Alternative stable states and human-environment interactions in forest mosaics and ecotones by Kirsten Henderson A Thesis presented to The University of Guelph In partial fulfillment of requirements for the degree of Doctor of Philosophy in Environmental Sciences Guelph, Ontario, Canada c Kirsten Henderson, September, 2016 ABSTRACT ALTERNATIVE STABLE STATES AND HUMAN-ENVIRONMENT INTERACTIONS IN FOREST MOSAICS AND ECOTONES Kirsten Henderson Advisor: Professor Madhur Anand University of Guelph, 2016 Co-Advisor: Professor Chris Bauch Alternative stable states result in drastic vegetation shifts, without much ad- vanced warning. Although common in theory, they are often more difficult to observe in ecological systems and the mechanisms responsible for shifts remain largely un- known. The fragility of bistable systems when faced with large external forces, such as human influence or climate, provides a plausible explanation for the elusiveness of alternative stable states. The work presented here aims to provide quantifiable mech- anisms for threshold responses in forest mosaic and ecotone systems. First, a model of tree-grass recruitment is parameterized with empirical data to gain an under- standing of mechanisms driving vegetative shifts. The following chapter uses climate moisture index to model vegetation structure and stability in a boreal/temperate forest/non-forest system. In addition to climate, anthropogenic land-use change is considered to be one of the greatest stresses on natural ecosystems. Landowner be- haviour, from questionnaire data, is incorporated into a human-environment coupled model to determine the influence of conservation and economic values on landscape dynamics. In both human and environment driven tree-grass systems, recruitment is a limiting factor and as such has a fundamental role in characterizing vegetative state shifts. After reviewing multiple tree-grass dynamical models and calibrating the model with empirical data, we find that soil moisture content acts as a proxy for fire and precipitation regimes, providing a simple, yet descriptive, recruitment func- tion. The findings from the boreal/temperate climate model support an alternative framework for conceptualizing alternative stable states and suggest that future cli- mate regimes may eclipse bistability. Likewise, human influences disrupt the natural bistable characteristic of forest-grassland mosaics. Generally, strong human influence precludes bistability. However, long-term conservation behaviour reintroduces bista- bility, driven by rarity-based conservation rather than natural processes. Ecosystem interactions are intricate and the presence of alternative stable states often adds to the complexity. This thesis highlights the fragility of bistable systems, particularly under strong environmental and human forces. Therefore, models that include empir- ical data and mechanistic processes, which enhance our understanding of biological processes, are key to determining what triggers shifts in states and assessing the vulnerability of a system to change. Contents List of Figures . vii List of Tables . ix Preface . .x Chapter 1: General Introduction . .1 1 Alternative stable states . .2 1.1 Alternative stable states in ecology . .2 1.2 Modelling alternative stable states . .4 2 Subtropical tree-grass mosaics . .4 2.1 Landscape patterns . .4 2.2 Socio-ecological drivers in landscape shifts . .5 3 Boreal/temperate forest and grassland ecotones . .6 3.1 Landscape patterns . .6 3.1.1 Boreal forest . .6 3.1.2 Sub-boreal . .6 3.1.3 Subalpine . .6 3.2 Socio-ecological drivers in landscape shifts . .7 4 Influence of climatic conditions . .7 5 Human Influence . .9 5.1 Human-environment interactions . .9 5.2 Human behaviour and perception . .9 5.3 Human activity in forest-grassland mosaic . 10 6 Thesis outline . 10 Chapter 2: Modelling bistability in tree-grass ecosystems: the recruitment function revisited . 25 1 Introduction . 26 2 Materials and Methods . 28 2.1 Model Review . 28 2.2 Recruitment Models . 29 2.2.1 Phenomenological fire-driven recruitment . 29 2.2.2 Empirically calibrated recruitment function . 30 iv 2.3 Tree-grass cover model . 38 2.4 Analysis . 40 3 Results . 40 4 Discussion . 43 Chapter 3: Bistability or fragility: the influence of climatic conditions on vegetation shifts in a managed forest . 54 1 Introduction . 55 2 Methods . 58 2.1 Case Study . 58 2.2 Model . 60 2.3 Analysis . 62 2.3.1 Current Forest Composition in the Face of Climate Change . 63 2.3.2 Alternative Reforestation Strategies . 63 3 Results . 64 3.1 Equilibrium Forest Cover . 64 3.1.1 Current Forest Composition in the Face of Climate Change . 66 3.1.2 Alternative Stable States . 68 3.1.3 Alternative Reforestation Strategies . 69 4 Discussion . 70 S1 Supplementary Information Text . 79 S1.1 TACA seedling survival . 79 S1.2 Mortality function . 80 S1.3 Converting evapotranspiration to temperature . 81 S2 Supplementary Figure . 84 Chapter 4: Landowner perceptions of the value of natural forest and natural grassland in a mosaic ecosystem in southern Brazil . 85 1 Introduction . 86 2 Methods . 89 2.1 Study area . 89 2.2 Sampling and analysis of land use . 90 3 Results . 91 3.1 Composition and Preference . 91 3.2 Composition and preference change . 92 3.3 Land use influence . 93 3.4 Regional composition . 93 3.5 Restoration . 94 v 4 Discussion . 94 5 Conclusion . 99 Chapter 5: Seeing the Forest for the Grasslands and the People: Sustain- ability of Coupled Human-Environment Mosaic Ecosystems . 119 1 Introduction . 120 2 Materials and Methods . 123 2.1 Study System . 123 2.2 Environment (Land-Use & Natural Dynamics) Model . 124 2.3 Human Behaviour Model . 125 2.4 Analysis . 127 3 Results . 127 3.1 Base Case . 127 3.2 Changing Conservation Values . 128 3.3 Changing Economic and Conservation Discount Rates . 130 3.4 Changing Economic and Conservation Discounting Time Hori- zons . 130 3.5 Extent of Human Influence . 133 4 Discussion . 134 S1 Supporting Information: Materials and Methods . 143 S1.1 Environment (Land-Use & Natural Dynamics) Model . 143 S1.2 Human Behaviour Model . 144 S1.3 Parameter Selection . 146 S1.3.1 Base Case Study . 146 S1.3.2 Changing Conservation Values . 147 S1.3.3 Changing Economic and Conservation Discount Rates 148 S1.3.4 Discounting Economic and Conservation Time Horizons148 S1.3.5 Weak, Moderate and Strong Human Influence . 148 Chapter 6: General Conclusions . 151 A1 MATLAB Code . 159 A1.1 Forest cover: climate moisture index . 159 A1.2 Forest cover: human influence . 164 vi List of Figures 1.1 Shift in equilibrium state. .3 1.2 Forest-grassland mosaic southern Brazil. .5 2.1 Comparing recruitment functions: phenomenological, empirical and mechanistic . 37 2.2 Soil moisture content increases recruitment. 41 2.3 The modified recruitment function (r(T )) decreases the region of bista- bility, when comparing analogous parameters from phenomenological and mechanistic models. 42 2.4 Comparing tree cover over a range of potential natural disturbance rates (v)................................... 43 2.5 Comparing tree cover over a range of potential maximum recruitment rates (α and c).............................. 44 2.S1 Fermi-Dirac distribution in tree-grass systems. 52 2.S2 Intersection of recruitment and disturbance dynamics. 53 2.S3 Parameter plane showing decreased region of bistability over a range of α and v parameters . 53 3.1 Conceptual framework of alternative stable states through changes in variables and changes in parameters. 57 3.2 Map of Tree Farm Licence 48 (TFL 48). Distributed along the Alberta Plateau and in the Rocky Mountains in northeastern British Columbia. 59 3.3 Simplified conceptual diagram of forces and feedbacks on forest cover in our study region (BC boreal/temperate forests). 60 3.4 Forest equilibrium over a range of climate moisture indices. 66 3.5 Parameter plane demonstrating forest cover over a range of potential temperature and precipitation values . 67 3.6 Comparing region of bistability between forest cover modelled with climate change and without climate change. 69 3.7 Alternative reforestation strategies shift the threshold response of for- est cover under varying climatic conditions. 70 3.S1 Logistic regression curves for each drought class species. 80 vii 3.S2 he rate at which forest is replaced with planted (α) or natural (β) forest shifts the threshold response of forest cover (F) to climate moisture index (CMI). 84 4.1 Location of the study site.Map showing the 30 properties surveyed in the S~aoFrancisco de Paula region. 90 4.2 Required incentive to convert cropland to native vegetation. 95 5.1 Human-environment coupling. The negative relationship between hu- mans and natural ecosystems is driven primarily by competition with agricultural land (a). Alternatively, sustainable human behaviour (i.e., valuation of natural ecosystem services) leads to positive changes in natural ecosystems when natural ecosystems are rare (b), whereas the perception of.