UNIVERSIDADE ESTADUAL PAULISTA “JÚLIO DE MESQUITA FILHO” INSTITUTO DE BIOCIÊNCIAS – RIO CLARO unesp PROGRAMA DE PÓS-GRADUAÇÃO EM ECOLOGIA E BIODIVERSIDADE SPATIAL INSIGHTS FOR MANAGEMENT AND CONSERVATION OF ECOSYSTEM SERVICES AND BIODIVERSITY FOR A SUSTAINABLE FUTURE JULIA CAMARA DE ASSIS Fevereiro - 2020 JULIA CAMARA DE ASSIS SPATIAL INSIGHTS FOR MANAGEMENT AND CONSERVATION OF ECOSYSTEM SERVICES AND BIODIVERSITY FOR A SUSTAINABLE FUTURE Orientador: Prof. Dr. Milton Cezar Ribeiro Fevereiro, 2020 Tese apresentada ao Instituto de Biociências do Câmpus de Rio Claro, Universidade Estadual Paulista, como parte dos requisitos para obtenção do título de doutora em Ecologia e Biodiversidade. A848s Assis, Julia Camara Spatial insights for management and conservation of ecosystem services and biodiversity for a sustainable future / Julia Camara Assis. -- Rio Claro, 2020 131 p. : il., tabs. Tese (doutorado) - Universidade Estadual Paulista (Unesp), Instituto de Biociências, Rio Claro Orientador: Milton Cezar Ribeiro 1. Serviços ambientais. 2. Biodiversidade Conservação. 3. Sustentabilidade. I. Título. Sistema de geração automática de fichas catalográficas da Unesp. Biblioteca do Instituto de Biociências, Rio Claro. Dados fornecidos pelo autor(a). Essa ficha não pode ser modificada. To past, current and future generations. Ao meu pai e à minha mãe, à minha irmã e ao meu irmão, aos meus dois preciosos sobrinhos. ACKNOWLEDGEMENTS This study was financed in part by the Coordenação de Aperfeiçoamento de Pessoal de Nível Superior-Brasil (CAPES) - Finance Code 001. My PhD journey relied on the support of numerous agencies and individuals. Here is the space I try to acknowledge them all. I’ll do my best to be fair. I thank CAPES/NUFFIC Project 048/13 for the sandwich scholarship, that provided me with a remarkable experience abroad. Special thanks to Larissa Barreto and Miltinho, for choosing me. But also, thanks to Lars Hein who accepted to host me at Wageningen University and Research in the Netherlands. Thanks to Fundação de Amparo à Pesquisa do Estado de São Paulo – FAPESP (processo nº 2013/23457-6) and PROAP for funding part of my conferences expenses, where I could share my research results with my peers and enlarge my network. This included my participation in the Latin American ESP in Campinas/2018, the ESP10 World Conference in Germany/2019 and the 29th ICCB in Malaysia/2019. Thanks to the fundamental parental funding that allowed my persistence during the first 18 months of my PhD. Without my parents financial (and emotional) support, I would have given up that first year. Thanks to UNESP - Rio Claro and the PPG Ecologia e Biodiversidade. From Ivana to all the other administrative employees. Special thanks to all professors and teachers I met over the years. Thanks to the Department of Ecology, for embracing me. To Cris and the kind people I used to meet everyday and who kept things working, from lab and field technicians, to the cleaning ladies, the gardeners and the security crew. Thanks to Jean Paul Metzger and LEPaC for the opportunities to interact and engage in amazing scientific debate. Thanks for the invitation to participate the Landscape Workshop in April/2018 where I met the Australian team and had the excellent opportunity to be in charge of the spatial theoretical paper. It was a great pleasure to work with Camila Hohlenwerger, Larissa Boesing and Paula Prist. Also, a special thanks to Jonathan Rhodes from University of Queensland. Thanks to Lars Hein and the ESA team from Wageningen University and Research, for such a productive exchange in the Netherlands. I learned a lot from you Lars, you gave me the fantastic opportunity to improve my meeting skills and become a more straightforward scientist. Thank you, ESA team, for all the coffee breaks (with or without cake), the borrels and social events, specially Sinterklaas (where I learned a lot about Dutch festivities) and that end-of-the-year-gathering at Dolf’s place (even though I hate saté, Eka managed to provide delicious alternative dish for me). Thank you, ESA PhD colleagues, for the lunch times and plenty of cappuccinos together at Orion, for the potluck dinners and all the conversation and support from each of you. It was my greatest pleasure to engage all of you into a more pleasing and not-so-isolated journey (this is how it felt for me). Being in the Netherlands meant a lot not only to my development as a researcher, but as a human being in this planet. Thanks to my WUR friends Danny, Beatriz, Lucas and Trinity, who welcomed me in Asserpark. Thanks to my flatmates Dogan, Thomas, Marco and Maria. Thanks to my Toastmasters pals, who helped me evolve as a public speaker, specially Zsolt, Eva and Ruben. Thanks to all Ecuadorian and other Spanish-speaking friends, Giovanny, Gabriel, Luis, Cristina, Lorena. Thanks to my Brazilian friends Nathalia and Daniela, along with Julia, Olivia and Gabriela. Nat, thanks for sharing the “peroba oil” and so many memorable experiences and trips! Dani, thanks for embracing me and being the safest harbor. Thanks to my ESA colleagues Gabriela, Lena, Anna, Jerry, Leonardo. Special thanks to the friendship of Lucy. Eka, your presence made all workdays more pleasant; you are adorable. Sarahi, not enough words to thank you for your hospitality, I’ll never forget how I became ‘your responsibility’; it felt like instant friendship, I carry you in my heart. Confidence, thank you for letting me in, it meant a world. Shahid, you brought diversity to my life; you are a joyful blessed soul. Maddy, our friendship feels like home; I’m sure we have much to share in life yet, hope to meet you soon. Also, thanks for the reunion in November/2019. It was a great motivational push meeting most of you again. I hope we meet again soon enough. Thanks to LEEC, the apple of my eyes. Its dynamic and collaborative environment along with the multitude of personalities that were part of it through my years in Rio Claro provided me with the most pleasing place in the planet to work at. It was an honor to be part of the team. There, whenever I felt lost, I knew I was never alone. Thanks to my dear friends Giordano Ciocheti and Ricardo Sartorello who showed me how to persist and survive the last stretch of my PhD, even before it had begun. Thanks to my precious students: those from random classes, those from the Agroecosystem course (little did they know they were my personal experiment), and to Fernanda Saude (for teaching me how to guide), Isabela Cassimiro (for dedication and sweetness), Luiggi Galbiate (for the long talks and perseverance) and Francielli Carvalho (for taking in my eagerness and setting the pace). You taught me a lot about myself and my wishes for the future. Thanks to my EcoSysters and the powerpuff girls’ analogy. It makes so much sense! Gabi and Rafa, I believe that our encounter was the most fortunate interference of the universe and because of that our PhD journeys were three times better. You gave me courage and backed me up from that meeting on (and imagine some of us didn’t like each other at first…). You make me believe teamwork is the way to go and cooperation is the solution to most problems. Thanks to my LEEC/Rio Claro friends that each, on their own way, supported me, accompanied me, learned with me and helped me out (in a random order because there is no way they could possibly be compared in importance): Renata for sharing our home and life for almost 6 years; Flavia for your experience and wisdom along with your practical and straightforward view of life; Camila for your cuteness, blossoming and unconditional friendship and companionship; Milene for showing the way ahead and learning together about our strengths and weaknesses; Paula for opening up and teaching how important it is to rely on each other, from our struggles abroad to our calm LEEC days; Natalia for giving a sweet-sour delicious flavour to our very calm routine in Rio Claro; Julia Oshima for easing my SDM challenges and being the best example of kindness and professionalism; Patricia for reading the cards, sharing your sister and bringing your unique profile to this crazy academia environment; Vanesa for cultural exchange and disguised kindness, I love to have gone through distinct parts of our journeys together; Claudia for not saying much but showing how to be strong and go for it, with all your quiet accomplishments; Ana for gaining space little by little, and allowing me to get close enough to know you better, you certainly pushed me forward; João for being my best match (and Felipe along too), our reciprocity is unique; Lucas for the sweetest presence and constancy; Andre “fii” for the comradeship and one-to-one conversations about life; Bruno Borges for sharing feelings, doubts and space; Andre Regolin for the happy hours; Andre Tavares for the help and assurance when I most needed a push; Bernardo for teaching that everything is going to be okay; Erison for sharing the walk and the struggle till the last minute; Urucum for the multi-task and know- it-all skills; Ricardo Cerri for the deep conversations on the future of our careers; Matavelli for all the coffee breaks and unburdening talks; Blanca and Ruben (los 2 Ruben), for bringing the Rumba Latina into those joyful parties; José Fernando and Federico for being my favorite dance partners in Rio Claro; Thais Arruda for saving me from my stressed body and soul. Anyway, there were so many beings… Felipe Martello, Thadeu Sobral, Maurício Vancine, Annia Susin, Andreia Magro, Fábio Barros, Vinicius Tonetti, Karlla Barbosa, Fernandinha, Cleber e Barbara, Paloma Marques, Marina Giubbina, Iara Lopes, Gabi Rezende, Dani Ramos, Larissa Sayuri, José Wagner, Helena Prado, Marcela Barbosa, Berger Medeiros, Diego Braga, Victor Krepski, Laura Honda, Fernando Lima, Juliana Silveira, Calebe Mendes, Pavel, Paola Tokumoto, Kalinka, Ju Coelho... So many beings that were part of the team at some point. I thank you all for crossing my life. Also, thanks to so many others who crossed my path along this journey here and elsewhere. It’s unfortunate that most of us leave this (best) part to the end; my memory is not good enough to recall all the important people that crossed my way and walked along with me even for a minute… Don’t take it personal if your name is not here. Special thanks to my supervisor, Miltinho. You bewitched me and took me to Neverland. I learned how not to fear when taking daring steps, when to say no and when to be patient. I also understood how huge the challenges are ahead. Thanks for “family first”. I don’t know what exactly you taught me, and I may never, but the most valuable lesson I take from our master-apprentice experience is how to open doors: be it waiting for the right time, knocking, or with both feet in a flying kick. Special gratitude to two of my non- Rio Claro and non- academic friends, who brought meaning to my life outside my PhD: Fefe and Manoel. Even though our meetings were not as frequent as I wish, they were meaningful. There are not enough words to properly thank my family for all the support over these years. I am very grateful to both my parents. My mother, Fernanda, for her unconditional love and emotional support; and my father, Antonio, for understanding my endeavor from the perspective of someone who has gone through this path (even if under very different circumstances and with distinct goals). My love for you both has given me strength to overcome all the challenges. My dearest sister and friend, Marilia, who has become my idol and example of how to succeed in life, working hard after enjoying all the good share (or is it otherwise?). My mysteriously beloved brother, Guilherme, whose lack of words prevents me from learning more, but at the same time provides me with daily opportunities to learn how to respect our differences while searching for a common ground. To my two most loved nephews, Rafael and Felipe, and their mother, Karina, for the gift of their loving existence. To all my other relatives, who taught me the valuable links we share are my most valuable treasures, I thank you all. “in order to change an existing imagined order, we must first believe in an alternative imagined order.” Yuval Noah Harari Sapiens: A Brief History of Humankind, p. 118 RESUMO INFORMAÇÕES ESPACIAIS PARA GESTÃO E CONSERVAÇÃO DOS SERVIÇOS ECOSSISTÊMICOS E DA BIODIVERSIDADE PARA UM FUTURO SUSTENTÁVEL Compreender como os padrões e processos da paisagem afetam a biodiversidade e os serviços ecossistêmicos (ES) – benefícios que as pessoas recebem da natureza – é fundamental para prever suas respostas às mudanças no uso da terra. Para avançar na pesquisa, desenvolver e fornecer informações espaciais apropriadas para as tomadas de decisão, utilizei modelagem e proposições teóricas para expandir o conhecimento sobre serviços ecossistêmicos, biodiversidade e o papel desempenhado pelos seres humanos na conservação da paisagem e seu manejo sustentável. No capítulo 1, considerando a teoria atual e as lacunas de conhecimento acerca de ES, eu propus um framework para entender melhor os efeitos da estrutura da paisagem no fluxo de ES, um dos componentes da cadeia de provisão de ES. Com base na literatura, discuti os efeitos da composição e configuração da paisagem nos fluxos de ES, de acordo com seus mecanismos. Enfatizei a necessidade de caracterizar a oferta, o fluxo e a demanda de cada ES, a fim de quantificar e abordar completamente os aspectos espaciais da provisão de ES. No Capítulo 2, identifiquei hotspots de biodiversidade usando Modelos de Distribuição de Espécies (SDM) baseados em clima e paisagem para indicar áreas prioritárias para conservação e aquelas ameaçadas por mudanças rápidas no uso da terra, impulsionadas pelo desmatamento e expansão agrícola. Os hotspots de biodiversidade representam as áreas com potencial para hospedar várias espécies-chave e, portanto, estas são regiões-chave para serem protegidas. A seleção das espécies e a interpretação dos resultados dos modelos foram realizadas para o Maranhão (um estado brasileiro), enquanto a coleta e modelagem dos dados foram realizadas para um limite ecológico mais amplo. Indiquei as áreas com maior sobreposição de hotspots de espécies e não abrangidas pelas áreas protegidas existentes como prioritárias para ações de conservação. Reconheci a necessidade urgente de monitorar áreas com expansão agrícola. Por fim, concluí a tese com uma síntese das principais contribuições, recomendações para futuras pesquisas e uma reflexão sobre as lições aprendidas durante o desenvolvimento do meu doutorado. PALAVRAS-CHAVE: estrutura da paisagem; fluxo de serviço ecossistêmico; área prioritária para conservação; modelo de distribuição de espécie; Maranhão; sustentabilidade ABSTRACT SPATIAL INSIGHTS FOR MANAGEMENT AND CONSERVATION OF ECOSYSTEM SERVICES AND BIODIVERSITY FOR A SUSTAINABLE FUTURE Understanding how landscape patterns and processes affect biodiversity and ecosystem services (ES) – the benefits people receive from nature – is fundamental to predict their responses to land use change. To advance research, develop, and deliver appropriate spatial information for decision-making, I used models and theory to expand the knowledge about ecosystem services, biodiversity and the role played by humans in landscape conservation and sustainable management. In Chapter 1, taking into account the current theory and knowledge gaps related to ES, I proposed a framework to better understand the effects of landscape structure on the flow component of the ES supply chain. Based in the literature, I discussed the effects of landscape composition and configuration in the ES flows, according to their mechanisms. I emphasized the need to characterize the supply, flow and demand of each ES in order to quantify and thoroughly address the spatial aspects of ES provision. In Chapter 2, I identified biodiversity hotspots using climate- and landscape-based Species Distribution Models (SDM) to indicate priority areas for conservation and those areas threatened by fast land use change driven by deforestation and agriculture expansion. Biodiversity hotspots allowed the recognition of areas with the potential to host several key species and, therefore, those are key regions to be protected. The selection of species and the interpretation of model outputs were conducted for Maranhão, a state in Brazil, while data collection and modelling itself were carried out for a broader ecological boundary. I indicated the priority areas to be conserved as those with higher overlap of species hotspots and not encompassed by existing protected areas. I recognize urgent need to monitor areas with agriculture expansion. Lastly, I concluded the thesis with a synthesis of the main contributions, recommendations for further research and a reflection on the lessons learned during the development of my PhD. KEYWORDS: landscape structure; ecosystem service flow; priority area for conservation; Species Distribution Model; Maranhão; sustainability RESÚMEN INFORMACIÓN ESPACIAL PARA GESTIÓN Y CONSERVACIÓN DE LOS SERVICIOS DE ECOSISTEMAS Y DE LA BIODIVERSIDAD PARA UN FUTURO SOSTENIBLE Comprender cómo los patrones y procesos del paisaje afectan la biodiversidad y los servicios del ecosistema (ES) – los beneficios que las personas reciben de la naturaleza – es fundamental para predecir sus respuestas al cambio en el uso de la tierra. Para avanzar en la investigación, desarrollar y entregar información espacial apropiada para la toma de decisiones, utilicé modelos y teoría para expandir el conocimiento sobre los ES, la biodiversidad y el papel que juegan los humanos en la conservación del paisaje y su gestión sostenible. En el Capítulo 1, teniendo en cuenta la teoría actual y las brechas de conocimiento relacionadas con los ES, propuse un marco para comprender mejor los efectos de la estructura del paisaje en el componente de flujo de la cadena de provisión de ES. Con base en la literatura, discutí los efectos de la composición y configuración del paisaje en los flujos de ES, de acuerdo con sus mecanismos. Destaqué la necesidad de caracterizar la oferta, el flujo y la demanda de cada ES para cuantificar y abordar a fondo los aspectos espaciales de la provisión de ES. En el Capítulo 2, identifiqué los puntos críticos de biodiversidad utilizando los modelos de distribución de especies (SDM) basados en el clima y el paisaje para indicar áreas prioritarias para la conservación y aquellas áreas amenazadas por el cambio rápido del uso de la tierra impulsado por la deforestación y la expansión de la agricultura. Los puntos críticos de biodiversidad permitieron reconocer las áreas con el potencial de albergar varias especies clave y, por lo tanto, las regiones clave a proteger. La selección de especies y la interpretación de los resultados del modelo se realizaron para el estado brasileño de Maranhão, mientras que la recolección de datos y el modelado en sí se llevaron a cabo para un límite ecológico más amplio. Indiqué las áreas prioritarias a ser conservadas como aquellas con mayor superposición de puntos críticos de especies y no abarcadas por áreas protegidas existentes. Reconocí la necesidad urgente de monitorear áreas con expansión agrícola. Finalmente, concluí la tesis con una síntesis de las principales contribuciones, recomendaciones para futuras investigaciones y una reflexión sobre las lecciones aprendidas durante el desarrollo de mi doctorado. PALABRAS CLAVE: estructura del paisaje; flujo de servicio ecosistémico; área prioritaria para la conservación; modelo de distribución de especie; Maranhão; sustentabilidad SUMMARY BACKGROUND ................................................................................................................................................ 23 MOVING THE GEAR AND PUSHING THE KNOWLEDGE FRONTIER ......................................... 29 THESIS OUTLINE........................................................................................................................................... 33 CHAPTER 1 – A CONCEPTUAL FRAMEWORK TO LINK LANDSCAPE STRUCTURE AND ECOSYSTEM SERVICES ............................................................................................................................... 35 ABSTRACT ....................................................................................................................................................... 36 CONTEXT .......................................................................................................................................................... 37 1. INTRODUCTION .................................................................................................................................... 38 2. DECOMPOSING THE ECOSYSTEM SERVICE CONCEPT ......................................................... 38 3. HOW TO LINK LANDSCAPE STRUCTURE TO ECOSYSTEM SERVICES? .......................... 40 4. ES FLOW MODULATED BY LANDSCAPE STRUCTURE .......................................................... 43 5. THE SPATIAL RELATIONSHIP BETWEEN ECOSYSTEM SERVICE COMPONENTS ..... 45 6. CHALLENGES AND RESEARCH AGENDA .................................................................................... 50 GLOSSARY ........................................................................................................................................................ 54 CHAPTER 2 – IDENTIFYING HOTSPOTS FOR BIODIVERSITY CONSERVATION: FROM ECOLOGICAL TO ADMINISTRATIVE BOUNDARIES ........................................................................ 59 ABSTRACT ....................................................................................................................................................... 61 1. INTRODUCTION .................................................................................................................................... 62 2. MATERIAL AND METHOD ................................................................................................................ 65 3. RESULTS .................................................................................................................................................. 72 4. DISCUSSION ............................................................................................................................................ 78 5. CONCLUSION ......................................................................................................................................... 82 SUPPLEMENTARY MATERIAL ................................................................................................................. 83 SYNTHESIS AND RECOMMENDATIONS .............................................................................................. 97 REFERENCES ............................................................................................................................................... 107 APPENDIX 1 ................................................................................................................................................. 123 ADDITIONAL ACTIVITIES ....................................................................................................................... 125 LIST OF PUBLICATIONS .......................................................................................................................... 129 ABOUT THE AUTHOR ............................................................................................................................... 131 23 BACKGROUND Humanity is a pervasive species that managed to reach every corner of the planet. Even if there were hidden, or unexplored venues, if any of us acknowledges its existence, one could argue it does not exist at all. Nature and Conservation Conservation is a human action – be it to protect nature from humans or to spare it for humans themselves. Mace (2014) provoked this debate by showing the changes over time of conservation goals and the science underpinning them. According to her, for some decades, conservation science was targeted at sparing pristine natural assets from human interference. The Anthropocene has come to shift that framing into a more integrative one, in which people are viewed as part of nature. Thus, science has had to dive into human-nature relationships and envision their sustainable co-existence (Johnson et al., 2017). The complexity of social-ecological systems, once we frame people and nature together (i.e. a biocentric worldview), demands further collaboration between social and natural sciences (Bennet et al. 2015). So far, sustainability has evolved as the most promising idea to guide us into the future. Sustainability Science has managed to evolve the concept of sustainability since the 1980’s (Sartori et al., 2014), however, its conversion into practice is yet to be broadly incorporated by humanity (Purvis et al., 2019). Strong sustainability argues that our “finite planet cannot sustain human life with an economy that intends unlimited growth” (Morandín-Ahuerma et al., 2019). Also, sustainability relies on consumption behavior that do not deteriorate ecosystems and leads to social equity (Morandín-Ahuerma et al., 2019). Having economic values over culture is driving humanity past planetary boundaries. In great effort to reverse this trend and rethink human activities, the Sustainable Development Goals (SDG) were proposed to guide human actions towards a sustainable future (UN, 2016), addressing human population growth, the role of economy in serving us (and not otherwise), greenhouse gases emissions, pollution, biodiversity loss and the spread of invasive alien species, among several other goals (Griggs et al., 2013; Morandín- Ahuerma et al., 2019; Ripple et al., 2017). Changing and adjusting our societal, institutional and individual behaviour towards sustainable habits is the way forward to guarantee nature’s limits are not surpassed (Griggs et al., 2013; Rockström et al., 2009). 24 Science Goals, Economic ‘Rules’ and Political Decisions If we could evaluate the positions and responsibilities assumed by scientists, economists and politicians in human societies, we could find that scientists are usually engaged with facts and knowledge derived from sound scientific analysis (Bradford, 2017); while economists are mostly driven by people’s behaviour and choices in different contexts (Guerry et al., 2015); and politicians would search for ways to please those who have the power to keep them in power and prioritize decisions based on public opinion (i.e. populism; Dal Bó et al., 2017). Who rules the world? I wish the answer was “Girls”, but we are not there yet. Instead, most nations have political leaders that respond to international markets and aim at economic growth. Rarely, governments main goals are to promote equity and pay off historical debts towards unprivileged groups. Even more unexpected is to have politicians that fight historical privileges. And even when parts of society disagree with politicians’ behaviour, very little can they do to demand changes. Recently, however, participatory approaches have emerged to give voice to different groups interested in a matter so that decisions and plans could address diverse interests (Arnstein, 2019). Yet, most decisions still rely on economic profit over environmental viability, and the two overrule social justice. Building a dam that harms indigenous communities, building a road that passes over a low-income neighborhood, undertaking new mineral exploitation and annulling a protected area for that: these are political decisions, pushed by economic pressure from private companies in detriment of less privileged communities. This is how the world works. Should scientists become politicians, should economists remember they are social scientists, or should politicians become extinct? Integration among social, environmental and economic sciences are leading the way towards better decision-making (Guerry et al., 2015). From the science perspective, bridging the gap between science and society and the gap between theory and practice can put humanity on the sustainability path (Griggs et al., 2013). The role of scientists in overcoming the gap between science and policy has usually been performed as translating scientific discoveries to decision makers. Decades have passed to prove this is not enough. Communicating scientific advances to the general public and discussing the social implications of research outcomes has recently begun to be one of the scientist’s duties (Dickson, 2010). Scientists have the power to go further 25 and have stakeholders be part of the construction of knowledge since the beginning: from the conception of a research project to its closure (Elena M Bennett et al., 2015). Therefore, engaging with on the ground practitioners, NGOs, communities, and any other groups of interest has been explored through participatory approaches (Angelstam et al., 2019a) and the co-production of knowledge “to ensure that interventions and policies have appropriate impact and can operate across multiple temporal and spatial scales” (Elena M. Bennett et al., 2015). The Ecosystem Service Framework The concept of Ecosystem Services (ES) emerged in line with the concept of strong sustainability, which primarily claims the non-substitutability of natural capital (Ekins et al., 2003), and the creation of a shared vision of what a sustainable society would actually look like (Costanza et al., 1997). It derived from Westman’s (1977) idea of nature’s services and was later coined as ES by Ehrlich and Ehrlich (1981). Despite some variation in the concept, ES has drawn attention to the benefits that ecosystems generate for humans (Costanza et al., 1997; Daily et al., 1997; De Groot et al., 2002). Such benefits, that sustain and fulfill human life, result from the interactions among living and non-living elements of the ecosystems and human-engineered components of social-ecological systems (Daily et al., 1997; Guerry et al., 2015). More than that, the concept of ES “provides a valuable framework to define and analyse linkages and dependencies between natural and human systems” (Burkhard et al., 2010). Biodiversity plays a pivotal role for ecosystem functions maintenance, and therefore, for ES provision (Harrison et al., 2014). Boosted by the relevance of biodiversity and ES for the human wellbeing, some notable initiatives spread out to society the outlook of the challenging paradigm of sustainability: the Millennium Ecosystem Assessment (MA, 2005), The Economics of Ecosystems and Biodiversity (TEEB, 2010) and the Intergovernmental Platform on Biodiversity and Ecosystem Services (IPBES, 2012; see Box 1 for more information). All of them converged in the combination of economy, ecology and society for the sustainable stewardship of natural capital (Angelstam et al., 2019a). The integration of participatory methods to improve ES frameworks, in a variety of social contexts, has proven to enhance social learning, build capacity and trust as well as work as a tool to mediate power relations (Davies et al., 2015). In the local level, it can offer an alternative source of income via Payment for Ecosystem Services(PES; Wang et 26 al., 2017). Another advantage in the use of ES frameworks is that they can be used to evaluate progress towards sustainability (to monitor the SDGs and the Aichi Targets; CBD, 2010), by monitoring the state and trends of ES provision and biodiversity (Geijzendorffer et al., 2017). Box 1: Main international initiatives that spread out to society the challenges in the pursuit of sustainability. MA – Millennium Ecosystem Assessment (MA, 2005) The first comprehensive global assessment of the implications of ecosystem change for human wellbeing. It can be also considered the scientific basis for action needed to enhance the conservation and sustainable use of those systems and their contribution to people. ES are grouped into a) provision (e.g., food and water supply), b) regulation (e.g., climate regulation), c) cultural (e.g., recreational, spiritual activities), and d) support (e.g., photosynthesis). All these services performed for free by ecosystems are widely assumed to contribute to poverty alleviation and the degradation of these services is also assumed to result in negative effects on human wellbeing (Tallis et al., 2008). In this sense, a fully developed economics should routinely incorporate the value of ES into its analyses (Chapin et al., 2010; Polasky, 2012). TEEB – The Economics of Ecosystems and Biodiversity (TEEB, 2010) Five years after the MA a new global initiative emerged, TEEB, which focused on “making nature’s values visible”. Its principal objective is to mainstream the values of biodiversity and ES into decision-making at all levels. It aims to achieve this goal by following a structured approach to valuation that helps decision- makers recognize the wide range of benefits provided by ecosystems and biodiversity, demonstrate their values in economic terms and, where appropriate, suggest how to capture those values in decision-making. IPBES – Intergovernmental Platform on Biodiversity and Ecosystem Services (IPBES, 2012) Despite the complexity of ES identification, quantification and valuation advances in data acquisition, methods and technology regarding ES approach have been consistent since the MA (Guerry et al., 2015). The next step requires the inclusion of such advances in the implementation of action plans, management and policies. The IPBES was proposed as a mechanism to further strengthen the science-policy interface on biodiversity and ES. IPBES contributes to existing processes that ensure that decisions are made on the basis of the best available scientific information on conservation, sustainable use of biodiversity and ES. It was established in 2012 as an independent intergovernmental body open to all member countries of the United Nations. The members are committed to building IPBES as the leading intergovernmental body for assessing the state of the planet's biodiversity, its ecosystems and the essential services they provide to society. More recently, they incorporated the concept of Nature’s Contributions to People (NCP) to broaden perspectives about the relatioships between people and nature and include different worldviews (Díaz et al., 2018). 27 Spatial Planning Matters Spatial planning provides the basis for the development and implementation of good governance by integrating legal, economic, social, and rights-based instruments (Albert et al., 2020). It provides the on-the-ground link for economic, social and environmental policies. Furthermore, the use of spatial instruments in decision making across scales emerges as a powerful tool to balance diverse interests (Angelstam et al., 2019a) and promote the engagement of stakeholders in planning decisions (Albert et al., 2020). One of the challenges highlighted by Albers et al. (2020) to target biodiversity and ES in spatial planning is having spatial information in appropriate resolution for multiple scales (i.e. different spatial resolution and extents of influence; Jackson and Fahrig, 2015). Besides, understanding how landscape patterns and processes affects biodiversity and ES is key to predict their responses to land use change and deliver appropriate spatial information (Jones et al., 2013). The inclusion of ES and biodiversity as targets in spatial planning requires not only tools capable of analysing trade-offs (Jones et al., 2013) and dealing with power asymmetries (Reed et al., 2017), but also a deeper understanding of how landscape patterns affects both of them (Jones et al., 2013). Ruckelshaus et al. (2015) listed some lessons learned from 20 cases of application of biodiversity and ES information in decision-making. Among them, they recognized that creating a mutual understanding of ES and biodiversity-related information in an iterative science-policy process increases credibility and legitimacy of information to support decisions, despite being time consuming (Ruckelshaus et al., 2015). They also indicate that engaging with stakeholders to know what they need and empowering local experts enhance credibility and ownership of ES models and information resulting in effective contribution for the decision-making process (Ruckelshaus et al., 2015). 28 Keep pushing. Matt Might1 The Illustrated Guide to a Ph.D.2 1 http://matt.might.net/ Accessed in Jan/2020. 2 http://matt.might.net/articles/phd-school-in-pictures/ Accessed in Jan/2020. 29 MOVING THE GEAR AND PUSHING THE KNOWLEDGE FRONTIER A conceptual framework to link landscape structure and ecosystem services Contributions from Chapter 1 Economic demands and society’s desires are drivers of land use modification with consequences to environmental conditions. Although there is a strong relation between land cover and ecosystems capacity to provide ES, some services may not present such a straightforward relationship, depending more on the condition and spatial pattern of land cover or even on social and economic configuration (Remme et al., 2014). In those cases, different data types – such as ecosystem condition and produced capital or human labor – can be combined to better represent ES in a given area (Hein, 2010). The ES supply chain encompasses the components and processes of the provision of ES, which combines the supply (ecosystems assets) and the demand (people’s needs) through flow (the ‘delivery’ of the service that is converted into benefit). Even though measuring and quantifying ES has been an intensively developing field (Willemen et al., 2015), further research is necessary to advance our understanding on how landscape structure affects the ES supply chain. Compelled by Michell et al. (2015) and their proposition on how landscape fragmentation per se (sensu Fahrig, 2003) can influence ES supply and flow, the purpose of Chapter 1 is to further explore how other landscape structure attributes (i.e. composition and configuration) may affect ES flow. Based on the literature and much discussion with my collaborators, I developed a theoretical framework hypothesising on the effects of landscape structure on ES flow. For this, we considered both the proportion of supply and demand within a landscape, and how ES flow mechanisms may work. Among the potential contributions that may emerge from Chapter 1 are: 1) provoking the debate on how to address landscape influences on ES flow in future research; 2) building consensus regarding the components of the ES supply chain; and 3) advancing the understanding of landscape effects on them to assist spatial planning. 30 Identifying hotspots for biodiversity conservation: from ecological to administrative boundaries Contributions from Chapter 2 The expansion of agricultural land and other human activities (e.g., mining, transport infrastructure and urbanization) in developing countries has been positively reflected in economic growth. On the other hand, fast modification of landscapes causes ecosystem degradation that changes ecosystem functioning and their capability of generating services (Hein et al., 2016) that benefit human beings. Most tropical regions worldwide have gone through rapid land use change and biodiversity loss (Alroy, 2017). Within the current institutional decision-making framework, mainly based on economic profits, ecosystems and biodiversity are undervalued (UNEP-WCMC, 2015). Pushing the economic paradigm to the edge, the United Nations, along with other organizations, have developed the System of Environmental Economic Accounting (SEEA). The SEEA has been tested worldwide under the Experimental Ecosystem Accounting to develop and test measures of environmental assets, ecosystem stocks and services. In the future, the SEEA will be coupled with the System of National Accounts (SNA), that measures the economic activity in an integrated set of macroeconomic accounts. Ultimately, the goal is to shift the current economic paradigm that does not consider the degradation of environmental resources towards a different perspective that combines economic indicators with environmental ones (for further information on the SEEA EEA see Appendix 1). To accomplish this challenge, governments have been encouraged to develop the thematic accounts, which includes separately accounting for: air emissions; agriculture, forestry and fisheries; energy; water3; land use; ecosystem; biodiversity; and others. The proposition of Chapter 2 emerged to further develop the Biodiversity Account in tropical regions. The use of up to date modelling techniques to spatially quantify species diversity can be applied over time to set a baseline (using the climatic niche combined with ‘original’ land cover) and estimate changes over time (according to variation in land use). 3 The Brazilian Institute of Geography and Statistics (IBGE) has developed the SEEA for water, published in 2018, referring to the period between 2013 to 2015. Access in: Jan/2020. Available at: < https://www.ibge.gov.br/estatisticas/economicas/contas-nacionais/20207-contas-economicas- ambientais-da-agua-brasil.html?=&t=o-que-e>. https://www.ibge.gov.br/estatisticas/economicas/contas-nacionais/20207-contas-economicas-ambientais-da-agua-brasil.html?=&t=o-que-e https://www.ibge.gov.br/estatisticas/economicas/contas-nacionais/20207-contas-economicas-ambientais-da-agua-brasil.html?=&t=o-que-e 31 Even though the use of spatial information about biodiversity to prioritize areas for conservation is not new, it is still necessary in regions that do not have this type of information. The contributions of Chapter 2 are: 1) the use of a combined modelling approach to inform priority areas for conservation based on the potential spatial allocation of biodiversity hotspots; 2) the coupling of ecological and administrative boundaries to such approach; 3) the potential collaboration among experts to build up species databases for regions under intense land use change; and 4) the further development of the biodiversity accounts in the scope of the SEEA EEA. 32 33 THESIS OUTLINE To develop my research, I used models and theory to expand the knowledge about ES, biodiversity and the role played by humans in landscape conservation and sustainable management. As I adventured myself through the ES field, I realized the complexity of human-nature related approaches and the two chapters presented here reflect my interdisciplinary nature. I was pushed by the constant desire of contributing to environmental decisions and management aimed at sustainable land use. In the ES framework, I recognized a fertile field for the integration of environmental, social and economic perspectives. I also realized the challenge presented by the complexity of this framework in bridging the gap between scientists and practitioners. Chapter 1 initiated as a collective construction, for which I took the lead and based on much discussion, organized a theoretical proposition for understanding the effects of landscape structure on the flow component of the ES supply chain. Based in the literature, we discussed the effects of landscape composition and configuration in the ES flow. We emphasized the need to characterize the supply, flow and demand of each ES in order to quantify and thoroughly address how landscape structure can shape the ES flow. Chapter 2 was developed to spatially address the status of biodiversity using Species Distribution Models (SDM) and the EcoLand approach, that combines climate- and landscape-based SDM. We adopted a broader modelling area based on ecological characteristics of species (to avoid model overfitting) and analysed the outputs for a more restrict spatial administrative boundary at the state level. We combined high values of the EcoLand as hotspots for several species and evaluated their utility in the identification of priority and threatened areas in face of deforestation and agriculture expansion. After the two chapters, there is a closing section, in which I present a synthesis, recommendations and a personal reflection about the PhD challenges. Lastly, I present the additional activities I developed during my PhD, followed by my list of publications and a brief story about myself. 34 35 CHAPTER 1 – A CONCEPTUAL FRAMEWORK TO LINK LANDSCAPE STRUCTURE AND ECOSYSTEM SERVICES Julia Camara Assis, Rafaela Aparecida Silva, Gabriela Teixeira Duarte, André Luiz Batista Tavares, Milton Cezar Ribeiro Spatial Ecology and Conservation Lab, UNESP (LEEC), UNESP- Rio Claro, São Paulo, Brazil Camila Hohlenwerger, Andrea Larissa Boesing, Paula R. Prist, Eduarda Romanini, Julia Rodrigues Barreto, Jean Paul Metzger Department of Ecology, Institute of Biosciences, University of São Paulo, SP, Brazil Martine Maron, Jonathan R Rhodes School of Earth and Environmental Sciences, The University of Queensland, Brisbane, Australia 36 ABSTRACT The spatial arrangement of supply and demand along with landscape composition and configuration can determine the flow of ES. Therefore, understanding the effects of landscape structure in the components of the ES supply chain is fundamental to promote adequate landscape management that can, at the same time, guarantee sustainable provision of ES and the maintenance of ecosystem functions and biodiversity. In this research we investigate how ES flow can vary according to landscape structure, being facilitated or prevented to occur depending on the amount and spatial arrangement of ES supply and demand, as well as other landscape features. We addressed, besides habitat fragmentation per se, other landscape attributes – such as isolation, network complexity and matrix resistance – modulate the ES flow. Moreover, we postulated that the spatial arrangement of supply and demand areas determines the intensity of flow and whether maximum ES flow is attainable. We assume that certain landscape attributes will influence ES flow positively and others negatively, depending on the ES evaluated. Highly fragmented landscapes are expected to facilitate flow to a certain degree for pollinators, shaping pollination at local and landscape levels. Additionally, increasing isolation between supply and demand areas will decrease flow of pest control. Complementarily, network complexity may facilitate flow not depending on distance, but on existing connections, as is the case of outdoor recreation, whereas anthropogenic matrix resistance may be the limiting aspect in other cases. However, supply and demand can be overlapped in space and other elements may occur in the landscape, also interfering in ES flow. Thorough assessments of ES must account for the ES supply chain as a whole and further understand flow mechanisms in order to evaluate how landscape management can effectively optimize ES provision. Keywords: Ecosystem Service Flow; Landscape Composition; Landscape Configuration Potential Target Journals: Ecosystem Services; Trends in Ecology & Evolution; People and Nature 37 CONTEXT Origin - the workshop During five days in April 2018, I participated in the Workshop “Linking Landscape Structure to Ecosystem Services”, held at the Institute of Advanced Studies of the University of São Paulo. It was organized by Prof. Dr. Jean Paul Metzger and his team of the Landscape Ecology and Conservation Lab (LEPaC). The focus of the workshop was to promote a debate and collaboration among its participants to tackle the relationship between landscape structure and the components of the ES supply chain (supply-flow- demand). FAPESP (Project #217/50015-5) funded the event. Conception of the idea Extensive debate was carried out along those days and stablishing a common language was the first main challenge in the pursuit of new ideas and ways to further develop our understanding on how landscape influences ES. Several possible paths emerged during the very fruitful discussion. In the end of the workshop, the most successful ideas were intended to become papers. Out of three, I ended up in charge of one of them which focused on a spatial approach to further explore how landscape composition and configuration influence ES components. Embodiment – proposition and development As the first author in a varied group of collaborators, I struggled to translate into a manuscript outline the goals discussed during the workshop. After virtual meetings, several visits to LEPaC in São Paulo and plenty of reading, little by little, the manuscript gained shape. 38 1. INTRODUCTION Intense human-induced landscape modification – mostly led by urbanization and agriculture expansion – deteriorates ecosystems conditions and directly affects biodiversity and ecosystem functions (Johnson et al., 2017; Maxwell et al., 2016). Consequently, the provision of ecosystem services (ES) that sustain human wellbeing are declining globally (IPBES, 2019). Increasing human population is accompanied by higher demands for resources, such as food, water and energy (de Amorim et al., 2018; de Fraiture and Wichelns, 2010; Godfray et al., 2010). Effective landscape management is required to guarantee the sustainable use of environmental assets, to assure the maintenance of adequate amounts of ES supply and to safeguard equity in the access of ES (Berbés-Blázquez et al., 2016; de Groot et al., 2010). The ES supply chain is composed of three components: supply, demand, and flow. The spatial arrangement of supply and demand along with landscape composition and configuration can determine the flow of ES (Bagstad et al., 2014). Therefore, understanding the effects of landscape structure in the components of the ES supply chain is fundamental to promote adequate landscape management that can, at the same time, guarantee sustainable provision of ES and the maintenance of ecosystem functions and biodiversity. The supply and the demand components of the ES chain have been more extensively addressed in spatial terms, but not the flow. In this research we investigate how ES flow can vary according to landscape structure (composition and configuration), being facilitated or prevented to occur depending on the amount and spatial arrangement of ES supply and demand as well as other elements throughout the landscape. 2. DECOMPOSING THE ECOSYSTEM SERVICE CONCEPT The ES concept is a plural concept, able to navigate from a boundary object (i.e. adaptable to context and robust enough to enable communication) to a set of institutionalized infrastructures to operationalize the framework (Ainscough et al., 2019). Due to its complexity and interdisciplinarity, the characterization of ES as an applicable framework in distinct socio-environmental contexts has been widely discussed in the literature (Costanza et al., 2017). Decomposition of the ES supply chain has become more consensual as research digs deep in quantification and valuation methods and requires the distinction between the steps for ES provision. It becomes clear that the actual realization of an ES will only occur with the flow of ES from supply areas to demand areas 39 (Burkhard et al., 2014; Schröter et al., 2018; Villamagna et al., 2013). Hereafter, we assume the provision of ES depends on these three components: a) supply as the ecosystem capacity or potential to provide a given service; b) flow as the actual realization of a service or its translation into benefits for people; and c) demand as the service needed, desired or required by people (see Glossary for more details). While natural sciences embrace the supply side of ES (Martínez-Harms and Balvanera, 2012), broader participation of social sciences has further developed the demand side of ES (Wolff et al., 2017, 2015). Still, an interdisciplinary approach between economics, natural, and social sciences has been fundamental to integrating ES components and to address ES flow (Baró et al., 2016; Felipe-Lucia et al., 2015; Geijzendorffer et al., 2015; Palomo et al., 2013; Schröter et al., 2018). Interdisciplinarity is essential in this field, mostly because ES flows oscillate from geophysical and ecological processes (e.g., air mass movement for air pollution filtering; and the movement of organisms for pollination and biological pest control; de la Barrera et al., 2016; Grote et al., 2016; Medeiros et al., 2019) up to cultural appropriation (e.g., symbolism related to nature bonding; Fish et al., 2016; Schirpke et al., 2018), human agency and economic transactions (e.g., crop transportation and markets; Yu et al., 2013). Despite some variation in the definition of ES flow, there is a certain consonance in the way researchers address it. They may address the spatial matches and mismatches between supply and demand using the term ES flow itself (as in Baró et al., 2016; Felipe- Lucia et al., 2015; Ortiz et al., 2018; Palomo et al., 2013; Schirpke et al., 2019; Wolff et al., 2015) or, in other cases, using different terms (e.g., “actual use” in Schröter et al., 2012, “match” in Schulp et al., 2014). Some studies also deal with the spatial congruence of ES supply and demand, without directly addressing ES flows (see Geijzendorffer et al., 2015; Stürck et al., 2014; Willemen et al., 2012). Consolidation in the definition of ES flow is an ongoing process, and the understanding of its role in management and planning is of utmost importance to advance the discussion in the ES field. Environmental planning to secure the sustainable use of multiple ES depends on our comprehension of landscape relationships with ES supply, flow and demand and the spatial links between these ES components. Landscape management to optimize ES flow can be further developed by understanding (i) the spatial scale of ES components (Lamy et al., 2016; Mitchell et al., 2015b; Schirpke et al., 2019a; see Glossary), (ii) the spatial 40 (mis)matches between supply and demand (Baró et al., 2017; Geijzendorffer et al., 2015; Syrbe and Walz, 2012) and (iii) how landscape structure modulates ES flows (Baró et al., 2016) and therefore ES provision. Here we propose a spatially explicit framework relating landscape structure to ES focusing on ES flow. 3. HOW TO LINK LANDSCAPE STRUCTURE TO ECOSYSTEM SERVICES? Several studies evaluated the relationship between landscape and specific ES, limiting their analysis to how certain landscape attributes affect ES supply (for a few examples see Hodder et al., 2014; Inkoom et al., 2018; Liu et al., 2017; for reviews see Duarte et al., 2018; Lamy et al., 2016; Mitchell et al., 2013; Verhagen et al., 2016). There is a recurrent recognition of trade-offs between provisioning and regulating services as the landscape is managed to increase the supply of one type of ES over the other (Jopke et al., 2015; Raudsepp-Hearne et al., 2010). A simple explanation lies in the fact that most regulating services are provided by natural ecosystems whereas provisioning services – such as food, timber and fiber – come mostly from agroecosystems that are expanded at the expense of natural areas. In this example, increasing the provision of food would lead to decreases in pollination services (see meta-analysis on landscape structure effects on ES by Duarte et al., 2018). However, consideration of ES components separately (supply, flow and demand) can reveal more complex relationships between different ES, as well as how landscape structure shapes ES (Schirpke et al., 2019a). On the demand component of ES, there is evidence that demand for multiple ES direct or indirectly drives landscape modifications (Willemen et al., 2012). These drivers usually manifest in the form of economic pressures on crop, animal, and industrial production, resulting in land-use change (von Haaren et al., 2019). Main drivers of forest loss worldwide include agricultural expansion for commodities and food production as well as forestry and wildfire (Curtis et al., 2018). As a consequence of increasing provisioning services several regulating services are threatened without being accounted for (e.g., flood control, water regulation and carbon storage). Consequently, a number of attempts to manage land use and reduces the chances of undesired changes emerge in the form of policies, governance or market instruments. There are cases in which different spatial scales of supply and demand or a lack of a thorough description of the ES flow mechanism limit the assessment of ES instead of promoting an integrative spatial approach for landscape structure and ES components (cf. 41 (Serna-Chavez et al., 2014). Still, the application of the ES framework has led to a broader understanding of decoupled ES (i.e. when supply and demand occur at different spatial scales or landscapes, see Glossary) and, therefore, international or interregional flows have become more tangible (see Kastner et al., 2011; Koellner et al., 2019; Schröter et al., 2018). Despite relevant advances in integrative approaches for non-decoupled ES (i.e. when the ES supply and demand occur within a landscape, see Glossary), the understanding of how landscape structure affects flow is still superficial. We know that distance-dependent ES flows (e.g., pollination, in which the distance from crops to natural habitats matters; (Brosi et al., 2008; Ferreira et al., 2015) are usually most affected by landscape configuration (Baró et al., 2016; Mitchell et al., 2014), especially by the arrangement of supply and demand in space and which land uses lie between them (Serna-Chavez et al., 2014). Landscape structure has been commonly addressed through habitat loss and fragmentation, landscape complexity, spatial heterogeneity and connectivity, however not always using similar metrics and hindering comparison between studies (Duarte et al., 2018; Fahrig et al., 2011; Laterra et al., 2012; Mitchell et al., 2015a, 2013). Serna- Chavez et al. (2014) suggest that landscape metrics can assist the assessment of ES flow, once anthropogenic or natural barriers can affect landscape permeability for ES flows. Mitchell et al. (2015b) made a theoretical proposition about the effects of landscape fragmentation per se (sensu Fahrig, 2003) on ES supply and flow, without explicitly addressing ES demand. However, the understanding of ES flow, especially for non- decoupled ES, depends on the recognition of demand areas that benefit from ES flow (Serna-Chavez et al., 2014). For a better understanding on how landscape structure influences ES components (supply, flow and demand) and thus ES provision, going beyond only the effects of habitat fragmentation per se (Fahrig, 2003; Mitchell et al., 2015b), we need to fulfil several knowledge gaps: a) understanding the socio-economic drivers of landscape modification; b) how to predict trends and variation of ES responses to landscape structure over time; and c) critically assess synergies and trade-offs between emerging patterns of multiple ES. Thus, by further understanding the spatial distribution of supply and demand and the mechanisms of flow we can comprehensively understand how to manage landscapes to enhance supply and promote the sustainable flow of ES to meet existing demands. 42 Based on the ES theoretical framework (Figure 1), here we address how the ES supply and demand ratio (see Glossary), their spatial arrangement, and landscape structure modulate ES flow. For this we a) demonstrate how different landscape features can affect ES flow; b) integrate neutral areas in the landscape composition exploring the supply/demand spatial relationship including supply/demand overlap; and c) point out the challenges and propose a research agenda to enable landscape management to optimize ES flow. Figure 1: Conceptual model linking landscape structure and components of the ES supply chain. Landscape structure (both composition and configuration) can directly influence the condition of ecosystems (1) (e.g., their amount in the landscape or the quality of ecological processes within them) and, therefore, their capacity to supply ES (2). Landscape structure can also modulate the actual flow of ES (3), shaping their direction, permeability and amount (4). The demand fulfilment depends on the flow of ES and the conversion of ES supply into a benefit for humans (5). We argue that demand is mostly influenced by social aspects (6) and can be one of the drivers (7, 8) of changes in landscape configuration and composition. Other drivers (8) such as climate change, infrastructure development and urban expansion, can also lead to landscape modification (9). Similarly, political and economic instruments (8), like Payment for Ecosystem Services (PES), can be used to manage landscapes (9) with the purpose of increasing benefits to people via supply enhancement (1) or flow facilitation (3). (Adapted from Mitchell et al., 2015b). 43 4. ES FLOW MODULATED BY LANDSCAPE STRUCTURE ES flow can be modulated by the spatial arrangement of ES supply and demand areas. Therefore, identifying and mapping such elements is a first step to clarify how ES flow occurs within landscapes. The way each ES is delivered to people (whether it depends on a specific medium to be transferred or not) is crucial to identify which landscape attributes to account for. Several researches explicitly describe ES flow mechanisms (Table S1), many of which are modulated by landscape structure. For instance, the flow of regulating services, like pollination and biological pest control within agroecosystems, depends on the effects of the landscape on pollinator communities and their ability to move. Whereas the flow of cultural services such as outdoor recreation depends mostly on transportation infrastructure connecting people to green spaces (Balmford et al., 2015). To further explore the variation of ES flow we must investigate how it is modulated by the landscape and understand which landscape attributes may boost or hinder ES flow depending on the ES and the spatial distribution of supply and demand areas. For example, the erosion control service relies on a slope to be delivered, as well as on the location of vegetated areas (supply) and human settlements (demand) (Baró et al., 2017). In this case, the ES flow consists of risk reduction, meaning that the prevention of erosive processes (flow) benefits downslope human settlements. Modifying land use by removing natural vegetation uphill to create pastures, for example, would reduce the supply and flow for this regulating service, regardless to sparing adjacent downhill vegetation. Fragmentation or the aggregation of supply in this case could entirely modify the flow of the erosion control. Not only the amount of supply but how it is arranged may interfere in how much service is provided. In a steep slope landscape, the more aggregated the uphill vegetation is, the more it prevents erosion. However, in a hilly but not so steep region, a certain degree of fragmentation may optimize the amount of service provided. In another case, pollination is modulated by the capacity of pollinators that inhabit surrounding forest patches (supply) to move to a pollination-dependent crop (demand) to perform the pollination service by visiting multiple flowers. In this example, a more fragmented landscape may result in more contact (edge) between supply and demand facilitating the access of pollinators to crops to perform pollination (Figure 2a). In less fragmented landscapes with a continuous habitat patch adjacent to a large homogeneous 44 pollination-dependent crop, core crop areas would be more difficult to be reached by pollinators due to distance and possibly to movement resistance imposed on pollinators by crop matrix. Recreation services usually depend on people’s mobility via any means of transportation, be it terrestrial, aerial or waterway. As an illustration, consider a park with natural attractions such as an impressive waterfall (supply). Visitation of the waterfall will depend on peoples’ will to visit this location (demand) and their ability to get there (i.e. the existence of roads, bus lines, bike paths or the proximity to a train station). In this example, ES flow occurs when people commute from their homes to the park. Distance from residence to the park may act as a limiting factor (Figure 2b), that can be overcome by the existence of transportation infrastructure (Figure 2c). However, the spatial arrangement of transportation infrastructure can be heterogeneous and focused on a particular demand, for instance, connecting the park to the main nearby urban centre. In this case, it could be easier for people that live in the city to access the park instead of inhabitants of a closer nearby rural neighbourhood, due to accessibility (Figure 2c). Moreover, timber, as a provision ES, depends on forest condition and tree stocks (supply), on logging activity, and transport capacity (ES flow) to be properly delivered (Figure 2c). In this example, demand for timber could be any factory or industry that uses timber as a raw material (Kastner et al., 2011). Besides, the matrix could pose a resistance to the access to timber (Figure 2d). If a certain tree species was to be reached within a dense forest, relief, streams and understory vegetation can be an obstacle to reach and transport timber to nearby roads. Or else, the presence of a navigable river can facilitate access to certain remote areas. In these examples we illustrated how landscape configuration and the presence of certain elements within the landscape can facilitate or hinder the ES flow depending on its mechanism and the arrangement of supply and demand in space. In summary, we expect that combined landscape characteristics will present different effects on the flow, and the complexity of such combinations must be explored in further research. 45 Figure 2: Landscape management to optimize flow must rely on the effects of landscape configuration and composition on ES flow. We assume that certain landscape attributes will influence ES flow positively and others negatively, depending on the ES evaluated. Higher levels of fragmentation per se are expected to increase interspersion between supply and demand, facilitating flow to a certain degree for pollination (a). Increasing isolation between supply and demand areas will decrease flow of pest control (b). Complementarily, network complexity may facilitate flow not depending on distance, but on existing connections, as is the case of outdoor recreation (c). Likewise, matrix resistance may be the limiting aspect in other cases (d). 5. THE SPATIAL RELATIONSHIP BETWEEN ECOSYSTEM SERVICE COMPONENTS 5.1. Supply-demand ratio and flow variation for rival and nonrival ESs Given a hypothetical landscape in which all the space is occupied by ES supply or demand areas, varying from being totally occupied by demand to totally occupied by 46 supply, the maximum flow of a non-rival ES (see ES rivalry in Glossary) is not limited by the amount of supply once the flow can continuously occur without decreasing the supply. In such cases, the spatial arrangement of supply and demand in the landscape may prevent the maximum flow to occur (Figure 3a), leaving unattended demand. Using microclimate regulation as an example of non-rival ES, consider an urban green space capable of reducing heat islands within 100-m buffer area in a neighborhood (ES supply), no matter how many people stand in this buffered area (ES demand), the maximum flow will not be limited by demand. Therefore, if the maximum amount of demand possible is placed within the 100m buffer, the flow will be enough to satisfy it. Notice that in this case (and in others for which distance matters), the spatial distribution of supply can restrict the flow of the service depending on distance between supply and demand areas (Figure 3a and c). On the other hand, the maximum flow of a rival ES is limited by supply (Figure 3b). Using organic fruit produced in a community urban garden as an example of rival ES, the amount of supply will determine the demand that can be fulfilled. As demand for those organic fruit increases along with the number of people in the local community, the flow will be limited by the amount of fruit produced (Figure 3b). In a landscape perspective, the arrangement of supply and demand areas determines the intensity of flow and whether maximum ES flow is attainable. In the case of microclimate regulation, the flow will vary depending on the interspersion and size of green areas. The more interspersed (i.e. fragmented) among households and sidewalks the higher the flow of the service (black filled symbols in Figure 3a and c). If all green space is concentrated in only one area or portion of the urban landscape, the temperature will be lower in this area and its 100-m surroundings but all the rest of the neighborhood further than that will not receive the flow (unfilled symbols in Figure 3a and c). For the provision of organic fruit, people living further from the garden could have more limited access to the fruits, while people living adjacent to it could have more ready-to-go access. In this example, ES flow can be modulated by distance between supply and demand and maybe the more centralized the garden is the higher the flow (grey and black filled circles in Figure 3b and c). 47 Figure 3: Variation of flow along a demand/supply ratio for (a) non-rival and (b) rival ecosystem services. Here we use a (c) variation in the degree of fragmentation between demand and supply units to illustrate how it may affect flow (a and b). Maximum flow (dashed line) is limited by supply for rival ecosystem services (b). Blue shades illustrate flow variation under different landscape structure (a and b). White areas below dashed line (maximum flow) represent unmet demand when supply is available (a and b); grey area above dashed line represents unmet demand due to supply unavailability (b). Black, grey and unfilled symbols represent possible landscapes with a varying degree of fragmentation and D/S ratio. Examples of flow representation illustrate the difference between non-rival and rival ES: for the first, flow from the same supply unit may attend multiple demand units while for the second, flow from one supply unit to a demand unit prevents the flow to another demand unit. 5.2. Neutral areas interference in ES In a real landscape, the existence of supply and demand areas is usually accompanied by pervasive neutral areas (i.e. any land cover or land use that is not a supply or demand area for a given ES), different from what is shown in Figure 4a. In such cases, the supply-demand ratio is affected by a decrease in demand areas (Figure 4b), a decrease in supply areas (Figure 4c) or in both (Figure 4d). Using pollination as an example of a regulation service, in a landscape with forest remnants inhabited by pollinators (supply areas) and agricultural lands with a pollination-dependent crop (demand), the existence 48 of pastures dominated by an exotic grass species could be considered a neutral area for the pollination service, once it is not supply or demand area (Greenleaf et al., 2007; Ricketts, 2001). In this example, the amount (composition) and spatial arrangement (configuration) of landscape elements determine whether pollinators that inhabit the forest remnants can reach the crops and perform the service of pollination (i.e. ES flow). If pastures were not present, the configuration of habitat and crop would modify ES flow, possibly more interspersed landscapes would have higher ES flow (darker blue in Figure 4a) and less fragmented landscapes would have lower ES flow (lighter blue in Figure 4a). In a landscape with neutral areas substituting demand areas (Figure 4b), more habitat with some pasture and less crops, because the amount of crop (demand) is lower, maximum flow decreases accordingly (dashed line in Figure 4b). If pastures remain spatially between habitat and crops – and depending on crop depth in m – they may pose resistance to pollinators movement and prevent them from reaching core crop areas, consequently lowering ES flow (lighter blue flow in Figure 4b). On the other hand, when neutral areas substitute supply areas (Figure 4c), less habitat can still be enough to enable maximum ES flow depending on the landscape configuration. Lower amounts of habitat can still allow high ES flow when there is a high amount of edge between habitat and crops (darker blue areas in Figure 4c). However, if there is pasture in between them, again, ES flow will be reduced. Most certainly, landscapes with higher amounts of pastures would hinder pollinator movement from habitat to crops because the chance of pastures being present between them is higher (Figure 4d). However, ES flow would depend on the configuration of such elements. Using the microclimate regulation example, if there were empty lots (neutral areas) among the households in the neighborhood, that same amount of green space would be capable of supplying that same amount of service, however because demand is reduced, so is the maximum flow (dashed line in Figure 4b). In this same example, if neutral areas replace supply (Figure 4c) and demand remains the same, maximum flow can still be enough to fulfill demand. We understand that the arrangement (i.e. landscape configuration) of households, green spaces and empty lots (demand, supply and neutral areas respectively) will determine ES flow (Figure 4). 49 Figure 4: Variation in the amount of neutral areas (N) will modify the maximum flow of ecosystem service (dashed lines) within a landscape. We exemplify a landscape with ecosystem service demand (D) and supply (S) areas in the absence of neutral areas (a); a landscape with neutral areas replacing and thus reducing demand areas (b); a landscape with neutral areas replacing and thus reducing supply areas (c); and finally a landscape with neutral areas replacing and reducing both demand and supply areas (d). 5.3. Spatial overlap can expand supply-demand ratios and ES flow In other cases, supply and demand areas may be totally or partially overlapped in the landscape (Figure 5). Nonetheless, ES flow may still be affected by supply and demand spatial arrangement. As an example, the service of pest control depends on the existence of natural enemies that inhabit forest remnants (supply) and perform this regulation service in nearby crops (demand) (Boesing et al., 2017). In this case, ES flow would vary according to contact and distance between forest and crop (as the variation in the blue areas in Figure 5a). Spatial overlap can be exemplified for the same service in agroforestry systems which can host pest controllers (supply) and at the same time can be the demand area for pest control (Figure 5c) (Perfecto et al., 2004). A combination of both cases, with forest remnants, crop areas and agroforestry, represents a partial overlap (Figure 5b). For 50 biological pest control (Medeiros et al., 2019), the arrangement of supply and demand in space, as well as the overlap will determine flow. The closer supply and demand are the higher the flow (Blanche et al., 2006). When overlapped, we expect maximum flow to be the highest (compare blue flow variation areas in Figure 5a with 5b and 5c). On the other hand, the less overlap between supply and demand are, the lower the flow between them (as in the light blue area in Figure 5a). Figure 5: Landscapes may contain spatial overlap between ecosystem service demand (D) and supply (S) areas. For landscapes in which there is no overlap, maximum flow of ecosystem services (dashed line) may occur between supply and demand areas (a). Flow can vary below maximum according to other landscape attributes (blue gradient). In some cases, there might be a partial overlap between supply and demand areas (b) or a more extreme total overlap may take place (c). 6. CHALLENGES AND RESEARCH AGENDA Using landscape planning to optimize flow between supply and demand might be a feasible way to achieve more sustainable practices and long-term ES provision. If research efforts integrate all three components of the ES supply chain (supply, flow and demand) in the assessments and improve the understanding on how landscape patterns influence each one of them, landscape can be managed more efficiently to optimize flow, enhance supply and regulate demand for ES. In order to advance our framework, some challenges must be overcome: 51 6.1. Quantifying supply and demand variation using comparable units In our framework we assume a non-variation of supply and demand per area unit. However, ecosystems conditions are heterogeneous and highly influenced by landscape composition and configuration, which have important consequences for ES provision. Thus, when accounting for ES supply one must take the variation of ecosystem condition into consideration. Likewise, ES demand per area unit depends on the development of methods and tools to estimate how much service people need by household or other spatial demand unit. Demand will vary depending on population density, consumption habits and culture. Such quantification and mapping should be further developed in collaboration between natural and social sciences. Besides, one may argue that landscape structure may also affect demand. Thus, another challenge remains: how do environmental aspects influence demand allocation? Once we have supply and demand quantified and mapped, understanding flow mechanisms and the spatial match between supply and demand will allow the management and sustainable plan of ES provision within a landscape or even outside of it through interregional flows for decoupled ESs. Another aspect to be considered is how high demands and flow for certain ES may end up degrading the supply (decreasing its quantity and quality). This is more straightforward for rival ES, but it may also occur for non-rival ES. For example, high visitation rates (flow) in protected areas can degrade hiking paths as well as the attractiveness of the place, thus degrading supply of the recreational service. Controlling the access of visitors (flow) may be enough to prevent degradation of supply and ensuring the long-term flow of this service. 6.2. Managing demand to enable optimal flow Demand can be managed by the development of people’s awareness about ES provision and therefore enabling their engagement in the development of policies to guarantee ES provision in the future. Demand management may become feasible with the recognition of societies needs and priorities regarding ES provision. However, for this purpose government, institutions and other interests must be aligned in the pursuit of collective wellbeing and sustainable practices. A few aspects to be considered, other than natural and social sciences joining efforts, are: a) how feasible it is to manage demand; b) how to develop and improve methods to assess people’s desires and needs; and, c) how 52 to value ES and benefits, not exclusively in monetary terms, so that governments can excel in serving citizens’ needs, while institutions can seek more sustainable practices. Therefore, linking this to a landscape perspective is also necessary to elucidate how the fulfillment of a local demand depends on local supply or on the supply of other landscapes (further regions). 6.3. Exploring how neutral areas in the landscape affect ES flow What we call neutral areas here are in fact the landscape matrix, or land-uses other than those that characterize supply or demand areas. Neutral areas may be different for each ES, so we must consider this variation when dealing with multiple ES. Moreover, although not captured in our framework, neutral areas vary on how much resistance and/or reduction they can cause on flow. For example, different agricultural areas can impose different movement resistance to birds (Barros et al., 2019; Boesing et al., 2017), thus the presence of a particular agricultural area between supply and demand areas can increase or decrease biological pest control services provided for demand areas by these natural enemies (Medeiros et al., 2019). Therefore, elucidating the different roles of such neutral areas in the connection between supply and demand and the modulation of flow for different ES is the next step to be taken. For this, describing and clarifying flow mechanisms (i.e. those mediated by animal movement, people’s movement, mass movement, transport infrastructure, trading and market prices) and how they vary across neutral areas will guide the further comprehension of the role of anthropogenic matrix attributes (e.g., diversity, resistance, similarity, contact, distance). 6.4. Understanding the scale of effect on flow management Managing landscapes to improve ES flow requires not only the comprehension of flow mechanisms but also the adoption of multi-scale approaches. ES flows within a landscape or region is not limited by administrative boundaries. The multi-scale ES flow management depend on people, governments and institutions to be on the same page, specially when ES supply and demand are decoupled. Knowing and recognizing who will be affected by management decisions is a promising way to go and improve spatial planning by including landscape effects on ES flows, as pointed out in this study. 53 6.5. Incorporating synergisms and trade-offs in the spatial explicit framework Identifying synergies and trade-offs when evaluating ES bundles has implications for landscape management with ES maintenance goals, including enhancing supply, optimizing flow and coupling it with demand. The proposed spatial framework emphasises that ES synergies and trade-offs can be fully captured with consideration of all three ES components and their relationship with landscape structure. Therefore understanding how landscape composition and configuration affect ES supply (e.g., Duarte et al., 2018) is essential, but not enough. Researchers must also address how landscape structure relates to ES demand and flow to predict the consequences of landscape management to ES provision. Likewise, existing research focusing on multiple services (e.g., Lamy et al., 2016; Mitchell et al., 2014), ES bundles (e.g., Baró et al., 2017), and synergies and trade-offs (e.g., Power, 2010; Schirpke et al., 2019a) also evidenced the urge to integrally address the spatial aspect of ES supply, flow and demand to inform environmental planning and sustainable use of ES. 6.6. Integrating temporal variation on ES assessments Temporal variation in ecosystem conditions and consequently in ES supply brings consequences to flow and demand fulfilment. Nonetheless, demand also varies over time due to social aspects affecting people’s desires and needs from nature. This perspective has gained attention in planning initiatives for sustainable futures due to the fact that in order to guarantee long-term ES provision, we need to account for variations of both supply and demand through time and thus its effect on ES provision. This means that not only the spatial variation influences ES provision but also the way each component changes over time. Despite the very long way to go in the development of spatio-temporal approaches, scientists’ awareness of such temporal variations may be the necessary heads up for management and policy development for the present time. Another promising strategy to account for temporal variation in ES provision is incorporating the ES framework into both, the conservation and restoration agendas. This could assist the development of a sustainable management of ES over time. Because supply can be both protected or ‘created’ (i.e. restored, regenerated or recovered; Chazdon, 2008) or enhanced to assist unfulfilled demand, the restoration agenda could incorporate ES. More specifically on the restoration of ecosystems, once restored areas 54 have the potential to provide several ES, choosing where to restore can optimize ES flow. Similarly, research efforts could address whether there is a minimum amount of supply to guarantee long-term ES flow without affecting ES supply maintenance (Hein et al., 2016). GLOSSARY Ecosystem Service Supply Chain: parts of ecosystem service realization that generates benefits to people from ecosystems (Tallis et al., 2012). These components consist of the ecosystem service supply, ecosystem service flow and ecosystem service demand. Ecosystem Service Supply: the potential of an ecosystem to provide an ecosystem service irrespective of it being used, recognized or valued by humans (Mitchell et al., 2015b; Tallis et al., 2012). Even though ES supply per spatial unit may vary according to the ecosystem type and its condition, here we refer to supply as the area of the ecosystem that has the potential to supply a certain ES, without accounting for such variation. Ecosystem Service Flow: the delivery of an ecosystem service to people; the actual production of a service experienced by people (Villamagna et al., 2013); the transfer of service from supply to demand through space and time (Bagstad et al., 2013; Serna- Chavez et al., 2014). ES flow characteristics are specific to each service (Serna- Chavez et al., 2014). Ecosystem Service Demand: the amount of service required or desired by people (Villamagna et al., 2013; Wolff et al., 2017, 2015), even in cases in which they are not aware of such needs. ES demand can be expressed in terms of risk reduction, preferences and values, direct use or consumption of goods and services (Brander and Crossman, 2017; Wolff et al., 2017, 2015). Here we assume the demand area as the location of people who require or desire the ES in question despite variations between these areas, such as population, or cultural differences Ecosystem Service Provision: the delivery of a service to be used or enjoyed by people (Mitchell et al., 2015b; Villamagna et al., 2013). 55 Benefit: positive change in people’s wellbeing (Tallis et al., 2012) as a result of ES flow from ES supply meeting an existing ES demand. Neutral Area: an area in the landscape that is neither supply or demand for the ES in question, but that can contribute to ES flow (i.e. facilitate, hinder or be neutral to it). Here we assume that neutral areas do not interfere in supply and demand even though in landscape contexts such interactions are prone to generate variability. Ecosystem Service Rivalry and Excludability: to be rival means that the use of an ecosystem service by one person precludes the use of it by another person (Fisher et al., 2009). To be excludable means that one person can keep another from using or accessing an ecosystem service (Fisher et al., 2009). Marketed goods are usually rival and excludable; however, it is possible to fit ecosystem services along a continuum from rival to non-rival and from excludable to non-excludable. Decoupled Ecosystem Service: when the supply of and demand for an ES occur at different spatial scales or in different landscapes (Burkhard et al., 2014; Fisher et al., 2009). Inversely, non-decoupled ES is when an ES has its supply and demand occurring within the same landscape. The spatial relationship between supply and demand has been previously classified as in situ, omnidirectional, and directional (Burkhard et al., 2014). Ecosystem Condition: refers to the integrity and state or functioning of an ecosystem, ultimately determining its capacity to supply ES (Burkhard and Maes, 2017; UN, 2014). Landscape Structure: spatial composition (the number of existing elements) and configuration (how they are arranged in space) of the landscape (Wu, 2013). Spatial Scale: the extent and grain of the physical structure of the system and the distance and range in which its processes occur (Dungan et al., 2002). Supply and Demand Ratio: the proportion between the amount of ecosystem service supply area and the amount of ecosystem service demand area, here considered for non- decoupled ES. Even though we acknowledge the existence of variation in supply and demand per area unit, here we do not take this variation into account. 56 Table S1: Summarized characterization of ecosystem service components (supply, flow and demand), measurements and proxies used in previous research. ECOSYSTEM SERVICE SUPPLY FLOW DEMAND Provisioning Services Water provision for irrigationa location of groundwater basins and their annual recharge infrastructure for extraction, distribution and irrigation; Basin recharge rate (soil infiltration capacity and topography) agricultural areas dependent on groundwater irrigation Urban agricultureb organoponics and urban agriculture crop yield of vegetables and fruits minimum recommended intake of fruits and vegetables for all inhabitants in residential areas Drinking water provisionc water availability water usage water requirement Grassland biomassc forage production the actual used fraction of forage energy demand of forage feeding livestock species Fuel woodc increment of above- ground biomass in forests timber removals (forest accessibility and in-site conditions) fuel wood requirements based on statistics of consumption, building areas and heating degree days Regulating Services Filtration of surface waterc potential nitrogen filtration capability amount of nitrogen filtered by ecosystems amount of nitrogen introduced in ecosystems Protection of areas against avalanches, mudslides and rockfallsc forest area with a protective effect forest area with a protective effect for human infrastructure area of human infrastructure located in avalanche and rock- fall hazard zones Climate regulationa primarily, tropical (including peat forests) and boreal forest (hotspots) tropical, sub-tropical and mid-latitude agricultural areas vulnerable to changes in precipitation and temperature vegetation and soil interactions with the atmosphere (photosynthesis, soil and plant evapotranspiration, soil heterotrophic respiration, albedo, etc.) Carbon sequestrationc the annual rate of CO2 sequestration by above- and below-ground biomass in forests equal to supply the annual rate of CO2 emissions 57 Air purificationd urban vegetation (trees and shrubs) air quality improvement by removing pollutants the magnitude of pressures on air quality and population exposed to these pressures Pollinationa natural and semi- natural habitats foraging ranges of pollinator species, pollen transportation and deposition, abundance and effectiveness of pollinating organisms agricultural areas dependent on or profiting from biotic pollination Erosion regulation soil classification, soil organic matter, standing vegetation control of landslides and amount of sediment retained by ecosystems in stakeholder areas risk of soil erosion Cultural Services Outdoor recreationd availability of recreational sites proximity from people’s homes to recreational sites population density (assuming all inhabitants have similar desires) Outdoor recreationc recreation opportunities provided by ecosystems visitation rates density of potential beneficiaries derived from the number of residents and overnight stays of tourist Outdoor Recreationb urban green spaces considering threshold values of crowding inhabitants of urban areas from which dwellers can reach urban green spaces considering access and a maximum travel distance the minimum availability of recreational green spaces within maximum distances from home including all urban residential zones Symbolic speciesc spatial distribution of habitats of selected symbolic plants and animals mapping of the occurrence of selected symbolic plants and animals in hotel names not mentioned a Serna-Chavez et al., 2014; bOrtiz et al., 2018; c Schirpke et al., 2019a; d Baró et al., 2016 58 59 CHAPTER 2 – IDENTIFYING HOTSPOTS FOR BIODIVERSITY CONSERVATION: FROM ECOLOGICAL TO ADMINISTRATIVE BOUNDARIES Julia Camara Assis1, Julia Emi Faria Oshima1, Vinicius Rodrigues Tonetti1, João Carlos Castro Pena1, Thadeu Sobral-Souza1, Tadeu Gomes de Oliveira2, Larissa Barreto3, Lars Hein4, Milton Cezar Ribeiro1 1 Spatial Ecology and Conservation Lab, UNESP-Rio Claro, São Paulo, Brazil 2 Universidade Estadual do Maranhão, UEMA, São Luís, Maranhão, Brazil 3 Universidade Federal do Maranhão, UFMA, São Luís, Maranhão, Brazil 4 Wageningen University & Research, Wageningen, The Netherlands 60 LIST OF COLLABORATORS AND THEIR PARTICIPATION*: Supervisors: Larissa Barreto, Lars Hein, Milton C Ribeiro, Modelling procedures: Julia Emi Faria Oshima, Maurício Humberto Vancine, Thadeu Sobral-Souza Data gathering: Vinicius Rodrigues Tonetti Manuscript review: João Carlos Castro Pena, Maurício Humberto Vancine Mammal data providers: Alexine Keuroghlian; Ana Paula Carmignoto; Edsel Amorim; Fernando Tortato; Paul Colas Rosas; Rafael Hoogestein; Babi Zimbres; Ricardo Machado; Carlos Peres; Edson Souza; Maria Luisa da Silva Pinto Jorge; Rodrigo S. P. Jorge; Francesca Belem Lopes Palmeira; Cristiano T. Trinca; Leandro Silveira; Anah Jácomo; Giselle Bastos Alves; Natalia Torres; Cazuza Junior; Grasiela Porfírio; Guilherme Ferreira; Teresa Anacleto Data and expert consultation for mammals: Julia E F Oshima, Tadeu Gomes de Oliveira; Paloma Marques; Alessandra Bertassoni; Carmen Barragan; Elildo Junior; Gerson Buss; Jim Patton; Marcos de Souza Fialho Bird data providers from WikiAves: CERRADO: Breno Pinheiro; Cristóvão Pereira; Douglas Meyer; Eric Nando Vieira; Fabricio Teresa; Fernanda Fernandex; Fernando Bittencourt; Leonardo Victor; Lucas Alves2; Marcos A Lima; Matheus Santos 2; Nicodemus Rosa; Paulo Vale (Poloca); Rafael Grisostenes; Ronaldo Francisco AMAZON: Adriel Raach; Almir Távora; Felipe Bittioli; Ian Thompson; Kurazo Okada; Luciano Faria CERRADO AND AMAZON: Emerson Kaseker; Igor Camacho; Thiago Silveira; Victor Castro Data and expert feedback for birds: Sidnei de Melo Dantas (for Amazon species); Vitor Carneiro de Magalhães Tolentino (for Cerrado species) * All the collaborators listed above will be co-authors in the manuscript to be submitted 61 ABSTRACT Biodiversity conservation is vital to maintain ecological functions that provide ecosystem services that sustain human life. Most tropical regions with vast biodiversity have incomplete species databases. Even in these cases, Species Distribution Models (SDM) have been widely applied in the design of conservation and land management actions. In this study, we proposed the use of a climate- and landscape-based SDM approach to identify biodiversity hotspots to aid management decisions in face of fast land use change. Maranhão State, one of the frontiers of deforestation and agriculture expansion in Brazil, was used as a study case. We selected nine bird and twelve mammal species of ecological relevance in the Amazon and Cerrado areas of the state and adopted a broader modelling background, according to species distributions range. We used the ensemble of five replicates of four different algorithms to come up with a frequency probability of each species occurrence for climatic and landscape variables. We considered areas with high climate and high landscape suitability as species hotspots. Then, we quantified the spatial overlap of these hotspots and analysed their spatial distribution regarding recent land use conversion and existing protected areas network. Our results showed that 90% of the Maranhão State is a hotspot for at least one species, whereas ~30% of its area is a hotspot for more than five species. Agricultural areas in 2018 had ~31% of overlap with hotspots for more than five species, which may be critical for biodiversity conservation in these regions. Out of the total area with some degree of legal protection, 48.2% hosts hotspots for more than five species. We indicated the priority areas to be conserved as those with higher overlap of species hotspots and not encompassed by existing protected areas. We recognize urgent need to monitor areas with agriculture expansion. Biodiversity hotspots allowed the recognition areas with the potential to host several species and therefore the identification of key regions to be protected. Keywords: Species Distribution Modelling; Maranhão; Neotropical biodiversity; Conservation Biogeography Potential target journals: Conservation Biology; Biodiversity Conservation; Perspectives in Ecology and Conservation 62 1. INTRODU