Using models and connectivity analysis to the predict the current and future patterns of invasion by Silver-wattle (Acacia dealbata Link) in the Northwest Iberian Peninsula

Using models and connectivity analysis to predict the current and future patterns of invasion by Silverwattle (Acacia dealbata Link) in Northwest Iberian Peninsula Alice Lucas Roxo Mestrado em Ecologia, Ambiente e Território Departamento de Biologia 2013 Orientador Professor Doutor João Honrado Professor Auxiliar na Faculdade de Ciências da Universidade do Porto e Investigador no Centro de Investigação em Biodiversidade e Recursos Genéticos CIBIO -UP Faculdade de Ciências da Universidade do Porto Coorientador Doutora Joana Vicente Investigadora no Centro de Investigação em Biodiversidade e Recursos Genéticos CIBIO –UP Faculdade de Ciências da Universidade do Porto FCUP Using models and connectivity analysis to predict the current and future patterns of invasion by Silver-wattle (Acacia dealbata Link) in Northwest Iberian Peninsula 2 Acknowledgments First of all, I would like to thank Professor João Honrado for the opportunity, the support and encouragement he gave me throughout this year. I also want to thank Joana Vicente, for all the availability, support, guidance and encouragement provided this past year. To Rui Fernandes, for all the help, availability, support, companionship and friendship throughout this project. To all the people from the research group of Biodiversity and Conservation Ecology, specially Angela Lomba, Paulo Alves, Bruno Marcos, João Gonçalves, António Monteiro, Sofia Vaz and Ana Rita Silva, for the way they received me and for all the assistance they provided. To Frederico Santarém and Veronica Onofre, for the latter 6yr friendship, for being the best friends you can have, and always being there for me, for all the follies and because without them my journey through this master would not be the same thing. To all my friends, especially Ismael Macaia, Porfirio Costa, Diogo Silvério, Ricardo Felicio, André Gonçalves, Joana Rocha, Margarida Henrique, João Moura, João Santos, Sara Moutinho and all Biofocas for all the memorable moments and all the laughs. To Miguel Friães, for all the support, help, dedication and company this past year, for being present in the worst days, and for making my days better. Finally, but most importantly to my family, especially my parents, Ana and Paulo Sobral, and my brother Tomas, for all they have done for me, for all the support, dedication and care, because without them I wouldn’t be who I am today and to be, above all, the best friends you can possibly have. FCUP Using models and connectivity analysis to predict the current and future patterns of invasion by Silver-wattle (Acacia dealbata Link) in Northwest Iberian Peninsula 3 This reserach was developed at FCUP within project “Biodiversidad Vegetal Amenazada Galicia-Norte de Portugal. Conocer, gestionar e implicar” (0479_BIODIV_GNP_1_E), funded by "2º Eixo prioritário – Cooperação e Gestão Conjunta e Meio Ambiente, Património e Gestão de Riscos" of the "Programa Operacional de Cooperação Transfronteiriça Espanha – Portugal (POCTEP) 20072013". FCUP Using models and connectivity analysis to predict the current and future patterns of invasion by Silver-wattle (Acacia dealbata Link) in Northwest Iberian Peninsula 4 Abstract Biological invasions are one of the major promoters of biodiversity loss worldwide. Invasions by alien species are thought enhanced by changes in climate and disturbance regimes, as well as by other environmental shifts. Preserving native biodiversity and ecosystems from invasion by alien species requires comprehensive studies and measures to anticipate impacts, and to protect species and habitats of high conservation value. This can be achieved using species distribution models (SDM) to predict the distribution of invasive alien species (IAS) in a region of interest, both for current conditions and under scenarios of future environmental changes. In this research, we modeled the distribution of Acacia dealbata Link (an aggressive IAS) in the cross-border context of Galicia and North of Portugal, in order to predict and explain the current distribution of Acacia dealbata and to forecast possible impacts of future climate change scenarios on that distribution. We applied a combined predictive modeling (CPM) approach, where we fitted species distribution models using subsets of predictors previously classified as acting at regional or local scales, that were spatially combined to obtain the final projections. These combined models predict a wider variety of potential species responses, providing more informative projections of species distributions and dynamics than traditional models. This approach was complemented by connectivity analyses of the distribution of the invader, to evaluate current and future conflicts between this IAS and high conservation value areas. Model projections suggest that the distribution of Acacia dealbata in the study area will increase in the future, if climate changes scenarios are confirmed, and that these changes in environment will increase the connectivity of the species distribution in the area. Therefore, using CMPs and connectivity analysis it is possible to support resource prioritization for anticipation and monitoring of IAS impacts. CMPs can support measures to prevent invasions in this region as well as cross-border management in which both borders will work together and focus their efforts to protect and conserve specific important habitats. Key words: Acacia; Climate change; Protected areas; Combined predictive modelling; Species distribution models. FCUP Using models and connectivity analysis to predict the current and future patterns of invasion by Silver-wattle (Acacia dealbata Link) in Northwest Iberian Peninsula 5 Resumo As invasões biológicas são um dos principais promotores da perda de biodiversidade a nível global. Considera-se que a invasão por parte de espécies exóticas pode ser estimulada por alterações climáticas e dos regimes de perturbação, bem como por outras mudanças ambientais. Preservar a biodiversidade nativa e os ecossistemas das invasões por espécies exóticas requer estudos exaustivos e medidas para antecipar os impactos das mesmas e proteger as espécies e habitats de elevado valor de conservação. A utilização de modelos de distribuição de espécies (SDMs) permite prever a distribuição de espécies exóticas invasoras (EEI) numa certa região, tanto para as condições atuais como para cenários futuros de mudanças ambientais. Neste trabalho, modelámos a distribuição de Acacia dealbata Link (uma invasora agressiva) num contexto transfronteiriço (Galiza e Norte de Portugal), de forma a explicar a atual distribuição da Acacia dealbata e a antecipar os possíveis efeitos das alterações climáticas nesta região. Desenvolvemos um modelo preditivo combinado (MPC), em que os modelos de distribuição foram ajustados usando subconjuntos de variáveis anteriormente classificadas como importantes à escala regional ou à escala local, tendo os dois submodelos sido espacialmente combinados de forma a obter as projeções finais. Estes modelos combinados fornecem uma maior variedade de potenciais respostas das espécies, proporcionando projeções da distribuição e dinâmica de espécies mais informativas que os modelos. Esta abordagem foi complementada com análises de conectividade da distribuições da espécie invasora, para avaliar os conflitos atuais e futuros entre a sua distribuição e áreas de elevado valor de conservação. Os resultados mostram que a distribuição de Acacia dealbata na área de estudo irá aumentar no futuro, se as previsões de alterações climáticas se confirmarem, e que estas alterações no ambiente irão aumentar a conectividade da distribuição da espécie na área. Desta forma, através da utilização de MPCs e análises de conectividade identificámos as áreas com maior probabilidade de serem afetadas por esta invasora, com particular destaque para a rede regional de áreas protegidas, e é possível apoiar a priorização de recursos para antecipação e monitorização dos impactos das EEIs. Esta abordagem pode apoiar medidas de prevenção de invasões nesta região e pode também auxiliar a gestão transfronteiriça das invasões biológicas e de outras ameaças atuais e futuras à integridade ecológica dos habitats e à conservação da sua biodiversidade. FCUP Using models and connectivity analysis to predict the current and future patterns of invasion by Silver-wattle (Acacia dealbata Link) in Northwest Iberian Peninsula 6 Palavras – chave: Acacia; Alterações climáticas; Áreas protegidas; Modelos preditivos combinados; Modelos de distribuição de espécies. FCUP Using models and connectivity analysis to predict the current and future patterns of invasion by Silver-wattle (Acacia dealbata Link) in Northwest Iberian Peninsula 13 1. Introduction 1.1. Climatic change and threats to biodiversity The Earth's biosphere is facing an increasing degradation and loss of diversity due to human activities and other anthropogenic pressure on the global environment (Vitousek et al., 1977). To understand these changes it’s important to be aware of the concept of biodiversity. Biodiversity is the variability among living organisms from all sources and the ecological complexes of which they are part, which includes diversity within species, between species as well as of ecosystems (Millenium Ecosystem Assessment, 2005). Biodiversity is fundamental to ecosystem structure and functioning, and it supports the broad spectrum of goods and services that humans derive from natural systems (Staudinger et al., 2012). Human alteration of the global environment has induced the sixth extinction event in history of life (Dirzo and Raven, 2003; Chapin et al., 2000) promoting changes in the global distribution of organisms (Chapin et al., 2000). Modifications in species distribution can change both ecosystem processes, and the resilience of ecosystems to environmental changes (Staudinger et al., 2012, Chapin et al., 2000), with thoughtful consequences in ecosystem services (Chapin et al., 2000). The major causes of biodiversity decline are: land use changes, invasive species, pollution, climate change and overexploitation resulting from economic, sociopolitical, cultural, demographic, technological, and other indirect drivers (Millennium Ecosystem Assessment, 2005). 1.2. Biological invasions as a driver of biodiversity change Biological invasions can be defined as the introduction and spread of exotic organisms into regions outside of their native range (Méndez et al., 2011), and they are among the most important human-driven agents of change in natural environments worldwide (Narščius et al., 2012). The rate of biological invasions is quickly increasing last years (Essl et al., 2011) and invasive species are causing negative effects on both biodiversity and human well-being (Essl et al., 2011). Impacts of invader species include: modifications in composition, structure and functioning of the ecosystems (Terereai et al., 2013), habitat changes by elimination and/or reduction of native species (Narščius et al., 2012), biotic homogenization, loss of biodiversity and loss of ecosystem services in many parts of the world (Castro-Díez et al., 2011). Plants invasions have become recognized as a major threat to both natural and human-exploited ecosystems worldwide (Hobbs and Humphries, 1995), as they affect FCUP Using models and connectivity analysis to predict the current and future patterns of invasion by Silver-wattle (Acacia dealbata Link) in Northwest Iberian Peninsula 14 ecosystem structure and function (Lake and Leishman, 2004). Human activities helped (intentionally and accidentally) that many plant species grow outside their native ranges, some of these non-native plants overcome several barriers, becoming invasive (Bradley et al. 2010). To a successful invasion many different mechanisms are involved (Gulezian and Nyberg, 2010) and many different stages/ecological processes need to be overcome at different spatial scales (Brown et al., 2008; Richardson et al., 2000). The first barrier corresponds to a major geographical barrier that the plant overcomes through human help (Figure 1, Barrier A); after, the species is called casual. The second barrier corresponds to an environmental filter, involving biotic and abiotic factors (Figure 1; Barrier B). Then, the species needs to overcome the reproduction barrier (Figure 1; Barrier C). After overcoming these three first barriers, a taxon is considered successfully naturalized. A species or population can be called invasive if overcomes the local/regional dispersal barriers (Figure 1; Barrier D), the environmental barrier(s) in human modified or alien-dominated vegetation (Figure 1; Barrier E), and the environmental barriers in natural or semi-natural vegetations (Figure 1; Barrier E; Richardson et al., 2000). Moreover, the extent of invasion by alien species is a function of ecosystem-level properties (invasibility), propagule pressure, characteristics of the invasive species (invasiveness), and the properties of the individual native species in the ecosystem (Westphal et al., 2008). FCUP Using models and connectivity analysis to predict the current and future patterns of invasion by Silver-wattle (Acacia dealbata Link) in Northwest Iberian Peninsula 15 Figure 1 - Stages of invasion process and their representative barriers (from Richardson et al., 2000). Alien invaders are often “passengers” privileged (Gaertner and Richardson, 2011) by other environmental changes (Gaertner and Richardson, 2011), such as rising temperature, increased atmospheric carbon dioxide (CO2), extreme events in precipitation, nitrogen (N) deposition, and disturbances associated with changes in land use or land cover (Bradley et al., 2009). For example, changes in climatic conditions might directly influence the likelihood of alien species in a territory, and also affect the naturalization success (Walther et al., 2009). Synergic effects between alien invaders and global changes difficult the individualization between both the effects of alien species on the native ecosystems, and the effects of the disturbance that lead to the initial plant invasion (Gaertner and Richardson, 2011). 1.3. The specific case of invasive acacias The invasive alien plants belonging to genus Acacia are among the most common invaders in many parts of the world (González-Muñoz et al., 2012), affecting ecosystems structure and functioning worldwide (Lorenzo et al., 2010), and are already present in sensitive habitats in Europe (coastal dunes, river courses, natural parks and biosphere reserves; Lorenzo et al., 2010). Therefore invasions by Acacia genus Barriers: A B C DE F 1. Geographic 2. Environmental (local) 3. Reproduction 4. Dispersal 5. Environmental (Disturbed habitats) 6. Environmental (Natural habitats) Alien NaturalizedCasual Invasive FCUP Using models and connectivity analysis to predict the current and future patterns of invasion by Silver-wattle (Acacia dealbata Link) in Northwest Iberian Peninsula 16 (wattles) represents a threat to natural habitats through competition and replacement of native species, homogenization of the community, and loss of native biodiversity (Fuentes-Ramírez et al., 2011). Some characteristics of this genus promote the threat to native ecosystems and species, particularly the high colonization capacity, e.g. in places disturbed by fire, harvesting and other types of human disturbance (FuentesRamírez et al., 2011). Some key traits are present in this genus promoting the dominance in competitive interactions with native species as: rapid growth rates and ability to out-compete native plants; capacity to accumulate high volumes of biomass; large, persistent seed banks, and the capacity to fix nitrogen (Le Maitre et al., 2011). There are at least eight wattle species naturalized and potential invasive plants in the south of Europe (Lorenzo et al., 2010): A. dealbata, A. melanoxylon, A. retinodes, A. saligna, A. karroo, A. pycnantha, A. mearnsii, and A. longifolia. Three of the Acacia species: A. dealbata; A. melanoxylon, and A. longifolia are the most problematic invaders in France, Portugal, Spain and Italy, particularly in conservation areas, being A. dealbata the most widespread species (Lorenzo et al., 2010). 1.4. Protected areas According to International Union Conservation of Nature (IUCN, 2008), a protected area is “a clearly defined geographical space, recognized, and managed, through legal or other effective means, to achieve the long term conservation of nature with associated ecosystem services and cultural values” (http://www.iucn.org/). Protected areas comprises more than 12.7% of the planet’s land surface (Geldmann et al., 2013) being an important tool and strategy to maintain habitat integrity and species diversity, ensuring the maintenance of natural processes across landscapes and ecosystems (http://www.cbd.int/protected/overview/). The designation and maintenance of protected areas is the most important and efficient conservation strategy worldwide (Kleinbauer et al., 2010), nonetheless the importance of nature reserves varies depending the regions of the world, related to the degree of anthropic transformation of the territory (Pysek et al., 2002). Although protected areas remain a crucial key to global diversity conservation strategies, they remain susceptible to anthropogenic changes and consequences (Spear et al., 2013), especially to alien plant invasion. Alien invader species are very problematic for conservation value areas due to the potentially threat to the persistence of native species (Uddin et al., 2013). The degree of a reserve invasion can be related in some cases with the total number of visitors FCUP Using models and connectivity analysis to predict the current and future patterns of invasion by Silver-wattle (Acacia dealbata Link) in Northwest Iberian Peninsula 17 (Pysek et al., 2002). Recently Pysek et al. (2002) shown that nature reserves all over the world are invaded about half as often as sites outsides reserves. 1.5. Using models to achieve and anticipate patterns of invasion Species distribution models (SDM) are empirical models that relate field observations to environmental predictors variables centered on the statistical or theoretical derived response surfaces (Guisan and Thuiller, 2005). The use of SDMs has grown in the last two decades due to the recognized application value in applied scientific fields as ecosystem management and biodiversity conservation (Rodríguez-Reya et al., 2013). SDMs can also be applied in fundamental ecological questions as, such as the ecological impacts of climate and land-use changes in biological invasions (Vicente et al., 2011; Guisan and Thuiller, 2005). Many scientific studies in invasions applied SDMs as a tool, for example: to assess species invasion and proliferation, to model species assemblages (biodiversity, composition), to quantify the environmental niche of the species, to support appropriate management plans for species recovery (mapping suitable sites for species reintroduction), to support conservation planning and reserve selection (suggesting surveyed sites of high potential of occurrence for rare species; Guisan and Thuiller, 2005). Despite many recent applied purposes of SDMs related to climate change and conservation planning, the use of these tools in theoretical ecology and evolution is re-emerging (Guisan and Thuiller, 2005). According to Nentwig et al. 2008, the main research fields of biological invasions can be classified as: pathways of biological invasions, traits of a good invader, patterns of invasion and invasibility, ecological impacts of biological invasions, economy and socio-economy of biological invasions, and finally, prevention and management of biological invasions. One of the major goals in biological invasion studies is to understand which and how biotic and abiotic factors constrain the spread of invasive species, determining current and future potential distributions (Wang and Jackson, 2011; Farashi et al., 2013). Research efforts must be directed to find ways to anticipate invasions in order to protect global biodiversity from the effects of alien species (Vicente et al., 2011; Jones, 2012). This goal can be partially achieved determining the potential geographic extent of invasions, requiring predictive tools to assess the geographic distribution of invasive species under current and future environmental conditions (Vicente et al., 2011; Farashi et al., 2013). These predictions can simplify the early detection of invasive species and maximize monitoring efficiency (Jones, 2012). FCUP Using models and connectivity analysis to predict the current and future patterns of invasion by Silver-wattle (Acacia dealbata Link) in Northwest Iberian Peninsula 18 Ongoing climate and land-use changes are forecasted to boost the invasion of alien species in several habitats (Vicente et al., 2011) being important to have a multi-scale approaches due to most of the pathways of species introductions result from human activities at several spatial and temporal scales (Vicente et al., 2011; Razgour et al., 2011). Although most of the studies using SDMs considered environmental predictors at a single grain size and fixed spatial extent, factors driving the distribution and abundance of organisms can act at different scales (Vicente et al., 2011; Guisan and Thuiller, 2005). For example: topography, geomorphology, human land-use or biotic interactions are usually considered to operate at regional and local scales while climate operate at the global scale (Vicente et al., 2011; Elith and LeathwicK, 2009). There are several available modelling tools (Thuiller et al., 2009), being Biomod2 one of the newest and accurate statistical package. Biomod2 is implemented in the R statistical software (Georges and Thuiller, 2013), allowing the calibration of ensemble forecasting species distributions, overcoming a range of uncertainties and assumptions in the methodological models. This package allows the calibration of a model applying 10 different modelling techniques, and to forecast the potential species distribution under different environmental conditions (climate and land use change scenarios; Thuiller et al. 2009). Due to the complex nature of species range dynamics and the related factors acting at multiple spatial scales, modelling species distribution is a challenging procedure (Vicente et al., 2011; Guisan and Thuiller, 2005). 1.6. Objectives In this work we applied a combined predictive modelling (CPM) framework to: (1) identify the areas that are today potentially invaded by Silver-wattle (Acacia dealbata Link) in Northwest Iberian Peninsula (Galicia and North of Portugal); and (2) forecast the effects of climate changes on the distribution of this species by projecting the models for future conditions. Our rationale was that the application of CMPs to predict the current and future patterns of invasion can support measures to prevent invasions in the study region and support cross-border management to protect important habitats from invasion. FCUP Using models and connectivity analysis to predict the current and future patterns of invasion by Silver-wattle (Acacia dealbata Link) in Northwest Iberian Peninsula 19 2. Methods 2.1. Study area The study area is located in the Northwest Iberian Peninsula (Fig. 2), covering the North Portugal and Galicia. This area presents a high plant biodiversity, with more than 3000 species and subspecies, many of them endemic of this region (http://www.biodiversidade.eu/pt/info/proxecto/). This region is experiencing a strong environmental disturbance promoted by many factors, like invasive species, creating a serious threat to conservation of biodiversity (http://www.biodiversidade.eu/pt/info/proxecto/). Figure 2 – Location of the study area in Europe (a) and Iberian Peninsula (b) with digital elevation model detail (c). The area extent is 35017 km², covering the transition between the Mediterranean and Atlantic biogeographic regions (Rívas-Martínez et al., 2004). The study area is very heterogeneous, with elevation ranging from the sea level (west) to 2059 m in the eastern mountains, with the major rivers, as Douro, Tua, and Cenza, running from east to west. The geology of the mountains is overwhelmed by acid igneous and FCUP Using models and connectivity analysis to predict the current and future patterns of invasion by Silver-wattle (Acacia dealbata Link) in Northwest Iberian Peninsula 20 metamorphic rocks that form acid soils (Roucoux et al., 2005). The study area presents a very complex land cover (more than 11 classes of land cover, according to the European Environment Agency) and a large network of protected sites: Parque Nacional Peneda-Geres, Parque Natural de Montesinho and Parque Natural de Alvão in Portugal and Serra Candán and Pena Trevinca in Spain. Currently this region has a temperate climate (Roucoux et al., 2005) and carries the full force of the westerly winds bringing cyclones from the Atlantic (Roucoux et al., 2005). The precipitation levels in the area are the highest in the whole peninsula (Roucoux et al., 2005) (up to 3000mm per year in the eastern mountain summits). The annual mean temperatures ranges from ca. 5°C to ca. 17°C, where the summers are cool (18-22ºC (Roucoux et al., 2005) and winters are mild and frost-free (10-13ºC around the coast) (Roucoux et al., 2005). 2.2. Test species and occurrence data The test species is an alien invader plant from the Leguminosae family, Acacia dealbata Link (Fig. 3). This wattle species is original from the southeast of Australia and found in a wide range of different habitats, from coastal to subalpine regions, and from high rainfall to arid inland areas, growing in tropical, subtropical and warm temperate regions (Lorenzo et al., 2009). Figure 3 – Acacia dealbata Link. This species can be found in areas with over 500 mm rainfall, typically at altitudes from 350-1000m above the sea level (Lorenzo et al., 2009). Introduced in Europe in the 1820s (Carballeira and Reigosa, 1999), and planted as an ornamental plant in southern Europe, which offered favorable climates for this development, with sufficient sun exposure an little frosts, becoming widely naturalized in this area (Lorenzo et al., FCUP Using models and connectivity analysis to predict the current and future patterns of invasion by Silver-wattle (Acacia dealbata Link) in Northwest Iberian Peninsula 21 2009). In the Southwest Europe this species occurs in riparian zones, water courses and sunny edges of pinewoods, or on the south and west facing slopes, where it forms dense stands that choke in the natural vegetation and it often invades area under intensive agriculture uses and further away from the sea than other Acacia species present in these regions (Lorenzo et al, 2009). Acacia dealbata was introduced in the Iberian Peninsula in the second half of the 19th century (Sanz Elorza et al., 2004) and became a problematic species in Portugal and Spain, where it is threatening the native flora and becoming a serious environmental problem. In addition to its great colonizing capacity it leads to a very low covering of undergrowth species caused by its alellopathic ability (Vicente et al, 2011). The occurrence dataset for Portugal was collected from previous scientific works (Vicente et al., 2013) and Spain data were and provided by Direccion Xeral de Conservacion da Natureza da Xunta da Galicia. Sampling sites were selected using a random-stratified strategy (Vicente et al., 2013), employing climate (mean annual temperature) and bedrock as stratifying layer. Sampling units were cells of 1km2 in which presents where collected between March and April 2011. 2.3. Selection and classification of environmental predictors We selected 53 predictors that, by expert knowledge and according to previous reporting in scientific literature, operate as determinants of ecology and distribution of the target species (see annex 1). To avoid correlation, we applied the Spearman rank correlation coefficient (non-parametric test), to select the predictors with lowest correlation between each other. We use the software STATISTICA to select the predictors and only those with correlation lower than 0.5 were considered (Vicente et al., 2011). Therefore we select 17 environmental predictors to calibrate the models (Table 1 describes the predictors chosen). Climate predictors were obtained using information from the WorlClim database in raster format with a spatial resolution of 1km2 while the remaining environmental predictors were obtained from detailed environmental maps. The environmental predictors were classified a priori as “regional” or “local” based on ecological theory and their spatial scale of variation using a statistical predictor classification based on spatial autocorrelation (Vicente et al., 2011).Using point pattern statistic (PPSA—spdep R package; available at http://cran.r-project.org/web/ packages/spdep) we calculated Geary’s c autocorrelation measure for all predictor variables for increasing neighbourhood distances. At least, we performed a hierarchical FCUP Using models and connectivity analysis to predict the current and future patterns of invasion by Silver-wattle (Acacia dealbata Link) in Northwest Iberian Peninsula 22 clustering based on a Euclidean distance matrix of all Geary’s values and the two groups in the final classification tree were very well separated corresponding to predictors with local versus regional influence (Vicente et al., 2011). FCUP Using models and connectivity analysis to predict the current and future patterns of invasion by Silver-wattle (Acacia dealbata Link) in Northwest Iberian Peninsula 29 Figure 5 – Possible combinations for species distributions dynamics using combined models (Vicente et al., 2011). The letters A, B, C and D refers to the different possible combinations (Aregional and local suitability, Bonly regional suitability. Conly local suitability and Dno suitability). 2.7. Current and future conflicts with protected areas The models of the species were used to forecast the current and future species distributions and their conflict with the protected areas present in the study area, by spatially overlapping the potential distribution maps with a map of the protected area, corresponding to Natura 2000 networks. With this projection it's possible to determine the potential threats to the protected areas in the near future, as well as the possible impacts that may arise from them. By overlapping the models of the species with a map of protected areas (corresponding to Nature Network 2000 – ICNF for Portugal and to Espacios Naturales Protegidos en España for Spain) the distribution maps of combined models allow a detection of present and future invasion areas. FCUP Using models and connectivity analysis to predict the current and future patterns of invasion by Silver-wattle (Acacia dealbata Link) in Northwest Iberian Peninsula 30 2.8. Connectivity of the distribution of the invasive species First, we calculated the temporal changes in spatial connectivity of predicted suitable areas for the test species across the test area, using the connectivity index developed by Randy et al., 2009 (Vicente et al., 2013). We considered value of species presence: probability of 1 for the response type A, probability of 0.5 for response type B and C, and probability of 0 for response type D. The connectivity index attains a maximum value of 1 when all the cells surrounding a focal suitable cell are also suitable (Vicente et al., 2013). We quantified the changes in the spatial relationship between protected areas and the connectivity of predicted areas using the raster calculator in ArcGIS 9.3 (ESRI, 2010). Then, we analysed the trend of connectivity for the specie in space (full area and inside protected areas) and time (2020 and 2050). At last, a Spearman rank correlation test was used to identify significant relations between connectivity and protection value. FCUP Using models and connectivity analysis to predict the current and future patterns of invasion by Silver-wattle (Acacia dealbata Link) in Northwest Iberian Peninsula 31 3. Results 3.1. Predictions and determinants of current distributions The potential distribution of the species Acacia dealbata under current conditions is illustrated in Figure 6. The combined predictive model was calibrated using 17 environmental variables "a priori" classified as "regional" and "local" (see Table 2). Figure 6 - Spatial projection of Acacia dealbata potential distribution using combined models, under current conditions (2000). . FCUP Using models and connectivity analysis to predict the current and future patterns of invasion by Silver-wattle (Acacia dealbata Link) in Northwest Iberian Peninsula 32 Table 2 – Environmental predictors selected to calibrate the combined models. The partial regional model was calibrated with 7 environmental variables, and an AUC value of 0.849 was obtained, while the partial local model was calibrated with 10 environmental variables with a final AUC value of 0.887. From the 7 regional environmental predictors, the most important for partial regional model calibration were Minimum Temperature of Coldest Month (MTCM, with a value of 0.394), Annual Precipitation (AP, with a value of 0.282) and Temperature Annual Range (TAR, with a value of 0.238). Considering the partial local model, between the total of 10 local predictors, the most important were Density of local road network (DensRoad, with a value of 0.392), percentage of arable land (pArL, with a value of 0.173 )and percentage cover of mixed forest (pMixFo, with a value of 0.197). Under current climate conditions (2000), the potential percentage of predicted area for Acacia dealbata in areas with regional and local suitability, type A, was 18.3%. For areas with only-regional suitability, type B, the percentage of predicted area 27% while in the areas with only local suitability, type C, the percentage of predicted area was Predictors Description Scale of variation AP Annual Precipitation Regional DistRiv Distante to main river Regional GPP Mean gross annual primary productivityMM Regional MTCM Min Temperature of Coldest Mont Regional pCamb % of cambissoils (per 1km2) Regional PS Precipitation Seasonality Regional TAR Temperature Annual Range Regional DensRiv Density of local hydrographic network Local DensRoad Density of local road network Local Distroad Distance to roads Local MPAR Mean perimeter-area ratio Local MSI Mean shape index Local PArL % cover of arable land (per 1km2) Local PArS % cover of artificial stands (per 1km2) Local pBiFo % cover of broad-leaf forest (per 1km2) Local pCoFo % cover of conifer forest (per 1km2) Local pMixFo % cover of mix forest (per 1km2) Loca FCUP Using models and connectivity analysis to predict the current and future patterns of invasion by Silver-wattle (Acacia dealbata Link) in Northwest Iberian Peninsula 33 14%. Finally, for areas predicted to be unsuitable for the species (unsuitable local and regional conditions), type D, the percentage of predicted area was 40.7% (Table 3). Table 3 – Predicted percentage of the study area occupied by each “occupancy type” in the combined models, for the test species, under current conditions (2000). Occupancy types Percentage of predicted área Type A - regional and local suitability 18.3% Type B - only regional suitability 27.0% Type C - only local suitability 14.0% Type D - no suitability 40.7% 3.2. Forecast of future distributions The spatial distribution of the test species under future climate conditions for the year 2020 is illustrated in Figure 7. The combined models for the A1 scenario (Fig. 7-b) and for the B2 scenario (Fig. 7-d) were obtained by overlapping the spatial projections of the local and the regional partial models. Figure 7Spatial projections under current conditions (2000 - a) and future conditions - 2020 A1 scenario (b) and Bb2 scenario (d) and 2050 A1 scenario (c) and B2 scenario (e). FCUP Using models and connectivity analysis to predict the current and future patterns of invasion by Silver-wattle (Acacia dealbata Link) in Northwest Iberian Peninsula 34 Under future conditions (2020) considering A1 climatic scenario, predicted area with local and regional suitability, type A, was 22.3%.The percentage for areas with only regional suitability, type B, was 33.3%, and for areas with only local suitability, type C was 10%. Finally, unsuitable regional and local areas, type D, were predicted as 34.4%. For B2 climatic scenario, type A (both regional and local suitability) predicted area was 20%, type B (regional suitability) was 28.9%, for type C (local suitability) was 12.3% and for type D (both regional and local unsuitability) was 38.8% (Table 4). Table 4 - Predicted percentage of the study area occupied by each “occupancy type” in the combined models, for the test species, under future conditions (2020), for the a1 and b2 scenarios. Percentage of predicted area Occupancy types A1 B1 Type A - regional and local suitability 22.3% 20.0% Type B - only regional suitability 33.3% 28.9% Type C - only local suitability 10.0% 12.3% Type D - no suitability 34.4% 38.8% A spatial distribution for the species Acacia dealbata, under future climate conditions for the year 2050 is presented in the figure 7, where the combined model for the A1 scenario (fig. 7c) and for the B2 scenario(fig 7e). In table 5 is presented the estimated percentage of the study area occupied by each “occupancy types”. Areas predicted with both regional and local suitability (Type A) was 26,5% in A1 scenario and 22.8% in B2 scenario. The percentage of area predicted to be type B (only regional suitability) was 46% in the a1 scenario and 36.4% in the b2 scenario, and the percentage of areas predicted as type C (only local suitability) was 5.8% in the a1 scenario and 9,4% in b2. Areas predicted as unsuitable local and regional conditions (Type D) was 21.7% and 31.4% in the a1 and b2 scenarios, respectively. Table 5 - Predicted percentage of the study area occupied by each “occupancy type” in the combined models, for the test species, under future conditions (2050), for the a1 and b2 scenarios. Percentage of predicted area Occupancy types A1 B2 Type A - regional and local suitability 26.5% 22.8% Type B - only regional suitability 46.0% 36.4% Type C - only local suitability 5.8% 9.4% Type D - no suitability 21.7% 31.4% FCUP Using models and connectivity analysis to predict the current and future patterns of invasion by Silver-wattle (Acacia dealbata Link) in Northwest Iberian Peninsula 35 3.3. Range dynamics between years 2000 and 2050 Based on the potential distribution of the species predicted for the years of 2000, 2020 and 2050, we analysed the potential range dynamics for the species Acacia dealbata between the time periods: 2000-2020; 2000-2050; and 2020-2050. Predicted changes in range dynamics for the species Acacia dealbata between 2000-2020, 2000-2050 and 2020-2050 (under climate change scenarios) are shown in Figure 8. For the three time periods, predominate dynamic obtained was the response type “no change” (see Table 6). In both scenarios, for the responses type “Improvement of the conditions” and “Colonization”, from the periods 2000-2020 to 2000-2050 an increase of the response was observed, and for the period of 2020-2050 a decrease was observed compared to the time period 2000-2050. For the response type “Deterioration of the conditions”, for both scenarios, a decrease trend was observed from the period of 2000-2020 to the period of 2000-2050. From time period 2000-2050 to 2020-2050 the obtained values were the same. Finally, for the response type “extinction”, a decreasing of the values was observed from the time period 2000-2020 to 2000-2050 and also from the time period of 2000-2050 to 2020-2050. Table 6 – Dynamic of environmental suitability for the test species from combined modes, under climate change scenarios, for the periods 2000-2020, 2000-2050 and 2020-2050. Scenarios Dynamics No change Improvement Deterioration Extinction Colonization A1B 2000-2020 77.9% 5.1% 1.0% 4.9% 11.1% 2000-2050 68.1% 8.6% 0.3% 2.0% 2.1% 2020-2050 80.0% 4.5% 0.3% 1.3% 13.9% B2 2000-2020 89.9% 2.7% 0.9% 4.3% 6.2% 2000-2050 79.1% 5.2% 0.6% 2.9% 12.2% 2020-2050 84.2% 3.4% 0.6% 2.2% 9.6% FCUP Using models and connectivity analysis to predict the current and future patterns of invasion by Silver-wattle (Acacia dealbata Link) in Northwest Iberian Peninsula 36 Figure 8 – Changes in the potential and dynamic of the test species between 2000-2020 (for the A1 scenario – aand the B2 scenario - d), 2000-2050 (for the A1 scenario – band the B2 scenario - e) and 2020-2050 (for the A1 scenario – c - and the B2 scenario - f). 3.4. Current and future conflicts with protected areas In Table 7 we present the percentage of area of Natura 2000 sites occupied by the different response types for the test species Acacia dealbata, for current (2000) and future (2020 and 2050) environmental conditions (Figure 9).For the climatic scenario A1, the percentage of predicted area with the response types A (regional and local suitability for the species) within Natura 2000 sites increases in the year 2020 and 2050. Considering predicted type B (regional suitability) and D (both unsuitable regional and local) inside the Natura 2000 sites, it’s predicted no changes for the year 2020 followed by an increase of predicted percentage of type B (regional suitability), and a decrease in the type D (both regional and local unsuitability), for 2050. The percentage of occupancy of the type C (local suitability) inside the Natura 2000 sites decreases both years. FCUP Using models and connectivity analysis to predict the current and future patterns of invasion by Silver-wattle (Acacia dealbata Link) in Northwest Iberian Peninsula 37 For the climatic scenario B2, a decrease of the percentage of occupancy of the species inside Natura 2000 sites can be observed for the year of 2020 for response types A (regional and local suitability) and B (regional suitability), followed by an increase in the year 2050. An increase of the percentage is observed for response types C (local suitability) and D (both regional and local unsuitability) for the year 2020, while in 2050 the percentage decreases. Table 7– Percentage of occupancy in Natura 2000 for each occurrence type (regional and local suitability (type A), only regional suitability (type B), only local suitability (type C), and no suitability (type D), for the test species, for current (2000) and future (2000, 2050) conditions. Percentage of predicted area Occupancy type 2000 2020 a1 2020b2 2050a1 2050b2 Type A - regional and local suitability 9.0% 10.6% 8.6% 13.9% 11.0% Type B - only regional suitability 18.9% 18.9% 15.4% 34.5% 22.2% Type C - only local suitability 11.8% 10.2% 12.2% 6.9% 9.8% Type D - no suitability 60.3% 60.3% 63.8% 44.7% 57.0% Figure 9Spatial projections of invasion within Natura 2000 sites (protected areas for combined models, under current (2000) and future conditions (2020 and 2050). FCUP Using models and connectivity analysis to predict the current and future patterns of invasion by Silver-wattle (Acacia dealbata Link) in Northwest Iberian Peninsula 38 Predominate dynamic type predicted by the combined models in each time period within Natura 2000, was “No change” (see Fig 10 and Table 8). For both climatic scenarios, responses types “deterioration of the conditions” and “extinction” decreased from the time period 2000-2020 to 2000-2050 and from the time period 2000-2050 to 2020-2050. Considering the A1 climatic scenario, the response types “Improvement of the conditions” and “Colonization”, from the time period 2000-2020 to 2000-2050, was increasing and then, from the time period 2000-2050 to 2020-2050, was decreased. Considering the B2 climatic scenario, from the time period 2000-2020 to 2000-2050 and 2000-2050 to 2020-2050, an increasing was observed. Table 8 – Percentage of occupancy in Natura 2000 of each dynamic type (no change, improvement of the conditions, deterioration of the conditions, extinction and colonization), for the test species, for the periods 2000-2020, 2000-2050 and 2020-2050. Scenarios Dynamics No change Improvement Deterioration Extinction Colonization A1B 2000-2020 82.2% 2.8% 1.2% 6.9% 6.9% 2000-2050 70.4% 5.5% 0.6% 3.9% 19.6% 2020-2050 79.3% 3.5% 0.1% 0.7% 16.4% B2 2000-2020 84.4% 1.3% 1.6% 8.1% 4.6% 2000-2050 84.3% 2.8% 0.8% 4.4% 7.7% 2020-2050 85.9% 2.9% 0.5% 2.0% 8.7% FCUP Using models and connectivity analysis to predict the current and future patterns of invasion by Silver-wattle (Acacia dealbata Link) in Northwest Iberian Peninsula 45 4.4. CPMs, connectivity of invader distributions, and protected areas Connectivity analysis is useful in ecology research (Vicente et al., 2013), and it elucidates patterns, thereby providing important insights into the processes that drive plant invasions in habitats (Minor et al., 2009). Surfaces with high connectivity represent potential dispersal corridors and areas with high connectivity will be the ones where management may be less effective, as local control measures may be counteracted by immigration and propagule pressure from neighbouring invaded areas (Vicente et al., 2013). In this study, we used a connectivity index based on the nearest neighbours that is considered a crude connectivity metric, although simpler to obtain. However, combined predictive models projections allowed to weight the potential suitable areas for the Acacia dealbata species by evaluating and quantification of the availability of both regional conditions and suitable local habitat. Areas with high connectivity of the species will be the ones with less effective management, as local control measures may be counteracted and propagule pressure from neighboring invaded areas (Vicente el al., 2013). For the test species, connectivity was predicted to increase in the future. Although connectivity of suitable areas for the species inside protected areas is lower than across the full areas (because protected areas coincide with areas of high elevation) the changes in environmental conditions predicted for the future may change the connectivity of the species (Vicente et al., 2013). In conclusion, the connectivity analysis showed that protected areas will potentially suffer a high pressure from Acacia dealbata under future environmental conditions. FCUP Using models and connectivity analysis to predict the current and future patterns of invasion by Silver-wattle (Acacia dealbata Link) in Northwest Iberian Peninsula 46 5. Conclusions Using a combined predictive modeling approach to forecast the current and future distribution of Acacia dealbata in Northwest Iberian Peninsula, as well as assessing its current and future conflicts with protected areas, we conclude that:  Models predict that the area with suitable conditions of climate (regional) and habitat/landscape (local) will increase in the future, if the climate change scenarios are confirmed;  The species currently invades nature reserves in the study area, and models predict that the percentage of occupancy will increase in the future; and  Connectivity analysis showed that changes in future environment may increase the connectivity of Acacia dealbata distribution in the test area. From an applied perspective, our results suggest that:  Spatial predictions of the models provide an important basis for the efficient control of this invasive species in invaded areas and for the early detection of new areas of invasion;  CMP projections can support measures to prevent invasions in this region and promote cross-border management of invasions; and  Connectivity analyses are important in invasion ecology, as they can contribute to prioritize management actions considering regional dispersal corridors and the probability of recolonization after local control measures have been undertaken. FCUP Using models and connectivity analysis to predict the current and future patterns of invasion by Silver-wattle (Acacia dealbata Link) in Northwest Iberian Peninsula 47 References Barrows, C.W., Murphy-Mariscal, M.L. (2012). Modelling impacts of climate change on Joshua trees at their southern boundary: How scale impacts predictions. Biological Conservation 152, 29–36. Beaumont L.J., Gallagher R.V., Downey P.O., (2009) Modelling the impact of Hieraciumspp. on protected areas in Australia under future climates. Ecography 32:757-764. Becker, T., Dietz, H., Billeter, R.,Buschmann, H. (2005). Altitudinal distribution of alien plant species in the Swiss Alps. Perspectives in Plant Ecology, Evolution and Systematics 7, 173–183. Bradley, B.A., Blumenthal, D.M., Wilcove, D.S., Ziska, L.H. (2010) . Predicting plant invasions in an era of global change. Trends in Ecology and Evolution Vol.25 No.5, 311. Broennimann, O. and Guisan, A. (2008). Predicting current and future biological invasions: both native and invaded ranges matter. Biol. Lett. 4, 585–589. Brown, K.A., Spector, S, WU, W. (2008). Multi-scale analysis of species introductions: landscape and demographic models to improve management decisions about nonnative species. Journal of Applied Ecology, 45, 1639–1648. Carballeira, A., Reigosa, M.J. (1999). Effects of natural leachates of Acacia dealbata Link in Galicia (NW Spain). Bot. Bull. Acad. Sin. 40, 87-92. Castro – Díez, P., Gogoy, O., Saldanã, A., Richardson, D.M. (2011). Predicting invasiveness of Australian acacias on the basis of their native climatic affinities, life history traits and human use. Diversity and Distributions 17, 934–945. Catford, J.A., Vesk, P.A., White, M.D., Wintle. B.A. (2011).Hotspots of plant invasion predicted by propagule pressure and ecosystem characteristics. Diversity and Distributions, (Diversity Distrib.) 17, 1099–1110. Chapin III, F.S., Zavaleta, E.S., Eviner, V.T., Naylor, R.N., Vitousek, P.M., Reynolds, H.L.; Hooper, D.U., Lavorel, S., Sala, O.E., Hobbie, S.E., Mack, M.C., Díaz, S. (2000).Consequences of changing Biodiversity. NATURE 405. Crossman, N.D., Bryan, B.A., Cooke, D.A (2011).An invasive plant and climate change threat index for weed risk management: Integrating habitat distribution pattern and dispersal process. Ecological indicators 11 1 8 3 – 1 9 8.Dirzo, R., Raven, P.H. (2003). Global State of Biodiversity and Loss. Annu. Rev. Environ. Resour. 28, 137–67. Elith, J., Leathwick, J.R. (2009). Species Distribution Models: Ecological Explanation and Prediction Across Space and Time . Annu. Rev. Ecol. Evol. Syst. 40, 677-697. FCUP Using models and connectivity analysis to predict the current and future patterns of invasion by Silver-wattle (Acacia dealbata Link) in Northwest Iberian Peninsula 48 Essl, F., Milasowszky, N., Thomas Dirnböck, T. (2011). Plant invasions in temperate forests: Resistance or ephemeral phenomenon? Basic and Applied Ecology 12, 1–9. Evangelista, P.H., Kumar, S., Stohlgren, t.j., Young, N.E. (2011). Assessing forest vulnerability and the potential distribution of pine beetles under current and future climate scenarios in the Interior West of the US. Forest Ecology and Management 262, 307–316. Fernández, M., Hamilton, H., Alvarez, O., Guo, Q. (2012). Does adding multi-scale climatic variability improve our capacity to explain niche transferability in invasive species? Ecological Modelling 246, 60– 67. Fischer, D., Thomas, S.M., Niemitz, F., Reineking, B., Bierkuhnlein, C. (2011). Projection of climatic suitability for Aedes albopictus Skuse (Culicidae) in Europe under climate change conditions. Global and Planetary Change 78 , 54–64. Fuentes-Ramírez, A., Pauchard, A., Cavieres, L.A., García, R.A. (2011). Survival and growth of Acacia dealbata vs. native trees across an invasion front in south-central Chile. Forest Ecology and Management 261, 1003–1009. Gallardo, B. and Aldridge, D.C. (2013). Evaluating the combined threat of climate change and biological invasions on endangered species. Biological Conservation 160, 225–233. Gaertner, M., Richardson, D.M., Prirett, S.D.J. (2011). Effects of Alien Plants on Ecosystem Structure and Functioning and Implications for Restoration: Insights from Three Degraded Sites in South African Fynbos. Environmental Management 48, 57– 69. Gaikwad, J., Wilson, P.D., Ranganthan, S. (2011). Ecological niche modeling of customary medicinal plant species used by Australian Aborigines to identify speciesrich and culturally valuable areas for conservation. Ecological Modelling 222, 3437– 3443. Gallien, L., Douzet, R., Pratte, S., Zimmermann, N.E., Thuiller, W. (2012). Invasive species distribution models - how violating the equilibrium assumption can create new insights. Global Ecology and Biogeography, (Global Ecol. Biogeogr.) 21, 1126–1136. Geldmann, J., Barnes; M., Coad, L., Craigie, I.D., Hockings, M., Burgess, N.D. (2013). Effectiveness of terrestrial protected areas in reducing habitat loss and population declines. Biological Conservation 161, 230-238. Guisan, A., Thuiller, W. (2005). Predicting species distribution: offering more than simple habitat models. Ecology Letters 8, 993–1009. Guisan, A., Zimmermann, N.E. (2000). Predictive habitat distribution models in ecology. Ecological modelling 135, 147-186. FCUP Using models and connectivity analysis to predict the current and future patterns of invasion by Silver-wattle (Acacia dealbata Link) in Northwest Iberian Peninsula 49 Gulezian, P.Z., Nyberg, D.W. (2010). Distribution of invasive plants in a spatially structured urban landscape. Landscape and Urban Planning 95, 161-168. González-Muñoz, N., Costa-Tenorio, M., Espigares, T. (2012).Invasion of alien Acacia dealbata on Spanish Quercus robur forests: Impact on soils and vegetation. Forest Ecology and Management 269, 214–221. Farashi, A.; Kaboli, M. and Karami, M. (2013). Predicting range expansion of invasive raccoons in northern Iran using ENFA model at two different scales. Ecological Informatics 15, 96–102. Hobbs, R.J., Humphries, S.E. (1995). An Integrated Approach to the Ecology and Management of Plant Invasions. Conservation Biology, Pages 761-770 Volume 9, No. 4. Jones, C.C. (2012). Challenges in predicting the future distributions of invasive plant species. Forest Ecology and Management 284, 69–77. Kleinbauer, I., Dullinger, S., Peterseil, J., Essl, F. (2010). Climate change might drive the invasive tree Robinia pseudacacia into nature reserves and endangered habitats. Biological Conservation 143, 382–390. Lake, J.C., Leishman, M.R. (2004). Invasion success of exotic plants in natural ecosystems: the role of disturbance, plant attributes and freedom from herbivores. Biological Conservation Volume 117, Issue 2, Pages 215–226. Le Maitre, D.C., Gaertner, M., Marchante, E., Ens, E., Holmes, P.M., Pauchard, A., O’Farrell, P.J., Rogers, A.M., Blanchard, R., Blignaut, J., Richardson, D.M. (2011). Impacts of invasive Australian acacias: implications for management and restoration. Diversity and Distributions, (Diversity Distrib) 17, 1015–1029. Liu, C., White, M., Newell, G., Griffione (2013). Species distribution modelling for conservation planning in Victoria, Australia. Ecological Modelling 249, 68– 74. Lomba A., Pellissier, L., Randin , C., Vicente, J. Moreira, F., Honrado, J., Guisan, A. (2010). Overcoming the rare species modelling paradox: A novel hierarchical framework applied to an Iberian endemic plant. Biological Conservation 143, 2647– 2657 Lorenzo, P., González, L., Reigosa, M.J. (2010). The genus Acacia as invader: the characteristic case of Acacia dealbata Link in Europe. Ann. For. Sci. 67, 101. Méndez, V., Llopis, I., Campos, D., Horsthemkc, W. (2011). Effect of environmental fluctuations on invasion fronts. Journal of Theoretical Biology 281, 31–38. Millennium Ecosystem Assessment (2005). Ecosystems and human well-being: synthesis report. Island Press, Washington, DC. Online at http://www.millenniumassessment.org/ FCUP Using models and connectivity analysis to predict the current and future patterns of invasion by Silver-wattle (Acacia dealbata Link) in Northwest Iberian Peninsula 50 Minor, E.S., Tessel, S.M., Engelhardt, K.A.M., Lookingbill, T.R. (2009). The role of landscape connectivity in assembling exotic plant communities: A network analysis. Ecology. 90, 1802-1809. Moilanen, A., Nieminen, M., 2002. Simple connectivity measures in spatial ecology. Ecology. 83, 1131–1145 Narščius, A., Olenin, S., Zaiko, A., Minchin, D. (2012). Biological invasion impact assessment system: From idea to implementation. Ecological Informatics 7, 46–51. Ohlemuller, R., Walker, S., Wilson, B. (2006). Local vs regional factors as determinants of the invasibility of indigenous forest fragments by alien plant species. OIKOS 112: 493_/501. Pino, J., Font, X., Carbó, J., Javé, M., Pallaès, L. (2005). Large-scale correlates of alien plant invasion in Catalonia (NE of Spain). Biological Conservation 122, 339–350. Pysek, P., Jarosik, V., Kucera, T. (2002). Patterns of invasion in temperate nature reserves. Biological Conservation 104, 13–24. Razgour, O., Hammer, J., Jones, G. (2011). Using multi-scale modelling to predict habitat suitability for species of conservation concern: The grey long-eared bat as a case study. Biological Conservation 144, 2922–2930. Richardson, D.M., Pysek, P., Rejmanek. M., Barbour, M.G., Panetta, F.D., C.J. West (2000). Naturalization and invasion of alien plants: concepts and definitions. Diversity and Distributions 6, 93–107. Rivas-Marínez S., Penas, A., Díaz, T.E. (2004). Biogeographic Map of Europe. Cartographic Service, University of Léon, Spain. Robinson, T. P., Van Klinken, R. K., Metternicht, G. (2010). Comparison of alternative strategies for invasive species distribution modeling. Ecological Modelling 221, 2261– 2269. Rodríguez-Rey, M., Jiménez-Valverde, A., Acevedo, P. (2013). Species distribution models predict range expansion better than chance but not better than a simple dispersal model. Ecological Modelling 256, 1– 5. Roucoux, K.H., Abreu, L., Shackleton, N.J., Tzedakis, P.C. (2005). The response of NW Iberian vegetation to North Atlantic climate oscillations during the last 65kyr. Forest Ecology and Management 297, 84–93. Saume, I., Kowarik. I. (2010). Urban rivers as dispersal corridors for primarily winddispersed invasive tree species. Landscape and Urban Planning 94, 244–249. Sanz-Elorza, M., Dana, E.D., Sobrino, E. (2004). Atlas de las plantas alóctonas invasoras de Espanã. Madrid: Dirección General para la Biodiversidad. Staudinger, M.D., Grimm, N.B., Staudt, A., Carter, L.S., Chapin III, F.S., Kareiva, P., Ruckelshaus, M., Stein, B.A. (2012). Impacts of Climate Change on Biodiversity, FCUP Using models and connectivity analysis to predict the current and future patterns of invasion by Silver-wattle (Acacia dealbata Link) in Northwest Iberian Peninsula 51 Ecosystems, and Ecosystem Services: Technical Input to the 2013 National Climate Assessment. Cooperative Report to the 2013 National Climate Assessment. 296 p. Spear, P., Foxcroft, L.C., Vezuidenhout, H., McGeoch, M.A. (2013). Human population density explains alien species richness in protected areas. Biological Conservation 159, 137 - 147. Terereai, F., Gaertner, M., Jacobs, S.M., Ricardson, D.M. (2013). Eucalyptus invasions in riparian forests: Effects on native vegetation community diversity stand structure and composition. Forest Ecology and Management 297, 84–93. Thuiller, W., Richardson, D.M., Pysek, P., Midgley, G.F., Hughes, G.O., Rouge, M. (2005).Niche-based modelling as a tool for predicting the risk of alien plant invasions at a global scale. Global Change Biology 11, 2234–2250. Thuiller, W., Lafourcade, B., Engler, R., Araujo, M.B. (2009). BIOMOD, a platform for ensemble forecasting of species distributions. Ecography 21, 369-373. Thuiller, W., Georges D., Engler R. (2013). Ensemble platform for species distribution modeling. Trethowan,P.D., Robertson,M.P., McConnachie, A.J. (2011). Ecological niche modelling of an invasive alien plant and its potential biological control agents. South African Journal of Botany 77, 137–146. Uddin, M.B., Steinbauer, M.J., Jentsch, A., Mukul, S.A., Beierkuhnlein, C. (2013). Do environmental attributes, disturbances and protection regimes determine the distribution of exotic plant species in Bangladesh forest ecosystem? Forest Ecology and Management 303, 72–80. Vicente, J., Alves, P., Randin, C., Guisan, A., Honrado, J. (2010). What drives invasibility? A multi model inference test and spatial modelling of alien plant species richness patterns in northern Portugal. Ecography 33: 1081-1092. Vicente, J., Randin, C.F., Gonçalves, J., Metzer. M.J., Lomba, A., Honrado. J., Guisan, A. (2011). Where will conflicts between alien and rare species occur after climate and land-use change? A test with a novel combined modelling approach. Biol Invasions 13, 1209–1227. Vicente, J.R., Fernandes, R.F., Randin, C.F., Broennimann, O., Gonçalves, J., Marcos, B., Pôças, I., Alves, P., Guisan, A., Honrado, J.P (2013). Will climate change drive alien invasive plants into areas of high protection value? An improved model-based regional assessment to prioritize the management of invasions (in press) a Vicente J., Pereira, H.M., Randin, C.F., Gonçalves, J., Lomba, A., Alves, P., Metzger, M., Cezar, J., Guisan, A. and Honrado, J. (2013): A multi-model dissimilarity framework FCUP Using models and connectivity analysis to predict the current and future patterns of invasion by Silver-wattle (Acacia dealbata Link) in Northwest Iberian Peninsula 52 to evaluate the effects of environment and dispersal on alien plant invasions. Perspectives in Plant Ecology, Evolution and Systematics. (in press) b Vicente, J., Randin, C.F., Pottier, J., Lomba, A., Gonçalves, J., Broennimann, O., Honrado, J. and Guisan, A. (2013). A framework for assessing the scale of influence of environmental factors on ecological patterns. Methods in Ecology and Evolution (in revision (?) c Vicente, J.R., Pinto, A.T., Araujo, M.B., Verburg, P.H. Lomba, A., Randin, C.F., Guisan, A., Honrado, J.P. (2013). Using Life Strategies to Explore the Vulnerability of Ecosystem Services to Invasion by Alien Plants. Ecosystems 16: 678–693 d Vitousek, P.M., Mooney, H.A., Lubcenco, J., Melillo, J.M. (1977). Human domination of Earth’s ecosystems. Science 277, 494-499. Walther, G., Roques, A., Hulme, P. E., Sykes, M. T., Petr Pysek, P., Kuhn, I., Zobel, M., Bacher, S., Botta-Dukat, Z., Bugmann, H., Czucz, B., Dauber, J., Hickler, T., Jarosik, V., Kenis, M., Klotz. S., Minchin, D., Moora, M., Nentwig, W., Ott, J., Panov, V. E., Reineking, B., Robinet, C., Semenchenko, V., Solarz, W., Thuiller, W., Vila, M., Vohland, K., Settele, J. (2009). Alien species in a warmer world: risks and opportunities. Trends in Ecology and Evolution Vol.24 No.12. Wang, L., Jackson, D.A. (2011). Modeling the establishment of invasive species: habitat and biotic interactions influencing the establishment of Bythotrephes longimanus. Biol Invasions 13, 2499–2512. Westphal, M.I., Browne, M., MacKinnon, K., Noble, I. (2008). The link between international trade and the global distribution of invasive alien species. Biol Invasions, 10:391–398. Internet references: CBD, Convention ON Biological Diversity (1992), Website: www.cbd.int, assessed in: 24/04/13. IUCN, the International Union for Conservation of Nature (2000), Website: www.iucn.org, assessed in: 03/05/13. Biodiversidade ameazada, Website: www.biodiversidade.eu, assessed in: 21/02/13.Delivering Alien Invasive Species Inventories for Europe, Website: http://www.europe-aliens.org/pdf/Acacia_dealbata.pdf, assessed in: 02/03!13 FCUP Using models and connectivity analysis to predict the current and future patterns of invasion by Silver-wattle (Acacia dealbata Link) in Northwest Iberian Peninsula 53 Annex Annex 1: Environmental predictors select to apply the Spearman rank correlation coefficient. Variable type Predictors Description References ANT Annual Mean Temperature Broennimann and Guisan, 2008 Climate AP Annual Precipitation Barrows and MurphyMaiscal, 2012 I Isothermality Castro-Díez et al., 2011 MDR Mean Diurnal Range Crossman and Cooke, 2011 MTCM Min Temperature of Coldest Month Evangelista et al., 2011 MTCQ Min Temperature of Coldest Qaurter Férnandez et al., 2012 ; Fischer et al., MTDQ Mean Temperature of Driest Quarter Fischer et al., 2011 MTWM Max Temperature of Warmest Month Gaikwad et al., 2011; Gallien et al., 2012 MTWaQ Mean Temperature of Warmest Quarter Godoy et al., 2008 MTWeQ Mean Temperature of Wettest Quarter Ohlemuller et al., 2006; Pino et al., 2005 PCQ Precipitation of Coldest Quarter Trethowan et al., 2011; Thuiller et al. 2005 PDM Precipitation of Driest Month Vicente et al., 2010; Vicente et al., 2011 PDQ Precipitation of Driest Quarter Vicente et al., 2013a; Vicente et al., 2013b PS Precipitation Seasonality Vicente et al., 2013c; Vicente et al., 2013d PWM Precipitation of Wettest Month Vicente et al., 2013ª PWaQ Precipitation of Warmest Quarter PWeQ Precipitation of Wettest Quarter TAR Temperature Annual Range TS Temperature Seasonality Land cover pArL % cover of arable land (per 1 km²) (Landscape pArS % cover of artificial stands (per 1 km²) Chytry et al., 2008 composition) pBiFo % cover of broad-leaf forest (per 1 km²) http://www.europe-aliens.org pCoFo % cover of conifer forest (per 1 km²) g/pdf/Acacia_dealbata.pdf pHAa % cover of heterogeneous agricultural areas (per 1 km²) Ohlemuller et al., 2006 pMiFo % cover of mix forest (per 1 km²) Pino et al., 2005 pOs % cover of open spacie with small or no vegetation (per 1 km²) Vicente et al., 2013a pPas % cover of pastures land (per 1 km²) Vicente et al., 2013d pPeCr % cover of permanent crops (per 1 km²) pSHv % cover of scrub and/or herbaceous vegetation associations (per 1 km²) Pwa % cover of water (per 1 km²) Transport DensRail Density of local rail network Catford et al. 2011 and DensRiv Density of local hydrographic network Vicente et al., 2010; Vicente et al., 2011 hydrographic DensRoad Density of local road network Vicente et al., 2013a; Vicente et al., 2013b network DistRail Distance to rails Vicente et al., 2013c; Vicente et al., 2013d DistRiv Distance to rivers DistRoad Distance to roads Soil pCamb % of cambissoils (per 1 km²) FCUP Using models and connectivity analysis to predict the current and future patterns of invasion by Silver-wattle (Acacia dealbata Link) in Northwest Iberian Peninsula 54 pFluvi % of fluvissoils (per 1 km²) pHist % of histossoils (per 1 km²) pLit % of lithossoils (per 1 km²) pLuvi % of luvissoils (per 1 km²) Vicente et al., 2013a pPodz % of podzossoils (per 1 km²) Vicente et al., 2013b pRank % of rankerssoils (per 1 km²) pRego % of regossoils (per 1 km²) Lanscape Ed Edge Density Structure GPP Mean gross annual primary productivity Lomba et al., 2010 MPAR Mean perimeter-area ratio Vicente et al., 2010 MPFD Mean patch fractal dimensions Vicente et al., 2013a; Vicente et al., 2013b MPS Mean patch size Vicente et al., 2013c; Vicente et al., 2013d MSI Mean shape index Vicente et al., 2013ª NumP Number of patches PSSD Patch size standart deviation TE Total Edge