Biodiversity Feedbacks and the Sustainability of Social-Ecological Systems Anne-Sophie Lafuite

Biodiversity feedbacks and the sustainability of social-ecological systems Anne-Sophie Lafuite

To cite this version:

Anne-Sophie Lafuite. Biodiversity feedbacks and the sustainability of social-ecological systems. Ecol- ogy, environment. Université Paul Sabatier - Toulouse III, 2017. English. NNT : 2017TOU30289. tel-01949635

HAL Id: tel-01949635 https://tel.archives-ouvertes.fr/tel-01949635 Submitted on 10 Dec 2018

HAL is a multi-disciplinary open access L’archive ouverte pluridisciplinaire HAL, est archive for the deposit and dissemination of sci- destinée au dépôt et à la diffusion de documents entific research documents, whether they are pub- scientifiques de niveau recherche, publiés ou non, lished or not. The documents may come from émanant des établissements d’enseignement et de teaching and research institutions in France or recherche français ou étrangers, des laboratoires abroad, or from public or private research centers. publics ou privés. THÈSETHÈSE

En vue de l’obtention du DOCTORAT DE L’UNIVERSITÉ DE TOULOUSE

Délivré par : l’Université Toulouse 3 Paul Sabatier (UT3 Paul Sabatier)

Présentée par : Anne-Sophie LAFUITE Érosion de la biodiversité & durabilité des systèmes socio-écologiques

JURY Jérôme CHAVE DR CNRS Président du Jury Madhur ANAND Pr Membre du Jury Anne-Sophie CRÉPIN Dr Membre du Jury Safa MOTESHARREI Dr Membre du Jury Luc Doyen DR CNRS Membre du Jury Michel LOREAU DR CNRS Membre du Jury

École doctorale et spécialité : SEVAB : Écologie, biodiversité et évolution Unité de Recherche : Station d’Écologie Théorique et Expérimentale (UMR 5321) Directeur de Thèse : Michel LOREAU Rapporteurs : Madhur ANAND et Anne-Sophie CRÉPIN Abstract

In this thesis, I intend to develop theoretical interdisciplinary approaches that provide new insights into the sustainability of coupled social-ecological systems (SESs). Espe- cially, I focus on the long-term consequences of the current biodiversity crisis on human demography. Human–nature interactions occur over many spatial and temporal scales, and mismatches between the scales of human dynamics and ecological processes can be detrimental to sustainability. I use dynamical system modelling to represent human– nature interactions through a feedback loop, that includes human economy, demography, and biodiversity dynamics. Human population growth and natural habitat destruction drive species to extinction, thus aﬀecting the provisioning of essential ecosystem services. In turns, many of these services indirectly or directly aﬀect agricultural productivity, thus threatening the provisioning of food to a growing human population. First, I show that temporal mismatches resulting from a time-delayed loss of biodiversity can generate unsustainable human population cycles. Model analysis enlightens the existence of critical thresholds in the dynamical variables of the system. Remaining within those sustainable boundaries allows avoiding overshoot-and-collapse population crises that greatly reduce human well-being, and thus sustainability. Second, I explore the relationships between human behavioral change and the adaptive capacity of SESs. I show that temporal mismatches postpone desirable behavioral changes and increase the probability of an abrupt shift towards an unsustainable basin of attraction. Policies that help reduce temporal mismatches, i.e., accelerate behavioral changes or reduce extinction debts, are thus crucially needed. The third chapter explores the potential of land-use management directed towards the conservation of natural habitats to reduce temporal mismatches, and shows that

i the economic internalization of biodiversity feedbacks can help prevent or mitigate unsustainable crises. This thesis emphasizes the role of feedbacks and scales in human–nature interactions, and highlights the importance of foresight for the long-term sustainability of human societies. Policy implications regarding cross-scale eﬀects are discussed, so as to deﬁne integrative approaches fostering adaptability and sustainability. Results call for a better understanding and assessment of both time delays and spatial processes, and urge for the development of integrative management approaches accounting for human demography, socio-economic and long-term ecosystem dynamics, at various spatial and temporal scales.

ii Résumé général

La perte des habitats naturels et leur fragmentation est responsable d’une crise d’extinction de la biodiversity inédite, qui réduit à son tour l’approvisionnement en services écosys- témiques difficilement substituables. La plupart de ces services influencent directement ou indirectement la productivité agricole, menaçant ainsi l’accès à l’alimentation d’une population humaine en pleine expansion. De plus, la destruction des habitats ne mène pas instantanément les espèces à l’extinction. Des espèces vouées à l’extinction peuvent mettre des décennies à s’éteindre, ce qui génère d’importantes dettes d’extinction. Les mécanismes sous-jacents à ces dettes reposent sur des processus écologiques impliquant plusieurs échelles spatiales, et leurs conséquences s’étendent sur des échelles temporelles allant de la décennie au siècle. Les effets d’échelle sont omniprésents dans les dynamiques des systèmes socio-écologiques (SSEs). En particulier, l’interaction entre des variables lentes et rapides peut amplifier des feedbacks non durables et mener à des effets inatten- dus, comme des transitions abruptes vers des états indésirables. La présente thèse propose d’explorer les conséquences sur la durabilité des SSEs de ces délais temporels. Le travail développe des approches interdisciplinaires et théoriques novatrices, incluant des éléments importants d’économie, de démographie humaine et de dynamique sociale et écologique caractérisant ces systèmes. Les résultats obtenus sont utilisés pour proposer plusieurs pistes de politiques publiques permettant de mieux prendre en compte ces effets d’échelle tout en limitant leurs conséquences négatives et en favorisant l’adaptabilité des SSEs. Un premier modèle couplant de manière stylisée la démographie de la population humaine, l’économie de marché et une érosion différée de la biodiversité affectant la productivité agricole, est d’abord présenté. Son analyse montre l’existence de seuils critiques dans les

iii variables dynamiques du système. Maintenir la taille de population humaine et la perte de biodiversité en deçà des seuils durables précédemment identifiés permet d’éviter des effondrements type “overshoot-and-collapse”, durant lesquels le bien-être humain est fortement réduit, ainsi que la durabilité du système. Mais la capacité d’adaptation des SSEs n’est pas à sous-estimer et peut permettre de maintenir la durabilité d’un SSE, notamment via des changement de comportement humain. Le chapitre suivant utilise ces seuils de durabilité pour définir une norme de consommation durable, dont l’adoption permet de préserver la durabilité du système. Cependant, lorsque les choix des consommateurs se basent sur leur perception à court terme de l’état de dégradation environnementale, le modèle montre que les délais temporels engendrent de la procrastination vis-à-vis des changements de comportement à effectuer. Cette procrastination peut mener le système à basculer brusquement dans un bassin d’attraction non durable. Dans une telle situation, seule une vision de long terme peut permettre d’éviter un tel changement de régime à large échelle. Certaines politiques publiques peuvent s’avérer utiles pour accélerer le changement des comportements de consommation en améliorant l’accès à l’information des consommateurs. Les politiques publiques peuvent aussi directement réduire les dettes d’extinction en conservant les habitats naturels, par exemple via la mise en place de taxes foncières limitant la destruction des habitats naturels, ou la création de réserves naturelles. Mais de telles politiques doivent alors prendre en compte les caractéristiques spatiales des SSEs, et notamment les effets d’échelle. Cette thèse démontre l’importance des effets d’échelle temporels sur la durabilité de long terme des SSEs. Une meilleure compréhension de ces effets d’échelles et de leurs conséquences sur les systèmes couplés homme-nature est nécessaire. Le développement d’approches de gestion intégratives doit prendre en compte les feedbacks existant entre la démographie humaine, les systèmes socio-économiques et la dynamique des écosystèmes, à différentes échelles spatiales et temporelles.

iv Remerciements / Acknowledgments

The work made during this PhD project has beneﬁted from insightful advice, help and support from numerous people I would like to warmly thank here / J’aimerais remercier ici toutes les personnes grâce à qui cette thèse a été rendue possible.

Tout d’abord, merci à Michel, mon directeur de thèse, de m’avoir laissée libre dans le choix de mon sujet de thèse, et de m’avoir accordé sa conﬁance. Merci pour ta patience et ton soutien tout au long de cette (longue) thèse!

Merci également à François Salanié pour les discussions et conseils avisés sur les aspects économiques de mon modèle, et pour avoir répondu présent à chacun de mes appels au secours.

I would also like to thank all the members of the “Human Spatial Behavior, Biodi- versity and Ecosystem Services“ group, for insightful comments and discussions at the beginning of my PhD project; Charles, Ann, Jean, Ingela, and Claire. A special thank to David, for useful comments on my ﬁrst manuscript, and for the great time we spent crack climbing together in Arizona.

Merci à tous les collègues et amis que j’ai eu la chance de cotoyer durant ces quelques années dans ces belles Pyrénées Ariégeoises. Merci à Maryse et Christian pour leur accueil formidable, et tous ces bons moments partagés dans leur magniﬁque gîte. Merci à tous les copains de la première heure, Tomas, Shaopeng, Mat, Theresa, Jarad, Aisha, Elie, James, Yann et Tom, pour ces ”kiki“ mémorables, les leçons de salsa/rock et les cours de guitare, les cols à vélo avant le passage du tour de France, les bivouacs et toutes ces expériences culinaires (et excursions hospitalières) en votre compagnie. A mes compagnons ”outdoor”, Alexis, Gauderic, Gougou, Jeﬀ, Jonathan et Kévin, merci de m’avoir sorti le nez de

v ma thèse pour aller grimper ou crapahuter en montagne ! Merci à Robinou pour les expériences d’escalade arboricole, les tentative de vol de canoë ratées, et les schoko bons suprise qui redonnent le sourire :) Merci à Anne-So (n°1) pour les pauses thé remplies de conseils, astuces et potins, ainsi qu’à Audrey pour ses nombreux conseils avisés sur mon premier article, et les discussions autour du projet Homme-Nature en général. Et enfin, merci à Arnaud pour son écoute et son soutien sans faille, ses cadeaux étranges, et ses bons petits plats. Et parceque ma vie ne s’est pas limitée à Moulis et son CNRS ces dernières années, merci à ma copine de toujours (ou presque), kikette, et à tous les autres copains niçois, lyonnais ou parisiens, pour leurs encouragements. Merci à Evrard de m’avoir donné la chance de l’accompagner à trois reprises en expédition. Merci pour ces expériences qui ont changé ma vie, m’ont permis de rencontrer des personnes extraordinaires et de tisser des liens d’amitié très forts. En particulier, merci à Aurélie ”moineaute“ pour tes magnifiques aquarelles qui me rappellent ces moments inoubliables au milieu de la forêt du Sulawesi, et à Tanguy, the spécialiste des crottes d’ours des Pyrénées. Un merci tout particulier à Yanick, pour ta présence malgré la distance, pour m’avoir transmis un peu de ton énergie inépuisable, et permis de prendre un peu de hauteur. Enfin, merci à ma famille pour son soutien, et en particulier à mes grand-parents, pour leur intérêt et leurs questions incessantes. J’espère que vous allez enfin comprendre ce que j’ai fait pendant toutes ces années! Sans oublier Tana et Youk, pour leur joie de vivre au quotidien...

vi Preface

This thesis of the SEVAB doctoral school (Toulouse III University) was conducted at the Centre for Biodiversity Theory and Modelling, at the CNRS Theoretical and Experimental Ecology Station in Moulis (France). The PhD was directed by Michel Loreau, and funded by the Midi-Pyrénées region and the TULIP Laboratory of Excellence (ANR-10-LABX- 41).

The three chapters of this thesis are prepared for journal contribution. Chapter 1 is published in Ecological Modelling, and Chapter 2 is accepted in Proceedings of the Royal Society B. The remaining article is in preparation for the Journal of Environmental Economics and Management (Chapters 3).

Chapter 1 1 presents a model that couples human demography, market economics and biodiversity loss, before exploring the consequences of ecological time delays on the long- term sustainability of this system. Anne-Sophie Lafuite and Michel Loreau conceived the research, in collaboration with François Salanié (Toulouse School of Economics). Anne- Sophie Lafuite conducted the model analysis and led the writing of the manuscript. Michel Loreau reviewed the manuscript.

Chapter 2 2 explores the potential of social norms in enforcing sustainability, when their evolution depends on the short-term human perception of environmental degradation. Anne-Sophie Lafuite, Claire de Mazancourt and Michel Loreau conceived the research. Anne-Sophie Lafuite analyzed the model and prepared the manuscript. All authors reviewed it critically.

1Lafuite, A.-S. and Loreau, M. (2017). Time-delayed biodiversity feedbacks and the sustainability of social-ecological systems. Ecological Modelling 351:96-108 2Lafuite, A.-S., de Mazancourt, C. and Loreau, M. (2017). Delayed behavioural shifts undermine the sustainability of social-ecological systems. Proceedings of the Royal Society B: Biological Sciences 284:20171192

vii Chapter 3 3 focuses on potential regulatory policies that could prevent or limit the negative consequences of ecological time delays on human societies. Analyses and simulations were conducted by Anne-Sophie Lafuite and Gonzague Denise. Anne-Sophie Lafuite prepared the manuscript; Michel Loreau and Gonzague Denise reviewed it.

3Lafuite, A.-S., Denise, G., and Loreau, M. (submitted). Sustainable land use management under biodiversity lag eﬀects.

viii Contents

Abstract i

Résumé général iii

Remerciements / Acknowledgments v

Preface vii

List of Figures xi

List of Tables xii

General Introduction 1 Context ...... 2 Human-nature interactions in the Anthropocene ...... 10 Towards integrative sustainability approaches ...... 13 Objectives, methods and main questions ...... 18 Bibliography ...... 22

1 Time-delayed biodiversity feedbacks and the sustainability of social-ecological systems 37 Chapter outline ...... 38 Abstract ...... 39 Introduction ...... 40 1 Methods ...... 42 2 Results and Discussion ...... 50 3 Conclusions ...... 60 Bibliography ...... 64

Appendices 69 1.A Endogenous or Exponential Technological Change ...... 70 1.B Economic derivations ...... 72 1.C Dynamical System Analysis ...... 75 1.D Eﬀect of Land Operating Costs on Sustainability ...... 77

2 Delayed behavioral shifts undermine the sustainability of social-ecological systems 79 Chapter outline ...... 80 Abstract ...... 81 Introduction ...... 82

ix 1 Model description ...... 85 2 Results ...... 93 3 Discussion and conclusions ...... 99 Bibliography ...... 104

Appendices 110 2.A Functions and aggregate parameters ...... 111 2.B Parameters deﬁnition, units and defaults values ...... 111 2.C Dynamical system analysis ...... 112 2.D Sensitivity analysis ...... 115

3 Sustainable land-use management under biodiversity lag eﬀects 118 Chapter outline ...... 119 Abstract ...... 120 Introduction ...... 121 1 A simple land - biodiversity - demography model ...... 124 2 A natural land depletion tax ...... 127 3 Dynamical system analysis ...... 130 4 Optimal land conversion policy ...... 135 5 Conclusions and discussion ...... 141 Bibliography ...... 145

Appendices 151 3.A General Market Equilibrium ...... 152

General discussion 157 Synthesis of the results ...... 158 Perspectives ...... 163 General conclusion ...... 173 Bibliography ...... 174

x List of Figures

A Population projections and global land use change...... 4 B The Global Living Planet Index...... 6 C Local biodiversity loss...... 6 D Extinction and functioning debts...... 16

1.1 Coupling between human and ecological dynamics ...... 44 1.2 Transient dynamics with varying relaxation coefficients ()...... 53 1.3 Sustainability sensitivity analysis...... 58 1.4 Land-use scenarios...... 60 A1.1 Effect of endogenous and exponential technological change on sustainability. 71 A1.2 Effect of the land operating cost (κ) on sustainability...... 78

2.1 Coupling between human, social and ecological dynamics ...... 86 2.2 Combined effects of ostracism and the difference between sustainable and unsustainable norms on stability...... 95 2.3 Effect of the initial conditions on the long-term equilibria...... 96 2.4 Effect of varying extinction debts on transient dynamics and stability. . . . 98 2.5 Combined effects of the extinction debt and social pressure on long-term sustainability...... 99 A2.1 Distribution of the sensitivity of the sustainability criterion ∆...... 117

3.1 Model summary...... 125 3.2 Eﬀect of a land tax τ on the equilibrium features of the model...... 134 3.3 Land-use management scenarios...... 139 3.4 Overshoot management scenarios...... 140

xi List of Tables

1.1 Parameters’ deﬁnition and default values ...... 51

A2.1 Functions and aggregate parameters ...... 111 A2.2 Deﬁnition, units and default values of the parameters ...... 111

3.1 Deﬁnition and default values of the parameters...... 129 3.2 Functions and aggregate parameters...... 130

xii GENERAL INTRODUCTION

1 GENERAL INTRODUCTION

Context

Nature and society have long been perceived as separate and independent. This historical view was initiated during the Enlightment era [1], during which developments in science and technology enhanced people’s abilities to control or transform nature, “from desert wilderness to cultivated garden” [2]. The forces of industrialization and urbanization further split humans from their environments, apparently decoupling human population growth from natural constraints. At that time, this decoupling between humans and nature was already questioned by Malthus, who enlightened the limits of exponential population growth on a finite planet [3]. However, views of humans as limited by nature remained marginal [4], and change of paradigm started only recently [5]. There is now mounting evidence that humans and nature influence each other through bidirectional interactions [6, 7, 8]. Yet the coupling of human societies and ecosystem dynamics is not fully recognized nor accounted for in scientific research and public policies [9]. In this coupled human–nature dynamics, human demography and biodiversity play major roles.

Human population growth and scenarios

Human population has been growing tremendously since the 1950s, mainly due to advances in public health, massive consumption of fossil fuels, and increases in agricultural productivity. The rate of population growth has been declining since the 1980s, while the absolute total numbers kept increasing. As of mid-2017, the world population reached 7.6 billion, and its annual growth rate has fallen from 1.24 % to 1.10% between 2005 and 2017, due to fertility declines in most areas of the world. Between 2010 and 2015, countries where fertility is below replacement levels accounted for 46 % of the global population [10].

Under the most likely United Nations population scenarios, the world population is projected to reach 8.6 billion in 2030, and to increase further to 9.8 billion in 2050 and 11.2 billion by 2100 [10]. The same scenarios suggest that global population is likely to continue to rise later in the century, with only little chance (23 %) to stabilize or begin

2 GENERAL INTRODUCTION to fall before 2100.

However, these population scenarios are highly sensitive to future fertility and population aging assumptions. Several demographic uncertainties, such as international migration and the structure of families, cast doubts upon the reliability of these population projections [11]. Other demographic models that expand conventional approaches via probabilistic treatments of fertility, mortality, and migration support a peak population forecast at around 2070 followed by a decline [12]. Despite demographic transition and fertility reductions [10, 13], a stabilization of the world population seems unlikely this century [14, 15].

Moreover, human population projections have poorly accounted for potential ecological feedbacks on human population growth so far, despite the scale of impact of human activities on the environment [16]. Changes in the natural resources and the Earth System (e.g., climate change) have important feedback eﬀects on human societies [9]. However, current models do not incorporate these critical feedbacks, and often take human demography and economic growth as exogenous drivers, using estimates such as the UN population projections.

Accounting for the exhaustion of non-renewable natural resources on the industrial sector, the recent “Limits to Growth” updated report [17] predicts an overshoot and eventual collapse of food production, human population and living standards by 2030- 2040, under the business-as-usual scenario. Similarly, another study predicts the decline of agricultural production between 2030-2060 due to water scarcity [18]. More generally, concerns about a potential global collapse of our society [19] due to overshooting Earth’s limits, such as biodiversity and essential ecosystem services, are raised in a number of studies [20, 7, 8, 21].

Population growth and natural habitat destruction

Human use of land has transformed ecosystems across most of the terrestrial biosphere for millennia [23]. The conversion of natural lands to croplands, pastures, and urban areas represents the most visible form of human impact on the environment [24], with

3 GENERAL INTRODUCTION

Figure A: Human population growth and global land-use change. Left panel: Pop- ulation of the world, estimates (1950-2017) and medium-variant projection (2017-2100) with 95% conﬁdence intervals. Extracted from the United Nations’ World Population Prospects: the 2017 Revision [10]. Right panel: Projected land-use changes until 2100. Extracted from Newbold et al. (2015) [22].

40% of Earth’s land surface being currently under agriculture [25], and 75% experiencing measurable human pressures [26]. These pressures are rapidly intensifying in biodiversity- rich places [27], since most land conversion occurs in the tropics through forest conversion to agriculture [28, 29].

In these biodiverse and ecologically fragile areas, high fertility and associated rapid population growth directly contribute to land conversion, such as in Central America [30]. A reason for this high fertility is that children constitute an asset to farm families that are often short on labor [31], and guarantee families a certain number of surviving children despite high rates of child mortality [32]. Such vicious circles of sustained high fertility in the face of declining environmental resources [33] can result in “tragedy of the commons” situations, where overuse of natural resources leads to their collapse [34].

More generally, the relationship between human population dynamics and natural habitat loss is highly dependent on the scale of analysis [30]. At local scales, the eﬀect can be ambiguous [35, 36], and population dynamics usually acts in concert with other signiﬁcant factors such as local institutions, policies, economic globalization, and cultural change [37, 38].

At the global level, human population growth and consumption are the two major drivers of humanity’s ecological footprint [16]. Between 2000 and 2016, the human pop-

4 GENERAL INTRODUCTION ulation has increased by 23% and the world economy has grown 153%, while the human footprint has increased by just 9%, due to increased resource use efficiency [26]. Build- ing upon such efficiency gains, some growth economists advocate for the potential of economic growth to improve environmental quality, following an initial phase of deterioration [39]. However, the benefits of rising resource use efficiency and the demographic transition [40, 41, 13] can be undermined by the global dietary transition towards more land demanding diets [42, 43], as well as by perverse economic rebound effects due to a higher opportunity cost of land conservation [44, 45], both resulting in further habitat destruction, despite a slower human population growth and a higher production efficiency.

Impacts of land conversion on species richness

Domination of ecosystems by humans is increasingly taking over the primary productivity of land (30% of terrestrial productivity), mainly through land conversion and deforestation [46]. Natural habitat loss is threatening more and more species and the ecosystem functions that they provide [47]. Global species extinction rates are about 1000 times the likely background rate of extinction [48, 49]. The Living Planet Index shows a 58% decline in populations monitored between 1970 and 2012 (Fig. B). This trend is likely to continue in the 21st century [50], and could result in the loss of two thirds of species populations by 2020 compared to 1970 levels [51, 52].

At local spatial scales, the worst impacted habitats show an average 76.5% reduction in local species richness (Fig. C), compared to a 13.6% average reduction globally [22]. Despite recent data syntheses ﬁnding mixed evidence for patterns of net species loss at local spatial scales, thus arguing for a no net change in local species richness [53, 54], there is growing evidence for a negative relationship between land use changes and local biodiversity loss [55, 56].

Additionally to natural habitat loss, habitat fragmentation is causing rapid species loss [57, 58] and alteration of ecosystem functioning [59]. Fragmentation of natural habitat ampliﬁes the negative eﬀect of habitat loss on species extinctions [60], and may double biodiversity loss from deforestation in tropical forests [61].

5 GENERAL INTRODUCTION

Figure B: The Global Living Planet Index shows a decline of 58 % (range: - 48 to -66 %) between 1970 and 2012. Trend in population abundance for 14,152 populations of 3,706 species monitored across the globe between 1970 and 2012. The white line shows the index values and the shaded areas represent the 95 % conﬁdence limits surrounding the trend. Extracted from the WWF report (2016) [51].

Figure C: Net change in local richness caused by land use and related pressures by 2000. Extracted from Newbold et al. (2015) [22].

6 GENERAL INTRODUCTION

Biodiversity, ecosystem functioning and agricultural productivity

Species richness, i.e., the numbers of species, is a well-studied surrogate for several other dimensions of biodiversity, such as taxonomic/phylogenetic distinctiveness, and functional diversity [62]. Species extinctions are altering key processes important to the productivity of Earth’s ecosystems [63], and further species loss will accelerate change in ecosystem processes [64, 65]. Local species richness underpins many ecosystem functions, such as the biogeochemical processes that regulate the Earth system [66]. Thus, changes in biodiversity that alter ecosystem processes can have profound consequences for services that humans derive from ecosystems [46]. The ecosystem consequences of local species loss are as quantitatively significant as the direct effects of several global change stressors, such as ozone, acidification, elevated CO2, and nutrient pollution [67]. Quantitative assessments of the relationship between biodiversity and ecosystem process rates show clear evidence of a positive effect of biodiversity for most ecosystem services at the community level [68] and larger spatial scales [69]. Complementarity between species’ patterns of resource use can increase average rates of productivity and nutrient retention [63]. Most experiments indicate that the relationship between ecosystem process rates and species richness saturates at higher levels of diversity, suggesting redundancy. However, although species may appear functionally redundant when one function is considered under one set of environmental conditions, many species are needed to maintain multiple functions at multiple times and places in changing environments [70]. Such a complementarity between species and functions across time scales results in a stabilizing effect of diversity on ecosystem processes [71].

Evidence that land conversion and biodiversity loss are threatening the provisioning of essential ecosystem services [6] has contributed to raise concerns about the capacity of the biosphere to provide goods and services in the long term [72], especially regarding the growing demand for food production [73]. Important biodiversity-dependent services to agricultural production include biological control of pests [74], crop pollination [75], erosion limitation [76], water quality [77] and soil quality [78].

Globally, about 10% of agricultural yields are estimated to be destroyed by animal

7 GENERAL INTRODUCTION pests, and estimated crop losses to pests have significantly increased in the last 40 years, despite a 15- to 20-fold increase in chemical pesticide use [79]. Successful biological control of pests by natural enemies is thus of key economic and ecological importance [80, 81], and is responsible for an estimated 50–90% of the biological pest control occurring in crop fields [82]. Along with biodiversity loss, the simplification of agricultural landscapes in favor of a domination by arable cropland affects the exchange of services between crop and non-crop habitats [83, 74]. Landscape complexity has been shown to favor both the diversity and abundance of natural enemies [84], compared to simple and intensely cultivated landscapes [74, 83]. However, its effect on service provision compared to simple landscapes can be ambiguous [85], due to changes in trophic interactions between diverse enemy assemblages [86].

Similarly, there is mounting evidence that pollinator decline has signiﬁcant economic consequences in many agricultural areas, with a total economic value of pollination estimated at €153 billion, or 9.5% of the value of the world agricultural production used for human food in 2005 [75]. Pollinator decline also aﬀects consumers surplus, especially regarding vegetable, fruits, edible oil crops, and nuts, up to €50 billions in each cate- gory. Consumption of these goods is predicted to sharply decline compared to current consumption levels. Functional group diversity of bee pollinators increases crop yield through niche complementarity [87]. Natural habitat loss and isolation reduce pollination services despite abundance of honey bees [88], emphasizing the importance of wild pollinator conservation.

The soil organism community also aﬀects land productivity through a wide range of ecosystem services that are essential to the sustainable function of natural and managed ecosystems [78]. Soil organisms have direct and indirect impacts on agricultural productivity, e.g., through carbon and nutrient cycles and decomposition, soil structure modiﬁcation, and food web interactions [89]. Soil erosion is a major environmental threat to the productive capacity of agriculture, and enhances nutrient loss, which reduces the fertility of remaining soils [90], with important economic consequences [76]. Each year about 10 million ha of cropland are lost due to soil erosion, 10 to 40 times faster than the

8 GENERAL INTRODUCTION rate of soil renewal, thus imperiling future human food security and environmental quality [91]. Enhancing functional diversity in agroecosystems can help secure crop protection and soil fertility [92]. Conservation of biodiversity can also help manage nutrient uptake and storage, and improve water quality by limiting excessive nutrient loading of water bodies [77], since diverse systems capture a greater proportion of biologically available resources such as nitrogen.

Trade-oﬀ between land conversion and food production

Land use changes have enabled humans to appropriate an increasing share of the planet’s resources, but they have undermined the capacity of ecosystems to sustain food production, maintain freshwater and forest resources, and regulate climate and air quality. Thus, we face the challenge of managing trade-oﬀs between immediate human needs and maintaining the capacity of the biosphere to provide goods and services in the long term [72]. Given the current failure to feed humanity, some authors are pessimistic about the capacity of the planet to make the projected 9.7 billion population food-secure and healthy in 2050 [93].

Indeed, in order to meet the world’s future food security and sustainability needs, food production must grow substantially while, at the same time, agriculture’s environmental footprint must shrink dramatically [94, 95]. Solutions include halting agricultural expansion, closing ‘yield gaps’ on underperforming lands, increasing cropping eﬃciency, reducing waste and shifting diets [95, 96]. Together, these strategies could double food production while greatly reducing the environmental impacts of agriculture [42].

However, these solutions poorly account for the dependence of agricultural production on biodiversity. Conventional intensification tends to disrupt beneficial functions of biodiversity by assuming that biodiversity in agroecosystems is functionally negligible, and that technological improvements can substitute for ecosystem services [97]. Intensi- fication of agriculture by use of high-yielding crop varieties, fertilization, irrigation, and pesticides has contributed substantially to the tremendous increases in food production over the past 50 years, at the expense of water quality, toxicity from pesticides, nitrous

9 GENERAL INTRODUCTION oxide emissions, and degradation of habitat for biodiversity [98].

Intensification can also exacerbate land-clearing in the absence of appropriate policies and enforcement, as a consequence of economic rebound effects [44]. Thus, a “land sparing” view [99, 100, 101], i.e., the segregation of land for nature and for production, may not always benefit biodiversity conservation in the long run [102]. Moreover, such a “land sparing” view fails to account for real-world complexity, and especially the dependence of agricultural production on biodiversity-dependent ecosystem services at various spatial scales [103, 104]. The preservation of natural habitats and the services they provide, while enhancing food production and facing the looming land scarcity, is a central challenge for sustainability [44] that requires developing holistic ways to conceptualise challenges related to food, biodiversity, and land scarcity [105, 106].

Human-nature interactions in the Anthropocene

Nature and society as a coupled system

Until recently, the domains of “society” and “nature” were seen as separate and independent. These distinctions have become diﬃcult to maintain given the intensity and scale of the human activities on Earth. The early 70s have witnessed a rising awareness about the ﬁniteness of non-renewable natural resources and the long term consequences of exponential economic and population growth [107, 4]. Although criticized for their relatedness to the Malthusian theory [3], and for underestimating the forward-looking behavior of people [108] and the potential of economic growth to compensate for declines in environmental quality [109, 110, 111, 112], these early studies paved the way for further investigations, and brought human demography into the debate.

The Brundtland report [5] highlighted new environmental concerns, such as deforestation, desertiﬁcation, biodiversity loss, climate change, as well as the necessity to view economy as a subsystem of the larger ecological system. The report also popularized the idea of “sustainability” and “sustainable development“ as a requirement for intra and inter generational justice [113].

10 GENERAL INTRODUCTION

“Sustainable development is a development that meets the needs of the present without compromising the ability of future generations to meet their own needs.”

The Brundtland Report (1987) [5]

The Brundtland report challenged many of the fundamental assumptions of neoclassical economics, such as the assumption that human-made capital is a near-perfect substitute for natural resources. However, its general deﬁnition of sustainable development focusing on human ”needs“ was not straightforward to formalize into economic or ecological terms, thus leading to a variety of sustainability deﬁnitions, both within economics and ecological research.

Economic sustainability and human well-being

Some of the first formal attempts to analyze the new debate in terms of economics defined sustainability as the “non-declining utility of a representative member of society for millennia into the future“ [114]. Following conventional reasoning on sustainability, some authors have argued that internalizing resource market inefficiencies and environmental externalities would allow for a sustainable allocation of human-made and natural capital [115]. This view called “weak sustainability” is seeking to maintain constant the sum of human-made and total natural capital, thus allowing for substitution between them. An alternative approach to sustainable development has focused on natural capital assets, suggesting that they should not decline through time [116]. This “strong sustainability” view seeks to maintain intact natural capital and man-made capital separately, highlighting the limited substitutability between them, and recognizing that natural capital may also provide indirect services to human economies [117].

With the emergence of a sustainability literature following the publication of the Brundtland Report [118, 119], mainstream neoclassical economists have started recognizing that economic growth could not in the long term compensate for declines in environmental quality [120], while introducing the idea of a ﬁnite, but dynamic, human carrying capacity, including many resource limitations originating from available water,

11 GENERAL INTRODUCTION energy, and other ecosystem goods and services [121, 120].

Ecological sustainability and ecosystem resilience

The discovery of multiple basins of attraction in ecosystems in the 1960–1970s inspired environmental scientists to challenge the dominant stable equilibrium view [122]. In the stable equilibrium view, resilience, or “engineering resilience”, measures the time it takes the system to recover from a perturbation. Early focus on near-equilibrium stability has led to inappropriate command-and-control resource management, such as maximum sustainable yields, in situations where slowly changing variables can lead to abrupt changes [123].

Conversely, “ecosystem resilience” has been deﬁned as the propensity of a system to retain its organizational structure following perturbation, and the magnitude of disturbance that can be absorbed before the system shifts towards an alternative equilibrium [123]. This view of ecosystems as resilient and multi stable inspired the deﬁnition of ecological sustainability [124], and the mounting evidence of regime shifts in natural ecosystems [125, 126] has paved the way for the study of social-ecological regime shifts [127, 128, 129]. However, this view of the resilience and stability of ecosystems is heavily debated within ecological researchers for being inconsistent and one-dimensional, thus unable to capture the multidimensional nature of ecological stability [130].

An ecological economics of sustainability

Despite these criticisms [130], the resilience perspective is increasingly used as an approach for understanding the dynamics of social–ecological systems [131]. It emphasizes non- linear dynamics, thresholds, how periods of gradual change interplay with periods of rapid change and how such dynamics interact across temporal and spatial scales [132]. Interactional approaches, such as social-ecological systems (SESs), adaptive cycles [133] and political ecology [134, 135], consider ecological and social components as distinct but interacting across spatial and temporal scales, through various feedbacks. Especially, their focus is on the adaptive capacity of human-nature systems [136], and the implications of

12 GENERAL INTRODUCTION non-convexities [137], slow-fast processes [138] and regime shifts [139] for management.

However, economic sustainability and social-ecological resilience approaches are still largely disjoint [140], since the eﬃciency requirements of the economic sustainability often conﬂict with the stability requirements of the ecological sustainability approach [113, 141]. The development of an ecological economics of sustainability [119] that accounts for biodiversity, non-linearities, the value of ecosystem services [142] and proposes to include natural capital in the concept of wealth [143], has started bridging the gap between economic and ecological sustainability views, though it is still isolated from the main body of contemporary economics and ecological research [144].

Despite these recent advances, feedback loops between population growth and the natural capital are still poorly accounted for by modern growth economists.

Towards integrative sustainability approaches

Mounting evidence of the unsustainability of our current development path calls for the development of an integrative theory for long-term ecological, economic and demographic changes, in order to unravel the complexity of coupled human and natural systems [145]. These changes concern rapidly and slowly changing processes, e.g., ecological, social, cultural [133], that vary across space, time, and organizational units, exhibit nonlinear dynamics with thresholds [138], reciprocal feedback loops [9], time lags [139], resilience [133] and surprises [127, 129].

Developing theory for sustainable futures requires a model of (1) how human and ecological processes interact across temporal scales, (2) how changes in land use and ecosystems feed back on human demography, and (3) how human societies adjust their behavior in response to perceived changes in ecosystems.

Cross-scale biodiversity feedbacks in coupled SESs

Cross-scale inﬂuences have been increasingly studied due to the growing impact of humans on the planet [146]. Scale is usually deﬁned in terms of spatial and temporal dimensions

13 GENERAL INTRODUCTION

[147], and many of the problems resulting from human-nature interactions can be framed as scale mismatches [148]. Scale mismatches occur when the spatial or temporal reach of management does not align with the spatial or temporal reach of the problem being managed, e.g., biodiversity loss [149]. For example, local habitat destruction can generate strong and non-linear eﬀects on ecosystem service provision [150] and stability [151] at larger spatial scales. Such cross-scale spatial eﬀects can lead to “cascading thresholds”, i.e., the tendency of the crossing of one threshold to induce the crossing of other thresholds [152].

Recent studies have focused on spatial scale mismatches, their implications for agro- environmental policies [153] and the conservation of ecosystem services [154]. Only a few studies, however, have considered the consequences of temporal scale mismatches [155]. Given that the scale of human activities increasingly couples fast and slow processes [156], temporal cross-scale interactions cannot be neglected anymore [157]. Sustainability studies require a shift from an economic short-term perspective (years to decades) focusing on fast moving variables, towards a long-term perspective (decades to centuries) accounting for the interaction between slow- and fast-changing variables.

Interaction between slow- and fast-moving processes have been extensively studied in ecology. Slow-fast interactions can lead to sudden shifts in aquatic and terrestrial systems [158]. Studies on lakes [137], coral reefs [138], forests [159] and arid lands [160] have shown that smooth change can be interrupted by sudden drastic switches to a contrasting state [125]. Gradual and slow changes in temperature, nutrient loading, habitat fragmentation or biotic exploitation might have little eﬀect until a threshold is reached at which a large shift occurs [126]. Such shifts can be attributed to alternative stable states [161], and result in large reductions in human well-being that are diﬃcult to reverse [127]. Since alternative states are often undesirable from a sustainability perspective, the consequences of such regime shifts for the management of natural resources has been extensively studied in several systems, including shallow lakes [137] and coral reefs [138]. These studies have emphasized the importance of preserving resilience and building adaptive capacity in systems with slow-fast processes exhibiting potential regime shifts [139].

14 GENERAL INTRODUCTION

In predator-prey systems, slow-fast dynamics can result in boom-bust cycles, e.g., between insect pests and boreal forests [162]. Interaction between a rapidly growing human population and a slowly regenerating natural resource can lead to similar boom- bust cycles, and is one of the explanations for the collapse of the civilization on Easter Island [163]. Time delays between human and ecological dynamics may thus lead to unexpected overshoot-and-collapse cycles in the long run. Time delays can emerge from land conversion and fragmentation due to a decrease in the relaxation rate of population dynamics [164, 58], or resistant life-cycle stages [165]. As a result, species extinctions are postponed in time (Fig. D.a), which generates extinction debts [166, 167]. Relaxation rates decrease with habitat fragmentation [168], until the extinction threshold of the population is reached [169]. Further habitat fragmentation then increases relaxation rates, so that extinction debts are generally larger in large habitat fragments [57].

Estimates suggest that 80% of the species extinctions in the Amazon are still pending [170], which may increase the number of 20th-century extinctions in bird, mammal, and amphibian forest-speciﬁc species by 120%, i.e., more than 140 forest-speciﬁc vertebrates [171]. In Europe, such time delays may have led to underestimate population declines of plants and animals facing medium-to-high extinction risks, since these are more closely matched to indicators of socioeconomic pressures from the early or mid-, rather than the late, 20th century [172]. The negative impact of human activities on current biodiversity may not become fully realized until several decades into the future, which constitutes a challenge for biodiversity conservation as well as a window of conservation opportunities [173]. However, restoration measures also have to account for immigration lags, i.e., time- delayed recovery of biodiversity, that can be even larger than extinction debts [174, 175].

As a result of the relationship between biodiversity and ecosystem functioning, these extinction debts translate into time-delayed loss of biodiversity-dependent ecosystem services (Fig. D.b), such as carbon storage [176], nutrient cycling and biomass [59], and pollination service [177]. Given the importance of biodiversity-dependent ecosystem services for humanity and agricultural production, such functioning debts generate a time-delayed feedback loop between biodiversity loss and human societies. Since lag eﬀects in key

15 GENERAL INTRODUCTION

Figure D: Extinction debt (a) and biodiversity-ecosystem functioning relationship (b). The grey areas capture the magnitude of the extinction and functioning debts, as the diﬀerence between current and long-term species richness (a) or ecosystem function (b), for a given level of habitat destruction (a) or species loss (b). Extracted from Isbell et al. (2015) [176].

ecosystem processes can reinforce undesirable feedbacks [138] and increase vulnerability to social-ecological traps [178], accounting for time delays in current conservation policies and population projections appears crucial.

Breaking the taboo of human demography

The interface of population growth and deterioration of the local resource base has been much neglected by modern demographers and development economists [33], who point to the accumulation of capital and technological progress to discredit potential malthusian overshoot-and-collapse population crises [3, 144]. However, accounting for population growth shows that the accumulation of manufactured capital, knowledge and human capital (health and education) has not compensated for the degradation of natural capital in South Asia and sub-Saharan Africa, as well as in the United Kingdom and the United States [179].

Looking back at human history shows evidence of repetitions of catastrophic disrup- tions [180, 181, 182] followed by long periods of reinvention and development [183, 184]. These crises result from a conjunction of social and environmental events, that reinforce persistent mismatches between the responses of people and their social and ecological

16 GENERAL INTRODUCTION conditions [185]. Understanding the history of how humans have interacted with the rest of nature can help clarify the options for managing our increasingly interconnected global system [186]. Some authors even call for the search for organizing principles in secular demographic cycles to transform history into an analytical, predictive science [187]. Study of the collapse of ancient human societies has become a growing ﬁeld of research [188, 189], with Easter Island being the most famous and well-studied example of societal collapse [190]. Overexploitation of the main natural resource of this paciﬁc island, palm tree, led to an overshoot-and-collapse population crisis that is well reproduced by a classical predator-prey model [163], with a slowly regenerating resource relative to a fast growing human population.

Among the many extensions to this predator-prey model [191], some have shown that technological progress [192] and human foresight [193] would not have prevented the collapse of Easter Island civilization. These studies suggest that modern societies may be vulnerable to such crises as well, and advocate for the importance of demographic adjustments as a mean to prevent such crises [194]. Other extensions emphasize the role of governance failure, tipping points in ecological systems, and positive feedbacks between the economy and the environment, as preconditions for a crisis [195]. More recently, economic stratiﬁcation was also highlighted as a potential cause of collapse in unequal societies [196]. However, models that account for biodiversity feedbacks on human demography are still lacking [9].

Accounting for human behavior and perception

Another crucial aspect of human-nature interaction is human behavioral change and adaptive capacity [133, 197]. Historically, human responses to crises have allowed societies to prevent or adapt to social ecological traps [198]. Including human behavioral change in coupled ecological-economic models can greatly aﬀect the dynamic of the system [199] and is a central aspect of the adaptive capacity and resilience of coupled human-nature systems [131]. Among the many human behavioral theories available [200], rational economic behavior assumes that human decisions seek to maximize short-term ﬁnancial returns and

17 GENERAL INTRODUCTION consumption utility. The behavior of the population arises from the aggregation of individual identical behaviors. Prices and foresight are assumed to transmit present and future resource scarcities to human decisions. However, many externalities of human activities that aﬀect public goods, such as biodiversity, are not reﬂected by market prices. When not internalized through appropriate economic tools, such externalities can result in the overexploitation of common resources [34].

Early research on the interaction between the environment and human populations have thus emphasized the need for government control in order to prevent the overexploitation of natural resources. However, recent empirical evidence has shown that local communities can achieve sustainable resource use through cooperative self-governance [201]. Successful communities often establish social norms, i.e., rules of shared behavior, that protect common natural resources [202] or help achieve group interests [203].

The enforcement of cooperation strongly hinges on the ecological characteristics of SESs. Previous experimental and theoretical studies have emphasized the role of resource productivity and mobility [204] as well as temporal variability [205, 206] on the robustness of cooperation. Time-delayed eﬀects can also reinforce procrastination regarding behavioral shifts [207], and lowering of conservation standards, also known as the “shifting baseline syndrome [208, 209]. The evolution of social norms, by aﬀecting feedbacks and drivers of SESs, may thus lead to large-scale behavioral shifts and help reinforce conventional governmental control [210].

Objectives, methods and main questions

General objective

Societies and ecosystems interact over many spatial and temporal scales [147] and through several feedbacks [9]. However, social and ecological scales are not always aligned [155], and many of the problems encountered by societies in managing natural resources arise because of a mismatch between the scale of management and the scale(s) of the ecological processes being managed [133]. Such scale mismatches can contribute to a decrease in

18 GENERAL INTRODUCTION social-ecological resilience and human well-being [155].

Especially, time delays between ecological and human dynamics, such as extinction debts, generate temporal scale mismatches that aﬀect the feedback loop controlling the dynamics of the system. The aim of this thesis is to investigate theoretically the consequences of such time-delayed biodiversity feedbacks on the sustainability of social-ecological systems.

Theoretical modelling of social-ecological systems

Several approaches have been used to study the sustainability of coupled social-ecological systems [133]. Among them, theoretical modelling allows exploring the long term consequences of the interaction between dynamical variables through abstract simulations. These simulations allow assessing the qualitative eﬀect of a changing ecological, social, or economic parameters of an SES. Such theoretical approaches can help identify important feedbacks and counter-intuitive eﬀects, and design policies to foster sustainability. However, there is a lack of theoretical SES models that simultaneously account for human demography, behavioral change and biodiversity feedbacks, in a sustainability perspective.

In order to account for the feedback loop between biodiversity loss and human population growth, I build upon the literature on human-nature interactions modeled as dynamical predator-prey systems [163, 195, 191, 194, 196]. However, conversely to classical predator-prey models where the natural resource of interest is directly exploited by the predator [163], here I assume an indirect coupling between humans and biodiversity, through the feedback of biodiversity loss on agricultural production. This modelling approach thus challenges classical studies of the trade-oﬀ between food production and biodiversity conservation [99, 101] that poorly account for the biodiversity feedback on agricultural productivity [6].

Human behavior is modeled both through the rationality of economic decisions, and the conformance to social norms. In the basic model structure (chapter 1), the behavior of the population results from the aggregation of individual economic decisions regarding their agricultural and industrial consumptions [194]. In turns, total demand drives

19 GENERAL INTRODUCTION production and land conversion in order to meet the market equilibrium between supply and demand. Since consumption behaviors are not only driven by economic rationality, but are also inﬂuenced by others’ behaviors [206], I extend this basic model structure to account for the inﬂuence of social norms on human decisions (chapter 2).

The ecological compartment of the model includes evidence from theoretical and experimental ecological research. First, the dependence of species extinction on habitat loss is derived from a classical species-area relationship (SAR), one of the best-known patterns in ecology [211]. Power-law SARs (S = cAz) have received good support from ecological data and theory [212], and have been widely used to predict the change of species richness S with area A, where c and z are empirical constants [213]. One general pattern revealed by SAR studies is that, all else being equal, larger areas are expected to support more species. Underlying mechanisms include the null hypothesis of random placement and passive sampling [214], the area per se hypothesis, i.e., lower extinction probability with larger areas [215], and the habitat diversity hypothesis, i.e., larger areas having greater habitat diversity [211].

SARs only capture, however, the equilibrium number of species of a system, but ignore the transient dynamics of species extinctions, which can be time-delayed [166, 169]. Ac- counting for such extinctions debts can be done through the relaxation rates of ecological communities [170]. Theoretical and experimental evidence suggest that these relaxation rates are proportional to the diﬀerence between current and long-term species richness, as captured by SARs [164, 166, 169]. Time-delayed loss of biodiversity is then assumed to feed back on human dynamics through a reduction in agricultural productivity. Finally, we assume a positive and saturating relationship between biodiversity and ecosystem processes, based upon mounting empirical and theoretical evidence [68, 216].

The value of such a minimal conceptual model lies in its analytical tractability and the possibility of fully understanding its dynamics across the parameters’ space. Analysis of both its transient, i.e., far-from-equilibrium, and the equilibrium features allows exploring the qualitative eﬀects of the ecological, social, technological or economic parameters.

20 GENERAL INTRODUCTION

Main questions

In the following chapters, I aim at investigating the long-term consequences of a time- delayed biodiversity feedback on the sustainability of SESs. In chapter 1, I seek to derive a condition on parameters’ space under which the classical definition of sustainability, i.e., non-decreasing human well-being over time [5, 114], is met. Since human well-being depends on biodiversity through the feedback of biodiversity- dependent ecosystem services on agricultural consumption, this sustainability condition integrates both economic and ecological constraints. Such an integrative condition allows assessing the qualitative effects of various parameters of the sustainability of the SES, as well as the existence of critical thresholds in the variables of the system. In chapter 2, I build upon the results of chapter 1 to identify sustainable features of SESs, and investigate conditions under which norm-driven human behavioral change can enforce such sustainable features. I also question the robustness of behavioral changes to time-delayed biodiversity loss, and explore the relationship between time delays and resilience. Despite the potential of norm-driven behavioral change in enforcing large scale shifts [210], government control remains an important tool to enforce sustainability [34]. How- ever, uncertainty regarding natural processes, such as the precise temporal dynamics of biodiversity loss, can undermine the efficiency of public policies. In chapter 3, I explore the efficiency of a land-use management policy in enforcing sustainability, through the internalization of the externality of biodiversity loss on agricultural production, and despite imperfect information regarding ecological time delays.

21 GENERAL INTRODUCTION

Bibliography

1. J Vining, M.S. Merrick, and E.A. Price. The distinction between humans and nature: Human perceptions of connectedness to nature and elements of the natural and unnatural. Human Ecology Review, 15(1), 2008.

2. C. Merchant. Uncommon Ground: Rethinking the Human Place in Nature, chapter Rein- venting Eden: Western culture as a recovery narrative., pages 132–159. New York: W.W. Norton & Company, 1996.

3. T.R. Malthus. An essay on the principle of population. J. Johnson, London, 1798.

4. D. H. Meadows, D. L. Meadows, J. Randers, and W. W. Behrens III. The limits to growth. Universe Books, New York, US, 1972.

5. G. Brundtland, M. Khalid, S. Agnelli, S. Al-Athel, B. Chidzero, L. Fadika, V. Hauﬀ, I. Lang, M. Shijun, M. Morino de Botero, M. Singh, S. Okita, et al. Our Common Future (The Brundtland Report). Oxford University Press, 1987.

6. Bradley J Cardinale, J Emmett Duﬀy, Andrew Gonzalez, David U Hooper, Charles Per- rings, Patrick Venail, Anita Narwani, Georgina M Mace, David Tilman, David A Wardle, Ann P Kinzig, Gretchen C Daily, Michel Loreau, James B Grace, Anne Larigauderie, Di- ane S Srivastava, and Shahid Naeem. Biodiversity loss and its impact on humanity. Nature, 486(7401):59–67, 2012. ISSN 00280836. doi: 10.1038/nature11148.

7. J Rockstrom and et al. A safe operating space for humanity. Nature, 461:472–475, 2009.

8. W. Steﬀen, K. Richardson, J. Rockstrom, S. E. Cornell, I. Fetzer, E. M. Bennett, R. Biggs, S. R. Carpenter, W. de Vries, C. A. de Wit, C. Folke, D. Gerten, J. Heinke, G. M. Mace, L. M. Persson, V. Ramanathan, B. Reyers, and S. Sorlin. Planetary boundaries: Guiding human development on a changing planet. Science, 347(6223):1259855–1259870, 2015.

9. S Motesharrei, J Rivas, E Kalnay, G.R. Asrar, A.J. Busalacchi, R.F. Cahalan, M.A. Cane, R.R. Colwell, K Feng, R.S. Franklin, K Hubacek, F Miralles-Wilhelm, T Miyoshi, M Ruth, R Sagdeev, A Shirmohammadi, J Shukla, J Srebric, V.M. Yakovenko, and N Zeng. Mod- eling sustainability: population, inequality, consumption, and bidirectional coupling of the earth and human systems. National Science Review, 3(4):470–494, 2016. doi: 10.1093/n- sr/nww081.

10. United Nations. World Population Prospects: the 2017 Revision Highlights. New York: UN., 2017.

11. Joel E Cohen. Human population: the next half century. Science, 302(5648):1172–1175, 2003. ISSN 1095-9203. doi: 10.1126/science.1088665.

12. W Lutz, W Sanderson, and S Scherbov. The end of world population growth. Nature, 412: 543–545, 2001. doi: 10.1038/35087589.

13. John Bongaarts. Human population growth and the demographic transition. Philosophical transactions of the Royal Society of London B: Biological sciences, 364(1532):2985–90, 2009. ISSN 1471-2970. doi: 10.1098/rstb.2009.0137.

14. Corey J A Bradshaw and Barry W Brook. Human population reduction is not a quick ﬁx for environmental problems. Proceedings of the National Academy of Sciences, 111(46): 16610–16615, 2014.

22 GENERAL INTRODUCTION

15. Patrick Gerland, a. E. Raftery, H. Ev Ikova, Nan Li, Danan Gu, Thomas Spoorenberg, Leontine Alkema, Bailey K Fosdick, Jennifer Chunn, Nevena Lalic, Guiomar Bay, Thomas Buettner, G. K. Heilig, and J. Wilmoth. World population stabilization unlikely this century. Science, 346(6206):234–237, 2014.

16. T Dietz, E.A. Rosa, and R York. Driving the human ecological footprint. Frontiers in Ecology and the Environment, 5(1):13–18, 2007.

17. D. H. Meadows, D. L. Meadows, and Randers. Limits to growth: the 30 year update. EarthScan, London, Sterling, VA, 2004.

18. S Suweis, A Rinaldo, A Maritan, and D’Odorico. Water-controlled wealth of nations. Proceedings of the National Academy of Sciences, 110(11):4230–4233, 2012.

19. P R Ehrlich and a H Ehrlich. Can a collapse of global civilization be avoided? Proceedings of the Royal Society of London B: Biological Sciences, 280(1754):20122845, 2013.

20. Mathis Wackernagel, Niels B Schulz, Diana Deumling, Alejandro Callejas Linares, Martin Jenkins, Valerie Kapos, Chad Monfreda, Jonathan Loh, Norman Myers, Richard Norgaard, and Jørgen Randers. Tracking the ecological overshoot of the human economy. Proceedings of the National Academy of Sciences of the United States of America, 99(14):9266–71, 2002.

21. Tim Newbold, Lawrence N. Hudson, Andrew P. Arnell, Sara Contu, Adriana De Palma, Simon Ferrier, Samantha L. L. Hill, Andrew J. Hoskins, Igor Lysenko, Helen R. P. Phillips, Victoria J. Burton, Charlotte W. T. Chng, Susan Emerson, Di Gao, Gwilym Pask-Hale, Jon Hutton, Martin Jung, Katia Sanchez-Ortiz, Benno I. Simmons, Sarah Whitmee, Han- bin Zhang, Jörn P. W. Scharlemann, and Andy Purvis. Has land use pushed terrestrial biodiversity beyond the planetary boundary? A global assessment. Science, 353:288–291, 2016.

22. T Newbold, L.N. Hudson, S.L.L. Hill, S Contu, I Lysenko, R.A. Senior, L Börger, D.J. Bennett, A Choimes, B Collen, J Day, A De Palma, S Dr’iaz, S Echeverria-Londono, M.J. Edgar, A Feldman, M Garon, M.L.K. Harrison, T Alhusseini, D.J. Ingram, Y Itescu, J Kattge, V Kemp, L Kirkpatrick, M Kleyer, D.L.P. Correia, C.D. Martin, S Meiri, M Novosolov, Y Pan, H.R.P. Phillips, D.W. Purves, A Robinson, J Simpson, S.L. Tuck, E Weiher, H.J. White, R.M. Ewers, G.M. Mace, J.P.W. Scharlemann, and A Purvis. Global eﬀects of land use on local terrestrial biodiversity. Nature, 520:45–50, 2015.

23. E.C. Ellis, J.O. Kaplan, D.Q. Fullet, S Vavrus, K.K. Goldewijk, and P.H. Verburg. Used planet: A global history. Proceedings of the National Academy of Sciences, 110(20):7978– 7985, 2013.

24. W B Meyer and B L Turner. Human Population Growth and Global Land-Use/Cover Change. Annual Review of Ecology and Systematics, 23(1):39–61, 1992. ISSN 0066-4162. doi: 10.1146/annurev.es.23.110192.000351.

25. EW Sanderson, M Jaiteh, MA Levy, KH Redford, AV Wannebo, and G Woolmer. The human footprint and the last of the wild. BioScience, 52(10):891–904, 2002.

26. O Venter, E.W. Sanderson, A. Magrach, J.R. Allan, J. Beher, K.R. Jones, H.P. Possing- ham, W.F. Laurance, P. Wood, B.M. Fekete, M.A. Levy, and J.E.M Watson. Sixteen years of change in the global terrestrial human footprint and implications for biodiversity conservation. Nature Communications, 7(12558), 2016.

23 GENERAL INTRODUCTION

27. William F. Laurance, Jeﬀrey Sayer, and Kenneth G. Cassman. Agricultural expansion and its impacts on tropical nature. Trends in Ecology & Evolution, 29(2):107–116, 2014.

28. G McGranahan, PJ Marcotullio, X Bai, D Balk, T Braga, and et al. Ecosystems and Human Well-being: Current Status and Trends., chapter Urban systems. Washington, DC: Island, 2005.

29. M C Hansen, P V Potapov, R Moore, M Hancher, S a Turubanova, a Tyukavina, D Thau, S V Stehman, S J Goetz, T R Loveland, a Kommareddy, a Egorov, L Chini, C O Justice, and J R G Townshend. High-resolution global maps of 21st-century forest cover change. Science, 342(6160):850–3, 2013.

30. DL Carr and RE Bilsborrow. Population and land use/cover change: a regional comparison between central america and south america. J Geogr Educ, 43:7–16, 2001.

31. JC Caldwell and P Caldwell. The cultural context of high fertility in sub-saharan africa. Popul Dev Rev, 13(3):409–437, 1987.

32. M Cain. Fertility as an adjustement to risk. Popul Dev Rev, 9(4):688–702, 1983.

33. P Dasgupta. Handbook of Environmental Economics, chapter Population, Poverty, and the natural environment. North Holland, Amsterdam, 2003.

34. G Hardin. The tragedy of the commons. Science, 162:1243–1248, 1968.

35. AS Mather and CL Needle. The relationships of population and forest trends. Geogr J, 166:2–13, 2000.

36. HJ Geist and EF Lambin. Proximate causes and underlying driving forces of tropical deforestation. BioScience, 52(2):143–151, 2002.

37. Eric F. Lambin, B. L. Turner, Helmut J. Geist, Samuel B. Agbola, Arild Angelsen, John W. Bruce, Oliver T. Coomes, Rodolfo Dirzo, Günther Fischer, Carl Folke, P. S. George, Kather- ine Homewood, Jacques Imbernon, Rik Leemans, Xiubin Li, Emilio F. Moran, Michael Mortimore, P. S. Ramakrishnan, John F. Richards, Helle Skånes, Will Steﬀen, Glenn D. Stone, Uno Svedin, Tom A. Veldkamp, Coleen Vogel, and Jianchu Xu. The causes of land- use and land-cover change: Moving beyond the myths. Global Environmental Change, 11 (4):261–269, 2001.

38. A de Sherbinin, D Carr, S Cassels, and L Jiang. Population and environment. Annu Rev Environ Resour, 32:345–373, 2007.

39. GM Grossman and AB Krueger. Economic growth and the environment. The Quaterly Journal of Economics, 110(2):353–377, 1995.

40. O. Galor and D.N. Weil. Population, Technology, and Growth: From Malthusian Stagna- tion to the Demographic Transition and beyond. The American Economic Review, 90(4): 806–828, 2000.

41. James A. Brander. Viewpoint: Sustainability: Malthus revisited? Canadian Journal of Economics, 40(1):1–38, 2007.

42. D. Tilman and M Clark. Global diets link environmental sustainability and human health. Nature, 515:518–522, 2014. doi: 10.1038/nature13959.

24 GENERAL INTRODUCTION

43. Peter Alexander, Mark D.A. Rounsevell, Claudia Dislich, Jennifer R. Dodson, Kerstin Engström, and Dominic Moran. Drivers for global agricultural land use change: The nexus of diet, population, yield and bioenergy. Global Environmental Change, 35:138–147, 2015.

44. Eric F. Lambin and P. Meyfroidt. Global land use change, economic globalization, and the looming land scarcity. Proceedings of the National Academy of Sciences of the United States of America, 108(9):3465–72, 2011. ISSN 10916490. doi: 10.1073/pnas.1100480108.

45. B Phalan, Rhys E Green, Lynn V Dicks, G Dotta, C Feniuk, A Lamb, B.B.N. Strassburg, D.R. Williams, Erasmus K.H.J. zu Ermgassen, and A Balmford. How can higher-yield farming help to spare nature ? Science, 351(6272):450–451, 2016.

46. F.S.III Chapin, E.S. Zavaleta, V.T. Eviner, R.L. Naylor, P.M. Vitousek, H.L. Reynolds, D.U. Hooper, S. Lavorel, O.E. Sala, S.E. Hobbie, M.C. Mack, and S. Díaz. Consequences of changing biodiversity. Nature, 405:234–242, 2000.

47. Peter M Vitousek, Harold a Mooney, Jane Lubchenco, and Jerry M Melillo. Human Domination of Earth’ s Ecosystems. Science, 277:494–499, 1997. ISSN 00368075. doi: 10.1126/science.277.5325.494.

48. S.L. Pimm, C.N. Jenkins, R Abell, T.M. Brooks, J.L. Gittleman, L.N. Joppa, P.H. Raven, C.M. Roberts, and J.O. Sexton. The biodiversity of species and their rates of extinction, distribution and protection. Science, 344:1246752, 2014.

49. G Ceballos, P.R. Ehrlich, A.D. Barnosky, A Garcia, R.M. Pringle, and T.M. Palmer. Accelerated modern human–induced species losses: Entering the sixth mass extinction. Science Advances, 1(5):e1400253, 2015. doi: 10.1126/sciadv.1400253.

50. Henrique M Pereira, Paul W Leadley, Vânia Proença, Rob Alkemade, Jörn P W Scharle- mann, Juan F Fernandez-Manjarrés, Miguel B Araújo, Patricia Balvanera, Reinette Biggs, William W L Cheung, Louise Chini, H David Cooper, Eric L Gilman, Sylvie Guénette, George C Hurtt, Henry P Huntington, Georgina M Mace, Thierry Oberdorﬀ, Carmen Re- venga, Patrícia Rodrigues, Robert J Scholes, Ussif Rashid Sumaila, and Matt Walpole. Scenarios for global biodiversity in the 21st century. Science, 330(6010):1496–1501, 2010.

51. WWF. Living planet report 2016. 2016.

52. D.P. Tittensor, M Walpole, S.L.L. Hill, D.G. Boyce, G.L. Britten, N.D. Burgess, S.H.M. Butchart, P.W. Leadley, E.C. Regan, R. Alkemade, R. Baumung, C Bellard, L Bouwman, N.J. Bowles-Newark, A.M. Chenery, W.W.L. Cheung, V Christensen, H.D. Cooper, A.R. Crowther, M.J.R. Dixon, A Galli, V Gaveau, R.D. Gregory, N.L. Gutierrez, T.L. Hirsch, R Höft, S.R. Januchowski-Hartley, M Karmann, C.B. Krug, F.J. Leverington, J Loh, R.K. Lojenga, K Malsch, A Marques, D.H.W. Morgan, P.J. Mumby, T Newbold, K Noonan- Mooney, S.N. Pagad, B.C. Parks, H.M. Pereira, T Robertson, C Rondinini, L Santini, J.P.W. Scharlemann, S Schindler, U.R. Sumaila, L.S.L Teh, J van Kolck, P Visconti, and Y Ye. A mid-term analysis of progress toward international biodiversity targets. Science, 346(6206):241–244, 2014. doi: 10.1126/science.1257484.

53. M Vellend, L Baeten, Isla H. Myers-Smith, S.C. Elmendorf, R Beauséjour, C.D. Brown, P De Frenne, K Verheyen, and S Wipf. Global meta-analysis reveals no net change in local-scale plant biodiversity over time. Proceedings of the National Academy of Sciences, 110(48):19456–9, 2013.

25 GENERAL INTRODUCTION

54. M Dornelas, N.J. Gotelli, B McGill, H Shimadzu, F Moyes, C Sievers, and A.E. Magurran. Assemblage time series reveal biodiversity change but not systematic loss. Science, 344: 296–299, 2014.

55. A Gonzalez, B.J. Cardinale, G.R.H Allington, J Byrnes, K.A. Endsley, D.G. Brown, D.U. Hooper, F Isbell, M.I. O’Connor, and M Loreau. Estimating local biodiversity change: a critique of papers claiming no net loss of lacol diversity. Ecology, 97(8):1949–60, 2016.

56. Mark Vellend, Maria Dornelas, Lander Baeten, Robin Beauséjour, Carissa D. Brown, Pieter De Frenne, Sarah C. Elmendorf, Nicholas J. Gotelli, Faye Moyes, Isla H. Myers-Smith, Anne E. Magurran, Brian J. McGill, Hideyasu Shimadzu, and Caya Sievers. Estimates of local biodiversity change over time stand up to scrutiny. Ecology, 98(2):583–590, 2017.

57. G Ferraz, G.J. Russell, P.C. Stouﬀer, R.O. Bierregaard, S.L. Pimm, and T.E. Lovejoy. Rates of species loss from amazonian forest fragments. Proceedings of the National Academy of Sciences, 100(24):14069–14073, 2003.

58. G. Ferraz, J.D. Nichols, J.E. Hines, P.C. Stouﬀer, Jr. Bierregaard, and T.E. Lovejoy. A large-scale deforestation experiment: Eﬀects of patch area and isolation on Amazon birds. Science, 315(5809):238–241, 2007.

59. N. M. Haddad, L. a. Brudvig, J. Clobert, K. F. Davies, a. Gonzalez, R. D. Holt, T. E. Lovejoy, J. O. Sexton, M. P. Austin, C. D. Collins, W. M. Cook, E. I. Damschen, R. M. Ewers, B. L. Foster, C. N. Jenkins, a. J. King, W. F. Laurance, D. J. Levey, C. R. Mar- gules, B. a. Melbourne, a. O. Nicholls, J. L. Orrock, D.-X. Song, and J. R. Townshend. Habitat fragmentation and its lasting impact on Earth’s ecosystems. Science Advances, 1 (2):e1500052–e1500052, 2015. ISSN 2375-2548. doi: 10.1126/sciadv.1500052.

60. I Hanski, G.A. Zurita, M.I. Bellocq, and J. Rybicki. Species-fragmented area relationship. Proceedings of the National Academy of Sciences, 110(31):12715–12720, 2013.

61. J Barlow, G.D Lennox, J Ferreira, E Berenguer, A. C. Lees, Ralph Mac Nally, James R. Thomson, Silvio Frosini de Barros Ferraz, Julio Louzada, Victor Hugo Fonseca Oliveira, Luke Parry, Ricardo Ribeiro de Castro Solar, Ima C. G. Vieira, Luiz E. O. C. Aragão, Rodrigo Anzolin Begotti, Rodrigo F. Braga, Thiago Moreira Cardoso, Raimundo Cosme de Oliveira Jr, Carlos M. Souza Jr, Nárgila G. Moura, Sâmia Serra Nunes, João Vic- tor Siqueira, Renata Pardini, Juliana M. Silveira, Fernando Z. Vaz-de Mello, Ruan Carlo Stulpen Veiga, Adriano Venturieri, and Toby A. Gardner. Anthropogenic disturbance in tropical forests can double biodiversity loss from deforestation. Nature, 535: 144–147, 2016. doi: 10.1038/nature18326.

62. Forest Isbell, Andrew Gonzalez, Michel Loreau, Jane Cowles, Sandra Díaz, Andy Hector, Georgina M. Mace, David A. Wardle, Mary I. O’Connor, J. Emmett Duﬀy, Lindsay A. Turnbull, Patrick L. Thompson, and Anne Larigauderie. Linking the inﬂuence and dependence of people on biodiversity across scales. Nature, 546:65–72, 2017.

63. D. U. Hooper, F. S. Chapin, J. J. Ewel, A. Hector, P. Inchausti, S. Lavorel, J. H. Lawton, D. M. Lodge, M. Loreau, S. Naeem, B. Schmid, H. Setälä, A. J. Symstad, J. Vandermeer, and D. A. Wardle. Eﬀect of biodiversity on ecosystem functioning: a consensus of current knowledge. Ecological Monographs, 75(1):3–35, 2005.

64. D.A. Wardle, R.D. Bardgett, R.M. Callaway, and W.H. Van der Putten. Terrestrial ecosystem responses to species gains and losses. Science, 332:1273–1277, 2011.

26 GENERAL INTRODUCTION

65. C Perrings, S Naeem, F.S. Ahrestani, D.E. Bunker, P Burkill, G Canziani, T Elmqvist, J.A. Fuhrman, F.M. Jaksic, Z Kawabata, A Kinzig, G.M. Mace, H Mooney, A.-H. Prieur- Richard, J Tschirhart, and W Weisser. Ecosystem services, targets, and indicators for the conservation and sustainable use of biodiversity. Front. Ecol. Environ., 9:512–520, 2011.

66. M Loreau, S Naeem, P Inchausti, J Bengtsson, J.P. Grime, A Hector, D.U. Hooper, M.A. Huston, D Raﬀaelli, B Schmid, D Tilman, and D.A. Wardle. Biodiversity and ecosystem functioning: current knowledge and future challenges. Science, 294:804–808, 2001.

67. D.U. Hooper, E.C. Adair, B.J. Cardinale, J.E.K Byrnes, B.A. Hungate, K.L. Matulich, A Gonzalez, J.E. Duﬀy, L Gamfeldt, and M.I. O’Connor. A global synthesis reveals biodiversity loss as a major driver of ecosystem change. Nature, 486:105–108, 2012.

68. P Balvanera, A.B. Pfisterer, N Buchmann, J.-S. He, T Nakashizuka, D Raffaelli, and B Schmid. Quantifying the evidence for biodiversity effects on ecosystem functioning and services. Ecology Letters, 9(10):1146–1156, 2006.

69. U Brose and H Hillebrand. Biodiversity and ecosystem functioning in dynamic landscapes. Philosophical Transactions of the Royal Society B, 371:20150267, 2016.

70. F Isbell, V Calcagno, A Hector, J Connolly, W.S. Harpole, P.B. Reich, M Scherer-Lorenzen, B Schmid, D Tilman, J van Ruijven, A Weigelt, B.J. Wilsey, E.S. Zavaleta, and M Loreau. High plant diversity is needed to maintain ecosystem services. Nature, 477:199–202, 2011.

71. Michel Loreau and Claire de Mazancourt. Biodiversity and ecosystem stability: A synthesis of underlying mechanisms. Ecology Letters, 16(s1):106–115, 2013.

72. Jonathan A Foley, Ruth Defries, Gregory P Asner, Carol Barford, Gordon Bonan, Stephen R Carpenter, F Stuart Chapin, Michael T Coe, Gretchen C Daily, Holly K Gibbs, Joseph H Helkowski, Tracey Holloway, Erica A Howard, Christopher J Kucharik, Chad Monfreda, Jonathan A Patz, I Colin Prentice, Navin Ramankutty, and Peter K Snyder. Global consequences of land use. Science (New York, N.Y.), 309(5734):570–574, 2005.

73. G Daily, P Dasgupta, B Bolin, P Crosson, J du Guerny, P Ehrlich, C Folke, A-M Jans- son, B-O Jansson, N Kautsky, A Kinzig, S Levin, K-G Mäler, P Pinstrup-Andersen, and B Walker. Food production, population growth, and the environment. Science, 281:1291– 1292, 1998.

74. FJ Bianchi, CJ Booij, and T Tscharntke. Sustainable pest regulation in agricultural landscapes: A review on landscape composition, biodiversity and natural pest control. Pro- ceedings of the Royal Society B, 273(1595):1715–1727, 2006.

75. N Gallai, J Salles, J Settele, and B Vaissière. Economic valuation of the vulnerability of world agriculture confronted with pollinator decline. Ecological Economics, 68(3):810–821, 2009.

76. D Pimentel, C Harvey, P Resosudarmo, K Sinclair, D Kurz, M McNair, S Crist, L Shpritz, L Fitton, R Saﬀouri, and R Blair. Environmental and economic costs of soil erosion and conservation beneﬁts. Science, 267(5201):1117–23, 1995.

77. B Cardinale. Biodiversity improves water quality through niche partitioning. Nature, 472: 86–89, 2011.

78. E Barrios. Soil biota, ecosystem services and land productivity. Ecological Economics, 64 (2):269–285, 2007.

27 GENERAL INTRODUCTION

79. EC Oerke. Crop losses to pests. Journal of Agricultural Sciences, 144(1):31–43, 2006.

80. WJ Lewis, JC van Lenteren, SC Phatak, and JH Tumlinson. A total system approach to sustainable pest management. Proceedings of the National Academy of Sciences USA, 94 (23):12243–12248, 1997.

81. JE Losey and M Vaughan. The economic value of ecological services provided by insects. Bioscience, 56(4):311–323, 2006.

82. D Pimentel. Environmental and economic costs of the application of pesticides primarily in the united states. Environ Dev Sustain, 7(2):229–252, 2005.

83. R Chaplin-Kramer, ME O’Rourke, EJ Blitzer, and C Kremen. A meta-analysis of crop pest and natural enemy response to landscape complexity. Ecology Letters, 14(9):922–932, 2011.

84. MM Gardiner, DA Landis, C Gratton, CD DiFonzo, M O’Neal, JM Chacon, MT Wayo, NP Schmidt, EE Mueller, and GE Heimpel. Landscape diversity enhances biological control of an introduced crop pest in the north-central usa. Ecological Applications, 19(1):143–154, 2009.

85. C Thies and T Tscharntke. Landscape structure and biological control in agro-ecosystems. Science, 285(5429):893–895, 1999.

86. Emily A Martin, Björn Reineking, Bumsuk Seo, and Ingolf Steﬀan-Dewenter. Natural enemy interactions constrain pest control in complex agricultural landscapes. Proceed- ings of the National Academy of Sciences, 110(14):5534–5539, 2013. doi: 10.1073/p- nas.1215725110.

87. P Hoehn, T Tscharntke, J Tylianakis, and I Steﬀan-Dewenter. Functional group diversity of bee pollinators increases crop yield. Proceedings. Biological sciences / The Royal Society, 275(1648):2283–91, 2008.

88. L Garibaldi, I Steffan-Dewenter, C Kremen, J Morales, R Bommarco, S Cunningham, L Carvalheiro, N Chacoff, J Dudenhffer, S Greenleaf, A Holzschuh, R Isaacs, K Krewenka, Y Mandelik, M Mayfield, L Morandin, S Potts, T Ricketts, H Szentgyrgyi, B Viana, C Westphal, R Winfree, and A Klein. Stability of pollination services decreases with isolation from natural areas despite honey bee visits. Ecology Letters, 14(10):1062–1072, 2011.

89. IT Handa, R Aerts, F Berendse, MP Berg, A Bruder, O Butenschoen, E Chauvet, MO Gess- ner, J Jabiol, M Makkonen, BG McKie, B Malmqvist, ET Peeters, S Scheu, B Schmid, J van Ruijven, VC Vos, and S Hättenschwiler. Consequences of biodiversity loss for litter decomposition across biomes. Nature, 509:218–21, 2014.

90. D Pimentel and N Kounang. Ecology of soil erosion in ecosystems. Ecosystems, 1(5): 416–426, 1998.

91. D Pimentel. Soil erosion: a food and environmental threat. Environment, development and sustainability, 8(1):119–137, 2006.

92. M.A. Altieri. The ecological role of biodiversity in agroecosystems. Agriculture, Ecosystems and Environment, 74(1-3):19–31, 1999.

93. Paul R. Ehrlich and John Harte. Opinion: To feed the world in 2050 will require a global revolution. Proceedings of the National Academy of Sciences, 112(48):14743–14744, 2015.

28 GENERAL INTRODUCTION

94. FAO. How to Feed the World in 2050. Insights from an expert meeting at FAO, (1):1–35, 2009.

95. Jonathan A. Foley, Navin Ramankutty, Kate A. Brauman, Emily S. Cassidy, James S. Ger- ber, Matt Johnston, Nathaniel D. Mueller, Christine O/’Connell, Deepak K. Ray, Paul C. West, Christian Balzer, Elena M. Bennett, Stephen R. Carpenter, Jason Hill, Chad Mon- freda, Stephen Polasky, Johan Rockstrom, John Sheehan, Stefan Siebert, David Tilman, and David P. M. Zaks. Solutions for a cultivated planet. Nature, 478:337–342, 2011.

96. T Kastner, M.J.I Rivas, W Koch, and S Nonhebel. Global changes in diets and the consequences for land requirements for food. Proceedings of the National Academy of Sciences, 109(18):6868–6872, 2012.

97. E. Boserup. The conditions of agricultural growth. Population Studies, 20(1):1–108, 1965.

98. P Matson, W Parton, and M Swift. Agricultural intensiﬁcation and ecosystem properties. Science, 277(5325):504–509, 1997.

99. Rhys E Green, Stephen J Cornell, Jörn P W Scharlemann, and Andrew Balmford. Farming and the fate of wild nature. Science, 307(5709):550–555, 2005.

100. Robert M. Ewers, Jörn P W Scharlemann, Andrew Balmford, and Rhys E. Green. Do increases in agricultural yield spare land for nature? Global Change Biology, 15(7):1716– 1726, 2009.

101. Ben Phalan, Malvika Onial, Andrew Balmford, and Rhys E Green. Reconciling food production and biodiversity conservation: land sharing and land sparing compared. Science, 333(6047):1289–1291, 2011.

102. L. R. Carrasco, C. Larrosa, E. J. Milner-Gulland, and D. P. Edwards. A double-edged sword for tropical forests. Science, 346(6205):38–40, 2014.

103. Teja Tscharntke, Yann Clough, Thomas C. Wanger, Louise Jackson, Iris Motzke, Ivette Perfecto, John Vandermeer, and Anthony Whitbread. Global food security, biodiversity conservation and the future of agricultural intensiﬁcation. Biological Conservation, 151(1): 53–59, 2012.

104. C.D. Mendenhall, D.S. Karp, C.F.J. Meyer, E.A. Hadly, and G.C. Daily. Predicting biodiversity change and averting collapse in agricultural landscapes. Nature, 509:213–217, 2014.

105. Joern Fischer, David J. Abson, Van Butsic, M. Jahi Chappell, Johan Ekroos, Jan Hanspach, Tobias Kuemmerle, Henrik G. Smith, and Henrik von Wehrden. Land sparing versus land sharing: Moving forward. Conservation Letters, 7(3):149–157, 2014. ISSN 1755-263X. doi: 10.1111/conl.12084. URL http://dx.doi.org/10.1111/conl.12084.

106. Jan Christian Habel, Mike Teucher, Berthold Hornetz, Ralph Jaetzold, Josphert N. Ki- matu, Sichangi Kasili, Zachariah Mairura, Ronald K. Mulwa, Hilde Eggermont, Wolf- gang W. Weisser, and Luc Lens. Real-world complexity of food security and biodiversity conservation. Biodiversity and Conservation, 24(6):1531–1539, 2015.

107. P.R. Ehrlich. The population bomb. Ballantine Books, New York, 1968.

108. R Solow. Is the end of the world at hand? Challenge, 16(1):39–50, 1973.

109. E Boserup. The conditions of agricultural growth. London:Earthscan, 1965.

29 GENERAL INTRODUCTION

110. P Dasgupta and G Heal. The optimal depletion of exhaustible resources. Review of Eco- nomic Studies, 41(5):3–28, 1974.

111. J Stiglitz. Growth with exhaustible natural resources: Eﬃcient and optimal growth paths. Review of Economic Studies, 41(5):123–137, 1974.

112. R.M. Solow. Intergenerational equity and exhaustible resources. The Review of Economic Studies, 41(5):29–45, 1974.

113. D Pearce. Foundations of an ecological economics. Ecological Modelling, 38(1-2):9–18, 1987.

114. J. Pezzey. Sustainability: an interdisciplinary guide. Environmental values, 1:321–362, 1992.

115. R Solow. An almost practical step toward sustainability. Resources Policy, 19(3):162–172, 1993.

116. D.A. Pearce, A Markandya, and E.B. Barbier. Blueprint for a Green Economy. London, Earthscan, 1989.

117. R Costanza and H.E. Daly. Natural capital and sustainable development. Conservation Biology, 6(1):37–46, 1992.

118. E.B. Barbier. The concept of sustainable economic development. Environmental Conser- vation, 14(2):101–110, 1987.

119. R Costanza. Ecological Economics: the science and management of sustainability. Columbia University Press, New York, 1991.

120. K Arrow, B Bolin, R Costanza, P Dasgupta, C Folke, C.S. Holling, Bengt-Owe Jansson, Si- mon Levin, K.-G. Mäler, C Perrings, and D Pimentel. Economic growth, carrying capacity, and the environment. Science, 268:520–521, 1995.

121. J E Cohen. Population growth and earth’s human carrying capacity. Science, 269(5222): 341–346, 1995.

122. C.S. Holling. Resilience and stability of ecological systems. Annual Review of Ecology and Systematics, 4:1–23, 1973.

123. C.S. Holling. Engineering within ecological constraints., chapter Engineering resilience versus ecological resilience., pages 31–44. National Academy, Washington, D.C., USA., 1996.

124. C.S. Holling. Sustainable Development of the Biosphere., chapter Resilience of ecosystems; local surprise and global change., pages 292–317. Cambridge University Press, Cambridge, 1986.

125. M Scheﬀer, S Carpenter, J.A. Foley, C Folke, and B. Walker. Catastrophic shifts in ecosystems. Nature, 413:591–596, 2001.

126. M Scheﬀer and S.R. Carpenter. Catastrophic regime shifts in ecosystems: linking theory to observation. Trends in Ecology and Evolution, 18(12):648–656, 2003.

127. M Scheﬀer. Critical Transitions in Nature and Society. Princeton and Oxford, 2009.

30 GENERAL INTRODUCTION

128. S.J. Lade, A. Tavoni, S.A. Levin, and M Schlüter. Regime shifts in a social-ecological system. Theoretical Ecology, 6(3):359–372, 2013. doi: 10.1007/s12080-013-0187-3.

129. Chris T Bauch, Ram Sigdel, Joe Pharaon, and Madhur Anand. Early warning signals of regime shifts in coupled human–environment systems. Proceedings of the National Academy of Sciences, 113(51):14560–14567, 2016.

130. I Donohue, H Hillebrand, J.M. Montoya, O.L. Petchey, S.L. Pimm, M.S. Fowler, K Healy, A.L. Jackson, M Lurgi, D McClean, N.E. O’Connor, E O’Gorman, and Q Yang. Navigating the complexity of ecological stability. Ecology Letters, 19(9):1172–1185, 2016.

131. C Folke, J Colding, and F Berkes. Navigating social-ecological systems: building resilience for complexity and change., chapter Synthesis: building resilience and adaptive capacity in social–ecological systems. Cambridge University Press, New York, 2003.

132. C Folke. Resilience: the emergence of a perspective for social-ecological systems analyses. Global Environmental Change, 16:253–267, 2006.

133. L.H. Gunderson and C.S. Holling. Panarchy: understanding transformations in human and natural systems. Island Press, 2001.

134. CL Jolly. Four theories of population change and the environment. Popul Environ, 16(1): 61–90, 1994.

135. LC Gray and WG Moseley. A geographical perspective on poverty-environment interactions. Geogr J, 171(1):9–23, 2005.

136. S.R. Carpenter and W.A. Brock. Adaptive capacity and traps. Ecology and society, 13(2): 40, 2008.

137. KG. Mäler, A. Xepapadeas, and A. De Zeeuw. The Economics of Non-Convex Ecosystems., chapter The economics of Shallow Lakes., pages Springer, Dordrecht. 2004.

138. A.-S. Crepin. Using fast and slow processes to manage resources with thresholds. Envi- ronmental and Resource Economics, 36:191–213, 2007.

139. A.-S. Crepin, R Biggs, S Polasky, M Troell, and A de Zeeuw. Regime shifts and management. Ecological Economics, 84:15–22, 2012.

140. S Derissen, M.F. Quaas, and S Baumgärtner. The relationship between resilience and sustainability of ecological-economic systems. Ecological Economics, 70:1121–1128, 2011.

141. M Common and C Perrings. Towards an ecological economics of sustainability. Ecological economics, 6(1):7–34, 1992.

142. P Dasgupta, SA Levin, and J Lubchenco. Economic pathways to ecological sustainability: challenges for the new millennium. BioScience, 50(4):339–345, 2000.

143. P Dasgupta. Human well-being and the natural environment. Oxford, UK: Oxford Univer- sity Press., 2001.

144. P Dasgupta. Nature in economics. Environmental and Resource Economics., 39(1):1–7, 2008.

31 GENERAL INTRODUCTION

145. Jianguo Liu, Thomas Dietz, Stephen R. Carpenter, Marina Alberti, Carl Folke, Emilio Moran, Alice N. Pell, Peter Deadman, Timothy Kratz, Jane Lubchenco, Elinor Ostrom, Zhiyun Ouyang, William Provencher, Charles L. Redman, Stephen H. Schneider, and William W. Taylor. Complexity of coupled human and natural systems. Science, 317: 1513–1516, 2007.

146. S.A. Levin. Fragile dominion: complexity and the commons. Perseus Books, Reading, MA, 1999.

147. S.A. Levin. The problem of pattern and scale in ecology. Ecology, 73:1943–1967, 1992.

148. A.M. Guerrero, R.R.J. McAllister, J Crocorna, and K.A. Wilson. Scale mismatches, conservation planning, and the value of social-network analyses. Conservation Biology, 27(1): 35–44, 2013.

149. F Padt, P Opdam, N Polman, and C Termeer, editors. Scale-sensitivity Governance of the Environment. Wiley-Blackwell, 2014.

150. M.G.E. Mitchell, E.M. Bennett, and A. Gonzalez. Strong and non-linear eﬀects of fragmentation on ecosystem service provision at multiple scales. Environmental Research Letters, 10(9), 2015.

151. S Wang and M Loreau. Biodiversity and ecosystem stability across scales in metacommu- nities. Ecology Letters, 19(5):510–518, 2016.

152. A Kinzig, P Ryan, M Etienne, H Allison, T Elmqvist, and B.H. Walker. Resilience and regime shifts: Assessing cascading eﬀects. Ecology and Society, 11(1):20, 2006.

153. C Pelosi, M Goulard, and G Balent. The spatial scale mismatch between ecological processes and agricultural management: Do diﬃculties come from underlying theoretical frameworks? Agriculture, Ecosystems & Environment, 139(4):455–462, 2010.

154. A Satake, T.K. Rudel, and A Onuma. Scale mismatches and their ecological and economic eﬀects on landscapes: A spatially explicit model. Global Environmental Change, 18(4): 768–775, 2008.

155. G.S. Cumming, D.H.M. Cumming, and C.L. Redman. Scale mismatches in social-ecological systems. Ecology and Society, 11(1):14, 2006.

156. J. Liu, V. Hull, M. Batistella, R. DeFries, T. Dietz, F. Fu, T.W. Hertel, R.C. Izaurralde, E.F. Lambin, S. Li, L.A. Martinelli, W.J. McConnell, E.F. Moran, R. Naylor, Z. Ouyang, K.R. Polenske, A Reenberg, G. de Miranda Rocha, C.S. Simmons, P.H. Verburg, P.M. Vitousek, F. Zhang, and C. Zhu. Framing sustainability in a telecoupled world. Ecology and Society, 18(2):26, 2013.

157. R.V. O’Neill, J.R. Kahn, and C.S. Russell. Economics and ecology: the need for detente in conservation ecology. Conservation ecology, 2(1):4, 1998.

158. S Rinaldi and M Scheﬀer. Geometric analysis of ecological models with slow and fast processes. Ecosystems, 3(6):507–521, 2000.

159. A.-S. Crepin. Multiple species boreal forests: what faustmann missed. Environmental and Resource Economics, 26(4):625–646, 2003.

160. S Kéﬁ, M Rietkerk, M van Baalen, and M Loreau. Local facilitation, bistability and transitions in arid ecosystems. Theoretical Population Biology, 71(3):367–379, 2007.

32 GENERAL INTRODUCTION

161. S Carpenter, W Brock, and P Hanson. Ecological and social dynamics in simple models of ecosystem management. Conservation Ecology, 3(2):4, 1999.

162. S Rinaldi. Limit cycles in slow-fast forest-pest models. Theoretical Population Biology, 41 (1):26–43, 1992.

163. James A. Brander and M. Scott Taylor. The Simple Economics of Easter Island: A Ricardo- Malthus Model of Renewable Resource Use. American Economic Review, 88(1):119–138, 1998.

164. J.M. Diamond. Biogeographic kinetics: estimation of relaxation times for avifaunas of southwest paciﬁc islands. Proc.Nat.Acad.Sci. USA, 69(11):3199–3203, 1972.

165. K Hylander and J Ehrlen. The mechanisms causing extinction debts. Trends in Ecology and Evolution, 28(6):341–346, 2013.

166. David Tilman, Robert M. May, Clarence L. Lehman, and Martin A. Nowak. Habitat destruction and the extinction debt. Nature, 371(6492):65–66, 1994.

167. J Krauss, R Bommarco, M Guardiola, R.S. Heikkinen, A Helm, M Kuussaari, R Lindborg, E Öckinger, M Pärtel, J Pino, J Pöyry, K.M. Raatikainen, A Sang, C Stefanescu, T Teder, M Zobel, and I Steﬀan-Dewenter. Habitat fragmentation causes immediate and time- delayed biodiversity loss at diﬀerent trophic levels. Ecology Letters, 13(5):597–605, 2010.

168. Otso Ovaskainen and Ilkka Hanski. Transient dynamics in metapopulation response to perturbation. Theoretical population biology, 61(3):285–295, 2002.

169. Ilkka Hanski and Otso Ovaskainen. Extinction debt at extinction threshold. Conservation Biology, 16(3):666–673, 2002.

170. O. R. Wearn, D. C. Reuman, and R. M. Ewers. Extinction Debt and Windows of Conser- vation Opportunity in the Brazilian Amazon. Science, 337(6091):228–232, 2012.

171. Isabel M.D. Rosa, M.J. Smith, O.R. Wearn, D Purves, and R.M. Ewers. The environmental legacy of modern tropical deforestation. Current Biology, 26:2161–2166, 2016. doi: 10.1016/j.cub.2016.06.013.

172. S. Dullinger, F Essl, W Rabitsch, K.-H. Erb, S Gingrich, H Haberl, K Hülber, V Jarosik, F Krausmann, I Kühn, J Pergl, P Pysek, and P.E. Hulme. Europe’s other debt crisis caused by the long legacy of future extinctions. Proceedings of the National Academy of Sciences, 110(18):7342–7347, 2013.

173. M Kuussaari, R Bommarco, R.K. Heikkinen, A Helm, J Krauss, R Lindborg, E Ockinger, M Partel, J Pino, F Roda, C Stefanescu, and T Teder. Extinction debt: a challenge for biodiversity conservation. Trends in Ecology and Evolution, 24(10):564–571, 2009.

174. S.T. Jackson and D.F. Sax. Balancing biodiversity in a changing environment: extinction debt, immigration credit and species turnover. Trends in Ecololgy and Evolution, 25(3): 153–160, 2010.

175. Tobias Naaf and Jens Kolk. Colonization credit of post-agricultural forest patches in NE Germany remains 130-230years after reforestation. Biological Conservation, 182:155–163, 2015.

176. Forest Isbell, David Tilman, Stephen Polasky, and Michel Loreau. The biodiversity- dependent ecosystem service debt. Ecology Letters, 18(2):119–134, 2015.

33 GENERAL INTRODUCTION

177. S Cusser, J.L. Neﬀ, and S Jha. Land use change and pollinator extinction debt in exurban landscapes. Insect Conservation and Diversity, 8(6):562–572, 2015.

178. J.E. Cinner. Social-ecological traps in reef ﬁsheries. Global Environmental Change, 21(3): 835–839, 2011.

179. P Dasgupta. Nature and the economy. Journal of Applied Ecology, 44(3):475–487, 2007.

180. C.L. Redman. Human impact on ancient environments. University of Arizona Press, Tucson, 1999.

181. W.H. McNeill. Something new under the sun: an environmental history of the twentieth- century. Norton, New York, 2000.

182. J Diamond. Collapse: How Societies choose to fail or survive. Penguin, 2005.

183. R.McC Adams. Complexity in archaic states. Journal of Anthropological Archaeology, 20 (3):345–360, 2001.

184. Thomas E Currie, Simon J Greenhill, Russell D Gray, Toshikazu Hasegawa, and Ruth Mace. Rise and fall of political complexity in island south-east asia and the paciﬁc. Nature, 467:801–804, 2010.

185. W.J. Boonstra and F.W. de Boer. The historical dynamics of social-ecological traps. Ambio, 43(3):260–274, 2014. doi: 10.1007/s13280-013-0419-1.

186. R. Costanza, L Graumlich, W Steﬀen, C Crumley, J Dearing, K Hibbard, R Leemans, C Redman, and D Schimel. Sustainability or collapse: what can we learn from integrating the history of humans and the rest of nature? Ambio, 36:522–527, 2007.

187. P Turchin. Arise "cliodynamics". Nature, 454:34–35, 2008.

188. J.A. Tainter. The Collapse of Complex Societies. Cambridge University Press, Cambridge, UK, 1988.

189. Clive Ponting. A green History of the World: The Environment & the Collapse of Great Civilizations. Number 35. Penguin, 1991.

190. J.M. Diamond. Easter’s end. Discovery, 16:62–69, 1995.

191. Rafael Reuveny. Taking Stock of Malthus: Modeling the Collapse of Historical Civilizations. Annual Review of Resource Economics, 4(1):303–329, 2012.

192. R. Reuveny and C. S. Decker. Easter Island: Historical anecdote or warning for the future? Ecological Economics, 35(2):271–287, 2000.

193. David H. Good and Rafael Reuveny. The fate of Easter Island: The limits of resource management institutions. Ecological Economics, 58(3):473–490, 2006.

194. John M. Anderies. Economic development, demographics, and renewable resources: a dynamical systems approach. Environment and Development Economics, 8(02):219–246, 2003.

195. M. Scott Taylor. Innis lecture: Environmental crises: Past, present, and future. Canadian Journal of Economics, 42(4):1240–1275, 2009.

34 GENERAL INTRODUCTION

196. Safa Motesharrei, Jorge Rivas, and Eugenia Kalnay. Human and nature dynamics (HANDY): Modeling inequality and use of resources in the collapse or sustainability of societies. Ecological Economics, 101:90–102, 2014.

197. A Tavoni and S Levin. Managing the climate commons at the nexus of ecology, behaviour and economics. Nature Climate Change, 4:1057–1063, 2014.

198. W.J. Boonstra, E Bjorkvik, and L.J. Haider. Human responses to social-ecological traps. Sustainability Science, 11(6):877–889, 2016. doi: 10.1007/s11625-016-0397-x.

199. J.M. Anderies. On modelling human behavior and institutions in simple ecological economic systems. Ecological Economics, 35:393–412, 2000.

200. M Schlüter, A Baeza, G Dressler, K Frank, J Groeneveld, W Jager, M.A. Janssen, R.R.J McAllister, B Müller, K Orach, N Schwartz, and N Wijermans. A framework for map- ping and comparing behavioural theories in models of social-ecological systems. Ecological Economics, 131:21–35, 2017.

201. E Ostrom. Governing the commons: the evolution of institutions for collective action. New York, NY: Cambridge University Press, 1990.

202. E Ostrom. Collective Action and the Evolution of Social Norms. Journal of Economic Perspectives, 14(3):137–158, 2000.

203. K Nyborg and M Rege. On social norms: the evolution of considerate smoking behavior. Journal of Economic Behavior and Organization, 52:323–340, 2003. doi: 10.1016/S0167- 2681(03)00031-3.

204. E Ostrom. A general framework for analyzing sustainability of social-ecological systems. Science, 325:419–422, 2009. doi: 10.1126/science.1172133.

205. A Tavoni, M Schlüter, and S Levin. The survival of the conformist: Social pressure and renewable resource management. Journal of Theoretical Biology, 299:152–161, 2012.

206. M Schlüter, A Tavoni, and S Levin. Robustness of norm-driven cooperation in the commons. Proceedings of the Royal Society B: : Biological Sciences, 283:20152431, 2016.

207. M Scheﬀer, F Westley, and W Brock. Slow response of societies to new problems: causes and costs. Ecosystems, 6(5):493–502, 2003.

208. D Pauly. Anecdotes and the shifting baseline syndrome of ﬁsheries. Trends in Ecology and Evolution, 10(10):430, 1995.

209. S.K. Papworth, J Rist, E.J. Coad, and E.J. Milner-Gulland. Evidence for shifting baseline syndrome in conservation. Conservation Letters, 2(2):93–100, 2009. doi: 10.1111/j.1755- 263X.2009.00049.x.

210. K Nyborg, J.M. Anderies, A Dannenberg, T Lindahl, C Schill, M Schlüter, W.N. Adger, K.J. Arrow, S. Barrett, S. Carpenter, F.S. Chapin III, A.-S. Crépin, G. Daily, P. Ehrlich, C. Folke, W. Jager, N Kautsky, S.A. Levin, O.J. Madsen, S. Polasky, M. Scheﬀer, B. Walker, E.U. Weber, J. Wilen, A. Xepapadeas, and A de Zeeuw. Social norms as solutions. Science, 354(6308):42–43, 2016. doi: 10.1126/science.aaf8317.

211. M.L. Rosenzweig. Species diversity in space and time. Cambridge University Press, 1995.

35 GENERAL INTRODUCTION

212. J Dengler. Which function describes the species-area relationship best? a review and empirical evaluation. Journal of Biogeography, 36:728–744, 2009.

213. O Arrhenius. Species and area. Journal of Ecology, 9:95, 1921.

214. E.F. Connor and E.D. McCoy. The statistics and biology of the species-area relationship. American Naturalist, 113:119–130, 1979.

215. R.H. Mac Arthur and E.O. Wilson. The theory of island biogeography. Monographs in Population Biology, 1967.

216. A O’Connor, M.I.and Gonzalez, J.E.K Byrnes, B.J. Cardinale, J.E. Duﬀy, L. Gamfeldt, J.N. Griﬃn, D. Hooper, B.A. Hungate, A Paquette, P.L. Thompson, L.E. Dee, and K.L. Dolan. A general biodiversity-function relationship is mediated by trophic level. Oikos, 126(1):18–31, 2017.

36 CHAPTER 1

Time-delayed biodiversity feedbacks and the sustainability of social-ecological systems

Anne-Sophie Lafuite 1, Michel Loreau 1

As published in Ecological Modelling Volume 351, 2017, pages 96–108

1Centre for Biodiversity Theory and Modelling, Theoretical and Experimental Ecology Station, CNRS and Paul Sabatier University, Moulis, France

37 CHAPTER 1

Chapter outline

Models capturing the dynamic feedback loop between human population growth and biodiversity loss across several temporal and spatial scales are lacking. Chapter 1 takes on a dynamical system approach in order to capture the interaction between human societies, natural habitat destruction and biodiversity loss. Our modelling approach includes insights from market economics, spatial ecology, and human demography. This work aims at investigating the long-term consequences of time-delayed biodiversity feedbacks on the sustainability of human societies.

38 CHAPTER 1

Abstract

The sustainability of coupled social-ecological systems (SESs) hinges on their long-term ecological dynamics. Land conversion generates extinction and functioning debts, i.e. a time-delayed loss of species and associated ecosystem services. Sustainability theory, however, has not so far considered the long-term consequences of these ecological debts on SESs. We investigate this question using a dynamical model that couples human demography, technological change and biodiversity. Human population growth drives land conversion, which in turn reduces biodiversity-dependent ecosystem services to agricultural production (ecological feedback). Technological change brings about a demographic transition leading to a population equilibrium. When the ecological feedback is delayed in time, some SESs experience population overshoots followed by large reductions in biodiversity, human population size and well-being, which we call environmental crises. Using a sustainability criterion that captures the vulnerability of an SES to such crises, we show that some of the characteristics common to modern SESs (e.g. high production eﬃciency and labor intensity, concave-down ecological relationships) are detrimental to their long-term sustainability. Maintaining sustainability thus requires strong counteracting forces, such as the demographic transition and land-use management. To this end, we provide integrative sustainability thresholds for land conversion, biodiversity loss and human population size - each threshold being related to the others through the economic, technological, demographic and ecological parameters of the SES. Numerical simulations show that remaining within these sustainable boundaries prevents environmental crises from occurring. By capturing the long-term ecological and socio-economic drivers of SESs, our theoretical approach proposes a new way to deﬁne integrative conservation objectives that ensure the long-term sustainability of our planet.

Keywords: Biodiversity ; ecological economics ; ecosystem services; extinction debt; social-ecological system ; sustainability

39 CHAPTER 1

Introduction

Current trends in human population growth [1, 2] and environmental degradation [3] raise concerns about the long-term sustainability of modern societies, i.e. their capacity to meet their needs “without compromising the ability of future generations to meet their own needs“ [4]. Many of the ecosystem services supporting human systems are underpinned by biodiversity [5], and current species extinction rates threaten the Earth’s capacity to keep providing these essential services in the long run [6, 7]. The long-term ecological feedback of ecosystem services on human societies has been largely ignored by neo-classical economic theory, mainly due to the focus on short-term feedbacks, and the assumption that ecosystem processes can be substituted for by human capital (e.g. tools and knowledge) and labor, thereby releasing ecological checks on human population and economic growth [8, 9]. In particular, tremendous increases in agricultural productivity resulted in more than 100% rise in aggregate food supply over the last century [10].

However, the substitution of human capital for natural resources, also called the "technology treadmill" [11], is currently facing important limitations. One such limitation is land scarcity, as the remaining arable land reserve might be exhausted by 2050 [12]. Moreover, recent projections suggest a slowdown in the growth rate of agricultural Total Factor Productivity (TFP), which measures the effect of technological inputs on total output growth relative other inputs [13]. Technological improvements may not compensate for arable land scarcity [14], thus questioning the potential for continued TFP growth in the future [15, 16]. Another limitation comes from the loss of many biodiversity-dependent ecosystem services which play a direct or indirect role in agricultural production, such as soil formation [17], nutrient cycling, water retention, biological control of pests [18] and crop pollination [19, 20]. Substituting ecological processes with energy and agrochemicals has mixed environmental impacts, with unintended consequences such as water use disturbance, soil degradation, and chemical runoff. These effects are responsible for a slowdown in agricultural yield growth since the mid-1980s [21] and have adverse consequences on biodiversity, human health [22] and the stability of ecosystem processes [23].

40 CHAPTER 1

Moreover, biodiversity does not respond instantaneously to land-use changes. Habitat fragmentation [24] increases the relaxation time of population dynamics [25] - i.e. how fast a species responds to environmental degradation. As a consequence, species extinctions are delayed in time, which generates an extinction debt [26] and a biodiversity-dependent ecosystem service debt [27]. These ecological debts may persist for more than a century, and increase as species get closer to extinction [28]. The accumulation of these ecological debts may have long-term eﬀects on modern human societies.

Such time-delayed ecological feedbacks have been neglected by the most inﬂuential population projection models [29, 30, 31]. Yet, environmental degradation can have catastrophic consequences for human societies even without any delayed eﬀect [32, 33]. A well- known example is Easter Island, this Polynesian island in which civilization collapsed during the 18th century due to overpopulation, extensive deforestation and overexploitation of its natural resources [34]. In order to investigate the mechanisms behind this collapse, Brander and Taylor [35] modeled the growth of the human population as endogenously driven by the availability of natural resources, the depletion of which was governed by economic constraints. Their model was essentially a Lotka-Volterra predator-prey model, which is familiar to ecologists, with an economic interpretation. It showed that one of Easter Island’s ecological characteristics - the particularly low renewal rate of its forests - may be responsible for the famine cycles which brought forward the collapse of this civilization.

Given the unprecedented rates of current biodiversity and ecosystem service loss [6], accounting for their long-term feedback on modern human societies appears crucial. In an attempt to delimit safe thresholds for humanity, a 10% loss in local biodiversity was defined as one of the core ”planetary boundaries“ which, once transgressed, might drive the Earth system into a new, less desirable state for humans [36, 37]. However, in practice, the biodiversity threshold above which ecosystem processes are significantly affected varies among ecosystems [38, 39], and is correlated to other thresholds such as land-use change [40]. The definition of integrative thresholds thus requires that we consider the interaction between the economic and ecological components of systems.

41 CHAPTER 1

In order to explore the long-term consequences of ecological debts for human societies, we build upon Brander and Taylor’s framework, but we allow the human population to produce its own resources through land conversion. In our approach, terrestrial natural habitats provide essential ecosystem services to the agricultural lands - which are assumed to be unsuitable to biodiversity. Biodiversity and human population dynamics are coupled through the time-delayed eﬀect of biodiversity-dependent ecosystem services on agricultural production (ecological feedback). It may be viewed as a minimal social-ecological system (SES) model that couples basic insights from market economics and spatial ecology. From market economics, we derive human per capita consumptions and the rate of land conversion. From spatial ecology, we use a classical species-area relationship (SAR) to capture the dependence of species diversity on the remaining area of natural habitat, and account for ecological debts through the relaxation rate of communities following habitat loss.

We investigate the behavior of the system at equilibrium analytically, and then numerically evaluate the trajectories to the equilibrium. We show that the transient dynamics of an SES depends on its ecological, economic, demographic and technological parameters. Some SESs experience large population overshoots followed by reductions in biodiversity, human population size and well-being - which we call ”environmental crises“. We then analytically derive an integrative sustainability criterion that captures the vulnerability of an SES to such crises. This criterion allows assessing the eﬀects of some parameters on the long-term sustainability of the SES, and deriving integrative land conversion and biodiversity thresholds.

1 Methods

1.1 Coupling human and ecological dynamics

We model the long-term dynamics of three variables: the human population (H), technological eﬃciency (T) and biodiversity (B).

42 CHAPTER 1

  H˙ = µ(B, T) H    T˙ = σ T(1 − T/T ) (1)  m    ˙ B = − [B − S(H)]

The human population endogenously grows at a rate µ(B, T), which is explicitly de- ﬁned as a function of the per capita agricultural and industrial consumptions in section 1.1.1. Technological eﬃciency is assumed to follow logistic growth at an exogenous rate

σ, until a maximum efficiency Tm is reached (section 1.1.2). We use an economic general equilibrium framework to derive per capita human consumptions at market equilibrium, i.e. when production supply equals the demand of the human population (section 1.1.3). Using these consumptions, we derive a proportional relationship between land conversion and the size of the human population (section 1.1.4). Land conversion affects biodiversity through a change in the long-term species richness S(H). Current biodiversity B reaches its long-term level S(H) at a relaxation rate (section 1.2). Biodiversity-dependent ecosystem services then feed back on agricultural production and affect the per capita agricultural consumption and the human growth rate, µ(B, T). Model structure is summarized in Fig. 1.1.

1.1.1 Human demography

The interaction between human population, technology and income has been mainly studied by endogenous growth theory, which distinguishes three phases of economic development [41, 42]: (1) a Malthusian regime with low rates of technological change and high rates of population growth preventing per capita income to rise; (2) a Post-Malthusian Regime, where technological progress rises and allows both population and income to grow, and (3) a Modern Growth regime characterized by reduced population growth and sustained income growth [43]. Transition to this third regime results from a demographic transition which reverses the positive relationship between income and population growth.

In order to consider the basic linkages between human demography and economics, the

43 CHAPTER 1

Human population (H)

) Economic y ,,T 2 (T B ) ( y 1 market L L 1 2 Agriculture Technology Industry (T) A 1 A 2 fS(B) S Economic market Biodiversity (B)

Converted land ) (H A A(H)=A1+A2 A S (H ) Natural habitat

1-A(H)=A3

Figure 1.1: Coupling between human and ecological dynamics Black boxes: production sectors; grey boxes: dynamical variables; white box: auxiliary economic model; dashed lines: production inputs (labor Li, land Ai and technology T ); solid lines: per capita outputs (yi); grey line: biodiversity-dependent ecosystem services (ecological feedback, fS); black dotted line: eﬀect of land conversion on biodiversity; circle: total land divided into converted land (A1 + A2) and natural habitat (A3), where A1 + A2 + A3 = 1.

44 CHAPTER 1 growth rate of the human population is assumed to depend on the per capita consumptions of agricultural y1 and industrial goods y2 [44]:

min µ = µmax(1 − exp(y1 − y1))exp(−b2y2) (2)

min where µmax is the maximum human population growth rate, and y1 is the minimum per capita agricultural goods consumption, such that human population size increases if y1 >

min min y1 , and decreases if y1 < y1 . b2 is the sensitivity of µ to industrial goods consumption, and thus captures the strength of the demographic transition. A higher agricultural goods consumption increases the net human growth rate while a higher industrial goods consumption eventually limits the net growth rate of the human population. Note that both y1 and y2 vary with the states of the system (section 1.1.3). The system reaches

min its long-term equilibrium (y1 = y1 ) when further technological change and habitat conversion no longer increase total agricultural production, i.e. no longer compensate for the negative ecological feedback on agricultural production.

1.1.2 Technological change

Technological change is central to explain the transition from a Malthusian to a modern human population growth regime. Technology is often captured through the Total Factor Productivity term (TFP) of a production function, which accounts for eﬀects of technological inputs on total output growth relative to the other inputs, i.e. labor and land in our model. Accelerating TFP growth in recent years partially compensated for the slowing down in input growth (especially land) and allowed total output growth to maintain itself around 2% per year [45]. However, recent reviews suggest that agricultural TFP growth is slowing down in a number of countries [13] and that this trend is likely to continue [15, 16].

In our model, technological change increases production efficiency in the agricultural and industrial sectors, leading to higher productions for a given level of inputs (i.e. converted land and labor). We assume a logistic growth of production efficiency towards a maximum efficiency denoted by Tm, where σ is the exogenous rate of technological change

45 CHAPTER 1

(system (11)). We explore other forms of technological change in Appendix 1.A.

1.1.3 Instantaneous market equilibrium

We assume a closed market where total labor is given by the size of the human population. Consumption and production levels are derived by solving for a general market equilibrium, where prices of the production inputs and consumption goods vary endogenously. As the economic dynamics are much faster than the demographic and ecological dynamics, the market is assumed to reach an equilibrium between supply and demand instantaneously.

On the demand side, the human population is assumed to be homogeneous, i.e. composed of H identical agents. Per capita consumptions of agricultural and industrial goods

(y1 and y2, resp.) are derived from the maximization of a utility function, U(y1, y2), which is a common economic measure of the satisfaction experienced by consumers (Appendix 1.B.1):

η 1−η U(y1, y2) = y1 y2 (3)

where y1 is per capita agricultural demand, y2 is per capita industrial demand, and η is the preference for agricultural goods.

On the supply side, a total quantity Yi of goods is produced by sector i, using two production inputs, labor Li, and converted land Ai (Appendix 1.B.2):

α1 1−α1 α2 1−α2 Y1 = TL1 A1 fS(B) Y2 = TL2 A2 (4)

where production eﬃciency is captured by the variable T, αi is the relative use of labor compared to land in sector i, and the function fS(B) measures provisioning of biodiversity- dependent ecosystem services to agricultural production (ecological feedback). Such Cobb-Douglas production functions with constant returns to scale are a common assumption of growth models and allow for the substitution of land by labor. Potential ecological feedbacks on industrial production were ignored.

46 CHAPTER 1

When demand equals supply, the per capita consumptions are (Appendix 1.B.3):

y1(B, T) = γ1 fS(B) T/Tm y2(T) = γ2 T/Tm (5)

where γ1 and γ2 are explicitly deﬁned as functions of the parameters of the system in Table 1.1.

The per capita consumption of agricultural goods y1 is thus subject to the trade-off between the negative effect of decreasing biodiversity-dependent ecosystem services, and the positive effect of technological change. Conversely, the per capita consumption of industrial goods y2 monotonously increases with technological efficiency - as does the strength of the demographic transition.

1.1.4 Land conversion dynamics

Rising agricultural and industrial productions require the conversion and maintenance of land surfaces A1 and A2 respectively, at a cost of κ units of labor per unit of converted area. Land conversion reduces the area of natural habitat A3, where A1 + A2 + A3 = 1.

Let us denote by A the converted land, i.e. A = A1 + A2. At the market equilibrium, the relationship between A and the human population size H is (Appendix 1.B.4):

H = φA (6) where φ represents the density of the human population on converted land, and is explicitly deﬁned as a function of the economic parameters in Table 1.1. The converted area A decreases with the land operating cost κ, since high operating costs reduce the incentive to convert natural habitat. Proportional relationships between human population sizes and converted surfaces are commonly observed in local and regional developing economies, where subsistence agriculture remains strong and the transition to a modern growth regime is not achieved yet [46].

47 CHAPTER 1

1.2 Spatio-temporal dynamics of biodiversity and ecosystem ser-

vices

We need two types of relationships, which capture (1) the dependence of biodiversity upon the area of natural habitat A3 and the size of the human population (S(H)), and (2) the dependence of ecosystem services upon biodiversity (fS(B)). Since species richness can be related to the area and spatial characteristics of natural habitats [47], we choose species richness as a biodiversity measure in our system.

1.2.1 Species-area relationship (SAR)

SARs are commonly used tools in ecological conservation to assess the eﬀect of natural habitat destruction on species richness [48]. A common way to describe the decrease of

z species richness with habitat loss is to use a power function S = c A3, where c and z represent the intercept and the slope in log-log scale. By choosing a unitary intercept

(c = 1) and under the constraint A1 + A2 + A3 = 1, the number of species S(H) that can be supported in the long run on an area A3 of natural habitat can be rewritten:

  z  (1 − H/φ) for H/φ ≤ 1 S(H) = (7)   0 for H/φ > 1

In terrestrial systems, the slope z typically ranges from 0.1 to 0.4 [49, 50], with values depending on ecosystem characteristics and species response to habitat loss. In our model, the long-term species richness S(H) thus decreases when the human population grows, and equals zero if the human population exceeds its maximum viable size φ.

We assume, following empirical [51, 52] and theoretical [47, 53] expectations, that the rate of community relaxation to this long-term richness is proportional to the diﬀer- ence between current richness, B, and long-term richness S(H) (system (11)), where the relaxation rate is [53].

48 CHAPTER 1

1.2.2 Biodiversity-ecosystem service relationship (BES)

Species richness has a positive eﬀect on the level of many regulating services [5], among which the pollination and pest regulation services are particularly important to agricultural production - since the production of over 75% of the world’s most important crops and 35% of the food produced is dependent upon animal pollination [54].

The relationship between biodiversity and ecosystem services can be captured by a power function [55, 56] :

Ω fS(B) = B (8)

where Ω < 1, since the shape of the function fS is mostly concave-down in terrestrial systems [57].

1.3 Model summary

System (11) can be rewritten as:

  ˙ min Ω H = µmaxH 1 − exp(y − γ1B T/Tm) exp(−b2γ2T/Tm)  1    ˙ T = σT(1 − T/Tm)  (9)   z   − [B − (1 − H/φ) ] for H/φ < 1  ˙  B =      − B for H/φ > 1

The aggregate parameters γ1, γ2, and φ result from the economic derivations of sections 1.1.3, 1.1.4 and Appendix 1.B. All parameters and aggregate parameters are explicitly deﬁned in Table 1.1. In the short term, technological change and human population growth drive land conversion and increase human consumption and well-being. In the long term, the time-delayed biodiversity loss reduces the supply of biodiversity-dependent ecosystem services to agricultural production (ecological feedback). The ecological relaxation rate generates a time lag between human and ecological dynamics, mediated through the ecological feedback. In the next section, we investigate the consequences of this time lag on the long-term

49 CHAPTER 1 sustainability of an SES.

2 Results and Discussion

2.1 Analysis of the dynamical system

For a given set of parameters and initial conditions, the dynamics of our SES is driven by the interaction between the dynamical variables summarized in eq. (9). In section 2.1.1, we characterize the steady states that can potentially be reached by the SES in the long term, and the necessary condition for their stability. We then simulate in section 2.1.2 the transient dynamics of the SES over time, for various time lags between the ecological and human dynamics ().

2.1.1 Steady states and stability condition

A steady state is reached by the dynamical variables when the system is at equilibrium, i.e. when H˙ = T˙ = B˙ = 0 (Appendix 1.C.1). There are two potential steady states (H,T,B) for our system: (1) an undesirable equilibrium (0,Tm, 1), when the parameters do not allow the human population to persist in the long term, and (2) a desirable equilibrium

∗ ∗ (H ,Tm,B ):

1 1 ∗ min Ωz ∗ min Ω H = φ 1 − y1 /γ1 B = y1 /γ1 (10)

For a given set of parameters, only one of these equilibria is stable and reached by the

∗ ∗ SES (Appendix 1.C.2). The desirable equilibrium (H ,Tm,B ) is stable if condition (11) is met:

min γ1 > y1 (11)

Condition (11) captures the capacity of an SES to persist in the long term, given its technological and economic characteristics (Table 1.1). The higher γ1, the higher the initial human growth rate and the lower the biodiversity at equilibrium B∗. When condition

50 CHAPTER 1

Dynamical variables Initial value H Human population 10−3 T Technology 10−3 B Biodiversity 1 Parameters Default values Economic parameters η Agents preference for agricultural goods 0.5 α1 Labor intensity in the agricultural sector 0.3 α2 Labor intensity in the industrial sector 0.9 Technological parameters Tm Maximum technological eﬃciency varies σ Rate of technological change 0.5 κ Land operating cost 0.35 Demographic parameters µmax Maximum growth rate 1 min y1 Minimum per capita agricultural consumption 0.5 b2 Sensitivity to industrial goods’ consumption 0.2 Ecological parameters Ω Concavity of the BES relationship 1 z Concavity of the SAR 0.3 Relaxation rate 0.6 Aggregate parameters φ κ/(1 − α1η − α2(1 − η)) α1 1−α1 γ1 ηTmα1 ((1 − α1)/κ) α2 1−α2 γ2 (1 − η)Tmα2 ((1 − α2)/κ) 0 min −b2γ2 θ 4Ωzy1 e 1 0 min Ωz θ θ (γ1/y1 ) − 1

∆ − µmaxθ

Table 1.1: Deﬁnition and default values of the parameters and dynamical variables. αi and η ∈ [0; 1]. When possible, parameter values have been derived from the literature (Ω, z, , alphai, η), or calibrated using historical population and land conversion min trends (µmax, σ, κ, y1 , b2). Other parameter values are chosen so as to keep the system feasible, i.e. with positive population sizes.

51 CHAPTER 1

(11) is (not) met, the SES reaches the desirable (undesirable) equilibrium in the long term. According to eq. (10), natural habitat (1 − H/φ) is never entirely converted at equilibrium, since H/φ < 1.

Condition (11) only guarantees that the desirable equilibrium is reached in the long term, but tells nothing about the trajectories to the equilibrium, and especially about the eﬀect of the relaxation rate .

2.1.2 Transient dynamics for varying relaxation rates

In order to test for the eﬀect of a time-delayed ecological feedback on the transient dynamics of an SES, we numerically evaluate the trajectories to the equilibrium and increase the lag between the human and ecological dynamics by decreasing .

When the relaxation time is negligible (1/ → 0), biodiversity responds instantaneously to habitat conversion (B = S(H)) and the ecological feedback on agricultural production is instantaneous. In this case, Fig.1.2.A shows that transient trajectories converge monotonically to the viable equilibrium (H∗,B∗).

Introducing a time-delay between the dynamics of humans and biodiversity by decreasing leads to damped oscillations during the transient dynamics to the equilibrium (Fig.1.2.B). Moreover, if the relaxation rates of species extinction and recovery are distinct (e.g. recovery takes longer than extinction), these damped oscillations can lead to the collapse of the human population (Fig.1.2.C).

These damped oscillations result from the repetition of three successive phases (Fig.1.2.D): (1) Human population growth and biodiversity decline (µ ≥ 0). During the initial growth phase, the human population reaches its equilibrium size H∗. But at this point, the biodiversity debt is not paid oﬀ yet, i.e. biodiversity and ecosystem services are still in excess (B > B∗). As a consequence, the per capita consumption is not at equilibrium

min min (y1 > y1 ), and human population keeps growing until y1 = y1 . (2) Environmental crisis and population decline (µ ≤ 0). When biodiversity eventually reaches its equilibrium value (B = B∗), both human population growth and habitat

min conversion stop (y1 = y1 ). However, the human population now exceeds its equilibrium

52 CHAPTER 1

A B Post−Malthusian 1 Growth Regime 1 1 1 B* Modern Growth Regime 0.5 0.5 0.5 0.5 H* Biodiversity Human population

0 0 0 0 0 200 400 600 800 1000 0 200 400 600 800 1000 Time Time * C D B

1 1 1 (1)

(2) 0.5 0.5 0.5 H* Biodiversity Human population Human population (3) 0 0 0 0 200 400 600 800 1000 0.6 0.7 0.8 0.9 1 Time Biodiversity

Figure 1.2: Transient human population and biodiversity dynamics with varying relaxation coeﬃcients (). Parameters are given in Table 3.1, with Tm = 1. A: negligible relaxation time ( = 10); B: high relaxation time ( = 0.02); C: diﬀerential relaxation times for biodiversity loss (− = 0.02) and recovery (+ = −/20); D: phase plane trajectories for biodiversity and human population. Grey curve: negligible relaxation time ( = 10); black curve: high relaxation time ( = 0.02) leading to the repetition of phases 1 (thick curve), 2 (dashed curve) and 3 (dotted curve) until the equilibrium (B∗,H∗) is reached; double-arrow: maximum amplitude of the transient crisis.

53 CHAPTER 1 size (H > H∗). This population overshoot results in an over-conversion of natural habitat which generates an additional biodiversity debt. While this debt is paid oﬀ, human

min growth rate becomes negative (y1 < y1 ) and human population declines. (3) Human population and biodiversity recovery (µ ≥ 0). As the human population declines, the converted surface decreases and allows for the recovery of biodiversity. When biodiversity is not in deﬁcit anymore (B = B∗), the human population can reach its equilibrium size, H∗.

2.2 Sustainability sensitivity analysis

2.2.1 Sustainability conditions

Transient environmental crises are undesirable from a sustainability perspective, since these collapses result from a decrease in per capita agricultural consumption, and thus in human utility (eq. (3)) - which we use as a proxy for human well-being. Let us define an SES as sustainable when its transient dynamics monotonously converge to the long-term equilibrium (i.e. no crisis) and the system experiences a net growth in human utility. Condition (12) ensures that technological change is sufficient to compensate for the effect of biodiversity loss on human utility:

min η Tm/T (0) > (γ1/y1 ) (12) where T (0) is the initial technological eﬃciency. When condition (12) is met, human utility at equilibrium is higher than the initial human utility (U ∗ > U(0)), but environmental crises may still lead to a decrease in human utility and population size during the transient dynamics. Using dynamical system properties (see Appendix 1.C.2), we derive a general condition for the absence of environmental crises:

> θµmax (13)

54 CHAPTER 1 where θ is explicitly deﬁned as a function of the economic, demographic, technological and ecological parameters of the SES in Table 1.1. Condition (13) implies that a given SES is not vulnerable to environmental crises if its ecological dynamics () is fast enough compared to the growth of its human population (≈ µmax), for a given set of parameters (θ).

Let us rewrite condition (13) as a sustainability criterion, ∆ = − θµmax. From the expression of θ in Table 1.1, it is straightforward to deduce the eﬀect of some parameters on ∆. In particular, ∆ decreases with the preference for agricultural goods, η, since it increases the global demand for agricultural goods. ∆ increases with the land operating cost κ, since high operating costs reduce the incentive to convert natural habitat (see

Appendix 1.D for more details). The strength of the demographic transition (b2) also increases the sustainability criterion ∆, by reducing the net growth rate of the human population.

In the next sections, we explore the effect of the other parameters through numerical simulations. We use the default parameters of Table 1.1 such that conditions (11) and (12) are met, i.e. the SES can reach its desirable equilibrium, and human well-being at equilibrium is higher than the initial well-being. A given SES is sustainable (unsustainable) if its parameters satisfy ∆ > 0 (∆ < 0) and condition (12). Any increase (decrease) in ∆ has a positive (negative) effect on long-term sustainability. We then assess the effect of some characteristics of SESs on their steady states (H∗,B∗) and sustainability (∆) by varying a single parameter at a time: the maximum technological efficiency (Tm), the agricultural labor intensity (α1) and the concavity of the BES relationship (Ω). For each of these parameters, we also plot the effect of the strength of the demographic transition b2.

2.2.2 Eﬀect of technological change

In our model, production eﬃciency increases with technological change until a maximum technological eﬃciency Tm. Increasing Tm results in larger human population sizes and less biodiversity at equilibrium (Fig. 1.3.A). When the demographic transition is weak

55 CHAPTER 1

(b2 = 0.2), the eﬀect of technological change on sustainability is negative (Fig. 1.3.D), and the amplitude of the transient crises rises with technological change (Fig. 1.3.A and D). On the other hand, technological change has a positive eﬀect on per capita and total human utility at equilibrium (Fig. 1.3.G), since it increases industrial consumption (y2) and the size of the human population at equilibrium H∗.

However, for a stronger demographic transition (b2 = 2), the eﬀect of technological change on sustainability switches from negative to positive at Tm = 4 (dashed curve in

Fig. 1.3.D). Further increasing Tm allows the industrial consumption y2 to reduce the net human growth rate to the point where the system does not experience environmental crises

(∆ > 0 for Tm > 10). Note that b2 only aﬀects the transient dynamics of the system, and thus does not prevent high levels of biodiversity loss when technological eﬃciency rises.

2.2.3 Eﬀect of labor intensity

As labor intensity α1 captures the relative use of labor compared to land in the agricultural sector, a moderate increase in labor intensity (α1 < 0.7) reduces land conversion requirements and benefits both biodiversity and sustainability (Fig.1.3.B and E). How- ever, further increasing labor intensity (α1 > 0.7) increases agricultural productivity to the point where it lowers agricultural prices and increases the demand for natural habitat conversion. This rebound effect thus allows the human population and total human utility to rise again, while biodiversity, sustainability and per capita human utility decrease (Fig.1.3.B, E and H). A stronger demographic transition positively affects the sustainability criterion for all labor intensities, and thus partially mitigates the negative rebound effect (dashed curve, Fig.1.3.E). Note that the shape of the relationship between sustainability and labor intensity depends on the land operating cost, κ. If κ = 1, the bell-shaped curve is centered on α1 = 0.5. If κ > 1 (κ < 1, resp.), the sustainability criterion is maximum at α1 < 0.5 (α1 > 0.5, resp.). Indeed, higher operating costs reduce the incentive to convert natural habitat and increase sustainability (Appendix 1.D), thus reducing the sustainability-optimal labor intensity.

56 CHAPTER 1

2.2.4 Eﬀect of the concavity of ecological relationships

The concavity of the ecological relationships, i.e. the SAR and BES relationships, is captured by their parameters z and Ω, respectively. Since both parameters have similar eﬀects on ∆, we only present the results for Ω. Commonly observed concave-down BES relationships (Ω < 1) lead to highly populated and biodiversity-poor SESs at equilibrium (Fig.1.3.C). These SESs are characterized by a higher total human utility (Fig.1.3.I) and a lower sustainability (Fig.1.3.F) than SESs with concave-up BES relationships. A stronger demographic transition increases sustainability for both concave-down and -up BES relationships (dashed curve, Fig.1.3.F).

Modern SESs may thus be particularly vulnerable to the current rise in technological efficiency and reduction in labor intensity - unless the demographic transition is strong enough to counteract their negative effects on long-term sustainability. Sustainable SESs are characterized by higher levels of biodiversity, lower human population sizes, industrial consumption and consumption utility at equilibrium, compared to unsustainable SESs. In section 2.3, we derive sustainability thresholds for land conversion, biodiversity and human population size, and explore their efficiency in preserving the long-term sustainability of SESs through numerical simulations.

2.3 Application to land-use management

2.3.1 Integrative sustainability threshold for land conversion

Our sustainability condition ∆ > 0 can be rewritten as A < AS (Appendix 1.C.3), where

AS is the sustainable land conversion threshold:

0 ! θ µmax AS = 1 − 0 (14) + θ µmax

θ0 is explicitly deﬁned as a function of the economic, technological and ecological parameters of the SES in Table 1.1. AS represents the maximum area of natural habitat that a given SES can convert without becoming unsustainable, and depends on the economic

57 CHAPTER 1

A B C ) )

* 1.5 1.5 1 1 *

1 1 1 1

0.5 0.5

0.5 0.5 0.5 0.5

0 0 0 0 0 0 Eq. biodiversity (B Eq. human pop. (H 0 5 10 15 20 0 0.5 1 0 0.5 1 1.5 2

D E F 2 5 ) 1 ∆ 0 0 0 −2

−1 −5 −4

Sust. criterion ( −6 −2 −10 0 5 10 15 20 0 0.5 1 0 0.5 1 1.5 2

G H I 10 1.5 1.5 ) 10 *

) 0.5 0.5 * .U * 1 1

5 5 0.5 0.5 Eq. utility (U

0 0 0 0 0 0 Total utility (H 0 5 10 15 20 0 0.5 1 0 0.5 1 1.5 2 Max. techn. efficiency (T ) Labor intensity (α ) Ω m 1 BES concavity ( )

Figure 1.3: Effect of varying parameter values on the steady states, the sustainability criterion and human well-being at equilibrium. Parameters are given in Table 1.1, with Tm = 1.8 (B-E-H) and Tm = 1.2 (C-F-I). A to C: Effect of Tm, α1, and Ω on the human population size (black curve) and biodiversity (grey curve) at equilibrium, when demographic transition is weak (b2 = 0.2). Grey areas represent the maximum amplitude of the transient crises as defined in Fig.2.D. D to F: Effect of Tm, α1 and Ω on the sustainability criterion (∆ = − θµmax), when demographic transition is weak (solid curve, b2 = 0.2) and strong (dashed curve, b2 = 2). Parameter values for which ∆ is positive correspond to sustainable trajectories. G to I: Effect of Tm, α1, and Ω on per capita human well-being (U ∗, black curve) and total human well-being (H∗.U ∗, grey curve) at equilibrium, when demographic transition is weak (b2 = 0.2). Grey areas represent the maximum amplitude of the transient variations in well-being.

58 CHAPTER 1

and ecological characteristics of the SES. In particular, AS decreases as the ecological relaxation rate decreases. As a result, the larger the ecological relaxation time (i.e. the lower ), the more natural habitat a SES needs to preserve in order to remain sustainable. Using this land conversion threshold AS, it is also possible to derive a sustainable

z biodiversity threshold BS = (1 − AS) , and a human population threshold HS = φAS.

2.3.2 Land-use scenarios

In this section, we explore the eﬃciency of the land conversion threshold (eq.(14)) in preventing environmental crises. To do so, we consider two alternative land-use scenarios: (1) no restriction on land conversion, and (2) conservation of an area of natural habitat

1 − AS, e.g. through the creation of a protected area. In order to stop land conversion at the sustainable threshold AS without (directly) limiting the dynamics of the human population, we have to deﬁne converted land A as a fourth dynamical variable:

  ˙  H/φ for A ≤ AS A˙ = (15)   0 else

Fig.1.4 shows that, in an unsustainable SES (i.e. ∆ < 0), the ﬁrst scenario generates transient environmental crises (Fig.1.4.A), while the second scenario prevents environmental crises from occurring (Fig.1.4.B). A precautionary approach to natural habitat conservation appears necessary to prevent biodiversity and human population from exceeding their own sustainable thresholds, thus resulting in environmental crises in the long run.

59 CHAPTER 1

A B 2 2 2 2 ) ) * 1.5 1.5 1.5 1.5 *

1 1 1 1

0.5 0.5 0.5 0.5 Eq. biodiversity (B Eq. human pop. (H

0 0 0 0 0 20 40 60 80 100 0 20 40 60 80 100 Time Time

Figure 1.4: Land-use scenarios. For both scenarios, parameters are given in Table 1.1, with Tm = 5. A: No restriction on land conversion; dynamics of an unsustainable SES (∆ < 0) with no land-use management. B: Implementation of the sustainable land conversion threshold; dynamics of the same SES (∆ < 0) when the sustainable threshold of natural habitat 1 − AS is set aside.

3 Conclusions

We explore the long-term dynamics of coupled SESs by modelling the main reciprocal feedbacks between human population and biodiversity, and show that the temporal trajectories of the human population and biodiversity are not necessarily monotonous. In particular, for large ecological relaxation times, we observe the emergence of transient environmental crises, i.e. large reductions in biodiversity, human population size and well- being. More complex population projection models calibrated with data for the world population, industrialization, pollution, food production and non-renewable resources obtained similar ”overshoot and collapse“ scenarios [29]. Recent updates even suggest that the business-as-usual trajectory of our societies may be approaching such a global collapse [30, 31]. However, neither these population projection models nor neo-classical economic models consider any long-term ecological feedback of biodiversity on agricultural production, which may worsen these crises scenarios.

The present work thus sheds new light upon the importance of accounting for ecological time lags when studying the sustainability of SESs and deﬁning management policies.

60 CHAPTER 1

Accumulating data on extinction debts [52, 58, 59, 60, 61] and on the relationship between biodiversity and the provisioning of ecosystem services [5, 56] make it possible to include these time-delayed ecological feedbacks in more complex simulation models, such as existing models coupling the economy and climate [62, 63]. The ability of our conceptual model to perform population projections is limited, since many of our economic and ecological parameters do not vary endogenously. For example, the ecological relaxation rate is known to decrease with habitat loss and fragmentation [28], and population density on converted land φ rose between 1993 and 2009 [64], since the rate of land degradation (9%) was lower than that of the human population and economic growth (resp. +23% and +153%).

Although detailed policy-relevant population scenarios would require a simulation model of greater complexity, our toy-model allows assessing the qualitative effects of some parameters of SESs on their long-term sustainability. It provides us with an integrative sustainability criterion that captures the vulnerability of an SES to environmental crises, given its economic, demographic, technological and ecological characteristics. Using this criterion, we show that some of the characteristics common to modern SESs, such as high technological efficiency and low labor intensity, are detrimental to their long-term sustainability. Even though land-use efficiency reduces land conversion in the short term, it also increases the time lag between the dynamics of humans and biodiversity and the vulnerability of the SES to environmental crises. Indeed, a more efficient use of land resources releases ecological checks on human population growth, which eventually increases both land conversion and the extinction debt. Moreover, the commonly observed concave-down SARs and BES relationships exacerbate this effect, since a given amount of habitat conversion translates into a lower loss of biodiversity and ecosystem services with concave-down than with concave-up relationships.

We show that counteracting forces such as the demographic transition and natural habitat conservation can mitigate environmental crises - provided that the ecological relaxation rates and the sensitivity of the human growth rate to industrial consumption are high enough. Indeed, the demographic transition is often presented as the solution to

61 CHAPTER 1 accelerate economic development and reduce environmental impact in developing countries [65, 34]. However, recent projections cast doubt upon its stabilizing potential [66]. Our results also show that, in order to eﬃciently counteract time-delayed biodiversity feedbacks, natural habitat conservation should take place when biodiversity is still abundant. Alternatively, habitat restoration can also help mitigating crises, and century-long extinction debts may be seen as windows of conservation opportunity [52]. However, our results suggest that we should not rely on habitat restoration, since restoration delays are often longer than extinction debts [67, 68], which worsens environmental crises.

The lack of such considerations about ecological time lags in current conservation policies calls for a precautionary approach to natural habitat conservation. To this end, we provide an integrative land conversion threshold that captures the long-term ecological dynamics of species confronted with the destruction of their habitat, given the economic, technological and demographic parameters of an SES. The smaller the ecological relaxation rate ( → 0), the more natural habitat should be preserved in order to avoid environmental crises. Since relaxation rates decrease with habitat loss and fragmentation [28], the sustainable amount of natural habitat may increase when more and more natural habitat is lost.

Regarding rates of current biodiversity loss [69], it is thus crucial to determine how much natural habitat we need to preserve in the long run. Recent attempts to define global conservation thresholds for biodiversity and land use [37] neglect the interaction between the social and ecological components of SESs. For instance, as biodiversity moves closer to its own potential threshold, it reduces the land-use change threshold [39]. There is also a lack of context-dependency for the application of these thresholds at local scales, as SESs with different sets of economic, technological and ecological characteristics present different thresholds. Our theoretical model of a coupled SES proposes a way to move beyond these limitations. Our integrative thresholds for land conversion and biodiversity loss are not defined independently from each other, but related through the system’s parameters. As a consequence, it is possible to predict how a change in one of the thresholds affects the other, and to define context-dependent land-use policies that

62 CHAPTER 1 prevent environmental crises from occurring. Despite its simplicity, our model may be seen either as a thought experiment about the sustainability of our planet, or as a representation of a local SES - which could be connected to other SESs by considering trade, human migrations and species dispersal. Improvements to the model include adding property rights and trade in order to gain realism in land-use change dynamics [70], and internalizing the value of ecosystem services via payments for ecosystem services and taxes on land conversion [71]. Finally, our results emphasize the critical need for a better assessment of ecological time lags and how they feed back on human systems. Reducing uncertainties on ecological time lags effects is crucial to be able to define integrative thresholds and foster sustainability of our planet, which may be seen as a global social-ecological system [7]. The pursuit of such objectives requires taking into account the main internal feedbacks and specificities of coupled social- ecological systems, which dynamical system modelling allows.

63 CHAPTER 1

Bibliography

1. Joel E Cohen. Human population: the next half century. Science, 302(5648):1172–1175, 2003. ISSN 1095-9203. doi: 10.1126/science.1088665.

2. Patrick Gerland, a. E. Raftery, H. Ev Ikova, Nan Li, Danan Gu, Thomas Spoorenberg, Leontine Alkema, Bailey K Fosdick, Jennifer Chunn, Nevena Lalic, Guiomar Bay, Thomas Buettner, G. K. Heilig, and J. Wilmoth. World population stabilization unlikely this century. Science, 346(6206):234–237, 2014.

3. Peter M Vitousek, Harold a Mooney, Jane Lubchenco, and Jerry M Melillo. Human Domination of Earth’ s Ecosystems. Science, 277:494–499, 1997. ISSN 00368075. doi: 10.1126/science.277.5325.494.

4. G. Brundtland, M. Khalid, S. Agnelli, S. Al-Athel, B. Chidzero, L. Fadika, V. Hauﬀ, I. Lang, M. Shijun, M. Morino de Botero, M. Singh, S. Okita, et al. Our Common Future (The Brundtland Report). Oxford University Press, 1987.

5. Bradley J Cardinale, J Emmett Duﬀy, Andrew Gonzalez, David U Hooper, Charles Perrings, Patrick Venail, Anita Narwani, Georgina M Mace, David Tilman, David A Wardle, Ann P Kinzig, Gretchen C Daily, Michel Loreau, James B Grace, Anne Larigauderie, Diane S Srivastava, and Shahid Naeem. Biodiversity loss and its impact on humanity. Nature, 486 (7401):59–67, 2012. ISSN 00280836. doi: 10.1038/nature11148.

6. Henrique M Pereira, Paul W Leadley, Vânia Proença, Rob Alkemade, Jörn P W Scharle- mann, Juan F Fernandez-Manjarrés, Miguel B Araújo, Patricia Balvanera, Reinette Biggs, William W L Cheung, Louise Chini, H David Cooper, Eric L Gilman, Sylvie Guénette, George C Hurtt, Henry P Huntington, Georgina M Mace, Thierry Oberdorﬀ, Carmen Re- venga, Patrícia Rodrigues, Robert J Scholes, Ussif Rashid Sumaila, and Matt Walpole. Scenarios for global biodiversity in the 21st century. Science, 330(6010):1496–1501, 2010.

7. P R Ehrlich and a H Ehrlich. Can a collapse of global civilization be avoided? Proceedings of the Royal Society of London B: Biological Sciences, 280(1754):20122845, 2013.

8. E. Boserup. The conditions of agricultural growth. Population Studies, 20(1):1–108, 1965.

9. P Dasgupta and G Heal. The optimal depletion of exhaustible resources. Review of Economic Studies, 41(5):3–28, 1974.

10. J Schmidhuber and F. N Tubiello. Global food security under climate change. Proceedings of the National Academy of Sciences, pages 19703–19708, 2007. ISSN 0027-8424, 1091-6490. doi: 10.1073/pnas.0701976104.

11. J. Pezzey. Sustainability: an interdisciplinary guide. Environmental values, 1:321–362, 1992.

12. Eric F. Lambin and P. Meyfroidt. Global land use change, economic globalization, and the looming land scarcity. Proceedings of the National Academy of Sciences of the United States of America, 108(9):3465–72, 2011. ISSN 10916490. doi: 10.1073/pnas.1100480108.

13. P Kumar, S Mittal, and M Hossain. Agricultural Growth Accounting and Total Factor Productivity in South Asia: A Review and Policy Implications. Agricultural Economics Research Review, 21:145–172, 2008.

14. M Zeigler and A Steensland. Global Agricultural Productivity Report (GAP Report). Global Harvest Initiative, 2016.

64 CHAPTER 1

15. R.J. Gordon. Is U.S. Economic Growth Over? Faltering Innovation Confronts the Six Headwinds. 18315, 2012.

16. R. Shackleton. Total factor productivity growth in historical perspective. Working Paper 2013-01, 2013.

17. E Barrios. Soil biota, ecosystem services and land productivity. Ecological Economics, 64 (2):269–285, 2007.

18. Emily A Martin, Björn Reineking, Bumsuk Seo, and Ingolf Steﬀan-Dewenter. Natural enemy interactions constrain pest control in complex agricultural landscapes. Proceedings of the National Academy of Sciences, 110(14):5534–5539, 2013. doi: 10.1073/pnas.1215725110.

19. N Gallai, J Salles, J Settele, and B Vaissière. Economic valuation of the vulnerability of world agriculture confronted with pollinator decline. Ecological Economics, 68(3):810–821, 2009.

20. L Garibaldi, I Steffan-Dewenter, C Kremen, J Morales, R Bommarco, S Cunningham, L Car- valheiro, N Chacoff, J Dudenhffer, S Greenleaf, A Holzschuh, R Isaacs, K Krewenka, Y Man- delik, M Mayfield, L Morandin, S Potts, T Ricketts, H Szentgyrgyi, B Viana, C Westphal, R Winfree, and A Klein. Stability of pollination services decreases with isolation from natural areas despite honey bee visits. Ecology Letters, 14(10):1062–1072, 2011.

21. P. L. Pingali. Green Revolution: Impacts, limits, and the path ahead. Proceedings of the National Academy of Sciences, 109(31):12302–12308, 2012.

22. Jonathan A Foley, Ruth Defries, Gregory P Asner, Carol Barford, Gordon Bonan, Stephen R Carpenter, F Stuart Chapin, Michael T Coe, Gretchen C Daily, Holly K Gibbs, Joseph H Helkowski, Tracey Holloway, Erica A Howard, Christopher J Kucharik, Chad Monfreda, Jonathan A Patz, I Colin Prentice, Navin Ramankutty, and Peter K Snyder. Global consequences of land use. Science (New York, N.Y.), 309(5734):570–574, 2005.

23. Michel Loreau and Claire de Mazancourt. Biodiversity and ecosystem stability: A synthesis of underlying mechanisms. Ecology Letters, 16(s1):106–115, 2013.

24. N. M. Haddad, L. a. Brudvig, J. Clobert, K. F. Davies, a. Gonzalez, R. D. Holt, T. E. Lovejoy, J. O. Sexton, M. P. Austin, C. D. Collins, W. M. Cook, E. I. Damschen, R. M. Ewers, B. L. Foster, C. N. Jenkins, a. J. King, W. F. Laurance, D. J. Levey, C. R. Mar- gules, B. a. Melbourne, a. O. Nicholls, J. L. Orrock, D.-X. Song, and J. R. Townshend. Habitat fragmentation and its lasting impact on Earth’s ecosystems. Science Advances, 1 (2):e1500052–e1500052, 2015. ISSN 2375-2548. doi: 10.1126/sciadv.1500052.

25. Otso Ovaskainen and Ilkka Hanski. Transient dynamics in metapopulation response to perturbation. Theoretical population biology, 61(3):285–295, 2002.

26. David Tilman, Robert M. May, Clarence L. Lehman, and Martin A. Nowak. Habitat destruction and the extinction debt. Nature, 371(6492):65–66, 1994.

27. Forest Isbell, David Tilman, Stephen Polasky, and Michel Loreau. The biodiversity- dependent ecosystem service debt. Ecology Letters, 18(2):119–134, 2015.

28. Ilkka Hanski and Otso Ovaskainen. Extinction debt at extinction threshold. Conservation Biology, 16(3):666–673, 2002.

29. D. H. Meadows, D. L. Meadows, J. Randers, and W. W. Behrens III. The limits to growth. Universe Books, New York, US, 1972.

65 CHAPTER 1

30. D. H. Meadows, D. L. Meadows, and Randers. Limits to growth: the 30 year update. EarthScan, London, Sterling, VA, 2004.

31. G. Turner. Is Global Collapse Imminent? MSSI Research Paper No. 4, 2014.

32. J Diamond. Collapse: How Societies choose to fail or survive. Penguin, 2005.

33. Clive Ponting. A green History of the World: The Environment & the Collapse of Great Civilizations. Number 35. Penguin, 1991.

34. James A. Brander. Viewpoint: Sustainability: Malthus revisited? Canadian Journal of Economics, 40(1):1–38, 2007.

35. James A. Brander and M. Scott Taylor. The Simple Economics of Easter Island: A Ricardo- Malthus Model of Renewable Resource Use. American Economic Review, 88(1):119–138, 1998.

36. R.J. Scholes and R Biggs. A biodiversity intactness index. Nature, 434:45–49, 2005.

37. W. Steﬀen, K. Richardson, J. Rockstrom, S. E. Cornell, I. Fetzer, E. M. Bennett, R. Biggs, S. R. Carpenter, W. de Vries, C. A. de Wit, C. Folke, D. Gerten, J. Heinke, G. M. Mace, L. M. Persson, V. Ramanathan, B. Reyers, and S. Sorlin. Planetary boundaries: Guiding human development on a changing planet. Science, 347(6223):1259855–1259870, 2015.

38. D.U. Hooper, E.C. Adair, B.J. Cardinale, J.E.K Byrnes, B.A. Hungate, K.L. Matulich, A Gonzalez, J.E. Duﬀy, L Gamfeldt, and M.I. O’Connor. A global synthesis reveals biodiversity loss as a major driver of ecosystem change. Nature, 486:105–108, 2012.

39. G Mace, B Reyers, R Alkemade, R Biggs, F Chapin III, S Cornell, S Diaz, S Jennings, P Leadley, P Mumby, A Purvis, R Scholes, A Seddon, M Solan, W Steﬀen, and G Woodward. Approaches to deﬁning a planetary boundary for biodiversity. Global Environmental Change, 28:289–297, 2014.

40. T Oliver. How much biodiversity loss is too much? Nature, 353(6296):220–221, 2016.

41. O. Galor and D.N. Weil. Population, Technology, and Growth: From Malthusian Stagnation to the Demographic Transition and beyond. The American Economic Review, 90(4):806– 828, 2000.

42. T Kogel and A Prskawetz. Agricultural Productivity Growth and Escape from the Malthu- sian Trap. Journal of Economic Growth, 6:337–357, 2001.

43. P.F. Peretto and S Valente. Growth on a ﬁnite planet: resources, technology and population in the long run. Journal of Economic Growth, 20(3):305–331, 2015.

44. John M. Anderies. Economic development, demographics, and renewable resources: a dynamical systems approach. Environment and Development Economics, 8(02):219–246, 2003.

45. K.O. Fuglie. Is a slowdown in agricultural productivity growth contributing to the rise in commodity prices? Agricultural Economics, 39:431–441, 2008.

46. W B Meyer and B L Turner. Human Population Growth and Global Land-Use/Cover Change. Annual Review of Ecology and Systematics, 23(1):39–61, 1992. ISSN 0066-4162. doi: 10.1146/annurev.es.23.110192.000351.

47. R.H. Mac Arthur and E.O. Wilson. The theory of island biogeography. Monographs in Population Biology, 1967.

66 CHAPTER 1

48. H.M. Pereira, G. Ziv, and M. Miranda. Countryside Species–Area Relationship as a Valid Alternative to the Matrix-Calibrated Species–Area Model. Conservation Biology, 28(3): 874–876, 2014.

49. E.F. Connor and E.D. McCoy. The statistics and biology of the species-area relationship. American Naturalist, 113:119–130, 1979.

50. K.A. McGuiness. Equations and explanations in the study of species-area curves. Biological Reviews, 59:423–440, 1984.

51. G. Ferraz, J.D. Nichols, J.E. Hines, P.C. Stouﬀer, Jr. Bierregaard, and T.E. Lovejoy. A large-scale deforestation experiment: Eﬀects of patch area and isolation on Amazon birds. Science, 315(5809):238–241, 2007.

52. O. R. Wearn, D. C. Reuman, and R. M. Ewers. Extinction Debt and Windows of Conser- vation Opportunity in the Brazilian Amazon. Science, 337(6091):228–232, 2012.

53. J.M. Diamond. Biogeographic kinetics: estimation of relaxation times for avifaunas of southwest paciﬁc islands. Proc.Nat.Acad.Sci. USA, 69(11):3199–3203, 1972.

54. A.-M. Klein, Bernard E, James H Cane, Ingolf Steﬀan-Dewenter, Saul A Cunningham, Claire Kremen, and Teja Tscharntke. Importance of pollinators in changing landscapes for world crops. Proceedings of the Royal Society of London B: Biological Sciences, 274(1608): 303–313, 2007.

55. C Kremen. Managing ecosystem services: what do we need to know about their ecology? Ecology Letters, 8(5):468–479, 2005.

56. A O’Connor, M.I.and Gonzalez, J.E.K Byrnes, B.J. Cardinale, J.E. Duﬀy, L. Gamfeldt, J.N. Griﬃn, D. Hooper, B.A. Hungate, A Paquette, P.L. Thompson, L.E. Dee, and K.L. Dolan. A general biodiversity-function relationship is mediated by trophic level. Oikos, 126 (1):18–31, 2017.

57. Camilo Mora, Roberto Danovaro, and Michel Loreau. Alternative hypotheses to explain why biodiversity-ecosystem functioning relationships are concave-up in some natural ecosystems but concave-down in manipulative experiments. Scientiﬁc reports, Nature, 4:5427, 2014.

58. S. Dullinger, F Essl, W Rabitsch, K.-H. Erb, S Gingrich, H Haberl, K Hülber, V Jarosik, F Krausmann, I Kühn, J Pergl, P Pysek, and P.E. Hulme. Europe’s other debt crisis caused by the long legacy of future extinctions. Proceedings of the National Academy of Sciences, 110(18):7342–7347, 2013.

59. Isabel M.D. Rosa, M.J. Smith, O.R. Wearn, D Purves, and R.M. Ewers. The environmental legacy of modern tropical deforestation. Current Biology, 26:2161–2166, 2016. doi: 10.1016/j.cub.2016.06.013.

60. J Barlow, G.D Lennox, J Ferreira, E Berenguer, A. C. Lees, Ralph Mac Nally, James R. Thomson, Silvio Frosini de Barros Ferraz, Julio Louzada, Victor Hugo Fonseca Oliveira, Luke Parry, Ricardo Ribeiro de Castro Solar, Ima C. G. Vieira, Luiz E. O. C. Aragão, Rodrigo Anzolin Begotti, Rodrigo F. Braga, Thiago Moreira Cardoso, Raimundo Cosme de Oliveira Jr, Carlos M. Souza Jr, Nárgila G. Moura, Sâmia Serra Nunes, João Victor Siqueira, Renata Pardini, Juliana M. Silveira, Fernando Z. Vaz-de Mello, Ruan Carlo Stulpen Veiga, Adriano Venturieri, and Toby A. Gardner. Anthropogenic disturbance in tropical forests can double biodiversity loss from deforestation. Nature, 535:144–147, 2016. doi: 10.1038/nature18326.

67 CHAPTER 1

61. K Norris. Ecology: the tropical deforestation debt. Current Biology, 26:R756–R777, 2016.

62. W.D. Nordhaus. Optimal Greenhouse-gas Reductions and Tax Policy in the "DICE" model. The American Economic Review, 83(2):313–317, 1993.

63. W.D. Nordhaus. Economic aspects of global warming in a post-Copenhagen environment. Proceedings of the National Academy of Sciences, 107(26):11721–11726, 2010.

64. O Venter, E.W. Sanderson, A. Magrach, J.R. Allan, J. Beher, K.R. Jones, H.P. Possingham, W.F. Laurance, P. Wood, B.M. Fekete, M.A. Levy, and J.E.M Watson. Sixteen years of change in the global terrestrial human footprint and implications for biodiversity conservation. Nature Communications, 7(12558), 2016.

65. John Bongaarts. Human population growth and the demographic transition. Philosophical transactions of the Royal Society of London B: Biological sciences, 364(1532):2985–90, 2009. ISSN 1471-2970. doi: 10.1098/rstb.2009.0137.

66. Corey J A Bradshaw and Barry W Brook. Human population reduction is not a quick ﬁx for environmental problems. Proceedings of the National Academy of Sciences, 111(46): 16610–16615, 2014.

67. Tobias Naaf and Jens Kolk. Colonization credit of post-agricultural forest patches in NE Germany remains 130-230years after reforestation. Biological Conservation, 182:155–163, 2015.

68. Martine Maron, Richard J. Hobbs, Atte Moilanen, Jeﬀrey W. Matthews, and Kimberly Christie. Faustian bargains? Restoration realities in the context of biodiversity oﬀset policies. Biological Conservation, 155:141–148, 2012.

69. Tim Newbold, Lawrence N. Hudson, Andrew P. Arnell, Sara Contu, Adriana De Palma, Simon Ferrier, Samantha L. L. Hill, Andrew J. Hoskins, Igor Lysenko, Helen R. P. Phillips, Victoria J. Burton, Charlotte W. T. Chng, Susan Emerson, Di Gao, Gwilym Pask-Hale, Jon Hutton, Martin Jung, Katia Sanchez-Ortiz, Benno I. Simmons, Sarah Whitmee, Han- bin Zhang, Jörn P. W. Scharlemann, and Andy Purvis. Has land use pushed terrestrial biodiversity beyond the planetary boundary? A global assessment. Science, 353:288–291, 2016.

70. Eric F. Lambin, B. L. Turner, Helmut J. Geist, Samuel B. Agbola, Arild Angelsen, John W. Bruce, Oliver T. Coomes, Rodolfo Dirzo, Günther Fischer, Carl Folke, P. S. George, Kather- ine Homewood, Jacques Imbernon, Rik Leemans, Xiubin Li, Emilio F. Moran, Michael Mor- timore, P. S. Ramakrishnan, John F. Richards, Helle Skånes, Will Steﬀen, Glenn D. Stone, Uno Svedin, Tom A. Veldkamp, Coleen Vogel, and Jianchu Xu. The causes of land-use and land-cover change: Moving beyond the myths. Global Environmental Change, 11(4): 261–269, 2001.

71. Robert Costanza, Rudolf de Groot, Paul Sutton, Sander van der Ploeg, Sharolyn J. Ander- son, Ida Kubiszewski, Stephen Farber, and R. Kerry Turner. Changes in the global value of ecosystem services. Global Environmental Change, 26(1):152–158, 2014.

68 Appendix

69 CHAPTER 1

1.A Endogenous or Exponential Technological Change

In this appendix, the assess the eﬀect of other forms of technological change on the dynamics of the system.

Technological efficiency T represents the Total Factor Productivity (TFP) of the industrial and agricultural sectors. Following recent concerns regarding the potential for continued TFP growth in the future (Gordon, 2012; Shackelton, 2013), we choose to model technological change as a logistic function, such that technological efficiency grows at an exogenous rate σ, until it reaches a maximum technological efficiency (Tm).

˙ T = σT (1 − T/Tm) (16)

This function allows us to retain analytical tractability, and for instance, derive analytical sustainability criteria. We could also have modeled technological change as endogenous, e.g. driven by the rate of human population growth :

˙ T = σ(µ)T (1 − T/Tm) (17) where σ(µ) is a function of the human growth rate, for instance σ(µ) = µ(B,T ). The behaviour of the system is not qualitatively aﬀected by this assumption. However, the system becomes analytically intractable.

Another possibility is to assume an exponential technological progress, i.e. technological eﬃciency keeps growing instead of reaching a maximum technological eﬃciency

Tm: T˙ = σT (18)

In this case, the system has no global equilibrium (H∗,T ∗,B∗) and no analytical results. However, as technological change increases the strength of the demographic transition, it reduces the growth rate of the human population (µ(T,B) → 0) along with land conversion, so that the human population and biodiversity reach a stationary state in the

70 CHAPTER 1 long run.

In both cases (endogenous or exponential technological change), however, our results would not be qualitatively affected. Indeed, with endogenous technological change and a strong demographic transition (i.e. high value of the sensitivity of the human growth rate to industrial consumption b2), the effect of the maximum technological efficiency Tm on the vulnerability to environmental crises switches from negative to positive and reduces the amplitude of the transient environmental crises (Fig. A1.1.A). This corresponds to the dashed curve in fig.3.D (b2 = 2). Similarly, with an exponential technological change, the effect of varying labor intensity α1 is non-linear (Fig. A1.1.B), as it first reduces environmental crises before increasing human population size again, through an economic rebound effect. This corresponds to fig.3.B and E.

A B 1 1 10 10 9 9 ) )

* 8 8 * 7 7

6 6

0.5 0.5 5 5 4 4

3 3 Eq. biodiversity (B

Eq. human pop. (H 2 2

1 1

0 0 0 0 0 10 20 30 40 50 60 0 0.2 0.4 0.6 0.8 1 Max. tech. efficiency (T ) Labor intensity (α ) m 1

Figure A1.1: Effect of endogenous (A) and exponential (B) technological change on long-term sustainability. A: effect of an endogenous technological change (σ = µ(B, T)) on the transient and long-term dynamics of the system, when the maximum technological efficiency Tm varies. Grey areas represent the amplitude of the transient environmental crises. Parameters are given in Table 1, with Ω = 2 and b2 = 0.5.; B: effect of an infinite technological change (T˙ = σ) on the transient and long-term dynamics of the system, when the labor intensity α1 varies. Parameters are given in Table 1, with Ω = 2, b2 = 0.1 and σ = 0.1.

71 CHAPTER 1

1.B Economic derivations

In this section, the aim is to derive the per capita agricultural and industrial consumptions at the market equilibrium, i.e. when the production supply equals the demand of the human population. The human population is assumed to be composed of identical agents, with preferences η. Each agent allocates part of his revenue w to buy agricultural goods, and the rest to buy industrial goods, which generates an aggregate demand for the human population H (section 3.A). On the supply side, ﬁrms produce agricultural and industrial goods given the costs of production inputs, i.e. land and labor, which gives the aggregate supply of the production sectors (section 3.A). Section 3.A solves for the general equilibrium model, and derives the per capita consumptions at the market equilibrium. Section 3.A then derives the rate of land conversion at the economic market equilibrium.

1.B.1 Consumer Optimization

Overall, each agent supplies one unit of labor, so that his revenue is the wage w. Let

U(y1, y2) be the individual consumer well-being, where y1 and y2 are per capita consumption rates of agricultural and industrial goods. Agents maximize their well-being

η 1−η U(y1, y2) = y1 y2 under the revenue constraint p1y1 + p2y2 ≤ w, where p1 and p2 are the prices of agricultural and industrial goods respectively, w the per capita income, and η the preference for agricultural goods. To solve this maximization problem, we deﬁne the Lagrangian:

L ≡ U(y1, y2) − Λ[p1y1 + p2y2 − w]

First order conditions are:

72 CHAPTER 1

  ηU(y1,y2) ∂L/∂y1 = y − Λp1 = 0 1 (19)  (1−η)U(y1,y2) ∂L/∂y2 = − Λp2 = 0  y2

Adding both conditions yields U(y1, y2)/Λ = p1y1 + p2y2 = w, and solving (19) for y1 and y2, and substituting for U(y1, y2)/Λ yields the aggregate demands:

D D p1Y1 = ηwH p2Y2 = (1 − η)wH (20)

1.B.2 Firms Optimization

Firms in sectors 1 and 2 produce a quantity Yi (i = 1, 2) of output, using labor Li and land Ai. A units of land and H units of labor are available. Agricultural firms occupy A1 units of land, and industrial firms occupy A2 units of land, such that A = A1 + A2. This comes at an operating cost (including clearance and maintenance) of κ units of labor per unit of land, so that each firm maximizes a profit Πi = piYi −wLi −κwAi. The production functions Yi are:

α1 1−α1 α2 1−α2 Y1 = T fS(B) L1 A1 Y2 = T L2 A2 (21)

where fS(B) is the ecological feedback, T captures production eﬃciency and αi is labor intensity in sector i. Therefore, ﬁrst order conditions are:

  αipiYi ∂Πi/∂Li = L − w = 0 i (22)  (1−αi)piYi ∂Πi/∂Ai = − κw = 0  Ai

Adding both lines of (23) gives the total supply in sector i:

S piYi = w(Li + κAi) (23)

73 CHAPTER 1

1.B.3 Economic General Equilibrium

D S Markets’ clearing for the agricultural and industrial sectors piYi = piYi (eq. (21) and (24)) yields:

  ηH = L1 + κA1 (24)  (1 − η)H = L2 + κA2

Input factors in each sector verify Li/αi = κAi/(1 − αi) (see eq. (23)), so that the equilibrium land allocations are:

A1 = Hη(1 − α1)/κ A2 = H(1 − η)(1 − α2)/κ (25)

and the equilibrium labor allocations are:

L1 = Hηα1 L2 = H(1 − η)α2 (26)

1.B.4 Land Conversion

The allocation of total labor H between agricultural production (L1), industrial production

(L2) and land conversion (κA) writes L1 + L2 + κA = H. Replacing L1 and L2 by their optimal allocations (eq.(26)), we deduce the relationship A = H/φ between the human population size H and the converted land A, where the density of the human population per unit of converted land is:

κ φ = (27) 1 − α1η − α2(1 − η)

74 CHAPTER 1

1.B.5 Per capita consumptions

Using the equilibrium allocations of labor Li and land Ai, the per capita agricultural and industrial consumptions Y/H (eq.(22)) can be rewritten:

Ω y1 = γ1B T/Tm y2 = γ2T/Tm (28)

where

α1 1−α1 α2 1−α2 γ1 = ηTmα1 ((1 − α1)/κ) γ2 = (1 − η)Tmα2 ((1 − α2)/κ) (29)

1.C Dynamical System Analysis

In this section, we derive the analytical expressions of the equilibria, and the conditions for their stability, as well as a condition for the sustainability of the temporal trajectories, i.e. non-decreasing human well-being over time.

1.C.1 Equilibria and Stability Condition

∗ ∗ The Jacobian matrix at the viable equilibrium (H ,Tm,B ) writes:   ∗ Ω−1∗ ∗ Ω∗  0 µmaxγ1exp(−b2γ2)ΩH B µmaxγ1exp(−b2γ2)H B      J ∗ = − z (1 − H∗/φ)z−1 − 0   φ     0 0 −σ 

∗ The eigenvalues λi (i = 1 : 3) of the system are solutions of Det(J − λI) = 0, where

∗ Det(J − λI) = −(σ + λ)(λ( + λ) + θµmax/4) (30)

and z θ = 4 (1 − H∗/φ)z−1γ exp(−b γ )ΩH∗BΩ−1∗ (31) φ 1 2 2

The ﬁrst eigenvalue is λ1 = −σ, and the two others are solution of the characteristic

2 equation λ + λ + θµmax/4 = 0, which discriminant is:

75 CHAPTER 1

D = ( − θµmax) (32)

√ √ λ2 = (− − D)/2 λ3 = (− + D)/2 (33)

The viable equilibrium is stable if all the eigenvalues have negative real parts, i.e. √ ∗ ∗ D < , which gives a stability condition for the viable equilibrium (H ,Tm,B ):

∗ min θ > 0 ⇔ H < φ ⇔ γ1 > y1 (34)

1.C.2 Sustainability Condition

min For parameters such that the viability condition γ1 > y1 is met, there are two possible transient dynamics depending on the eigenvalues of the system: (1) for real eigenvalues, a monotonous convergence to equilibrium, and (2) for complex conjugate eigenvalues, damped oscillations. The former case stands for a sustainable system, while the later one stands for an environmental crisis. The eigenvalues are real if and only if the discriminant of the characteristic equation is positive, so that a sustainability condition for our system is:

> θµmax (35)

A sustainability criterion for our system is ∆ = − θµmax, where ∆ > 0 stands for sustainable trajectories.

1.C.3 Sustainability Thresholds

Using the relationship between the biodiversity and human population size at equilibrium H∗ = φA∗, θ (eq.(31)) can be rewritten as a function of the equilibrium converted area A∗. The sustainability criterion ∆ then rewrites:

0 ! ∗ θ µmax A ≤ 1 − 0 (36) + θ µmax

76 CHAPTER 1

0 min −b2γ2 where θ = 4Ωzy1 e , and is the ecological relaxation rate. Over a sustainable trajectory (∆ > 0), converted land never exceeds its equilibrium level, i.e. A ≤ A∗.

Condition (36) is thus equivalent to A ≤ AS, where AS is the sustainable land conversion threshold of our system. Similarly, the sustainability criterion can be written as a human or a biodiversity threshold as follows:

∗ ∗ z H ≤ φAS B ≥ (1 − AS)

1.D Eﬀect of Land Operating Costs on Sustainability

This section presents the qualitative eﬀect of the parameter κ, which captures the cost of natural habitat conversion, on the dynamics of the system.

The exploitation of converted land comes with operating costs at each period of time, which include the initial cost of natural habitat conversion, and the cost of land maintenance. Operating costs κ are in units of human labor per unit of converted land. High operating costs are beneﬁcial to the long-term sustainability of the SES (Fig.A1.2.B), since they reduce the incentive to convert natural habitat - in a similar manner to taxes on converted land. Therefore, biodiversity at equilibrium increases with κ (Fig.A1.2.A) while consumption utility decreases (Fig.A1.2.C). The eﬀect of κ on the size of the human population at equilibrium H∗ is non-linear (Fig.A1.2.A), since low values reduce the

min population density φ, and high values make the system unviable (γ1 → y1 ), i.e. natural habitat conversion becomes to expensive.

77 CHAPTER 1

A 1 1 ) ) * *

0.5 0.5 Eq. biodiversity (B Eq. human pop. (H 0 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 B 2

) 0 ∆

−2

−4

−6

Sust. criterion ( −8

−10 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 C 0.7 )

* 0.65

0.6

Eq. utility (U 0.55

0.5 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 Land conversion cost (κ) Figure A1.2: Effect of the land conversion cost (κ). A: Effect of varying κ on the ∗ ∗ viable equilibria (H ,B ), with default parameters of Table 3.1 and Tm = 1.8. The grey area represents the amplitude of the transient crises. B: Effect of κ on the sustainability criterion (∆). Values of κ for which ∆ is positive correspond to sustainable trajectories. C: Effect of κ on human well-being at equilibrium. The grey area represents the amplitude of the transient variations in well-being.

78 CHAPTER 2

Delayed behavioral shifts undermine the sustainability of social-ecological systems

Anne-Sophie Lafuite 1, Claire de Mazancourt 1, Michel Loreau 1

Accepted in Proceedings of the Royal Society B.

1Centre for Biodiversity Theory and Modelling, Theoretical and Experimental Ecology Station, CNRS and Paul Sabatier University, Moulis, France

79 CHAPTER 2

Chapter outline

Considering the lack of models capturing the dynamic feedback loop between human population growth and biodiversity loss, human behavior and social aspects were not considered in Chapter 1. However, human behavioral change underpins the adaptive capacity of social-ecological systems, and can strongly aﬀect their long-term dynamics. Chapter 2 aims at investigating the potential of social norms in enforcing sustainability, when behavioral change is based on perceived environmental degradation. This work beneﬁts from recent studies on norm-driven cooperation in the use of natural resources.

80 CHAPTER 2

Abstract

Natural habitat loss and fragmentation generate a time-delayed loss of species and associated ecosystem services. Since social-ecological systems (SESs) depend on a range of ecosystem services, lagged ecological dynamics may affect their long-term sustainability. Here, we investigate the role of consumption changes in sustainability enforcement, under a time-delayed ecological feedback on agricultural production. We use a stylized model that couples the dynamics of biodiversity, technology, human demography and compliance to a social norm prescribing sustainable consumption. Compliance to the sustainable norm reduces both the consumption footprint and the vulnerability of SESs to transient overshoot-and-collapse population crises. We show that the timing and interaction between social, demographic and ecological feedbacks govern the transient and long-term dynamics of the system. A sufficient level of social pressure (e.g. disapproval) applied on the unsustainable consumers leads to the stable coexistence of unsustainable and sustainable or mixed equilibria, where both defectors and conformers coexist. Under bistability conditions, increasing time delays reduces the basin of attraction of the mixed equilibrium, thus resulting in abrupt regime shifts towards unsustainable pathways. Given recent evidence of large ecological relaxation rates, such results call for farsightedness and a better understanding of lag effects when studying the sustainability of coupled SESs.

Keywords: Biodiversity ; lagged feedback ; regime shifts ; social-ecological system ; social norms ; sustainability

81 CHAPTER 2

Introduction

Early research on the interaction between human populations and their environment emphasized the need for government control in order to prevent the overexploitation of common pool natural resources [1]. However, subsequent research has shown that local communities can achieve sustainable resource use through cooperative self-governance [2]. Successful communities often establish social norms, i.e. rules of shared behavior, that protect common natural resources [3] or help achieve group interests [4]. Such regulatory mechanisms are “bottom-up” processes, as opposed to classical “top-down” government control. Both types of regulation involve, however, sanctioning mechanisms that seek to internalize the externalities of human activities, be they moral incentives (e.g. social exclusion) in the ﬁrst case, or economic instruments (e.g. taxes, subsidies) and regulatory policies in the second [5, 6].

Human behavioral change can significantly affect the dynamics of social-ecological systems (SESs) [7], and is a central aspect of their adaptability and resilience [8, 9]. The evolution of social norms affects feedbacks and drivers of SESs, potentially leading to large-scale behavioral shifts [10]. Such shifts may allow escape from social-ecological traps, i.e. persistent mismatches between the responses of people and their ecological conditions, that are undesirable from a sustainability perspective [11].

The establishment of sustainable social norms strongly hinges on the ecological characteristics of SESs. Previous experimental and theoretical studies have emphasized the role of resource productivity and mobility [12] as well as temporal variability [13, 14] on the robustness of cooperation. Evidence from the literature on natural resource management shows that the interaction between fast and slow ecosystem processes aﬀects the optimal management strategy [15], while inappropriate management may reinforce undesirable feedbacks and push the SES into a social-ecological trap [16]. However, the consequences of mismatches between slow and fast social-ecological processes on the robustness of cooperation remain an open question.

Extinction debts, i.e. time-delayed loss of species following habitat destruction, can

82 CHAPTER 2 emerge from the spatial dynamics of SESs [17]. Indeed, land conversion and fragmentation alter spatial ecological processes and the rate of relaxation of natural communities, i.e. the rate of change of species richness in response to habitat destruction [18]. Recent evidence suggests increasing ecological relaxation rates, generating large extinction debts [19]. As an example, 80% of the species extinctions in the Amazon are estimated to be pending [20], which may increase the number of 20th-century extinctions in bird, mammal, and amphibian forest-speciﬁc species by 120% [21]. In European landscapes, studies ﬁnd that extinctions lag well behind contemporary levels of socioeconomic pressures [22], the current number of threatened species being better explained by socio-economic indicators from the early or mid-20th century [23].

The accumulation of these extinction debts generates functioning debts [24] that postpone the negative eﬀect of biodiversity loss on ecosystem processes. Since many of the ecosystem services that play a direct or indirect role in agricultural production depend on biodiversity [25, 26, 27, 28], current species extinction rates [29, 30, 31] do not only threaten the long-term provisioning [32, 33] and stability of ecosystem processes [34, 35, 36], they also generate a time-delayed feedback loop between humans and nature [37]. In the long run, such time-delayed biodiversity feedbacks may result in large environmental crises, i.e. overshoot-and-collapse population cycles [37], similar to the famine cycles that have been observed in extinct societies [38].

Characteristics common to the majority of modern agricultural systems were found to increase the vulnerability of SESs to such crises [37]. Among these characteristics are a high production efficiency and a low labor share per unit of agricultural good, due to the substitution of technology (e.g. machines, fertilizers and pesticides) for human labor and ecosystem services. Recent evidence suggests that land use efficiency has been rising at the global scale [39]. However, it is not clear whether these efficiency gains will help save natural habitats and biodiversity in the long run, due to economic rebound effects [40], i.e. when lower prices stimulate demand and higher yields raise profits, encouraging further agricultural expansion [41]. Such a decoupling between human population growth and ecological dynamics can thus reinforce unsustainable feedbacks [42].

83 CHAPTER 2

Shifting consumption has, however, been identiﬁed as a major strategy that could allow doubling food production while greatly reducing the environmental impacts of agriculture [43, 44]. Norm-driven consumption changes towards more environmentally-friendly agricultural goods, whose production relies more on ecosystem services and labor than on technology, may thus play a key role in ensuring the long-term sustainability of SESs at large scales. However, the magnitude of time-delayed ecological feedbacks may postpone the required behavioral changes, and push (or keep) the global SES into a social-ecological trap [42].

The aim of this article is to investigate the effects of time-delayed biodiversity loss on the establishment of sustainable social norms. To this end, we develop a dynamical system model of an endogenously growing human population divided into norm-following and norm-violating consumers, that share a common stock of land and associated biodiversity. Rising consumption demand of the human population drives production supply and natural habitat conversion through market constraints. The model thus differs from related common-pool resource systems that only consider a constant population of harvesters and a single resource [13]. The present model builds upon previous work [37], where the growth rate of the human population depends on the consumptions of industrial and agricultural goods, as well as on the strength of the demographic transition governed by technological change. The time-delayed loss of biodiversity-dependent ecosystem services then acts as a lagged feedback on agricultural productivity that can push the system into an overshoot-and-collapse crisis [38]. In the following, we present the model structure and show that allowing for consumers’ behavioral change generates bistability between sustainable and unsustainable equilibria, and thus the potential for regime shifts. We then explore different scenarios of social pressure and extinction debts, and conclude with a discussion of our results.

84 CHAPTER 2

1 Model description

1.1 Coupling human demography, biodiversity and social dy-

namics

We model a population of consumers, whose demand for agricultural and industrial goods requires the conversion of their common natural habitat. Our SES model describes the long-term interaction between four dynamical variables (Fig.2.1): the human population (H), technological efficiency (T), biodiversity (B) and the proportion of sustainable consumers, hereafter “conformers” (q). Conformers, by complying to a sustainable norm prescribing the consumption of environmentally-friendly agricultural goods, reduce their footprint in terms of natural habitat destruction and long-term biodiversity loss. Total habitat is gradually converted towards agricultural and industrial lands. The remaining natural habitat supports a community of species (biodiversity) that provides a range of ecosystem services to agricultural production [35]. Loss of natural habitat leads to time-delayed species extinctions, thus reducing both the common-pool biodiversity and long-term agricultural productivity [45]. Such a lagged feedback on agricultural production can result in long-term environmental crises characterized by overshoot-and-collapse population cycles (Fig.2.4.c). These crises transiently reduce human well-being [37], thus threatening intergenerational equity and sustainability [46]. Since the vulnerability of SESs to lag effects increases with natural habitat destruction and biodiversity loss [37], a sufficient proportion of conformers reducing their consumption footprint may help limit land conversion while preserving the long-term sustainability of the SES. The following sections present the main features of our dynamical system. Further details about the economic derivations can be found in Lafuite & Loreau (2017) [37], from which the model is extended.

85 CHAPTER 2

α α (a) u (b) s 1 ) Human population (H) * 0.8 Conformers Defectors 0.6 (q) w(q) (1-q) 0.4 ) ,,q y ,,T ) α 2 (T B (q ) 0.2 ( α 2 y 1 Eq. biodiversity (B 0 Agriculture Technology Industry 0 0.2 0.4 0.6 0.8 1 (T) (c) Ω 1 B -α 0.5 (q -α 2 q) 1 ) Biodiversity ∆ n 0 iio (B) s d rr d e n v Converted land a v L n L o A(H,q) c −0.5 S( H, q) Sust. criterion ( Natural habitat −1 1-A(H,q) 0 0.2 0.4 0.6 0.8 1 Labor intensity (α)

Figure 2.1: Coupling between human, social and ecological dynamics, and definition of the sustainable and unsustainable consumption norms. (a) Model summary Black boxes: production sectors; grey boxes: dynamical variables; dashed lines: production inputs (labor, land and technology), with α(q) being the share of labor compared to land to produce one unit of agricultural good; solid lines: per capita consumptions of agricultural and industrial goods, y1(B, T, q) and y2(T); grey dotted lines: ecological feedback; double arrow: social sanctioning (e.g. ostracism); circle: total land divided into converted land A(H, q) and natural habitat, which supports a long-term species richness S(H, q). All functions are explicitly defined in the main text and in Table S2 (electronic supplementary material). (b) Effect of labor elasticity on equilibrium biodiversity. Grey areas represent the amplitude of the transient environmental crises. (c) Effect of labor elasticity on sustainability. The sustainability criterion ∆ is derived in [37]. ∆ > 0 stands for sustainable transient trajectories, i.e. no environmental crises. The sustainability-optimal agricultural labor elasticity αs maximizes both ∆ and ∗ the biodiversity at equilibrium, B . The unsustainable labor elasticity αu is chosen so that αu > αs and ∆(αu) < 0.

86 CHAPTER 2

1.2 Human consumption and technological change

Human consumption is related to the production of agricultural and industrial goods through an auxiliary economic model, which is assumed to be at a moving market equilibrium. Following previous work [47, 37], the eﬀects of both biodiversity and technology on agricultural (j = 1) and industrial (j = 2) productions, e.g. ecosystem services, chemicals and machines, are captured by the total factor productivity (TFP) term of

Cobb-Douglas production functions, with labor Lj and land Aj as inputs (eq.(1)).

Ω α 1−α α2 1−α2 Y1 = B T L A Y2 = T L A (1) | {z } 1 1 |{z} 2 2 TFP TFP where α and α2 are the labor elasticities of the agricultural and industrial sectors, respectively. The relationship between biodiversity and ecosystem services [35] is captured by a concave-down function of biodiversity, BΩ, with Ω < 1 [48], and the feedback of ecosystem services on industrial production is neglected. For simplicity’s sake, technology is taken as exogenous, i.e. independent from the human and ecological dynamics, and technological eﬃciency is assumed to follow a logistic growth at a rate σ towards a maximum eﬃciency,

Tm (eq.(4)). Such a logistic growth allows reproducing the past rise and current stagnation of the agricultural TFP [49]. Other forms of technological change, e.g. exponential or endogenous, do not qualitatively aﬀect the dynamics of the model [37].

˙ T = σ T (1 − T/Tm) (2)

Solving for the market equilibrium, i.e. when supply equals demand, gives the per capita industrial and agricultural consumptions as functions of biodiversity and technological eﬃciency. Industrial consumption, y2 = γ2T/Tm, varies with technological eﬃciency

Ω only, while agricultural consumption y1i = γ1iB T/Tm of conformers (i = s) and defectors (i = u) also depends on biodiversity-dependent ecosystem services. Both industrial and agricultural consumptions increase with technological eﬃciency, T. By increasing production eﬃciency (eq.(4)), technological change helps counterbalancing the feedback

87 CHAPTER 2 of biodiversity loss on agricultural productivity in the short term, thus ensuring that the consumption utility of consumers does not decrease with time [37]. γ1i and γ2 are functions of socio-economic parameters that capture the characteristics of agricultural and industrial productions (electronic supplementary material, Table S1). Using these parameters, a norm of sustainable agricultural practices γ1s is deﬁned in the next section.

1.3 A norm of sustainable consumption

The footprint of agricultural goods can be related to the parameters of their production function, and especially to the output elasticity of labor, hereafter denoted as α, and the output elasticity of land, which equals 1 − α (eq.(1)). In economics, output elasticity captures the percent change in production resulting from a 1% change in an input, and is a proxy for the relative share of inputs used in production. Thus, the higher α, the higher the labor force per unit of land used in agricultural production. Agricultural labor forces have been globally declining with the substitution of machines, fertilizers and pesticides for labor and ecosystem services, and the consequent rise in production eﬃciency [50] and economies of scale [51]. Conventional industrialized agricultural systems thus have lower labor elasticities α than environmentally-friendly systems, such as small-scale organic farming, where the substitution of labor and ecosystem services for technology is lower.

In previous work, labor elasticity has been related to the sustainability of SESs, in terms of their vulnerability to overshoot-and-collapse crises [37]. To do so, we have captured the transient dynamics of our SES by a sustainability criterion, ∆. This criterion captures the relative rate of change of biodiversity compared to the human population, since it is the diﬀerence between the ecological relaxation rate, , and the maximum growth rate of the human population, µ, as ∆ = − θµ. The respective roles of these parameters are detailed in the following sections. θ is a function of assessable ecological and economic parameters of the SES (electronic supplementary material, Table S1), and captures the rate of change of species richness in response to land conversion [18, 20]. ∆ > 0 means that the ecological dynamics is fast enough compared to the human dynamics ( > θµ), thus preventing transient overshoot-and-collapse crises. However, ∆ < 0 means that the

88 CHAPTER 2 ecological dynamics is much slower than the human dynamics ( < θµ), so that there is a high probability of experiencing transient crises. A sensitivity analysis of this criterion to the parameters of the system is presented in Appendix 2.D. Using this criterion, we show in Lafuite & Loreau [37] that a low labor elasticity, i.e. a low share of labor in production, or equivalently, a high substitution of human capital for technology, increases the vulnerability of SESs to lag eﬀects, while there exists an intermediate sustainability-optimal labor elasticity that maximizes both long-term biodiversity (Fig.2.1.a) and sustainability (Fig.2.1.b).

Let us deﬁne as αs the sustainability-optimal labor elasticity, and αu < αs an unsustainable labor elasticity chosen such that ∆(αu) < 0 (Fig.2.1.a). The expected agricultural labor elasticity then varies with the proportion of conformers as α(q) = qαs + (1 − q)αu. Through means of eco-labeling, consumers can either buy sustainable agricultural products (y1s), or follow their unsustainable consumption habits and buy unsustainable agricultural products (y1u). Since economic dynamics are much faster than ecological and demographic dynamics, we assume that agricultural and industrial production instantaneously follows consumers’ demand. Such a shift in agricultural production may not be met instantaneously due to inertia and production barriers [52], and farmers’ adaptability may have to be supported through adequate policy changes [53]. However, given the large time scales considered here, it seems reasonable to neglect such time delays with respect to the extent of extinction debts. Thus, in our system, a consumption shift towards sustainable goods, which, in turns, drives a shift towards more environmentally-friendly agricultural practices, may prevent environmental crises.

1.4 Human demography

The growth rate of human populations can be related to consumption levels [47] by capturing basic linkages between technology and human demography [54, 55, 56]. Following previous studies [47, 56], we assume that the human growth rate endogenously varies with the mean agricultural and industrial consumptions, so as to increase with agricultural consumption, and decrease with industrial consumption, capturing the eﬀect of the

89 CHAPTER 2 demographic transition.

ymin−y −b y H˙ = µ H 1 − e 1 1 e 2 2 (3)

min µ is the maximum growth rate, y1 is the minimum consumption threshold, y1 = q ·y1s +

(1 − q) · y1u is the average agricultural consumption, and b2 is the demographic sensitivity to industrial consumption. The strength of the demographic transition thus gradually increases with industrial consumption and limits human population growth [54].

Dependence of the human growth rate on consumption levels also allows coupling human demography with social changes regarding consumption choices. Indeed, conformers do not only have a lower consumption footprint than defectors, it can be shown that they also have a lower agricultural consumption level, i.e. y1s < y1u. As a result, conformers also have a lower reproduction rate compared to defectors. This can be interpreted as a quantity-quality trade-oﬀ in both consumption choices and the number of children, a mechanism which has been shown to partly explain the fertility reductions observed during the demographic transition [57]. Under our assumptions, shifting behaviors towards sustainable consumption habits thus reduces the growth rate of the human population, therefore increasing the sustainability of the SES.

1.5 Land conversion and biodiversity dynamics

The rate of land conversion is also derived at market equilibrium, as function of the dynamical variables of our system under sustainable and unsustainable labor elasticities, αu and αs (see [37] for more details about the economic derivations). For a given proportion of conformers q and human population H, converted area writes A(H, q) = H/φ, where

φ = qφs + (1 − q)φu is the mean population density on converted land, and φu and φs are explicitly deﬁned as function of the economic parameters of the SES in Table S1 (electronic supplementary material).

Natural habitat conversion results in time-delayed changes in species richness [58], so that the long-term species richness may be reached only after decades [20]. These extinction debts [17] are a result of many mechanisms [59] which lower the relaxation rates of communities [18]. We use a power-law species-area relationship to capture the

90 CHAPTER 2 dependence of long-term species richness on the remaining area of natural habitat [60, 61, 62, 63]. Since A(H, q) ∈ [0; 1], we allow the long-term species richness to vary between 1 (no habitat conversion) and 0 (all habitat is converted) by writing S(H, q) = (1−A(H, q))z, where the slope z ∈ [0; 1] ensures that the function is concave-down [48]. Following experimental and theoretical results [64, 20, 65, 18], we then assume that the rate of community relaxation is proportional to the diﬀerence between current biodiversity B and long-term species richness S(H, q).

B˙ = − [B − S(H, q)] (4) where mesures the relaxation rate of the community of species. The inverse of the relaxation coeﬃcient measures the time it takes to lose approximately 63% of the species that are doomed to extinction [18].

1.6 Social dynamics

Let us assume that the human population has identified the sustainability-optimal agricultural labor elasticity, αs (Fig.2.1.b and c). Restricting one’s consumption to sustainable agricultural goods has become a social norm, i.e. a shared rule of behavior. Recent studies demonstrate the importance of social norms on eating behaviors [66] and their role in shifting preferences towards healthy food [67, 68, 69]. The importance of dietary social norms is especially important in young adults [70], whose eating patterns typically become life-long habits [71]. Perception of others’ pro-environmental behavior was identified as the first step towards environmentally-friendly behavioral change [72].

Deviance from a social norm can lead to direct or indirect sanctioning from other members of the SES, be they important others or strangers [73]. Ostracism can result in social exclusion or poor reputation [74], thus decreasing the well-being of individuals. As a consequence, social pressure can reduce the well-being of defectors to the point where it becomes more proﬁtable for them to shift behavior in order to conform to the sustainable norm. A common way to approximate the well-being of consumers is through

91 CHAPTER 2 their consumption utility, which is a function of their per capita agricultural and industrial consumptions, and thus varies with the dynamical variables of the system. Let us denote

η 1−η the utility of a consumer of type i (i = {u, s}) as Ui = y1i y2 , where η is the preference for agricultural goods. Under our assumption that αs > αu, it can be shown that the consumption utility of defectors in the absence of social pressure, Uu, is always higher than the consumption utility of conformers, Us. Therefore, in the absence of social pressure, defectors have no incentive to shift their habits.

Following previous studies [13, 14], we assume that social pressure decreases the utility of defectors, Ud = Uu − w(q) · δU , so that it may become more proﬁtable for defectors to shift their consumption and comply to the sustainable norm. The severity of the os-

t·er·q tracism function, w(q) = wmaxe , increases with the proportion of conformers in the population, q, and depends on the maximum sanctioning wmax, the sanctioning eﬀective- ness threshold t, and the growth rate of the function, r. In addition to depending on the number of conformers in the community, graduated sanctioning and equity considerations leads conformers to act more strongly against defectors which consumption is the most unsustainable [3]. Thus, the lower αu and the larger the diﬀerence in consumption utilities between conformers and defectors, δU = (Uu − Us)/Uu, the stronger the social pressure.

The proportion of conformers then follows a replicator dynamics [13, 14], i.e. varies both with the proportion of conformers q, and the diﬀerence between the sustainable consumption utility, Us, and the average consumption utility, U = q · Us + (1 − q) · Ud, itself varying with the other dynamical variables of the system through the consumption of agricultural and industrial goods, y1i and y2.

˙q= q · [Us − U] = q · (1 − q) · (Uu − Us) · [w(q)/Uu − 1] (5)

Since Uu > Us in our model, a global dietary shift towards sustainable consumption ( ˙q > 0) is only possible if the severity of the social pressure is higher than the utility of defectors in the absence of social pressure, i.e. w(q) > Uu (eq.(5)).

In the following, our focus is on the potential of consumers’ behavioral change in preventing unsustainable trajectories, i.e. overshoot-and-collapse population crises leading

92 CHAPTER 2 to biodiversity-poor equilibria in the long run [37]. We ﬁrst analyze the dynamical system of equations (4), (2), (3) and (5), with a negligible ecological relaxation rate ( = 0.1). The consequences of lag eﬀects are explored in section 2.4.

2 Results

2.1 Social-ecological equilibria

∗ ∗ ∗ Our SES can have two types of equilibria (H ,B ,Tm, q ), hereafter denoted as viable (H∗ > 0 and B∗ < 1) or unviable (H∗ = 0 and B∗ = 1), when the economic parameters do not allow the human population to maintain itself in the environment [37]. Let

∗ ∗ ∗ us denote the viable equilibria as (Hi ,Bi ,Tm, qi ), with i = {u, s, c}. Among the viable

∗ equilibria, one is unsustainable (i = u), i.e. only defectors persist (qu = 0) and the transient dynamics includes overshoot-and-collapse population cycles under large extinction debts (Fig. 2.4.c). The other two types of viable equilibria are either fully sustainable

∗ (i = s) when only conformers persist (qs = 1), or partially sustainable (i = c) when

∗ both conformers and defectors coexist (qc ∈]0; 1[). The coexistence equilibrium satisﬁes

∗ ∗ ∗ w(qc ) = Uu (Bc ), for which there is no analytical solution. A general analytical solution for the unsustainable and fully sustainable equilibria is given in eq.(6), where the population density φi and γ1i (i = {u, s}) are explicitly deﬁned as functions of the parameters of the SES in Table S1 (electronic supplementary material).

1 min ! Ω 1 ∗ y1 ∗ ∗ z Bi = Hi = φi 1 − Bi (6) γ1i

Under our assumption that αs > αu, it can be shown that γ1s < γ1u, so that biodiversity at the sustainable equilibrium is higher than that at the unsustainable equilibrium, i.e

∗ ∗ Bs > Bu. However, population density is also higher at the sustainable equilibrium, i.e.

φs > φu, so that the human population size at equilibrium does not necessarily decrease with the proportion of conformers. Compliance to the sustainable consumption norm thus helps preserving biodiversity while not necessarily reducing the size of the human

93 CHAPTER 2 population.

2.2 Alternative stable states

A stability analysis of our SES model shows that two of the viable equilibria can be both stable at the same time, depending on the severity of the ostracism function compared to the consumption utility at equilibrium (electronic supplementary material, section 3). The per capita consumption utilities at the sustainable, unsustainable and coexistence

∗ min η 1−η equilibria are equal to U = (y1 ) γ2 . Compliance to the sustainable consumption norm thus does not reduce the long-term consumption utility.

The sustainable equilibrium is stable if the maximum ostracism w(1) is higher than the consumption utility that the defectors would have at the sustainable equilibrium,

∗ ∗ η ∗ i.e. w(1) > Uu(Bs ), where Uu(Bs ) = (γ1u/γ1s) U . Conversely, the unsustainable equilibrium is stable if the minimum ostracism w(0) is lower than the consumption utility

∗ ∗ ∗ at the unsustainable equilibrium, i.e. w(0) < Uu(Bu) where Uu(Bu) = U (Fig.2.2.a).

∗ ∗ Therefore, for intermediate consumption utilities, w(0) < Uu(Bu) < Uu(Bs ) < w(1), both the unsustainable and sustainable equilibria are stable ((U/S) region in Fig.2.2.b).

∗ ∗ For high consumption utilities, w(0) < U and w(1) < Uu(Bs ), ostracism is too weak to allow norm-driven behavioral change, and the unsustainable equilibrium is the only stable equilibrium that the SES can reach ((U) region in Fig.2.2.b), or both the unsustainable and mixed equilibria are stable ((U/M) region in Fig.2.2.b). Since there is no analytical expression for the mixed equilibrium, we are not able to derive any stability condition for this bistability region. However, Fig.2.2.b shows that the shift between the two bistable regions (U/M) and (U/S) depends on the footprint αu of unsustainable consumption.

The larger the footprint of defectors compared to conformers (αu << αs), the larger the bistability region (U/M) between the mixed and unsustainable equilibria and the smaller the bistability region (U/S). Thus, the larger the required behavioral change to shift from unsustainable habits (αu) towards sustainable habits (αs), the more diﬃcult it is to reach sustainability.

94 CHAPTER 2

(a) (b) 0.9 0.76 (U/S) 0.8 0.74

U (B*) 0.7 u s 0.72 (U/S) 0.6 0.7

U (B* ) 0.5 u u (U/M) 0.68 max

0.4 w 0.66 (U) Ostracism w(q) 0.3 0.64 (U/M)

0.2 0.62

0.1 0.6 (U)

0 0.58 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 0.06 0.08 0.1 0.12 0.14 α0.16 0.18 0.2 0.22 0.24 0.26 Conformers (q) u

Figure 2.2: Combined eﬀect of the ostracism parameters and the diﬀerence between sustainable and unsustainable norms on the stability of the equilibria. (a) Shape of the ostracism function for varying maximum ostracism wmax. (U) Weak sanctioning (wmax = 0.4) and stability of the unsustainable equilibrium only, such ∗ ∗ that w(0) < w(1) < Uu(Bu) < Uu(Bs ); (U/M) intermediate sanctioning (wmax = 0.64) and bistability of the unsustainable and the mixed equilibria; (U/S) strong sanctioning (wmax = 0.8) and bistability of the unsustainable and the sustainable equilibria, such ∗ ∗ that w(1) > Uu(Bs ) and w(0) > Uu(Bu). See table S1 in the electronic supplementary material for other parameter values. (b) Stable equilibria with varying maximum ostracism wmax and unsustainable labor elasticity αu. Regions (U), (U/M) and (U/S) correspond to the red, blue and black curves in (a), respectively.

2.3 Impact of the initial state of the SES

Depending on the parameters of the SES, the size of the human population at the sustainable equilibrium can be either higher (e.g. for Tm = 2) or lower (e.g. for Tm = 1.8) than at the unsustainable equilibrium. Let us now consider a situation where the human population size at the sustainable equilibrium is lower than that at the unsustainable equilibrium.

Fig.2.3 shows that, when there is bistability ((U/M) and (U/S) panels), the sustainable and mixed equilibria are only reached in the long run when the initial proportion of conformers is high enough. The stronger the ostracism, the lower the minimum proportion of conformers required for sustainability, i.e. the larger the sustainable basin of attraction. Gradually changing social parameters may thus push an initially unsustainable SES ((U) panel in Fig.2.3) towards a sustainable path ((U/S) panel in Fig.2.3), provided that the initial social capital is large enough.

95 CHAPTER 2

(U) (U/M) (U/S) 4 4 4 (0,H* ) u 3 3 (q*,H*) 3 * c c (0,H ) (0,H* ) (1,H*) u u s 2 2 2

1 1 1 Human pop. (H)

0 0 0 0 0.5 1 0 0.5 1 0 0.5 1 Conformers (q) Conformers (q) Conformers (q)

Figure 2.3: Eﬀect of the initial conditions on the long-term equilibria, for various ostracism strengths wmax. Initial proportions of conformers are q(0) = 0.2 and q(0) = 0.6. Cases (U), (U/M) and (U/S) correspond to the ostracism functions deﬁned in Fig. 2.2.a. Black curve: isocline H˙ = 0; dashed curve: isocline ˙q= 0; white dots: ∗ ∗ unstable equilibria; green dot: (stable) sustainable equilibrium, (Hs ,Bs , 1); red dot: (sta- ∗ ∗ ble) unsustainable equilibrium, (Hu,Bu, 0); blue dot: (stable) coexistence equilibrium, ∗ ∗ ∗ (Hc ,Bc , qc ); transient trajectories are represented by the blue curves. See table S1 in the electronic supplementary material for parameter values.

Under conditions of bistability, the type of equilibrium that will be reached in the long run thus depends on the rate of social change. In the following, we show that the rate of social change also depends on human perception of environmental changes and, in our case, extinction debts.

2.4 Impact of extinction debts on the eﬀectiveness of ostracism

We now explore the transient behavior of the SES with varying ecological relaxation rates, , for two of the initial conditions used in Fig.2.3, corresponding to two initial proportions of conformers q(0) = 0.2 and q(0) = 0.6, with the same human population size H(0) = 0.5. In order to better visualize transient environmental crises, we plot the null-clines and transient trajectories in the human-biodiversity phase plane (Fig.2.4.a). Ecological relaxation rates slow down the social dynamics by postponing the utility reduction of defectors, Uu, and therefore, their consumption shift towards sustainable habits (eq.(5)). When the extinction debt is moderate, transient dynamics towards the unsustainable and mixed equilibria show environmental crises, the amplitude of which is lower for the mixed equilibrium (Fig.2.4.b). The sustainable trajectories do not experience any overshoot-

96 CHAPTER 2 and-collapse behavior, even for high extinction debts ((U/S) panel in Fig.2.4.c), which confirms the relevance of our sustainability criterion. A high extinction debt leads to very large environmental crises over the unsustainable trajectories (Fig.2.4.c). Moreover, in the case of bistability between the unsustainable and mixed equilibria, all trajectories now reach the unsustainable equilibrium ((U/M) panel in Fig.2.4.c). Large ecological relaxation rates thus result in the loss of stability of the mixed equilibrium in favor of the unsustainable equilibrium. This result suggests a shift in the dominant social- ecological feedback for increasing relaxation rates. At low relaxation rates, the ecological dynamics is fast enough for the negative effect of environmental degradation on human well-being to result in a fast enough social changes, thus reinforcing sustainable feedbacks through an efficient social ostracism. However, large extinction debts slow down the ecological dynamics and postpone the negative ecological feedback on human well-being. This reduces the efficiency of social ostracism and results in a shift of the dominant feedback towards unsustainable feedbacks, i.e. increasing consumptions and decreasing labor intensities.

Fig.2.5 shows the combined impact of lag effects and social ostracism on the basins of attraction of the sustainable, mixed and unsustainable equilibria. Increasing both the initial proportion of conformers and the strength of the ostracism can push an initially unsustainable SES into the basin of attraction of the sustainable or mixed equilibria (Fig.2.5.a). However, decreasing the ecological relaxation rate reduces the basin of attraction of the mixed equilibrium in favor of the unsustainable equilibrium (Fig.2.5.b). The stability of the mixed equilibrium appears to be much more sensitive to ecological time-lags than that of the sustainable equilibrium. Thus, moderate behavioral changes leading to a mixed equilibrium may not be robust enough to ecological lag effects. These results suggest that only important behavioral changes allowing to reach the fully sustainable equilibrium may be able to counteract the destabilizing effect of ecological time lags. The extinction debt, by postponing the consequences of environmental degradation on human well-being, thus reduces the robustness of social change and norm-driven sustainability.

97 CHAPTER 2

(a) (U) (U/M) (U/S) 7 7 7

6 6 6

5 5 5

4 4 4 * * * * 3 3 (B ,H ) 3 (B ,H ) c c s s * * * * * * 2 (B ,H ) 2 (B ,H ) 2 (B ,H ) u u u u u u Human pop. (H) 1 1 1

0 0 0 0 0.5 1 0 0.5 1 0 0.5 1 (b) 7 7 7

6 6 6

5 5 5

4 4 * * 4 (B* ,H* ) (B ,H ) (B* ,H* ) u u u u u u 3 3 3

2 2 2 * * Human pop. (H) (B ,H ) (B*,H*) 1 1 c c 1 s s

0 0 0 0 0.5 1 0 0.5 1 0 0.5 1

6 6 6

5 5 5 * * * * (B ,H ) (B* ,H* ) (B ,H ) 4 u u 4 u u 4 u u 3 3 3

2 2 2

Human pop. (H) (B*,H*) 1 1 1 s s

0 0 0 0 0.5 1 0 0.5 1 0 0.5 1 Biodiversity (B) Biodiversity (B) Biodiversity (B)

Figure 2.4: Eﬀect of varying extinction debts and ostracism strengths wmax on transient dynamics and stability, for two initial proportions of conformers q(0). Cases (U), (U/M) and (U/S) correspond to the ostracism functions deﬁned in Fig. 2.2.a, with similar initial conditions, i.e. H(0) = 0.5, B(0) = (1 − H(0)/φ(q))z and q(0) = 0.2 or q(0) = 0.6. (a) low extinction debt ( = 0.1); (b) intermediate extinction debt ( = 0.0025); (c) large extinction debt ( = 0.0005); green dot: (stable) sustainable equilibrium; red dot: (stable) unsustainable equilibrium; blue dot: (stable) mixed equilibrium; transient trajectories are represented by the blue curves. See table S1 in the electronic supplementary material for other parameter values.

98 CHAPTER 2

(a) (b) 0.7 0.7 1

0.69 0.69 0.9

0.68 0.68 0.8 max 0.67 0.67 0.7

0.66 0.66 0.6

0.65 0.65 0.5

0.64 0.64 0.4

0.63 0.63 0.3

Maximum ostracism w 0.62 0.62 0.2

0.61 0.61 0.1

0.6 0.6 0 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 0 0.2 0.4 0.6 0.8 Initial prop. conformers q(0) Initial prop. conformers q(0)

Figure 2.5: Proportion of conformers at equilibrium (q∗) under the combined eﬀects of the initial proportion of conformers q(0) and maximum ostracism wmax, for an increasing extinction debt. Other initial conditions: H(0) = 0.5, and B(0) = (1 − H(0)/φ(q(0)))z. See table S1 in the electronic supplementary material for other parameter values. (a) low extinction debt ( = 0.2); (b) high extinction debt ( = 0.0005). Blue color represents a population of defectors (q∗ = 0), red color represents a population of conformers (q∗ = 1) and intermediate colors represent a coexistence of conformers and defectors (q∗ ∈ [0, 1]).

3 Discussion and conclusions

We investigate the robustness of norm-driven sustainability, as measured by a shift towards low-footprint consumption habits. Speciﬁcally, we focus on the robustness of SESs to time- delayed biodiversity losses caused by human-driven natural habitat destruction. Time- delayed ecological feedbacks are known to reinforce negative management feedbacks and potentially push SESs into social-ecological traps [15]. However, little research so far has investigated the long-term impacts of extinction debts on the sustainability of coupled SESs. Ecological studies of the anthropogenic impacts on resources or ecosystems often neglect changes in the size and behavior of the human population. Additionally, natural resources are often managed as decoupled from the ecosystems they are part of, and most socio-economic studies overlook the ﬁniteness and physical limits of natural systems. Modeling sustainability requires accounting for the bidirectional coupling between human and natural systems [75], and especially the feedback loop between human population growth and environmental degradation [37].

99 CHAPTER 2

In our model, this feedback loop is mediated through biodiversity-dependent ecosystem services to agricultural production. A human population exploits a shared land resource divided into natural habitat and converted agricultural and industrial lands. Natural habitat supports a community of species and provides a range of biodiversity-dependent regulatory services to agricultural production, which can itself be seen as a provisioning service. A norm of sustainable consumption is maintained through social sanctioning of unsustainable consumers. Increasing demand for sustainable consumption translates into more sustainable agricultural practices, which involve the use of a larger proportion of labor compared to land. Finally, human population growth is driven by the interaction between available agricultural resources, technological changes and social changes, thus adding to the growing literature modeling the interaction between human populations and their environments [38, 76].

Our approach thus diﬀers from the classical economic literature related to the internalization of intertemporal externalities of agricultural production [77, 6]. Though we also consider the pressure of a group over another as a driver of sustainable change, the penal- ties involved are not chosen optimally. This would require that each of the conformers had access to perfect information regarding the social damage associated with the unsustainable consumption of defectors, as well as the optimal social pressure to apply. The point of this article was not to derive an optimal regulation strategy, but to illustrate the potential of bottom-up consumption changes in driving sustainable shifts in production practices, given the long-term ecological dynamics of the SES.

The sustainable consumption norm is identiﬁed following Lafuite and Loreau’s (2017) sustainability criterion, which characterizes the vulnerability of an SES to transient “overshoot- and-collapse” population crises. This criterion captures the diﬀerence between the rates of ecological relaxation and human population growth, so that sustainable SESs have high enough ecological relaxation rates compared to the growth rate of their human populations. We verify here the validity of this sustainability criterion, showing that a shift towards more environmentally-friendly agricultural practices, i.e. characterized by a lower substitution of ecosystem services and labor for technology, decreases the vulner-

100 CHAPTER 2 ability of SESs to transient crises. Such a global shift towards sustainable agricultural practices would require reversing current trends of land-intensive and highly mechanized agricultural production towards more labor-intensive productions, e.g. small-scale agroecological farms. Growing evidence suggests that diverse small-scale agro-ecological farms increase carbon sequestration, support biodiversity, rebuild soil fertility and sustain yields over time, thus securing farm livelihoods, while competing with industrial agriculture in terms of total outputs, especially under environmental stress [78].

Under a negligible ecological time delay between natural habitat loss and biodiversity erosion, full sustainability is ensured when both social sanctioning and the proportion of conformers are large enough, and when the required behavioral change to shift from unsustainable to sustainable habits is not too large. Otherwise, a minority of defectors coexists with a majority of conformers at the mixed equilibrium. When social sanctioning and/or the proportion of conformers is too low, only defectors persist at equilibrium. This unsustainable equilibrium is always stable, so that there is bistability between the unsustainable and sustainable or mixed equilibria, when these are stable. These findings echo those of Tavoni et al. [13], who used a similar non-costly social sanctioning to study cooperation in the management of a single natural resource under variable environmental conditions. However, time delays have an opposite effect to resource variability, since temporal variability tends to decrease the mean resource level, thus increasing the probability of a behavioral shift towards norm compliance. Our model differs from Tavoni et al. [13] in many aspects; first, here we focus on the interaction between various ecosystem services, especially provisioning and regulatory services, instead of a single natural resource; second, these services feed back on the dynamics of the human population that uses these services, so that the human population varies endogenously with the state of the environment; lastly, social sanctioning affects consumers’ behavior, instead of pro- ducers’. The latter feature allows us to focus on the potential of consumers’ behavioral changes in establishing sustainability in coupled SESs. We could also have a modeled a reciprocal pressure between conformers and defectors, in which case our result would also depend on the relative strength of these pressures.

101 CHAPTER 2

Our study provides insights into the consequences of lag eﬀects for norm-driven sustainability. Biodiversity loss acts as a negative feedback on human well-being, through the loss of biodiversity-dependent regulatory services to agricultural production. A time- delayed biodiversity feedback thus maintains a high utility of defectors for a longer period of time. This time lag decreases the eﬃciency of social ostracism, thus delaying behavioral shift. Postponing the behavioral shift of defectors towards sustainable consumption for too long can make the mixed equilibrium totally unreachable, meaning that a tipping point has been crossed in terms of human population size and habitat destruction. The weaker stability of the mixed equilibrium and its propensity to regime shifts was already observed by Lade et al. [79]. Thus, under large time delays, the only way to reach sustainability is to reach the full-sustainability equilibrium, which requires much larger behavioral changes. However, given the widely observed coexistence of both conformers and defectors in small groups [80], such behavioral changes seem rather unlikely.

Moreover, theory suggests that relaxation rates are not constant, but increase with the extent of habitat destruction and fragmentation [58], thus further delaying the feedback of biodiversity-dependent ecosystem services on human societies [24]. In situations where habitat destruction leads to a strong increase in ecological relaxation rates, we would expect a decrease in sustainability, or a shift towards unsustainable development paths. An interesting extension to our work would thus be to use a spatially-explicit ecological model, in order to gain more realism regarding the temporal dynamics of ecological relaxation rates under habitat destruction, and study social-ecological regime shifts from a spatial perspective.

The emergence of tipping points and regime shifts in coupled SESs [79] is gaining increasing interest [81], with many implications for the adaptive management of SESs [82]. Regime shifts can lead to social-ecological traps, where unsustainable feedbacks reinforce each others and push the SES into an undesirable state [11]. Some authors suggest that humanity may be locked in a technological innovation pathway that reinforces such unsustainable feedbacks [42]. Time-delayed ecological feedbacks may also aﬀect the human perception of environmental changes, thus worsening the amnesia and shortsightedness

102 CHAPTER 2 observed in conservation science, known as the shifting baseline syndrome [83]. This syndrome refers to a shift over time in the expectation of what a healthy biodiversity baseline is, and can lead to tolerate incremental loss of species through inappropriate management [84]. Time delays can also be related to perceived environmental uncertainty, which has been shown to endanger the establishment of cooperation in SESs with common- pool resources [85]. Our results highlight the importance of accounting for the feedback loop between human demography, environmental degradation and behavioral changes when studying the long-term sustainability of coupled SESs. Especially, the temporal dynamics of coupled social-ecological processes matter, since time-delayed ecological feedbacks alter the human perception of environmental degradation and the pace of behavioral changes. Policies that enhance the adaptive capacity of social-ecological systems may thus beneﬁt from taking both social norms and time delays into account [10]. These insights also point to future research needs regarding the interplay of social, demographic and ecological long-term dynamics.

Authors’ contributions

A.-S.L., C. de M. and M.L. jointly designed the study, developed the model and analyzed and interpreted the model results. A.-S.L. drafted the manuscript; M.L. and C. de M. revised it critically. All authors gave ﬁnal approval for publication.

Acknowledgements

This work was supported by the TULIP Laboratory of Excellence (ANR-10-LABX-41) and the Midi-Pyrénées Region. We thank Matthieu Barbier, Kirsten Henderson and David Shanafelt for valuable discussions and helpful comments on earlier versions of the manuscript.

103 CHAPTER 2

Bibliography

1. G Hardin. The tragedy of the commons. Science, 162:1243–1248, 1968.

2. E Ostrom. Governing the commons: the evolution of institutions for collective action. New York, NY: Cambridge University Press, 1990.

3. E Ostrom. Collective Action and the Evolution of Social Norms. Journal of Economic Perspectives, 14(3):137–158, 2000.

4. K Nyborg and M Rege. On social norms: the evolution of considerate smoking behavior. Journal of Economic Behavior and Organization, 52:323–340, 2003. doi: 10.1016/S0167- 2681(03)00031-3.

5. W.D. Nordhaus. Optimal Greenhouse-gas Reductions and Tax Policy in the "DICE" model. The American Economic Review, 83(2):313–317, 1993.

6. J Pretty, C Brett, D Gee, R Hine, C Mason, J Morison, M Rayment, G Van Der Bijl, and T Dobbs. Policy challenges and priorities for internalizing the externalities of modern agriculture. Journal of Environmental Planning and Management, 44(2):263–283, 2001.

7. J.M. Anderies. On modelling human behavior and institutions in simple ecological economic systems. Ecological Economics, 35:393–412, 2000.

8. C Folke, J Colding, and F Berkes. Navigating social-ecological systems: building resilience for complexity and change., chapter Synthesis: building resilience and adaptive capacity in social–ecological systems. Cambridge University Press, New York, 2003.

9. B. Walker, C.S. Holling, S.R. Carpenter, and A Kinzig. Resilience, adaptability and trans- formability in social–ecological systems. Ecology and Society, 9(2):5, 2004.

10. K Nyborg, J.M. Anderies, A Dannenberg, T Lindahl, C Schill, M Schlüter, W.N. Adger, K.J. Arrow, S. Barrett, S. Carpenter, F.S. Chapin III, A.-S. Crépin, G. Daily, P. Ehrlich, C. Folke, W. Jager, N Kautsky, S.A. Levin, O.J. Madsen, S. Polasky, M. Scheﬀer, B. Walker, E.U. Weber, J. Wilen, A. Xepapadeas, and A de Zeeuw. Social norms as solutions. Science, 354(6308):42–43, 2016. doi: 10.1126/science.aaf8317.

11. W.J. Boonstra, E Bjorkvik, and L.J. Haider. Human responses to social-ecological traps. Sustainability Science, 11(6):877–889, 2016. doi: 10.1007/s11625-016-0397-x.

12. E Ostrom. A general framework for analyzing sustainability of social-ecological systems. Science, 325:419–422, 2009. doi: 10.1126/science.1172133.

13. A Tavoni, M Schlüter, and S Levin. The survival of the conformist: Social pressure and renewable resource management. Journal of Theoretical Biology, 299:152–161, 2012.

14. M Schlüter, A Tavoni, and S Levin. Robustness of norm-driven cooperation in the commons. Proceedings of the Royal Society B: : Biological Sciences, 283:20152431, 2016.

15. A.-S. Crepin. Using fast and slow processes to manage resources with thresholds. Environ- mental and Resource Economics, 36:191–213, 2007.

16. J.E. Cinner. Social-ecological traps in reef ﬁsheries. Global Environmental Change, 21(3): 835–839, 2011.

104 CHAPTER 2

17. David Tilman, Robert M. May, Clarence L. Lehman, and Martin A. Nowak. Habitat destruction and the extinction debt. Nature, 371(6492):65–66, 1994.

18. J.M. Diamond. Biogeographic kinetics: estimation of relaxation times for avifaunas of southwest paciﬁc islands. Proc.Nat.Acad.Sci. USA, 69(11):3199–3203, 1972.

19. N. M. Haddad, L. a. Brudvig, J. Clobert, K. F. Davies, a. Gonzalez, R. D. Holt, T. E. Lovejoy, J. O. Sexton, M. P. Austin, C. D. Collins, W. M. Cook, E. I. Damschen, R. M. Ewers, B. L. Foster, C. N. Jenkins, a. J. King, W. F. Laurance, D. J. Levey, C. R. Mar- gules, B. a. Melbourne, a. O. Nicholls, J. L. Orrock, D.-X. Song, and J. R. Townshend. Habitat fragmentation and its lasting impact on Earth’s ecosystems. Science Advances, 1 (2):e1500052–e1500052, 2015. ISSN 2375-2548. doi: 10.1126/sciadv.1500052.

20. O. R. Wearn, D. C. Reuman, and R. M. Ewers. Extinction Debt and Windows of Conser- vation Opportunity in the Brazilian Amazon. Science, 337(6091):228–232, 2012.

21. Isabel M.D. Rosa, M.J. Smith, O.R. Wearn, D Purves, and R.M. Ewers. The environmental legacy of modern tropical deforestation. Current Biology, 26:2161–2166, 2016. doi: 10.1016/j.cub.2016.06.013.

22. J Krauss, R Bommarco, M Guardiola, R.S. Heikkinen, A Helm, M Kuussaari, R Lindborg, E Öckinger, M Pärtel, J Pino, J Pöyry, K.M. Raatikainen, A Sang, C Stefanescu, T Teder, M Zobel, and I Steﬀan-Dewenter. Habitat fragmentation causes immediate and time-delayed biodiversity loss at diﬀerent trophic levels. Ecology Letters, 13(5):597–605, 2010.

23. S. Dullinger, F Essl, W Rabitsch, K.-H. Erb, S Gingrich, H Haberl, K Hülber, V Jarosik, F Krausmann, I Kühn, J Pergl, P Pysek, and P.E. Hulme. Europe’s other debt crisis caused by the long legacy of future extinctions. Proceedings of the National Academy of Sciences, 110(18):7342–7347, 2013.

24. Forest Isbell, David Tilman, Stephen Polasky, and Michel Loreau. The biodiversity- dependent ecosystem service debt. Ecology Letters, 18(2):119–134, 2015.

25. E Barrios. Soil biota, ecosystem services and land productivity. Ecological Economics, 64 (2):269–285, 2007.

26. Emily A Martin, Björn Reineking, Bumsuk Seo, and Ingolf Steﬀan-Dewenter. Natural enemy interactions constrain pest control in complex agricultural landscapes. Proceedings of the National Academy of Sciences, 110(14):5534–5539, 2013. doi: 10.1073/pnas.1215725110.

27. N Gallai, J Salles, J Settele, and B Vaissière. Economic valuation of the vulnerability of world agriculture confronted with pollinator decline. Ecological Economics, 68(3):810–821, 2009.

28. L Garibaldi, I Steffan-Dewenter, C Kremen, J Morales, R Bommarco, S Cunningham, L Car- valheiro, N Chacoff, J Dudenhffer, S Greenleaf, A Holzschuh, R Isaacs, K Krewenka, Y Man- delik, M Mayfield, L Morandin, S Potts, T Ricketts, H Szentgyrgyi, B Viana, C Westphal, R Winfree, and A Klein. Stability of pollination services decreases with isolation from natural areas despite honey bee visits. Ecology Letters, 14(10):1062–1072, 2011.

29. G Ceballos, P.R. Ehrlich, A.D. Barnosky, A Garcia, R.M. Pringle, and T.M. Palmer. Accel- erated modern human–induced species losses: Entering the sixth mass extinction. Science Advances, 1(5):e1400253, 2015. doi: 10.1126/sciadv.1400253.

105 CHAPTER 2

30. Tim Newbold, Lawrence N. Hudson, Andrew P. Arnell, Sara Contu, Adriana De Palma, Simon Ferrier, Samantha L. L. Hill, Andrew J. Hoskins, Igor Lysenko, Helen R. P. Phillips, Victoria J. Burton, Charlotte W. T. Chng, Susan Emerson, Di Gao, Gwilym Pask-Hale, Jon Hutton, Martin Jung, Katia Sanchez-Ortiz, Benno I. Simmons, Sarah Whitmee, Han- bin Zhang, Jörn P. W. Scharlemann, and Andy Purvis. Has land use pushed terrestrial biodiversity beyond the planetary boundary? A global assessment. Science, 353:288–291, 2016.

31. J Barlow, G.D Lennox, J Ferreira, E Berenguer, A. C. Lees, Ralph Mac Nally, James R. Thomson, Silvio Frosini de Barros Ferraz, Julio Louzada, Victor Hugo Fonseca Oliveira, Luke Parry, Ricardo Ribeiro de Castro Solar, Ima C. G. Vieira, Luiz E. O. C. Aragão, Rodrigo Anzolin Begotti, Rodrigo F. Braga, Thiago Moreira Cardoso, Raimundo Cosme de Oliveira Jr, Carlos M. Souza Jr, Nárgila G. Moura, Sâmia Serra Nunes, João Victor Siqueira, Renata Pardini, Juliana M. Silveira, Fernando Z. Vaz-de Mello, Ruan Carlo Stulpen Veiga, Adriano Venturieri, and Toby A. Gardner. Anthropogenic disturbance in tropical forests can double biodiversity loss from deforestation. Nature, 535:144–147, 2016. doi: 10.1038/nature18326.

32. Henrique M Pereira, Paul W Leadley, Vânia Proença, Rob Alkemade, Jörn P W Scharle- mann, Juan F Fernandez-Manjarrés, Miguel B Araújo, Patricia Balvanera, Reinette Biggs, William W L Cheung, Louise Chini, H David Cooper, Eric L Gilman, Sylvie Guénette, George C Hurtt, Henry P Huntington, Georgina M Mace, Thierry Oberdorﬀ, Carmen Re- venga, Patrícia Rodrigues, Robert J Scholes, Ussif Rashid Sumaila, and Matt Walpole. Scenarios for global biodiversity in the 21st century. Science, 330(6010):1496–1501, 2010.

33. P R Ehrlich and a H Ehrlich. Can a collapse of global civilization be avoided? Proceedings of the Royal Society of London B: Biological Sciences, 280(1754):20122845, 2013.

34. F.S.III Chapin, E.S. Zavaleta, V.T. Eviner, R.L. Naylor, P.M. Vitousek, H.L. Reynolds, D.U. Hooper, S. Lavorel, O.E. Sala, S.E. Hobbie, M.C. Mack, and S. Díaz. Consequences of changing biodiversity. Nature, 405:234–242, 2000.

35. Bradley J Cardinale, J Emmett Duﬀy, Andrew Gonzalez, David U Hooper, Charles Perrings, Patrick Venail, Anita Narwani, Georgina M Mace, David Tilman, David A Wardle, Ann P Kinzig, Gretchen C Daily, Michel Loreau, James B Grace, Anne Larigauderie, Diane S Srivastava, and Shahid Naeem. Biodiversity loss and its impact on humanity. Nature, 486 (7401):59–67, 2012. ISSN 00280836. doi: 10.1038/nature11148.

36. Michel Loreau and Claire de Mazancourt. Biodiversity and ecosystem stability: A synthesis of underlying mechanisms. Ecology Letters, 16(s1):106–115, 2013.

37. A.-S. Lafuite and M Loreau. Time-delayed biodiversity feedbacks and the sustainability of social-ecological systems. Ecological Modelling, 351:96–108, 2017. doi: 10.1016/j.ecolmodel.2017.02.022.

38. James A. Brander and M. Scott Taylor. The Simple Economics of Easter Island: A Ricardo- Malthus Model of Renewable Resource Use. American Economic Review, 88(1):119–138, 1998.

39. O Venter, E.W. Sanderson, A. Magrach, J.R. Allan, J. Beher, K.R. Jones, H.P. Possingham, W.F. Laurance, P. Wood, B.M. Fekete, M.A. Levy, and J.E.M Watson. Sixteen years of change in the global terrestrial human footprint and implications for biodiversity conservation. Nature Communications, 7(12558), 2016.

106 CHAPTER 2

40. Eric F. Lambin and P. Meyfroidt. Global land use change, economic globalization, and the looming land scarcity. Proceedings of the National Academy of Sciences of the United States of America, 108(9):3465–72, 2011. ISSN 10916490. doi: 10.1073/pnas.1100480108.

41. Robert M. Ewers, Jörn P W Scharlemann, Andrew Balmford, and Rhys E. Green. Do increases in agricultural yield spare land for nature? Global Change Biology, 15(7):1716– 1726, 2009.

42. F. Westley, P Olsson, C Folke, T Homer-Dixon, H Vredenburg, D Loorbach, J Thompson, M Nilsson, E Lambin, J Sendzimir, B Banarjee, V Galaz, and S van der Leeuw. Tipping towards sustainability: emerging pathways of transformation. Ambio, 40:762–780, 2011.

43. Jonathan A. Foley, Navin Ramankutty, Kate A. Brauman, Emily S. Cassidy, James S. Ger- ber, Matt Johnston, Nathaniel D. Mueller, Christine O/’Connell, Deepak K. Ray, Paul C. West, Christian Balzer, Elena M. Bennett, Stephen R. Carpenter, Jason Hill, Chad Mon- freda, Stephen Polasky, Johan Rockstrom, John Sheehan, Stefan Siebert, David Tilman, and David P. M. Zaks. Solutions for a cultivated planet. Nature, 478:337–342, 2011.

44. D. Tilman and M Clark. Global diets link environmental sustainability and human health. Nature, 515:518–522, 2014. doi: 10.1038/nature13959.

45. P. L. Pingali. Green Revolution: Impacts, limits, and the path ahead. Proceedings of the National Academy of Sciences, 109(31):12302–12308, 2012.

46. G. Brundtland, M. Khalid, S. Agnelli, S. Al-Athel, B. Chidzero, L. Fadika, V. Hauﬀ, I. Lang, M. Shijun, M. Morino de Botero, M. Singh, S. Okita, et al. Our Common Future (The Brundtland Report). Oxford University Press, 1987.

47. John M. Anderies. Economic development, demographics, and renewable resources: a dynamical systems approach. Environment and Development Economics, 8(02):219–246, 2003.

48. A O’Connor, M.I.and Gonzalez, J.E.K Byrnes, B.J. Cardinale, J.E. Duﬀy, L. Gamfeldt, J.N. Griﬃn, D. Hooper, B.A. Hungate, A Paquette, P.L. Thompson, L.E. Dee, and K.L. Dolan. A general biodiversity-function relationship is mediated by trophic level. Oikos, 126 (1):18–31, 2017.

49. M Zeigler and A Steensland. Global Agricultural Productivity Report (GAP Report). Global Harvest Initiative, 2016.

50. D.R. Lighthall. Farm structure and chemical use in the corn belt. Rural sociology, 60(3): 505–520, 1995.

51. M Duﬀy. Economies of size in production agriculture. Journal of Hunger and Environmental Nutrition, 4(3-4):375–392, 2009.

52. M.J. Pfeﬀer. Labor and production barriers to the reduction of agricultural chemical inputs. Rural sociology, 57(3):347–362, 1992.

53. T Garnett, M.C. Appleby, A Balmford, I.J. Bateman, T.G. Benton, P Bloomer, B Burlingame, M Dawkins, L Dolan, D Fraser, M Herrero, I Hoﬀmann, P Smith, P.K. Khornton, C Toulmin, S.J. Vermeulen, and H.C.J. Godfray. Sustainable intensiﬁcation in agriculture: Premises and policies. Science, 341:33–34, 2013.

54. O. Galor and D.N. Weil. Population, Technology, and Growth: From Malthusian Stagnation to the Demographic Transition and beyond. The American Economic Review, 90(4):806– 828, 2000.

107 CHAPTER 2

55. T Kogel and A Prskawetz. Agricultural Productivity Growth and Escape from the Malthu- sian Trap. Journal of Economic Growth, 6:337–357, 2001.

56. P.F. Peretto and S Valente. Growth on a ﬁnite planet: resources, technology and population in the long run. Journal of Economic Growth, 20(3):305–331, 2015.

57. A Fernihough. Human capital and the quantity-quality trade-oﬀ during the demographic transition. Journal of Economic Growth, 22(1):35–65, 2017.

58. Ilkka Hanski and Otso Ovaskainen. Extinction debt at extinction threshold. Conservation Biology, 16(3):666–673, 2002.

59. K Hylander and J Ehrlen. The mechanisms causing extinction debts. Trends in Ecology and Evolution, 28(6):341–346, 2013.

60. E.F. Connor and E.D. McCoy. The statistics and biology of the species-area relationship. American Naturalist, 113:119–130, 1979.

61. K.A. McGuiness. Equations and explanations in the study of species-area curves. Biological Reviews, 59:423–440, 1984.

62. David Storch, Petr Keil, and Walter Jetz. Universal species–area and endemics–area relationships at continental scales. Nature, 488(7409):78–81, 2012.

63. Joel Rybicki and Ilkka Hanski. Species-area relationships and extinctions caused by habitat loss and fragmentation. Ecology Letters, 16(SUPPL.1):27–38, 2013.

64. G. Ferraz, J.D. Nichols, J.E. Hines, P.C. Stouﬀer, Jr. Bierregaard, and T.E. Lovejoy. A large-scale deforestation experiment: Eﬀects of patch area and isolation on Amazon birds. Science, 315(5809):238–241, 2007.

65. R.H. Mac Arthur and E.O. Wilson. The theory of island biogeography. Monographs in Population Biology, 1967.

66. S Higgs and J Thomas. Social inﬂuences on eating. Current Opinion in Behavioral Sciences, 9:1–6, 2016. doi: 10.1016/j.cobeha.2015.10.005.

67. K Ball, R.W. Jeﬀery, G Abbott, S.A. McNaughton, and D Crawford. Is healthy behavior contagious: associations of social norms with physical activity and healthy eating. Interna- tional Journal of Behavioral Nutrition and Physical Activity, 7:86, 2010. doi: 10.1186/1479- 5868-7-86.

68. S. Mollen, R.N. Rimal, R.A.C. Ruiter, and G. Kok. Healthy and unhealthy social norms and food selection. ﬁndings from a ﬁeld-experiment. Appetite, 65:83–89, 2013. doi: 10.1016/j.appet.2013.01.020.

69. EM Templeton, MV Stanton, and J Zaki. Social norms shift preferences for healthy and unhealthy foods. PLoS ONE, 11(11):e0166286, 2016. doi: 10.1371/journal.pone.0166286.

70. J.E. Pelletier, D.J. Graham, and M.N. Laska. Social norms and dietary behaviors among young adults. American Journal of Health and Behaviour, 38(1):10.5993/AJHB.38.1.15, 2014. doi: 10.5993/AJHB.38.1.15.

71. Y Wang and T Lobstein. Worldwide trends in childhood overweight and obesity. Interna- tional Journal of Pediatric Obesity, 1(1):11–25, 2006.

108 CHAPTER 2

72. U Dahlstkand and A Biel. Pro-environmental habits: Propensity levels in behavioral change. Journal of Applied Social Psychology, 27(7):588–601, 1997.

73. L Balafoutas, N Nikiforakis, and B Rockenbach. Direct and indirect punishment among strangers in the ﬁeld. Proceedings of the National Academy of Sciences, 111(45):15924– 15927, 2014. doi: 10.1073/pnas.1413170111.

74. M. Milinski. Reputation, a universal currency for human social interactions. Philosophical transactions of the Royal Society of London B: Biological sciences, 371:20150100, 2016. doi: 10.1098/rstb.2015.0100.

75. S Motesharrei, J Rivas, E Kalnay, G.R. Asrar, A.J. Busalacchi, R.F. Cahalan, M.A. Cane, R.R. Colwell, K Feng, R.S. Franklin, K Hubacek, F Miralles-Wilhelm, T Miyoshi, M Ruth, R Sagdeev, A Shirmohammadi, J Shukla, J Srebric, V.M. Yakovenko, and N Zeng. Mod- eling sustainability: population, inequality, consumption, and bidirectional coupling of the earth and human systems. National Science Review, 3(4):470–494, 2016. doi: 10.1093/n- sr/nww081.

76. M. Scott Taylor. Innis lecture: Environmental crises: Past, present, and future. Canadian Journal of Economics, 42(4):1240–1275, 2009.

77. T.S.. Aidt. Political internalization of economic externalities and environmental policy. Journal of Public Economics, 69(1):1–16, 1998.

78. IPES-Food. From uniformity to diversity: a paradigm shift from industrial agriculture t diversiﬁed agroecological systems. Panel of Experts on Sustainable Food systems., 2016.

79. S.J. Lade, A. Tavoni, S.A. Levin, and M Schlüter. Regime shifts in a social-ecological system. Theoretical Ecology, 6(3):359–372, 2013. doi: 10.1007/s12080-013-0187-3.

80. E Fehr and U Fischbacher. Why social preferences matter: the impact of non-selﬁsh motives on competition, cooperation and incentives. Economic Journal, 112:C1–C33, 2002.

81. J.G. Polhill, T Filatova, M Schlüter, and A Voinov. Modelling systemic change in coupled socio-environmental systems. Environmental Modelling and Software, 75:318–332, 2016.

82. A.-S. Crepin, R Biggs, S Polasky, M Troell, and A de Zeeuw. Regime shifts and management. Ecological Economics, 84:15–22, 2012.

83. S.K. Papworth, J Rist, E.J. Coad, and E.J. Milner-Gulland. Evidence for shifting baseline syndrome in conservation. Conservation Letters, 2(2):93–100, 2009. doi: 10.1111/j.1755- 263X.2009.00049.x.

84. D Pauly. Anecdotes and the shifting baseline syndrome of ﬁsheries. Trends in Ecology and Evolution, 10(10):430, 1995.

85. C Schill, N Wijermans, M Schlüter, and T Lindahl. Cooperation is not enough exploring social-ecological micro-foundations for sustainable common-pool resource use. PLoS ONE, 11(8):e0157796, 2016. doi: 10.1371/journal.pone.0157796.

109 Appendix

110 CHAPTER 2

2.A Functions and aggregate parameters

Functions and aggregate parameters Deﬁnition

φi κ/(1 − αiη − α2(1 − η)) Population density on converted land

1−α2 α2 1−α2 γ2 (1 − η) Tm (α2) κ Max. per capita industrial consumption

1−αi αi 1−αi γ1i η Tm (αi) κ Max. per capita agricultural consumption

1 ! Ωz min γ1u −b2γ2 ∆ − 4Ωzy1 min − 1 e µ Sustainability criterion y1

1 ! Ωz min γ1u −b2γ2 θ 4Ωzy1 min − 1 e y1

Table A2.1: Functions and aggregate parameters. i = {u, s}

2.B Parameters deﬁnition, units and defaults values

Parameters Default values Units η Agents preference for agricultural goods 0.35 − αs Sustainable agricultural labor elasticity 0.5 − αu Unsustainable agricultural labor elasticity 0.15 − α2 Labor elasticity in the industrial sector 0.9 − −1 wmax Maximum ostracism varies H t Threshold eﬃciency −200 − r Rate of social change −30 − −α Tm Maximum technological eﬃciency 1.8 H σ Rate of technological change 0.1 H−α.t−1 κ Land operating cost 1 H µ Maximum growth rate 1 H.t−1 min −1 y1 Minimum per capita agricultural consumption 0.3 H b2 Sensitivity to industrial goods’ consumption 3.5 − Ω Concavity of the BES relationship 0.4 − z Concavity of the SAR 0.2 − Ecological relaxation rate 0.1 t−1

Table A2.2: Deﬁnition and default values of the parameters. H: units of labor; t: units of time.

111 CHAPTER 2

When available, parameter values are taken from the literature (, z, Ω, αi, η, t, r). Else, they are calibrated using historical trends, and in order to guarantee the feasibility of

min the equilibria, i.e. positive human population sizes (µ, b2, σ, y1 , κ). Finally, ostracism wmax is varied over a wide range of values for which the system remains feasible.

2.C Dynamical system analysis

In this appendix, we derive the analytical expressions of the equilibria, and conditions for their stability.

  min  ˙ y −y1 −b2y2 H = µ H 1 − e 1 e     ˙ z B = − [B − (1 − H/φ) ] (7)   ˙q= q(1 − q)(Us − Uu)(1 − w(q)/Uu)     ˙ T = σ T [1 − T/Tm]

Parameters and functions are summarized in Tables 2.1 and 2.2, with y1 = q y1s + (1 −

η 1−η q) y1u, φ = q φs + (1 − q) φu, the consumption utility Ui = y1iy2 (i = {u, s}), and the

terq ostracism function w(q) = wmaxe . Solving system (11) for H˙ = 0, B˙ = 0, T˙ = 0 and ˙q= 0 gives ﬁve equilibria: (1) a

∗ ∗ ∗ ∗ sustainable equilibrium, (Hs ,Bs ,Tm, 1), (2) an unsustainable equilibrium, (Hu,Bu,Tm, 0),

∗ ∗ ∗ (3) a mixed equilibrium, (Hc ,Bc ,Tm, qc ), and (4) two unviable equilibria, (0, 1,Tm, 0) and

(0, 1,Tm, 1).

∗ ∗ ∗ We ﬁrst evaluate the Jacobian matrix at the viable equilibria, (H , B , q , Tm) where q∗ = 1 or q∗ = 0. After simpliﬁcation, we obtain:

 ∗ ∗ ∗ ∗  J1 Ωγ1 ∗ γ1 J1 0 ∗ J (γ1s − γ1u)  B 1 Tm     J∗  − 2 − J ∗ φs−φu 0   φ∗ 2 φ∗2  J(H∗, B∗, q∗, T ) =   m    ∗   0 0 J3 0     0 0 0 −σ 

112 CHAPTER 2

∗ z−1 ∗ −b2γ2 ∗ ∗Ω ∗ H ∗ ∗ ∗ ∗ where J1 = µe H B , J2 = z 1 − φ∗ , and J3 = (Us(B )−Uu(B ))(1−2q )(1− w(q∗) ∗ ). Uu

The determinant D of this Jacobian matrix is the product of the four eigenvalues of the system. An equilibrium is locally stable if all its eigenvalues are negative, i.e. D > 0. In order to assess the local stability of the viable equilibria, lets ﬁrst derive the determinant

∗ ∗ ∗ of J(H , B , q , Tm):

1 min −b2γ2 ∗− z ∗ ∗ ∗ ∗ ∗ D = −σzΩµy1 e (B − 1)(1 − 2q )[δuw(q ) + Us(B ) − Uu(B )] (8)

∗ ∗ Uu(B )−Us(B ) where δ = ∗ . u Uu(B )

We obtain the determinant Ds of the Jacobian evaluated at the sustainable equilibrium

∗ ∗ ∗ by taking B = Bs and q = 1, so that:

∗− 1 min −b2γ2 z ∗ ∗ Ds = σzΩµy1 e (Bs − 1) δ [1 − w(1)/Uu(Bs )] (9)

∗ ∗ ∗ minη 1−η η ∗ where δ = (Us(Bs )−Uu(Bs )) = y1 γ2 (1−(γ1u/γ1s) ). Since γ1u > γ1s and Bs ∈ [0, 1], 1 ∗ ∗− z we deduce that δ < 0 and Bs − 1 > 0, so that the sign of Ds depends on the last term

∗ ∗ of eq. (9). The sustainable equilibrium (Hs ,Bs , 1,Tm) is thus locally stable (Ds > 0) if

∗ w(1) > Uu(Bs )

∗ 1−η min η η where Uu(Bs ) = γ2 y1 (γ1u/γ1s) .

∗ The determinant Du of the Jacobian evaluated at the unsustainable equilibrium (B =

∗ ∗ Bu and q = 0) is:

∗− 1 min −b2γ2 z ∗ ∗ Du = −σzΩµy1 e (Bu − 1) δ [1 − w(0)/Uu(Bu)] (10)

∗ ∗ ∗ minη 1−η η ∗ where δ = (Us(Bu)−Uu(Bu)) = y1 γ2 ((γ1s/γ1u) −1). Since γ1u > γ1s and Bu ∈ [0, 1], 1 ∗ ∗− z we deduce that δ < 0 and Bu − 1 > 0, so that the sign of Du depends on the last term

∗ ∗ of eq. (9). The unsustainable equilibrium (Hu,Bu, 0,Tm) is thus locally stable (Du > 0)

113 CHAPTER 2 if

∗ w(0) < Uu(Bu)

∗ 1−η min η where Uu(Bu) = γ2 y1 .

∗ Let us now evaluate the Jacobian matrix at the unviable equilibria, (0, 1, q , Tm) where q∗ = 1 or q∗ = 0. After simpliﬁcation, we obtain:   ∗  J0 0 0 0      −z/φ∗ − z φs−φu 0   φ∗2  J(0, 1, q∗, T ) =   m    ∗   0 0 J3 0     0 0 0 −σ

min ∗ ∗ −b2y2 y1 −γ1 where J0 = µe (1 − e ).

∗ The determinant D(0, 1, q ,Tm) writes:

min ∗ ∗ −b2γ2 y −γ ∗ ∗ D(0, 1, q ,Tm) = σµe (1 − e 1 1 )(1 − 2q )(Us(1) − Uu(1))[1 − w(q )/Uu(1)] (11)

η 1−η η 1−η where Us(1) = γ1sγ2 and Uu(1) = γ1uγ2 , so that Us(1) < Uu(1).

Therefore, the determinant D(0, 1, 0,Tm) is:

min −b2γ2 y −γ1u D(0, 1, 0,Tm) = σµe (1 − e 1 )(Us(1) − Uu(1))[1 − w(0)/Uu(1)] (12)

min Thus, when the viable equilibria are feasible, i.e. when y1 < γ1u < γ1s, the unviable

∗ equilibrium (0, 1, 0,Tm) is stable if w(0) > Uu(1). However, since Uu(1) > U , the unviable equilibrium (0, 1, 0,Tm) is only stable when the corresponding viable equilibrium

∗ ∗ (Hu,Bu, 0,Tm) is unstable.

Similarly, the determinant D(0, 1, 1,Tm) is:

min −b2γ2 y −γ1s D(0, 1, 1,Tm) = −σµe (1 − e 1 )(Us(1) − Uu(1))[1 − w(1)/Uu(1)] (13)

When the viable equilibria are feasible, the unviable equilibrium (0, 1, 1,Tm) is stable if

∗ ∗ w(1) < Uu(1). In this case, both viable (Hs ,Bs , 1,Tm) and unviable (0, 1, 1,Tm) equilibria

∗ can be stable at the same time, if Uu(Bs ) < w(1) < Uu(1).

114 CHAPTER 2

2.D Sensitivity analysis

The sensitivity of the sustainability criterion ∆ to a parameter par (sens∆par ) can be measured as: par ∂∆ sens∆ = · (14) par ∆ ∂par

sens∆par is negative when ∆ and par vary in opposite directions, and positive when they vary in the same direction. The higher the absolute value of sens∆par, the more sensitive ∆ is to par. We calculate sensitivity over a range of parameter values. For each of the 12 parameters of the model with no social pressure - from which the sustainability criterion is derived (see Lafuite & Loreau (2017) [37]), we use a set of 500 values uniformly distributed in an interval of 20 % around the baseline values used in numerical simulations (Table S2).

Fig.A2.1 shows the distribution of the sensitivity of the sustainability criterion ∆ to each of the parameters of the basic model [37]. The sensitivity of ∆ to parameters such

min as the minimum nutritional threshold y1 , the ecological relaxation rate , the cost of land conversion κ and the strength of the demographic transition b2, is positive. These parameters thus positively affect sustainability. Indeed, a higher nutritional threshold favors lower population sizes in the long run, thus leading to a lower land conversion pressure that preserves biodiversity and sustainability. A higher ecological relaxation rate means smaller extinction debts, thus a lower probability of population overshoots. A higher conversion cost reduces the incentives to convert natural habitat, while a stronger demographic transition prevents population overshoots by slowing down the growth of the human population. Conversely, the preference for agricultural goods η, the maximum rate of population growth µ and technological efficiency Tm negatively affect sustainability (sens∆ < 0), since these parameters exacerbate population overshoots (µ) or land conversion (Tm and η).

This sensitivity analysis conﬁrms the sensitivity of the system to extinction debts

min (), but also to nutritional requirements (y1 ), land conversion costs (κ), technological

min eﬃciency (Tm) and consumers’ preferences (η). Two of these parameters, η and y1 ,

115 CHAPTER 2 are expected to be relatively constant, since minimum nutritional requirements and consumers’ preference for agricultural goods are subject to biological constraints. The high sensitivity of ∆ to land conversion costs suggests that economic incentives such as land taxes may help improve the sustainability of the system.

116 CHAPTER 2

10 ∆

−5

−10 Sensitivity of sustainability criterion

−15

min y α b α T 1 ε κ 2 2 Ω z σ 1 µ m η

Figure A2.1: Distribution of the sensitivity of the sustainability criterion ∆. The sensitivity of a variable to a parameter is measured as the elasticity of the variable with respect to the parameter. For each parameter, the local sensitivity is calculated for 500 values uniformly distributed in an interval of ± 20% around the value used for numerical simulations (Table S2).

117 CHAPTER 3

Sustainable land-use management under biodiversity lag eﬀects

Anne-Sophie Lafuite 1, Gonzague Denise 2, Michel Loreau 1

In preparation for Journal of Environmental Economics and Management

1Centre for Biodiversity Theory and Modelling, Theoretical and Experimental Ecology Station, CNRS and Paul Sabatier University, Moulis, France 2Toulouse School of Economics, Toulouse, France

118 CHAPTER 3

Chapter outline

In Chapter 2, we have explored the potential of norm-driven behavioral change in enforcing sustainability, through the bottom-up emergence of large-scale behavioral shifts. A more conventional way to enforce change is through top-down government control. Chapter 3 aims at investigating the eﬃciency of evidence-based social planning in mitigating the negative consequences of land conversion on biodiversity and human well-being, despite imperfect information regarding the temporal dynamics of biodiversity loss. This work builds upon the tools developed by environmental and resource economics in order to internalize the negative externalities of human activities.

119 CHAPTER 3

Abstract

The destruction of natural habitats for agricultural production results in local biodiversity loss. However, many biodiversity-dependent ecosystem services directly or indirectly impact agricultural production. Land conversion thus results in a negative externality, mediated through the erosion of biodiversity. When the consequences of this externality are delayed in time, lack of internalization results in overshoot-and-collapse dynamics, which are undesirable from a sustainability perspective. Here, we show that the internalization of this externality through a land tax results in several win-win eﬀects in the long run. First, more biodiversity is preserved at equilibrium, which increases the carrying capacity and total well-being of the human population. Second, an optimal taxation path that maximizes the discounted sum of human utilities prevents or greatly alleviates such crises, thus increasing the sustainability of the system. Especially, this result holds in the case of imperfect information regarding the precise temporal dynamics of biodiversity loss, suggesting that the design of eﬃcient land-use management policies is possible despite incomplete ecological data. This study thus highlights the necessity of internalizing biodiversity-dependent externalities through economic incentives, especially under uncertainty regarding long-term ecological dynamics.

Keywords: Biodiversity ; ecological economics ; ecosystem services; extinction debt; foresight; social-ecological system ; sustainability; land tax

120 CHAPTER 3

Introduction

Human use of land has transformed ecosystems across most of the terrestrial biosphere for millennia [1]. The conversion of natural lands to croplands, pastures and urban areas represents the most visible form of human impact on the environment [2], with 40% of Earth’s land surface being currently under agriculture [3], and 75% experiencing measurable human pressures [4]. These pressures are rapidly intensifying in biodiversity-rich places, since most land conversion occurs in the tropics through forest conversion to agriculture [5, 6]. As a consequence, land use and land cover changes are among major drivers of biodiversity loss, at both local [7] and global scales [8].

In turn, biodiversity loss affects the provisioning of essential ecosystem services, such as pollination, pest control, nutrient cycling and erosion control [9], with consequences on many human activities, and especially for agricultural production [8]. Biodiversity loss is thus a major and underestimated feedback that may affect human population growth in the long run [10], while the magnitude of land use effects on both biodiversity and biodiversity-dependent services is raising concerns about the potential of land-use changes to push terrestrial biodiversity beyond the planetary boundary [11].

These impacts of land-use change on biodiversity are poorly reﬂected in market prices, and hence have been mostly ignored by decision-makers, despite their large cost for human economies, with an estimated value for the global ecosystem services of $145 trillion in 2011, and up to $20 trillion loss per year between 1997 and 2011 [12]. Loss of biodiversity-dependent ecosystem services thus constitutes a negative externality, that threatens intergenerational equity and sustainability [13].

At the global scale, land conversion is primarily driven by the growth of the human population [14], and arable lands are rapidly shrinking [15]. Recent evidence suggests that land use efficiency has been rising at the global scale [4]. However, such efficiency gains may not help save natural habitats and biodiversity in the long run, due to economic rebound effects, if lower prices stimulate demand and if higher yields raise profits, encouraging agricultural expansion [15]. By increasing the opportunity cost of conserva-

121 CHAPTER 3 tion, these eﬀects undermine the eﬃciency of regulatory environmental policies, such as government protected forests and natural habitats, in protecting biodiversity [16].

Land-sparing mechanisms that could help overcome these rebound effects include land zoning, incentive-based economic instruments (e.g. land taxes, subsidies and payments), spatially strategic intensification and voluntary standards [16]. Especially, incentive-based mechanisms such as land taxes may allow internalizing the externality of land conversion on biodiversity-dependent ecosystems services and agricultural production [17]. Such mechanisms are based on economic efficiency concepts, so as to achieve the maximum amount of resource protection for a given production level.

During the past decade, the European Union has widely used incentive-based mechanisms to reduce gas emissions from motor fuels and vehicles, but also plastic bags, landfill waste, batteries, pesticides, and fertilizers. Mounting evidence shows that taxes have helped reducing pollution and the consumption of natural resources in many cases, with a higher efficiency and at lower costs than conventional regulatory approaches [18]. How- ever, use of such negative price signals for environmentally damaging activities has been less spread in the US, where tax credits and deductions are favored. More generally, the low level of acceptance of taxes lies in interest group pressures and extensive data requirements (e.g., regarding the external costs of human activities), as well as inadequate sensitivity to issues of sustainability and scientific uncertainty.

Indeed, the eﬃciency of conventional taxes is limited by available scientiﬁc knowledge. This is especially true for the relationship between biodiversity-dependent ecosystem service loss and land use changes, for which there is still a high uncertainty regarding the long-term temporal dynamics of ecosystems in the context of accumulating extinction and functioning debts [19, 20, 21], i.e. the time-delayed loss of species and services following a change in land use. Moreover, conventional taxes do not necessarily guarantee intergenerational equity and sustainability, i.e. do not prevent the over-use of natural capital and reductions in human well-being over time [13, 22].

Some authors have thus proposed to deﬁne a broad natural capital depletion tax to ensure that resource inputs from the environment to the economy stay within planetary

122 CHAPTER 3 boundaries and are sustainable [23, 24, 25]. Implementation of such a tax would raise prices of natural resources, thus encouraging technological advances while slowing down the rate of environmental depletion [18]. Other authors have proposed a corrected version of the net national product that accounts for the eﬀect of agricultural land development on biodiversity, while ensuring a constant social welfare [26, 27].

However, these developments have poorly accounted for the temporal dynamics of biodiversity-dependent ecosystem service loss, and do not consider this negative feedback on human demography. Indeed, biodiversity-dependent agricultural consumption aﬀects human demography, resulting in a dynamic feedback loop between land conversion, biodiversity loss and human population growth [28]. Time delays between land conversion and biodiversity loss, mediated through extinction debts [19], result in a lagged feedback on agricultural production [29, 21, 20]. Such lag eﬀects can result in long-term environmental crises, i.e. overshoot-and-collapse population cycles [28], which reduce human well-being.

In this paper, we propose to assess the efficiency of a natural land depletion tax in securing sustainability and preserving biodiversity, despite uncertainty about the temporal dynamics of biodiversity loss. The paper is organized as follows. In section 1, we present a dynamical system model that couples human demography and technological change to biodiversity loss, through the effect of land conversion on the flow of biodiversity- dependent ecosystem services to agricultural production [28]. In section 2, the externality of land conversion on biodiversity is internalized through a natural land depletion tax τ per unit of converted land. We show how this tax affects the consumption levels, the ratio of the production inputs, and the rate of land conversion. In section 3, we analyze the effects of this tax on the long term equilibria and sustainability of the system, as captured by a criterion ensuring a non-decreasing human well-being over time. We show that a land tax can increase both biodiversity and total agricultural production at equilibrium, when the substitution of land for labor and ecosystem services has a net positive effect on total agricultural production. The land tax also reduces the vulnerability of the system to time delays, but its ability to prevent crises depends on its level at equilibrium, and thus on the land conversion policy. Section 4 derives the optimal land conversion policy designed

123 CHAPTER 3 by a foresighted planner, who aims to internalize the externality of land conversion on biodiversity under the assumption that the temporal dynamics of biodiversity is unknown. We illustrate the eﬃciency of such a policy in preserving biodiversity, increasing total production, and preventing the unsustainable consequences of time-delayed ecological feedbacks, despite incomplete scientiﬁc knowledge regarding their temporal dynamics.

1 A simple land - biodiversity - demography model

1.1 Substitution of natural capital for production inputs

We build upon the model of Lafuite & Loreau (2017) [28], that considers a population of consumers whose demand for agricultural (i = 1) and industrial (i = 2) goods requires the conversion of their common natural habitat. The two goods in the model are each produced using labor Li and land Ai, and we assume full-employment, i.e. total labor is captured by the size of the human population. Only converted land is capable of producing these goods, while land not converted for production remains as natural habitat capable of supporting species, and provides a range of biodiversity-dependent ecosystem services to agricultural production [9].

Total factor productivity increases with technological efficiency in both sectors, as well as with biodiversity-dependent ecosystem services in the agricultural sector. The ecosystem services provided by this community of species are assumed to increase with biodiversity and saturate at high levels of species richness, through a power-law relationship BΩ, where Ω ∈ [0, 1] [30]. Technological efficiency is also assumed to follow a logistic growth towards a maximum efficiency, Tm, in order to reproduce past agricultural productivity rise and current stagnation [31].

By using Cobb-Douglas production functions (eq.(1)), we allow for the partial substitution between production inputs (labor and land), but also between natural capital (biodiversity-dependent services) and technology.

Ω α1 1−α1 α2 1−α2 Y1 = TB L1 A1 Y2 = TL2 A2 (1)

124 CHAPTER 3

Human population (H)

) y Regulatory ,τ Economic 2 (T ,T , B τ ( market ) agency y 1 L 2 L 1 Agriculture Technology Industry (T) A f (B) 1 A 2 S Economic market Biodiversity (B) Land Converted land conversion S A(H,τ)=A +A (H 1 2 A(H,τ) , τ) Natural habitat 1-A(H, )=A τ 3

Figure 3.1: A simple land use, biodiversity and human demography model. τ: natural land depletion tax; y1: per capita agricultural consumption; y2: per capita industrial consumption; L: labor; A: land; S: species-area relationship; fS: biodiversity- dependent ecosystem services. Modiﬁed from Lafuite & Loreau (2017)

1.2 Dynamical system

The long-term behavior of the population is captured by a feedback loop between three dynamical variables: the human population H (eq.(2)), biodiversity B (eq.(3)), and technological eﬃciency T (eq.(4)).

ymin−y (B,T) −b y (T) H˙ = µH(1 − e 1 1 )e 2 2 (2)

B˙ = −(B − S(H)) (3) ˙ T = σT(1 − T/Tm) (4)

125 CHAPTER 3

Human demography can be related to the agricultural and industrial consumptions, y1(B, T) and y2(T) [32, 33, 34, 35]. These consumptions are derived at market equilibrium (Appendix 3.A):

Ω y1 = γ1B T/Tm y2 = γ2T/Tm (5)

where γ1 and γ2 are functions of the parameters of the system (Table 2). A higher agricultural production allows the population to grow at a maximum rate µ while the

min average per capita consumption is higher than a minimum consumption, y1 . Human growth is slowed down by a demographic transition factor, the strength of which increases with industrial production, technological eﬃciency, and a scaling parameter b2. The number of remaining species S is determined by a species–area curve relationship, S(H) = (1 − H/φ)z, where z is a constant parameter [36, 37, 38, 39], and φ is the density of the human population on converted land at market equilibrium (Table 2).

Human population growth results in land conversion that reduces the number of species S(H) that natural habitat can support, thus leading to species extinction. These extinctions are delayed in time [19, 40], hence the long-term species richness may be reached only after decades [41]. In order to account for this temporal dynamics, we build upon theoretical and experimental evidence regarding the relaxation rate of natural communities following habitat loss [42], which is proportional to the diﬀerence between current biodiversity B and long-term species richness S(H), and to a relaxation parameter .

1.3 Sustainability conditions

Model analysis allows deriving conditions for its stability and sustainability [28], i.e. non-

η 1−η declining human well-being over time [13], where well-being U = y1 y2 is a function of the agricultural and industrial consumptions, and the preference for agricultural goods η. Two necessary sustainability conditions have been identiﬁed that involve a suﬃciently high level of substitution of natural capital for technology on the one hand, and the resistance to transient overshoot-and-collapse population crises on the other.

Indeed, rising technological eﬃciency compensates for the negative feedback of biodiversity- dependent ecosystem services on agriculture if it is higher than the loss of services in terms

126 CHAPTER 3 of human well-being, i.e.

∗ Ωη Tm/T(0) > (B(0)/B ) (6)

However, this condition is not suﬃcient to ensure sustainability, since time delays in the feedback of biodiversity loss on human demography can result in unsustainable overshoot-and-collapse population cycles. This is the case when biodiversity loss is much slower than human population growth, i.e.

< σµ where σ is a function of the parameters of the system (Table 2).

In the following, we aim at assessing the eﬃciency of a land tax in preserving the sustainability of this system under a time-delayed biodiversity feedback on human population growth. First, we show how the tax aﬀects the consumption levels, conversion rate and long-term equilibria of the system.

2 A natural land depletion tax

2.1 Production

A tax τ per unit of converted area is added to the maintenance cost of κ units of labor per unit of land. At each period, the production proﬁt in sector i = {1, 2} is

Ω αi 1−αi Πi = piTB Li Ai − wLi − (κw + τ)Ai

with Li the human labor, Ai the exploited area in sector i, pi the price of the output, w the consumer wage, and αi the elasticity of labor. Proﬁt maximization gives the production supply for each sector (Appendix 3.A), and the relationship between input factors in each sector (eq.(7)) shows that a tax τ increases the optimal ratio of labor to land, thus

127 CHAPTER 3 generating an incentive to substitute land for labor.

Li/Ai = (κ + τ)αi/(1 − αi) (7)

2.2 Consumption

The total revenue of the tax τA, where A = A1 + A2 is total converted land, is then redistributed among the consumers, who are assumed to maximize their consumption utility U under their revenue constraint p1y1 + p2y2 ≤ w + τA/H. As a result of this redistribution, the total demand for agricultural and industrial goods (eq.(8)) increases with the land tax.

τA! τA! p Y D = η w + H p Y D = (1 − η) w + H (8) 1 1 H 2 2 H

At the equilibrium between supply and demand, the optimal allocations of labor and land (Appendix 3.A) provide the equilibrium consumptions in the agricultural and industrial sectors: φ κ + τ α1 φ κ + τ α2 y1τ = y1 y2τ = y2 (9) φτ κ φτ κ where y1 and y2 are the business-as-usual consumptions (eq.(5)), and φ and φτ are the densities of the human population on converted land, in the business-as-usual and the regulated cases respectively.

2.3 Land conversion

The density of the human population on converted land, φτ = H/A, also derives from the labor market equilibrium L1 + L2 + κA = H.

φ − κ! φ = φ + τ (10) τ κ where φ is the population density in the absence of regulation (Table 2). Note that φ > κ since αi ∈ [0, 1] and η ∈ [0, 1], so that φτ > φ. A tax τ per unit of converted land thus

128 CHAPTER 3 increases the human population density on converted land, which aﬀects the converted surface A = H/φτ , and the long-term number of species that the remaining natural habitat

z can support, S(H) = (1 − H/φτ ) .

Since φτ increases with τ, the eﬀect of the tax on consumptions (eq.(9)) is not straightforward, and depends on the economic parameters of the system. Moreover, a land taxation policy will only help preserve more natural habitats and biodiversity compared to a business-as-usual case, if the human population density on converted land φτ increases faster than the size of the human population, H. The next section explores the conditions under which this objective can be met, by studying the eﬀects of the tax on the equilibrium features of the model.

Parameters Default values Units Economic parameters η Agents’ preference for agricultural goods 0.5 − α1 Agricultural labor intensity varies − α2 Industrial labor intensity varies − δ Discount rate 0.04 t−1 Technological parameters −α Tm Maximum technological eﬃciency 1 H σ Rate of technological change 3 H−α.t−1 κ Land operating cost 0.2 H Demographic parameters µ Maximum growth rate 1 H.t−1 min −1 y1 Minimum per capita agricultural consumption 0.3 H b2 Sensitivity to industrial goods’ consumption 0.1 − Ecological parameters Ω Concavity of the BES relationship 0.4 − z Concavity of the SAR 0.3 − Ecological relaxation rate 1 t−1

Table 3.1: Deﬁnition and default values of the parameters and dynamical variables. H: units of labor; t: units of time.

129 CHAPTER 3

Functions and aggregate parameters Deﬁnition

φ κ Population density without regulation 1−α1η−α2(1−η)

φ−κ φτ φ + τ κ Population density with regulation

1−α2 α2 γ T (1 − η)αα2 1−α2 φ κ+τ Max. per capita industrial consumption 2τ m 2 κ φτ κ

1−α1 α1 γ T ηαα1 1−α1 φ κ+τ Max. per capita agricultural consumption 1τ m 1 κ φτ κ

1 ! Ωz min γ1τ −b2γ2τ ∆τ − 4Ωzy1 min − 1 e µ Sustainability criterion y1

Table 3.2: Functions and aggregate parameters expression and deﬁnition. The expressions in the unregulated case are obtained by taking τ = 0, so that γ1 = γ1(τ=0), γ2 = γ2(τ=0), ∆ = ∆(τ=0)

3 Dynamical system analysis

Here, we analyze the eﬀect of a land tax τ on the equilibria of the regulated system:

  ymin−y (B,T) −b y (T) H˙ = µ H 1 − e 1 1τ e 2 2τ    T˙ = σT(1 − T/T ) (11)  m    ˙ z B = − [B − (1 − H/φτ ) ]

Parameters and functions are summarized in Table 3.1 and 3.2.

3.1 Steady states and sustainability conditions

The equilibrium of the system is reached when technological eﬃciency is at its maximum

˙ min level Tm (T = 0), human consumption is at its equilibrium level y1 so that human population cannot grow anymore (H˙ = 0), and the extinction debt of biodiversity has been entirely paid, so that B = S(H) (B˙ = 0).

∗ ∗ There are two possible equilibria: (1) a desirable equilibrium, (Hτ ,Tm,Bτ ), and (2)

130 CHAPTER 3

an undesirable equilibrium, (0,Tm, 1).

1 min ! Ω 1 ∗ y1 ∗ ∗ z Bτ = Hτ = φτ (1 − Bτ ) γ1τ

where γ1τ and φτ are explicitly deﬁned in Table 3.2.

The sustainability conditions of the system now depend on the tax τ (Table 3.2):

!Ωη Tm B(0) > ∗ ∆τ > 0 (12) T(0) Bτ

3.2 Eﬀects of the tax on equilibrium features

The eﬀect of the tax τ on the equilibrium features of the model is mediated through the relationships γ1τ (τ), i.e. the level of substitution of land for natural capital in the agricultural production, and γ2τ (τ), i.e. the per capita level of industrial consumption

∗ at equilibrium. Indeed, the level of biodiversity at equilibrium Bτ directly depends on

∗ γ1τ , which in turn determines the level of human population Hτ , while γ2τ determines the level of industrial consumption, and thus the human well-being at equilibrium u∗ =

min η 1−η (y1 ) γ2τ .

The shapes of γ1τ (τ) and γ2τ (τ) depend on the economic parameters of the system, and especially on the labor elasticities in the agricultural and industrial sectors, α1 and

α2. Labor elasticity captures the increase in output resulting from a 1% increase in labor. Since the main effect of the tax is to increase the ratio of labor to land, varying labor elasticities between sectors result in differing effects of the taxation policy on the equilibrium features of the system.

3.2.1 Eﬀect on biodiversity, sustainability and population size

∂γ1τ The tax has a positive eﬀect on biodiversity if τ < 0. It can be shown that

∂(B∗ − B∗) τ > 0 for α ≤ α (13) ∂τ 1 2

131 CHAPTER 3 so that the tax τ always has a positive eﬀect on the long-term level of biodiversity, when the labor elasticity of the industrial sector is higher or equal to that of the agricultural sector (α2 ≥ α1), i.e. the most common situation in real-world systems.

Under the assumption that α1 ≤ α2, we distinguish two situations: (1) labor elasticity is higher in the industrial than in the agricultural sector (α2 > α1), so that γ1τ decreases with τ while γ2τ increases (Fig.3.2.A/C/E/G) and (2) labor elasticity in the industrial and agricultural sectors are similar (α2 ≈ α1), so that both γ1τ and γ2τ decrease with τ (Fig.3.2.B/D/F/H).

By increasing the ratio of labor to land, the tax reduces land conversion and allows preserving more biodiversity through the substitution of land for biodiversity, i.e. γ1τ decreases with τ (Fig.3.2.C and D). This higher biodiversity level at equilibrium ensures a higher sustainability of the system, which becomes less vulnerable to transient overshoot- and-collapse crises, as captured by our sustainability criterion ∆τ > 0 (Fig.3.2.G and H).

This reduction of land conversion is not only compensated by a higher natural capital, but also by a larger labor force, which increases the size of the human population (Fig.3.2.A and B). However, the eﬀect of the tax on the size of the human population at equilibrium is non-linear, since high tax levels reduce the incentive for land conversion to the point where it becomes economically unviable to convert land anymore, thus reducing the size of the human population at high tax levels (Fig.3.2.A and B).

3.2.2 Eﬀect on industrial consumption and human well-being

Distinction between cases (1) and (2) lies in the eﬀect of τ on the consumption of industrial goods at equilibrium, γ2τ , and thus on human well-being. When labor elasticity is higher in the industrial than in the agricultural sector (α2 > α1), land taxation increases industrial consumption at equilibrium compared to the business-as-usual case (Fig.3.2.E). However, if the industrial labor elasticity is lower (α2 ≈ α1), land taxation reduces industrial consumption (Fig.3.2.F).

The tax level required to achieve a positive sustainability criterion (Fig.3.2.G and H)

132 CHAPTER 3 or to maximize human well-being (Fig.3.2.A and B) is much higher in case (2) than in case (1), so as to compensate for the lower labor to land ratio of the industrial sector. Thus, the total labor force, i.e. the size of the human population, is also higher compared to case (1) (Fig.3.2.A and B). This increase in population size thus reduces both per capita industrial consumption and human-well-being, in case (2) compared to case (1) (Fig.3.2.E and F). Despite its positive eﬀects on biodiversity and sustainability, a land tax may thus reduce per capita well-being if the initial labor elasticities are too low, through a large increase in labor, i.e. population size.

3.2.3 Optimal tax vs. sustainability

A regulator seeking to maximize total human well-being through an optimal tax τ opt (Fig.3.2.A and B) thus needs to account for the economic structure of the system, and especially the relative values of α1 and α2. However, such an optimal tax may not prevent unsustainable crises from occurring, since low relaxation rates can reduce the sustainability of the system to the point where it becomes vulnerable to crises, i.e. if ∆τ=τ opt < 0 (Fig.3.2.G and H). Thus, the higher the extinction debt, the higher the tax on land conversion should be in order to avoid unsustainable trajectories. Time-delayed biodiversity loss may make the optimal tax level τ opt unsustainable, so that a land policy that does not account for such temporal delays may fail to prevent unsustainable crises. In the next section, the optimal land conversion policy is derived in the case of a foresighted regulator, and its eﬃciency in preventing crises is explored.

133 CHAPTER 3

α >> α α ≈ α 2 1 2 1 A B ) * τ ) 2 2 2 0.5 * τ .u * τ 1.5 1.5

1 1

0.5 0.5

0 0 0 0 Human pop. (H 0 0.5 1 1.5 2 0 0.5 1 1.5 2 Total utility (H τopt τopt

) C D * τ 1 1 1 1

0.5 0.5 0.5 0.5 1 τ γ

0 0 0 0

Biodiversity (B 0 0.5 1 1.5 2 0 0.5 1 1.5 2

E F 0.5 0.5 0.7 0.7 ) * τ

2 τ 0.4 0.4 0.5 0.5 γ Utility (u 0.3 0.3 0.3 0.3 0 0.5 1 1.5 2 0 0.5 1 1.5 2 )

τ G H

∆ 5 5 ε=1 ε=0.01 0 0 ε=1 ε=0.01 −5 −5 0 0.5 1 1.5 2 0 0.5 1 1.5 2 Sustainability ( Tax (τ) Tax (τ)

Figure 3.2: Eﬀect of a land tax τ on the equilibrium features of the model, for various labor elasticities α1 and α2.

134 CHAPTER 3

4 Optimal land conversion policy

Here, we allow the tax to vary with the dynamical variables of the system, so as to internalize the negative eﬀects of biodiversity loss on agricultural production, while maximizing human-well-being.

4.1 Analytical derivations

At each period, and for a population size H, a technological eﬃciency T, and a biodiversity level B = (1 − H/φτ ), we assume that a benevolent social planner aims at maximizing the total discounted utility of consumers,

η 1−η Uτ (B, T, H) = H · y1τ (B, T) y2τ (T) using a land tax τ per unit of converted land as control. This land tax limits the conversion of natural habitat so as to internalize the negative externalities on biodiversity and agricultural production, and varies with the size of the dynamical variables of the system.

The objective of the social planner is to maximize the present value of a continuous sum of discounted utilities, at an annual rate δ, subject to the dynamics of the human ˙ ˙ population, H (eq.(2)), technological change T (eq.(4)), the consumptions levels y1τ and y2τ , as well as to the loss of biodiversity, B = (1 − H/φτ ). Thus, we assume here that the social planner does not know the temporal dynamics of biodiversity loss, B˙ (eq.(3)). However, he accounts for the long term eﬀects of land conversion on biodiversity, through the use of a species-area relationship, which is one of the best-known patterns in ecology [43].

Z ∞ −δt max e Uτ (B, T, H) dt t0

135 CHAPTER 3 subject to

  min  ˙ y1 −y1τ (B,T) −b2y2τ (T) H = µ H 1 − e e    T˙ = σT(1 − T/T )  m   B = (1 − H/φ )z (14)  τ    Ω y1τ = γ1τ B T/Tm     y2τ = γ2τ T/Tm

The Hamiltonian function for this problem is

˙ ˙ H = Uτ (B, T, H) + λH H + λT T (15)

where λH and λT are adjoint variables. First order conditions are

∂H ∂H ∂H = 0, = δλ − λ˙ , = δλ − λ˙ (16) ∂τ ∂H H H ∂T T T and transversality conditions are

lim H(t) · λH (t) = 0 lim T(t) · λT (t) = 0 t→+∞ t→+∞

∂H Solving for the ﬁrst order condition ∂τ = 0 gives the optimal tax τ as a solution of the following equation:

!  !1−η  ∂y y min 1τ 2τ y −y1τ −b2y2τ η + λH µe 1  = (17) ∂τ y1τ ! !η! ∂y min y 2τ −b2y2τ y −y1τ 1τ b2λH µe (1 − e 1 ) − (1 − η) (18) ∂τ y2τ

∂H ˙ Solving for the ﬁrst order condition ∂H = δλH − λH gives the dynamics of the adjoint variable λH as a function of the other dynamical variables of the system (B, H, T) and the

136 CHAPTER 3 control τ:

" #! −b y ymin−y ymin−y ∂y1τ ∂Uτ λ˙ = λ δ − µe 2 2τ (1 − e 1 1τ )) + He 1 1τ − U − H H H ∂H τ ∂H

˙ We do not need the last condition, which gives the temporal dynamics λT , since technological efficiency varies exogenously, and thus does not depend on the other variables of the system, nor on the control τ. ˙ We then simulate system (11) along with the dynamics of the adjoint variable λH , by solving at each time step for τ using eq.(18). Thus, though the social planner does not account for time-delayed biodiversity loss (B˙ ), the numerical simulations do include the effect of extinction debts on agricultural production, human consumption and human demography. Simulations allow exploring the effects of the tax on the transient dynamics of the regulated system.

4.2 Numerical simulations

Parameters are chosen so as to meet the sustainability condition (6) in the business- as-usual case. This guarantees that the substitution of natural capital for technology is high enough to ensure a non-declining human well-being over time. Since the tax policy necessarily preserves more biodiversity at equilibrium, the condition remains true in the regulated case. Thus, we can focus on the consequences of ecological time delays for sustainability, by comparing their eﬀects on the regulated and business-as-usual scenarios.

In the case of a negligible extinction debt (e.g. = 1), both the regulated and business-as-usual trajectories are sustainable, i.e. do not experience transient overshoot- and-collapse population crises. Fig.3.3 conﬁrms the eﬀect of land taxation presented in the previous section, since the tax increases biodiversity (Fig.3.3.C) and human population size (Fig.3.3.A) at equilibrium, compared to the business-as-usual case.

The optimal tax increases with the size of the human population, before reaching its equilibrium level, previously denoted as τ opt (Fig.3.3.E). This regulatory policy increases the carrying capacity of the human population, i.e. the maximum population size that

137 CHAPTER 3 the environment can support (Fig.3.3.A). Moreover, the distance between the population equilibrium and the carrying capacity is larger in the regulated than in the business-as- usual case, a feature which increases the resistance of the system to time delays, i.e. its sustainability [28]. Indeed, for a higher extinction debt (e.g. = 0.005), the regulated system appears much more resistant to transient population crises compared to the business-as-usual scenario (Fig.3.3.B). The transient dynamics of the optimal taxation path also changes at high extinction debts, since the tax reaches higher levels during the initial growth phase of the human population (Fig.3.3.F) in order to counteract the faster population growth, and prevent it to overshoot its carrying capacity. The eﬃciency of the optimal taxation policy also holds for higher extinction debts, thus making the system very resistant to time- delayed feedbacks, despite incomplete knowledge regarding the precise temporal dynamics of biodiversity loss. Fig.3.4 shows the eﬀect of this taxation policy when the system is initially overshooting its carrying capacity φ, for various ecological relaxation rates . Implementation of the optimal tax stops the unsustainable population growth, through a high tax value which fosters land restoration (Fig.3.4.C). The resultant reduction in the size of the human population (Fig.3.4.A) leads to the tax value to decrease until the sustainable equilibrium is reached, while biodiversity is slowly recovering (Fig.3.4.B). A larger time delay in biodiversity recovery results in a longer degrowth phase of the human population, the size of which falls below its equilibrium value before increasing again. In the unregulated scenario, the initial overshoot results in more population growth (Fig.3.4.A) and biodiversity loss (Fig.3.4.B), especially at high extinction debts. The business-as-usual scenario leads to a larger long-term population reduction than in the regulated case. Policy regulation thus greatly alleviates long-term population crises in a system in overshoot - even when biodiversity recovery is very slow.

138 CHAPTER 3

ε=1 ε=0.005 A B 8 15 Tax Tax No tax No tax 6 10

5 2 Human pop. (H) 0 0 0 200 400 600 800 1000 0 500 1000 1500 2000

C D 1 1 Tax Tax 0.8 No tax 0.8 No tax

0.6 0.6

0.4 0.4

0.2 0.2 Biodiversity (B) 0 0 0 200 400 600 800 1000 0 500 1000 1500 2000

E F 6 6

5 5

4 4 τ ) 3 3

Tax ( 2 2

1 1

0 0 0 200 400 600 800 1000 0 500 1000 1500 2000 Time Time

Figure 3.3: Land-use management scenarios for varying ecological relaxation rates (). Dashed lines represent the carrying capacities φ of the human population in each case. Initial values: B(0) = 1; H(0) = 0.1; T (0) = 0.5; λH (0) = −12. Parameter min values: α1 = 0.5; κ = 1; µ = 0.1; y1 = 0.3; Ω = 0.4; z = 0.3; Tm = 1; δ = 0.04; σ = 0.3; η = 1.

139 CHAPTER 3

10 no tax ε=0.005 tax ε=0.001 8

2 Human pop. (H) 0 0 500 1000 1500 2000 2500 3000 3500

B 1 no tax ε=0.005 0.8 tax ε=0.001

0.6

0.4

0.2 Biodiversity (B)

0 0 500 1000 1500 2000 2500 3000 3500

C 3

2.5

2 τ ) 1.5

Tax ( 1

0.5

0 0 500 1000 1500 2000 2500 3000 3500 Time

Figure 3.4: Land-use management scenarios from an overshoot situation, and for varying ecological relaxation rates (). The dashed lines represent the carrying capacities φ and φτ of the human population. Initial values: B(0) = 0.2; H(0) = 5; T (0) = 0.5; λH (0) = −12. Parameter values are given in Table 1, except for µ = 0.11. Green curves: regulated system; black curves: business-as-usual; dotted curves: = 0.001; continuous curves: = 0.005.

140 CHAPTER 3

5 Conclusions and discussion

In this paper, we linked a simple general equilibrium market model with a dynamical system coupling human demography, technological change and time-delayed biodiversity loss, through the consumption of agricultural and industrial goods, and the time-delayed feedback of biodiversity-dependent ecosystem services on agricultural production. We then demonstrated the eﬀects of a natural land depletion tax on long-term biodiversity, human carrying capacity, well-being and sustainability of the system. Because the land tax increases the ratio of labor to land used in the agricultural and industrial productions, it fosters land-use intensiﬁcation through the substitution of land for labor in both sectors, but also for biodiversity-dependent ecosystem services in the agricultural sector.

Internalization of biodiversity-dependent externalities on agricultural production thus fosters ecologically-intensive agricultural systems, that preserve more biodiversity and can support a larger human population. To the extent that biodiversity is aligned with the sustainability of the system, i.e. its resistance to overshoot-and-collapse population crises resulting from a time-delayed loss of biodiversity-dependent ecosystem services [28], such a land taxation policy means more stability and sustainability. However, a land tax can have adverse consequences for the per capita human well-being through its effect on industrial consumption, especially in systems with a low industrial labor intensity, in which substitution of land for labor is not compensated by an increase in natural capital. This result may change when considering a symmetrical effect of biodiversity on both the agricultural and industrial sectors, since many industrial activities rely on natural services. Indeed, deforestation and biodiversity loss can affect regional and global climate [44] with feedbacks on hydrology [45] and other important provisioning services to the industrial sector, such as wood production and clean water.

The stabilizing eﬀect of the tax did not require a precise knowledge of the ecological dynamics of the system, and especially of the temporal dynamics of biodiversity loss. Our results thus suggest that lack of data and uncertainty about complex ecological dynamics should not prevent land-use management to adopt a precautionary approach to

141 CHAPTER 3 environmental uncertainty [46]. Sustainable land-use management policies should build upon well-known ecological patterns, such as species-area relationships [43], as well as recent advances in ecological research that allow making large-scale predictions up to continental or global scales, ranging from the future distribution of biological diversity to changes in ecosystem functioning and services [47, 48].

The model we developed in this paper was kept quite simple in order to make clear the basic logic of how economic incentives can affect land conversion, biodiversity conservation and sustainability. For example, the proportional relationship between human population growth and land conversion, resulting from the assumption of a constant maintenance cost of κ units of labor per unit of converted land, is unrealistic. Drivers of land use and land cover change are abundant, complex and scale-dependent [49]. At global scale, recent evidence that human population and the world economy are growing faster than the human footprint suggests a globally more efficient use of land [4], so that endogenizing efficiency gains and technological change along with economic growth appears essential in order to gain realism in the relationship between human population growth and land conversion.

Our model considered the use of converted land taxation as a way to preserve natural habitats. Property taxes have been used in several other contexts [50], as a source of revenue [51], a way of promoting urbanization [52] or conversely, in order to reduce the use of land for house building [53] and foster land-use efficiency [50]. However, despite the established efficiency of taxes as a way of internalizing the externalities of human activities [54], such as modern agriculture [55], land taxes are rarely used for conservation purposes, especially in rural areas [56]. Reasons include a higher riskiness of net farmer income, costly administration and informational requirements [57], difficulty to administer progressive tax rates based on land holdings, and political acceptability of negative price signals [51]. For these reasons, taxes remain marginal in both research and implementation on the internalization of ecosystem services into economic decisions [54], in comparison to national governmental payment programs [58], such as the green payments and subsidies implemented by the European Common Agricultural Policy, whose

142 CHAPTER 3 efficiency in preserving biodiversity appears limited [59, 60]. Strong political will is thus required to shift the current paradigm and improve the efficiency of agricultural policies, through a better identification and management of the conflicts between agriculture and biodiversity conservation [61] and the development of more integrated approaches to policy, land-use and biodiversity [62].

Moreover, despite mounting evidence of synergies between biodiversity and multiple ecosystem service conservation [63, 64, 65], scientific understanding of ecosystem production functions remains a limiting factor in incorporating natural capital into economic decisions [66]. Improving scientific knowledge of the multi-layered relationship between biodiversity and ecosystem services [67] is a necessary step towards the reconciliation of conventional conservation management focused on conserving specific species and habitats [68], with the management of multiple ecosystem services [69, 65] within an integrative approach [70].

There are a number of other ways that the model could be enriched, both on the economic side and on the ecology side. We assumed that the provisioning of services to productive lands depended on the total area of natural habitat only. In reality, service provisioning is spatially- and distance-dependent, since intermediate habitat heterogeneity and fragmentation is required to provide access to several services, such as pest control and pollination [71]. In turn, habitat fragmentation affects the viability of communities and generates extinction debts [40, 21]. Expanding the model in this way requires the spatialization and differentiation of economic incentives, and the distinction between local and regional or global species richness and services. Second, the assumption that species can only utilize natural habitat is also too restrictive, since certain species can persist on working landscapes (agricultural fields and managed forests). Finally, we have modeled utility as a function of the indirect effect of global species richness on private consumption. It is equally plausible that alternative measures of biodiversity enter the utility function, such as its cultural, spiritual and aesthetic value. Including these as arguments of utility would require a different objective function, but would not qualitatively alter the results of this paper.

143 CHAPTER 3

Enriching both the economic and ecological sides of the model could add insights and greater realism to the analysis of the links between land-use management, sustainability and biodiversity conservation. Specifically directing policy on the basis of this work will require going beyond the conceptual model presented here. Including the feedback of biodiversity on human demography into complex existing modeling tools, such as the Integrated Valuation of Ecosystem Services and Tradeoffs (InVEST) model [72], may help assess and design integrative food-biodiversity policies, in a sustainability perspective. InVEST already models multiple ecosystem services, biodiversity conservation, commodity production, and tradeoffs at landscape scales, but poorly accounts for the relationships between biodiversity and services, and ignores human demography. Other global Integrated Assessment Modeling frameworks have been developed to specifically address climate change [73] or water availability [74] issues. For example, the Dynamic Integrated Climate-Economy (DICE) model has been extensively used to inform optimal tax policies aimed at reducing greenhouse-gas emissions [75]. These more complex models, however, would still contain the basic insights on how time delays and land taxes affect sustainability and biodiversity conservation.

Acknowledgements

This work was supported by the TULIP Laboratory of Excellence (ANR-10-LABX-41) and the Midi-Pyrénées Region. We thank Matthieu Barbier, David Shanafelt and François Salanié for valuable discussions.

144 CHAPTER 3

Bibliography

1. E.C. Ellis, J.O. Kaplan, D.Q. Fullet, S Vavrus, K.K. Goldewijk, and P.H. Verburg. Used planet: A global history. Proceedings of the National Academy of Sciences, 110(20):7978– 7985, 2013.

2. W B Meyer and B L Turner. Human Population Growth and Global Land-Use/Cover Change. Annual Review of Ecology and Systematics, 23(1):39–61, 1992. ISSN 0066-4162. doi: 10.1146/annurev.es.23.110192.000351.

3. EW Sanderson, M Jaiteh, MA Levy, KH Redford, AV Wannebo, and G Woolmer. The human footprint and the last of the wild. BioScience, 52(10):891–904, 2002.

4. O Venter, E.W. Sanderson, A. Magrach, J.R. Allan, J. Beher, K.R. Jones, H.P. Possingham, W.F. Laurance, P. Wood, B.M. Fekete, M.A. Levy, and J.E.M Watson. Sixteen years of change in the global terrestrial human footprint and implications for biodiversity conservation. Nature Communications, 7(12558), 2016.

5. G McGranahan, PJ Marcotullio, X Bai, D Balk, T Braga, and et al. Ecosystems and Human Well-being: Current Status and Trends., chapter Urban systems. Washington, DC: Island, 2005.

6. M C Hansen, P V Potapov, R Moore, M Hancher, S a Turubanova, a Tyukavina, D Thau, S V Stehman, S J Goetz, T R Loveland, a Kommareddy, a Egorov, L Chini, C O Justice, and J R G Townshend. High-resolution global maps of 21st-century forest cover change. Science, 342(6160):850–3, 2013.

7. T Newbold, L.N. Hudson, S.L.L. Hill, S Contu, I Lysenko, R.A. Senior, L Börger, D.J. Ben- nett, A Choimes, B Collen, J Day, A De Palma, S Dr’iaz, S Echeverria-Londono, M.J. Edgar, A Feldman, M Garon, M.L.K. Harrison, T Alhusseini, D.J. Ingram, Y Itescu, J Kattge, V Kemp, L Kirkpatrick, M Kleyer, D.L.P. Correia, C.D. Martin, S Meiri, M Novosolov, Y Pan, H.R.P. Phillips, D.W. Purves, A Robinson, J Simpson, S.L. Tuck, E Weiher, H.J. White, R.M. Ewers, G.M. Mace, J.P.W. Scharlemann, and A Purvis. Global eﬀects of land use on local terrestrial biodiversity. Nature, 520:45–50, 2015.

8. Jonathan A Foley, Ruth Defries, Gregory P Asner, Carol Barford, Gordon Bonan, Stephen R Carpenter, F Stuart Chapin, Michael T Coe, Gretchen C Daily, Holly K Gibbs, Joseph H Helkowski, Tracey Holloway, Erica A Howard, Christopher J Kucharik, Chad Monfreda, Jonathan A Patz, I Colin Prentice, Navin Ramankutty, and Peter K Snyder. Global consequences of land use. Science (New York, N.Y.), 309(5734):570–574, 2005.

9. Bradley J Cardinale, J Emmett Duﬀy, Andrew Gonzalez, David U Hooper, Charles Perrings, Patrick Venail, Anita Narwani, Georgina M Mace, David Tilman, David A Wardle, Ann P Kinzig, Gretchen C Daily, Michel Loreau, James B Grace, Anne Larigauderie, Diane S Srivastava, and Shahid Naeem. Biodiversity loss and its impact on humanity. Nature, 486 (7401):59–67, 2012. ISSN 00280836. doi: 10.1038/nature11148.

10. S Motesharrei, J Rivas, E Kalnay, G.R. Asrar, A.J. Busalacchi, R.F. Cahalan, M.A. Cane, R.R. Colwell, K Feng, R.S. Franklin, K Hubacek, F Miralles-Wilhelm, T Miyoshi, M Ruth, R Sagdeev, A Shirmohammadi, J Shukla, J Srebric, V.M. Yakovenko, and N Zeng. Mod- eling sustainability: population, inequality, consumption, and bidirectional coupling of the earth and human systems. National Science Review, 3(4):470–494, 2016. doi: 10.1093/n- sr/nww081.

145 CHAPTER 3

11. Tim Newbold, Lawrence N. Hudson, Andrew P. Arnell, Sara Contu, Adriana De Palma, Simon Ferrier, Samantha L. L. Hill, Andrew J. Hoskins, Igor Lysenko, Helen R. P. Phillips, Victoria J. Burton, Charlotte W. T. Chng, Susan Emerson, Di Gao, Gwilym Pask-Hale, Jon Hutton, Martin Jung, Katia Sanchez-Ortiz, Benno I. Simmons, Sarah Whitmee, Han- bin Zhang, Jörn P. W. Scharlemann, and Andy Purvis. Has land use pushed terrestrial biodiversity beyond the planetary boundary? A global assessment. Science, 353:288–291, 2016.

12. Robert Costanza, Rudolf de Groot, Paul Sutton, Sander van der Ploeg, Sharolyn J. Ander- son, Ida Kubiszewski, Stephen Farber, and R. Kerry Turner. Changes in the global value of ecosystem services. Global Environmental Change, 26(1):152–158, 2014.

13. G. Brundtland, M. Khalid, S. Agnelli, S. Al-Athel, B. Chidzero, L. Fadika, V. Hauﬀ, I. Lang, M. Shijun, M. Morino de Botero, M. Singh, S. Okita, et al. Our Common Future (The Brundtland Report). Oxford University Press, 1987.

14. T Dietz, E.A. Rosa, and R York. Driving the human ecological footprint. Frontiers in Ecology and the Environment, 5(1):13–18, 2007.

15. Eric F. Lambin and P. Meyfroidt. Global land use change, economic globalization, and the looming land scarcity. Proceedings of the National Academy of Sciences of the United States of America, 108(9):3465–72, 2011. ISSN 10916490. doi: 10.1073/pnas.1100480108.

16. B Phalan, Rhys E Green, Lynn V Dicks, G Dotta, C Feniuk, A Lamb, B.B.N. Strassburg, D.R. Williams, Erasmus K.H.J. zu Ermgassen, and A Balmford. How can higher-yield farming help to spare nature ? Science, 351(6272):450–451, 2016.

17. M Cropper and W Oates. Environmental economics: A survey. Journal of Economic Literature, 30(2):675–740, 1992.

18. Robert Costanza, John H Cumberland, Herman Daly, Robert Goodland, Richard B Nor- gaard, Ida Kubiszewski, and Carol Franco. An Introduction to Ecological Economics, Second Edition. CRC Press, 2014.

19. David Tilman, Robert M. May, Clarence L. Lehman, and Martin A. Nowak. Habitat destruction and the extinction debt. Nature, 371(6492):65–66, 1994.

20. Forest Isbell, David Tilman, Stephen Polasky, and Michel Loreau. The biodiversity- dependent ecosystem service debt. Ecology Letters, 18(2):119–134, 2015.

21. N. M. Haddad, L. a. Brudvig, J. Clobert, K. F. Davies, a. Gonzalez, R. D. Holt, T. E. Lovejoy, J. O. Sexton, M. P. Austin, C. D. Collins, W. M. Cook, E. I. Damschen, R. M. Ewers, B. L. Foster, C. N. Jenkins, a. J. King, W. F. Laurance, D. J. Levey, C. R. Mar- gules, B. a. Melbourne, a. O. Nicholls, J. L. Orrock, D.-X. Song, and J. R. Townshend. Habitat fragmentation and its lasting impact on Earth’s ecosystems. Science Advances, 1 (2):e1500052–e1500052, 2015. ISSN 2375-2548. doi: 10.1126/sciadv.1500052.

22. J. Pezzey. Sustainability: an interdisciplinary guide. Environmental values, 1:321–362, 1992.

23. R Costanza. Ecological Economics: the science and management of sustainability. Columbia University Press, New York, 1991.

24. R Costanza and H.E. Daly. Natural capital and sustainable development. Conservation Biology, 6(1):37–46, 1992.

146 CHAPTER 3

25. C Perrings. Ecological sustainability and environmental control. Structural Change and Economic Dynamics, 2(2):272–295, 1991.

26. John M Hartwick. The Economics and Ecology of Biodiversity Decline., chapter Decline in biodiversity and risk-adjusted net national product. Cambridge University Press, 1995.

27. A Endres and V Radke. Land use, biodiversity, and sustainability. Journal of Economics, 70(1):1–16, 1999.

28. A.-S. Lafuite and M Loreau. Time-delayed biodiversity feedbacks and the sustainability of social-ecological systems. Ecological Modelling, 351:96–108, 2017. doi: 10.1016/j.ecolmodel.2017.02.022.

29. P. L. Pingali. Green Revolution: Impacts, limits, and the path ahead. Proceedings of the National Academy of Sciences, 109(31):12302–12308, 2012.

30. A O’Connor, M.I.and Gonzalez, J.E.K Byrnes, B.J. Cardinale, J.E. Duﬀy, L. Gamfeldt, J.N. Griﬃn, D. Hooper, B.A. Hungate, A Paquette, P.L. Thompson, L.E. Dee, and K.L. Dolan. A general biodiversity-function relationship is mediated by trophic level. Oikos, 126 (1):18–31, 2017.

31. M Zeigler and A Steensland. Global Agricultural Productivity Report (GAP Report). Global Harvest Initiative, 2016.

32. O. Galor and D.N. Weil. Population, Technology, and Growth: From Malthusian Stagnation to the Demographic Transition and beyond. The American Economic Review, 90(4):806– 828, 2000.

33. T Kogel and A Prskawetz. Agricultural Productivity Growth and Escape from the Malthu- sian Trap. Journal of Economic Growth, 6:337–357, 2001.

34. John M. Anderies. Economic development, demographics, and renewable resources: a dynamical systems approach. Environment and Development Economics, 8(02):219–246, 2003.

35. P.F. Peretto and S Valente. Growth on a ﬁnite planet: resources, technology and population in the long run. Journal of Economic Growth, 20(3):305–331, 2015.

36. E.F. Connor and E.D. McCoy. The statistics and biology of the species-area relationship. American Naturalist, 113:119–130, 1979.

37. K.A. McGuiness. Equations and explanations in the study of species-area curves. Biological Reviews, 59:423–440, 1984.

38. David Storch, Petr Keil, and Walter Jetz. Universal species–area and endemics–area relationships at continental scales. Nature, 488(7409):78–81, 2012.

39. Joel Rybicki and Ilkka Hanski. Species-area relationships and extinctions caused by habitat loss and fragmentation. Ecology Letters, 16(SUPPL.1):27–38, 2013.

40. Ilkka Hanski and Otso Ovaskainen. Extinction debt at extinction threshold. Conservation Biology, 16(3):666–673, 2002.

41. O. R. Wearn, D. C. Reuman, and R. M. Ewers. Extinction Debt and Windows of Conser- vation Opportunity in the Brazilian Amazon. Science, 337(6091):228–232, 2012.

42. J.M. Diamond. Biogeographic kinetics: estimation of relaxation times for avifaunas of southwest paciﬁc islands. Proc.Nat.Acad.Sci. USA, 69(11):3199–3203, 1972.

147 CHAPTER 3

43. M.L. Rosenzweig. Species diversity in space and time. Cambridge University Press, 1995.

44. J. Shukla, C. Nobre, and P. Sellers. Amazon deforestation and climate change. Science, 247:1322–25, 1990.

45. L.S. Lima, M.T. Coe, B.S. Soares Filho, S.V. Cuadra, Livia C.P. Dias, Marcos H Costa, Leandro S Lia, and Hermann O Rodrigues. Feedbacks between deforestation, climate, and hydrology in the southwestern amazon: implications for the provision of ecosystem services. Landscape Ecology, 29(2):261–274, 2014.

46. R Costanza and C Perrings. A ﬂexible assurance bonding system for improved environmental management. Ecological Economics, 2(1):57–75, 1990.

47. Owen L. Petchey, Mikael Pontarp, Thomas M. Massie, Sonia Kéﬁ, Arpat Ozgul, Maja Weilenmann, Gian Marco Palamara, Florian Altermatt, Blake Matthews, Jonathan M. Levine, Dylan Z. Childs, Brian J. McGill, Michael E. Schaepman, Bernhard Schmid, Piet Spaak, Andrew P. Beckerman, Frank Pennekamp, and Ian S. Pearse. The ecological forecast horizon, and examples of its uses and determinants. Ecology Letters, 18(7):597–611, 2015.

48. Forest Isbell, Andrew Gonzalez, Michel Loreau, Jane Cowles, Sandra Díaz, Andy Hector, Georgina M. Mace, David A. Wardle, Mary I. O’Connor, J. Emmett Duﬀy, Lindsay A. Turn- bull, Patrick L. Thompson, and Anne Larigauderie. Linking the inﬂuence and dependence of people on biodiversity across scales. Nature, 546:65–72, 2017.

49. DL Carr and RE Bilsborrow. Population and land use/cover change: a regional comparison between central america and south america. J Geogr Educ, 43:7–16, 2001.

50. Richard Miller Bird and Naomi Enid Slack, editors. International Handbook of Land and Property Taxation. Edward Elgar, 2004.

51. J Skinner. Prospects for agricultural land taxation in developing countries. World Bank Economic Review, 5(3):493–511, 1991.

52. Wallace E Oates and Robert M Schwab. The impact of urban land taxation: the pittsburgh experience. National Tax Journal, 50(1):1–21, 1997.

53. B Needham. Land taxation, development charges, and the eﬀects on land-use. Journal of Property Research, 17(3):241–257, 2000.

54. R Pirard. Market-based instruments for biodiversity and ecosystem services: A lexicon. Environmental Science & Policy, 19-20:59–68, 2012.

55. J Pretty, C Brett, D Gee, R Hine, C Mason, J Morison, M Rayment, G Van Der Bijl, and T Dobbs. Policy challenges and priorities for internalizing the externalities of modern agriculture. Journal of Environmental Planning and Management, 44(2):263–283, 2001.

56. J Skinner. If agricultural land taxation is so eﬃcient, why is it so rarely used? World Bank Economic Review, 5(1):113–133, 1991.

57. S Lockie. Market instruments, ecosystem services, and property rights: Assumptions and conditions for sustained social and ecological beneﬁts. Land Use Policy, 13:90–98, 2013.

58. S Schomers and B Matzdorf. Payments for ecosystem services: A review and comparison of developing and industrialized countries. Ecosystem Services, 6:16–30, 2013.

148 CHAPTER 3

59. David Kleijn, Frank Berendse, Ruben Smit, and Niels Gilissen. Agri-environment schemes do not eﬀectively protect biodiversity in dutch agricultural landscapes. Nature, 413:723–725, 2001.

60. D Kleijn and W.J. Sutherland. How eﬀective are european agri-environment schemes in conserving and promoting biodiversity? Journal of Applied Ecology, 40(6):947–969, 2003.

61. K Henle, D Alard, J Clitherow, P Cobb, L Firbank, T Kull, D McCracken, R.F.A. Moritz, J Niemela, M Rebane, and D Wascher. Identifying and managing the conﬂicts between agriculture and biodiversity conservation in europe–a review. Agriculture, Ecosystems & Environment, 124(1-2):60–71, 2008.

62. E.H.A. Mattison and K Norris. Bridging the gaps between agricultural policy, land-use and biodiversity. Trends in Ecology and Evolution, 20(11):610–616, 2005.

63. P Balvanera, G C Daily, P R Ehrlich, T H Ricketts, S-A Baileys, S Kark, C Kremen, and H Pereira. Conserving biodiversity and ecosystem services. Science, 291(5511):2047, 2001.

64. S Macfadyen, S A Cunningham, A C Costamagna, and N A Schellhorn. Managing ecosystem services and biodiversity conservation in agricultural landscapes: are the solutions the same? Journal of Applied Ecology, 49(3):690–694, 2012.

65. Z Austin, A McVittie, D McCracken, A Moxey, D Moran, and P C L White. The co-beneﬁts of biodiversity conservation programmes on wider ecosystem services. Ecosystem Services, 20:37–43, 2016.

66. G Daily. Ecosystem services: From theory to implementation. Proceedings of the National Academy of Sciences, 105(28):9455–9456, 2008.

67. G M Mace, K Norris, and A H Fitter. Biodiversity and ecosystem services: a multilayered relationship. Trends in Ecology & Evolution, 27(1):19–26, 2012.

68. T M Brooks, R A Mittermeier, G A B da Fonseca, J Gerlach, M Hoﬀmann, J F Lamoreux, C G Mittermeier, J D Pilgrim, and A S L Rodrigues. Global biodiversity conservation priorities. Science, 313(5783):58–61, 2006.

69. B Reyers, S Polasky, H Tallis, H A Mooney, and A Larigauderie. Finding common ground for biodiversity and ecosystem services. BioScience, 62(5):503–507, 2012.

70. K Norris. Biodiversity in the context of ecosystem services: the applied need for systems approaches. Philos Trans R Soc Lond B Biol Sci., 367(1586):191–199, 2012.

71. M.G.E. Mitchell, E.M. Bennett, and A. Gonzalez. Strong and non-linear eﬀects of fragmentation on ecosystem service provision at multiple scales. Environmental Research Letters, 10(9), 2015.

72. Erik Nelson, Guillermo Mendoza, James Regetz, Stephen Polasky, Heather Tallis, DRichard Cameron, Kai MA Chan, Gretchen C Daily, Joshua Goldstein, Peter M Kareiva, Eric Lons- dorf, Robin Naidoo, Taylor H Ricketts, and MRebecca Shaw. Modeling multiple ecosystem services, biodiversity conservation, commodity production, and tradeoﬀs at landscape scales. Modeling multiple ecosystem services, biodiversity conservation, commodity production, and tradeoﬀs at landscape scales, 7(1):4–11, 2009.

73. J Carmichael, J Tansey, and J Robinson. An integrated assessment modeling tool. Global Environmental Change, 14(2):171–183, 2004.

149 CHAPTER 3

74. M Hejazi, J Edmonds, L Clarke, P Kyle, E Davies, V Chaturvedi, M Wise, P Patel, J Eom, K Calvin, R Moss, and S Kim. Long-term global water projections using six socioeconomic scenarios in an integrated assessment modeling framework. Technological Forecasting and Social Change, 81:205–226, 2014.

75. W.D. Nordhaus. Optimal Greenhouse-gas Reductions and Tax Policy in the "DICE" model. The American Economic Review, 83(2):313–317, 1993.

150 Appendix

151 CHAPTER 3

3.A General Market Equilibrium

Consumer Optimization

The number of agents is captured by the population variable, H. Overall, each agent supplies one unit of labor, and his revenue is the wage w. A tax τ per unit of converted land A is implemented by a social planner. Let U(y1, y2) be the individual consumer utility, where y1 and y2 are per capita consumption rates of agricultural and industrial

η 1−η goods. Agents are assumed to maximize their utility U(y1, y2) = y1 y2 , where η is the preference for agricultural goods. After redistribution of the total revenue of the tax τA among the agents, the revenue constraint writes p1y1 + p2y2 ≤ w + τA/H, where p1 and p2 are the prices of agricultural and industrial goods respectively.

To solve this maximization problem, we deﬁne the Lagrangian:

L ≡ U(y1, y2) − Λ[p1y1 + p2y2 − w − τA/H]

First order conditions are:

  ηU(y1,y2) ∂L/∂y1 = y − Λp1 = 0 1 (19)  (1−η)U(y1,y2) ∂L/∂y2 = − Λp2 = 0  y2

Adding both conditions yields U(y1, y2)/Λ = p1y1 + p2y2 = w + τA/H, and solving (19) for y1 and y2, and substituting for U(y1, y2)/Λ yields the per capita demands:

τA! τA! p yD = η w + p yD = (1 − η) w + (20) 1 1 H 2 2 H

Thus, the aggregate demand for agricultural and industrial goods writes:

152 CHAPTER 3

τA! τA! p Y D = η w + H p Y D = (1 − η) w + H (21) 1 1 H 2 2 H

Firms Optimization

Firms in sectors 1 and 2 produce a quantity Yi (i = 1, 2) of output, using labor Li and land Ai. A units of land and H units of labor are available. Agricultural ﬁrms occupy

A1 units of land, and industrial firms occupy A2 units of land, such that A = A1 + A2. This comes at an operating cost of κ units of labor per unit of land, so that each firm maximizes a profit Πi = piYi − wLi − (κw + τ)Ai. The production functions Yi are:

Ω α1 1−α1 α2 1−α2 Y1 = T B L1 A1 Y2 = T L2 A2 (22) where BΩ is a concave-down function capturing the ecological feedback of biodiversity- dependent ecosystem services on agricultural production, with Ω < 1, T captures the agricultural total factor productivity, i.e. production eﬃciency, and αi is labor intensity in sector i. Therefore, ﬁrst order conditions are:

  αipiYi ∂Πi/∂Li = L − w = 0 i (23)  (1−αi)piYi ∂Πi/∂Ai = − κw − τ = 0  Ai

Adding both lines of (23) gives the total supply in sector i:

S piYi = wLi + (κw + τ)Ai (24)

Dividing both lines of (23) gives the relationship between input factors in each sector:

Li/αi = (κ + τ)Ai/(1 − αi) (25)

153 CHAPTER 3

Economic General Equilibrium

To close the system, let us choose the wage w as the numeraire and set it equal to 1.

D S Markets’ clearing for the agricultural and industrial sectors piYi = piYi (eq. (21) and (24)) yields:

  τA η 1 + H H = L1 + (κ + τ)A1 (26)   τA (1 − η) 1 + H H = L2 + (κ + τ)A2

Dividing both lines of eq.(26) gives:

η L + (κ + τ)A = 1 1 (27) 1 − η L2 + (κ + τ)A2

Using eq. (25) to replace L1 and L2 in eq.(27), and noting that A1 + A2 = A, we derive land allocations at market equilibrium, as functions of the area of converted land A:

η(1 − α1) (1 − η)(1 − α2) A1 = A A2 = A (28) 1 − α1η − α2(1 − η) 1 − α1η − α2(1 − η)

Using eq.(25), the labor allocations at the market equilibrium are:

ηα1 (1 − η)α2 L1 = (κ + τ) A L2 = (κ + τ) A (29) 1 − α1η − α2(1 − η) 1 − α1η − α2(1 − η)

Land Conversion

Adding both lines of (26) gives the labor market equilibrium L1 +L2 +κA = H. Replacing

L1 and L2 by their optimal allocations, we deduce the relationship between the human population size H and the converted land A:

A = H/φτ (30)

154 CHAPTER 3

where the density of the human population on converted land φτ is:

φ − κ! φ = φ + τ (31) τ κ and φ is the population density in the absence of regulation:

κ φ = (32) 1 − α1η − α2(1 − η)

Note that φ > κ since αi ∈ [0; 1] and η ∈ [0; 1], so that φτ > φ. A tax τ per unit of converted land thus increases the human population density on converted land.

Per capita consumptions

Using eq.(30) so as to replace A in the equilibrium allocations of labor Li (eq.(29)) and land Ai (eq.(28)), we now obtain the inputs’ allocations as functions of the dynamical variable H:

φ κ + τ φ κ + τ L1 = α1ηH L2 = α2(1 − η)H (33) φτ κ φτ κ and

φ (1 − α1)η φ (1 − α2)(1 − η) A1 = H A2 = H (34) φτ κ φτ κ

The per capita agricultural and industrial consumptions Yi/H (eq.(22)) can then be rewritten as functions of the dynamical variables of the system B and T as follows:

φ κ + τ α1 φ κ + τ α2 y1τ (B, T) = y1(B, T) y2τ (T) = y2(T) (35) φτ κ φτ κ where the per capita consumptions in the absence of regulation are:

1−α1 1−α2 1 − α1 1 − α2 y (B, T) = ηαα1 BΩT y (T) = (1 − η)αα2 T (36) 1 1 κ 2 2 κ

Since φ < φτ , the eﬀect of a tax τ on the per capita consumptions is not straightforward

155 CHAPTER 3

and will depend on the parameters of the system, and especially on labor intensity αi. In systems with low labor intensities, a tax on converted land will rather reduce agricultural consumption and population growth, while it could have the opposite eﬀect on systems with higher labor intensities.

156 GENERAL DISCUSSION

157 GENERAL DISCUSSION

Synthesis of the results

In the first chapter of this thesis (Chapter 1), a social-ecological system model including the feedback of biodiversity loss on human demography is described. This simple model is used to perform abstract simulations and qualitatively assess the effects of parameters on the long-term dynamics of the system, and especially the potential effects of time delays due to extinction debts. It retains the basic structure of a predator-prey model, where human population growth indirectly depletes biodiversity through the destruction of natural habitats. The feedback loop closes through the effect of biodiversity-dependent ecosystem services loss on agricultural production. Human population consumption and land conversion are derived from an economic general equilibrium model, assuming an instantaneous equilibrium between demand and supply of consumption goods. The basic model thus allows to jointly simulate the temporal dynamics of biodiversity loss and human population growth, constrained by economic rationality and demographic processes such as the demographic transition.

This conceptual model allows exploring the long-term consequences of time-delayed biodiversity loss on the dynamics of a human population. The model reproduces classical cyclic dynamics of predator-prey models, when there is a time delay between the dynamics of the predator and that of the prey [98]. Growth of the predator population decimates the slowly regenerating prey, thus resulting in a food shortage and the degrowth of the predator population, until the prey population has regrown large enough for the predator to grow again. A similar mechanism has been invoked to explain the collapse of the human population on Easter Island, where the main variety of palm tree was characterized by a low regeneration rate compared to the varieties found on other Paciﬁc Islands [13]. In a coupled SES, such oscillations result in a decrease in human well-being, and are thus incompatible with sustainability requirements [16]. Given mounting evidence of accumulating extinction and functioning debts, this eﬀect of time delays that generates temporal mismatches between human and ecological timescales should be better recognized and accounted for.

158 GENERAL DISCUSSION

One advantage of the model’s simplicity lies in the clear understanding of the eﬀects of the ecological, economic and technological parameters on both transient dynamics and equilibrium features. Analytical results allow identifying parameters’ space that make the system more vulnerable to such overshoot-and-collapse crises, as well as thresholds in the dynamical variables beyond which the system starts to oscillate. Due to the coupling of the dynamical variables within the SES model, each threshold is related to the others, so that crossing one of them leads to crossing the others, in a “cascading eﬀect” [54]. Improving knowledge about such thresholds is essential to reduce uncertainty and inform biodiversity conservation, land-use management and demographic policies in an integrative way.

Technological change makes our model SES more vulnerable to crises, when it increases the efficiency of agricultural production. This destabilizing effect of technological change has already been emphasized in typical Easter Island systems, regarding resource harvesting efficiency, carrying capacity and intrinsic growth rate [96]. In our system, a higher production efficiency is equivalent to a higher substitution between human capital and biodiversity-dependent services. This negative effect of technological efficiency thus suggests that conventional technology-intensive agriculture may threaten production systems with overshoot-and-collapse crises in the long run. A higher decoupling from natural constraints thus appears detrimental to sustainability, despite the classical view of growth economists on this topic [11].

Our results thus contrast with the commonly accepted view that technological change will automatically solve humanity’s environmental sustainability problems [109]. This view is based on several misunderstandings about the effects of technological change on resource-use. “Efficiency technologies” can increase resource-use efficiency, as in our model, while “extraction technologies” and “consumption technologies”, raise the scale of resource extraction and per capita resource consumption. Our results show that the benefits of efficiency technologies can be compensated for by increases in consumption and land conversion, known as rebound effects, and leading to long-term negative impacts on sustainability. Similar observations show that other types of technological changes can also be detrimental, and empirical records show that the net effect of technological advances

159 GENERAL DISCUSSION has been a continued increase in global per capita resource-use, waste generation, and emissions [71].

Conversely to technological efficiency, labor elasticity reduces the vulnerability of the SES to crises, which is consistent with the effect of technological efficiency, since a higher labor elasticity can be interpreted as a lower reliance on technology, i.e. a lower substitution of technology for natural services and labor. There is mounting evidence that recou- pling agricultural systems and natural ecosystem services through ecologically-friendly practices is key to meet future food demand through the sustainable intensification of agricultural production [32, 36]. Our results show that this is especially important with regard to long-term ecological dynamics and pending extinction debts [118, 5, 99].

Our results also emphasize the role of the demographic transition in alleviating transient crises. By slowing down the growth of the human population, the demographic transition reduces the delay between the dynamics of the human population and biodiversity loss, and hence the vulnerability to overshoot-and-collapse crises. However, the observed rapidity and strenght of the demographic transition may not allow counteracting time-delayed feedbacks, since population projections cast doubts upon the efficiency of the demographic transition in halting human population growth this century [40, 12]. Interaction between the demographic transition and technological progress results in a non-linear effect of technology on sustainability, since a very high technological efficiency increases industrial consumption and strengthens the demographic transition to the point where the net effect of technological change switches from negative to positive. However, this result is highly dependent upon the assumption that there is no biodiversity feedback on the industrial sector. Deforestation and biodiversity loss can affect regional and global climate [107] with feedbacks on hydrology [59] and other important provisioning services to the industrial sector, such as wood production and clean water. Including a feedback of biodiversity loss on the industrial sector would thus make technological change less likely to have a net positive effect on sustainability.

Adding social interactions to the basic structure presented in Chapter 1 enriches the dynamics of the model, by allowing for alternative stable states to coexist, and for regime

160 GENERAL DISCUSSION shifts between sustainable and unsustainable basins of attraction to occur (Chapter 2). We consider social mechanisms that foster norm-driven consumption changes, such as disapproval or ostracism, and a consumption norm corresponding to sustainable agricultural practices, as identified in Chapter 1. The demand of norm-following consumers for such sustainable products drives changes in agricultural practices, as a result of market equilibrium between supply and demand. A high enough efficiency of the social pressure and a large enough proportion of norm-following consumers in the population lead to large-scale shifts towards more sustainable consumption and production practices. Such shifts greatly reduce the vulnerability of SESs to unsustainable crises, thus confirming the role of human behavioral change in the adaptive capacity of SESs [37] and in the pre- vention of social-ecological traps [19]. However, short term perception of environmental degradation and uncertainty over time-delayed ecological feedbacks can sap the efficiency of human behavioral change in enforcing sustainability. Increasing the time delay between human and ecological dynamics reduces the resilience of the sustainable equilibrium in favor of the resilience of the unsustainable equilibrium, thus favoring abrupt regime shifts towards the unsustainable basin of attraction. Foresighted social change is thus required to counteract large time-delayed feedbacks and prevent social-ecological traps.

Not only can bottom-up social change, arising from the aggregation of individual behaviors, help enforce sustainability at large scales [106, 81, 119, 104, 79], top-down government control is also essential to prevent the overexploitation of natural resources [46], internalize ecosystem services into economic decisions [94, 91, 105] and enforce property rights [80]. Especially, foresighted government control can help counteract the negative consequences resulting from shortsighted economic decisions. In our model, foresighted land taxation alleviates the negative consequences of biodiversity loss in the long run, by fostering the substitution of labor, land and technology for biodiversity-dependent ecosystem services, thus preserving more natural habitats and biodiversity than in the business-as-usual scenario (chapter 3). Our results contrast with Easter Island’s studies [13] suggesting that an optimal resource management with inﬁnite horizon would not have prevented this historical overshoot-and-collapse population crisis, resulting from the inter-

161 GENERAL DISCUSSION action between a fast growing human population and a slowly regenerating resource [41]. The present results thus reinforce recent evidence about the need to develop incentive- based mechanisms for the simultaneous conservation of biodiversity and ecosystem services in order to foster sustainability [4, 115, 61, 78, 97, 3], and highlight the eﬃciency of such mechanisms to counteract time-delayed ecological feedbacks, especially in the absence of precise knowledege about the temporal dynamics of these feedbacks. In the following, some perspectives of this work in social-ecological systems modelling and sustainability research are identiﬁed and discussed.

162 GENERAL DISCUSSION

Perspectives

Overall, our model has a structure upon which future improvements can be built. In the following, we identify potential improvements to the formalization of the main relationships of the model, including the relationship between (1) human consumption and demography, (2) population growth and land conversion, (3) land conversion and biodiversity, and (4) habitat fragmentation and ecosystem services.

Human consumption and demography

Even though various types of technological change did not qualitatively affect our results (chapter 1, Appendix), endogenizing technological change along with economic growth and fertility choices appears necessary in a predictive perspective. In similar models, capital accumulation [1] and economic stratification [72] have destabilizing effects, i.e. favor population cycles and collapses. Conversely, human foresight regarding natural resources’ exploitation has a stabilizing effect, although discounting may reduce this effect [41, 24]. Endogenous income and fertility dynamics can be destabilizing in case of a low substitutability between labor and natural resources, driving the economy towards either demographic explosion or collapse [84]. Assuming general equilibrium for the economic sector has major flaws, although most economic models make this assumption. A possible extension of this work would be to develop a non-equilibrium economic sector that does not necessarily converge to an optimal equilibrium. These results highlight the importance of accounting for realistic economic representation when modelling human-nature systems. Accounting for recent theoretical and empirical evidence regarding changing fertility choices and death rates with economic development and accumulation of human capital, such as the child quantity-quality trade-off during the demographic transition [33], or increased death rates due to resource scarcity, would also allow for a more realistic modelling of fertility choices.

More realism can also be added to the relationships between human demography, consumption, and biodiversity. First, in our model, the biodiversity feedback on human

163 GENERAL DISCUSSION population growth only occurs through the level of agricultural consumption. Biodiver- sity also provides other services than agricultural production, such as carbon sequestration and climate regulation, cultural, recreational and spiritual services [17], all of which affecting human well-being, mortality and fertility choices. Second, accounting for the effect of consumption quantity and quality, on birth and death rates separately, could enrich the behavior of the model [1]. Indeed, standard quantitative consumption metrics are inadequate to capture the nutritional value of agricultural goods, especially regarding the erosion of essential dietary nutrients due to the widespread selection of high-yielding and nutrient-poor cereals, such as wheat, rice, and maize [27]. Nutrient-poor diets, which are higher in refined sugars, refined fats, oils and meats, are detrimental to human health, since they increase incidence of type II diabetes, coronary heart disease and other chronic non-communicable diseases that lower global life expectancies [112].

Thus, agricultural intensification may not allow meeting the nutritional requirements of the global population, and may generate significant feedbacks on human mortality. Furthermore, alternative diets that offer substantial health benefits, such as pescetar- ian, mediterranean and vegetarian diets, could help reduce land conversion and resultant species extinctions [112]. Such healthy diets include a larger part of fresh fruits and vegeta- bles, whose production benefits from biodiversity-dependent services, such as pollination. Less intensive agricultural practices relying more on biodiversity-dependent services than on toxic chemicals, could thus benefit both human health and biodiversity. These complex feedbacks between agricultural practices, biodiversity conservation, food quality and human health may affect the behavior of our conceptual model, and need to be accounted for in order to produce reliable forecasts.

Population growth and land conversion

Drivers of land use and land cover change are abundant, complex and scale-dependent [20]. At global scale, recent evidence that human population and the world economy are growing faster than the human footprint suggests a globally more eﬃcient use of land [117]. However, economic mechanisms such as the displacement, rebound, cascade, and

164 GENERAL DISCUSSION remittance effects that are amplified by economic globalization can accelerate land conversion despite land intensification [57]. Endogenizing efficiency gains and technological change along with economic growth thus appears essential in order to gain realism and predictive power.

At local scales, the effect of human population growth can be ambiguous [64, 39], and population dynamics usually acts in concert with other significant factors such as local institutions, policies and cultural change [58, 25]. Additionally, economic globalization and rising connectivity between distant SESs can lead to the local specialization in production that drives specialization in ecosystems and their associated biodiversity, leading to significant declines in both local and global biodiversity, especially when endemism is high, i.e. when trading partners contain dissimilar species [92].

In the same vein, human migrations and spatial movements are steadily increasing across regions of the world, with consequences for local labour markets and land-use patterns [75]. Thus, accounting for the spatial human and economic cross-scale eﬀects appears crucial to adequately capture the relationship between human population growth and land conversion.

Land conversion and biodiversity

Another important improvement relates to the spatialization of the ecological compartment of the model. Indeed, habitat loss and fragmentation changes the connectivity and size of fragments of natural habitat, thus affecting the spatial dynamics of populations and making species cross their extinction threshold. Such thresholds depend both on species traits, such as dispersal and colonization, and on the spatial configuration of habitat patches [45]. As a result, habitat fragmentation significantly amplifies the negative effect of habitat loss on species richness, so that accounting for this effect increases the predictive power of species-area curves [100, 44].

Fragmentation also has a non-linear eﬀect on the relaxation rate of populations [82]. Habitat fragmentation slows down the transient dynamics of viable populations, i.e. populations that are above their extinction threshold, and accelerates the dynamics of doomed

165 GENERAL DISCUSSION ones, i.e. those that are below their extinction threshold [45]. As a result, species extinctions are expected to occur faster on smaller habitat fragments [34]. Thus, the relaxation rate of communities is not a constant parameter, but endogenously varies with the spatial conﬁguration of habitat fragments.

Lastly, our model assumption that recovery debts are equivalent to extinction debts is unrealistic. First, land degradation can lead to irreversibility of the land conversion process. Soil erosion enhances nutrient loss, which reduces the fertility of soils [89], with important economic consequences [90]. Each year about 10 million ha of cropland are lost due to soil erosion, 10 to 40 times faster than the rate of soil renewal [88]. Second, when recovery is possible, recovering and restored ecosystems have lower species abundances and diversity, and less cycling of carbon and nitrogen than ‘undisturbed’ ecosystems, due to the accumulation of a recovery debt [70]. Such recovery debts can be much larger than extinction debts, especially due to a time-delayed recolonization e.g. 130-230 years after reforestation in Germany [74]. Recolonization not only depends on the dispersal abilities of species between distant habitat patches, but also on the state of biodiversity at other spatial scales, highlighting the importance of a cross-scale spatial perspective [43].

Adoption of a spatially-explicit approach appears crucial in order to account for the complex relationships between habitat loss and biodiversity in a more realistic way.

Habitat fragmentation and ecosystem services

Spatialization is also a crucial aspect of the provisioning of ecosystem services in real- world landscapes [114]. First, landscapes have long been represented as split between natural habitats supporting biodiversity on one side, and inhospitable human-dominated lands on the other [42, 86]. However, real-world landscapes are a mix of natural, semi- natural and productive lands, with bidirectional interactions between them. Productive land hosts species that are new to the landscape and may propagate to proximate patches of natural habitats, while agricultural chemicals impact wild species and nutrient ﬂows. Besides, natural patches provide multiple ecosystem services to surrounding ﬁelds, such as crop production, pest regulation, decomposition, carbon storage, soil fertility, water

166 GENERAL DISCUSSION quality regulation [66] and pollination [63], but also dis-services such as pest damage and competition for water or pollination between natural habitats and crops [120].

Fragmentation generally has negative effects on ecosystem service supply, through its negative effect on species richness. However, fragmentation can have positive or negative effects on service flow, through an increased proximity between crops and the natural habitat patches providing services [67]. Recent advances on connectivity-ecosystem service provision relationships [68] highlight the potential for highly non-linear effects of fragmentation on service provisioning at local scales [69]. Another cause of non-linearity is the crossing of extinction thresholds beyond a critical amount of habitat loss, that can lead to the collapse of a service in agroecosystems, such as pollination [53]. Such non- linearities can result in cascading effects when crossing a local threshold results in the crossing of other thresholds at larger spatial scales [54].

Second, spatial aspects can modify the positive relationship between biodiversity and services reported by most experimental studies. For example, pest control usually increases with landscape complexity, which increases both species diversity and the proximity between crops and natural habitats patches hosting natural enemies [9, 38]. However, this effect can be ambiguous [111]. Natural enemies usually have a strong positive response to landscape complexity, but this effect does not necessarily translate into pest control, since pest abundances do not always show significant response to landscape complexity [21]. A possible explanation may be that the relationship between pest control and landscape complexity is also driven by changes in negative natural enemy interactions, such as predation from birds. Thus, by altering natural enemy interactions, landscape complexity can provide ecosystem services as well as disservices [62].

In a spatially-explicit context, the relationships between biodiversity, ecosystem services supply and ecosystem services flow are not necessarily linear. Accounting for landscape spatial configuration, distance-dependence in service provisioning, trade offs among multiple ecosystem services, non-linearities and cross-scale interactions thus appears essential in order to adequately capture the complexity of real-world agroecosystems, in relation to the sustainability of SESs relying on them.

167 GENERAL DISCUSSION

Towards predictive social-ecological models

Human demography is a major taboo in public opinion, as well as in studies of the sustainability of modern human societies [24]. Several roots of this taboo include the rise of individualism, deep cultural values surrounding fertility and the idea that progress and ’Providence’ will provide for all of our needs [47], which lead to a contradiction with the necessity of women’s access to family planning and healthcare as a universal right for women. Ecology thus threatens fundamental and almost religious values [48].

A great part of the uncertainty in our current understanding and projections of human population growth, and hence of the sustainability of human-nature interactions, lies in biodiversity feedbacks on human demography [71]. Incomplete knowledge of the temporal dynamics of biodiversity loss and of the relationships between biodiversity and essential ecosystem services has impaired realistic representation of some processes, which are however critical for a full understanding of these feedbacks. Hence, despite observations of time delays in species extinctions and predictions of increasing species loss and consequences for ecosystem functioning, the feedback of biodiversity on human societies is still represented in a crude way [76]. Surprisingly, the consequences of land use changes on biodiversity appear less studied than those of climate change [113], despite the major threats of future human population growth for natural habitat conversion and massive biodiversity loss driven by land-use changes.

It is critically needed to develop predictive models that include the feedback of biodiversity loss on human demography, especially regarding the potential consequences of time delays highlighted in this thesis. Extinction debts pose a signiﬁcant but often un- recognized challenge for biodiversity conservation [55, 31], while constituting windows of conservation opportunities such as habitat restoration and landscape management [118]. Reducing uncertainty regarding the role and importance of biodiversity is crucial to inform policy-makers, and halt the current extinction crisis. Recent advances in ecological research now allow making large-scale predictions up to continental or global scales, ranging from the future distribution of biological diversity to changes in ecosystem functioning

168 GENERAL DISCUSSION and services [85, 52]. Next-generation social-ecological models [93, 35] should thus better integrate both human demography and detailed biodiversity feedbacks, along with uncertainty and collective action [2], in order to improve their predictive power in a changing world [73].

Moreover, the development of predictive social-ecological system models appears crucial given the mounting evidence for social-ecological regime shifts [56, 101, 6]. Increasing risk of such abrupt regime shifts towards often less desirable and more resilient states has fostered the development of various modeling approaches, such as statistical, system dynamics, equilibrium and agent-based modelling. However, these approaches are largely disjunct, and do not necessarily consider the criteria that are empirically studied, thus making testing and application hard [35]. Yet some of these approaches provide useful insights into the management of SESs at risk of regime shifts [23].

Accounting for the feedback of biodiversity on human demography requires transdisci- plinary collaboration involving ecological researchers, economists, social and climate scientists, and demographers. The Integrated Valuation of Ecosystem Services and Tradeoffs (InVEST) framework [76] models multiple ecosystem services, biodiversity conservation, commodity production, and tradeoffs at landscape scales, but poorly accounts for the relationships between biodiversity and services, and ignores human demography. Other global Integrated Assessment Modeling frameworks have been developed to specifically address climate change [18] or water availability [49] issues. For example, the Dynamic Integrated Climate-Economy (DICE) model has been extensively used to inform optimal tax policies aimed at reducing greenhouse-gas emissions [77]. Many Integrated Assess- ment Models (IAMs) have been developed to describe the key processes in the interaction of human development and the natural environment. Biodiversity should be included in existing IAMs, as a dynamic variable coupled to other sectors with feedbacks. Indeed, the Global Change Assessment Model (GCAM) couples the economy, energy sector, land use and water with a climate model, but ignores biodiversity feedbacks [30]; the Integrated Global System Modeling (IGSM) framework simulates the evolution of economic, demographic, trade and technological processes, and the resulting greenhouse gas emissions,

169 GENERAL DISCUSSION conventional air and water pollutants, and land-use/land-cover change, through coupled sub-models of physical, dynamical and chemical processes [108, 95]; the Model for En- ergy Supply Strategy Alternatives and their General Environmental Impact (MESSAGE) focuses on the global economy and its main sectors (energy, agriculture, forestry), thus neglecting biodiversity feedbacks [65]; the Integrated Model to Assess the Global Envi- ronment (IMAGE) represents interactions between society, the biosphere and the climate system, but considers population, economy, policy and technology as external drivers [110]; Modelling International Relationships in Applied General Equilibrium (MIRAGE) is a multi-sectoral and multi-regional computable general equilibrium model dedicated to trade policy analysis [26]. Future research should build upon the strengths of each of these IAMs in order to fully account for the complexity of coupled social-ecological systems.

Population ethics and inequality

A major difficulty in defining integrative biodiversity-food security policies lies in their ethical consequences. Indeed, a policy seeking to maximize total agricultural production may lead to favor large human populations with a low average well-being [83]. This para- dox, known as the “repugnant conclusion”, has been extensively studied in population ethics. Solutions to this problem include defining a minimal well-being as a constraint to the maximization problem [10]. Equivalently, in our model we set a minimum consumption level, below which human population starts declining, thus preventing “repugnant” population equilibria in the long run. Yet such a threshold remains arbitrary and may still pose ethical problems in real-world situations.

Another ethical difficulty lies in defining sustainable policies that also guarantee equity and poverty eradication [50]. Income inequality among members of a society, leading to economic stratification, causes social and economic instability [87] and has played a central role in the collapse of past civilizations [29]. Unequal distribution of wealth has grown extremely fast since about 1950, along with population and consumption [15, 14]. The average per capita material and energy use in developed countries is higher than in developing countries by a factor of 5 to 10, and about 50% of the world’s people live on less

170 GENERAL DISCUSSION than $3 per day, 75% on less than $8.50, and 90% on less than $23 [71]. In agreement with the historical record, models have shown that economic stratiﬁcation can lead to collapse, whereas collapse can be avoided if resources are distributed equitably [72]. Piketty [87] proposes a global system of progressive wealth taxes to help reduce inequality and avoid the vast majority of wealth coming under the control of a tiny minority.

On uncertainty and risk in a changing world

The complexity and multiplicity of feedbacks that drive the dynamics of SESs make their study and understanding challenging, and the associated uncertainty can undermine decision-making and postpone management shifts. Thus, recognition of the insurance value of biodiversity is crucial to improve decision-making under uncertainty, and foster the adoption of a precautionary approach to this problem. Indeed, in addition to its direct and indirect use or non-use values [17], biodiversity also has an insurance value, since productivity and stability increase with biodiversity under variable environmental conditions [51] and in a spatially heterogeneous environment [60]. This natural insurance value can be seen as a substitute for ﬁnancial insurance [7].

Ecosystem resilience has also been interpreted as an economic insurance to changes in the provisioning of ecosystem services, the value of which increases with the level of resilience of the system [8]. However, ecosystem resilience does not imply sustainability, since resilience and sustainability are independent concepts [28]. Thus, more criteria than just resilience have to be taken into account when designing policies for the sustainable development of social-ecological systems, and resilience should not be confused with sustainability. Yet the concept of insurance value of both biodiversity and resilience can be useful to foster decision-making under uncertainty. Several precautionary approaches to environmental uncertainty have been proposed, such as ﬂexible environmental assurance bonding systems [22].

Another challenge to decision-making lies in our perception of environmental changes, taking place over spatial and temporal scales far beyond the scales which we directly per- ceive. Consequences of global changes, such as climate change, are diﬃcult to apprehend

171 GENERAL DISCUSSION and understand without the help of the scientific community. This is even more relevant for time-delayed biodiversity erosion, since it often does not directly affect the individual responsible for its erosion. Such a disconnection between the time and location of an action and its consequences, and the resultant decoupling between human and nature dynamics, is detrimental to the effective regulation of such “tragedy of the commons” situations [46]. However, even if a problem is recognized by experts, the time lag before any societal change in opinion can be significant, due to inertia and hysteresis processes, and especially in presence of a weak central decision-making authority or influential lobbys. Such situations cause abrupt but late shifts in opinion [103]. When consequences of current actions are delayed in time, as is the case with biodiversity loss, opinion shifts may be postponed even more, and come too late to prevent catastrophic outcomes. The consequences of time delays highlighted in this thesis should thus be better recognized and accounted for.

172 GENERAL DISCUSSION

General conclusion

Minimal conceptual models are eﬃcient tools to test hypotheses, develop theory and direct empirical research, and their limited number of parameters allows complete exploration of their behavior [102]. However, many essential aspects are usually neglected in simple models, leading to a trade-oﬀ between generality and predictive ability that makes testing and validation very hard. As a consequence, the popularity of abstract models remains marginal among policy makers and ecologists [102]. Conceptual models are, however, an essential step towards understanding the dynamics of larger complex models used to provide quantitative assessments and make predictions [116]. Despite quantitative limitations, minimal conceptual models can provide useful qualitative predictions. For example, despite uncertainty regarding the precise temporal dynamics and strength of biodiversity feedbacks, there is no doubt that time delays increase the vulnerability of human societies to overshoot natural resources. Such quantitative uncertainties should not prevent informed actions from halting current biodiversity loss and natural habitat destruction. A precautionary and farsighted approach is required to foster the necessary shifts in opinion and decision-making, which should be implemented across scales [80] and account for cross-scale interactions [43, 52], be they temporal or spatial. Both modelling exercise and experimental approaches feed knowledge and enlighten choices regarding such a complex adaptive system that represent human-nature interactions.

173 GENERAL DISCUSSION

Bibliography

1. John M. Anderies. Economic development, demographics, and renewable resources: a dynamical systems approach. Environment and Development Economics, 8(02):219–246, 2003.

2. John M Anderies. Understanding the dynamics of sustainable social-ecological systems: Human behavior, institutions, and regulatory feedback networks. Bulletin of Mathematical Biology, 77(2):259–280, 2014.

3. Z Austin, A McVittie, D McCracken, A Moxey, D Moran, and P C L White. The co- beneﬁts of biodiversity conservation programmes on wider ecosystem services. Ecosystem Services, 20:37–43, 2016.

4. P Balvanera, G C Daily, P R Ehrlich, T H Ricketts, S-A Baileys, S Kark, C Kremen, and H Pereira. Conserving biodiversity and ecosystem services. Science, 291(5511):2047, 2001.

5. J Barlow, G.D Lennox, J Ferreira, E Berenguer, A. C. Lees, Ralph Mac Nally, James R. Thomson, Silvio Frosini de Barros Ferraz, Julio Louzada, Victor Hugo Fonseca Oliveira, Luke Parry, Ricardo Ribeiro de Castro Solar, Ima C. G. Vieira, Luiz E. O. C. Aragão, Rodrigo Anzolin Begotti, Rodrigo F. Braga, Thiago Moreira Cardoso, Raimundo Cosme de Oliveira Jr, Carlos M. Souza Jr, Nárgila G. Moura, Sâmia Serra Nunes, João Vic- tor Siqueira, Renata Pardini, Juliana M. Silveira, Fernando Z. Vaz-de Mello, Ruan Carlo Stulpen Veiga, Adriano Venturieri, and Toby A. Gardner. Anthropogenic disturbance in tropical forests can double biodiversity loss from deforestation. Nature, 535: 144–147, 2016. doi: 10.1038/nature18326.

6. Chris T Bauch, Ram Sigdel, Joe Pharaon, and Madhur Anand. Early warning signals of regime shifts in coupled human–environment systems. Proceedings of the National Academy of Sciences, 113(51):14560–14567, 2016.

7. S Baumgärtner. The insurance value of biodiversity in the provision of ecosystem services. Natural Resource Modeling, 20(1):87–127, 2007.

8. S Baumgärtner and S Strunz. The economic insurance value of ecosystem resilience. Ecolocial Economics, 101:21–32, 2014.

9. FJ Bianchi, CJ Booij, and T Tscharntke. Sustainable pest regulation in agricultural landscapes: A review on landscape composition, biodiversity and natural pest control. Pro- ceedings of the Royal Society B, 273(1595):1715–1727, 2006.

10. Charles Blackorby, Walter Bossert, and David Donaldson. Critical-Level Utilitarianism and the Population-Ethics Dilemma. Economics and Philosophy, 13(02):197–230, 1997.

11. E Boserup. The conditions of agricultural growth. London:Earthscan, 1965.

12. Corey J A Bradshaw and Barry W Brook. Human population reduction is not a quick ﬁx for environmental problems. Proceedings of the National Academy of Sciences, 111(46): 16610–16615, 2014.

13. James A. Brander and M. Scott Taylor. The Simple Economics of Easter Island: A Ricardo- Malthus Model of Renewable Resource Use. American Economic Review, 88(1):119–138, 1998.

174 GENERAL DISCUSSION

14. M Branko. Global income inequality in numbers: In history and now. Global Policy, 4(2): 198–208.

15. M Branko. Worlds Apart: Measuring International and Global Inequality. Princeton, NJ: Princeton University Press, 2005.

16. G. Brundtland, M. Khalid, S. Agnelli, S. Al-Athel, B. Chidzero, L. Fadika, V. Hauﬀ, I. Lang, M. Shijun, M. Morino de Botero, M. Singh, S. Okita, et al. Our Common Future (The Brundtland Report). Oxford University Press, 1987.

17. Bradley J Cardinale, J Emmett Duﬀy, Andrew Gonzalez, David U Hooper, Charles Per- rings, Patrick Venail, Anita Narwani, Georgina M Mace, David Tilman, David A Wardle, Ann P Kinzig, Gretchen C Daily, Michel Loreau, James B Grace, Anne Larigauderie, Di- ane S Srivastava, and Shahid Naeem. Biodiversity loss and its impact on humanity. Nature, 486(7401):59–67, 2012. ISSN 00280836. doi: 10.1038/nature11148.

18. J Carmichael, J Tansey, and J Robinson. An integrated assessment modeling tool. Global Environmental Change, 14(2):171–183, 2004.

19. S.R. Carpenter and W.A. Brock. Adaptive capacity and traps. Ecology and society, 13(2): 40, 2008.

20. DL Carr and RE Bilsborrow. Population and land use/cover change: a regional comparison between central america and south america. J Geogr Educ, 43:7–16, 2001.

21. R Chaplin-Kramer, ME O’Rourke, EJ Blitzer, and C Kremen. A meta-analysis of crop pest and natural enemy response to landscape complexity. Ecology Letters, 14(9):922–932, 2011.

22. R Costanza and C Perrings. A ﬂexible assurance bonding system for improved environmental management. Ecological Economics, 2(1):57–75, 1990.

23. A.-S. Crepin, R Biggs, S Polasky, M Troell, and A de Zeeuw. Regime shifts and management. Ecological Economics, 84:15–22, 2012.

24. P Dasgupta. Nature in economics. Environmental and Resource Economics., 39(1):1–7, 2008.

25. A de Sherbinin, D Carr, S Cassels, and L Jiang. Population and environment. Annu Rev Environ Resour, 32:345–373, 2007.

26. Y Decreux and H Valin. Mirage, updated version of the model for trade policy analysis with a focus on agriculture and dynamics. CEPII Working Paper, (2007-15), 2007.

27. R DeFries, J Fanzo, R Remans, C Palm, S Wood, and T.L. Anderman. Metrics for land- scarce agriculture. Science, 349(6245):238–240, 2015.

28. S Derissen, M.F. Quaas, and S Baumgärtner. The relationship between resilience and sustainability of ecological-economic systems. Ecological Economics, 70:1121–1128, 2011.

29. J Diamond. Collapse: How Societies choose to fail or survive. Penguin, 2005.

30. J. Edmonds and J. Reilly. Global Energy: Assessing the Future. Oxford University Press, New York, 1985.

175 GENERAL DISCUSSION

31. F Essl, S Dullinger, W Rabitsch, P.E. Hulme, P Pysek, J.R.U Wilson, and D.M. Richard- son. Delayed biodiversity change: no time to waste. Trends in Ecology and Evolution, 30 (7):375–378, 2015.

32. FAO. How to Feed the World in 2050. Insights from an expert meeting at FAO, (1):1–35, 2009.

33. A Fernihough. Human capital and the quantity-quality trade-oﬀ during the demographic transition. Journal of Economic Growth, 22(1):35–65, 2017.

34. G Ferraz, G.J. Russell, P.C. Stouﬀer, R.O. Bierregaard, S.L. Pimm, and T.E. Lovejoy. Rates of species loss from amazonian forest fragments. Proceedings of the National Academy of Sciences, 100(24):14069–14073, 2003.

35. T Filatova, J. Gary Polhill, and S van Ewijk. Regime shifts in coupled socio-environmental systems: Review of modelling challenges and approaches. Environmental Modelling & Software, 75:333–347, 2016.

36. Jonathan A. Foley, Navin Ramankutty, Kate A. Brauman, Emily S. Cassidy, James S. Ger- ber, Matt Johnston, Nathaniel D. Mueller, Christine O/’Connell, Deepak K. Ray, Paul C. West, Christian Balzer, Elena M. Bennett, Stephen R. Carpenter, Jason Hill, Chad Mon- freda, Stephen Polasky, Johan Rockstrom, John Sheehan, Stefan Siebert, David Tilman, and David P. M. Zaks. Solutions for a cultivated planet. Nature, 478:337–342, 2011.

37. C Folke, J Colding, and F Berkes. Navigating social-ecological systems: building resilience for complexity and change., chapter Synthesis: building resilience and adaptive capacity in social–ecological systems. Cambridge University Press, New York, 2003.

38. MM Gardiner, DA Landis, C Gratton, CD DiFonzo, M O’Neal, JM Chacon, MT Wayo, NP Schmidt, EE Mueller, and GE Heimpel. Landscape diversity enhances biological control of an introduced crop pest in the north-central usa. Ecological Applications, 19(1):143–154, 2009.

39. HJ Geist and EF Lambin. Proximate causes and underlying driving forces of tropical deforestation. BioScience, 52(2):143–151, 2002.

40. Patrick Gerland, a. E. Raftery, H. Ev Ikova, Nan Li, Danan Gu, Thomas Spoorenberg, Leontine Alkema, Bailey K Fosdick, Jennifer Chunn, Nevena Lalic, Guiomar Bay, Thomas Buettner, G. K. Heilig, and J. Wilmoth. World population stabilization unlikely this century. Science, 346(6206):234–237, 2014.

41. David H. Good and Rafael Reuveny. The fate of Easter Island: The limits of resource management institutions. Ecological Economics, 58(3):473–490, 2006.

42. Rhys E Green, Stephen J Cornell, Jörn P W Scharlemann, and Andrew Balmford. Farming and the fate of wild nature. Science, 307(5709):550–555, 2005.

43. L.H. Gunderson and C.S. Holling. Panarchy: understanding transformations in human and natural systems. Island Press, 2001.

44. I Hanski, G.A. Zurita, M.I. Bellocq, and J. Rybicki. Species-fragmented area relationship. Proceedings of the National Academy of Sciences, 110(31):12715–12720, 2013.

45. Ilkka Hanski and Otso Ovaskainen. Extinction debt at extinction threshold. Conservation Biology, 16(3):666–673, 2002.

176 GENERAL DISCUSSION

46. G Hardin. The tragedy of the commons. Science, 162:1243–1248, 1968.

47. Garrett Hardin. Ecology and the death of providence. Journal of religion and society, 15 (1):57–68, 1980.

48. Garrett Hardin. Naked Emperors: Essays of a Taboo-Stalker. Los Altos, 1982.

49. M Hejazi, J Edmonds, L Clarke, P Kyle, E Davies, V Chaturvedi, M Wise, P Patel, J Eom, K Calvin, R Moss, and S Kim. Long-term global water projections using six socioeconomic scenarios in an integrated assessment modeling framework. Technological Forecasting and Social Change, 81:205–226, 2014.

50. K Hubacek, G Baiocchi, K Feng, and A Patwardhan. Poverty eradication in a carbon constrained world. Nature Communications, 8(1).

51. F Isbell, V Calcagno, A Hector, J Connolly, W.S. Harpole, P.B. Reich, M Scherer-Lorenzen, B Schmid, D Tilman, J van Ruijven, A Weigelt, B.J. Wilsey, E.S. Zavaleta, and M Loreau. High plant diversity is needed to maintain ecosystem services. Nature, 477:199–202, 2011.

52. Forest Isbell, Andrew Gonzalez, Michel Loreau, Jane Cowles, Sandra Díaz, Andy Hector, Georgina M. Mace, David A. Wardle, Mary I. O’Connor, J. Emmett Duﬀy, Lindsay A. Turnbull, Patrick L. Thompson, and Anne Larigauderie. Linking the inﬂuence and dependence of people on biodiversity across scales. Nature, 546:65–72, 2017.

53. Timothy H. Keitt. Habitat conversion, extinction thresholds, and pollination services in agroecosystems. Ecological Applications, 19(6):1561–1573, 2009. ISSN 1939-5582. doi: 10.1890/08-0117.1. URL http://dx.doi.org/10.1890/08-0117.1.

54. A Kinzig, P Ryan, M Etienne, H Allison, T Elmqvist, and B.H. Walker. Resilience and regime shifts: Assessing cascading eﬀects. Ecology and Society, 11(1):20, 2006.

55. M Kuussaari, R Bommarco, R.K. Heikkinen, A Helm, J Krauss, R Lindborg, E Ockinger, M Partel, J Pino, F Roda, C Stefanescu, and T Teder. Extinction debt: a challenge for biodiversity conservation. Trends in Ecology and Evolution, 24(10):564–571, 2009.

56. S.J. Lade, A. Tavoni, S.A. Levin, and M Schlüter. Regime shifts in a social-ecological system. Theoretical Ecology, 6(3):359–372, 2013. doi: 10.1007/s12080-013-0187-3.

57. Eric F. Lambin and P. Meyfroidt. Global land use change, economic globalization, and the looming land scarcity. Proceedings of the National Academy of Sciences of the United States of America, 108(9):3465–72, 2011. ISSN 10916490. doi: 10.1073/pnas.1100480108.

58. Eric F. Lambin, B. L. Turner, Helmut J. Geist, Samuel B. Agbola, Arild Angelsen, John W. Bruce, Oliver T. Coomes, Rodolfo Dirzo, Günther Fischer, Carl Folke, P. S. George, Kather- ine Homewood, Jacques Imbernon, Rik Leemans, Xiubin Li, Emilio F. Moran, Michael Mortimore, P. S. Ramakrishnan, John F. Richards, Helle Skånes, Will Steﬀen, Glenn D. Stone, Uno Svedin, Tom A. Veldkamp, Coleen Vogel, and Jianchu Xu. The causes of land- use and land-cover change: Moving beyond the myths. Global Environmental Change, 11 (4):261–269, 2001.

59. L.S. Lima, M.T. Coe, B.S. Soares Filho, S.V. Cuadra, Livia C.P. Dias, Marcos H Costa, Leandro S Lia, and Hermann O Rodrigues. Feedbacks between deforestation, climate, and hydrology in the southwestern amazon: implications for the provision of ecosystem services. Landscape Ecology, 29(2):261–274, 2014.

177 GENERAL DISCUSSION

60. M Loreau, N Mouquet, and A Gonzalez. Biodiversity as spatial insurance in heterogeneous landscapes. Proceedings of the National Academy of Sciences, 100:12765–12770, 2003.

61. S Macfadyen, S A Cunningham, A C Costamagna, and N A Schellhorn. Managing ecosystem services and biodiversity conservation in agricultural landscapes: are the solutions the same? Journal of Applied Ecology, 49(3):690–694, 2012.

62. Emily A Martin, Björn Reineking, Bumsuk Seo, and Ingolf Steﬀan-Dewenter. Natural enemy interactions constrain pest control in complex agricultural landscapes. Proceed- ings of the National Academy of Sciences, 110(14):5534–5539, 2013. doi: 10.1073/p- nas.1215725110.

63. Kyle T. Martins, A Gonzalez, and Martin J Lechowicz. Pollination services are mediated by bee functional diversity and landscape context. Agriculture, Ecosystems & Environment, 200:12–20, 2015.

64. AS Mather and CL Needle. The relationships of population and forest trends. Geogr J, 166:2–13, 2000.

65. S. Messner and M. Strubegger. User’s guide for message iii. International Institute for Applied Systems Analysis, WP-95-69, 1995.

66. Matthew G. E. Mitchell, Elena M. Bennett, and Andrew Gonzalez. Forest fragments modulate the provision of multiple ecosystem services. Journal of Applied Ecol- ogy, 51(4):909–918, 2014. ISSN 1365-2664. doi: 10.1111/1365-2664.12241. URL http://dx.doi.org/10.1111/1365-2664.12241.

67. Matthew G.E. Mitchell, Andrés F. Suarez-Castro, Maria Martinez-Harms, Martine Maron, Clive McAlpine, Kevin J. Gaston, Kasper Johansen, and Jonathan R. Rhodes. Reframing landscape fragmentation’s eﬀects on ecosystem services. Trends in Ecology & Evolution, 30(4):190–198, 2015.

68. M.G.E. Mitchell, E.M. Bennett, and A. Gonzalez. Linking landscape connectivity and ecosystem service provision: Current knowledge and research gaps. Ecosystems, 16(5): 894–908, 2013.

69. M.G.E. Mitchell, E.M. Bennett, and A. Gonzalez. Strong and non-linear eﬀects of fragmentation on ecosystem service provision at multiple scales. Environmental Research Letters, 10(9), 2015.

70. D. Moreno-Mateos, E. B. Barbier, P. C. Jones, H. P. Jones, J. Aronson, J. A. López-López, Michelle L McCrackin, Paula Meli, D Montoya, and J. M. Rey Benayas. Anthropogenic ecosystem disturbance and the recovery debt. Nature Communications, 8:14163, 2017.

71. S Motesharrei, J Rivas, E Kalnay, G.R. Asrar, A.J. Busalacchi, R.F. Cahalan, M.A. Cane, R.R. Colwell, K Feng, R.S. Franklin, K Hubacek, F Miralles-Wilhelm, T Miyoshi, M Ruth, R Sagdeev, A Shirmohammadi, J Shukla, J Srebric, V.M. Yakovenko, and N Zeng. Mod- eling sustainability: population, inequality, consumption, and bidirectional coupling of the earth and human systems. National Science Review, 3(4):470–494, 2016. doi: 10.1093/n- sr/nww081.

72. Safa Motesharrei, Jorge Rivas, and Eugenia Kalnay. Human and nature dynamics (HANDY): Modeling inequality and use of resources in the collapse or sustainability of societies. Ecological Economics, 101:90–102, 2014.

178 GENERAL DISCUSSION

73. Nicolas Mouquet, Yvan Lagadeuc, Vincent Devictor, Luc Doyen, Anne Duputié, Damien Eveillard, Denis Faure, Eric Garnier, Olivier Gimenez, Philippe Huneman, Franck Jabot, Philippe Jarne, Dominique Joly, Romain Julliard, Sonia Kéﬁ, Gael J. Kergoat, Sandra Lavorel, Line Le Gall, Laurence Meslin, Serge Morand, Xavier Morin, Hélène Morlon, Gilles Pinay, Roger Pradel, Frank M. Schurr, Wilfried Thuiller, and Michel Loreau. Review: Predictive ecology in a changing world. Journal of Applied Ecology, 52(5):1293–1310, 2015.

74. Tobias Naaf and Jens Kolk. Colonization credit of post-agricultural forest patches in NE Germany remains 130-230years after reforestation. Biological Conservation, 182:155–163, 2015.

75. United Nations. World Population Prospects: the 2017 Revision Highlights. New York: UN., 2017.

76. Erik Nelson, Guillermo Mendoza, James Regetz, Stephen Polasky, Heather Tallis, DRichard Cameron, Kai MA Chan, Gretchen C Daily, Joshua Goldstein, Peter M Kareiva, Eric Lonsdorf, Robin Naidoo, Taylor H Ricketts, and MRebecca Shaw. Modeling multiple ecosystem services, biodiversity conservation, commodity production, and tradeoﬀs at landscape scales. Modeling multiple ecosystem services, biodiversity conservation, commodity production, and tradeoﬀs at landscape scales, 7(1):4–11, 2009.

77. W.D. Nordhaus. Optimal Greenhouse-gas Reductions and Tax Policy in the "DICE" model. The American Economic Review, 83(2):313–317, 1993.

78. K Norris. Biodiversity in the context of ecosystem services: the applied need for systems approaches. Philos Trans R Soc Lond B Biol Sci., 367(1586):191–199, 2012.

79. K Nyborg, J.M. Anderies, A Dannenberg, T Lindahl, C Schill, M Schlüter, W.N. Adger, K.J. Arrow, S. Barrett, S. Carpenter, F.S. Chapin III, A.-S. Crépin, G. Daily, P. Ehrlich, C. Folke, W. Jager, N Kautsky, S.A. Levin, O.J. Madsen, S. Polasky, M. Scheﬀer, B. Walker, E.U. Weber, J. Wilen, A. Xepapadeas, and A de Zeeuw. Social norms as solutions. Science, 354(6308):42–43, 2016. doi: 10.1126/science.aaf8317.

80. E Ostrom. Governing the commons: the evolution of institutions for collective action. New York, NY: Cambridge University Press, 1990.

81. E Ostrom. Collective Action and the Evolution of Social Norms. Journal of Economic Perspectives, 14(3):137–158, 2000.

82. Otso Ovaskainen and Ilkka Hanski. Transient dynamics in metapopulation response to perturbation. Theoretical population biology, 61(3):285–295, 2002.

83. Derek Parﬁt. Overpopulation and the Quality of Life. In Torbjörn Tännsjö and Jesper Ryberg, editors, The Repugnant Conclusion: Essays on Population Ethics, pages 7–22. Springer Netherlands, 2004.

84. P.F. Peretto and S Valente. Growth on a ﬁnite planet: resources, technology and population in the long run. Journal of Economic Growth, 20(3):305–331, 2015.

85. Owen L. Petchey, Mikael Pontarp, Thomas M. Massie, Sonia Kéﬁ, Arpat Ozgul, Maja Weilenmann, Gian Marco Palamara, Florian Altermatt, Blake Matthews, Jonathan M. Levine, Dylan Z. Childs, Brian J. McGill, Michael E. Schaepman, Bernhard Schmid, Piet Spaak, Andrew P. Beckerman, Frank Pennekamp, and Ian S. Pearse. The ecological forecast horizon, and examples of its uses and determinants. Ecology Letters, 18(7):597–611, 2015.

179 GENERAL DISCUSSION

86. Ben Phalan, Malvika Onial, Andrew Balmford, and Rhys E Green. Reconciling food production and biodiversity conservation: land sharing and land sparing compared. Science, 333(6047):1289–1291, 2011.

87. T Piketty. Capital in the Twenty-First Century. Harvard University Press, 2014.

88. D Pimentel. Soil erosion: a food and environmental threat. Environment, development and sustainability, 8(1):119–137, 2006.

89. D Pimentel and N Kounang. Ecology of soil erosion in ecosystems. Ecosystems, 1(5): 416–426, 1998.

90. D Pimentel, C Harvey, P Resosudarmo, K Sinclair, D Kurz, M McNair, S Crist, L Shpritz, L Fitton, R Saﬀouri, and R Blair. Environmental and economic costs of soil erosion and conservation beneﬁts. Science, 267(5201):1117–23, 1995.

91. R Pirard. Market-based instruments for biodiversity and ecosystem services: A lexicon. Environmental Science & Policy, 19-20:59–68, 2012.

92. S Polasky, C Costello, and C McAusland. On trade, land-use, and biodiversity. Journal of Environmental Economics and Management, 48:911–925, 2004.

93. J.G. Polhill, T Filatova, M Schlüter, and A Voinov. Modelling systemic change in coupled socio-environmental systems. Environmental Modelling and Software, 75:318–332, 2016.

94. J Pretty, C Brett, D Gee, R Hine, C Mason, J Morison, M Rayment, G Van Der Bijl, and T Dobbs. Policy challenges and priorities for internalizing the externalities of modern agriculture. Journal of Environmental Planning and Management, 44(2):263–283, 2001.

95. R.G. Prinn. Development and application of earth system models. Proc. Nat. Ac. Sc., 110 (suppl. 1):3673–3680, 2012.

96. R. Reuveny and C. S. Decker. Easter Island: Historical anecdote or warning for the future? Ecological Economics, 35(2):271–287, 2000.

97. B Reyers, S Polasky, H Tallis, H A Mooney, and A Larigauderie. Finding common ground for biodiversity and ecosystem services. BioScience, 62(5):503–507, 2012.

98. S Rinaldi. Limit cycles in slow-fast forest-pest models. Theoretical Population Biology, 41 (1):26–43, 1992.

99. Isabel M.D. Rosa, M.J. Smith, O.R. Wearn, D Purves, and R.M. Ewers. The environmental legacy of modern tropical deforestation. Current Biology, 26:2161–2166, 2016. doi: 10.1016/j.cub.2016.06.013.

100. Joel Rybicki and Ilkka Hanski. Species-area relationships and extinctions caused by habitat loss and fragmentation. Ecology Letters, 16(SUPPL.1):27–38, 2013.

101. M Scheﬀer. Critical Transitions in Nature and Society. Princeton and Oxford, 2009.

102. M Scheﬀer and J Beets. Ecological models and the pitfalls of causality. Hydrobiologia, 275 (1):115–124, 1994.

103. M Scheﬀer, F Westley, and W Brock. Slow response of societies to new problems: causes and costs. Ecosystems, 6(5):493–502, 2003.

180 GENERAL DISCUSSION

104. M Schlüter, A Tavoni, and S Levin. Robustness of norm-driven cooperation in the commons. Proceedings of the Royal Society B: : Biological Sciences, 283:20152431, 2016.

105. S Schomers and B Matzdorf. Payments for ecosystem services: A review and comparison of developing and industrialized countries. Ecosystem Services, 6:16–30, 2013.

106. R Sethi and E Somanathan. The evolution of social norms in common property resource use. American Economic Review, 86:766–788, 1996.

107. J. Shukla, C. Nobre, and P. Sellers. Amazon deforestation and climate change. Science, 247:1322–25, 1990.

108. A.P. Sokolov, C.A. Schlosser, S. Dutkiewicz, S. Paltsev, D.W. Kicklighter, H.D. Jacoby, R.G. Prinn, C.E. Forest, J.M. Reilly, C. Wang, B. Felzer, M.C. Saroﬁm, J. Scott, P.H. Stone, J.M. Melillo, and J. Cohen. The mit integrated global system model (igsm) version 2: Model description and baseline evaluation. Joint Program Report Series, 124, 2005.

109. R.M. Solow. Intergenerational equity and exhaustible resources. The Review of Economic Studies, 41(5):29–45, 1974.

110. E Stehfest, D van Vuuren, L Bouwman, T Kram, R Alkemade, M Bakkenes, H Biemans, A Bouwman, M den Elzen, J Janse, P Lucas, J van Minnen, C Müller, and A Gerdien Prins. Integrated assessment of global environmental change with IMAGE 3.0: Model description and policy applications. PBL Netherlands Environmental Assessment Agency, 2014.

111. C Thies and T Tscharntke. Landscape structure and biological control in agro-ecosystems. Science, 285(5429):893–895, 1999.

112. D. Tilman and M Clark. Global diets link environmental sustainability and human health. Nature, 515:518–522, 2014. doi: 10.1038/nature13959.

113. Nicolas Titeux, Klaus Henle, Jean-Baptiste Mihoub, Adrián Regos, Ilse R. Geijzendorﬀer, Wolfgang Cramer, Peter H. Verburg, and Lluís Brotons. Biodiversity scenarios neglect future land-use changes. Global Change Biology, 22(7):2505–2515, 2016.

114. Teja Tscharntke, Yann Clough, Thomas C. Wanger, Louise Jackson, Iris Motzke, Ivette Perfecto, John Vandermeer, and Anthony Whitbread. Global food security, biodiversity conservation and the future of agricultural intensiﬁcation. Biological Conservation, 151(1): 53–59, 2012.

115. R K Turner and G C Daily. The ecosystem services framework and natural capital conservation. Environmental and Resource Economics, 39(1):25–35, 2008.

116. E.H. Van Nes and M. Scheﬀer. A strategy to improve the contribution of complex simulation models to ecological theory. Ecological Modelling, 185:153–164, 2005.

117. O Venter, E.W. Sanderson, A. Magrach, J.R. Allan, J. Beher, K.R. Jones, H.P. Possing- ham, W.F. Laurance, P. Wood, B.M. Fekete, M.A. Levy, and J.E.M Watson. Sixteen years of change in the global terrestrial human footprint and implications for biodiversity conservation. Nature Communications, 7(12558), 2016.

118. O. R. Wearn, D. C. Reuman, and R. M. Ewers. Extinction Debt and Windows of Conser- vation Opportunity in the Brazilian Amazon. Science, 337(6091):228–232, 2012.

181 GENERAL DISCUSSION

119. E Yoeli, M Hoﬀman, D.G. Rand, and M.A. Nowak. Powering up with indirect reciprocity in a large-scale ﬁeld experiment. Proceedings of the National Academy of Sciences, 110 (Supplement 2):10424–10429, 2013. doi: 10.1073/pnas.1301210110.

120. Wei Zhang, Taylor H. Ricketts, Claire Kremen, Karen Carney, and Scott M. Swinton. Ecosystem services and dis-services to agriculture. Ecological Economics, 64(2):253–260, 2007.

182 AUTEUR: Anne-Sophie LAFUITE Directeur de thèse: Michel LOREAU

TITRE: Érosion de la biodiversité et durabilité des systèmes socio-écologiques

RÉSUMÉ: Ecosystèmes et sociétés humaines interagissent de façon bidirectionnelle, notamment via la perte de biodiversité et de services ecosystémiques. Au sein de cette boucle de rétroaction, les interactions d’échelles (spatiales et temporelles) mènent à des découplages qui peuvent réduire le bien-être humain et la durabilité des systèmes socio- écologiques (SSEs). Par une approche de modélisation théorique, nous explorons les conséquences de long terme de telles interactions d’échelle. Nous montrons que le décou- plage temporel dû aux dettes d’extinction peut mener à des effondrements, qui concernent les SSEs les plus efficaces. De plus, ces découplages temporels diffèrent les changements nécessaires et réduisent la capacité d’adaptation du système, rendant plus probables des transitions soudaines vers des trajectoires non durables. Cependant, la conservation des habitats naturels et l’internalisation économique des conséquences de la perte de biodi- versité permettent d’éviter ou de réduire ces crises. Cette étude met en évidence le rôle des rétroactions et des interactions d’échelles dans les SSEs, et insiste sur l’importance d’une vision de long terme pour la durabilité des sociétés humaines.

MOTS-CLÉS: biodiversité; économie de l’écologie; dette d’extinction; durabilité; échelle; eﬀondrement; feedback; résilience; services écosystémiques; système socio-écologique. DISCIPLINE: Écologie, Biodiversité et Évolution

LABORATOIRE D’ACCUEIL: Centre de Théorie et Modélisation de la Biodiversité, Station d’Écologie Théorique et Éxpérimentale (UMR CNRS 5321)

183 AUTHOR: Anne-Sophie LAFUITE PhD Supervisor: Michel LOREAU

TITLE: Biodiversity feedbacks and the sustainability of social-ecological systems

ABSTRACT: Human-nature interactions form a feedback loop that is driven by the loss of biodiversity-dependent ecosystem services. These interactions occur over many spatial and temporal scales, and mismatches between the scales of human dynamics and ecological processes can contribute to a decrease in human well-being and sustainability. I investigate theoretically the long-term consequences of biodiversity feedbacks on the sustainability of social-ecological systems (SESs). I show that temporal mismatches resulting from extinction debts can generate unsustainable human population cycles, especially in the most technology-intensive SESs. Moreover, temporal mismatches postpone desirable behavioral changes and reduce resilience, thus increasing the probability of abrupt regime shifts towards unsustainable trajectories. However, natural habitat conservation, e.g. through land set aside or the economic internalization of biodiversity feedbacks, can help prevent or mitigate such crises. This thesis thus emphasizes the role of feedbacks and scales in human-nature interactions, and highlight the importance of foresight for the long-term sustainability of human societies.

KEYWORDS: biodiversity; collapse; ecological economics; ecosystem services; extinction debt; feedback; resilience; scale; social-ecological system; sustainability. DISCIPLINE: Ecology, Biodiversity and Evolution

HOST LABORATORY: Centre for Biodiversity Theory and Modelling, Theoretical and Experimental Ecology Station, Moulis, France.

184