-02-04. http://www.icnf.pt/portal/icnf/ docref/enf, (Accessed 20.07.2016). Davis, L., Johnson, K., 1987. Forest Management. McGraw-Hill, New York. Dias, S.S., Ferreira, A.G., Gonc¸alves, A.C., 2008. Definic¸ão de zonas de aptidão para espécies florestais com base em características edafo-climáticas. Silva Lusit. 16, 17–35. FAO, 2013. The State of Mediterranean Forests 2013 (accessed 09.03.2015) www.fao.org/docrep/017/i3226e/i3226e.pdf. Fernandes, P.M., Luz, A., Loureiro, C., 2010. Changes in wildfire severity from maritime pine woodland to contiguous forest types in the mountains of northwestern Portugal. For. Ecol. Manage. 260, 883–892, http://dx.doi.org/10. 1016/j.foreco.2010.06.008. Fernandes, P.M., Loureiro, C., Guiomar, N., Pezzatti, G.B., Manso, F.T., Lopes, L., 2014. The dynamics and drivers of fuel and fire in the Portuguese public forest. J. Environ. Manage. 146, 373–382, http://dx.doi.org/10.1016/j.jenvman.2014. 07.049. Fernandes, P., 2009. Combining forest structure data and fuel modelling to classify h hazard in Portugal. Ann. For. Sci. 66 (4), 415, http://dx.doi.org/10.1051/ forest/2009013. Geri, F., Rocchini, D., Chiarucci, A., 2010. Landscape metrics and topographical determinants of large-scale forest dynamics in a Mediterranean landscape. Landsc. Urban Plan. 95, 46–53, http://dx.doi.org/10.1016/j.landurbplan.2009. 12.001.

Geri, F., Rocchini, D., Chiarucci, Godinho-Ferreira, P., Azevedo, A., Rego, F., 2005. Forest typology map for Portugal (in portuguese). Silva Lusit. 13 (1), 1–34. Grant, A., Worrel, R., Wilson, S., Ray, D., Mason, B., 2012. Achieving Diversity in Scotland’s Forest Landscapes. Forestry Commission Scotland Practice Guide. Forestry Commission, Edinburg. ICNF, 2013. IFN6–Áreas Dos Usos Do Solo E Das Espécies Florestais De Portugal Continental. Resultados Preliminares. Instituto da Conservac¸ão da Natureza e das Florestas, Lisboa (accessed 09.03.2015) http://www.icnf.pt/portal/ florestas/ifn/ifn6#dad. ICNF, 2015. Cartografia Nacional De áreas Ardidas (formato “shapefile”): 1990–1999; 2000–2009; 2010; 2011; 2012; E 2013. Instituto da Conservac¸ão da Natureza e Florestas, Lisboa (accessed 28.07.2015) http://www.icnf.pt/ portal/florestas/dfci/incendios-florestais/info-geo. IGP, 2004. Cartografia De Risco De Incêndio Florestal. Relatório Do Distrito De Viseu. Instituto Geográfico Português (accessed 18.07.2016) http://scrif.igeo. pt/cartografiacrif/riscoviseu/Viseu.html. IGeoE, 1993. Carta militar de Portugal 1:25000 − Continente. Instituto Geográfico do Exército. https://www.igeoe.pt/index. php?id=186&p=1&distrito=5&escala=&extracto=278, (accessed 28.07.2015). Jones, N., de Graaff, J., Rodrigo, I., Duarte, F., 2011. Historical review of land use changes in Portugal (before and after EU integration in 1986) and their implications for land degradation and conservation, with a focus on Centro and Alentejo regions. Appl. Geogr. 31, 1036–1048, http://dx.doi.org/10.1016/j. apgeog.2011.01.024. Kerr, G., 1999. The use of silvicultural systems to enhance the biological diversity of plantation forests in Britain. Forestry 72, 191–205, http://dx.doi.org/10. 1093/forestry/72.3.191. ˜ J., Lloret, F., 2010. Feedbacks Loepfe, L., Martinez-Vilalta, J., Oliveres, J., Pinol, between fuel reduction and landscape homogenisation determine fire regimes in three Mediterranean areas. For. Ecol. Manage. 259, 2366–2374, http://dx. doi.org/10.1016/j.foreco.2010.03.009. Moreira, F., Vaz, P., Catry, F., Silva, J.S., 2009. Regional variations in wildfire susceptibility of land-cover types in Portugal: implications for landscape management to minimize fire hazard. Int. J. Wildl. Fire 18, 563–574, http://dx. doi.org/10.1071/WF07098. Moreira, F., Viedma, O., Arianoutsou, M., Curt, T., Koutsias, N., Rigolot, E., Barbati, A., Corona, P., Vaz, 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. Nunes, M.C.S., Vasconcelos, M.J., Pereira, J.M.C., Dasgupta, N., Alldredge, R.J., Rego, F.C., 2005. Land cover type and fire in Portugal: do fires burn land cover selectively? Landsc. Ecol. 20, 661–673, http://dx.doi.org/10.1007/s10980-0050070-8. Oliveira, S., Moreira, F., Boca, R., San-Miguel-Ayanz, J., Pereira, J.M.C., 2014. Assessment of fire selectivity in relation to land cover and topography: a comparison between Southern European countries. Int. J. Wildl. Fire 23, 620–630, http://dx.doi.org/10.1071/WF12053. Parente, J., Pereira, M.G., 2016. Structural fire risk: the case of Portugal. Sci. Total Environ. 573, 883–893, http://dx.doi.org/10.1016/j.scitotenv.2016.08.164. Peng, J., Wang, Y., Zhang, Y., Wu, J., Li, W., Li, Y., 2010. Evaluating the effectiveness of landscape metrics in quantifying spatial patterns. Ecol. Indic. 10 (2), 217–223, http://dx.doi.org/10.1016/j.ecolind.2009.04.017. Piussi, P., Farrell, E.P., 2000. Interactions between society and forest ecosystems: challenges for the near future. For. Ecol. Manage. 132, 21–28, http://dx.doi.org/ 10.1016/S0378-1127(00)00376-5. Seijo, F., Millington, J.D.A., Gray, R., Sanz, V., Lozano, J., García-Serrano, F., Sangüesa-Barreda, G., Camarero, J.J., 2015. Forgetting fire: traditional fire knowledge in two chestnut forest ecosystems of the Iberian Peninsula and its implications for European fire management policy. Land Use Policy 47, 130–144, http://dx.doi.org/10.1016/j.landusepol.2015.03.006. Serra, P., Pons, X., Saurí, D., 2008. Land-cover and land-use change in a Mediterranean landscape: a spatial analysis of driving forces integrating biophysical and human factors. Appl. Geogr. 28, 189–209, http://dx.doi.org/10. 1016/j.apgeog.2008.02.001. Silva, J., Moreira, F., Vaz, P., Catry, F., Ferreira, P., 2009. Assessing the relative fire proneness of different forest types in Portugal. Plant Biosyst. 143, 597–608, http://dx.doi.org/10.1080/11263500903233250. ˜ F., Gutiérrez, D., 2010. Impacts of changes in Teixido, A.L., Quintanilla, L.G., Carreno, land use and fragmentation patterns on Atlantic coastal forests in northern Spain. J. Environ. Manage. 91, 879–886, http://dx.doi.org/10.1016/j.jenvman. 2009.11.004. Tortora, A., Statuto, D., Picuno, P., 2015. Rural landscape planning through spatial modelling and image processing of historical maps. Land Use Policy 42, 71–82, http://dx.doi.org/10.1016/j.landusepol.2014.06.027.