|UFDC Home||myUFDC Home | Help|
This item has the following downloads:
1 ANALYSIS OF TRADE OFFS AMONG CARBON SEQUESTRATION, TIMBER PRODUCTION AND WATER YIELD IN NORTH FLORIDA SLASH PINE FORESTS, USA By RONALD CADEMUS A THESIS PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PAR TIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE UNIVERSITY OF FLORIDA 2013
2 2013 R onald C ademus
3 To my mother and my father
4 ACKNOWLEDGMENTS First of all I would like to tha nk my family: my parents and my brothers, Wilson, Rodne, Widnord and Lever; my cousins, Loulou, Naomie and Jean Yves; my uncle Gerard, for their love and encouragement from thousand miles away. Especially, I would like to thank Dr. Francisco Escobedo for a ll his advice and support during my master program. His patience, comprehension, motivation and above all friendship, this thesis work would have not been possible. I also would like to thank Dr. Amr Abd Elrahman and Dr. Matthew Cohen for their valuable c omments and contributions to help improve this thesis document, and also for their teaching that helped me develop my critical thinking I want to thank Watershed Initiative for National Natural Resources and Environment ( WINNER ) p roject of the United Stat es Agency for International development (USAID) Survey study and the School of Natural Resources and Environment (SNRE) for financial support to realize this work and my master program, especially Dr. Stephen Humphrey, Florence Sergile and Melissa Wokasch This research would not have been done without support and contributions from Dr. Daniel McLaughlin, Dr. Carlos Gonzalez Benecke and Dr. Nilesh Timilsina. I also would like to thank Alexandra Je an Charles for supporting and encouraging me in difficult times. I also want to thank those special friends: Sonia Delphin, Olivia Jeanne, Yamil Rodriguez, Symithe Steeve Julien, Cassandra Coulanges, Peggy and Roger Sedlaceck for trusting me and give me all the motivation I ne ed to finish this work. I want to thank my colleagues Arthur, Beneche, Dakson, Isnel, Lidwine, Pascale, Reginald and Dr. Lemane Delva, for being part of this process. I also want to thank my friends in Gainesville: Angelica Martha Oscar, Ivelisse Silvia Sofia, Kimmel, Marco, and Cristobal
5 TABLE OF CONTENTS p age ACKNOWLEDGMENTS ................................ ................................ ................................ .. 4 LIST OF TABLES ................................ ................................ ................................ ............ 7 LIST OF FIGURES ................................ ................................ ................................ .......... 8 ABSTRACT ................................ ................................ ................................ ..................... 9 CHAPTER 1 INTRODUCTION ................................ ................................ ................................ .... 11 Ecosystem Service Framework ................................ ................................ .............. 11 Carbon Sequestration, Timber Production and Water Yield from Forest Ecosystems ................................ ................................ ................................ ......... 13 Timber Production in the US Southeast Region ................................ ..................... 15 Forest Ecosystem Structures and Water Yield ................................ ....................... 16 Spatial Analyses and Trade offs ................................ ................................ ............. 19 Study Objective ................................ ................................ ................................ ....... 22 2 METHODOLOGY ................................ ................................ ................................ ... 25 Study Areas ................................ ................................ ................................ ............ 25 Estimating Ecosystem Services ................................ ................................ .............. 26 Forest Inventory and Analysis (FIA) Program ................................ ................... 26 Carbon Sequestration Estimation ................................ ................................ ..... 28 Timber Volume Estimation ................................ ................................ ............... 30 Water Yield Estimation ................................ ................................ ..................... 30 Trade off Analysis ................................ ................................ ................................ ... 32 Statistical Analysis of Drivers ................................ ................................ .................. 36 3 RESULTS ................................ ................................ ................................ ............... 48 Ecosystem Services Estimation and Interaction ................................ ..................... 48 Effect of Drivers on Individual Ecosystem Service ................................ .................. 49 Effect of Drivers on Ecosystem Service Interactions ................................ .............. 50 4 DISCUSSION ................................ ................................ ................................ ......... 57 Overview ................................ ................................ ................................ ................. 57 Ecosystem Services Estimation and Interaction ................................ ..................... 57 Effect of Drivers on Individual Ecosystem Service ................................ .................. 60 Effect of Drivers on Ecosystem Service Interactions ................................ .............. 62
6 5 CONCLUSION ................................ ................................ ................................ ........ 64 REFERENCES ................................ ................................ ................................ .............. 69 BIOGRAPHICAL SKETCH ................................ ................................ ............................ 78
7 LIST OF TABLES Table page 2 1 Description of the forest inventory and analysis (FIA) attributes and other input data used in the estimation of the ecosystem services .............................. 39 2 2 Descriptive statistics of the quantitative variables and drivers used in the analysis ................................ ................................ ................................ .............. 42 3 1 Descriptive statistics of the level of provision of each eco system service used in the interaction analysis. Note: Level 1= low, 2= Medium and 3= High. ........... 52 3 2 Percentage of plot in each category of interactions between carbon sequestration, timber vol ume and water yield. ................................ .................. 52 3 3 Parameter estimates of the predictors of net carbon sequestration.. .................. 53 3 4 Parameter estimates of the predictors of timber volume.. ................................ .. 53 3 5 Parameter estimates of the predictors of water yield.. ................................ ........ 53 3 6 Effect likelihood ratio tests and parameter estimates for synergy (1) and trade off (0) interactions. ................................ ................................ .................... 54
8 LIST OF FIGURES Figure page 1 1 Examples of relationship between sets of ecosystem services, due to interactions between them or due to response to the same driver... .................. 24 2 1 Forest Inventory Analysis (FIA) plot design ................................ ........................ 43 2 2 Study Area: Northeastern and Northwestern Forest Inventory Analysis (FIA) Survey Units, Florida, USA ................................ ................................ ................. 44 2 3 Prediction of mean annual leaf area index (LAI) from stand basal area for Florida slash pine stands. ................................ ................................ ................... 44 2 4 Relationship between evapotranspiration transpiration ratio and leaf area index (LAI) for slash pine stands in the so utheastern coastal region.. ................ 45 2 5 Water yield as function of net carbon sequestration for forest inventory analysis (FIA) slash pine plots, in Florida northeastern and northwestern .......... 45 2 6 Water yield as function of timber volume for forest inventory analysis (FIA) slash pine plots, in Florida northeaster n and northwestern survey units ............ 46 2 7 Distribution of the normalized values on a 0 1 scale for the variables (a net carbon sequestration, b timber volume, c water yield). ................................ .... 46 2 8 Diagram of the in teraction classification framework ................................ ............ 47 3 1 Net carbon sequestration rate provision levels for forest inventory analysis (FIA) slash pine plots, in Florida northeaster n and northwestern survey units ... 54 3 2 Timber volume provision levels for forest inventory analysis (FIA) slash pine plots, in Florida northeastern and northwestern survey units, 2002 2011. ......... 55 3 3 Water yield provision levels for forest inventory analysis (FIA) slash pine plots, in Florida northeastern and northwestern survey units, 2002 2011. ......... 55 3 4 Ecosystem services interactions among net carbon sequestration, timber volume and water yield for forest inventory analysis (FIA) slash pine plots ....... 56
9 Abstract of Thesis Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Master of Science ANALYSIS OF TRADE OFFS AMONG CARBON SEQUESTRATION, TIMBER PRODUCTION AND WATER YIELD IN NORTH FLORIDA SLASH PINE FORESTS, USA By Ro nald Cademus May 2013 Chair: Francisco Escobedo Major: Interdisciplinary Ecology Managing forests solely for carbon sequestration and timber production objectives might have negative effects on the provision of water due to high losses from evapotranspi ration. Therefore information on the interactions among these three ecosystems services and how they are spatially bundled can provide useful insights for land management decision making. This study used a US Forest Services inventory dataset and computed leaf area index (LAI) to quantify levels of provision and service bundles of carbon sequestration, timber volume and water yield for slash pine sites in North Florida Moreover, using a ranking classification approach we determined spatially explicit int eractions among the services as well as the effect of drivers such as stand age, site productivity, silvicultural treatments, ownership and disturbance regime on individual ecosystem service s and on the interactions ( i.e., trade off/synergy). Results ind icated that growth in biomass reduced water yield during the study period. Neverthe less, this trade off varied across space, as revealed by the lack of correlation in the model of water yield as function of net carbon sequestration or timber volume. Specif ic areas of synergy among the 3 ecosystem services were found. A lso, the
10 results indicated that, although the effect of some drivers was not statistically signi ficant on individual services, all the drivers analyzed affect ed the intera ction among the serv ices with stand age, silvicultural treatment and site quality, the most significant. Finally, the fra mework developed in this study can be used to assess and manage natural ecosystems for multiple and optimal p rovision of services to people.
11 CHAPTER 1 I NTRODUCTION Ecosystem Service Framework The term ecosystem services has been increasingly used and cited in several scientific publications (Assessment, 2003; Costanza et al., 1998; Daily, 1997; Daily, 2000; Guo, Xia o, & Li, 2000) Nevertheless, this term is often used interchangeably with ecological processes, ecosystem functions or environmental services ecosystem goods, etc. (Bennett, Peterson, & Gordon, 2009; Daily, 1997; de Groot, Alkemade, Braat, Hein, & Willemen, 2010) Understanding the differences between those terms is key for communication among users of the ecosystem service concept (de Groot et al., 2010) Ecosystem services are the benefits human receive as a result of the na tural processes from the structures of an ecosystem (Daily, 2000; de Groot, Wilson, & Boumans, 2002; Egoh et al., 2007) However, because of the human controls on the biosphere, many ecosystem services are under th reat (Heal, Daily, Ehrlich, & Salzman, 2001) The Millennium Ecosystem Assessment (2003), has suggested four main categories for ecosystem services: provisioning services, including fresh water, timber production and all other direct goods or services provided to human being ; regulating services, such as flood control and climate regulation; cultural services, related to recreation and aesthetic; and finally, supporting services, which are the natural processes upon which the other categories are dependent, such as nutrient c ycling, soil formation, and primary production. Although there has been a growing interest in the scientific study of ecosystem services for conservation planning and management (de Groot et al., 2010) since the Millennium Ecosystem Assessment in 2005 (Carpenter,
12 2005) a lot of effort is still needed to fully understand the concept and to systematically characterize the concept in measurable biophysical terms (Daily, 2000) Daily (1997), proposed four important eleme nts to better comprehend the ecosystem service framework. First, it is important to identify the stocks of ecosystem structures that provide the ecosystem services by mapping their geographic location. Another element is the ecological characterization, wh ich is a key element that relates to the ecosystem the inherit interdependency between ecosystem services (e. g, carbon sequestra tion, timber and water) and the level at which the production of different services to gether can be sustainable. Finally, it is imperative to monitor changes in the provision of ecosystem services by using appropriate indicator s (e. g., carbon sequestration water quantity). The quantification of ecosystem services is also an important step to the process of ecological compensation by which owners of biological resources (e.g., forest property owners) can be provided with financial benefits for conserving or managing their properties to provide services to other s (Guo et al., 2000) For example, Guo et al. (200), used simulation models coupled with spatial analysis to quantitatively determine the capacity of forest ecosystem to regulate the water flow within a watershed and also included an economic valuation. Also Egoh, Reyers, Rouget, and Richardson (2011) quantitatively evaluated the bundle of ecosystem services supplied by grass lands in South Africa and suggested that when identifying the right target area for a given ecosystem service, conservation interventions tend to be less challenging and more cost effective, especially when areas that provide that service are clustered.
13 I ntegrating ecosystem services into conservation planning can catalyze interest for developing sustainable approaches to assess the congruence and relevance of multiple ecosystem management goals (Balvanera et al., 2 001; Egoh et al., 2007) The biodiversity approach, for example, considers species and habitats as providers of ecosystem goods and material s to society (Carpenter, 2005) and support for ecosystem functions that regulate the sustainability and viability of the services provided (Egoh et al., 200 7) Mapping of ecosystem services, which includes elements of biodiversity and ecological processes, can help identify providers (forest species) of ecosystem services (Kremen, 2005) by assessing and mapping their functional contribution (B alvanera et al., 2001) and can provide explicit arguments for inclusion in environmental assessments (Costanza et al., 1998; Kremen, 2005) Carbon Sequestration Timber Production and Water Yield from Forest E cosy stems In the past century the global environment has been drastically impacted by human activity. For example, there has been an increase in the carbon dioxide concentrations in the atmosphere, inducing climate change (Friedlingstein, Dufresne, Cox, & Rayner, 2003) primarily due to the burning of fossil fuels inappropriate forest management and other anthrop ogenic activities C limate change, together with the severe reduction of biodiversity, has fueled the interest to manage natural ecosystems, while accounting for their multifunctionality (Schwenk, Donovan, Keeton, & Nunery, 2012) Along with wetland and grassland habitats, terrestrial forest ec osystems represent a bigger reservoir of carbon than the atmospheric stocks (Lal, 2004) Given the potential of forest ecosystem to sequester and store carbon dioxide in their living
14 biomass, afforestation and reforestation (hereon jointly referred to as forestation) programs are one of the strategies used by instr uments such as the Kyoto Protocol to mitigate global climate change (Jackson et al., 2005; Jindal, Swallow, & Kerr, 2008) Land use patterns involving reforestation or afforestation have potential impact on other e cosystem services and functions such as stream flow, food security, and loss of biodiversity (Canadell & Raupach, 2008) In addition, carbon sequestration and storage in terrestrial forest biomass can be widely recognized when the social value is accounte d for (Tallis et al., 20 11) This importance and value of forest ecosystems in climate change mitigation and their corresponding social value due to avoided disturbance from non emission (Nordhaus, 2007; Tol, 2009) make these ecosystem service, and therefore the carbon cycle, the focus of many scientific publications (Amthor, 1995; Anderson et al., 2010; Canadell & Raupach, 2008; Dixon et al., 1994; Falkowski et al., 2000; Friedlingstein et al., 2003) The carbon stored in terrestrial forest ecosystems has different fates: it can be released into t he atmosphere, sequestered or conserved in the soil (Brown et al., 1996). The amounts of carbon biomass present in woody plants are products of important ecosystems processes. Indeed, the carbon is produced by photosynthesis and consumed through ecosystem respiration dead material harvesting, etc (Brown, Schroeder and Kern 1999). One of the challenges for forest managers is to provide incentive to landowners for improvement of their forest land to retain more carbon (Subak, 2002) The biomass accumulation in aboveground parts of live trees and the biomass present in the underground coarse roots represents carbon offsets (Dwivedi, Alavalapati, Susaeta, & S tainback, 2009) Although its impact on climate change is
15 positive as an offset strategy to reduce carbon dioxide in the atmosphere, forest management with carbon storage purposes is still under threat since the carbon stored in forest biomass may be rele ased back to the atmosphere should a disturbance occur (Canadell and Raupach 2008). Timber Production in the US Southeast Region Biomass in the form of timber production is also an important ecosystem service because it can generate profit to landowners a nd communities that properly manage forest plantations (Tallis et al., 2011) The rate at which timber is harvested influences the sustainability of this service as well other services related to the functioning of forest ecosystems. Forest management is important both economica lly and ecologically to the US Southeast region, especially in North Florida (Hendry & Gholz, 1986) This activity contributes greatly to the supply of timber product (Dwivedi et al, 2009). Also, many studies recognized the role of pine plantations in the Southeast of United States for their contribution to carbon sequestration (Shan, Morris, & Hendrick, 2001) The forest gross domes tic product ( GDP ) (Brown et al., 2012) In 2010, approximately 50% (nearly 17.4 million acres) of Florida was cover ed by forests. Trends in ownership and land use changes affect how forest lands are managed in Florida (Brown et al., 2012) The Nonindustrial Private For est s (NIPF) account for 65% (around 10.3 million acres) of timberland s (i.e., lands available for timber production) in Florida. However, public ownership i s on the rise, from 28% in 2007 to 30% (4,751,700 acres) in 20 10 (Brown et al., 2012)
16 Forest Ecosystem Structures and Water Y ield Although carbon and timber are receiving a lo t of attention for their driving effects on climate change and the energy budget and their market and profit opportunities respectively, water remains one of the most valuable resources for people on Earth. Via its cycle, water drives most of the biolog ical interactions between all ecosystems on Earth (Chapin Iii & Matson, 2011) Becaus e of this control on living biomass, the hydrologic cycle can also influence the carbon cycle and thus the energy budget (Chahine, 1992) Therefore, a strong interaction exists between the carb on and the hydrologic cycles that warrant a useful approach by studying them together. Hydrologic services provided by terrestrial ecosystems can be defined as the benefits people receive from the impact of these ecosystems on fresh water (Bennett et al., 2009; Brauman, Daily, Duarte, & Mooney, 2007) Vegetation plays a role in both the carbon (GrNzweig, Lin, Rotenberg, Schwartz, & Yakir, 2003; Kirschbaum, 2003) and hydrolog ical cycles (Farley, Jobbgy, & Jackson, 2005; Scheffer, Holmgren, Brovkin, & Claussen, 2005) Through photosynthesis, plants fix the chemical (Churkina, Running, Schloss, & Intercomparison, 1999) energy required by their tissues to sequester carbon in the terrestrial ecosystems (Chapin Iii & Matson, 2011) Just as plant distribution and composition influence the hydrologic cycle (Dunn & Mac kay, 1995; Gerten, Schaphoff, Haberlandt, Lucht, & Sitch, 2004) water controls the productivity and distribution (Stephenson, 1990) of plant communities in the terrestrial ecosystems. Within a given watershed, an increase in forest cover will reduce the surface water flow as a consequence of an incre ase in evapotranspiration, whereas a deforested catchment will increase runoff (Bosch & Hewlett, 1982; Brown, Podger, Davidson, Dowling, & Zhang, 2007; Brown, Zhang,
17 McMahon, Western, & Vertessy, 2005) These mechan isms are driven mainly by the composition and pattern of leaf area (Neilson, 1995), through stomatal conductance and transpiration rates (Skiles & Hanson, 1994) and rooting behavior (Milly, 1997) Evapotranspiration is one of the key ecosystem processes that relate carbon sequestration to water yield and subsequent water supply and climate regulation services. For example, the latent heat flux that drives the transfer of water to the atmosphere and measured in energy units is identical to evapotranspiration, measured in units of water (Chapin Iii & Matson, 2011) Some authors have attempted to model evapotranspiration ( ET ) using a dimension analysis method (describing a process using dimensionless group of variables) to estimate ET from forest ecosystems at multiple temporal scales. For example Running et al. (1989) map ped evapotranspiration in western Montana (USA), predominated by coniferous forests and a comprehensive assessment was made by Zhang et al. (2001), who estimated the mean evapotranspiration response following the conversion of grasslands to a forested cove r. Leaf are index (LAI) is defined as the ratio of leaf area to ground cover for broadleaf plant canopies or projected needle forests (Iiames, Congalton, Pilant, & Lewis, 2008; Myneni, Ramakrishna, Nemani, & Running 1997; Running & Coughlan, 1988) and is one of the most widely used forest attributes for parameterization of forest ecosystem function models (Fassnacht, Gower, MacKenzie, Nordheim, & Lillesand, 1997; Grier & Run ning, 1977; Jensen & Binford, 2004; Myneni et al., 1997) the main forest structure characteristic that allows the exchange of CO2 and water vapor, between the atmosphere and forest canopies (Fassnach t et al., 1997; Iiames et
18 al., 2008) In addition, the relationship between leaf area index (LAI) and water availability has been widely investigated (Vose et al., 1994) For example, a study conducted in Australi a (1999), (Watson, Vertessy, & Grayson, 1999) fou nd that: i) LAI is positively correlated with evapotranspiration, which in turn negatively affects water yield; ii) In a region where soil and water are not limited, the relationship between evapotranspiration and LAI is linear. Also, for most ecosystems i t is widely recognized that biomass production is related to the photosynthetic surface area (Gholz, 1982) Generally an increase in net primary production (NPP) is associated with increase in LAI, which induces a higher water loss from interception and transpiration (Gerten et al., 2004) Additionally, LAI can be spatially analyzed, thus facilitating the identification of areas of interest to better study the interaction between carbon storage and water yield. Some studies hav e used LAI to map photosynthetic vegetation structure s (Kucharik et al., 2000; Running et al., 1989) Therefore, LAI can be considered the common driver between water yield and the aboveground carbon stored in forest ecosystems and used to identify potential areas where trade offs might be occurring. The assessment of the pattern s of LAI development at the stand level is often realized using a chronosequence approach, which is the estimation o f LAI from stands, classified according to different age classes (Clark, Gholz, & Castro, 2004; Gholz & Clark, 2002; Vose et al., 1994) Also, some studies use models derived from field data from chronosequence approaches or from repeated observations on the same stand (McLaughlin, Kaplan, & Cohen, 2012) Models developed by Gholz and Clark (2002), supported by McLaughlin et al. (2012), from studies on slash pine, showed that LAI is
19 hig hly correlated with stand age. Also, Gholz (1982), Grier and Running (1977), graphed LAI versus water balance, LAI versus biomass production, and NPP (Net Primary Production) versus water balance. Their studies found that LAI and biomass are positively cor related, LAI and water balance, as well as NPP and water balance, negatively correlated. Spatial Analyses and Trade offs Additionally, u nderstanding the inter actions among bundles of ecosystem services represents a challenge in land management decision ma king (Bennett et al., 2009; Rodrguez, 2006) Bennett et al. (200 9) provides three reasons that support the need to better understand the se interact ions among ecosystem services: 1) Focusing management of natural systems to produce one ecosystem service can substantially alter the provision of other important services (Bennett et al., 2009) Millennium Ecosystem Assessm ent 2005 and Diaz and Rosenberg 2008) ; 2) M anagement practices can reduce trade offs between sets of ecosystem services by focusing on the natural processes that generate the services (Pretty et al., 2006); and 3) L ack of information of the processes behin d the interactions between ecosystem services can lead to unexpected changes (Gordon, Peterson, & Bennett, 2008) Ecosystem services are in interaction (Figure 1 1) when the level of provision of one service directly affects the level of another service (Bennett et al. 2009). The interaction can be unidirectional in the case where the supply of one service alters the provision of the other but there is no reciprocity. Interactions among services can also be bidirectional when the impact is reciprocal, i.e., the provision of service affects the provision of the other and vice versa. Furthermore, interact ion can be negative or positive, where p ositive interactions happen when the occurrence of one service
20 increases the provision of the other Interactions and negative when t he level of provision of one service decreases the level of supply of the other s Thus t production of one compromises the provision of the other. An opposite response to the same driver can also create trade off between two ecosystem services. Meanwhile, a synergy occurs when the production of the 2 ecosystem services simultaneously increase due to interaction between them or because how they respond to a particular driver (Bennett et al. 2009, Schwenk et al. 2012, Nelson et al. 2009). An analysis of the trade offs among key ecosystem services can provide useful information on how one service can affect the provision others (Brauman et al., 2007) Usually, spatial information at the landscape scale only focuses on the land u se land cover (LULC) change, neglecting the biophysical function behind those issues (de Groot et al., 2010) To better understand the trade offs and synergies among a set of ecosystem services, it is imperative to analyze the landscape drivers behind their relationshi p (Bennett et al., 2009) The study of relationship between ecosystem services and the values that they generate is important to address decision making that can help manage trade offs in land use change (de Groot et al., 2010) To date, local and regional studies that account for these relationships are still limited (de Groot et a l., 2010) Therefore, st udies on the assessment of a land use/ cover capacity to provide a bundle of ecosystem services and the different drivers that influence trade off can provide useful information to ecosystem managers. Furthermore, decision making can be more efficient if, besides relying on location of features and functions, it can also be based on the quantification of the bundle of
21 provision of services (Troy & Wilson, 2006) Mapping ecosystem services can require analysis of extensive spatial data when focused on a regional scale, which may lead to the use of available models from literature that developed underlying ecosystem functions (de Groot et al., 2010) Mapping of ecosystem service approach is used by many other authors in other parts of the world. (Timilsina et al., 2012) introduced a spatial framework to identify priority areas for carbon storage in the state of Florida. Spec ifically, his study consisted of mapping hotspots of carbon storage from a range of forest types and management schemes. Gimona and van der Horst (2007) investigated multifunctionality hotspots for three bundled ecosystem services in Scotland by using a we ighting scheme approach to create benefit maps for the three services. Other approaches may use the mapping process as a step to identify important areas for biodiversity conservation, areas that provide important ecosystem services and therefore assess c onflict and spatial congruence (Egoh et al., 2007; Polasky et al., 2008) Thus, spatially explicit analyses are crucial to understand the biophysical processes behind the production of ecosystem services ( (Brown & Schroeder, 1999) Understanding the multifunctionality of landscape systems can help identify areas of trade offs and synergies, and therefore better manage conflicts (Chan, Shaw, Cameron, Underwood, & Daily, 2006; Egoh et al., 2008; Egoh, Reyers, Rouget, & Richardson, 2011; Groot et al., 2007) Also, the approach of mapping ecosystem functions and the relevance of this exercise remain a challenge (Willemen et al. 2008, de Groot et al. 2010). Therefore use of Geographic Inf ormation System (GIS) (Egoh, Reyers, Rouget, Bode, & Richardson, 2009; Gimona & van der Horst, 2007) coupled
22 with available observation al data (e. g. Forest Inventory data ) can help develop models that identify loca l patterns or bundle s of geographic phenomena and to better quantify ecosystem services for resources allocation purposes (Troy & Wilson, 2006) Study Objective Ecosystem services are the final, measurable out puts from ecosystem functions in both managed and natural forests that directly benefit humans. So, u nderstandi ng the relationship between different sets of ecosystem services (Kareiva, Heather, H., C., & Stephen, 2011; Kremen, 2005) through spatial analysis is possible by combining available field data with models in a geog raphic information platform (de Groot et al., 2010; Guo et al., 2000) Therefore, t he purpose of this study is to examine, spatially and statistically, carbon sequestration, timber production and water yield intera ctions in slash pine ecosystems in North Florida. More specifically, since forest plantations use water to sequester carbon and store it in their biomass, we proposed to assess the trade offs between carbon sequestration, timber production and water yield In order to accomplish this we first used inventory data from the FIA, available functional models and the literature for slash pine in Florida to quantify a bundle of ecosystem service s: carbon sequestration, timber production, and water yield. In this study we used two models developed by McLaughlin et al. (2012) where : the first one predicts leaf area index (LAI) from stand basal area and the second one, stand level water use as a function of LAI. With the ecosystem services quantified at the plot l evel a GIS based classification algorithm was used to assess the interaction among the level of pro vis ion of 3 key ecosystem services during a defined study period (2002 2010): water yield, carbon sequestration, and timber volume growth. In this classific ation the ecosystem
23 services were ranked in 3 classes, high medium and low, then an equation was used to estimate a three digit code, which identifies where there were trade off s or synergies among the three services. Furthermore, we determined the effect of biophysical drivers, such as stand age, and others such as silvicultural treatments and disturbance regime, on these ecosystem services and their interactions Therefore, we propose the following questions and hypotheses: Question 1: Are specific areas of slash pine a re as in north Florida reducing water yield? Hypothesis 1: Higher r eduction in water yield can be identified by FIA plots with higher carbon sequestration r ates and timber volumes. Question 2: How do forest structure (e. g. basal area, and stand age), management (e. g., site quality, silvicultural treatments ) and other human and ecological factors (e.g., ownership and disturbance regime s) drive carbon sequestration, timber production and water yield interactions? Hypothesis 2 : Drivers that positively affect carbon sequestration rates and/or volume of timber, will negatively affect water yield (Bennett et al., 20 09; McLaughlin et al., 2012) The results of this framework can be useful because they can help identify areas that maximize the provision of these three ecosystem services. Furthermore, the re levance of this framework is the application of its results in the development of management alternatives that maintain consistent levels of ecosystem service provision while not compromising certain management or conservation goals.
24 Figure 1 1. Examples of relationship between set s of ecosystem services, due to interactions between them or due to response to the same driver. The black arrows show positive effects and the grey negative effects [Adapted from Bennett, E. M., G. D. Peterson and L. J. Gordon (2009) Understanding relationships among multiple ecos ystem services (Page 1397, Figure 2). Ecology Letters, 12, 1394 1404
25 CHAPTER 2 METHODOLOGY Study Areas The study area was the northern region of Florida because it is the most forested part of the state (Carter & Jokela, 2003) The study area encompasses two of the four Florida Forest Inventory and Analysis ( FIA ) survey units: the Northeast and Northwest. In 2007, these regions covered 6.6 and 5.5 million acres, respectively which represent 76% of the total of timberland surveyed in Florida in the cycle 8 period (i.e., inventory carried out over the period 2002 to 2007) Northern Florida is dominated by pine flatwoods ecosystems (Timilsina et al., 2012) but other potential natural vegetation occurrences are: sand pine ( Pinus clausa ), in the deep sand and xeric zones; bald cypress ( Taxodium distichum L. ), water tupelo ( Nyssa aquatica L), and other hardwoods species, including laurel oak ( Quercus laurifolia ), live oak ( Quercus geminata ), sweetbay ( Magnolia virginiana ), sweetgu m ( Liquidambar styraciflua ), spruce pine ( Pinus glabra ), red maple ( Acer rubrum L. ), etc., in the f l ood plain s Due to human impact over the past century, the dominant longleaf pine ( Pinus palustris ) ecosystems have been replaced by slash pine ( Pinus elli ottii ), which represent around 69% of conifers in Florida Because of its capacity to grow fast and its adaptation to low fertility soils of US Coastal plain (Gholz et al., 1991) Slash pine (Pinus elliottii) is one of the dominant planted species in that region (Shan et al., 2001) One of the advantages of Slash pine plantations is that management and subsequent ecosyst em models are easier to be developed due to the homogenous structure with relatively simple understory species composition (Cropper Jr & Gholz, 1993) However, leaf area index ( LAI ) which
26 regulate s water availability, varies greatly across slash pine plantation in North Florida (Gholz & Clark, 2002) Slash pine forests are found mostly on low er elevations, where the natural habitats are formed over Quartzipsamments, nutrient poor and mostly poorly drained soils, which contain deep loamy or clayey particles at the subsurface (Hendry & Gholz, 1986) The climate is humid subtropical with a mean annual temperature of 19 C, and the average precipitation varies from 1000 to 1500 mm annually (McNab et al., 2005) The surface water system is characterized by fresh water springs and natural lakes which are found on limestone rocks formations, and the presence of some major rivers (McNab et al., 2005) A common disturbance regime is fire that occurs in moderate or low intensity in sand pine and longleaf pine areas (McNab et al., 2005) Timberland class represents 92% of the total forest cover in Florida, in 2010. The yellow pine forest group (i.e., slash pine, loblolly pine, lo ngleaf pine); which occupies 46% of timberland in Florida, is the dominant forest type group. Slash pine is the most domina nt yellow pine type group and also the dominant forest type, when considered individually. It re presents 30% of the total timberland in the state (4.8 million acres). In Florida, The Northeastern and Northwest ern FIA survey units contain 76% of all merchan table volume (14.7 billion cubic feet). Although yellow pine plantation in the Northwest unit has doubled from 1995 to 2007, the Northeast still accommodates the majority of planted yellow pine (Brown et al., 2012) Estimati ng Ecosystem Services Forest Inventory and Analysis ( FIA ) Program The Forest Service Inventory and Analysis (FIA) of the U.S. Department of Agricu lture (USDA) is a program dictated by the McSweeney McNary Forest Research
28 the litera ture reviewed; some parameters we re from the FIA database (e. g., basal area, stand age). This dataset was used mainly because of its public availability. Moreover, since the focus of this research is on net biomass change over time (i.e. ne t carbon sequestration), the fact that FIA remeasurement s were matched at the tree level (Woudenberg et al., 2010) is a useful contribution o f this study and for future research. Carbon Sequestration Estimation The carbon estimates in this analysis are based on data from the F I A inventor ies carried out during cycle 8 and 9. The cycle 8 inventory include s data from inventory 2002 to 2007, except 2005 while the cycle 9 contains inventories f r om inventory year 2009 to 2011 Often, the market for carbon credits focuses on changes in carbon storage (Hoover, Birdsey, Heath, & Stout, 2000) therefore, our objective was net carbon sequestration over a time period. We used slash pine stands in the Northeaste rn and Northwestern survey units and plot identification number to match plots sampled in cycle 8 to the ones remeasured in cycle 9. The result was 377 study plots, which are well distributed over the entire study area (Figure 2 2 ). We downloaded the FIA p lot level dataset, available in shape file format, from the Florida Geographic Data Library ( http://www.fgdl.org ). Additionally, since the plot dataset does not include carbon estimates at the tree level, we obtained tr ee level data from the FIA websites: http://www.fia.fs.fed.us/tools data/ The plot level shapefiles provide data on the condition of the plot and information such as ownership class, stand age, site pr oductivity class, disturbance, silvicultural treatment regimes, measured basal area, and understory and aboveground carbon. The tree level dataset provides information on each individual tree sampled and includes the status condition (whether the tree is l ive, felled, or dead), estimated carbon values measured from the dry
29 biomass portions of the tree (stump, bole, sapling, roots, etc.) for both aboveground and underground components, and estimated timber volumes according to the market standard (Woudenberg et al., 2010) Furthermore, with the tree level dataset, we use Microsoft Excel filter tool to select plot level standing live and dead trees based on if the status condition clas s (STATUSCD) was equal to 1 ( i.e., live trees ) or 2 (standing dead). Moreover, using the FIA column name ( CN ) which is a unique number to identify a sampled tree, we match ed the 2002 2007 and the 2009 2011 data. Since the carbon values provided by th e FI A database at the tree level are in pounds of carbon per tree, these per tree values were converted to megagram s or ton s of carbon per hectare (Mg C/ha) using the conversion factor TPA_UNADJ ( i.e. trees per acre unadjusted), which represent a theoretical number of a specific tree in 1 acre ( 0.404686 ha) given the plot size ( Table 2 1) These values were: i) 0.999 for trees in macroplots, ii) 6.01 for trees in subplots and iii) 74.96 for trees in microplots (Woudenberg et al., 2010). Using the pivot table t ool in Microsoft Excel, we summed the values of all trees per plot. In addition, this tree table was matched to the plot table using the plot column number ( PLT_CN ) a unique number that link a tree record to the plot record. The carbon storage value for e ach plot is the sum of the aboveground carbon in trees and was derived from the following equation (1) : (2 1) where is the dry biomass in the merchantable bole; the d ry biomass in the tree stump ; the d ry biomass in the top of the tree ; d ry biomass of saplings and the Dry biomass of woodland tree species
30 To determine the net annual carbon sequestration between measurement cycles we used the following equation (2) : (2 2) where, CSQTNET was the net annual Carbon sequestered during the time period, in megagrams of carbon per hectare per year (MgC/ha/yr). This value can be a negative number, since this a net change in carbon storage. Both carbon storage (CSTG1) in the first cycle and carbon storage (CSTG2) in the second cycle were in megagram of carbon per hectare. Finally, REMPER was the remeasurement period, the number of years between measurements for remeasured plots. Timber Volume Estimation We used the same tree level procedure as for carbon sequestration estimation to calculate the timber volume during the period of study. Since generally commercial timber production is a valuable commodity (Tallis et al., 2011) we used timber volume estimates in cycle 9 as the ecosystem service of interest. The attribute from the FIA database considered was the VOLCFSND which is defined as s ound cubic foot volume in the merchantable stem of trees. This measurem ent was recorded only for tree diameter s at Breast Height (i.e.DBH; 1.4m) greater than 5 inches (Woudenberg et al., 2010) We converted tree level values (cubic feet per tree) into cubic meter per hectare, usi ng the conversion factor TPA_UNADJ. Using the pivot table tool in Microsoft Excel, we summed the values of all trees per plot This tree table was matched to the plot table using the PLT_CN, as we did previously for carbon sequestration estimates. Water Y ield Estimation Since (LAI) is identified as key driver for the ecosystem services being studied as explained above, and because LAI measurements are not reported in the FIA database,
31 our estimations of water yield were based on modeled relationships betwe en (LAI) and specific stand structure attributes. Several forest structures influence the amount of leaves available that directly affect forest productivity and water yield. Gonzalez Benecke et al. (2012) developed an integrated model where LAI is predict ed from stand index, tree density and basal area. This model captures many factors that could influence LAI at the stand level and also uses data from 15 year time period and was measured frequently (Gonzalez Benecke, Jokela, & Martin, 2012) However, McLaughli n et al. (2012) developed a non linear model for slash pine stands in Florida, where mean annual leaf area index is predicted from stand basal area (Figure 2 3); thus, we estimated LAI, using McLaughlin et al. (2012) model. Moreover, actual evapotranspi ration was estimated using forest structure data for Slash pine ecosystems in Florida based on McLaughlin et al. (2012) who developed models that relate ecosystem water use and forest stand structure. Their results revealed that LAI is the best predictor of water use, which therefore can be used to predict water yield I n the studies reviewed by McLaughlin et al. (2012) different m ethods were used to measure evapotranspiration ( ET ) (Table 1), as ET was estimated using either Eddy Covariance (EC) measureme nts (Bracho et al., 2008; Gholz & Clark, 2002; Knowles, 1996) model simulation (Ewel & Gholz, 1991) or Eddy Covariance (EC) measurements combined with model simulation (Liu, Riekerk, & Gholz, 1998) We used the relationship b etween leaf are index (LAI) and the ratio of evapotranspiration to precipitation (ET/PPT), developed in the same study by McLaughlin et al. (2012) to calculate the annual water yield ed by the FIA plots in cycle 9
32 (2009 2011) First of al l, evapotranspirati on precipitation (ET PPT ) ratio was obtained from the linear regression model (Figure 2 4) using equation 2 3 : (2 3) w here ET/PPT is the ratio of evapotranspiration to precipitation and LAI, leaf area index in m 2 m 2 W ater yield for each plot was computed using the following equation : (2 4) w here WY is water yield in cubic meter per hectare, on a year basis (m3/ha/year) per year; ET/PPT is the ratio of evapotranspiration to prec ipitation, which was previously predicted from LAI (dimensionless); MAP is mean annual precipitation in meter (m), from a 10 year period (2001 2011), which includes our study time period (2002 2011). These precipitation data were obtained from a grid downl oaded from the PRISM Climate Group at Oregon State University website and are presented in Table 2 1 ( http://www.prism.oregonstate.edu/products/matrix.phtml ) The precipitation values w ere The estimates in this study, however do not consider the resulting water yield that could be observed at the ground level or surface water runoff (McLaughlin et al., 2012) Also the water yield es timates do not account for any changes (e.g., seasonal climate fluctuations) that could have occurred during the study period considered. Trade off Analysis As hypothesized, trade offs are expected between growth in biomass (i.e., carbon sequestration and timber volume growth) and water yield (Farley et al., 2005; Jackson et al., 2005) therefore we used our estimated ecosystem service values to develop two simplified regression models with Excel to explore relation ships between
33 water yield and net carbon sequestration and/or timber volume. As displayed in figure s 2 5 and 2 6 the graphs are showing a regression line that could be used to define a production frontier for assessing trade offs between net carbon sequ estration and water yield or between timber volume and water yield. However, as revealed by small adjusted R 2 (0.14 and 0.29), there is a weak relationship between dependent and independent variables, suggesting there is a variability across sites. Also, t here is no indication of the relative three way interaction among the level of provision within the ecosystem service bundle. Therefore another method was used to analyze the interactions between the ecosystem services. The ranking method is a common metho d used to rank the level of production of several ecosystem service goals (timber, carbon or water) and to prioritize areas where a particular goal is being achieved at the highest level when compared to others (Carr & Z wick 2007) Thus, this framework essentially can be used to classify the level of pro vision of a finite set of ecosystem services. Therefore, we used this ranking method as the conceptual basis for a spatial classification framework for ecosystem service tradeoffs. The framework was derived based on the Land Use Conflict Identification Strategy (LUCIS) model developed by Carr and Zwick (2007). Their objective was to classi fy lands based on suitability analysis in order to determine preferences or appropriateness for agriculture, conservation or urban uses. In contrast to the LUCIS model which used raster analysis; our approach was based on vector (point feature) analysis. T he reason was because our variables represent discreet phenomena and therefore cannot be interpolated in space (Mitchell, 2009)
34 Since, the data for the 3 ecosystem services present dif ferent ranges in values ( Table 2 2) we normalized the values usi ng a scale from 0 to 1 (Figure 2 7) by dividing all the values by the maximum values for each ecosystem service. This same approach was used by Carr and Zwick (2007) to determine land use preferences and conflicts bet ween agriculture, conservation and urban areas. In ArcGIS Version 10, we classified the normalized values using the Natural Break (Jenks) classification method, as opposed to Manual, Equal Interval and Standard Deviation classification methods, because the data values of the three services are not normally distributed, as suggested by Carr and Zwick (2007). Specifically, the natural breaks or Jenks natural breaks is a data classification method based on Jenks optimization procedure (Mennis & Liu, 2005) The algorithm used by this method arranges the values into different classes, relying on an iterative process where different breaks in the dataset are used to minimize the variance within classes and maximize the variance between classes, as much as possible (Brewer & Pickle, 2003) This method produced 3 classes for each service, which were coded respectively 1, 2 and 3. The codes 3, 2 and 1 represent high, medium and low, respectively. Since we were working with net estimates of the variables, the negative values were classified as 1 (low). Moreover, we determined interaction (i.e. synergy of tradeoff) codes by : ( 2 5)
35 where IC is the three digit code defining the type of interaction (a priori defined); CSL, the carbon sequestration level; TVL, the timber volume level and WYL, the water yield level. The output codes for the interactions were a series of numbers between 111 and 333 (Figure 2 8) which were further classified into plots where one of the ecosystem service is dominant over the others ( i.e., trade off) and plots where at least two of the ecosystem services ar e dominant ( i.e., synergy) (Figure 3 4 ). Furthermore, in ArcGIS, we created maps which display the spatial distribution of the FIA plots where trade off s or synergies as classified using our interaction classification framework, occur between the 3 ecosys tem services (i.e. bundle) and areas where water yield is dominant compared to areas where carbon or timber is dominant. Table 3 2 presents the calculated values of the three services used in the classification framework. For the purpose of this framework we defined a synergy as the situation when two of the ecosystem services being analyzed in a bundle, are produced at the same or higher (high high, medium medium) level than the third one (Bennett et al., 2009; Raud sepp Hearne, Peterson, & Bennett, 2010) More explicitly, when two of the services are generated at: 1) the same level 3 (high) while the other is at the level 2 (medium) or 2) when two of the services are generated at the same level 2 (medium) while the other at the level 1 (low), or 3) when the three services are yielded at the same level, either 1 (low), 2 (medium) or 3 (high). On the other hand, we defined a trade off as the situation when one of the ecosystem services is dominant compared to the other one (Bennett et al., 2009; Raudsepp Hearne et al., 2010) In this case a trade off would be when one of the services is generated at the level 3 and the other two at
36 level 1 or 2, or when one of services is produc ed at the level 2 and the other ones at the level 1. Statistical Analysis of Drivers We also tested the effect of different drivers on the provision of each individual ecosystem service as well as the effect of the same drivers on the resulting interactio ns in the ecosystem service bundle, based on the codes generated from the classification framework. Among the drivers tested, was stand age (McLaughlin et al., 2012; Timilsina et al., 2012) This variable is to be interpreted carefully because of inconsistencies in the method used for its estimation by the FIA. Although the time period considered in this study is relatively short (7 years) considering the ecosystem services being analyzed, the stand age used from the second inventory period (2009 2011) ranged from 2 to 139 years with 75% being lower than 49 years, which provides a good range to test temporal variation of the ecosystem proc esses. Other variables used were ownership, site productivity class, disturbance regime and silvicultural treatments (Woudenberg et al., 2010) Ownership (land tenure) was reported as public (e.g. state, federal, national park service, national forest system, etc..), and private but were reclassified into 2 dummy variables, 0 for public ownership and 1 for private ownership. Variations among the public land tenure classes (e.g., Cor porate and n on governmental conservation organization s) were not accounted for. In addition, we assumed that the plots remain under the same land tenure status over the analysis period. The data analyzed consisted of approximately 72% of the plots managed under private ownership and 28, under public land. For the site productivity class, the slash pine forest stands studied were growing on sites quality classified as 1, 2, 3, 4, 5 and 6, with productivity ranging >15.8, 15.7 and
37 11.6, 8.4 and 11.5 6.0 an d 8.3, 3.5 and 5.9, and 1.4 and 3.4 cubic meter per hectare per year (m3/ha/year), respectively (Woudenberg et al., 2010) As far as disturbance regime, we considered disturbances r ep orted from the second year (2009 2010), that occurred from 2003 to 2011. The dataset reported disturbance regime s that have been caused by disease on sampling or seedling trees, fire that i s either prescribed or natural at the crown or ground level, and li vestock or animal grazing or any anthropogenic damage In the statistical analysis performed, we used two classes: 0 (undisturbed) when there is no presence of disturbance and 1 (disturbed) for plots damaged by any of the reported factors. Approximately 89 % of the plots being analyzed showed no presence of disturbance. Finally, we tested the influence of silvicultural treatments on the provision of the three ecosystem services (Jerome K, 2009; Shan et al., 2001) P lots we re treated using several stand improvement activities such as cutting, t he use of fertilizers, herbicides or other activities with the objective of enhanc ing th e commercial value of a stand (Woudenber g et al., 2010) About 78% of the plots were not treated, with only 22% receiving silvicultural treatments at least once during the time period between the inventory cycles. All the statistical analyses were described in this study were analyzed in the St JMP Pro 10 package We used two different types of statistical tests to analyze the effect of the drivers on the interaction among net carbon sequestration, timber volume and water yield. First, we used a multiple regressi on analysis, with a backward selection expressed by the following equation: (2 6)
38 where y represents the value of the dependent variable that is being explained, in this case net carbon sequestration, timber volum intercept of the graph described by Equation 4. And x1, x2, x3, x4, x5 and x6, are the independent variables, represent by stand age, silvicultural treatment, ownership, site productivity and disturbance regime, respec tively. The coefficients b1, b2, b3, b4, b5 and b6, represent the change in Y that correspond to a unit change in stand age, basal area, silvicultural treatment, ownership, site productivity and disturbance regime, respectively. H omoscedasticity of the re siduals was examin ed using scatter plots and the visual relationship between residual predicted values. Normality was examined with the normal quantile plot and the goodness of fit using the Shapiro Wilks test. When assumptions were not met, data were log transforme d for the three variables. A P value of <0.05 and parameter estimate s were used to interpret the statistical significance of a particular driver for net carbon sequestration, net volume growth and water yield. Second, we used a multiple logisti c regression to test the effect of these same drivers (e.g. stand age, basal area, silvicultural treatments, ownership, site quality and disturbance) on the interactions among net carbon sequestration and water yield as well as timber volume and water yiel d. Two dummy variables (i.e. 0 and 1) were analyzed using a similar approach as a study of forest carbon hotspots in Florida, by Timilsina et al. (2012). Specifically, we analyzed the interaction codes as 2 dummy variables; trade The results of the multiple logistic regression were interpreted using chi square of the Likelihood Ratio tests, parameter estimates and/or odds ratios, and all were tested for significance using =0.05.
39 Table 2 1 Descriptio n o f the f orest inventory and a nalysis (FIA) attributes and other input data used in the estimation of the ecosystem services Input data Description Range of values Units Source Leaf Are Index (LAI) Ratio of leaf area to ground cover for broadleaf plant c anopies or projected for needle forests 0.16 7.12 m2/m2 (McLaughli n et al., 2012) Evapotranspiratio n precipitation ratio (ET/PPT) Ratio of evapotranspiration and precipitation (PPT), which indicates the portion of precipitation that has been evaporated and/or lost through transpiration. 0.549 0.967 (McLaughli n et al., 2012) Mean annual precipitation (MAP) Average annual precipitation for the period 2001 2011 1.105 1.638 m http://www. prism.or egonstat e.edu/ Car bon Tree aboveground Carbon value in the aboveground portion of the tree, except foliage, recorded for live trees with a diameter greater than 1 inch. Period1: 0.00 122.700 Period2: 0.00 99.183 Mg/ha FIA Manual Inventory year Year in which the data w ere recorded. Period1: 2002 2007 Period2: 2009 2011 Year FIA Manual Owner class code A code indicating the status ownership of the plot during the inventory. 11 National Forest System 12 National Grassland 13 Other Forest Service 21 National Park Serv ice 22 Bureau of Land Management 23 Fish and Wildlife Service 24 Department of Defense/Energy 25 Other federal 31 State 32 Local (County, Municipal, etc.) 33 Other non federal public 46 Private FIA Manual
40 Table 2 1 Continued Input data Description Range of values Units Source Stand age The estimated age of the plot using field records and or/ local procedures. Period2: 2 to 139 years Year FIA Manual Site productivity class code A code indicating the potential growth in cubic meter per hectare per yea r, which is related to the capacity of a forest land to grow biomass. (1) 15.8+; (2) 11.6 15.7; (3) 8.4 11.5; (4) 6.0 8.3; (5) 3.5 5.9; (6) 1.4 3.4; (7) 0.0 1.3 m3/ha/ year FIA Manual Disturbance code A code indicating the type of disturbance tha t was happened since the last inventory of a plot or within the last 5 years, in case of a new plot. To be recorded, the disturbance must have affected at least 1 acre and have caused damages to 25% of the trees in the plot. 0 No visible disturbance 10 I nsect Damage 20 Disease Damage 30 Fire damage (from crown and ground fire, either prescribed or natural) 40 Animal Damage 41 Beaver 42 Porcupine 43 Deer/ungulate 44 Bear 45 Rabbit 46 Domestic animal/livestock 50 Weather Damage 51 Ice 52 Wind (includes h urricane, tornado) 53 Flooding (weather induced) 54 Drought 60 Vegetation (suppression, competition, vines) 70 Unknown / not sure / other 80 Human caused damage 90 Geologic disturbances FIA Manual
41 Table 2 1 Continued Input data Description Range of values Units Source Disturbance year The estimated year in which a disturbance type have happened. From 2003 to 2010 FIA Manual Stand treatment code A code indicating the type of stand treatment that was applied since the last inventory of a plot or w ithin the last 5 years, in case of a new plot. To be recorded, the treatment type must have affected at least 1 acre. 00 No observable treatment. 10 Cutting 20 Site preparation Any practices realized with the intention of preparing a site natural or artificial regeneration 30 Artificial regeneration Following a disturbance or treatment (usually cutting), a new stand where at least 50% of the live trees present resulted from planting or direct seeding. 40 Natural regeneration Following a disturba nce or treatment (usually cutting), a new stand where at least 50% of the live trees present (of any size) were established through the growth of existing trees and/or natural seeding or sprouting. 50 Other silvicultural treatment The use of fertilizers, herbicides, girdling, pruning, or other activities (not covered by codes 10 40) designed to improve the commercial value of the residual stand. FIA Manual Basal area of live trees Stand basal area of live trees which DBH or DRC is greater than 1 inch. Basal area m2/ha FIA Manual Remeasurement period Number of years between inventories for remeasured plots 1.9 8.9 year FIA Manual
42 Table 2 2 Descriptive statistics of the quantitative variables and drivers u sed in the analysis Variable Minimum Maximum Mean Std. Deviation Net Carbon Sequestration (Mg/ha/yr) 11.716 9.15 5 0.573 2.653 Timber Volume (m3/yr) 0.000 329.08 55.575 59.571 Water Yield (m 3 /ha/yr) 461.113 6298.73 2223.479 1230.033 Stand age (yeas) 2 138 33.873 24.044 N=377 for all variables
43 Figure 2 1 Forest Inventory Analysis (FIA) plot design [ Source: Woudenberg, S. W., B. L. Conkling, B. M. O'Connell, E. B. LaPoint, J. A. Turner and K. L. Waddel. 2010. The Forest Inventory and Analysis Database: Database Description ersion 4.0 for Phase 2 (Page 8, Figure2). ed. F. S. U.S. Department of Agriculture, Rocky Mountain Research Station. Fort Collins, CO.
44 Figure 2 2 Study Area: northeastern and n orthwestern forest inventory a nalysis (FIA) s urv ey u nits, Florida, USA Fi gure 2 3. Prediction of mean annual leaf area index ( LAI ) from stand basal area for Florida slash pine stands. [Borrowed from McLaughlin, D., Kaplan, D., and Cohen, M. 2012. Managing Forests for Increased Regional Water Yield.. (Page 34, Figure 3) Journal of the American Water Resources Association
45 Figure 2 4. Re lationship between evapotranspiration transpiration ratio and leaf area index (LAI) for s lash pine stands in the southeastern coastal region. [Borrowed from McLaughlin, D., D. Kaplan and M. Cohe n (2012) Managing Forests for Increased Regional Water Yield (Page 24, Figure 2b) Journal of the american water resources a ssociation. Figure 2 5 W ater yield as function of net carbon sequestration for forest inventory a nalysis (FIA) slash pine plots, in Florida northeastern and n orthwestern survey units, 2002 2011
46 Figure 2 6 Water yield as function of timber volume for forest inventory a nalysis (FIA) slash pine plots, in Florida northeastern and n orthwestern survey units, 2002 2011 Figu re 2 7. Distribution of the no rmalized values on a 0 1 scale f or the variables (a net carbon sequestration, b timber volume, c water yield) using a 3 class classification scheme g enerated by the natural b reaks algorithm from ArcGIS 10 Note, 1 = low, 2= medium and 3 = high provision levels.
47 Figure 2 8 Diagram of the i nteraction classification f ramework
48 CHAPTER 3 RESULTS Ecosystem Services Estimation and Interaction Accordin g to the descriptive statistic f or the v ariables reported in Table 3 1, during the period 2002 2010, the classification framework developed in this study estimates that around 72% of the plots provide low (mean= 0.511Mg C/ha/year), 22% medium (mean=2.685 Mg C/ha/year) and 6% high (mean=5.725 Mg C/ha/year), levels of C sequestration. As for timber volume, approximately 67% of the plots were classified as providing low (mean= 22.639 m 3 /ha ), 28% as medium (mean=101.538 m 3 /ha ) and 5% as high (mean=243.339 m 3 /ha ) levels. Fina lly, for water yield, nearly 55% of the plots were found in the low provision level, 35 as medium and 10 as high, with means of 1363.46, 2772.29 and 5001.95 m 3 /ha/year respectively. Overall, the classification framework indicated that for all the variables, the data values were s kewed left, as approximately 70% of the plots were classified as providing low levels for all the services. This similarity in the distribution of the data determines the direction (i.e., trade off or synergy) of the interaction between the services. However, as shown in F igures 3 1 to 3 3 the spatial distribution of high water yield values mostly overlap with low and medium values for net carbon sequestration and net timber volume growth. A portion of the plots studied (around 20%) showed low synergy interactions among the three services, i.e. the services were supplied at the low provision level of 1. Most of the trade off interactions found in this study were because of t he dominance of water yield (38% of the cases), timber volume (13 % ) and net carbon sequestration (over 11 % ). However, as r eported in T able 3 2, our framework did identify a few areas (only 1% ) where the 3 ecosystem services in the bundle were in synergy at the intermediate
49 provision level. Finally, in a few cases there was synergy bet ween timber and water (around 4%); between carbon and water (2% ); and bet ween carbon and timber (over 10% ). Mapping ecosystem service provision at the plot level shown in figure s 3 1, 3 2 and 3 3, displayed the spatial distribution of the three provision levels across the study area. Moreover, t he spatial distribution in Figure 3 4 displays the pattern of plots where trade offs (i.e. crosses) occur compared to plots where there is synergy (i.e. circles). In the bundle, synergistic interactions were identified among the three services or between pair of services. Effect of Drivers on Individual Ecosystem Service The results from the regression analysis indicated that stand age, treatment and site quality were significant drivers of net carbon sequestration, at p=0.0052, p<0.0001 and p<0.0023, res pectively (Table 3 3 ). Therefore, the older the forest stand, the lower the net increase in carbon sequestration. In addition, carbon sequestration decreased as a result of implementing silvicultural treatments. Furthermore, site quality was also a signifi cant predictor of net carbon sequestration, as higher productive sites (i.e. lower class codes) were associated with higher carbon sequestration rates. Although ownership was not a statistically significant driver, net carbon sequestration was associated more so with private forests than those under public land tenure. Finally disturbance regime was not statistically significant, but on average, disturbed plots positively increased net carbon sequestration. For timber volume, three of the drivers, stand age, silvicultural treatment and site quality, had a statistically significant effect on timber volume as indicated in Table 3 4, where timber volume was higher in older stands than in younger ones (p=0.0455).
50 Similarly to carbon sequestration, silvicult ural treatment and site quality had significant effect (both at p<0.0001) on this variable. Indeed, slash pine forest stands that received treatments show an increase in the merchantable timber volume. For managed (treated) slash pine forests, merchantable timber volume is more likely to increase, compared to unmanaged slash pine stands. Ownership and disturbance regime were not statistically significant drivers for timber volume. Delphin, ( 2012) found that timber production in most of forested areas in Nort h Florida was under low risk of damage due to hurricanes. Finally, for water yield, all of the drivers analyzed had a significant effect, ex cept disturbance regime (Table 3 5 ). Water yield significantly decreases as the slash pine stands become older (p<0. 0001). Also, greater water yield values were associated with treated plots (p<0.0001). Unlike for the other variables, ownership was a significant predictor. Slash pine stands managed under private ownership yielded more water, compared to public owned sla sh pine forests. Finally, in contrast, to net carbon sequestration and timber volume, higher water yield values were more associated to lower site quality. Effect of Drivers on Ecosystem Service Interactions Results suggest that all of the drivers analyze d were statistically significant and indicate the likelihood of having a synergy or tradeoff among the s ervices. As indicated in Table 3 6 the coefficient for age was negative, so as a forest stand gets older the probability of having a trade off between net carbon sequestration, timber volume and water yield increases. In contrast, plots that receive silvicultural treatment are more likely to indicate a synergy between the services. Also, stands managed under public ownership were more likely to indicate a synergy between the services than the stands managed under private ownership. Moreover, synergistic plots were more likely to be
51 associated with more productive sites. Finally, disturbance regime, which was not a significant driver when analyzing the ser vices individually, indicates that stands that have experienced disturbance are more likely to indicate synergy between the services.
52 Table 3 1. Descriptive s tatistics of the level of pro vision of each ecosystem service used in the interaction ana lysis Note: Level 1= low, 2= Medium and 3= High. Ecosystem Service Level N Percent Minimum Maximum Mean Std. Deviation Net Carbon Sequestration ( Mg C/ha/year) 1 271 71.9 11.716 1.571 0.511 2.175 2 83 22.0 1.613 4.042 2.685 0.698 3 23 6.1 4.294 9.15 5 5.725 1.201 Timber Volume ( m 3 /ha/ ) 1 252 66.8 0.000 62.16 22.639 17.961 2 108 28.4 62.21 168.49 101.538 28.072 3 18 4.8 169.18 329.08 243.454 50.339 Water Yield ( m 3 /ha/year ) 1 207 54.9 461.13 2050.292 1363.46 394.816 2 132 35.0 2064.207 3830.178 2772.29 508.843 3 38 10.1 3915.432 6298.726 5001.95 670.493 Table 3 2. Percent age of plot in each category of interaction s between carbon sequestration, timber volume and water y ield Note: Level 1= low, 2= Medium and 3= High Synergy Trade off Code Percentage of plots Code Percentage of plots 111 All 3 services are in low synergy 19.63% 112 Water is moderately dominant 28.12% 122 Moderate synergy between timber and water 3.71% 113, 213 Water is highly dominant 10.08% 221 Moderate synergy between carbon and timber 9.28% 121 Timber is moderately dominant 9.55% 212 Moderate synergy between carbon and water 2.12% 211 Carbon is moderately dominant 6.90% 222 All 3 services are in moderate synergy 1.06% 311, 312, 321 C arbon is highly dominant 4.77% 331 High synergy between carbon and timber 1.33% 231 Timber is highly dominant 3.45%
53 Table 3 3 Parameter estimates of the predictors of net carbon sequestration (n ote that: treatment: 0=untreated and 1=treated; ownersh ip: 0=public and 1=private; disturbance: 0=undisturbed, 1=disturbed; site quality=1, 2, 3, 4, 5 and 6). Predictor Estimate Std Error t Ratio Prob>|t| Intercept 4.0542135 0.130006 31.18 <.0001* Age ^ 0.002361 0.00084 2.81 0.0052* Treatment ^ [0/1] 0. 624561 0.047084 13.26 <.0001* Ownership [0/1] 0.0295234 0.044058 0.67 0.5032 Site Quality 0.071779 0.023397 3.07 0.0023* Disturbance ^ [0/1] 0.0194269 0.059434 0.33 0.7440 Effect of predictor is statistically significant at ^ Drive m easured in cycle 9 (2009 2011 ) Table 3 4 Parameter estimates of the predictors of timber volume (n ote that: treatment: 0=untreated and 1=treated; ownership: 0=public and 1=private; disturbance: 0=undisturbed, 1=disturbed; site quality= 1, 2, 3, 4, 5 and 6). Predictor Estimate Std Error t Ratio Prob>|t| Intercept 6.3382277 0.543221 11.67 <.0001* Age ^ 0.007042 0.003509 2.01 0.0455* Treatment ^ [0/1] 1.029454 0.196736 5.23 <.0001* Ownership [0/1] 0.067585 0.184095 0.37 0.7137 Site Quality 0.64122 8 0.097764 6.56 <.0001* Disturbance ^ [0/1] 0.1083232 0.248339 0.44 0.6630 Effect of predictor is statistically significant at ^ Drive measured in cycle 9 (2009 2011 ) Table 3 5 Parameter estimates of the predicto rs of water yield. (n ote that: treatment: 0=untreated and 1=treated; ownership: 0=public and 1=private; disturbance: 0=undisturbed, 1=disturbed; site quality=1, 2, 3, 4, 5 and 6). Predictor Estimate Std Error t Ratio Prob>|t| Intercept 684.15413 376.4522 1.82 0.0700 Age ^ 17.3 4145 2.431407 7.13 <.0001* Treatment ^ [0/1] 1113.7285 136.3381 8.17 <.0001* Ownership [0/1] 324.7228 127.5775 2.55 0.0113* Site Quality 442.70012 67.75018 6.53 <.0001* Disturbance ^ [0/1] 50.617268 172.0992 0.29 0.7688 Effect of predictor is statis tically significant at ^ Drive measured in cycle 9 (2009 2011 )
54 Table 3 6 Effect likelihood r atio tests and parameter estimates for synergy (1) and trade off (0) interactions between net carbon sequestration timber volume and water yield in Florida slash pine plots. (Note that: treatment: 0=untreated and 1=treated; ownership: 0=public and 1=private; disturbance: 0=undisturbed, 1=disturbed; site quality=1, 2, 3, 4, 5 and 6). Driver Estimates DF L R ChiSquare Prob>ChiSq Age ^ 0.0289166 1 29. 6760995 <.0001* Treatment ^ [0/1] 0.93621588 1 8.82217979 0.0044* Ownership [0/1] 0.596385 1 4.55557823 0.0363* Site Quality 0.48486631 1 10.5227632 0.0015* Disturbance ^ [0/1] 1.0417733 1 6.75663149 0.0139* Effect of predictor is statistically signific ant at ^ Drive measured in cycle 9 (2009 2011 ) Figure 3 1 Net carbon seques tration rate provision levels for forest inventory a nalysis (FIA) slash pine plots, in Florida northeastern and n or thwestern s urvey units, 2002 2011
55 Figure 3 2. Tim ber volume provision levels for forest inventory a nalysis (FIA) s l ash pine plots, in Florida northeastern and n ort hwestern survey units, 2002 2011 Figure 3 3. W ater yield provision levels for forest inventory a nalysis (FIA) s lash pine plots, in Florida northeastern and n or thwestern s urvey units, 2002 2011
56 Figure 3 4 Ecosystem s ervices interactions among net carbon sequestration, timber volum e and water yield for forest inventory a nalysis (FIA) s lash pine plots, in Florida northeastern and n or thwestern survey units, 2002 2011
57 CHAPTER 4 DISCUSSION Overview This study presents a framework for quantifying and analyzing interactions among ecosystem service bundles, and their levels of provision, at the landscape scale. Specifically this study quantified and explored interactions among three ecosystem services: net carbon sequestration, net volume growth and water yield and also analyzed the influence of different human and ecological drivers on the provision level of the service and the result ing interactions. While the results confirmed evidence of trade off between water yield and carbon sequestration and timber in slash pine forest in North Florida, during the period 2002 to 2011, our classification framework identified some areas with moder ate synergy between the three ecosystem services or between pair of services. These areas of synergy have implications for multiple use management activities and policies that aim to maintain desired provision levels of multiple ecosystem services. Further more, our analysis indicated that stand age, silvicultural treatment, and site productivity were the most important drivers that influence the provision of the ecosystem services and therefore, dictate the direction of the interaction (e.g., trade off or s ynergy) among them. These findings are discussed in the following sections. Ecosystem Services Estimation and Interaction During the period between the FIA inventory cycle 8 and 9, the slash pine stands inventoried sequestered between 11.7 and 9 0 (mean=0 .6) Mg C/ha/year. These findings might be underestimating the total value as the fate of harvested products, which delay carbon emissions (Tallis et al., 2011; Timilsina et al., 2012) was not accounted for in this study. Using our spati al classification framework, 72% of the plot
58 studied were classified as providing low ecosystem service levels, with an average of 0.5 11 Mg C/ha/year. Nearly 22 an d 6% of the plots were categorized as medium and high provision, with averages of 2.7 and 5. 7 Mg C/ha/year, respectively. While this might be a result of the classification method used, this concentration of the carbon values was consistent with the findings of Timilsina et al. (2012) which indicated that slash pine fores ts present a lower probability of being a hotspot for carbon storage, compared to other forest types (e.g., upland hardwood and oak hickory). In addition, the overall mean value of 0.6 Mg C/ha/year or the maximum value of 9.2 Mg C/ha/year, were slightly lo wer than carbon sequestration rates (10 12 Mg C/ha/year) reported by Shan et al. (2001), for a 17 year slash pine stand. This is due to the effect of stand age on aboveground carbon sequestration, suggesting that older forest sequester less carbon than you nger ones (Clark et al., 2004) as 33 years was the mean age for the plot analyzed in this study In the case of timber, the merchantable volume average d approximately 56 c ubic meter per hectare ( m 3 /ha ). The classification framework grouped 67 % of the slash plots as low pro vision, with an average of 22.6 m 3 /ha This same classifi cation categorized 28.4 and 4.8 % of the plots as medium (101.5 m 3 /ha ) and high ( 243.5 m 3 /ha ) provision. Since slash pine is considered the dominant softwood species in Florida (Brown et al., 2012; Shan et al., 2001) these values are within range of estimates reported by (Brown et al., 2012) for softwoods in Florida., which averaged 82.5 cubic meter per hectare ( m 3 /ha). During the period of the study there was a positive water yield for all the Forest Inventory Analysis (FIA) plots consi dered, with values ranging from 461.0 to 6298.7
59 m 3 /ha/year This range of water yield values is similar to findings by McLaughlin et al. (2012) that reports annual water values ranging from 500 to over 6000 m 3 /ha/year However, water yield did decrease as a result of growth in biomass (Figure 2 5 and 2 6 ). There is wide range of publications that support the paradigm of reduction in water yield (e.g., runoff) as consequence of forest plantations (Bosch & Hewlett, 198 2; Farley et al., 2005; Jackson et al., 2005; van Dijk & Keenan, 2007) However, when the ecosystem functions behind the service are not taken into account, potential management practices that could enhance water use efficiency from forest plantations ar e often not fully considered and adopted into management (Jerome K, 2009) This study, while not attempting to reject the well supported paradigm that forest plantations dec rease the provision of water yield, does provide a framework that helps identify areas where synergies can be found among different ecosystems services provided by a forest area dominated by a single tree species Other studies such as Bennett et al. (200 9) and Raudsepp Hearne et al. (2010) defined trade off among ecosystem services as the situation when the production of one service inhibits the provision of the other, i.e., the supply of one service increases while the supply of the other decreases. They also defined synergy as the situation when the provision of two ecosystem services increases or decrease simultaneously, which represent a win win situation (Power, 2010). Indeed, Raudsepp Hearne et al. (2010) identified trade off pattern s at the landscap e scale between regulation services (e.g., carbon sequestration) and p rovision services (e.g., water). However, studies such as these on interaction s among sets of ecosystem services often use correlation
60 coefficient s and graphics for their analysis. Theref ore our approach analyzing these ecosystem service interactions, provision levels, and their drivers is a contribution to these types of studies. Although most of the provision levels of the three ecosystem services were classified as low, only 19.6% of t he plots were categorized in the 111 (i.e., the proportions are classified using natural breaks in the data ( (Carr & Zwick, 2007) In addition, our analysis identified some plots with synergy between pairs of services. In the case of net carbon sequestration and timber volume interaction, the synergy or t rade off defined here in this study is based on a management goal stand point, as carbon and timber refer to the same tree biomass measured in different units. However, since the majority of above carbon biomass is encountered in merchantable part of a tre e bole (Brown et al., 2012) this categorization of the biomass is important in defining and managing the forest ec osystem for a specific goal (e.g., carbon or timber) and therefore, can be used to identify which areas in north Florida and which management activities can help better reduce the negative impact of natural and managed forests on water yield. Effect of Dr ivers on Individual Ecosystem S ervice Usually growth in forest stands is quantified using metrics such as increase in aboveground ground biomass (e.g., net annual carbon sequestration) or the increment in volume of the stand, e.g., net annual volume growt h (Arneth, Kelliher, McSeveny, & Byers, 1998) The results reported in this study suggest that stand age, silvicultural treatments, and site quality were all significant drivers of these attributes of growth in forest carbon biomass and merchantable timber volume. Control of competition at early age benefits growth in carbon biomass of managed forest stands. However, the same
61 growth can be obtained in older stand when the practice of understory removal is used (Shan et al., 2001). The significant effect of site quality on carbon sequestration and timber volume is plausible as this driver indicate the capacity of a land to grow biomass, on an annual basi s (Woudenberg et al., 2010) Usually studies discuss the negative effect of forest plantations (Farley et al., 2005; Sahin & Hall, 1996; Zhan g, Dawes, & Walker, 2001) on water yield without accounting for the specific forest structures and their functions associated with this effect (Jerome K, 2009) However, McL aughlin et al. ( 2012) suggested that management schemes that control key biophysical drivers can help increase water yield substantially in forested areas. Our study, while not proposing specific management strategies of certain biophysical drivers of wate r yield, found that stand age, silvicultural treatment ownership and site quality were significantly associated with water yield. The results suggested that higher stand age is associated with lower water yield values. This can be explained by the fact tha t leaf area index ( LAI ) increase s as a forest becomes older (McLaughlin et al., 2012) As indicate in Table 3 4 silvicultural treatment s are positively associated with higher water yield. This is because a reduction in forest biomass by thinning, for instance wil l reduce LAI and therefore, increases water yield (Bosch & Hewlett, 1982; Douglass, 1983; Farley et al., 2005; Hewlett & Hibbert, 1961) Moreover, ownership was significantly associated with increase in water yield In fact, our findings showed that slash pine stands managed under public ownership, on average, show higher increase in water yield than their counterparts managed under private land tenure. This is may be due to water yield management goals and silvicul tural activities (Shan et al., 2001)
62 managed for water (e.g., water management district lands) while forests under private land tenure in the Southern region are primarily managed for timber production (Heath, Smith, Woodall, Azuma, & Waddell, 2011) which indicate more frequent silvicultural activities, which ha ve direct effect s on water yield (McLaughlin et al., 2012) Effect of Drivers on Ecosystem S ervice Interactions An important part of this study was to analyze the influence of drivers on the likel ihood of a plot be ing associated with a tradeoff or a synergy All of the drivers analyzed in this study, significantly predicted this outco me. However, as shown in Table 3 6 stand age, treatment and site quality were the most significant drivers. This is because these three drivers were also significant predictor of the services individually. One method often used to analyze the interaction among multiple ecosystem services is testing how each service responds to a common driver. Opposite responses indic ate trade off while similar ones suggest synergy (Be nnett et al., 2009) As developed in the method section, LAI was identified as the common driver in all three ecosystem services evaluated in this study. However, LAI was not analyzed as it was directly used in the estimation of water yield. Although the results of this study support evidence from already published literature on interaction among forest ecosystem services (Bennett et al., 2009; Farley et al., 2005; Jackson et al., 2005) this study introduce s a cl assification framework that has some potential applications for ecosystem management, but there are some limitations. One limitation is reflected in the classification algorithm used to group the data. Given the continuous nature of the variables, the clas ses generated by the natural breaks algorithm hide underlying trends in the dataset. Although 377 plots were used in
63 the study, they were distributed across nearly 50,000 square kilometer (km 2 ), across north Florida, which suggests a relative small sample size. Also, the results of the framework are limited and dependent on the classification method used, which can influence the number of plots identified in the category as trade off or synergy. But, since the values of the ecosystem services were skewed an d therefore, non normal, the Natural Break classification method was the best suited (C arr & Zwick 2007) H owever, once again this lim its the use of and interpretation of the continuous data in other types of optimization and multi criteria analyses. Finally, another limitation is related to the size of the plots (1 acre) used in this analysis, which makes the contribution of a plot, in terms of increase in water yield at a watershed scale, appear to be very minor
64 CHAPTER 5 CONCLUSION Based on the conceptual framework which define s ecosystem services as the direct benefits from natural ecosystems to humans, t his study investigated in teractions between carbon sequestrations, timber volume production and water yield for slash pine forests in n orth Florida. This study is novel in that the provision levels of each of these 3 ecosystem services was quantified using georeferenced field dat a from the Forest Inventory Analysis (FIA) program of the US Forest Service In the calculation the net values of each services was estimated over a 7 year time period (2002 2011 ). Secondly, a classification framework was developed to determine which plot s exhibited tradeoff or synergy interactions among the three ecosystem services or between pair of services. And finally, the study tested the effect of some human and ecological drivers on the individual services and also on the interactions in an ecosyst em service bundle (e.g., trade off or synergy). R esults indicated that biomass measured as carbon sequestration or timber volume reduced water yield during the study period. Nevertheless, this trade off interaction varied across space, as revealed by the l ack of correlation in the model of water yield as function of net carbon sequestration or timber volume growth. The classification framework developed in this study also account ed for this spatio temporal variability, and therefore could be used to identif y interest where trade off and synergy occur Also, the results indicated that, although the effect of some drivers was not statistically significant on individual service s (e.g., disturbance regime and ownership on timber volume an d carbon sequestration ), all the drivers analyzed
65 affect the interaction among the services studied with stand age, treatment and site quality, the most significant. Generally, the management of natural resources tends to focus on a single resource or obj e ctive, overlooking the multiple function s of an eco system (Nelson et al., 2009) This approach often ignor es the capability of an ecosystem to, when properly managed, generate a wide range of and multiple, services to people, and therefore maximize profits as well as enhance sustainability (Bennett et al., 2009; Tallis & Polasky, 2009; Tallis et al., 2011) Thus, managing natural ecosystems properly and sustainably requires information o n how to reach multiple benefits, objectives, or desired provision level of different ecosystem service s in a bundle (Tallis et al., 2011) However, reaching desired levels of these multiple services and identifying their spatio temporal characteristics is difficult. Geospatial and statistical modeling however, as used in this study can provide useful information on how to manage the natural ecosystem in order to reduce trade offs (Tallis & Polasky, 2009) These objectives can also be understood using hierarchical sets of statement that define different goals (Carr & Zwick 2007) For example, a timber company may want to achieve a certain timber volume from its managed forests. But, an environmental institution involved in climate change regulation may want the forest to sequester the maximum amounts of carbon per year. Similarly, a private or public entity managing the supply of water resources may have an interest in knowing the quantity of water a forested area can yield every year. In these three cases, if the entities depend on the same ecosystems (e. g., managed forests) and a single institution wants to be responsible for its multiple management, o ne of the strategies to attain sustainability
66 would be to prioritize actions o n area s or ecosystem s where these goals (or ecosystem service provision levels) are provided at the highest level. As environmental protection becomes more important due to the a nthropogenic impact s on natural systems, ecosystem services emerge as a relevant conceptual framework to study and manage natural ecosystems and monitor the effects of these unprecedented rates of land use change (Daily, 2000) Some of the most important changes are the increase in carbon dioxide concentration in the atmosphere (Vitousek, Mooney, Lubchenco, & Melillo, 1997) and alteration of species composition (Tilman et al., 1997) These environmental changes, which affect ecosystem processes and the flow of goods and services, (de Groot et al., 2010) are felt worldwide (Daily, 2000) For example, natural ecosystems in North Florida, on ce dominated by Long Leaf pine have been replaced by Slash pine (Clark et al., 2004) While this change in forest cover can provide substantial benefits to one sector (fores t industry), it may affect the overall balance of the regional ecosystems and this affects other process (e.g., water balance, biodiversity). In the US southeastern lower coastal plain, environmental conditions (e.g., soil and wildfire) coupled with manage ment practices have greatly influenced the spatio temporal variability of biophysical structures, such as leaf area index (LAI), composition and stand densities. Similarly, water and energy fluxes are sensible to these change in landscape conditions (Gholz & Clark, 2002) Most of the studies that assess the interaction of multiple ecosystem services use biodiversity as a proxy. One of the key aspects of this study is to illustrate that even a single species forest can provide multiple ecosystem services. Unlike many of the studies cited above, our analytical approach is not using any simulat ion or weighting of
67 the ecosystem service considered. We let the data dictate the ranking (high, medium, and low) of the services individually by applying the same classification system. Future research that take into account the issues explained above to investigate provision levels and interactions among biomass growth and water quantity can improve this study. Also, water resources entities interested in knowing the impact of forest management on water yield by their watersheds might want to use datasets which include more frequent inventories than those by the FIA program One other approach for evaluat ing interaction s between multiple ecosystem services is multi criteria decision analysis (MCDA) or multiple criteria decision support. For example (Schwenk et al., 2012) used this analytical approach to evaluate forest management alternatives on the provision of carbon sequestration, timber production and biodiversity. Their methodology included weight assignment to different managem ent objectives that target a specific ecosystem service. R esults can be used to assess how multiple forest management approaches suppl y different (opposite) services and how to balance trade off s and maximize the provision of the ecosystem services. E nvi ro nmental planning, decision making, and management of natural ecosystems for multi functional uses can create synergies between conservation (e.g. water services or carbon sequestration ) and economic growth ( e.g., timber volume) To achieve that goal, it is important to assess the relationship between ecosystem management and the provision of all the linked services provided and therefore detect optimal management option (de Groot et al., 2010) For example, this study, which assesses spatial and temporal analysis of evapotranspiration which is an important ecosystem process that links carbon sequestration timber production and water yield,
68 as a result of afforestation, is compelling to better understand the tradeoffs involved with the co management of services (Balvanera et al., 2001; Kearns, Inouye, & Waser, 1998; Monica, 1998) This spatial assessment which includes mapping and visualization is one of the instruments often used to analyze management changes on natural syste ms. In conclusion most studies focus on the quantification or valuation of a single ecosystem service without accounting for the interactions among multiple ecosystem services and their different levels of provision The method used in this study identif ied these areas where land managers can have acceptable levels for three key ecosystem services and an understanding of the forest management variables driving that interaction. Therefore, results and methods from this work could be used to improve the und erstanding and use of the ecosystem service framework and also help managers who rely on spatial information to know how and where management can be most effective and efficient at providing ideal bundle s of ecosystem services provided by natural ecosystem s.
69 REFERENCES Amthor, J. S. 1995. Terrestrial higher plant response to increasing atmospheric [CO2] in relation to the global carbon cycle. Global Change Biology 1(4): 243 274. Anderson, R. G., Canadell, J. G., Randerson, J. T., Jackson, R. B., Hungate, B. A., Baldocchi, D. D., Ban Weiss, G. A., Bonan, G. B., Caldeira, K., Cao, L., Diffenbaugh, N. S., Gurney, K. R., Kueppers, L. M., Law, B. E., Luyssaert, S., & O'Halloran, T. L. 2010. Biophysical considerations in forestry for climate protection. Frontiers in Ecology and the Environ ment 9(3): 174 182. Arneth, A., Kelliher, F. M., McSeveny, T. M., & Byers, J. N. 1998. Net ecosystem productivity, net primary productivity and ecosystem carbon sequestration in a Pinus radiata plantation subject to soil water deficit. Tree Physiology 18 (12): 785 793. Assessment, M. E. 2003. Ecosystems and Human Well Being: A Framework for Assessment : Island Press. Balvanera, P., Daily, G. C., Ehrlich, P. R., Ricketts, T. H., Bailey, S. A., Kark, S., Kremen, C., & Pereira, H. 2001. Editorial: Conserving B iodiversity and Ecosystem Services. Science 291(5511): 2047. Bennett, E. M., Peterson, G. D., & Gordon, L. J. 2009. Understanding relationships among multiple ecosystem services. Ecology Letters 12(12): 1394 1404. Bosch, J. M. & Hewlett, J. D. 1982. A re view of catchment experiments to determine the effect of vegetation changes on water yield and evapotranspiration. Journal of Hydrology 55(1 4): 3 23. Bracho, R., Powell, T. L., Dore, S., Li, J., Hinkle, C. R., & Drake, B. G. 2008. Environmental and biolo gical controls on water and energy exchange in Florida scrub oak and pine flatwoods ecosystems. J. Geophys. Res. 113(G2): G02004. Brauman, K. A., Daily, G. C., Duarte, T. K. e., & Mooney, H. A. 2007. The Nature and Value of Ecosystem Services: An Overview Highlighting Hydrologic Services. Annual Review of Environment and Resources 32(1): 67 98. Brewer, C. A. & Pickle, L. 2003. Evaluation of methods for classifying epidemiological data on choropleth maps in series. Annals of the Association of American Geo graphers 92(4): 662 681. Brown, A. E., Zhang, L., McMahon, T. A., Western, A. W., & Vertessy, R. A. 2005. A review of paired catchment studies for determining changes in water yield resulting from alterations in vegetation. Journal of Hydrology 310(1 4): 28 61. Brown, A. E., Podger, G. M., Davidson, A. J., Dowling, T. I., & Zhang, L. 2007. Predicting the impact of plantation forestry on water users at local and regional
70 scales: An example for the Murrumbidgee River Basin, Australia. Forest Ecology and Man agement 251(1 2): 82 93. Brown, M. J., Nowak, J., Johnson, T. G., Oswalt, S. N., Chamberlain, J. L., & Barnard, Resource Bulletin SRS 188 Asheville, NC: United States Department of Agriculture Forest Service, Southern Research Station. Brown, S. L. & Schroeder, P. E. 1999. Spatial Patterns of Aboveground Production and Mortality of Woody Biomass for Eastern U.S. Forest. Ecological Applications 9(3): 968 980. Canadell, J. G. & Raupach, M. R. 2008. Managing Forests for C limate Change Mitigation. Science 320(5882): 1456 1457. Carpenter, S. R. 2005. Ecosystems and human well being: scenarios: findings of the Scenarios Working Group : Island Pr. Carr, M. & Zwick, P. 2007. Smart land use analysis: The LUCIS model. : ESRI Press. Carter, D. R. & Jokela, E. J. 2003. Florida's Renewable Forest Resources. Chahine, M. T. 1992. The hydrological cycle and its influence on climate. Nature 359(6394): 373 38 0. Chan, K. M. A., Shaw, M. R., Cameron, D. R., Underwood, E. C., & Daily, G. C. 2006. Conservation Planning for Ecosystem Services. PLoS Biol 4(11): e379. Chapin Iii, F. S. & Matson, P. A. 2011. Principles of terrestrial ecosystem ecology : Springer. Chur kina, G., Running, S. W., Schloss, A. L., & Intercomparison, T. P. O. F. T. P. N. M. 1999. Comparing global models of terrestrial net primary productivity (NPP): the importance of water availability. Global Change Biology 5(S1): 46 55. Clark, K. L., Gholz, H. L., & Castro, M. S. 2004. Carbon Dynamics along a Chronosequence of Slash Pine Plantations in North Florida. Ecological Applications 14(4): 1154 1171. Costanza, R., d'Arge, R., de Groot, R., Farber, S., Grasso, M., Hannon, B., Limb urg, K., Naeem, S., O'Neill, R. V., Paruelo, J., Raskin, R. G., Sutton, P., & van den Belt, M. 1998. The value of the world's ecosystem services and natural capital. Ecological Economics 25(1): 3 15. Coulston, J. W., Reams, G. A., McRoberts, R. E., & Smit h, W. D. 2006. Practical Considerations When Using Perturbed Forest Inventory Plot Locations To Develop Spatial Models: A Case Study. In G. T. Rep (Ed.), Proceedings of the
71 sixth annual forest inventory and analysis symposium Vol. WO 70: 126. Denver, CO: U.S. Department of Agriculture Forest Service. Cropper Jr, W. P. & Gholz, H. L. 1993. Simulation of the carbon dynamics of a Florida slash pine plantation. Ecological Modelling 66(3 4): 231 249. Daily, G. C. 1997. Nature's services : societal dependence o n natural ecosystems Washington: Island Press. Daily, G. C. 2000. Management objectives for the protection of ecosystem services. Environmental Science & Policy 3(6): 333 339. de Groot, R. S., Wilson, M. A., & Boumans, R. M. J. 2002. A typology for t he classification, description and valuation of ecosystem functions, goods and services. Ecological Economics 41(3): 393 408. de Groot, R. S., Alkemade, R., Braat, L., Hein, L., & Willemen, L. 2010. Challenges in integrating the concept of ecosystem servi ces and values in landscape planning, management and decision making. Ecological Complexity 7(3): 260 272. Delphin, S. 2012. Effect of Urbanization and Hurricanes on the Provision of Ecosystem Services in the Lower Suwannee River and Pensacola Bay Watersh eds, Florida, USA. : University of Florida. Dixon, R. K., Brown, S., Houghton, R. A., Solomon, A. M., Trexler, M. C., & Wisniewski, J. 1994. Carbon Pools and Flux of Global Forest Ecosystems. Science 263(5144): 185 190. Douglass, J. E. 1983. The Potential f or W ater Yield Augmentation f rom Forest Management i n t he Eastern United States1 JAWRA Journal of the American Water Resources Association 19(3): 351 358. Dunn, S. M. & Mackay, R. 1995. Spatial variation in evapotranspiration and the influence of land use on catchment hydrology. Journal of Hydrology 171(1 2): 49 73. Dwivedi, P., Alavalapati, J. R. R., Susaeta, A., & Stainback, A. 2009. Impact of carbon value on the profitability of slash pine plantations in the southern United States: an integrated lif e cycle and Faustmann analysis. Canadian Journal of Forest Research 39(5): 990 1000. Egoh, B., Rouget, M., Reyers, B., Knight, A. T., Cowling, R. M., van Jaarsveld, A. S., & Welz, A. 2007. Integrating ecosystem services into conservation assessments: A re view. Ecological Economics 63(4): 714 721. Egoh, B., Reyers, B., Rouget, M., Richardson, D. M., Le Maitre, D. C., & van Jaarsveld, A. S. 2008. Mapping ecosystem services for planning and management. Agriculture, Ecosystems & Environment 127(1 2): 135 140.
72 Egoh, B., Reyers, B., Rouget, M., Bode, M., & Richardson, D. M. 2009. Spatial congruence between biodiversity and ecosystem services in South Africa. Biological Conservation 142(3): 553 562. Egoh, B. N., Reyers, B., Rouget, M., & Richardson, D. M. 2 011. Identifying priority areas for ecosystem service management in South African grasslands. Journal of Environmental Management 92(6): 1642 1650. Ewel, K. C. & Gholz, H. L. 1991. A Simulation Model of the Role of Belowground Dynamics in a Florida Pine P lantation. Forest Science 37(2): 397 438. Falkowski, P., Scholes, R. J., Boyle, E., Canadell, J., Canfield, D., Elser, J., Gruber, N., Hibbard, K., Hgberg, P., Linder, S., Mackenzie, F. T., Moore Iii, B., Pedersen, T., Rosenthal, Y., Seitzinger, S., Smet acek, V., & Steffen, W. 2000. The Global Carbon Cycle: A Test of Our Knowledge of Earth as a System. Science 290(5490): 291 296. Farley, K. A., Jobbgy, E. G., & Jackson, R. B. 2005. Effects of afforestation on water yield: a global synthesis with implica tions for policy. Global Change Biology 11(10): 1565 1576. Fassnacht, K. S., Gower, S. T., MacKenzie, M. D., Nordheim, E. V., & Lillesand, T. M. 1997. Estimating the leaf area index of North Central Wisconsin forests using the landsat thematic mapper. Rem ote Sensing of Environment 61(2): 229 245. Friedlingstein, P., Dufresne, J. L., Cox, P. M., & Rayner, P. 2003. How positive is the feedback between climate change and the carbon cycle? Tellus B 55(2): 692 700. Gerten, D., Schaphoff, S., Haberlandt, U., L ucht, W., & Sitch, S. 2004. Terrestrial vegetation and water balance hydrological evaluation of a dynamic global vegetation model. Journal of Hydrology 286(1 4): 249 270. Gholz, H. L. 1982. Environmental Limits on Aboveground Net Primary Production, Leaf Area, and Biomass in Vegetation Zones of the Pacific Northwest. Ecology 63(2): 469 481. Gholz, H. L., Vogel, S. A., Cropper, W. P., Jr., McKelvey, K., Ewel, K. C., Teskey, R. O., & Curran, P. J. 1991. Dynamics of Canopy Structure and Light Interception in Pinus Elliottii Stands, North Florida. Ecological Monographs 61(1): 33 51. Gholz, H. L. & Clark, K. L. 2002. Energy exchange across a chronosequence of slash pine forests in Florida. Agricultural and Forest Meteorology 112(2): 87 102. Gimona, A. & van d er Horst, D. 2007. Mapping hotspots of multiple landscape functions: a case study on farmland afforestation in Scotland. Landscape Ecology 22(8): 1255 1264.
73 Gonzalez Benecke, C. A., Jokela, E. J., & Martin, T. A. 2012. Modeling the effects of stand develo pment, site quality, and silviculture on leaf area index, litterfall, and forest floor accumulations in loblolly and slash pine plantations. Forest Science : 1. Gordon, L. J., Peterson, G. D., & Bennett, E. M. 2008. Agricultural modifications of hydrologica l flows create ecological surprises. Trends in Ecology & Evolution 23(4): 211 219. Grier, C. G. & Running, S. W. 1977. Leaf Area of Mature Northwestern Coniferous Forests: Relation to Site Water Balance. Ecology 58(4): 893 899. Groot, J. C. J., Rossing, W. A. H., Jellema, A., Stobbelaar, D. J., Renting, H., & Van Ittersum, M. K. 2007. Exploring multi scale trade offs between nature conservation, agricultural profits and landscape quality A methodology to support discussions on la nd use perspectives. Agriculture, Ecosystems & Environment 120(1): 58 69. GrNzweig, J. M., Lin, T., Rotenberg, E., Schwartz, A., & Yakir, D. 2003. Carbon sequestration in arid land forest. Global Change Biology 9(5): 791 799. Guo, Z., Xiao, X., & Li D. 2000. An Assessment o f Ecosystem S ervices: Water Flow Regulation a nd Hydroelectric Power Production Ecological Applications 10(3): 925 936. Heal, G., Daily, G. C., Ehrlich, P. R., & Salzman, J. 2001. Protecting natural capital through ecosystem serv ice districts. Stan. Envtl. LJ 20: 333. Heath, L. S., Smith, J. E., Woodall, C. W., Azuma, D. L., & Waddell, K. L. 2011. Carbon stocks on forestland of the United States, with emphasis on USDA Forest Service ownership. Ecosphere 2(1): art6. Hendry, L. C. & Gholz, H. L. 1986. Aboveground Phenology in North Florida Slash Pine Plantations. Forest Science 32(3): 779 788. Hewlett, J. D. & Hibbert, A. R. 1961. Increases i n Water Yield a fter Several Types o f Forest Cutting. International Association of Scientif ic Hydrology. Bulletin 6(3): 5 17. Hoover, C. M., Birdsey, R. A., Heath, L. S., & Stout, S. L. 2000. How to Estimate Carbon Sequestration on Small Forest Tracts. Journal of Forestry 98(9): 13 19. Iiames, J. S., Congalton, R. G., Pilant, A. N., & Lewis, T E. 2008. Leaf Area Index (LAI) Change Detection Analysis on Loblolly Pine (Pinus taeda) Following Complete Understory Removal. Photogrammetric Engineering and Remote Sensing 74(11): 1389 1400.
74 Jackson, R. B., Jobbgy, E. G., Avissar, R., Roy, S. B., Bar rett, D. J., Cook, C. W., Farley, K. A., le Maitre, D. C., McCarl, B. A., & Murray, B. C. 2005. Trading Water for Carbon with Biological Carbon Sequestration. Science 310(5756): 1944 1947. Jenkins, J. C., Chojnacky, D. C., Heath, L. S., & Birdsey, R. A. 2 003. National Scale Biomass Estimators for United States Tree Species. Forest Science 49(1): 12 35. Jensen, R. R. & Binford, M. W. 2004. Measurement and comparison of Leaf Area Index estimators derived from satellite remote sensing techniques. Internation al Journal of Remote Sensing 25(20): 4251 4265. Jerome K, V. 2009. Managing water use from forest plantations. Forest Ecology and Management 257(2): 385 389. Jindal, R., Swallow, B., & Kerr, J. 2008. Forestry based carbon sequestration projects in Africa : Potential benefits and challenges. Natural Resources Forum 32(2): 116 130. Kareiva, P., Heather, T., H., R. T., C., D. G., & Stephen, P. 2011. Natural Capital New York: Oxford University Press Inc. Kearns, C. A., Inouye, D. W., & Waser, N. M. 1998. End angered Mutualisms: The Conservation of Plant Pollinator Interactions. Annual Review of Ecology and Systematics 29: 83 112. Kirschbaum, M. U. F. 2003. Can Trees Buy Time? An Assessment of the Role of Vegetation Sinks as Part of the Global Carbon Cycle. Cl imatic Change 58(1): 47 71. Knowles, L. J. 1996. Estimation of evapotranspiration in the Rainbow Springs and Silver Springs basins in north central Florida Tallahassee, Florida: U.S. Geological SurveyWater Resources Investigations Report 96 4024. Kremen, C. 2005. Managing ecosystem services: what do we need to know about their ecology? Ecology Letters 8(5): 468 479. Kucharik, C. J., Foley, J. A., Delire, C., Fisher, V. A., Coe, M. T., Lenters, J. D., Young Molling, C., Ramankutty, N., Norman, J. M., & Go wer, S. T. 2000. Testing the performance of a dynamic global ecosystem model: Water balance, carbon balance, and vegetation structure. Global Biogeochem. Cycles 14(3): 795 825. Lal, R. 2004. Soil Carbon Sequestration Impacts on Global Climate Change and F ood Security. Science 304(5677): 1623 1627. Liu, S., Riekerk, H., & Gholz, H. L. 1998. Simulation of evapotranspiration from Florida pine flatwoods. Ecological Modelling 114(1): 19 34.
75 McLaughlin, D., Kaplan, D., & Cohen, M. 2012. Managing Forests for In creased Regional Water Yield. Journal of the American Water Resources Association McNab, W. H., Cleland, D. T., Freeouf, J. A., Keys Jr, J. E., Nowacki, G. J., & Carpenter, C. A. 2005. Description of ecological subregions: sections of the conterminous Uni ted States. US Department of Agriculture Forest Service, Washington, DC Mennis, J. & Liu, J. W. 2005. Mining Association Rules in Spatio Temporal Data: An Analysis of Urban Socioeconomic and Land Cover Change. Transactions in GIS 9(1): 5 17. Milly, P. C. D. 1997. Sensitivity of greenhouse summer dryness to changes in plant rooting characteristics. Geophys. Res. Lett. 24(3): 269 271. Mitchell, A. 2009. The ESRI guide to GIS Analysis: Spatial Measurements & Statistics Monica, G. 1998. Ecological economic model for optimal mangrove trade off between forestry and fishery production: comparing a dynamic optimization and a simulation model. Ecological Modelling 112(2 3): 131 150. Myneni, R. B., Ramakrishna, R., Nemani, R., & Runnin g, S. W. 1997. Estimation of global leaf area index and absorbed par using radiative transfer models. Geoscience and Remote Sensing, IEEE Transactions on 35(6): 1380 1393. Nelson, E., Mendoza, G., Regetz, J., Polasky, S., Tallis, H., Cameron, D., Chan, K. M. A., Daily, G. C., Goldstein, J., Kareiva, P. M., Lonsdorf, E., Naidoo, R., Ricketts, T. H., & Shaw, M. 2009. Modeling multiple ecosystem services, biodiversity conservation, commodity production, and tradeoffs at landscape scales. Frontiers in Ecology and the Environment 7(1): 4 11. Nordhaus, W. D. 2007. A Review of the "Stern Review on the Economics of Climate Change". Journal of Economic Literature 45(3): 686 702. Polasky, S., Nelson, E., Camm, J., Csuti, B., Fackler, P., Lonsdorf, E., Montgomery, C ., White, D., Arthur, J., Garber Yonts, B., Haight, R., Kagan, J., Starfield, A., & Tobalske, C. 2008. Where to put things? Spatial land management to sustain biodiversity and economic returns. Biological Conservation 141(6): 1505 1524. Raudsepp Hearne, C ., Peterson, G. D., & Bennett, E. M. 2010. Ecosystem service bundles for analyzing tradeoffs in diverse landscapes. Proceedings of the National Academy of Sciences 107(11): 5242 5247. Rodrguez, J. P. 2006. Trade offs across Space, Time, and Ecosystem Ser vices. Ecology and society 11(1): 28.
76 Running, S. W. & Coughlan, J. C. 1988. A general model of forest ecosystem processes for regional applications I. Hydrologic balance, canopy gas exchange and primary production processes. Ecological Modelling 42(2): 125 154. Running, S. W., Nemani, R. R., Peterson, D. L., Band, L. E., Potts, D. F., Pierce, L. L., & Spanner, M. A. 1989. Mapping Regional Forest Evapotranspiration and Photosynthesis by Coupling Satellite Data with Ecosystem Simulation. Ecology 70(4): 10 90 1101. Sahin, V. & Hall, M. J. 1996. The effects of afforestation and deforestation on water yields. Journal of Hydrology 178(1 4): 293 309. Scheffer, M., Holmgren, M., Brovkin, V., & Claussen, M. 2005. Synergy between small and large scale feedbacks o f vegetation on the water cycle. Global Change Biology 11(7): 1003 1012. Schwenk, W. S., Donovan, T. M., Keeton, W. S., & Nunery, J. S. 2012. Carbon storage, timber production, and biodiversity: comparing ecosystem services with multi criteria decision an alysis. Ecological Applications 22(5): 1612 1627. Shan, J., Morris, L. A., & Hendrick, R. L. 2001. The Effects of Management on Soil and Plant Carbon Sequestration in Slash Pine Plantations. Journal of Applied Ecology 38(5): 932 941. Skiles, J. W. & Hans on, J. D. 1994. Responses of arid and semiarid watersheds to increasing carbon dioxide and climate change as shown by simulation studies. Climatic Change 26(4): 377 397. Stephenson, N. L. 1990. Climatic Control of Vegetation Distribution: The Role of the Water Balance. The American Naturalist 135(5): 649 670. Subak, S. 2002. Forest certification eligibility as a screen for CDM sinks projects. Climate Policy 2(4): 335 351. Tallis, H. & Polasky, S. 2009. Mapping and Valuing Ecosystem Services as an Approac h for Conservation and Natural Resource Management. Annals of the New York Academy of Sciences 1162(1): 265 283. Tallis, H. T., Ricketts, T., Guerry, A. D., Nelson, E., Ennaanay, D., Wolny, S., Olwero, N., Vigerstol, K., Pennington, D., Mendoza, G., Aukem a, J., Foster, J., Forrest, J., Cameron, D., Lonsdorf, E., Kennedy, C., Verutes, G., Kim, C. K., Guannel, G., Guide, The Natural Capital Project Stanford. Tilman, D., Knops, J., Wedin, D., Reich, P., Ritchie, M., & Siemann, E. 1997. The Influence of Functional Diversity and Composition on Ecosystem Processes. Science 277(5330): 1300 1302.
77 Timilsina, N., Escobedo, F. J., Cropper Jr, W. P., Abd Elrahman, A., Brandeis, T. J., De lphin, S., & Lambert, S. 2012. A framework for identifying carbon hotspots and forest management drivers. Journal of Environmental Management Tol, R. S. J. 2009. The Economic Effects of Climate Change. The Journal of Economic Perspectives 23(2): 29 51. T roy, A. & Wilson, M. A. 2006. Mapping ecosystem services: Practical challenges and opportunities in linking GIS and value transfer. Ecological Economics 60(2): 435 449. van Dijk, A. I. J. M. & Keenan, R. J. 2007. Planted forests and water in perspective. Forest Ecology and Management 251(1 2): 1 9. Vitousek, P. M., Mooney, H. A., Lubchenco, J., & Melillo, J. M. 1997. Human Domination of Earth's Ecosystems. Science 277(5325): 494 499. Vose, J. M., Dougherty, P. M., Long, J. N., Smith, F. W., Gholz, H. L., & Curran, P. J. 1994. Factors Influencing the Amount and Distribution of Leaf Area of Pine Stands. Ecological Bulletins (43): 102 114. Watson, F. G. R., Vertessy, R. A., & Grayson, R. B. 1999. Large scale modelling of forest hydrological processes and thei r long term effect on water yield. Hydrological Processes 13(5): 689 700. Woudenberg, S. W., Conkling, B. L., O'Connell, B. M., LaPoint, E. B., Turner, J. A., & Waddel, K. L. 2010. The Forest Inventory and Analysis Database: Database s Manual Version 4.0 for Phase 2. In F. S. U.S. Department of Agriculture, Rocky Mountain Research Station. (Ed.). Fort Collins, CO. Zhang, L., Dawes, W. R., & Walker, G. R. 2001. Response of mean annual evapotranspiration to vegetation changes at catchmen t scale. Water Resour. Res. 37(3): 701 708.
78 BIOGRAPHICAL SKETCH Ronald Cademus was born in Maniche Haiti H e graduated from Universidad ISA in Dominican Republic in 2008 with a bachelor degree in f orestry, where he learned the importance of the imp ortance of ecosystem services for a more sustainable world. Upon graduation Ronald went back to Haiti and joined the Fondation Seguin, the leading national environmental NGO in Haiti, working La Visite National Park. During 2 years Ronald led a conservati on project where he trained local farmers in tree nursery and soil conservation. In June 2010, he was awarded a 2 scholarship grant from the USAID/WINNER Project in Ha iti for his master program here at University of Florida. He received his master from the Interdisciplinary Ecology program from the School of Natural Resources and Environment (SNRE) under the direction of Dr. Francisco Escobedo.