Daley Dissertation 033007

BOSTON UNIVERSITY GRADUATE SCHOOL OF ARTS AND SCIENCES Dissertation CHANGES IN ECOHYDROLOGICAL FUNCTION DUE TO THE LOSS AND REPLACEMENT OF EASTERN HEMLOCK IN A NEW ENGLAND FOREST

by

MICHAEL J. DALEY

B.S., Siena College, 1998 M.S., SUNY Plattsburgh, 2000

Submitted in partial fulfillment of the requirements for the degree of Doctor of Philosophy 2007

© Copyright by MICHAEL J. DALEY 2007

Approved by

First Reader

__________________________________________________ Nathan G. Phillips, PhD Associate Professor of Geography and Environment Boston University

Second Reader __________________________________________________ Mark Friedl, PhD Associate Professor of Geography and Environment and Chairman, Department of Geography and Environment Boston University

Third Reader

__________________________________________________ Guido D. Salvucci, PhD Professor of Earth Sciences, Geography and Environment and Chairman, Department of Earth Sciences Boston University

ACKNOWLEDGMENTS The process of researching and writing a dissertation is an exciting and challenging experience. Overcoming the many challenges encountered is only possible with the help, guidance, and support of others. Their support has made completion of this dissertation not only possible, but also an enjoyable and rewarding process. Foremost, I would like to thank my advisor Nathan Phillips. The individual attention I have received over the past five years has been remarkable and for that I am extremely grateful. Nathan’s energy for ecosystem research is contagious and spending time in the field with him is a tremendous experience. I also appreciate the time spent in the car with Nathan traveling between Boston and Harvard Forest. During these trips we discussed ecosystem research and graduate study and I value the guidance and advice Nathan provided. The attention I have received from Nathan is above all expectations and a huge reason for my success at Boston University. I would like to thank my family. My wife, Samantha, is a very special person who I am lucky to have in my life. She has humored me throughout and even served as a “research assistant” building sap flux sensors, editing papers, providing statistical help, and most importantly swatting mosquitoes off my back while I wired up equipment in the field. I cannot express enough gratitude to her. I also wish to acknowledge the love and support I have received from Mom and Dad. Their encouragement throughout the years has meant so much. They have always supported my interest for science, whether it was for frogs or volcanoes in elementary school, fish in college, or now trees. I thank Pat and Chuck for engaging in conversation about my research and sharing stories from their PhD iv

experiences. And I acknowledge the support I have received from all of my siblings and recognize that I could not have completed this process without them. I would like to acknowledge the National Science Foundation Graduate Research Fellowship Program. The financial support provided through this program has provided the opportunity to pursue my graduate research. NSF funding also supported me as a research assistant in Nathan’s lab and provided the opportunity to travel to ESA conventions. Harvard Forest is an exciting place where I am fortunate to have had the opportunity to conduct research. The people at Harvard Forest have been so helpful and for that I am extremely thankful. First, I thank Julian Hadley. Julian has been a tremendous help in my research and I have enjoyed the opportunity to work with him. From sharing eddy covariance data to providing advice, Julian has been invaluable to my research. Also at Harvard Forest, I would like to thank individuals who have provided assistance with equipment and facilities. I thank Michael Scott and Lucas Griffith for their help with Bucky. I thank Paul Kuzeja for his help in the field and his encouragement. I also thank Laurie Chiasson and Edythe Ellin for their help in navigating the facilities. Special thanks go to Mark Friedl and Guido Salvucci from the Department of Geography and Environment. They have served on my committee from the early days. I appreciate the suggestions and advice they have provided through this process. I also thank Adrien Finzi from the Department of Biology. Adrien provided a unique perspective and I always appreciate the opportunity to discuss research with him. v

There are many fellow students who have made my experience at Boston University so enjoyable. I am extremely grateful to have had the opportunity to work with Cory Pettijohn. I had the fortune of spending countless hours with Cory in the woods at Harvard Forest. Cory has been extremely helpful to my research. He has helped establish my field sites, maintain my equipment, analyze my data, and edit my work. Most importantly, Cory has been a great colleague and friend. I would also like to acknowledge the folks in the plant journal club at Boston University including Vicki, Abe, Jenny, Eddie, Heidi, and Anne. I have appreciated the opportunity to spend time discussing the plant ecology field with you all and sharing stories from our experiences. I thank Steve, Carson, and Brenda. My first year at Boston University was intense and you all provided the support and friendship needed. Thank you. As I start my next adventure I wish to thank the individuals at Boston University who have provided help along the way. Thank you Tsia, Laura, Mike, Victoria, and Stephanie. I appreciate all you have done for me.

vi

CHANGES IN ECOHYDROLOGICAL FUNCTION DUE TO THE LOSS AND REPLACEMENT OF EASTERN HEMLOCK IN A NEW ENGLAND FOREST (Order No.

)

MICHAEL J. DALEY Boston University Graduate School of Arts and Sciences, 2007 Major Professor: Nathan G. Phillips, Associate Professor of Geography and Environment ABSTRACT The disturbance currently occurring across the northeastern United States from the invasive pest hemlock woolly adelgid (Adelges tsugae) provides a unique opportunity to study the impact of community composition on ecosystem function. The spread of hemlock woolly adelgid has resulted in the replacement of eastern hemlock (Tsuga canadensis), an evergreen climax species, by deciduous seral species, mainly black birch (Betula lenta). Ecosystem water cycling has a large potential to be impacted by replacement of hemlock, and was the focus of this research. Physiological characteristics of individual tree species were found to influence fluxes of water between vegetation and the atmosphere. In Chapter 2, I show that the response to step changes in environmental forces that drive transpiration are faster in black birch compared to eastern hemlock largely due to differences in capacitance. The average time constant in black birch was 8.6 minutes compared to 15.2 minutes in vii

hemlock. Chapter 3 describes a study that shows nighttime transpiration present in paper birch (Betula papyrifera) but absent in neighboring trees of other species. Over 10% of total daily sap flux was due to transpiration at night in paper birch. Whole-tree water use was scaled to the stand and ecosystem level in Chapters 4 and 5 to estimate the effects of replacement of eastern hemlock by black birch on ecosystem water use and streamflow. Differences were observed during the peak growing season as daily transpiration rates were 1.6 times greater in black birch compared to hemlock. Cumulative transpiration in black birch exceeded hemlock transpiration by 63 mm from June until October. A large impact on streamflow, due to enhanced transpiration, will occur during this period and model simulations suggest significant differences in August flows with species replacement. The results of this research suggest that shifts in physiology as stands transition from being hemlock dominated to being black birch dominated will alter the hydrologic function of the ecosystem. The results highlight the significance of tree physiology and the seasonality of leaf cover in regulating ecosystem water use and streamflow.

viii

TABLE OF CONTENTS Approval Page...................................................................................................................iii Acknowledgments.............................................................................................................iv Abstract.............................................................................................................................vii Table of Contents...............................................................................................................ix List of Tables.....................................................................................................................xi List of Figures...................................................................................................................xii 1.1 Goals of the Current Study...................................................................................3 1.2 Dissertation Structure............................................................................................4 References....................................................................................................................6 2.1 Introduction..........................................................................................................10 2.2 Methods...............................................................................................................13 2.2.1 Study Site and Species.................................................................................13 2.2.2 Experimental Treatments.............................................................................14 2.2.3 Estimation of R and C..................................................................................15 2.2.4 Wavelet Analysis..........................................................................................16 2.2.5 Diameter Variation ......................................................................................18 2.3 Results.................................................................................................................18 2.3.1 Estimates of τ...............................................................................................18 2.3.2 R, C, τ Model...............................................................................................19 2.3.3 Response to Environment............................................................................20 2.3.4 Wavelet Analysis..........................................................................................21 2.4 Discussion...........................................................................................................21 References..................................................................................................................26 3.1 Introduction.........................................................................................................45 3.2 Methods...............................................................................................................47 3.2.1 Study Site and Species.................................................................................47 3.2.2 Instrumentation and Sampling.....................................................................49 3.2.3 Partitioning into Recharge and Nighttime Transpiration.............................50 3.2.4 Nighttime Gas Exchange Measurements.....................................................51 3.3 Results.................................................................................................................53 3.4 Discussion...........................................................................................................56 3.4.1 Hydrologic Significance of Nighttime Transpiration..................................58 3.4.2 Ecophysiological Differences......................................................................61 3.5 Acknowledgments...............................................................................................64 References..................................................................................................................65 ix

4.1 Introduction.........................................................................................................82 4.2 Materials and Methods........................................................................................84 4.2.1 Study Sites...................................................................................................84 4.2.2 Transpiration Measurements........................................................................85 4.2.3 Data Analysis...............................................................................................88 4.2.4 Evapotranspiration measurements...............................................................90 4.3 Results.................................................................................................................93 4.3.1 Annual Evapotranspiration...........................................................................93 4.3.2 Transpiration................................................................................................93 4.4 Discussion...........................................................................................................95 4.5 Acknowledgments.............................................................................................100 References................................................................................................................101 Appendix A..............................................................................................................116 5.1 Introduction.......................................................................................................119 5.2 Methods.............................................................................................................120 5.2.1 Data ...........................................................................................................121 5.2.2 Precipitation Interval File..........................................................................122 5.2.3 Location ....................................................................................................122 5.2.4 Canopy.......................................................................................................123 5.2.5 Fixed Parameters........................................................................................124 5.2.6 Soil and Flow Parameters..........................................................................124 5.2.7 Transpiration and Evapotranspiration Estimates.......................................125 5.2.8 Flow Estimates...........................................................................................126 5.2.9 Sensitivity Plots.........................................................................................126 5.3 Results ..............................................................................................................126 5.4 Discussion.........................................................................................................129 References................................................................................................................133 Appendix A..............................................................................................................150 Appendix B..............................................................................................................154 Concluding Remarks................................................................................................160 References................................................................................................................163 References.......................................................................................................................164 Curriculum Vita..............................................................................................................186

x

LIST OF TABLES Table 1.1. Physiological characteristics of early- and late-successional plants..................8 Table 2.2 The number, average height, and average diameter at breast height of the trees measured in each stand......................................................................................................32 Table 2.3 The time constant of trees as determined using the misting step function technique............................................................................................................................33 Table 2.4 The time constant, resistance, and capacitance of sampled eastern hemlock and black birch trees. Time constant was determined using the misting step function technique. ..........................................................................................................................34 Table 3.5 The diameter at breast height (DBH), height, sapwood area, and sapwood depth of the trees selected for study at Harvard Forest, MA. ..........................................72 Table 3.6 Summary of nighttime transpiration and recharge during the 2003 growing season. Night Transpiration and Recharge are calculated as the mean percent of total daily flux for the growing season. ND = not detectable...................................................73 Table 4.7 The percent of basal area 0-100 m from the hemlock and deciduous eddy covariance towers............................................................................................................108 Table 4.8 Diameter at breast height (DBH), sapwood depth, sapwood area, and projected crown area of the trees selected for study at Harvard Forest, MA..................................109

xi

LIST OF FIGURES Figure 1.1 Distribution of hemlock woolly adelgid infestations in 2005 as reported by the USDA Forest Service. Source: From (USDA Forest Service 2007)..................................9 Figure 2.2 An example of a misting treatment in a black birch tree (B). Sap flux in a nontreated tree is shown for reference (A)..............................................................................37 Figure 2.3 An example of the exponential decay after applying a misting treatment in a black birch tree. Closed circles represent sap flux in the misted tree and open circles are a reference tree.....................................................................................................................38 Figure 2.4 The lag in sap flux (normalized by daily maximum sap flux rate) between sensors at the base of the live crown (open) and at breast height (closed) in black birch (A) and eastern hemlock (B)..............................................................................................39 Figure 2.5 Resistance and capacitance as functions of sapwood area in black birch (A) and eastern hemlock (B). It is important to note that capacitance in not an independent measurement. Open circles represent resistance and closed circles capacitance..............40 Figure 2.6 The response of a set of eastern hemlock (C) and black birch (D) trees to a sudden change in environmental conditions including vapor pressure deficit (A) and photosynthetically active radiation (B)..............................................................................41 Figure 2.7 The average response in the set of eastern hemlock and black birch to the environmental changes in Fig. 2.5. The top line represents the response in eastern hemlock and the bottom in black birch..............................................................................42

xii

Figure 2.8 The effect of sapwood area on the response time of trees to sudden changes in environmental conditions. Closed circles are black birch and open triangles are hemlock. Error bars represent one standard error..............................................................................43 Figure 2.9 Continuous wavelet transform in eastern hemlock (A) and black birch (B). The corresponding sap flux time series for eastern hemlock (grey line) and black birch (black line) are shown in C. The thick black line contours represents areas at the 5% significance level above a background spectrum..............................................................44 Figure 3.10 Sap flux at breast height, stomatal conductance, and transpiration during the night of September 1 in paper birch (closed circles), red oak (closed triangles), and red maple (open circles). Night is represented by the shaded area. The sap flux values reported for each species are the mean of 3 trees. Conductance and transpiration were measured at the leaf level..................................................................................................75 Figure 3.11 Diurnal sap flux at breast height for paper birch (closed circle), red oak (closed triangle) and red maple (open circle) for the dates during which leaf gas exchange was measured (1 Sept. to 2 Sept.). The sap flux values reported for each species are the mean of 3 trees. Shaded areas represent night. Vapor pressure deficit (D) (dashed line), and wind speed (solid line) for these dates are also shown...............................................76 Figure 3.12 Time course of sap flux in a sample tree of paper birch (A), red oak (C), and red maple (E) on September 1, 2004. Gray lines represent the normalized flux at the base of live crown and black lines represent sap flux at breast height. The flux at the base of the live crown was normalized to equal the 24-h sum of sap flux at breast height. The depletion and recharge of stored water in the bole is shown for the paper birch (B), red xiii

oak (D), and red maple tree (F). The water storage flux is calculated as the difference between flux at the base of live crown and the flux at breast height. Periods greater than zero indicate periods of recharge and negative values indication periods of water withdrawal.........................................................................................................................77 Figure 3.13 Daily pattern of water use as measured at breast height in paper birch, red oak, and red maple from days 234 to 238 (22 August to 26 August 2003). The sap flux values reported for each species are the mean of 3 trees. Nighttime on days 235 and 236 had high rates of sap flux that remained elevated through the night in paper birch. Also shown are patterns in the raw sap flux signal and environmental driving forces for transpiration including solar radiation and vapor pressure deficit (D). The corrected values in red oak account for the proportion of sensors not in conducting sapwood (Clearwater et al. 1999).....................................................................................................78 Figure 3.14 Sensitivity of sap flux measured at the base of the live crown in paper birch (A), red oak (B), and red maple (C) to vapor pressure deficit (D) during August 2003. Sensitivity at night (open circles) and during the day (closed circles) is shown. The best fit exponential rise to maximum curve (y = a(1-e-bx)) is shown for both daytime (black line) and nighttime (gray line)...........................................................................................80 Figure 3.15 Mean sap flux at breast height in paper birch (closed circle), red oak (closed triangle), and red maple (open circle) the night of August 16. Paper birch showed sensitivity to the increase in vapor pressure deficit (D) (thick line) while red oak and red maple had no response.......................................................................................................81

xiv

Figure 4.16 Seasonal patterns of (A) air temperature, (B) vapor pressure deficit (D), (C) photosynthetically active radiation (PAR), (D) transpiration per projected crown area and (E) evapotransiration (ET). Points represent mean daily values averaged for one week intervals. Black birch daily transpiration and ET are represented by closed circles and eastern hemlock by open. Error bars represent one standard error for transpiration and one standard deviation for ET. Plotted points represent weekly means..........................110 Figure 4.17 (A) Cumulative transpiration per projected crown area in black birch (closed) and eastern hemlock (open) from June to October. Triangles represent data filled using multiple regression models and circles represent measurements using sap flux sensors. (B) Cumulative evapotranspiration in the deciduous (closed) and eastern hemlock (open) stands based on estimates using the eddy covariance technique. Cumulative precipitation is also shown as a solid line....................................................111 Figure 4.18 Daily patterns in transpiration per projected crown area in black birch (closed) and eastern hemlock (open) during a period of the peak growing season. Error bars represent +/- 1 standard error...................................................................................112 Figure 4.19 Diurnal pattern of transpiration rates expressed per projected crown area in black birch and eastern hemlock during the peak growing season (21 June), late growing season (29 August), and fall (9 October). Black birch transpiration is represented by closed circles and eastern hemlock by open. Error bars represent one standard error.....113 Figure 4.20 The relationship between black birch and eastern hemlock daily transpiration during three periods of the growing season. During the peak growing season, the slope of the relationship is 1.6 (r2=0.89) while during the late growing season the slope is 0.99 xv

(r2=0.95) and 0.64 (r2=0.64) in the fall. Fitted lines represent best-fit linear regressions. .........................................................................................................................................114 Figure 4.21 Changes in transpiration per unit of photosynthetically active radiation (PAR) during the growing season. The best-fit curve (Sigmoidal) was found using the curve fitting function in SigmaPlot..................................................................................115 Figure 0.22 Mean daily solar radiation (A), maximum (closed) and minimum (open) air temperature (B) and vapor pressure deficit (C) by month for five years at Harvard Forest. Also shown, average monthly precipitation (D)..............................................................139 Figure 0.23 Estimated relative leaf area index by day of year for the eastern hemlock (open) and deciduous stand (closed) used in the simulation...........................................140

xvi

1

INTRODUCTION The interaction of vegetation and its physical environment can affect processes from the ecosystem to global scales (Bonan 1992, Foley et al. 2003). Vegetation exerts strong controls on the exchange of energy and mass between the terrestrial land surface and the atmosphere. Vegetation may function as a sink for rising global CO2 levels (Woodwell et al. 1998), alter the partitioning of energy to latent and sensible heat (Avissar and Pielke 1991), increase watershed evapotranspiration, and reduce water yield (Swank and Douglas 1974). As human activities increasingly alter natural processes, understanding the feedbacks between vegetation and its physical environment are critical. Extensive research has focused on global changes such as increases to atmospheric carbon dioxide and alterations to nitrogen cycling on the interaction between vegetation and the physical environment. These global change processes, while certainly significant in their potential impacts and worthy of such extensive attention, are predicted to take several decades to affect ecosystem processes; other mechanisms of global change that are less well understood are having an immediate effect. One mechanism, biological invasions, is now so widespread it must be considered a component of global environmental change (Vitousek 1997). The list of biological invaders across the globe is long and continues to grow. In the continental United States, 11.5% of the plant species are invasive while in Canada this number is closer to 22% (Vitousek 1997). The invasive plant species are not evenly distributed as some areas have a higher percentage of established plant invaders. For

2 example, in Hawaii, 42% of the plant species are invasive while only 9.9% are in Texas (Vitousek 1997). Even more alarming than invasive plants is the number of invasive pests and pathogens that threaten forest health. Mountain pine beetle, emerald ash borer, white pine blister rust, and asian longhorn beetle are just a few of the hundreds of invasives infesting North America. Biological invasions, whether from a plant or animal, can have a significant effect on ecosystem processes and dynamics. In the United States, many invasive pests and pathogens have altered ecosystem processes over the last century. For example, the exotic pathogen chestnut blight (Cryphonectria parasitica) eliminated American chestnut (Castanea dentata) from northeastern forests during the early 1900s. It is well understood that killed trees were replaced by oak and other hardwood species, but very little is known of the response of ecosystem functions to this replacement . One current biological invader, hemlock woolly adelgid, provides a unique opportunity to study the impact of community structure change on ecosystem processes. Hemlock woolly adelgid is an aphid-like pest, native to Asia that is currently infesting eastern hemlock trees (Tsuga candensis) across the northeastern United States (Fig. 1.1). Hemlock woolly adelgid sucks on the sap of needles of eastern hemlock near the point of needle attachment . This feeding, along with the injection of a toxic saliva, causes needles to drop due to dessication and, as a result, trees can die within four years of infestation . The spread of HWA has resulted in the replacement of eastern hemlock, an evergreen climax species, by forests of varying composition across the East Coast. For example, in Connecticut, black birch (Betula lenta), a deciduous seral species, has been the dominant

3 replacement stand . Other deciduous species such as red oak (Quercus rubra) and red maple (Acer rubrum) are also filling the gap left by hemlock mortality in Connecticut . As shown in Table 1.1, the replacement of eastern hemlock, a late successional species, by black birch represents an extreme change in the morphological, phenological, and physiological characteristics of the ecosystem. Eastern hemlock is an abundant tree species in its range, occupying over 1 x 106 ha from the southern Appalachians to southern Canada . Given the prevalence of eastern hemlock in ecosystems across the northeast, infestation by hemlock woolly adelgid will have significant impacts on the cycling of water, carbon, and other nutrients through ecosystems.

1.1 GOALS OF THE CURRENT STUDY Shifts in the phenological and physiological characteristics of an ecosystem will affect exchanges of energy, carbon, and water. The magnitude and timing of fluxes are largely regulated by the morphology, phenology, and physiology of each species. This research addressed the influence of the phenology and physiology of individual species on ecosystem water use. Two species in particular, black birch and eastern hemlock, were the focus of this study. Black birch is the dominant replacement species in hemlock stands dying from hemlock woolly adelgid across New England . Differences in physiological attributes are likely to scale to affect ecosystem processes as black birch replaces eastern hemlock stands. The data presented here is intended to present baseline data of water use in eastern hemlock trees and black birch trees. Additionally, this research explores the influence of individual species, with unique physiologies, on

4 ecosystem processes. To examine the effect of species on ecosystem water use, this research will utilize a mechanistic approach in which processes at the whole-tree scale will be integrated to summarize processes that occur at larger spatial scales. To achieve this objective, the research will be divided into three main components: •

Whole-Tree Hydraulics and Physiology

•

Stand Level Water Use

•

Process-based Hydrological Modeling

1.2 DISSERTATION STRUCTURE In Chapter 2, whole-tree hydraulic controls in eastern hemlock and black birch were investigated. By applying a series of environmental perturbations, whole-tree resistances, capacitances, and time constants were determined from time series sap flux data. With knowledge of whole-tree resistance, capacitance and time constants, these hydraulic controls can be extended to examine potential alterations in the cycling of water due to a shift in the composition of a watershed. In Chapter 3, I examine physiological attributes of deciduous species at Harvard Forest and report observations of nighttime water use. The results in this paper highlight the unique physiological characteristics of individual species and how those characteristics regulate exchanges between vegetation and the atmosphere. Whole-tree water use in eastern hemlock and black birch is scaled to the stand level in Chapter 4. In this chapter, I report daily water use in eastern hemlock and black birch across seasons. Additionally, annual evapotranspiration in a hemlock and deciduous

5 stand are compared. This chapter highlights the influence of tree physiology on vegetation water use. Chapter 5 utilized the Brook90 hydrology model to simulate the replacement of eastern hemlock by black birch on monthly flows. Field measurements such as transpiration, stomatal conductance, and leaf area were used to parameterize the model. The model was run to predict changes in monthly and annual streamflow and evapotranspiration as a result of species replacement.

6

REFERENCES Avissar, R. and R.A. Pielke 1991. The impact of plant stomatal control on mesoscale atmospheric circulations. Agricultural and Forest Meteorology 54:353-372. Bonan, G.B., D. Pollard and S.L. Thompson 1992. Effects of boreal rorest vegetation on global climate. Nature 359:716-718. Foley, J.A., M.H. Costa, C. Delire, N. Ramankutty and P. Snyder 2003. Green surprise? How terrestrial ecosystems could affect earth's climate. Frontiers in Ecology and the Environment 1:38-44. Liebhold, A.M., W.L. MacDonald, D. Bergdahl and V. Mastro 1995. Invasion by exotic forest pests: a threat to forest ecosystems. Forest Science Monograph 30:1-49. McClure, M.S. 1991. Density-dependent feedback and population-cycles in adelgesTsugae (Homoptera, adelgidae) on Tsuga-canadensis. Environmental Entomology 20:258-264. McWilliams, W. and T. Schmidt 2000. Composition, structure, and sustainability of hemlock ecosystems in eastern North America. In Proceedings: Symposium on sustainable management of hemlock ecosystems in eastern North America Eds. K. McManus, K. Shields and D. Souto. USDA Forest Service, Durham, NH. Orwig, D.A. and D.R. Foster 1998. Forest response to the introduced hemlock woolly adelgid in southern New England, USA. Journal of the Torrey Botanical Society 125:60-73. Swank, W.T. and J.E. Douglas 1974. Streamflow greatly reduced by converting hardwood stands to pine. Science 185:857-859.

7 USDA Forest Service 2007. Hemlock woolly adelgid infestations 2005. Available from http://www.na.fs.fed.us/fhp/hwa/maps/hwa_2005.jpg [accessed 31 January 2007]. Vitousek, P.M., C.M. Dantonio, L.L. Loope, M. Rejmanek and R. Westbrooks 1997. Introduced species: A significant component of human-caused global change. New Zealand Journal of Ecology 21:1-16. Woodwell, G.M., F.T. Mackenzie, R.A. Houghton, M. Apps, E. Gorham and E. Davidson 1998. Biotic feedbacks in the warming of the earth. Climatic Change 40:495-518. Young, R., K. Shields and G. Berlyn 1995. Hemlock woolly adelgid (Homoptera:Adelgidae): Stylet bundle insertion and feeding sites. Annals of the Entomological Society of America 88:827-835.

8 Table 1.1. Physiological characteristics of early- and late-successional plants.

Attribute Seeds dispersal in time Secondary (induced) dormancy Seed germination enhanced by light fluctuating temperatures high NO-3 concentrations inhibited by far-red light high CO2 Light saturation intensity Light compensation point Efficiency at low light Photosynthetic rates Respiration rates Transpiration rates Stomatal and mesophyll resistances Resistance to water transport Acclimation potential Recovery from resource limitation Ability to compress environmental extremes Physiological response breadth Resource acquisition rates Material allocation flexibility

Early successional plants

Late succession plants

long common

short uncommon?

yes yes yes

no no no?

yes yes high high low high high high low low high fast

no no low low high low low low high high low slow

high broad fast high

low? narrow slow? low?

Source: From Bazzaz 1979; Reprinted, with permission, from the Annual Review of Ecology and Systematics, Volume 10 (c)1979 by Annual Reviews www.annualreviews.org

9

Figure 1.1 Distribution of hemlock woolly adelgid infestations in 2005 as reported by the USDA Forest Service. Source: From (USDA Forest Service 2007).

10

CHANGES IN ECOSYSTEM FUNCTION DUE TO THE LOSS AND REPLACEMENT OF TSUGA CANADENSIS

2.1 INTRODUCTION As each species exerts unique physiological controls on environmental resource use, changes in the structure and composition of ecosystems may scale to influence ecosystem level processes . The structure and composition of ecosystems are increasingly under pressure from factors such as climate change (Cramer et al. 2001), nitrogen deposition (Manning et al. 2006), and invasive pests and pathogens (Ellison et al. 2005). These pressures may alter ecosystem composition and have a significant impact on processes such as the cycling of carbon (Kueppers and Harte 2005, Fahey et al. 2005), nutrients (Finzi et al. 1998) and water (Hornbeck et al. 1993). Here we focus on impacts of ecosystem change due to an invasive pest on water flux. Across the northeastern USA, the invasive pest hemlock woolly adelgid (HWA; Adelges tsugae Annand) is currently altering the structure and composition of forested ecosystems. HWA is an aphid-like pest that feeds on eastern hemlock trees (Tsuga canadensis (L). Carr) and usually results in mortality in 4-7 years (. In large parts of its range, the spread of HWA has resulted in the replacement of eastern hemlock, an evergreen climax species, by deciduous species, such as black birch (Betula lenta L.), a deciduous seral species . This change represents a large change in the phenology, morphology, and physiology of the ecosystem. While others have examined the impact of

11 this ecosystem change on energy and nutrient fluxes (Jenkins et al. 1999; Stadler et al. 2006), in this study we focus on alterations to the water cycle, and in particular, vegetation hydraulic controls of water flux. The physiology of vegetation in an ecosystem, coupled with the environmental forces, determine the magnitude and dynamics of water transport from the soil to the atmosphere. Based on an analogy to Ohm’s law, the physiological controls on water flux within a whole-tree include hydraulic resistance (R) and capacitance (C). R and C together dictate vegetation water flow dynamics. Extending Ohm’s law, the product of R and C determine the time constant (τ) which describes the inertia of water flow within vegetation to environmental perturbations:

τ = R×C

Eqn. 1

With knowledge of R, C, and τ, the dynamic interactions between climate, soil, vegetation, and hydrology may be described more accurately by accounting for lags between atmospheric and vegetation processes. C contributes to homeostasis in the diurnal and seasonal water balance of trees (Sholz et al. 2007) and vegetation τ may be on the same order of magnitude as characteristic rainfall-runoff dynamics . Understanding R, C, and τ dynamics will increase our ability to understand the influence of vegetation on the movement of water through ecosystems. Transient response analysis (Ogata 1990) provides a novel technique to evaluate whole-tree R, C, and τ. In this methodology, the response of transpiration to whole-tree perturbations can be evaluated to extract information on the physiological controls in

12 individual species and tree size classes. Environmental perturbations in the form of step and ramp functions can be applied and the response can be evaluated to determine wholetree τ (Phillips et al. 2004). Estimates of τ coupled with measurements of R or C can be combined to estimate the complementary variable to provide an independent estimate of C or R for comparison with other measurements. The purpose of this study was to evaluate and compare ecohydrologic function of forests dominated by two tree species through estimates of whole-tree R, C, and τ in eastern hemlock and black birch stands. Black birch is expected to be the dominant replacement species in declining hemlock ecosystems across northeastern, USA (Orwig and Foster 1998, Kizlinski et al. 2002, Catovsky and Bazzaz 2000). To gain an understanding of ecohydrological response due to the loss and replacement of eastern hemlock, this research considered whole-tree hydraulic controls while indicating the potential effects of these controls at the ecosystem scale. Environmental perturbations were applied to whole tree systems and transient response analysis was used to derive estimates of hydraulic variables. Based on distinct life histories and physiological characteristics, the controls on water transport exerted by eastern hemlock and black birch are expected to be significantly different. We hypothesized that R, C, and τ are species dependent and the loss of eastern hemlock and replacement by black birch will alter ecosystem ecohydrological function as a result of decreasing ecosystem R, C and τ.

13 2.2 METHODS 2.2.1 Study Site and Species Eastern hemlock and black birch trees were measured at Harvard Forest, Petersham, Massachusetts, United States (42°32’N, 72°10’W, elevation 340 m). Eastern hemlock is one of the most long-lived and shade-tolerant species found in its range while black birch is an early-successional shade intolerant species . Study sites were established in five stands varying in species composition and height (Table 2.1). The stands labeled LYF, RD, and IRR were mixed deciduous stands containing both eastern hemlock and black birch trees. The stand labeled HM was a hemlock dominated stand and LPH was a black birch dominated plot in a mixed deciduous stand. A total of 26 trees were selected for measurement, varying in size and stand location (Table 2.1). To measure water movement in the xylem, Granier style sap flux sensors (Granier 1985) were installed in all trees. In smaller trees, one 20 mm sap flux sensor was installed at breast height in each tree. In larger trees, two 20 mm sap flux sensors were installed on opposite sides. A sap flux profile with depth was created for larger trees as explained in Chapter 4. In addition, one 20 mm sap flux sensor was positioned at the base of the live crown in two trees of each species. Environmental data was obtained from the Fisher Meteorological Station at Harvard Forest. The station was located within 1 km of all stands and equipped with CS500 temperature and relative humidity probes (Campbell Scientific, Logan, UT), a LICor 190 SA quantum sensor to measure photosynthetically active radiation (LI-COR, Lincoln, NE), and a Met One 385 heated rain gage to measure precipitation (Met One, Grant Pass, OR).

14 2.2.2 Experimental Treatments Perturbations in the form of step functions, pulse functions, impulse functions, and ramp functions can be used to analyze whole-tree hydraulic systems (Phillips et al. 2004). In this study, we utilized step and ramp functions to evaluate the whole-tree hydraulic systems. Step Functions Step functions describe a sudden change in the value of an input variable to a system, followed by maintenance of this value until the system reaches a steady state response. In this study, step changes were induced by misting the canopy on clear days (Fig. 2.1) and maintaining the mist until a steady state response was reached (Fig. 2.2). Several techniques were used to apply experimental misting treatments depending on the size and location of the trees. In larger trees, a transportable aerial work platform (Scanlift, Joensuu, Finland) was used in misting treatments. Water was applied using either a pressure washer connected to an electric pump, a gas powered pump, or a backpack firefighting pump. All treatments to smaller trees were done using the backpack firefighting pump. Misting perturbations were applied on clear days to a total of 14 trees monitored with sap flux sensors. Using time series analysis techniques as discussed by , τ was estimated based on sap flux response to these perturbations. A three parameter exponential decay function was fit to the data from the start of the misting treatment to determine τ:

15

y = y 0 + ae

−

t τ

Eqn. 1

where y is sap flux rate, y0 is the initial sap flux rate, and a is the sap flux rate at which steady flux is reached, t is time (minutes), and τ is time constant. Ramp Functions On very clear days, increases in solar radiation produce a nearly ramp-like increase in morning latent heat flux. This signal propagates with a lag through the wholetree hydraulic system to the sensors positioned at breast height. τ can be approximated by the ramp function and the lag between sensors positioned at the base of the live crown and at breast height (Phillips et al. 2004). In this study, τ was evaluated using the ramp function in one black birch (DBH = 27.3 cm) and two eastern hemlock trees (DBH = 44.6 and 45.7 cm) on consecutive clear days in 2004 and 5 consecutive clear days in 2005. 2.2.3 Estimation of R and C At the tree scale, the research determined R, C, and τ of hydraulic flow for eastern hemlock and black birch over a range of sizes. R is a measure of the constrictions on water transport in vegetation and is a function of hydraulic conductivity and stomatal behavior. R was determined by the following: R= (ψ c - ψ s) / E

Eqn. 2

where ψc equals the canopy leaf water potential, ψs equals the soil water potential, and E is the transpiration rate. The transpiration rate (E) was measured using sap flux sensors installed in a sample of five hemlock and six black birch trees. The canopy leaf water

16 potential (ψc) was measured from a sample of representative canopy leaves using a pressure chamber instrument (PMS Instruments, Albany, OR). Soil water potentials (ψs) were obtained by measuring predawn leaf water potential and adjusting for sample height. Hydraulic capacitance (C) represents the temporary storage of water in vegetation and may have an influence on the dynamics of water transport. Once R and τ were estimated, C was determined from: C = τ/R

Eqn. 3

2.2.4 Wavelet Analysis A possible approach towards assessing our estimates of time constants in black birch and eastern hemlock is to decompose the sap flux signal into its corresponding frequency domain representation. In that manner, we can independently analyze the various frequencies that are present in the signal. One of the possible ways to go about this is to perform a Fourier analysis, which breaks down a signal into constituent sinusoids of different frequencies. The drawback with Fourier analysis is that the time information is lost once transformed into the frequency domain. It is difficult to interpret or tell as to when a particular event took place. Wavelet analysis transforms a signal into a two dimensional form that allows for analysis of frequency variations over time. Wavelet transform of a continuous time series can be done using a continuous or a discrete mode (Grinsted et al. 2004, Torrence and Compo 1998). In this study, the continuous wavelet transform is performed over the hemlock and black birch time series to extract the low signal to noise ratio signals. Specifically, “signal” refers to the low

17 frequency variation of the tree sap flux. The low frequency in turn is the diurnal cycle of the time series. “Noise” refers to the high frequency variation of the tree sap flux data and this corresponds to the perturbations induced by environmental factors. The high frequency variation occurs at a time scale in the order of minutes. The continuous wavelet transform localizes the frequency of the signal in the time space and essentially represents localized power or energy content in the time-frequency space. Continuous wavelet transform was performed on three days of sap flux data from neighboring eastern hemlock and black birch tree. Data was manipulated to remove zero flow during the night. Time constants in the selected trees were estimated using the misting technique described above. In wavelet analysis, the Y-axis represents the time period which is almost half the actual time period of the signal. Therefore, the numbers displayed on the Y-axis are half the actual time period of the signal. This is because of the fact, that while taking the wavelet transform, the scales are chosen as fractional powers of two.

Eqn. 4

Eqn. 5

where so is the smallest resolvable scale and J determines the largest scale. The should be chosen so that the equivalent Fourier period is approximately 2 ∂t. The choice of a sufficiently small ∂j depends on the width in the spectral-space of the wavelet function. In

18 our time series, N is the total number of samples (e.g. 3000), ∂t = 1min, ∂j =0.125. Putting the values of N, ∂t, ∂j and =2 ∂t, J can be calculated, which in our case corresponds to a total of 1024 scales ranging from 1min to 3516 min (~ 2.5 days). This value of ∂j appears adequate to provide a smooth picture of wavelet power. 2.2.5 Diameter Variation During misting treatments, the tension in the water column is relaxed and xylem can expand. Xylem diameter variations were monitored in one black birch tree during a misting treatment on 25 August 2005. Independent measurements of sap flux and xylem variation were recorded during the misting treatment. Xylem variation was measured using pressure transducers as described in Sevanto et al. 2001. An exponential decay curve (Eqn. 2) was fit to the xylem variation data to determine τ.

2.3 RESULTS 2.3.1 Estimates of τ Three independent techniques were used to investigate the time constant in eastern hemlock and black birch trees. Utilizing the step function after applied treatments, τ was found to range between 9.4 min. to 24.8 min. in eastern hemlock trees (Table 2.2). Larger time constants were measured in the larger trees in eastern hemlock and diameter at breast height was a significant predictor of τ (p<0.05, R2 = 0.87). In black birch, there was relatively little variation in estimates of τ in the trees measured. τ ranged from 5.9 min. to 10.5 min. (Table 2.2). When controlling for tree size (DBH), species was a significant predictor of time constant (p<0.05, R2 = 0.74). On 25 August 2005, sap flux

19 and diameter variations were independently logged during an applied step function treatment in a black birch tree. Analysis of the sap flux data indicated a τ of 10.5 min. (Tree 9, Table 2.2) compared to 11.2 min. based on analysis of diameter variations (data not shown). And the ramp function was used to evaluate τ on clear mornings. Applying the ramp function, the lag between sensors at the base of the live crown and sensors at breast height revealed a mean time constant of 13.3 min. (SD=7.8) in the black birch tree compared to 34.6 min. (SD=10.4) in the eastern hemlock. A t-test on the equality of means was run and, based on estimates from using the ramp function, τ was significantly different between species (Pr(T < t) = 0.0000, t = -6.2320). The lag between high and low sensors on a clear morning is illustrated in Figure 2.3 for a representative black birch and eastern hemlock tree. 2.3.2 R, C, τ Model Whole-tree R, C and τ were estimated in five eastern hemlock and six black birch trees (Table 2.3). When controlling for sapwood area, species is not a significant predictor C and R. Based on sapwood area, R and C are not significantly different between eastern hemlock and black birch. The natural log of sapwood area is a significant predictor of the natural log of resistance in both black birch (p<0.05, R2 = 0.97) and eastern hemlock ((p<0.05, R2 = 0.97) (Fig. 2.4). In eastern hemlock, R in the largest tree (47.8 cm d.b.h.) was 0.091 h MPa kg-1 while R in the smallest tree (5.4 cm d.b.h.) was 5.70 h MPa kg-1. R in the largest black birch (27.4 cm d.b.h.) was 0.096 h MPa kg-1 while the smallest (4.5 cm d.b.h.) was 7.00 h MPa kg-1. C also reflected tree size in both species as the natural log of sapwood area was a significant predictor of the natural log of

20 capacitance in black birch (F( 1, 4) = 101.69, Prob > F = 0.0005, R-squared = 0.9622) and eastern hemlock (F( 1, 3) = 184.95, Prob > F = 0.0009, R-squared = 0.9840). In eastern hemlock, the C in the largest tree was 4.55 kg MPa-1 compared to 0.027 kg MPa-1 in the smallest. Similarly in black birch, the highest C was found in the largest tree (1.99 kg MPa-1) and the lowest C in the smallest tree (0.019 kg MPa-1). R and C were found to vary inversely with sapwood area (Fig. 2.4). Whole-tree R, C, and τ were evaluated on the same date in neighboring black birch and eastern hemlock trees positioned in the canopy (Tree 1 and Tree 7, Table 2.3). Although the two trees had different sapwood areas, daily transpiration rates were similar due to higher sap flux density in black birch. Transient response analysis to a misting perturbation indicated the whole-tree τ was 24.8 min. in the hemlock compared to 5.9 min. in the black birch tree. Measurements of whole-tree R were found to be nearly equal in both trees as the hemlock R was 0.091 h MPa kg-1 and the black birch was 0.097 h MPa kg-1. As a result of a higher τ, C was over 4.5 times greater in the hemlock tree compared to the black birch. C was 4.55 kg MPa-1 in the eastern hemlock compared to 1.01 kg MPa-1 in the black birch. 2.3.3 Response to Environment On 18 July 2005, vapor pressure deficit and photosynthetically active radiation rapidly decreased over a 1 hour time period (Fig. 2.5). The response of sap flux, as measured at breast height, was slower in eastern hemlock than black birch (Fig. 2.6). There was a significant difference in the mean τ between species (Pr(T < t) = 0.0007, t = -4.3404). During this environmental change, the mean time constant of eastern hemlock

21 was 70.6 min. compared to 40.9 in black birch. The response of a set of black birch and eastern hemlock trees to rapid environmental changes were evaluated on three separate dates and no effect of tree size on the response time was found (Fig. 2.7). 2.3.4 Wavelet Analysis Wavelet analysis was run on a two and a half day sample of transpiration data from black birch and eastern hemlock (Fig. 2.8). The estimated time constant was 24.8 minutes in the eastern hemlock and 5.9 minutes in the black birch (Tree2 and Tree10, Table 2.2). The variation in localized power with time corresponds with diurnal sap flux patterns in both species. At a time period near 512, which corresponds to an actual time period of 1024 minutes, we observe a significant band in both species. This low frequency signal corresponds to the diurnal sap flux curves shown in Fig. 2.8C. Comparing eastern hemlock and black birch spectrums, significant power is observed in black birch at low time periods while absent in eastern hemlock. Above a time period of 16, which corresponds to an actual time period of 30 minutes, little localized power is observed in eastern hemlock. However, in black birch, significant power is observed at nearly all time periods fading near a time period of 4 (8 minutes actual time). More power is present at higher frequencies in black birch. 2.4 DISCUSSION Our results suggest that the loss and replacement of mature eastern hemlock trees by black birch will result in a reduction of τ of the trees in the ecosystem. When controlling for tree size, species is a significant predictor of τ with larger values in eastern hemlock. Three independent estimate techniques suggest τ will decrease with the loss and

22 replacement of eastern hemlock. Differences in physiological characteristics, such as τ, of hemlock and black birch trees will likely scale to impact ecosystem functions including soil moisture dynamics and discharge of streams. Differences in time constants of vegetation are likely to scale to have an effect on the hydrologic dynamics of an ecosystem. Evapotranspiration rates have been found to be 1.6 times greater in black birch stands compared to eastern hemlock (see Chapter 4) and we expect this to influence soil moisture and streamflow dynamics with species replacement. However, in addition to the effect of evapotranspiration, here we suggest τ, R, and C also influence other components of ecosystem water balance including soil moisture and streamflow. Trees with a larger τ have more recharge or absorption of water into a tree when transpiration stops due to a perturbation such as a rain event (Fig. 2.6). In Table 2.6, we do a prototypical example to illustrate the impact of τ on water balance. During a perturbation that lasts three times τ, 0.07 mm is recharge in a tree with a τ of 10 min. compared to 0.22 mm in a tree with a τ of 30 min. During this period, 3 times more water is removed from the soil as recharge in the tree with a larger τ. This transfer of water may alter soil moisture conditions affecting hydrologic dynamics. As a result of differences in τ, following rain events we expect the time to peak flow to decrease and the magnitude of peak flow to increase as a result of hemlock replacement by black birch in small catchments. Whole-tree estimates of R, C, and τ are appropriate for modeling soil-plantatmosphere processes at the stand level. Variations in R and C of individual tree

23 components, such as sapwood, branches and leaves, have been examined in many species including Tsuga (e.g. Tyree et al. 1991). However, whole-tree estimates of R, C, and τ are limited in the literature. Estimating C is difficult and destructive and, as a result, whole tree-estimates of C are limited. Here, we demonstrate a novel method to derive C through the use of whole-tree τ. Although estimates of C are not independent, this methodology allows us to examine R, C, and τ in several size classes of multiple species. The range of C over several size classes estimated in this study (Table 2.3) is consistent with C estimated in Pinus, Quercus, and Fagus in previous studies (Table 2.5). Meinzer et al. 2004 suggested that tropical trees maintain a constant time constant across size class and inversely regulate resistance and capacitance. In our study, R and C were inversely related in both eastern hemlock and black birch (Fig. 2.4). In both species, high R was a characteristic of small trees while low R was seen in larger trees. In black birch, we also found consistent time constants, around 10 minutes, across all sizes measured (Table 2.2). The misting treatments did not reveal consistent τ across size in eastern hemlock as τ was much greater in larger trees (Table 2.2). However, the eight trees in the hemlock stand responded similarly to sudden changes in environmental conditions despite differences in sapwood area (Fig. 2.7). There remains the possibility that the ratio of leaf area to sapwood area may influence this pattern in τ. Leaf area was not measured in the current study however, work by McDowell et al. (2002) found that the ratio of leaf area to sapwood area declines with tree height in several coniferous species including Pseudotsuga menziesii var. menziesii and Pinus ponderosa. This shift in

24 leaf area to sapwood areas with height may occur in eastern hemlock but not black birch and may explain the size related pattern in τ. Black birch and eastern hemlock have widely different hydraulic architectures that impact R, C, and τ. Black birch is a diffuse porous species. Water is carried through vessel elements that are typically 15-150 µm in diameter and 100 mm long (Hunt et al. 1991). In eastern hemlock, tracheids conduct water and are typically 20-50 µm and 1-3 mm long (Zimmermann 1983). Hydraulic resistance in eastern hemlock occurs in the bordered pits between adjacent tracheids while resistance in black birch is a function of vessel radius based on the Hagen-Poiseuille law (Jeje and Zimmermann 1979). Our estimates suggest species is not a significant predictor of R when controlling for tree size. This is surprising as, based on hydraulic architecture, we would expect greater resistance in eastern hemlock compared to black birch. The continuous wavelet transform decomposes the sap flux signal in eastern hemlock and black birch transpiration in the time-frequency domain (Fig. 2.8). At low frequencies (high period), both species show significant localized power or energy. However, at higher frequencies, differences between species are apparent. As the eastern hemlock tree has a higher τ, the R and C of the tree function as filter of high frequency environmental signals. The τ of the eastern hemlock tree shown in Figure 2.8A was estimated at 24.8 minutes. The continuous wavelet transform supports this estimate as very little power is observed at time periods below 16 (32 minutes actual time period). Also supporting our estimates of τ, the wavelet transform reveals more response in black

25 birch at higher frequencies. Significant power was present in black birch at time periods as low as 4 (8 minutes actual time). Hunt et al. (1991) discussed how scaling whole tree estimates of τ to the ecosystem level may be a practical method to resolve regional scaling issues in plant water flow models. The ecosystem time constant gives an indication of the amount of time it would take vegetation to respond to a change incorporating lags between transpiration and absorption. The results of our analysis indicate that the timing between transpiration and absorption occurs more rapidly in black birch ecosystems compared to eastern hemlock (Table 2.2). Based on the argument of Hunt et al. (1991), soil-plantatmosphere continuum models with time steps greater than estimates of τ of the dominant vegetation will incorporate errors. Here, we demonstrate that greater lags between transpiration and absorption exist in eastern hemlock and lags must be incorporated in models with time steps faster than 30 minutes. In this work, we have utilized a novel method to estimate whole-tree R, C, and τ. Combined, these physiological characteristics give an indication of the lag between transpiration and absorption from the soil. As trees with greater τ have more absorption following a perturbation, we expect an impact on soil moisture dynamics. In this study, we compared R, C, and τ in black birch and eastern hemlock and our results suggest the loss and replacement of eastern hemlock will decrease the lag between transpiration and absorption. Further research on whole-tree R, C, and τ should expand on species studied and investigate the impact of leaf area to sapwood area ratios.

26

REFERENCES Burns, R.M. and Honkala, B.H. 1990. Silvics of North American trees. 1. Conifers; 2. Hardwoods. Agriculture Handbook 654. Vol. 2. USDA Forest Service, Washington, DC, 877 p. Catovsky, S. and F.A. Bazzaz 2000. The role of resource interactions and seedling regeneration in maintaining a positive feedback in hemlock stands. Journal of Ecology 88:100-112. Czikowsky, M.J. and D.R. Fitzjarrald 2004. Evidence of seasonal changes in evapotranspiration in eastern US hydrological records. Journal of Hydrometeorology 5:974-988. Cramer, W., A. Bondeau, F.I. Woodward, I.C. Prentice, R.A. Betts, V. Brovkin, P.M. Cox, V. Fisher, J.A. Foley, A.D. Friend, C. Kucharik, M.R. Lomas, N. Ramankutty, S. Sitch, B. Smith, A. White and C. Young-Molling 2001. Global response of terrestrial ecosystem structure and function to CO2 and climate change: results from six dynamic global vegetation models. Global Change Biology. 7:357-373. Edwards, W., P. Jarvis, J. Landsberg and H. Talbot 1986. A dynamic model for studying flow of water in single trees. Tree Physiology. 1:309-324. Ellison, A.M., M.S. Bank, B.D. Clinton, E.A. Colburn, K. Elliott, C.R. Ford, D.R. Foster, B.D. Kloeppel, J.D. Knoepp, G.M. Lovett, J. Mohan, D.A. Orwig, N.L. Rodenhouse, W.V. Sobczak, K.A. Stinson, J.K. Stone, C.M. Swan, J. Thompson, B. Von Holle and J.R. Webster 2005. Loss of foundation species: consequences

27 for the structure and dynamics of forested ecosystems. Frontiers in Ecology and the Environment 3:479-486. Fahey, T.J., T.G. Siccama, C.T. Driscoll, G.E. Likens, J. Campbell, C.E. Johnson, J.J. Battles, J.D. Aber, J.J. Cole, M.C. Fisk, P.M. Groffman, S.P. Hamburg, R.T. Holmes, P.A. Schwarz and R.D. Yanai 2005. The biogeochemistry of carbon at Hubbard Brook. Biogeochemistry 75:109-176. Finzi, A.C., C.D. Canham and N. Van Breeman 1998. Canopy tree-soil interactions within temperate forests: Species effects on pH and cations. Ecological Applications. 8:905-905. Granier, A. 1985. A New Method of Sap Flow Measurement in Tree Stems. Annales Des Sciences Forestieres 42:193-200. Grinsted, A., J.C. Moore and S. Jevrejeva 2004. Application of the cross wavelet transform and the wavelet coherence to geophysical time series. Nonlinear Processes in Geophysics 11:561-566. Hooper, D. and P. Vitousek 1997. The effects of plant composition and diversity on ecosystem processes. Science 277:1302-1305. Hooper, D., F. Chapin, J. Ewel, A. Hector, P. Inchausti, S. Lavorel, J. Lawton, D. Lodge, M. Loreau, S. Naeem, B. Schmid, H. Setala, A. Symstad, J. Vandermeer and D. Wardle 2005. Effects of biodiversity on ecosystem functioning: a consensus of current knowledge. Ecological Monographs 75:3-35.

28 Hornbeck, J.W., M.B. Adams, E.S. Corbett, E.S. Verry and J.A. Lynch 1993. Long-term impacts of forest treatments on water yield - a Summary for Northeastern USA. Journal of Hydrology 150:323-344. Hunt, E. and P.S. Nobel 1987. Non-steady state water flow for three desert perennials with different capacitances. Australian Journal of Plant Physiology. 14:363-375. Hunt, E.R., S.W. Running and C.A. Federer 1991. Extrapolating plant water-flow resistances and capacitances to regional scales. Agricultural and Forest Meteorology 54:169-195. Jeje, A.A. and M.H. Zimmermann 1979. Resistance to water flow in xylem vessels. Journal of Experimental Botany. 30:817-827. Jenkins, J., J. Aber and C. Canham 1999. Hemlock woolly adelgid impacts on community structure and N cycling rates in eastern hemlock forests. Canadian Journal of Forest Research 29:630-645. Kizlinski, M.L., D.A. Orwig, R.C. Cobb and D.R. Foster 2002. Direct and indirect ecosystem consequences of an invasive pest on forests dominated by eastern hemlock. Journal of Biogeography 29:1489-1503. Kueppers, L.M. and J. Harte 2005. Subalpine forest carbon cycling: Short- and long-term influence of climate and species. Ecological Applications 15:1984-1999. Landsberg, J.J., T.W. Blanchard and B. Warrit 1976. Studies on the movement of water through apple trees. Journal of Experimental Botany. 27:579-596. Loustau, D., P. Berbigier, P. Roumagnac, C. ArrudaPacheco, J.S. David, M.I. Ferreira, J.S. Pereira and R. Tavares 1996. Transpiration of a 64-year-old maritime pine

29 stand in Portugal .1. Seasonal course of water flux through maritime pine. Oecologia. 107:33-42. Magnani, F. and M. Borghetti 1995. Interpretation of seasonal-changes of xylem embolism and plant hydraulic resistance in Fagus-sylvatica. Plant Cell and Environment. 18:689-696. Manning, P., J.E. Newington, H.R. Robson, M. Saunders, T. Eggers, M.A. Bradford, R.D. Bardgett, M. Bonkowski, R.J. Ellis, A.C. Gange, S.J. Grayston, E. Kandeler, S. Marhan, E. Reid, D. Tscherko, H.C.J. Godfray and M. Rees 2006. Decoupling the direct and indirect effects of nitrogen deposition on ecosystem function. Ecology Letters 9:1015-1024. McClure, M.S. 1991. Density-dependent feedback and population-cycles in Adelgestsugae (Homoptera, Adelgidae) on Tsuga-canadensis. Environmental Entomology 20:258-264. McDowell, N., H. Barnard, B.J. Bond, T. Hinckley, R.M. Hubbard, H. Ishii, B. Kostner, F. Magnani, J.D. Marshall, F.C. Meinzer, N. Phillips, M.G. Ryan and D. Whitehead 2002. The relationship between tree height and leaf area: sapwood area ratio. Oecologia 132:12-20. Meinzer, F.C., S.A. James and G. Goldstein 2004. Dynamics of transpiration, sap flow and use of stored water in tropical forest canopy trees. Tree Physiology 24:901909.

30 Nobel, P.S. and P.W. Jordan 1983. Transpiration stream of desert species - resistances and Ccapacitances for a C-3, a C-4, and a CAM plant. Journal of Experimental Botany 34:1379-1391. Ogata K. 1990. Modern Control Engineering, 2nd Edition. Prentice Hall, New Jersey Orwig, D.A. and D.R. Foster 1998. Forest response to the introduced hemlock woolly adelgid in southern New England, USA. Journal of the Torrey Botanical Society 125:60-73. Phillips, N., A. Nagchaudhuri, R. Oren and G. Katul 1997. Time constant for water transport in loblolly pine trees estimated from time series of evaporative demand and stem sapflow. Trees-Structure and Function. 11:412-419. Phillips, N.G., R. Oren, J. Licata and S. Linder 2004. Time series diagnosis of tree hydraulic characteristics. Tree Physiology 24:879-890. Powell, D.B.B. and M.R. Thorpe 1977. Dynamic aspects of plant-water relations. In Environmental Effects of Crop Physiology Eds. J.J. Landsberg and C.V. Cutting. Academic Press, London, pp. 259-279. Running, S.W. 1980. Relating plant capacitance to the water erlations of Pinus contorta. Journal of Experimental Botany. 27:473-479. Shaw, E.M. 1994. Hydrology in Practice. Chapman & Hall, New York. Scholz, F.G., S.J. Bucci, G. Goldstein, F.C. Meinzer, A.C. Franco and F. MirallesWilhelm 2007. Biophysical properties and functional significance of stem water storage tissues in Neotropical savanna trees. Plant Cell and Environment. 30:236248.

31 Stadler, B., T. Muller and D. Orwig 2006. The ecology of energy and nutrient fluxes in hemlock forests invaded by hemlock woolly adelgid. Ecology 87:1792-1804. Steppe, K. and R. Lemeur 2007. Effects of ring-porous and diffuse-porous stem wood anatomy on the hydraulic parameters used in a water flow and storage model. Tree Physiology. 27:43-52. Tilman, D., J. Knops, D. Wedin, P. Reich, M. Ritchie and E. Siemann 1997. The influence of functional diversity and composition on ecosystem processes. Science 277:1300-1302. Tognetti, R., A. Longobucco and A. Raschi 1998. Vulnerability of xylem to embolism in relation to plant hydraulic resistance in Quercus pubescens and Quercus ilex cooccurring in a Mediterranean coppice stand in central Italy. New Phytologist. 139:437-447. Torrence, C. and G.P. Compo 1998. A practical guide to wavelet analysis. Bulletin of the American Meteorological Society 79:61-78. Tyree, M.T. and F.W. Ewers 1991. The Hydraulic Architecture of Trees and Other Woody-Plants. New Phytologist. 119:345-360. Zimmermann, M.H. 1983. Xylem Structure and the Ascent of Sap. Springer-Verlag, New York. 143 p.

32 Table 2.2 The number, average height, and average diameter at breast height of the trees measured in each stand.

Stand Code HM LPH RD RD IRR IRR LYF LYF

Species Hemlock Black Birch Hemlock Black Birch Hemlock Black Birch Hemlock Black Birch

n 8 8 1 1 4 2 1 1

Est. Height (m) 25 20 21 17 6 7 5 7

DBH (cm) 54.1 26.4 47.8 24.5 9.8 6.1 10.4 4.5

33

Table 2.3 The time constant of trees as determined using the misting step function technique.

Number 1 2 3 4 5 6 7 8 9 10 11 12 13

Stand HM RD HM IRR LYF IRR IRR LPH LPH RD IRR IRR LYF

Species Hemlock Hemlock Hemlock Hemlock Hemlock Hemlock Hemlock Black Birch Black Birch Black Birch Black Birch Black Birch Black Birch

DBH (cm) 52.9 47.8 44.6 11.3 10.4 6.3 5.4 27.4 22.8 24.5 6.7 5.5 4.5

τ 18.6 24.8 21.8 11.8 10.0 10.1 9.4 7.4 10.5 5.9 10.2 9.8 7.8

34 Table 2.4 The time constant, resistance, and capacitance of sampled eastern hemlock and black birch trees. Time constant was determined using the misting step function technique.

Num. 1 2 3 4 5 6 7 8 9 10 11

Species HM HM HM HM HM BB BB BB BB BB BB

DBH (cm) 47.8 6.3 5.4 10.4 11.3 5.5 24.5 27.4 22.8 6.7 4.5

Sapwood Area (cm2) 652 27.3 26.4 78.5 96.7 19.6 409 555.4 379.9 31.2 14.5

τ 24.8 10.1 9.4 10 11.8 9.8 5.9 7.4 10.5 10.2 7.8

R (h MPa kg-1) 0.091 4.65 5.70 1.058 0.497 2.52 0.097 0.0955 0.203 1.7075 7.0088

C (kg MPa-1) 4.55 0.036 .027 .157 0.396 0.0647 1.01 1.98769 0.862 0.0996 0.0185

35 Table 2.5. Previous estimates of whole-tree τ, C, and R in other species. Species Hilaria rigida

τ 0.2

C(kg Mpa-1) 0.00250

R (h MPa kg-1) 0.924

Encelia farinose

2.3

0.0420

0.2745

Pinus contorta Malus pumila

1.5 2.2

0.180 0.369

0.0529 0.0324

Pinus contorta Pinus pinaster Pinus taeda Anacardium excelsum Ficus insipida Schefflera morototoni

3.5 0.5-0.7 0.7 0.55 0.56 0.5

131.4 43.4

0.0042 0.013 0.014

Cordia alliodora Fagus sylvatica

0.51

Quercus rubur Fagus sylvatica Quercus pubescens

35.7 3.2 0.2123

0.16 0.050806

0.9398

0.029778 0.0083 – 0.00167. 0.00333 – 0.00667

Source Hunt and Nobel 1987; Nobel and Jordan 1983 Hunt and Nobel 1987; Nobel and Jordan 1983 Running 1980 Landsberg et al. 1976; Powell and Thorpe 1977 Edwards et al. 1986 Loustau et al. 1996 Phillips et al. 1997 Meinzer et al. 2004 Meinzer et al. 2004 Meinzer et al. 2004 Meinzer et al. 2004 Steppe and Lemeur 2007 Steppe and Lemeur 2007 Magnani and Borghetti 1995 Tognetti et al. 1998

36 Table 2.6 Prototypical example of the impact of the time constant assuming a starting flux rate of 0.45 mm h-1, a projected canopy area of 23.4 m2, sapwood depth of 11.3 cm, and sapwood area of 739 cm2. As sap flux should stop in 3 times τ minutes, the total mm of water flux into the tree is shown in the second column. The third column shows the total mm of water moving into the tree if the perturbation was only 30 min in duration. The final column assumes a daily transpiration rate of 4.5 mm and shows the fraction of daily transpiration that is recharge in a perturbation that lasts 3 times τ. τ (min)

mm per 3*τ

mm per 30 min.

10 30 60 120

0.07 0.22 0.44 0.87

0.07 0.14 0.18 0.20

Fraction of Daily Transpiration 0.02 0.05 0.10 0.19

37 Figure 2.2 An example of a misting treatment in a black birch tree (B). Sap flux in a nontreated tree is shown for reference (A).

Sap Flux (g m-2 s-1)

40 30 20 10 0 Sap Flux (g m-2 s-1)

A

B

30 20 10

174.2

174.4

174.6

174.8

175.0

38 Figure 2.3 An example of the exponential decay after applying a misting treatment in a black birch tree. Closed circles represent sap flux in the misted tree and open circles are a reference tree.

60

Sap Flux (g m

-2 -1 s )

50

40

30

20

10

0 0

12

24

36 Time (minutes)

48

60

72

39 Figure 2.4 The lag in sap flux (normalized by daily maximum sap flux rate) between sensors at the base of the live crown (open) and at breast height (closed) in black birch (A) and eastern hemlock (B).

B

Normalized Sap Flux

A

=4 minutes

201.26

201.28

201.30

=25 minutes

201.32

201.28

201.30

201.32

40 Figure 2.5 Resistance and capacitance as functions of sapwood area in black birch (A) and eastern hemlock (B). It is important to note that capacitance in not an independent

A

Black Birch

1

0

0 -2

-1 -2

-4

-3

B ln (resistance (h MPa kg-1 ))

2

2

Eastern Hemlock

2

2 1

0

0 -1

-2

-2 -3

-4

-4 3

4

5 ln (sapwood area (cm2))

6

7

ln (capacitance (kg MPa-1 ))

ln (resistance (h MPa kg-1 ))

3

ln (capacitance (kg MPa-1 ))

measurement. Open circles represent resistance and closed circles capacitance.

41 Figure 2.6 The response of a set of eastern hemlock (C) and black birch (D) trees to a sudden change in environmental conditions including vapor pressure deficit (A) and

g m-2 ground area

g m-2 ground area

PAR

D (kPA)

photosynthetically active radiation (B). 1.6 1.4 1.2 1.0 0.8 0.6 0.4 0.2

A

1600 1400 1200 1000 800 600 400 200

B

100

C

Hemlock

D

Black Birch

80 60 40 20 100 80 60 40 20 200.2

200.4

200.6

Decimal Date

200.8

201.0

42 Figure 2.7 The average response in the set of eastern hemlock and black birch to the environmental changes in Fig. 2.5. The top line represents the response in eastern hemlock and the bottom in black birch.

Day 200

Normalized Flux

Hemlock = 70.6 min. Black Birch 40.9 min.

20 40 60 80 100

Time Since Start of Drop in PAR/VPD (minutes)

43 Figure 2.8 The effect of sapwood area on the response time of trees to sudden changes in environmental conditions. Closed circles are black birch and open triangles are hemlock. Error bars represent one standard error.

Time Constant (min)

100 80 60 40 20 0 0

500

1000

1500

Sapwood Area (cm2)

2000

44 Figure 2.9 Continuous wavelet transform in eastern hemlock (A) and black birch (B). The corresponding sap flux time series for eastern hemlock (grey line) and black birch (black line) are shown in C. The thick black line contours represents areas at the 5% significance level above a background spectrum.

A

B

45

INTERSPECIFIC VARIATION IN NIGHTTIME TRANSPIRATION AND STOMATAL CONDUCTANCE IN A MIXED NEW ENGLAND DECIDUOUS FOREST

3.1 INTRODUCTION According to the optimization theory in plant physiology, stomata will vary over time to minimize water loss for a required amount of carbon gain (Cowan 1977). In C3 and C4 plants, there is no opportunity for carbon gain at night due to the lack of photosynthetically active radiation necessary for the light reactions of photosynthesis. Based on the optimization theory, stomata will close at night to prevent water loss when there is no opportunity for carbon gain. While stomatal closure at night is widely assumed, a number of studies have documented stomatal opening at night in many species and in a range of habitats (reviewed in Musselman and Minnick 2000). Nighttime stomatal conductance has significant implications for many physiological processes in trees. In many hydrologic models, stomatal conductance at night is often assumed to approach zero with solar radiation (e.g., Jarvis 1976, Bevin 1997). Many sap flow measurement systems assume zero flow at night (e.g., Catovsky and Bazzaz 2000, Phillips et al. 2003) and the validity of this assumption can substantially impact estimates of daytime sap flux. Nighttime water loss creates a disequilibrium between leaf and soil water potentials and thus limits the use of predawn leaf water potential to determine soil water potential in soil-plant-atmosphere continuum models (Bucci et al. 2004). Plants with nighttime stomatal conductance are more

46 susceptible to ozone damage as the detoxification of oxidants is more efficient in light when the photosynthetic electron transport is active (Mattsyek et al. 1995, Musselman and Minnick 2000, Grulke et al. 2004). Finally, isotope models used to partition ecosystem respiration are highly sensitive to nocturnal stomatal conductance because canopy conductance is a critical regulator of the stable carbon isotope signature of ecosystem respiration (δ13CR) (Cernusak et al. 2004, McDowell et al. 2004). Observations of water vapor exchange at night between plants and the atmosphere have been made through whole-tree sap flux measurements (Green et al. 1989, Hogg and Hurdle 1997, Benyon 1999), stomatal conductance measurements (Matssyek et al. 1995, Snyder et al. 2003), and through energy balance/Bowen ratio methods (Iritz and Lindroth 1994) in a range of habitats and species. Leaf-level gas exchange measurements are useful in that they can easily detect the presence of nighttime conductance and reveal directly whether exchange of water between leaves and the atmosphere is occurring. However, scaling and sampling issues make it difficult to assess the magnitude and timing of water exchange for the whole-tree. The use of sap flux sensors allow for reliable estimates of daily whole-tree water use. Green et al. (1989) found that on average about one-fifth of the daily sap flux in kiwifruit occurred at night. Nighttime sap flux was found to be 5% of the total water use in a Eucalyptus grandis (Hill ex Maiden) plantation (Benyon 1999). Measurements of nighttime sap flux do not necessarily indicate nighttime transpiration. The refilling of depleted water storage in the bole is a significant component of daily sap flux (Phillips et al. 2003). Using the heat pulse technique

47 Lopushinsky (1986) and Casparie et al. (1993) attributed nighttime sap flux to recharging depleted water in the xylem. However, the positive correlation between nighttime sap flow and vapor pressure deficit (D) suggests flow can be attributed to nighttime transpiration (Green et al. 1989, Hogg and Hurdle 1997). Oren et al. 2001 found that nighttime canopy conductance in Taxodium distichum L. is highly sensitive to D. Benyon (1999) concluded that measurements of nighttime sap flux in Eucalyptus could not have been the result of refilling of water stores because sap flux rates increased during the night after several hours of zero flow rates. Further, high levels of recharge would be expected following days with high cumulative sap flux and observations of night flux occurred on nights following days with low sap flux. The objective of this study was to detect nighttime sap flux in three species in a mixed deciduous forest through the use of both whole-tree sap flux sensors as well as measurements of leaf level stomatal conductance.

3.2 METHODS 3.2.1 Study Site and Species The study was conducted in the Prospect Hill tract of Harvard Forest, Petersham, MA, United States (42°32’N, 72°10’W, elevation 340 m). The study site was a mixed deciduous stand dominated by paper birch (Betula papyrifera Marsh.), northern red oak (Quercus rubra L.), red maple (Acer rubrum L.), black birch (Betula lenta L.), eastern hemlock (Thuga canadensis L.) and eastern white pine (Pinus strobus L.). The site burned 50 years ago resulting in the even aged hardwood forest found today. The forest is about 20 m tall and has a leaf area index of 4 to 4.5 the growing season (Julian Hadley,

48 personal communication). Soils in this portion of the Prospect Hill tract are classified as Typic Dystrochrepts, well-drained, sandy loams derived from glacial till, with average mineral soil depth and rooting zone of 0.75 m (Alison Magill, University of New Hampshire, personal communication). The mean annual temperature is 8.5°C and the site receives 105 cm of rain and 150 cm of snow annually. The trees selected for monitoring were located on Little Prospect Hill. The site was 0.5 km from the Fisher Meteorological Station. The station was equipped with environmental monitoring equipment to measure air temperature (Ta), relative humidity, precipitation, and solar radiation. Air temperature and relative humidity were measured with CS500 temperature and relative humidity probes (Campbell Scientific, Logan, UT, USA). Photosynthetically active radiation was measured with a Li-Cor 190 SA quantum sensor (Li-Cor, Lincoln, NE., USA). The meteorological data were sampled at 10 second intervals, and averaged and recorded at 60-min intervals. Meteorological data for use in this study was interpolated into 30-min intervals so as to match sap flux measurements. Three representative trees each of the dominant species, paper birch, red oak, and red maple, were selected for the study (Table 3.1). The selected trees were located within 10 m of a small access road. This allowed for canopy access using a mobile canopy access vehicle. The trees chosen were typical of the size class of the stand. Table 3.1 presents biometric data on trees used in this study.

49 3.2.2 Instrumentation and Sampling Constant-heat sap flux sensors (Granier 1985) were installed at breast height in each tree and at the base of live crown in accessible trees. Both the heated and reference probes were 20 mm long and contained a copper-constantan thermocouple junction. At breast height, a set of sensors was installed on opposite sides of each tree. One set of sensors was installed at the base of live crown in the bole just below live branches in all paper birch and red oak but only in one red maple tree due to limited canopy access. The sensors were sealed from precipitation and moisture with plastic containers and silicon caulk. The sensors were also surrounded with reflective bubble wrap to prevent direct solar heating and minimize environmentally-induced temperature variations. As a measure of water exchange, sap flux density (g H2O m-2sapwood area s-1) was calculated based on an empirical calibration equation (Granier 1985). Granier found that: u = 1.19 × 10 − 6 K 1.23

Eqn. 1

where u is sap flux density (m3m-2s-1) and K is related to the temperature difference between the two probes:

K=

(∆Tm − ∆T ) ∆T

Eqn. 2

where K is sap flux index, ∆T is the temperature difference between heated and reference probe, and ∆Tm is the temperature when there is no sap flux density (u = 0).

50 Data was sampled every 10-seconds, averaged and recorded at 30-min intervals. Measurements of sap flux density were made during 2003 and 2004 from May through September.

3.2.3 Partitioning into Recharge and Nighttime Transpiration Some of nighttime sap flux measured may be attributed to the recharge of depleted water stores. To assess the depletion and recharge of stored water, the 24-h sum of flux at the base of live crown in each tree was normalized so that it was equal to the 24-h sum of the flux at breast height. The ratio of 24-h flux at base of live crown to 24-h flux at breast height was multiplied by sap flux rates at base of live crown to normalize. Assuming the quantity of stored water in the crown is small compared to stem storage (Waring et al. 1979), the normalized flux at the base of live crown was assumed to represent transpiration. The sums of instantaneous differences between flux at breast height and transpiration represented change in stored water. Transpiration that exceeded flux at breast height represented water storage withdrawal. When flux at breast height exceeded transpiration, the excess water was refilling depleted water stores. Night was defined as hours during which solar radiation was less than 5.0 Wm-2. The sums of instantaneous changes in stored water were calculated for daytime and nighttime on each date during the 2003 growing season. Nighttime transpiration and recharge were calculated as the percentage of the total 24-h sap flux. This analysis assumes that 24-h sums of flux at the base of live crown equal flux at breast height (Shulze et al. 1985, Pallardy et al. 1995, Loustau et al. 1996, Martin et al.

51 1997, Phillips et al. 1997, Goldstein et al. 1998, Maherali and Delucia 2001). This method does not account for the possibility of long-term depletion of tree water content at seasonal time scales. The implications of this assumption are discussed in Phillips et al. (2003). If net daily water content is decreasing, nighttime transpiration estimates may be greater than we are estimating.

3.2.4 Nighttime Gas Exchange Measurements Leaf-level stomatal conductance and transpiration were measured on a clear late summer night using a LI-COR 6400 gas exchange system (Li-Cor). Measurements were obtained during clear weather conditions beginning at 1500 on September 1, 2004 and ending at 0800 the following morning. A mobile canopy access vehicle was used to allow for leaf-level measurements in the upper canopy of each species. Three leaves from different branches in the upper canopy of each species were sampled every three hours. The system was set up to closely match ambient environmental light, temperature, and humidity conditions. No light was used in the measurement chamber at night (21000500).

3.2.5 Leaf Energy Budget In order for nighttime transpiration to occur, there must be a source of energy. The latent heat flux (transpiration) must be balanced by net radiation, sensible heat flux, and energy storage:

52 0 = H + R + λE

Eqn. 3

where H is sensible heat flux, R is net radiation, and λE is latent heat flux. There are two components to net radiation, solar and thermal. Inside the leaf gas exchange chamber, the solar radiation was zero during our nighttime gas exchange measurements, however, there may be thermal net radiation from differences in the temperature of the chamber wall and the leaf temperature: Rthermal = 2εσ (Tw + 273) 4 − 2εσ (Tl + 273) 4

Eqn. 4

where ε is thermal emissivity of the leaf and σ is the Stephan-Boltzman constant. Transpiration (mmol m-2 s-1) must be converted to latent heat using:

λE = E × 44.1

Eqn. 5

where E is the transpiration rate (mmol m-2 s-1) and λE is latent heat in W m-2. The sensible heat flux can be found using: H = 2c p g b (Tl − Ta )

Eqn. 6

where cp is the heat capacity of air, gb is the boundary layer conductance, Tl is the temperature of the leaf and Ta is the air temperature. The leaf energy balance now becomes: 2εσ (Tw + 273) 4 − 2εσ (Tl + 273) 4 = 44.1E + 2c p g b (Tl − Ta ) If we allow: ∆T = Tl − Ta

Eqn. 8

and note for a small change in ∆T: Tl + 273) 4 ≈ (Ta + 273) 4 + 4(Ta + 273) 3 ∆T

Eqn. 9

Eqn. 7

53 we can solve for ∆T: 2εσ [(Tw + 273) 4 − (Ta + 273) 4 ] − 44.1E ∆T = 2c p g b + 8εσ (Ta + 273) 3

Eqn. 10

The leaf energy balance equations were used to calculate the difference between leaf temperature and air temperature. The leaf temperature could then be determined and the sensible heat flux calculated based on Eqn. 6.

3.3 RESULTS Sap flux at breast height remained elevated in paper birch on the night of September 1, 2004 when leaf gas exchange was measured (Fig. 3.1). Sap flux at breast height ceased shortly after sundown in red oak and red maple. In the 24-h period beginning at 0500, 13% of the total daily sap flux at breast height occurred at night in paper birch while only 6.6% in red oak and 2.4% in red maple (Fig. 3.2). Sap flux on this night in paper birch was slightly greater than the 2003 growing season mean (Table 3.2). Leaf-level measurements show that transpiration remained elevated in paper birch through the night and was about 50% of daytime rates (Fig. 3.1). High nighttime stomatal conductance levels were observed through the night in paper birch but not in red oak or red maple. Stomatal conductance only dropped 25% in paper birch after sunset reaching a minimum of 0.17 mol H2O m-2s-1. Stomatal conductance quickly dropped below 0.05 mol H2O m-2s-1 in red oak and red maple. All species utilized water stored in the bole for transpiration in the early hours of September 1, 2004 as the flux in the upper sensors was greater than sap flux at breast

54 height (Fig. 3.3). Recharge in paper birch occurred during daytime hours and water stores were again depleted at night during nighttime transpiration. Recharge did not occur at night in paper birch while the bole was recharged during night in red oak and red maple. In paper birch, the nighttime flux at the base of the live crown, which represents nighttime transpiration, was 23% of the total 24-h flux while the nighttime flux at breast height amounted to only 13% of total 24-h flux. Two consecutive nights (August 23 and August 24) during the 2003 growing season were found to have unusually high nighttime sap flux rates at breast height in paper birch (Fig. 3.4). The raw sap flux signal in red oak and red maple shows a stable baseline each night on the dates shown in the graph. However, in paper birch the baseline is depressed on the nights of August 23 and August 24. This depressed baseline indicates nighttime sap flux while the stable baselines of red oak and red maple indicate sap flows reach zero during the night. The environmental variables that drive nighttime transpiration are also shown in this figure. The vapor pressure deficit remained elevated on August 23 and August 24. Nighttime sap flux was highly correlated with D during the month of August in paper birch (Fig. 3.5). The relationship between sap flux at night and D did not appear to be different than the relationship during the day. The relationship between nighttime sap flux and D in red oak and red maple appeared significantly different than the relationship during the day. Direct evidence of nighttime transpiration through leaf gas exchange measurements were only made on the night of September 1. However, other indicators suggest that nighttime transpiration occurred throughout the growing season in paper

55 birch. For example, on the night of August 16 sap flux had almost stopped in paper birch (Fig. 3.6). Several hours after sunset, there was a rise in D. During the night, sap flux rates increased in paper birch in response to this increase in D. There was no response in sap flux in red oak or red maple, which remained at zero. Another indicator of nighttime transpiration is in the pattern of recharge. Recharge would be expected to be greatest on nights following high daytime sap flux. The cumulative sap flux on the night of September 1 was much greater than on other nights with greater cumulative daytime sap flux. The correlation between D and nighttime transpiration during the month of August strongly suggests nighttime transpiration occurs (Fig. 3.5). Sap flux at breast height at night was common in all species during the growing season (Table 3.2). The magnitude of nighttime sap flux varied between species. It is important to note that sap flux at breast height does not necessarily mean transpiration is occurring as much of this water is refilling depleted stores. Over the course of the growing season, nighttime sap flux, averaged across trees, represented 10.7% of the total sap flux at breast height in paper birch, 8.7% in red oak, and 2.7% in red maple. Partitioning sap flux into nighttime recharge and transpiration resulted in 10.3% of the total daily flux being attributed to nighttime transpiration in paper birch. The high percentage of nighttime flux that is transpiration is itself a function of the use of stored water to supply nighttime transpiration (Fig. 3.3). While the partitioning method indicated nighttime transpiration occurs in red oak and red maple, this flux was attributed to recharge as negligible nighttime transpiration was detected when leaf-level measurements were taken on a night with high evaporative demand.

56 The leaf energy balance was examined based on measurements with the LICOR6400 gas exchange instrument and Eqn. 10. During the night of 1 Sept., the leaf temperature was estimated to be 0.471 °C cooler than air temperature in paper birch due to nighttime transpiration (Table 3.3). Sensible heat largely balances the latent heat flux from paper birch leaves. In red oak and red maple, there was little difference between leaf temperature and air temperature. Thermal radiation from the instrument balance measured latent heat fluxes in red oak and red maple.

3.4 DISCUSSION This study is one of few to quantify interspecific changes in nighttime water vapor conductance in co-occurring tree species. Substantial nighttime stomatal conductance has been observed in other species of Betula, namely Betula pendula Roth (Matssyek et al. 1995, Günthardt-Goerg et al. 1997). Average nighttime transpiration as a proportion of daily water flux in paper birch in this study (10.3%) is comparable to values reported in the literature on other species. Nighttime water use was 5% of total daily water use in Eucalyptus grandis (Benyon 1999), 12.8% in Populus trichocarpa Torr. & Gray x P. deltoids Bartr ex. Marsh (Kim 2000), 6% in Malus sylvestris L., and 19% in Actinidia deliciosa Chev. (Green et al. 1989). On extreme nights, the percent of daily water use occurring at night was 53% in Populus (Kim 2000), 15% in Malus and 30% in Actinidia (Green et al. 1989). Measurements of nighttime sap flux in previous studies used similar Granier-style sensors. However, nighttime sap flux was not partitioned between recharge and

57 transpiration. The discharge and recharge of stored water may represent a substantial component of the daily sap flux in plants (Goldstein et al. 1998, Phillips et al. 2003). Incorporating an estimate of nighttime sap flux due to recharge, as we have done in this study, provides a more reliable estimate of the proportion of daily water use attributed to nighttime transpiration. There are some limitations to our partitioning method as sap flux in terminal branches was not measured. In red oak and red maple, the low stomatal conductance on a night with elevated D (Fig. 3.1) and the low sensitivity to D at night (Fig. 3.5 and Fig. 3.6), suggests that nighttime transpiration is negligible in these species. The nighttime sap flux measured in the sensors positioned at the base of the live crown thus was likely the recharge of depleted water stores in the terminal branches and leaves of the canopy. However, in paper birch, our partitioning method provides a reliable estimate of nighttime transpiration as fluxes at the base of the live crown exceeded fluxes at breast height indicating a withdrawal of stored water (Fig. 3.3). Environmental variables, such as D and wind speed, affected the timing and magnitude of nighttime transpiration in paper birch (Fig. 3.2). Iritz and Lindroth (1994) used an energy balance/Bowen ration method to measure evaporation in a Salix viminalis L. stand and found nighttime evaporation is controlled mainly by D and ventilation. Hogg and Hurdle (1997) found that nighttime sap flux was correlated with D in Populus tremuloides Michx. Nighttime D and wind speed explained the variation in nighttime sap flux in Eucalyptus (Benyon 1999), Malus, and Actinidia (Green et al. 1989). Contrary to these results, and our results for paper birch, D does not explain the variation in nighttime

58 sap flux in red oak and red maple well (Fig. 3.5). This is consistent with greater allocation of nighttime sap flux to refilling of depleted stores in those species. Although paper birch clearly showed evidence of substantial nighttime transpiration, evidence also suggests that some of this nighttime transpiration is itself derived from the use of stored water. Paper birch had a greater proportional decrease in sap flux than leaf conductance and transpiration on the night of September 1 (Fig. 3.1). The discrepancy between sap flux and leaf level measurements was probably not due to enhanced mixing of air in the measurement chamber, because estimates of bulk aerodynamic conductance at night were 10-20 times greater than leaf conductance (data not shown), indicating the canopy was well coupled to the air. Instead, the smaller proportional decrease in water flux at the leaf from day to night is consistent with the pattern in the bole where sap flux was greater at night in sensors positioned at the base of the live crown than at the base of the tree (Fig. 3.3). This indicates that nighttime transpiration is to some degree supported by stored water in woody tissues. If nighttime use of stored water is also derived from the crown itself, this would represent a stored water reservoir that we have neglected, and may reconcile the difference between our leaf level and sap flux measurements. Additionally, the discrepancy may be due to our sampling method as only leaves in the upper canopy were measured.

3.4.1 Hydrologic Significance of Nighttime Transpiration The sensitivity of nighttime transpiration and conductance in paper birch to D and wind speed raises a question of hydrologic significance. How often do environmental

59 conditions occur that can drive nighttime sap flux? Evaluating the relationship of nighttime transpiration to D in paper birch (Fig. 3.5) shows that a very low D may be sufficient to drive transpiration. Vapor pressure deficits are typically low during night in New England forests; however, elevated D also occurs on many nights. Additionally, on any typical night, D is substantial in the hours following sundown. Sap flux during this time is often assumed to be recharge; however, our analysis indicates substantial nighttime transpiration occurs during these hours. Paper birch has continental scale significance as it is one of the most widely distributed species in North America (Burns and Hankala 1990). The species is at the southern limit of its range in Massachusetts and one of many species found in New England deciduous forests, so nighttime transpiration is unlikely to have a drastic impact on the temperature and humidity of the microclimate. However, the species is a significant component of Canadian forests and has heavy density in British Columbia where there are about 250,000 ha of birch forests (Massie et al. 1994). Given the species composition of Canadian forests, further examination of paper birch and nighttime environmental conditions in homogenous landscapes is warranted. The mosaic of a landscape can vary including mixed stands, patchy islands, or homogenous stands. The underlying mosaic influences the feedbacks between fluxes of mass and energy and, therefore, landscape heterogeneity should be considered in surface energy fluxes at higher spatial resolutions (Avissar 1992). For example, nighttime transpiration was measured in a mixed deciduous stand at Harvard Forest. Only a small percentage of the trees found in this stand were actively transpiring at night (i.e. paper

60 birch). As a parcel of warm, dry air moves over the landscape, nighttime transpiration humidifies the air as a function, in part, of the vertical temperature, and thus humidity gradient in the turbulent transport environment. However, the mass of air added to the horizontally-advecting parcel of air from the transpiring trees is assumed to be much smaller compared to the same parcel of warm, dry air moving over a homogenous paper birch stand. As the parcel of air continues to move horizontally, the vertical humidity gradient in the mixed deciduous stand will be greater, and thus result in greater rate of exchange. The vertical humidity gradient between the air and canopy has been reduced in the homogenous stand resulting in less, if any, exchange even though stomata may be open. As the distance over the homogenous surface increases, the exchange will decrease and is a function of wind speed and diffusivity (McNaughton 1976). Global climate change has led to a shift in nighttime versus daytime temperatures. Night temperatures have increased more than day temperatures in the 20th century in northern latitudes (Folland et al. 2001) and unless atmospheric humidity increases, this may lead to elevated evaporative demand at night. Thus, in species like paper birch, increased evaporation demand during night may alter the diurnal balance of forest water use. If climate change does in fact increase nighttime D, Figure 3.5A suggests that sap flux in paper birch may simply increase as it does in relation to D during the daytime. However, caution must be used in extrapolating the sensitivity of sap flux to D at night to higher values currently observed only during the day. Indeed, analysis of nighttime sensitivity of maximum vapor conductance in paper birch to D does not support this simple interpretation. Using a boundary line analysis to isolate maximum vapor

61 conductance (here expressed as the proportional quantity Js/D) versus D (Oren et al. 2001), we found that the sensitivity of maximum canopy stomatal conductance to D is heightened at night. For example, at a D of 0.15 kPa, canopy stomatal conductance is the same at night as during daytime with optimal light conditions. However, at a D of 1.0, canopy stomatal conductance at night is only 16% of daytime values. The increase in nighttime sensitivity in paper birch is consistent with findings in other species found to transpire at night including Salix viminalils (Iritz and Lindroth 1994), Populus tremuloides (Hogg and Hurdle 1997) and Taxodium distichum (Oren et al. 2001).

3.4.2 Ecophysiological Differences The results of this study highlight the presence of ecophysiological differences between species and further illustrate how species composition can impact ecosystem function. Nighttime transpiration is often considered an unnecessary cost to a plant. With no prospect for carbon assimilation through photosynthesis, the loss of water at night is often thought of as a waste. Nighttime transpiration counters the notion that plants will maximize carbon gain and minimize water loss (Cowan 1977). The differences observed between species raises an interesting ecophysiological question. Why does paper birch have greater nighttime transpiration than red oak and red maple under the same environmental conditions? Incomplete closure of stomata at night may be an ecological strategy. Trees in the northeast typically grow in environments with low D at night as well as low risks of drought (Paulson et al. 1991). Partial closure may be an advantage to a tree to maximize

62 photosynthesis. With partial nighttime stomatal closure, trees could continue photosynthesis until sundown and resume higher rates of photosynthesis earlier in the morning than other competitors. This may allow trees to maximize the photosynthesis to transpiration ratio and avoid lags between assimilation and increased stomatal conductance (Oren et al. 2001). However, this strategy has risks during periods of drought. It has been found that red maple and red oak (Abrams 1990, are relatively drought tolerant in southeastern deciduous forests. Our results are consistent with these findings as the lack of nighttime transpiration in red oak and red maple may support drought tolerance in these species. Although we measured a limited number of species, the behavior by paper birch may also be a function of successional status. The tree is early successional and may “take risks” in terms of water conservation to accommodate high growth rates. Paper birch is shade intolerant while red oak is intermediate and red maple is tolerant (Burns and Hankala 1990). Tobiessen (1982) found that dark opening of stomata occurred as a function of relative shade tolerance. Dark stomatal opening was observed in shade intolerant trees but not in shade tolerant trees. The results of our study are consistent with these findings. The greatest nighttime transpiration and stomatal conductance occurred in paper birch, the most shade intolerant species. Nighttime transpiration and conductance was the least in red maple, the most shade tolerant of the three species. Functional sapwood contains living parenchyma cells that require oxygen for respiration. It has been speculated that respiration in wood parenchyma may be limited by insufficient oxygen supply (Gansert et al. 2001). Diffusion through intercellular gas

63 spaces may be insufficient to supply oxygen to deep parenchyma cells particularly in trees with deep sapwood. Substantial oxygen can be delivered to xylem parenchyma cells in the aqueous state through sap flux (Mancuso and Marras 2003). Gansert (2003) found that 60% of the oxygen delivered to the inner depths of sapwood in Betula pubescens Ehrh. was through dissolved oxygen in the xylem rather than through a gaseous pathway. The study also revealed that oxygen levels in the xylem reached minimum levels after sundown. Nighttime transpiration may play a critical role in delivering oxygen during this period of the diurnal cycle when oxygen concentrations can be critically low. Nighttime transpiration may provide a critical pathway for oxygen delivery to allow for parenchyma respiration in the deep sapwood of paper birch trees. By delivering oxygen to parenchyma cells, nighttime transpiration indirectly enhances nutrient transport as parenchyma cells function in nutrient transport in trees. This may be especially important to species such as paper birch with its relatively smooth, unbroken bark. Nighttime transpiration may also directly provide a benefit to paper birch in allowing greater mass flow nutrient uptake by roots compared to other species. In further development of this work, we plan to study these hypotheses.

64 3.5 ACKNOWLEDGMENTS This research was supported by a National Science Foundation Grant (Water Cycle Research EAR-0233643), a National Science Foundation Graduate Research Fellowship, and by the Harvard Forest Long-Term Ecological Research Program. We wish to thank Julian Hadley (Harvard Forest) for sharing eddy covariance tower data and for his gracious support of our resource needs at the Little Prospect Hill Tower Site. We thank the Harvard Forest staff, including Paul Kuzeja, Lucas Griffith, Edythe Ellin, Michael Scott and Adrien Fabos, for their continued support and assistance. We gratefully acknowledge Cory Pettijohn for assistance in collecting sap flux and nighttime stomatal conductance measurements. We also thank Margaret Barbour and Rachel Spicer for sharing unpublished data and for enlightening discussion.

65

REFERENCES Abrams, M.D. 1990. Adaptations and response to drought in Quercus species of North America. Tree Physiology 7:227-238. Avissar, R. 1992. Conceptual aspects of a statistical-dynamic approach to represent landscape subgrid-scale heterogeneities in atmospheric models. Journal of Geophysical Research-Atmospheres. 97:2729-2742. Baldocchi, D.D. and K.S. Rao 1995. Intra-field variability of scalar flux densities across a transition between a desert and an irrigated potato field. Boundary-Layer Meteorology. 76:109-136. Benyon, R.G. 1999. Nighttime water use in an irrigated Eucalyptus grandis plantation. Tree Physiology 19:853-859. Beven, K.J., (Ed.) 1997. Distributed modelling in hydrology: applications of TOPMODEL Ed. Wiley, Chichester, p. 330. Bucci, S.J., F.G. Scholz, J.L. Goldstein, F.C. Meinzer, J.A. Hinojosa, W.A. Hoffman and C.F. Augusto. 2004. Processes preventing nocturnal equilibrium between leaf and soil water potential in tropical savanna woody species. Tree Physiology 24:11191127. Burns, R.M., and B.H. Honkala 1990. Silvics of North America: 1. Conifers; 2. Hardwoods. Agriculture Handbook 654. U.S.D.A., Forest Service, Washington, DC. vol.2, p. 877. Caspari, H.W., S.R. Green and W.R.N. Edwards 1993. Transpiration of well-watered and water-stressed Asian pear trees as determined by lysimetry, heat-pulse, and

66 estimated by a Penman-Monteith model. Agricultural Forest Meteorolology 67:13-27. Catovsky, S. and F. Bazzaz. 2000. Contributions of coniferous and broad-leaved species to temperate forest carbon uptake: a bottom-up approach. Canadian Journal of. Forest Research 30:100-111. Cernusak, L.A., G.D. Farquhar, S.C. Wong and H. Stuart-Williams. 2004. Measurement and interpretation of the oxygen isotope composition of carbon dioxide respired by leaves in the dark. Plant Physiology 136:1-14. Clearwater, M.J., F.C. Meinzer, J.L. Andrade, G. Goldstein and N.M. Holbrook 1999. Potential errors in measurement of non-uniform sap flow using heat dissipation probes. Tree Physiology 19:681-687. Cowan, I.R. 1977. Stomatal behavior and environment. In Advances in Botanical Research Eds. R.D. Preston and H.W. Woodhouse. Academic Press, San Francisco, pp. 117-228. Federer, C. 1980. Paper birch and white oak saplings differ in responses to drought. Forest Science. 2:313-324. Folland, C.K., T.R. Karl, J.R. Christy, R.A. Clarke, G.V. Gruza, J. Jouzel, M.E. Mann, O. J., S. M.J. and S.-W. Wang 2001. Observed climate variability and change. In Climate Change: The Scientific Basis Ed. J.T. Houghton, et al. Cambridge Univ. Press, Cambridge, pp. 99-181. Gansert, D. 2003. Xylem sap flow as a major pathway for oxygen supply to the sapwood of birch (Betula pubescens Ehr.). Plant Cell Environment 26:1803-1814.

67 Gansert, D., Burgdorf, M. & Lösch, R. 2001. A novel approach to the in situ measurement of oxygen concentrations in the sapwood of woody plants. Plant Cell Environment 24:1055-1064. Goldstein, G., J.L. Andrade, F.C. Meinzer, N.M. Holbrook, J. Cavalier, P. Jackson and A. Celis. 1998. Stem water storage and diurnal patterns of water use in tropical forest canopy trees. Plant Cell Environment 21:397-406. Granier, A. 1985. Une nouvelle methode pour la mesure de flux de seve brute dans le tronc des arbres. Annals Forest Science 42:193-200. Green, S.R., K.G. McNaughton and B.E. Clothier 1989. Observations of nighttime water use in kiwifruit vines and apple trees. Agricultural Forest Meteorolology 48:251261. Grulke, N.E., R. Alonso, T. Nguyen, C. Cascio and W. Dobrowolski 2004. Stomata open at night in pole-sized and mature ponderosa pine: implications for O3 exposure metrics. Tree Physiology 24:1001-1010. Günthardt-Goerg, M.S., R. Matyssek, C. Scheidegger and T. Keller. 1993. Differentiation and structural decline in the leaves and bark of birch (Betula pendula) under low ozone concentration. Trees 7:104-114. Hogg, E.H. and P.A. Hurdle 1997. Sap flow in trembling aspen: implications for stomatal responses to vapor pressure deficit. Tree Physiology 17:501-509. Iritz, Z. and A. Lindroth 1994. Night-time evaporation from a short-rotation willow stand. Journal of Hydrology 157:235-245.

68 Jarvis, P.G. 1976. The interpretation of the variations in leaf water potential and stomatal conductance found in canopies in the field. Philosophical Transactions of the Royal Society of London Series B. 273:593 510. Kim, H. 2000. Transpiration of hybrid poplar in eastern Washington. In College of Forest Resources. University of Washington, Seattle, p. 107. Kramer, P.J. 1937. The relation between rate of transpiration and rate of absorption of water in plants. American Journal of Botany 24:10-15. Lopushinsky, W. 1986. Seasonal and diurnal trends of heat pulse velocity in Douglas-fir and ponderosa pine. Canadian Journal of Forest Research 16:814-821. Loustau, D., P. Berbigier, P. Roumagnac, C. Arruda-Pacheco, J.S. David, M.I. Ferreira, J.S. Pereira and R. Tavares. 1996. Transpiration of a 64-year-old maritime pine stand in Portugal. 1. Seasonal course of water flux through maritime pine. Oecologia 107:33–42. Maherali, H. and E.H. DeLucia. 2001. Influence of climate-driven shifts in biomass allocation on water transport and storage in ponderosa pine. Oecologia 129:481– 491. Mancuso, S. and A.M. Marras 2003. Different pathways of the oxygen supply in the sapwood of young Olea europaea trees. Planta online. DOI 10.1007/s00425-0020956-5. Martin, T.A., K.J. Brown, J. Èermák, R. Ceulemans, J. Kuèera, F.C. Meinzer, J.S. Rombold, D.G. Sprugel and T.M. Hinckley. 1997. Crown conductance and tree

69 and stand transpiration in a second- growth Abies amabilis forest. Canadian Journal of Forest Research. 27:797–808. Massie, M.R.C., E.B. Peterson, N.M. Peterson and K.A. Enns. 1994. An assessment of the stategic importance of the hardwood resource in British Columbia. BC Min. For., Victoria, BC, FRDA Rep. No. 221. Matssyek, R., M.S. Günthardt-Goerg, S. Maurer and T. Keller. 1995. Nighttime exposure to ozone reduces whole-plant production in Betula pendula. Tree Physiology 15:159-165. McDowell, N.G., D.R. Bowling, A. Schauer, J. Irvine, B.J. Bond, B.E. Law and J.R. Ehleringer. 2004. Associations between carbon isotope ratios of ecosystem respiration, water availability and canopy conductance. Global Change Biology 10:1767-1784. McNaughton, K.G. 1976. Evaporation and advection .2. Evaporation downwind of a boundary separating regions having different surface resistances and available energies. Quarterly Journal of the Royal Meteorological Society. 102:193-202. Musselman, R.C. and T.J. Minnick. 2000. Nocturnal stomatal conductance and ambient air quality standards for ozone. Atmospheric Environment 34:719-733. Oren, R., J. Sperry, G. Katul, D. Pataki, B. Ewers, N. Phillips and S. Schaeffer 1999. Survey and synthesis of intra- and interspecific variation in stomatal sensitivity to vapour pressure deficit. Plant Cell Environment 22:1515-1526. Oren, R., J.S. Sperry, B.E. Ewers, D.E. Pataki, N. Phillips and J.P. Megonigal 2001. Sensitivity of mean canopy stomatal conductance to vapor pressure deficit in a

70 flooded Taxodium distichum L. forest: hydraulic and non-hydraulic effects. Oecologia 126:21-29. Pallardy, S.G., J. Èermák, F.W. Ewers, M.R. Kaufmann,W.C. Parker and J.S. Sperry. 1995. Water transport dynamics in trees and stands. In Resource Physiology of Conifers. Eds. W.K. Smith and T.M. Hinckley. Academic Press, New York, pp 301–389. Pataki, D.E. and R. Oren 2003. Species differences in stomatal control of water loss at the canopy scale in a mature bottomland deciduous forest. Advances in Water Resources 26:1267-1278. Paulson, R.W., E.B. Chase, R.S. Roberts and D.W. Moody 1991. National water summary 1988-89—Hydrologic events and floods and droughts. Water-Supply Paper 2375. USGS (U.S. Geological Survey), Washington, D.C., p. 591. Phillips, N., A. Nagchaudhuri, R. Oren and G. Katul. 1997. Time constant for water transport in loblolly pine trees estimated from time series of evaporative demand and stem sapflow. Trees 11: 412–419. Phillips, N.G., M.G. Ryan, B.J. Bond, N.G. McDowell, T.M. Hinckley and J. Cermák. 2003. Reliance on stored water increases with tree size in three species in the Pacific Northwest. Tree Physiology 23:237-245. Schulze, E.D., J. Èermák, R. Matyssek, M. Penka, R. Zimmermann, F. Vasicek, W. Gries and J. Kuèera 1985. Canopy transpiration and water fluxes in the xylem of the trunk of Larix and Picea trees—a comparison of xylem flow, porometer and cuvette measurements. Oecologia 66:475–483.

71 Snyder, K.A., J.H. Richards and L.A. Donavan. 2003. Night-time conductance in C3 and C4 species: do plants lose water at night? Journal of Experimental Botany 54:861865. Tobiessen, P. 1982. Dark opening of stomata in successional trees. Oecologia 52:356-359. Tolk, J.A., S.R. Evett and T.A. Howell 2006. Advection influences on evapotranspiration of alfalfa in a semiarid climate. Agronomy Journal. 98:1646-1654. Waring, R., D. Whitehead and P. Jarvis 1979. The contribution of stored water to transpiration in Scot pine. Plant Cell Environment 2:309-318.

72 Table 3.5 The diameter at breast height (DBH), height, sapwood area, and sapwood depth of the trees selected for study at Harvard Forest, MA.

Tree Number 1 2 3 4 5 6 7 8 9

Species Paper birch Paper birch Paper birch Red oak Red oak Red oak Red maple Red maple Red maple

DBH (cm) 21.6 6.3 15.7 15.0 18.4 21.2 18.7 8.6 4.3

Sapwood Height Area (m) (cm2) 19.3 290.1 8.5 27.3 16.4 130.1 14.3 38.1 15.0 66.5 18.4 78.0 16.5 208.6 9.7 49.7 8.4 12.6

Sapwood Depth (cm) 6.3 3.0 3.7 0.9 1.3 1.3 5.4 3.6 2.0

73

Table 3.6 Summary of nighttime transpiration and recharge during the 2003 growing season. Night Transpiration and Recharge are calculated as the mean percent of total daily flux for the growing season. ND = not detectable.

Tree Number 1 2 3 4 5 6 7 8 9

Species Paper birch Paper birch Paper birch Red oak Red oak Red oak Red maple Red maple Red maple

Night Transpiration (%) 8.8 8.6 13.5 ND ND ND ND ND ND

Recharge (%) 0.22 0.54 0.43 10.1 7.6 8.9 3.1 2.4 2.5

74 Table 3.3 Components of the leaf energy balance at night for leaves measured using the LICOR-6400 on the night of 1 Sept. Estimates are from three leaves in one tree of each species. Standard deviations are shown in parenthesis. ∆T represents the estimated gradient between leaf temperature and air temperature.

Species Red Oak Red Maple Paper Birch

Latent Heat (W m-2) 6.1 (2.7) 7.8 (1.5) 83.4 (16.5)

∆T (Tleaf -Tair) 0.01 (0.02) -0.01 (0.01) -0.47 (0.11)

Sensible Heat (W m-2) 1.2 (2.6) -1.8 (2.1) -78.3 (18.2)

75

Figure 3.10 Sap flux at breast height, stomatal conductance, and transpiration during the night of September 1 in paper birch (closed circles), red oak (closed triangles), and red maple (open circles). Night is represented by the shaded area. The sap flux values reported for each species are the mean of 3 trees. Conductance and transpiration were measured at

Sap Flux (g H2O m-2sapwood s-1)

the leaf level.

40 30 20

Conductance to H2O (mol H2O m-2s-1)

10 0

0.3 0.2 0.1

Transpiration Rate (mmol H2O m-2s-1)

0.0 5 4 3 2 1 0 245.6

245.8

246.0 Day of Year

246.2

246.4

76

Figure 3.11 Diurnal sap flux at breast height for paper birch (closed circle), red oak (closed triangle) and red maple (open circle) for the dates during which leaf gas exchange was measured (1 Sept. to 2 Sept.). The sap flux values reported for each species are the mean of 3 trees. Shaded areas represent night. Vapor pressure deficit (D) (dashed line), and wind

40

30

20

10

0

1.5

2.0

1.0

1.5 1.0

0.5

0.5

0.0 245.2 245.4 245.6 245.8 246.0 246.2 246.4 Day of Year

0.0

Wind Speed (m s-1)

Sap Flux

D (kPa)

(g H2O m-2sapwood area s-1)

speed (solid line) for these dates are also shown.

77 Figure 3.12 Time course of sap flux in a sample tree of paper birch (A), red oak (C), and red maple (E) on September 1, 2004. Gray lines represent the normalized flux at the base of live crown and black lines represent sap flux at breast height. The flux at the base of the live crown was normalized to equal the 24-h sum of sap flux at breast height. The depletion and recharge of stored water in the bole is shown for the paper birch (B), red oak (D), and red maple tree (F). The water storage flux is calculated as the difference between flux at the base of live crown and the flux at breast height. Periods greater than zero indicate periods of recharge and negative values indication periods of water withdrawal.

10

B

A

30

5

20

0

10

-5

0

-10 10

D

C

20

5

15 0

10 -5

5 0

E

-10 10

F

20

5

15

0 10

-5

5

-10

0

245.4

245.8

Day of Year

245.4

245.8

Day of Year

Water Storage Flux

Sap Flux (g H2O m-2sapwood s-1)

40

78 Figure 3.13 Daily pattern of water use as measured at breast height in paper birch, red oak, and red maple from days 234 to 238 (22 August to 26 August 2003). The sap flux values reported for each species are the mean of 3 trees. Nighttime on days 235 and 236 had high rates of sap flux that remained elevated through the night in paper birch. Also shown are patterns in the raw sap flux signal and environmental driving forces for transpiration including solar radiation and vapor pressure deficit (D). The corrected values in red oak account for the proportion of sensors not in conducting sapwood (Clearwater et al. 1999).

Solar Radiation & Vapor Pressure Deficit

Elevated Nighttime Driving Force

Low Nighttime Driving Force

Paper Birch Raw Sap Flux Signal

2.0 1.5 1.0 0.5 0.0

D (kPa)

800 600 400 200 0 0.6

(mV)

Sap Flux

Solar Radiation -2 (W m )

79

Indicates Night Flux

0.5 0.4

Zero Flow

Sap Flux Through the Night

20

0 0.50 0.45 0.40 0.35 20 15 10 5

0.6

Red Oak Sap Flux Signal

Red Oak Sap Converted Sap Flux Values

60 45 30 15 0

Red Maple Sap Flux Signal

0.5 0.4

Sap Flux -2 -1 (g m s )

Sap Flux (mV)

Sap Flux Through the Night

60 Red Maple Sap Converted Sap Flux Values 40 20 0 234 235 236 237 238

Day of Year

Sap Flux with Correction -2 -1 (g m s )

(mV)

40

Sap Flux -2 -1 (g m s )

Sap Flux

Sap Flux -2 -1 (g m s )

60

Paper Birch Converted Sap Flux Values

80

Figure 3.14 Sensitivity of sap flux measured at the base of the live crown in paper birch (A), red oak (B), and red maple (C) to vapor pressure deficit (D) during August 2003. Sensitivity at night (open circles) and during the day (closed circles) is shown. The best fit exponential rise to maximum curve (y = a(1-e-bx)) is shown for both daytime (black line) and nighttime (gray line).

50

A

Paper Birch

40 30 20 2 Daytime (R =0.86) 2 Nighttime (R =0.66)

Sap Flux (g H2O m-2sapwood s-1)

10 0

B

Red Oak

20 15 10 2 Daytime (R =0.79) 2 Nighttime (R =0.10)

5 0 40

C

Red Maple

30 20 10

2 Daytime (R =0.84) 2 Nighttime (R =0.00)

0

0.0

0.5

1.0

1.5

D (kPa)

2.0

2.5

81 Figure 3.15 Mean sap flux at breast height in paper birch (closed circle), red oak (closed triangle), and red maple (open circle) the night of August 16. Paper birch showed sensitivity to the increase in vapor pressure deficit (D) (thick line) while red oak and red maple had no response.

Nighttime of August 16

1.0 0.8

6

Midnight

0.6

4 0.4 2

0.2

0

1900

0.0

2400

Hour

0500

D (kPa)

Sap Flux (g m-2s-1)

8

82

WATER USE BY EASTERN HEMLOCK (TSUGA CANADENSIS) AND BLACK BIRCH (BETULA LENTA): IMPLICATIONS OF EFFECTS OF THE HEMLOCK WOOLLY ADELGID

4.1 INTRODUCTION Biological disturbances such as those due to the introduction of exotic pests and pathogens can result in the selective loss and replacement of a tree species at a regional scale, resulting in significant changes to ecosystem composition and processes (Castello et al. 1995, Liebhold et al. 1995). As the physiological ecology of each species is unique, changes in ecosystem composition may impact ecosystem processes such as the cycling of carbon and water (Catovsky et al. 2002, Wedin and Tilman 1996). Across the northeastern United States, the exotic pest hemlock woolly adelgid (HWA, Adelges tsugae Annand) is rapidly altering the composition of eastern hemlock stands (Tsuga canadensis (L). Carr.). There is much interest in understanding the ecosystem impacts of the loss and replacement of eastern hemlock due to this exotic pest (e.g., Jenkins et al. 1996, Eschtruth et al. 2006). Evapotranspiration (ET) and its key component transpiration, have large potential to be impacted by replacement of hemlock, and are the focus of this study. The ecological community and ecosystem associated with eastern hemlock trees is unique. This coniferous evergreen species is one of the most long-lived and shadetolerant species found in its range. Trees may take 250 to 300 years to reach reproductive

83 maturity and may live to be over 800 years old . The dense stands that develop typically only allow small amounts of incoming sunlight to reach the understory , creating a unique microclimate. In addition, the soils under hemlock trees have characteristically low pH, high carbon to nitrogen ratios, and low rates of nitrogen mineralization and nitrification . Development of other species in the understory is severely constrained, often resulting in nearly pure hemlock stands. Due to these factors, eastern hemlock creates stable local conditions, modulates ecosystem processes, and is considered a foundation species (Ellison et al. 2005). Eastern hemlock stands are currently threatened by the exotic pest hemlock woolly adelgid (HWA), an aphid-like insect native to Asia that feeds on needles of eastern hemlock. Trees can die within 4–15 years of infestation (Orwig et al. 2002, . In New England, eastern hemlock is largely being replaced by black birch (Betula lenta L.), a deciduous seral species (Orwig and Foster 1998, Orwig et al. 2002, Catovsky and Bazzaz 2000). In Connecticut forests where heavy HWA infestation has occurred, up to 75% of trees found in replacement forests are black birch trees, although other deciduous species such as red oak (Quercus rubra L.) and red maple (Acer rubrum L.) have also been found in the gaps left by hemlock mortality. (Orwig 2002, Kittredge and Ashton 1990; Smith and Ashton 1993; Ward and Stephens 1996). The replacement of eastern hemlock, a late successional species, by black birch represents an extreme change in the morphological, phenological, and physiological characteristics of the ecosystem (Bazzaz 1979).

84 As forest infestations from exotic pests and pathogens are expected to increase over the next century (Enserink 1999), the decline of eastern hemlock provides a unique opportunity to increase our understanding of the impacts exotic pests and pathogens can have on ecosystem processes. The objective of this study was to assess the impact of the loss and replacement of eastern hemlock on key components of ecosystem water use. First, as more than two thirds of precipitation in the United States is returned to the atmosphere via transpiration from plants (Dunne and Leopold 1978), this research utilized whole-tree measurement techniques to estimate transpiration at the species level. Second, as the loss and replacement of eastern hemlock represents a shift from an evergreen coniferous forest to a broad-leaved deciduous forest, seasonal and annual evapotranspiration were estimated in an eastern hemlock and deciduous stand. Based on known attributes of early and late successional species (Bazzaz 1979), we hypothesized that transpiration during the growing season is greater in early successional, black birch compared to late successional, eastern hemlock. Further, during the non-growing season, evapotranspiration will be greater in hemlock stands than in deciduous stands due to hemlock’s maintenance of foliage.

4.2 MATERIALS AND METHODS 4.2.1 Study Sites All research was conducted in two stands in the Prospect Hill tract of Harvard Forest, Petersham, Massachusetts, United States (42°32’N, 72°10’W, elevation 340 m). Harvard Forest has an average annual precipitation of 1050 mm and a mean annual

85 temperature of 8.5°C. Study sites were established in a hemlock stand and in a plot dominated by black birch in the mixed deciduous stand on Little Prospect Hill. The hemlock stand is a climax forest dominated by eastern hemlock and is about 7 ha in area (Hadley and Schedlbauer 2002). As shown in Table 1, 84% of the basal area within 100 meters of the eddy covariance tower is eastern hemlock with the remainder of the stand consisting of scattered hardwoods and white pine (Pinus strobus L.). The area referred to as the Little Prospect Hill is a mixed deciduous stand dominated by red oak (Table 1). Other species found in the stand include black birch, red maple, paper birch (Betula papyrifera Marsh.), red pine (Pinus resinosa Ait.), white pine, and eastern hemlock. Because of the key role played by black birch in replacement of hemlock, within the Little Prospect Hill stand, we selected for measurement a plot dominated by black birch. In both the hemlock stand and the Little Prospect Hill stand, eddy covariance systems monitored ecosystem gas exchange including carbon and water fluxes. Both sites are located under 1.0 km from the Fisher Meteorological Station which is equipped with CS500 temperature and relative humidity probes (Campbell Scientific, Logan, UT), a LICor 190 SA quantum sensor to measure photosynthetically active radiation (PAR) (LICOR, Lincoln, NE), and a Met One 385 heated rain gage to measure precipitation (Met One, Grant Pass, OR). 4.2.2 Transpiration Measurements We estimated whole tree and crown transpiration with sap flux measurements. Constant-heat sap flux sensors (Granier 1985) were installed in eight eastern hemlock trees within the hemlock stand, within 20 m of the eddy covariance tower (Table 2).

86 Similarly, sap flux sensors were installed in eight black birch trees located in the Little Prospect Hill stand which were 300-400 m from the deciduous stand eddy covariance tower. The trees selected for measurement were located within 20m of the available power source. All trees were canopy dominants and chosen to be representative of dominant trees in the stand. The sensors were 20 mm long and contained a copperconstantan thermocouple junction. In each tree, at least two sets of sensors were installed on opposite sides of the tree to account for circumferential variability. All sensors were protected from precipitation and moisture by shielding with plastic containers. Additionally, the sensors were surrounded by reflective insulation to prevent direct solar heating and to minimize the effects of environmental temperature variations. Measurements of sap flux are widely used to estimate whole tree and canopy transpiration . Many studies assume sap flux is constant across the sapwood conducting area when scaling to the whole tree (e.g., Catovsky et al. 2002, Phillips et al. 2003). However, this simplifying assumption does not hold for many species . Using this scaling assumption may significantly overestimate the magnitude of transpiration as sap flux velocities are often greatest in the outer sapwood . To account for radial flow, inner sap flux sensors were installed in four trees of each species. Deep sap flux sensors were built based on the designs of Ford et al. (2004), James et al. (2002), and Spicer and Holbrook (2004). To capture difference in the radial profile in sap flux, 10 mm long sensors were positioned 3-4 cm and 4-5 cm from the cambium in black birch and 3-4 cm, 4-5 cm and 5-6 cm in eastern hemlock.

87 Sap flux (g H2O m-2 sapwood area s-1) was calculated using an empirical calibration equation (Granier 1985). Granier found that: u = 1.19 × 10 − 6 K 1.23

(Eq. 1)

where u is sap flux density (m3m-2s-1) and K is related to the temperature difference between the two probes: K=

(∆Tm − ∆T ) ∆T

(Eq. 2)

where K is sap flux index, ∆T is the temperature difference between heated and reference probe, and ∆Tm is the temperature when there is no sap flux density (u = 0). The zero sap flux condition was assumed to occur during extended periods of zero vapor pressure deficit at night. Data logging in both stands started in August, 2004. Data was collected every 30 seconds averaged into 1 minute intervals using a CR10X or CR23X datalogger (Campbell Scientific). Measurements collected from 24 August to 19 September (day of year 237 to 263) were referred to as late season throughout the analysis. Data collected from 20 September to 13 October (day of year 264 to 287) were labeled leaf fall season based on observations of leaf abscission in black birch during 2004 (J. O’Keefe, unpublished data). Damaged sensors were replaced in the spring of 2005 and operated until late July of the same season. The time period of 18 June to 26 July (day of year 169 to 207) was collectively referred to as the peak growing season. The peak growing season was defined based on the availability of data. Due to damage to multiplexors and solar

88 power issues, data was not collected in the black birch stand from 13 October to 18 June. On 27 July 2005, lightning damage ended data collection at the eastern hemlock site. 4.2.3 Data Analysis An analysis was conducted to calculate total sap flow using depth profiles based on the methods described by Lu et al. (2000). The total flow at breast height (g s-1) was found using: Ftotal = FIN + Fout

(Eq. 6)

Where Ftotal is the total flow for a cross-section of a tree, Fin is flow rate from a depth of 2 cm to the heartwood/sapwood boundary, and Fout is the flow rate at 0-2 cm depth as measured from the cambium. The outer flow rate Fout was calculated using the following: Fout = flux density × sapwood area 2

2

Fout = J s0 − 2 cm [πR − π ( R − 0.02) ]

(Eq. 7) (Eq. 8)

Where Js is the sap flux rate (g m-2 s-1) and R is the radius of the tree stem (m) minus the bark thickness. The flow rate at the inner depths (Fin) was calculated using: R

FIN =

∫ flux density × circumference dx

(Eq. 9)

0.02

Where R is the sapwood depth (m). The depth profile was created by calculating the Fd ratio which was found by dividing sap flux at a given depth by sap flux in the outer 0-2 cm of the tree. The Fd ratio was plotted against the depth beneath the cambium and a bestfit curve was found for each species using Sigma Plot (SPSS Inc., Chicago, IL). In eastern hemlock, the best fit curve was an exponential relationship (a+brx; r2=0.95) while

89 2

a second order polynomial was the best fit in black birch ( y 0 + ax + bx ; r2=0.55). The inner flow (Fin) could be found in eastern hemlock using: R

FIN =

∫J

x

s

(a + br )2π ( R − x)dx

(Eq. 10)

0.02

In black birch, inner flow was found using: R

FIN =

∫J

2

s

( y 0 + ax + bx )2π ( R − x )dx

(Eq. 11)

0.02

The total quantity of water flow in grams per day was found using total flow (Ftotal) and time in seconds. Many approaches to scaling whole-tree transpiration measurements to the stand level have been taken including using sapwood area (e.g. Phillips and Oren 2001), basal area (e.g. Teskey and Sheriff 1996), canopy position (e.g. Granier 1987), and leaf area (e.g. Hatton et al. 1995). Scaling transpiration based on projected crown area is frequently used in ecosystem water use studies to maintain a ground area basis particularly when measurements in pure plots are unattainable (e.g. Hatton et al. 1995, Oren et al. 1996, Catovsky et al. 2002). This scaling approach was selected for this study as our research question considered the replacement of an eastern hemlock stand by an equal area of black birch. The total flow in grams, as measured by our sap flux sensors, was divided by the projected crown area (m2) of the tree and converted to hydrologic units of mm. The projected crown area was found using the distance to the edge of the crown in eight directions from the stem. For gaps in transpiration data from June until August of 2004, we created a multiple regression model using transpiration, daytime vapor pressure deficit (VPD), and daytime PAR data from

90 June and July of 2005. The black birch model (R2=0.82; p<0.0001) and hemlock model (R2=0.83; p<0.0001) was run using VPD and PAR data from 2004 to fill in gaps (Table 4.3). The Student’s paired t-test was used to compare daily transpiration rates in eastern hemlock and black birch. Regression analysis was done using curve fit functions in Sigma Plot. 4.2.4 Evapotranspiration measurements The eddy covariance or eddy flux technique (Baldocchi et al. 1988) was used to measure ET from both the hemlock and deciduous stand on Little Prospect Hill. Sonic anemometers to measure three-dimensional wind (CSAT-3, Campbell Scientific Inc., Logan Utah USA) and intake ports for air samples were located approximately 5 m above the average canopy surface level at each site. An eddy flux footprint model (FSAM, Schmid 1994) was modified to estimate flux source areas for sampling points relatively close to the canopy surface. Footprint lengths for 80% of the measured gas fluxes during daytime typically ranged from of 200 to 800 m for the eddy flux system in the hemlock stand (Hadley and Schedlbauer 2002). However, the presence of scattered white pines at least 30 m tall, higher than the level of the eddy covariance system, probably limited footprint lengths to near the lower end of the 200-800 m range. The FSAM model indicated similar footprint lengths for gas fluxes measured during daytime at the Little Prospect Hill site, but without the limiting influence of emergent tall trees. As nearly all ET was observed to occur during daylight hours, these footprint lengths include the relevant source areas for ET estimates.

91 Air was drawn to closed path CO2/H2O analyzers (model LI-6262 or LI-7000, Licor Inc., Lincoln, NE USA) in instrument shelters near the ground, and data was logged at 5 hz. Covariances between vertical wind and CO2 and H2O concentrations were calculated every 30 minutes using deviations from 10-minute running means. The proper time interval between measurement of wind vectors and gas concentrations, to allow for air sample transport from the intake point to the analyzer (the “lag time”) was recalculated periodically by determining the lag times resulting in maximum covariance of H2O concentration with vertical wind. The coordinate plane of wind direction at both sites was rotated for each half-hour interval so that mean vertical wind was zero, to separate turbulent vertical transport from advection by mean flow along the streamlines. Eddy flux measurements using closed-path analyzers can underestimate ET, because small, high-frequency variations in water concentration are blurred during transport of air samples from the intake point above the forest to a gas analyzer at ground level. The extent to which this occurred was checked by a spectral correction procedure (Goulden et al. 1996), in which power spectra of sonic air temperature and CO2 and H2O concentrations are compared to determine a mathematical filtering for the temperature signal that matches the smoothing of CO2 and H2O signals due to gas mixing and adsorption/desorption of water vapor in the gas inlet line, plus electrical smoothing by the CO2/H2O analyzer. The ratio of original and smoothed temperature covariance with vertical wind is taken as an estimate of the CO2 or H2O covariance lost by attenuation of higher frequency variations. Corrections for loss of high-frequency flux using this procedure typically increased short-term and cumulative ET estimates by about 15%.

92 At both hemlock and LPH sites we also measured PAR, air temperature and relative humidity above the canopy every 60 seconds, with averages calculated every 30 minutes. Soil temperature at 10 cm depth at 5 locations near the base of each flux tower was measured and averaged at the same time intervals. Valid ET data from eddy flux measurements were often unavailable at both study sites. This was due to the locations of the eddy flux towers. In the hemlock stand the measurement tower is in the NE corner of the stand, so that SW wind was necessary for ET measurements, while on Little Prospect Hill, only upslope winds, with directions ranging from WSW to N, produce valid data. For the time periods when wind direction did not allow valid ET measurements, we estimated ET from statistical models created in SPlus (Insightful Inc., Seattle, WA)(Table 4.3). These models were derived from data collected in periods of appropriate wind direction and turbulence level, along with other environmental data collected at the flux towers. Data used in each model spanned a fairly short intervals (1-2 months), to reduce the chance that the important driving variables changed over the period of model derivation. In most periods of the year PAR explained 60 to 70% of the variation in ET at each site, and VPD explained about an additional 20% (Table 4.3). The coefficients for PAR and VPD changed over the course of the year, most dramatically during leaf development and leaf abscission in the deciduous forest (during May and June and in October, respectively), but also more slowly during aging of deciduous foliage in late summer to early fall. In spring and fall, water flux for the hemlock forest was affected by

93 the nighttime minimum temperature, which has been shown to affect leaf conductance in many conifers including hemlock (Smith et al. 1984, Hadley 2000a).

4.3 RESULTS 4.3.1 Annual Evapotranspiration The magnitude of transpiration and ET varied depending on the time of year as well as environmental conditions including vapor pressure deficit, air temperature, and photosynthetically active radiation (Fig. 4.1). In June and July 2004, average ET by the deciduous forest on Little Prospect Hill was nearly twice as great as for the hemlock forest (Fig. 4.1). The difference in ET rates between the forests declined in late summer. For most of the period from leaf abscission in mid-October until deciduous leaves were fully developed again the following June, deciduous and hemlock forests used similar quantities of water, despite the presence of foliage on the hemlock trees and none on the deciduous trees (Fig. 4.1E). For the year from July 2004 through June 2005, the deciduous forest used about 90 mm more water than the hemlock forest (Fig. 4.1E). From late June 2004 until heavy rains in the second half of September (after day 260), water use in the deciduous forest was approximately equal to incoming precipitation, while hemlock forest water use was about 100 mm less (Fig. 4.2B).

4.3.2 Transpiration During the peak growing season months of June and July, the mean transpiration

94 rate was significantly greater (p<.0000) in black birch than in eastern hemlock based on group mean comparison t-test over the period shown in Figure 4.3. Mean transpiration in black birch was 2.6 mm compared to 1.4 mm in hemlock. Cumulative transpiration per projected crown area was 69 mm greater in black birch compared to eastern hemlock during June through October (Fig 4.2A). As mean daytime temperature and mean daytime photosynthetically active radiation gradually declined over the growing season (Fig. 4.1) and leaf abscission started in black birch, shifts in the relative magnitude of daily transpiration between black birch and eastern hemlock were observed (Fig. 4.4). On 21 June, peak rates (expressed per projected crown area) were 1.1 x 10-4 mm s-1 (SE=2.6 x 10-5) in black birch compared to 6.7 x 10-5 mm s-1 (SE=1.5 x 10-5) in eastern hemlock. In late August, just prior to the start of leaf abscission, peak transpiration rates were nearly equal in the two species as the peak black birch rate was 9.4 x 10-5 mm s-1 (SE=2.0 x 10-5) while eastern hemlock had a peak rate of 8.0 x 10-5 mm s-1(SE=9.7 x 10-6). Later in this period, after about 40% of leaf fall in black birch (J. O’Keefe, unpublished data), there was a reversal in the relative magnitude of transpiration in black birch and eastern hemlock. On 9 October, peak transpiration rates were greater in eastern hemlock (7.8 x 10-5 mm s-1 SE=9.4 x 10-6) compared to black birch (4.7 x 10-5 mm s-1 SE=1.3 x 10-5). The effect of scaling by projected canopy area is illustrated in Appendix A. Total daily transpiration rates paralleled the patterns in peak transpiration. During the peak growing season were 1.6 times greater in black birch than eastern hemlock (Fig. 4.5). Daily transpiration was nearly equal during the late growing season in the two species. There was no significant difference in average daily transpiration in the late

95 season between black birch and eastern hemlock. Daily transpiration rates during the fall were greater in eastern hemlock. During this overall period there was a significant difference in transpiration between black birch and eastern hemlock: average daily transpiration was significantly greater (p=0.0026) in eastern hemlock. Total transpiration per unit PAR was 19% greater in eastern hemlock than in black birch during the leaf fall season (Fig. 4.6). This was a reversal of relative differences during the peak growing season, where transpiration per unit of PAR was greater in black birch. Fitting a sigmoidal curve shows a nearly linear decline in black birch transpiration per unit PAR near day 237 which we defined as the start of the late growing season. This decline was not seen in eastern hemlock until much later in the season near day 270.

4.4 DISCUSSION Black birch is expected to be the dominant replacement species in declining hemlock stands across New England (Orwig 2002, Catovsky and Bazzaz 2000). In this study, we estimated the impact of species replacement on key components of ecosystem water use through measurements of transpiration and ET. Our hypothesis that transpiration during the growing season is greater in early successional, black birch trees compared to late successional, eastern hemlock trees was supported. During late fall, winter, and early spring, ET was similar in the two forests. This is likely because of very low evaporation in the heavily shaded understory of the hemlock forest (Hadley 2000b) compared to the sun-exposed understory of the leafless deciduous forest. As a result, we

96 predict that loss and replacement of eastern hemlock by black birch has potential to alter the annual ecosystem water balance (Fig. 4.1). Our results indicate that a transition from a hemlock to a black birch dominated stand will have the greatest impact on water balance during the peak growing season (Fig. 4.1), when transpiration will be enhanced as a result of species replacement (Fig 4.2, Fig. 4.3 and Fig. 4.4). The difference in transpiration between species is a function of the physiological controls exerted by vegetation and environmental forces. Generally, late successional species have high leaf resistances and low transpiration rates (Bazzaz 1979). Maximum leaf resistance to water vapor in eastern hemlock was found to be twice the resistance in black birch during June and July in Harvard Forest (B. Hardiman and J. Hadley, unpublished data). Leaf resistance exerts large controls over whole-tree water flow (Jarvis and McNaughton 1986) and is likely a major factor in the greater transpiration rates measured in black birch during the peak growing season (Fig. 4.1, Fig. 4.3, Fig. 4.4 and Fig. 4.5). Differences in the hydraulic conductivity of sapwood may also affect transpiration rates (Lhomme 2001). The xylem anatomy in black birch is diffuse porous while eastern hemlock has a tracheid anatomy. Deciduous diffuse porous species are capable of having a ten-fold greater hydraulic conductivity than evergreen conifers (Zimmermann 1971). Annual water flux rates have been found to be greater in some evergreen tree species compared to deciduous species due to evergreens’ ability to transpire during the dormant season (Jarvis and Leverenz 1983, Schulze et al. 1977). However, measurements from our eddy covariance system indicate that in Harvard Forest, ET from eastern

97 hemlock ecosystems during the dormant season does not compensate for its lower flux rates during the growing season (Fig. 4.1). These results are consistent with conclusions by Catovsky et al. (2002) based on periodic transpiration measurements in hemlock, red oak, and red maple at Harvard Forest. Temperature, PAR, and VPD may regulate this pattern (Fig. 4.1). While transpiration rates are greater in eastern hemlock late in the growing season (Fig. 4.1 and Fig. 4.5), transpiration in both species is observed to decrease with air temperature, vapor pressure deficit, and PAR during this time (Fig. 4.1). The decline in transpiration per unit of PAR around day 240 (Fig. 4.6) is surprising in black birch. This date is in late August and still an opportune time for photosynthesis. Transpiration per unit D (data not shown) also decreased during this time suggesting leaf functioning was affected and the decline in transpiration was not simply due to changes in the environmental conditions. While transpiration is the dominant component of ecosystem ET, differences in interception evaporation between black birch and eastern hemlock may also influence ecosystem water balance. Interception was not measured in this study and drawing conclusions based on forest type are difficult as rates are regulated by the type and intensity of rainfall and other meteorological conditions (Crockford and Richardson 2000). During the dormant season, precipitation falls to the surface in the leafless deciduous stand while interception occurs in the canopy of the hemlock stand. However, from January through March when no transpiration was measured (Fig. 4.1), little difference in ET is observed between stands suggesting evaporation rates are similar.

98 Eddy covariance systems were used in this study to quantify the differences in ET between hemlock and deciduous forests at Harvard. Our conclusions are limited as the vegetation in the deciduous stand measured by the eddy covariance system is dominated by red oak, not black birch (Table 4.1). Eddy covariance data from a black birch dominated ecosystem is currently not available. Our measurements from the deciduous stand provide a good indication of the impact of species replacement on ET during the non-growing season as the interception and evaporation dynamics in a leafless red oak and black birch stand will be similar. During the growing season, water use in a red oak and black birch stand may also be similar as leaf conductance has been measured and found to be similar in red oak and black birch during the peak growing season (M. Daley, unpublished data). Abrupt disturbances are not new to northeastern U.S.A. ecosystems. A massive decline of eastern hemlock, hypothesized to be driven by climatic changes (Foster et al. 2007) but also by hemlock looper (Bhiry and Filion 1996, Davis 1981), occurred around 5000 years BP. Recovery from the disturbance took over 1000 years (Foster and Zebryk 1993). More recently, the exotic pathogen chestnut blight (Cryphonectria parasitica (Murrill) Barr) eliminated American chestnut (Castanea dentate Marsh.) from northeastern forests during the early 1900s. Unlike the replacement of chestnut by oaks and other, functionally similar hardwoods , the loss and replacement of eastern hemlock represents a substantial change in the transpiration processes of the ecosystem (Fig. 4.1 and Fig. 4.3). Based on water balance equations, an increase in transpiration must be

99 compensated by a combined reduction in stream outflow, ground-water outflow, and catchment water storage when holding precipitation and ground-water inflow constant. The impacts of hemlock woolly adelgid on water cycling will be substantial at the local level. Although only 4% of the timberland area is eastern hemlock forest type in New England, as a whole, in several counties of western Massachusetts, hemlock comprises over 12% of the timberland area (USDA Forest Service 2007). While this may represent a limited fraction of the forest area in major watersheds, small headwater watersheds are often dominated by eastern hemlock. In the Connecticut River Valley of Massachusetts, hemlock stands mapped from aerial images range in area from 10 ha to 242 ha, averaging 51 ha (Orwig et al. 2002). Stands this size can encompass the entire area of small watersheds, particularly in hilly terrain. Management plans must consider the potential shift in the timing and magnitude of water use as a result of eastern hemlock replacement. Our results indicate that a transition from a hemlock to a black birch dominated stand will result in 30% increase in stand water transpiration from June through October (Fig 4.2B). Small streams draining from hemlock dominated watersheds that would normally have a light or moderate flow during the late growing season will experience reduced flow and may cease flowing altogether as a result of increased water use by black birch trees. From June through early September, cumulative evapotranspiration in eastern hemlock was about 100 mm less than cumulative precipitation (Fig. 4.2). However, cumulative evapotranspiration from the deciduous stand was nearly equal to cumulative precipitation. Stream flow in small catchments may be unsustainable as ET and precipitation are balanced. Evans (2004)

100 found that small streams draining from hardwood forests are much more likely to dry up in summer than streams from hemlock forests. A shift in stream flow processes will affect stream macroinvertebrates (Snyder et al. 2002) and fish populations such as brook trout (Salvelinus fontinalis Mitchell) (Ross et al. 2003). Recreational opportunities and ecosystem services may be lost if vegetation water use is not considered in management plans for stands affected by hemlock woolly adelgid.

4.5 ACKNOWLEDGMENTS This research was supported by a National Science Foundation Grant (Water Cycle Research EAR-0233643), a National Science Foundation Graduate Research Fellowship to MD, and by the Harvard Forest Long- Term Ecological Research Program. We thank members of the Harvard Forest staff, including Paul Kuzeja, Lucas Griffith, Edythe Ellin, and Michael Scott, for their support and assistance. We also thank Rachel Spicer and Chelcy Ford for sharing advice on sap flux sensor construction.

101

REFERENCES Baldocchi, D.D., B.B. Hicks, and T.P. Meyers 1988. Measuring Biosphere-Atmosphere Exchanges of Biologically Related Gases with Micrometeorological Methods. Ecology 69:1331-1340. Bazzaz, F.A. 1979. The physiological ecology of plant succession. Annual Review of Ecology And. Systematics 10: 351-271. Bhiry, N. and L. Filion 1996. Mid-Holocene hemlock decline in eastern North America linked with phytophagous insect activity. Quaternary Research 45:312-320. Burns, R.M. and B.H. Honkala 1990. Silvics of North American trees. 1. Conifers; 2. Hardwoods. Agriculture Handbook 654. Vol. 2. USDA Forest Service, Washington, DC, 877p. Catovsky, S. and F.A. Bazzaz 2000. The role of resource interactions and seedling regeneration in maintaining a positive feedback in hemlock stands. Journal of Ecology 88:100-112. Catovsky, S., N.M. Holdbrook, and F.A. Bazzaz 2002. Coupling whole-tree transpiration and canopy photosynthesis in coniferous and broad-leaved tree species. Canadian Journal of Forest Research 32: 295-309. Castello, J.D., D.J. Leopold and P.J. Smallidge 1995. Pathogens, Patterns, and Processes in Forest Ecosystems. BioScience 45:16-24. Davis, M.B. 1981. Outbreaks of forest pathogens in Quaternary history. In Proceedings of the IV International Palynological Conference, Lucknow, pp. 216-227. Dunne, T. and. L.B. Leopold 1978. Water in Environmental Planning. W. H. Freeman

102 and Company: New York, pp. 1-818. Ellison, A., M. Bank, B. Clinton, E. Colburn, K. Elliott, C. Ford, D. Foster, B. Kloeppel, , J. Knoepp, G. Lovett, J. Mohan, D. Orwig, N. Rodenhouse, W. Sobczak, K. Stinson, J. Stone, C. Swan, J. Thompson, B.V. Holle, and J. Webster 2005. Loss of foundation species: consequences for the structure and dynamics of forested ecosystems. Frontiers in Ecology and the Environment 3:479-486. Enserink, M. 1999. Biological invaders sweep in. Science 285:1834-1836. Eschtruth, A.K., N.L. Cleavitt, J.J. Battles, R.A. Evans and T.J. Fahey 2006. Vegetation dynamics in declining eastern hemlock stands: 9 years of forest response to hemlock woolly adelgid infestation. Canadian Journal of Forest Research 36:1435-1450. Evans, R.A. 2004. Hemlock woolly adelgid and the disintegration of eastern hemlock ecosystems. Park Science 22: 53-56. Ford, C., M. McGuire, R. Mitchell, and R. Teskey 2004. Assessing variation in the radial profile of sap flux density in Pinus species and its effect on daily water use. Tree Physiology 24:241-249. Foster, D.R., D.H. Knight, and J.F. Franklin 1998. Landscape patterns and legacies resulting from large, infrequent forest disturbances. Ecosystems 1:497-510. Foster, D.R., W.W. Oswald, E.K. Faison, E.D. Doughty, and B.C.S. Hansen 2006. A climatic driver for abrupt mid-holocene vegetation dynamics and the hemlock decline in New England. Ecology 87:2959–2966.

103 Foster, D.R. and. T. M. Zebryk 1993. Long-term vegetation dynamics and disturbance history of a Tsuga-dominated forest in New England. Ecology 74:982-998. Goulden, M.L., J.W. Munger, S.M. Fan, B.C. Daube, and S.C. Wofsy 1996. Measurements of carbon sequestration by long-term eddy covariance: Methods and a critical evaluation of accuracy. Global Change Biology 2:169-182. Granier, A. 1985. Une nouvelle methode pour la mesure de flux de seve brute dans le tronc des arbres. Annals of Forest Science 42:193-200. Hadley, J.L. 2000a. Effect of daily minimum temperature on photosynthesis in eastern hemlock (Tsuga canadensis L.) in autumn and winter. Arctic Antarctic and Alpine Research 32:368-374. Hadley, J.L. 2000b. Understory microclimate and photosynthetic response of saplings in an old-growth eastern hemlock (Tsuga canadensis L.) forest. Ecoscience 7:66-72. Hadley, J.L. and J.L. Schedlbauer 2002. Carbon exchange of an old-growth eastern hemlock (Tsuga canadensis) forest in central New England. Tree Physiology 22:1079-1092. Hatton, T.J., Catchpole, E.A. and Vertessy, R.A. 1990. Integration of sapflow velocity to estimate plant water use. Tree Physiology 6:201-209. James, S., M. Clearwater, F. Meinzer, and G. Goldstein 2002. Heat dissipation sensors of variable length for the measurement of sap flow in trees with deep sapwood. Tree Physiology 22:277-283.

104 Jarvis, P.G. and J.W. Leverenz 1983. Productivity of temperate, deciduous and evergreen forests. In Encyclopedia of Plant Physiology, vol. 12D, Physiological Plant Ecology. Springer-Verlag, Berlin, pp. 234-280. Jarvis, P.G. and K.G. McNaughton 1986. Stomatal Control of Transpiration - Scaling up from Leaf to Region. Advances in Ecological Research 15:1-49. Jenkins, J.C., J.D. Aber and C.D. Canham 1999. Hemlock woolly adelgid impacts on community structure and N cycling rates in eastern hemlock forests. Canadian Journal of Forest Research 29:630-645. Kittredge, D.B. and P.M. Ashton 1990. Natural regeneration patterns in even-aged mixed stands in southern New England. Northern Journal of Applied Forestry 7:163-168. Lhomme, J.P. 2001. Stomatal control of transpiration: Examination of the Jarvis-type representation of canopy resistance in relation to humidity. Water Resources Research 37:689-699. Liebhold, A.M., W.L. MacDonald, D. Bergdahl, and V. Mastro 1995. Invasion by exotic forest pests: a threat to forest ecosystems. Forest Science Monograph 30:1-49. Lu, P., W. Muller and E. Chacko 2000. Spatial variations in xylem sap flux density in the trunk of orchard-grown, mature mango trees under changing soil water conditions. Tree Physiology 20:683-692. McClure, M.S. 1991. Density-dependent feedback and population cycles in Adelges tsugae (Homoptera: Adelgidae) on Tsuga canadensis. Environmental. Entomolology 20:258-264.

105 Mladenoff, D.J. 1987. Dynamics of nitrogen mineralization and nitrification in hemlock and hardwood treefall gaps. Ecology 68:1171-1180. Orwig, D. and D. Foster 1998. Forest response to the introduced hemlock woolly adelgid in southern New England, USA. Journal of Torrey Botanical Society 125:60-73. Orwig, D.A. 2002. Stand dynamics associated with chronic hemlock woolly adelgid infestations in southern New England. In Symposium on the Hemlock Woolly Adelgid in eastern North America proceedings. (eds. R.C. Reardon, Onken, B.P., and Lashomb, J.) pp. 36-47. New Jersey Agricultural Experiment Station Publication, New Brunswick, NJ. Orwig, D., D. Foster, and D. Mausel 2002. Landscape patterns of hemlock decline in New England due to the introduced hemlock woolly adelgid. Journal of Biogeography 29:1517-1530. Phillips, N.G., M.G. Ryan, B.J. Bond, N.G. McDowell, T.M. Hinckley, and J. Cermak 2003. Reliance on stored water increases with tree size in three species in the Pacific Northwest. Tree Physiology 23:237-245. Phillips, N., R. Oren, and R. Zimmerman 1996. Radial patterns of xylem sap flow in nondiffuse, diffuse- and ring-porous tree species. Plant Cell and Environment 19:983990. Ross, R. M., R. M. Bennett, C. D. Snyder, J.A. Young, D.R. Smith, and D.P. Lemarie 2003. Influence of eastern hemlock (Tsuga canadensis) on fish community structure and function in headwater streams of the Delaware River basin. Ecology of Freshwater Fish 12:60–65.

106 Schmid, H.P. 1994. Source areas for scalars and scalar fluxes. Boundary-Layer Meteorology 67:293-318. Schulze, E.D., M. Fuchs, and M.I. Fuchs 1977. Spacial Distribution of Photosynthetic Capacity and Performance in a Mountain Spruce Forest of Northern Germany .3. Significance of Evergreen Habit. Oecologia 30:239-248. Smith, W.K., D.R. Young, G.A. Carter, J.L. Hadley and G.M. McNaughton 1984. Autumn stomatal closure in six conifer species of the Central Rocky Mountains. Oecologia 63:237-242. Smith, D.M. and P.M. Ashton, 1993. Early dominance of pioneer hardwood after clearcutting and removal of advanced regeneration. Northern Journal of Applied Forestry 10: 14-19. Snyder, C., J. Young, D. Lemarie, and D. Smith, 2002. Influence of eastern hemlock (Tsuga canadensis) forests on aquatic invertebrate assemblages in headwater streams. Canadian Journal of Fisheries and Aquatic Sciences 59(2):262–275. Spicer, R. and N. Holbrook 2005. Within-stem oxygen concentration and sap flow in four temperate tree species: does long-lived xylem parenchyma experience hypoxia? Plant, Cell and Environment 28:192-201. USDA Forest Service. 2007. Forest inventory and analysis national program. Available from http://fia.fs.fed.us [accessed 17 January 2007]. Ward, J.S. and G.R. Stephens, 1996. Influence of crown class on survival and development of Betula lenta in Connecticut, U.S.A. Canadian Journal of Forest Research 26: 277-288.

107 Wedin, D.A. and D. Tilman 1996. Influence of nitrogen loading and species composition on the carbon balance of grasslands. Science 274:1720-1723. Wullschleger, S.D., F. Meinzer, and R. Vertessy 1998. A review of whole-plant water use studies in trees. Tree Physiology 18:499-512. Wullschleger, S. and A.W. King 2000. Radial variation in sap velocity as a function of stem diameter and sapwood thickness in yellow-poplar trees. Tree Physiology 20:511-518. Zimmermann, M.H. and C.L. Brown 1971. Trees: Structure and Function. SpringerVerlag, New York. 336 p.

108 Table 4.7 The percent of basal area 0-100 m from the hemlock and deciduous eddy covariance towers. Hemlock Deciduous Species Stand Stand Tsuga canadensis 84% 6% Betula lenta 3% 4% Quercus rubra 2% 54% Acer rubrum 4% 7% Pinus strobus 6% 11% Pinus resinosa 11% Betula papyrifera 3%

109 Table 4.8 Diameter at breast height (DBH), sapwood depth, sapwood area, and projected crown area of the trees selected for study at Harvard Forest, MA. Tree Species number 1 2 3 4 5 6 7 8 1 2 3 4 5 6 7 8

Hemlock Hemlock Hemlock Hemlock Hemlock Hemlock Hemlock Hemlock Black birch Black birch Black birch Black birch Black birch Black birch Black birch Black birch

DBH (cm)

Sapwood Depth (cm)

Sapwood Area (cm2)

39.3 77.7 67.8 44.6 52.9 52.9 45.7 52.2 12.2 30.5 18.9 42.5 22.8 24.2 33.0 27.4

4.9 9.7 8.5 5.6 6.6 6.6 5.7 6.5 6.0 7.1 6.3 11.1 9.1 7.3 11.3 7.3

458 1941 1377 667 901 897 652 884 113 506 237 1050 384 365 739 409

Projected Crown Area (m2) 10.0 65.0 33.6 21.5 31.2 32.1 28.9 43.0 23.4 21.7 24.2 31.6 10.7 15.2 23.4 14.9

110 Figure 4.16 Seasonal patterns of (A) air temperature, (B) vapor pressure deficit (D), (C) photosynthetically active radiation (PAR), (D) transpiration per projected crown area and (E) evapotransiration (ET). Points represent mean daily values averaged for one week intervals. Black birch daily transpiration and ET are represented by closed circles and eastern hemlock by open. Error bars represent one standard error for transpiration and

Air Temperature(C)

one standard deviation for ET. Plotted points represent weekly means. 30 20 10 0 -10 0.8 D (kPa)

A

B

0.6 0.4

C

D

E

4 3

Deciduous Estimated Annual ET = 417 mm Hemlock Estimated Annual ET = 327 mm

2 1

/0 1

09 /04 /0 1 10 /04 /0 1 11 /04 /0 1 12 /04 /0 1 01 /04 /0 1 02 /05 /0 1 03 /05 /0 1 04 /05 /0 1 05 /05 /0 1 06 /05 /0 1 07 /05 /0 1/ 05

0

08

ET (mm day )

-1

Deciduous Forest

(mm day-1)

PAR (umol s-1 m-2)

5 4 3 2 1

Black Birch

50 40 30 20 10

Transpiration

0.2

111 Figure 4.17 (A) Cumulative transpiration per projected crown area in black birch (closed) and eastern hemlock (open) from June to October. Triangles represent data filled using multiple regression models and circles represent measurements using sap flux sensors. (B) Cumulative evapotranspiration in the deciduous (closed) and eastern hemlock (open) stands based on estimates using the eddy covariance technique. Cumulative precipitation is also

Cumulative Cumulative

Evapotranspiration (mm) Transpiration (mm)

shown as a solid line.

400 300

A

200 100 0 400 300 200

B

100 0 180

200

220

240

260

280

Day of Year (2004)

day day day day

vs vs vs vs

hm mod mm hm act mm bb mod2 mm bb act2 mm

day vs LPH cumulative ET (mm) day vs Hemlock cumulative ET mm) day vs Cumulative ppt (mm)

112 Figure 4.18 Daily patterns in transpiration per projected crown area in black birch (closed) and eastern hemlock (open) during a period of the peak growing season. Error bars represent +/- 1 standard error.

Transpiration (mm day-1)

6

4

2

0

170

180

190

Day of Year

200

113 Figure 4.19 Diurnal pattern of transpiration rates expressed per projected crown area in black birch and eastern hemlock during the peak growing season (21 June), late growing season (29 August), and fall (9 October). Black birch transpiration is represented by closed circles and eastern hemlock by open. Error bars represent one standard error.

June 21, 2005

Transpiration (mm s-1)

140x10-6

October 9, 2004

August 29, 2004

120x10-6 100x10-6 80x10-6 60x10-6 40x10-6 20x10-6 0

172.4

172.6

172.8

242.4

242.6

242.8

283.4

283.6

283.8

284.0

114 Figure 4.20 The relationship between black birch and eastern hemlock daily transpiration during three periods of the growing season. During the peak growing season, the slope of the relationship is 1.6 (r2=0.89) while during the late growing season the slope is 0.99 (r2=0.95) and 0.64 (r2=0.64) in the fall. Fitted lines represent best-fit linear regressions.

Black Birch Transpiration (mm day-1)

6

5

4

3

2

Leaf Fall Season Late Growing Season Peak Growing Season

1

0 0

1

2

3

4

Eastern Hemlock Transpiration (mm day-1)

5

6

115 Figure 4.21 Changes in transpiration per unit of photosynthetically active radiation (PAR) during the growing season. The best-fit curve (Sigmoidal) was found using the curve fitting function in SigmaPlot.

0.008

Transpiration (mm) per Unit PAR(umol s-1 m-2)

Eastern Hemlock Black Birch 0.006

0.004

0.002

0.000 180

200

220

240

Day of Year

260

280

116

APPENDIX A Tables describing the major models used during different time periods to fill missing evapotranspiration data from the eddy covariance tower. TABLE A1. EVAPOTRANSPIRATION MODELS USED FOR JUNE 2004 – JUNE 2005 FOR THE HEMLOCK TOWER. Dates used Model n R2 p June 18 – ET= -0.33 + 1.829*10-3*PAR + 301 0.81 <0.0001 July 14 1.187*VPD July 14 ET= -0.32 + 1.480*10-3*PAR + 461 0.71 <0.0001 Sept 1 2.303*VPD Sept 1 ET= -0.89 + 0.05245*Hrs + 447 0.69 <0.0001 -4 Sept 30 9.17*10 *PAR + 2.835*VPD Sept 30 – ET= -1.48 + 0.1541*Hrs + 2.429*VPD 185 0.30 <0.0001 Oct 26 Oct 26 – ET= 0.34 + 4.597*10-4*PAR + 0.0292*Ta 306 0.60 <0.0001 Dec 31 + 0.0300*Ts – 0.0283*Ta.min – 0.3591*Frost Jan 1 – ET= -0.59 + 0.1262*Hrs - 2.853*10438 0.37 <0.0001 3 -4 April 5 *Hrssq + 6.941*10 *PAR - 0.05796*Ta + 0.05146*Ta.min - 0.1454*Frost April 6 – ET= -0.35 + 0.03459*Hrs + 1.092*10- 499 0.69 <0.0001 3 April 30 *PAR – 9.70*10-3*Ta.min + 0.3674*VPD + 5.01*10-5*Ta.min*PAR May 1 – ET= -1.71 + 0.3291*Hrs – 769 0.64 <0.0001 May 31 0.01271*Hrssq + 1.381*10-3*PAR + 0.6716*VPD June 1 – ET= -1.27 + 0.2246*Hrs – 531 0.75 <0.0001 June 30 0.008165*Hrssq + 0.02160*PAR + 0.7718* VPD

117

TABLE A2. EVAPOTRANSPIRATION MODELS USED FOR JUNE 2004 – JUNE 2005 FOR THE DECIDUOS TOWER ON LITTLE PROSPECT HILL. Dates used Model n R2 p -3 June 18 ET= -0.21 + 5.580*10 *PAR 460 0.67 <0.0001 July 14 July 14 ET= -0.97 + 4.638*10-3*PAR + 1.801*VPD 531 0.69 <0.0001 Sept 1 Sept 1 ET= -4.87+1.265*Hrs - 0.05341*Hrssq + 327 0.71 <0.0001 Sept 30 4.361*10-4*PAR – 0.7066*ln(PAR) + 3.769*VPD Sept 30 – ET= 8.35 – 0.0306*DOY + 240 0.30 <0.0001 -3 Oct 26 1.863*10 *PAR + 1.587*VPD Oct 26 – ET= -1.07 + 0.08538*Hrs + 499 0.36 <0.0001 -3 Dec 31 1.056*10 *PAR + 0.08625*Ts – 0.691*VPD Jan 1 – ET= -0.13 + 0.03506*Hrs + 1117 0.41 <0.0001 -4 April 5 6.922*10 *PAR + 0.01436*Ta.min April 5 – ET= -0.12 + 0.0949*Hrs – 0.003582*Hrssq + 396 0.47 <0.0001 -4 May 31 9.305*10 *PAR – 0.2080*VPD June 1 – ET= -0.91 + 0.03109*Hrs + 454 0.75 <0.0001 -3 June 30 3.873*10 *PAR + 0.8182*VPD

118

FIGURE A3. EVAPOTRANSPIRATION MODELS USED FOR JUNE 2004 – JUNE 2005 FOR THE DECIDUOS TOWER ON LITTLE PROSPECT HILL.

mm s-1 (integrated divided by proj area)

g s-1 (integrated over sapwood depth)

g m-2 sapwood areaa s-1

60

Eastern Hemlock

Black Birch A

50 40 30 20 10

B 3

2

1

C

172.2

172.4

172.6

172.8

173.0

172.2

172.4

172.6

172.8

173.0

150x10-6

100x10-6

50x10-6

0 172.2

172.4

172.6

172.8

119

SIMULATING SHIFTS IN ECOSYSTEM WATER BALANCE DUE TO AN INVASIVE PEST

5.1 INTRODUCTION Evapotranspiration from terrestrial land surfaces represents a significant component of the ecosystem water balance. Vegetation serves as a principal interface between the soil and the atmosphere, and exerts substantial control on hydraulic fluxes on time scales ranging from minutes to millennia (Hornberger et al. 2001). For example, in the United States more than two thirds of precipitation is returned to the atmosphere via transpiration from plants (Dunne and Leopold 1978). Understanding the nature, magnitude, and timing of water movement through vegetation is critical to a wide range of issues including the management and prediction of watershed water supply and dynamics (Bond et al. 2002), soil moisture dynamics (Salvucci et al. 1997), forest productivity and eco-physiology (Ryan and Yoder 1997), land-atmosphere exchanges (Sellers et al. 1997; Baldocchi et al.1998), and continental runoff (Gedney et al. 2006). Alterations to natural processes through human actions have raised concerns regarding potential feedbacks to the hydrologic cycle. Issues such as global climate change, deforestation, soil erosion, and pollution are increasingly under scrutiny, particularly with regard to the hydrologic cycle. Across the eastern, USA the invasive pest hemlock woolly adelgid (HWA; Adelges tsugae Annand) is currently altering natural processes in eastern hemlock forests (Tsuga canadensis (L). Carr). The impact of the invasive pest is seen from southeastern to northeastern, USA with infestations in over 15 states (USDA Forest Service 2006).

120 The spread of HWA may substantially alter the natural hydrological processes of ecosystems as the health of hemlock forests declines. Initially, infested hemlock forests experience a reduction in leaf area which decreases transpiration and increases throughfall (Stadler et al. 2005). Over a period of 4-7 years, hemlock trees die and other species compete for the gaps left behind (McClure 1991). In studies of replacement forests across Connecticut, the replacement species have been dominated by deciduous species; mainly black birch (Orwig and Foster 1998, Kizlinski et al. 2002). This replacement represents a dramatic change in the phenology and physiology of the ecosystem and feedbacks may affect fluxes of energy, moisture, and momentum between the terrestrial land surface and the atmosphere (Bonan et al. 2003) The purpose of this study was to assess the potential changes in the water balance of ecosystems affected by the invasive pest hemlock woolly adelgid. The process based model Brook90 (Federer 2002) was used to assess potential changes in the cycling of water as a result of species composition change. In this analysis, we modeled the impact of replacing an eastern hemlock forest with a deciduous stand dominated by black birch. The daily, monthly, and annual water balance was simulated for stands dominated by both species.

5.2 METHODS The Brook90 model is a process based water balance model that simulates the hydrologic cycle for a point or small catchment. The model can be used to study evapotranspiration, soil water movement, and runoff at a daily time step at the point

121 scale. The model relies on the Shuttleworth and Wallace (1985) modification of PenmanMonteith equation to estimate potential evapotranspiration. The redistribution of water in the soil is based on Darcy’s Law. Output from Brook90 includes daily estimates of streamflow. A simulation was adapted and run for a coniferous stand dominated by eastern hemlock and a deciduous stand dominated by black birch. There are six parameter files used in the Brook90 model. The location file describes the latitude and seasonal changes in leaf cover. The flow parameters describe the infiltration and drainage of water. The soil parameters depend on soil texture and layer depth and regulate available water for vegetation. The structure and physiology of the vegetation are described in the canopy file. Parameters in the fixed file also describe the physiology and structure of the vegetation as well as environmental parameters. Finally, the initial file sets the initial soil moisture conditions. As the effect of species replacement on infiltration and soil texture will not be significant, the flow and soil parameter files were not varied between simulations. However, the location, canopy, and fixed parameter files are dependent on the vegetation type and unique parameters were run in each condition. 5.2.1 Data The simulation was run using climate and phenology data from Harvard Forest, Petersham, Massachusetts, United States (42°32’N, 72°10’W, elevation 340 m). Daily meteorological variables required by the Brook90 model include solar radiation, vapor pressure, maximum temperature, minimum temperature, wind speed, and precipitation. These data were obtained from the Fisher Meterological Station at Harvard Forest. Daily records from January 2001 to December 2005 were used to run the model. The station is

122 equipped with a Vaisala HMP45C probe for temperature and relative humidity (Vaisala Oyj, Vantaa, Finland); a Met One 385 heated rain gage to measure precipitation (Met One, Grant Pass, OR); A Licor LI200X pyranometer (LI-COR, Lincoln, NE) to measure solar radiation; and a R.M. Young 05103 wind monitor (R.M. Young, Traverse City, MI). Meteorological data from 2001 was used to ramp up the model and was not included in the results. Average daily meteorological conditions by month for Harvard Forest are shown in Figure 5.1. Peak values of meteorological variables including solar radiation, vapor pressure deficit, and air temperature occur in June, July and August at Harvard Forest. Annual precipitation is distributed relatively evenly over each month of the year. 5.2.2 Precipitation Interval File Hourly precipitation from the Fisher Meteorlogical Station was compiled to create a precipitation interval file. The model was run using 24 precipitation intervals per day. 5.2.3 Location A relative leaf area index, leaf area normalized by maximum, based on day of year was created for both the hemlock and deciduous forests (Fig 5.2). The deciduous leaf area was determined based on black birch phenology data collected by John O’Keefe at Harvard Forest (O’Keefe, personal communication). Leaf area in eastern hemlock was reduced to 0.8 from October to May (Hadley and Schedlbauer 2002). In both deciduous and hemlock, the relative height, height normalized by maximum height, was not varied with day of year and the slope and aspect were run at 0° as recommended by Federer (2002).

123 5.2.4 Canopy Unless otherwise stated, parameters from the Conifer3 and Decid3 files provided by Federer (2002) were used as default. Details of the parameters in these files and their use in this study are provided in Appendix A. Field data was collected at Harvard Forest to estimate maximum stomatal conductance in each species. Maximum stomatal conductance was estimated using measurements taken with the LI-COR 6400 (LI-COR) on 16 August, 2006 (Table A6). Five leaves from a representative black birch and eastern hemlock tree were sampled under maximal conditions (PAR=1500 umol m2 s, tair=25 ° C, CO2=400 units) between 1240 and 1300 hours. In addition, supplementary leaf gas exchange from other time periods were evaluated as a check on of the gas exchange data collected on 16 August, 2006. The maximum leaf area index was estimated in at 4.4 in eastern hemlock by Hadley and Schedlbauer (2002) and 4.2 in the deciduous stand (Hadley, personal communication). McClugherty et al. (1982) measured live matter of fine roots in layers to the maximum rooting depth in a mixed hardwood stand at Harvard Forest. These measurements were used to provide estimates of the relative root density in soil layers. The biomass for each layer was divided by the total live biomass to obtain relative values. Fine root density data was not available at Harvard Forest for eastern hemlock trees. However, McClaugherty et al. (1982) measured root density in red pine (Pinus resinosa Ait.). The root density of this coniferous species was used to represent eastern hemlock root density in our simulation.

124 5.2.5 Fixed Parameters The CVPD parameter regulates the reduction in leaf conductance due to increasing vapor pressure deficit. This parameter was set to 0.9 for hemlock and 2.0 for black birch. Whole-tree transpiration measurements (Daley et al. 2007) were used for selection of this parameter. Granier-style sap flux sensors were installed in eight black birch and eight eastern hemlock trees at Harvard Forest. The sensors measured whole-tree water use from August through October in 2004 and from June through July in 2005. Daily estimates of transpiration scaled to the ground area were compared to model output of transpiration to tune the CVPD parameter. All remaining fixed parameters were from the DECID3 and CONIF3 parameters files. 5.2.6 Soil and Flow Parameters In running the model, soil and water flow parameters were not altered between simulations of hemlock and black birch forests. Details of soil and flow parameters are included in Appendix A. Flow parameters were set based on parameters supplied by Federer (2002) for Hubbard Brook Watershed 6. Soils at Harvard Forest are typically classified as Gloucester series; well drained, sandy loams derived from glacial till (Lyford 1964). Soil layers were created based on the Gloucester series as described by National Resources Conservation Service (USDA 2007). The top layers are largely organic material and parameters created by Federer (2002) at Hubbard Brook were used to describe these layers. The stone fraction in each layer was from descriptions of Gloucester series. Based on a sandy loam texture, soil hydraulic parameters were obtained from Clapp and Hornberger (1978). To assess the soil hydraulic parameters provided by Clapp and Hornberger (1978) representative cores from Little Prospect Hill

125 were collected from 15 to 30 cm depth. Water release curves were generated using tempe cells (Soil Moisture Systems, Tucson, AZ) and a pressure manifold (Fig. 5.3). Field capacity was estimated from the release curve based water content at a water potential of -33 kPa (Richards and Weaver 1944). The water content at field capacity was estimated at 0.270 which closely aligned with parameters provided by Clapp and Hornberger (1978) for sandy loam. Similarly, water content of saturated cores was 0.430 which also was in tight agreement with Clapp and Hornberger (1978) estimates. Conductivity and wetness were taken from Clapp and Hornberger (1978) for sandy loam soil. The model was run using 4 soil layers to a depth of 1 meter. 5.2.7 Transpiration and Evapotranspiration Estimates Transpiration was estimated in a sample of eight eastern hemlock trees located in a hemlock dominated stand monitored by an eddy covariance tower. In addition, transpiration in eight black birch trees was also measured in a mixed deciduous stand equipped with an eddy covariance tower. In each tree, multiple Granier style sap flux (Granier 1985) sensors were installed on opposite sides of trees. Sap flux with depth was also measured in several representative trees from each species. Whole-tree water use was estimated for each tree as described in Daley et al. 2007. Evapotranspiration was measured above an eastern hemlock dominated stand using the eddy covariance technique as described in Hadley and Schedlbauer 2002. Evapotranspiration data was available from June through October in 2004.

126 5.2.8 Flow Estimates Stream water levels were measured manually in a pipe located in the Upper Bigelow Brook catchment at Harvard Forest. The basal area of this 24 ha catchment is predominately eastern hemlock, however, white pine, red pine, and mixed deciduous species are also found in the catchment. Manual measurements were obtained beginning in May of 2005. Water level measurements were converted to stream discharge (l/s) using standard equations for pipes. Stream discharge was converted to flow (mm) using the catchment area and a daily time interval. 5.2.9 Sensitivity Plots Gonzalez (2000) measured the sensitivity of all parameters used in the Brook90 model for a ponderosa pine watershed in Arizona and identified the most sensitive parameters. Sensitivity plots for the parameters identified as more sensitive were run in this study and are included in Appendix B. 5.3 RESULTS The output of our simulation was compared to measurements of water flow, transpiration, and evapotranspiration collected at Harvard Forest. Manual measurements of water level were converted to flow units for comparison with modeled flow for a hemlock dominated stand (Fig. 5.4). Transpiration was measured in a sample of eight black birch and eight eastern hemlock trees using Granier style sap flux sensors. The trees were sampled for a 37 day period during the peak of the growing season. The slope of the regression between black birch transpiration and eastern hemlock was 1.6 (p<0.05, R2=0.89)(Fig. 5.5A). This slope closely matched the regression of modeled transpiration

127 in eastern hemlock and black birch which had a slope of 1.68 (p<0.05, R2=0.99) (Fig. 5.5B). However, the relationship between model and estimated evapotranspiration was much more scattered (Fig. 5.6A) (p<0.05, R2=0.31). Much of this scatter occurs on days with precipitation (Fig. 5.6C) (p>.0.05, R2=0.07) compared to days without precipitation (Fig. 56) (p<0.05, R2=0.82). Our model predicts many changes in the annual water budget as the result of the loss and replacement of eastern hemlock by black birch. Although there is no difference in annual water yield between ecosystems dominated by black birch or eastern hemlock, annual evapotranspiration is significantly greater in the hemlock stand (Pr(T < t) = 0.0092, t = -3.2125). Over the four years run in the model, average annual evapotranspiration in the eastern hemlock ecosystem was 547 mm compared to 503 in black birch. No significant difference in transpiration was observed. The difference in average annual evapotranspiration was largely due to differences in intercepted rain evaporation; 71 mm in black compared to 107 in eastern hemlock. Annually, intercepted rain evaporation was significantly greater in eastern hemlock than black birch (Pr(T < t) = 0.0000, t = -9.5328). Annual intercepted snow evaporation was also significantly different as black birch was only 4 mm compared to 29 mm in eastern hemlock (Pr(T < t) = 0.0007, t = -5.5529). Over 12 mm more soil evaporation occurred annually in the black birch scenario (Pr(T > t) = 0.0293, t = 2.3312) at an annual average rate of 73 mm. Similarly, there was a significant difference in annual snow evaporation as black birch evaporated more (Pr(T > t) = 0.0001, t = 8.0892).

128 Although there is a decrease in evapotranspiration during February, March, April, and May with species replacement, there is no significant difference in monthly flows during these months (Table 5.1, Fig. 5.7). When controlling for precipitation, species is a significant predictor of evapotranspiration during June, July, August, and September (Table 5.1, Fig. 5.7B). However, only during August are monthly flows predicted to be significantly lower as a result of species replacement (Fig. 5.7A). When controlling for precipitation, species is a significant predictor of flow in August (p<0.05, R2=0.82). Over the course of the peak growing season months (May through September), the cumulative effect of the increase in evapotranspiration was 68 mm as a result of species replacement (Fig. 5.11). During a dry period in 2003, the simulation for black birch predicted no streamflow for a five day period (Fig. 5.9). Flow was maintained in the eastern hemlock simulation, although still low. Eastern hemlock is a coniferous species that maintains foliage throughout the year. During the dormant season, while black birch is leafless, evapotranspiration was nearly double in the eastern hemlock stand. Predicted eastern hemlock evapotranspiration was 176 mm compared to 102 mm in black birch (Fig. 5.10). Transpiration and intercepted rain evaporation are the largest components of this flux (Fig 5.8). April, October, and November have low leaf cover in black birch (Fig. 5.2) but adequate environmental conditions to drive transpiration and other components of evapotranspiration in eastern hemlock (Fig. 5.1). Lower transpiration, intercepted rain evaporation, and intercepted snow evaporation are predicted during these months in the

129 black birch forest (Table 5.1, Fig. 5.8). Soil evaporation is greater during these months in the black birch stand as the canopy is still thin, allowing solar radiation to penetrate. During the growing season months of May through September, predicted total evapotranspiration is greater in the black birch stand (Fig. 5.11). Average black birch evapotranspiration was 397 mm compared to 329 mm in eastern hemlock. The predicted increase in evapotranspiration as the result of species replacement is largely due to an increase in transpiration (Fig. 5.11). Average transpiration during these months is 311 mm in black birch compared to 206 mm in eastern hemlock. 5.4 DISCUSSION The loss and replacement of eastern hemlock forests by black birch will alter the hydrologic cycle of the ecosystem through two main pathways. First, physiological attributes such as stomatal conductance and xylem anatomy will be altered as a result of species replacement. Second, there will be a shift in the seasonality of the leaf cover in the canopy, shifting from evergreen to deciduous. In this work, we used the water budget model

Brook90 to simulate the hydrologic impacts of the loss and replacement of eastern hemlock by black birch. Our simulation indicates few changes in monthly flow, however, many components of the monthly water budget will be significantly altered including transpiration, intercepted rain evaporation, and intercepted snow evaporation (Fig. 5.8). Many paired catchment experiments have been conducted to examine the effect of vegetation on water yield and evapotranspiration (reviewed in Bosch and Hewlett 1981). Some paired catchment experiments have examined the conversion of a forest from one species to another (reviewed in Brown et al. 2005). However, the conversion of eastern

130 hemlock catchments to deciduous dominated catchments has not been analyzed. Reductions in conifer vegetation cover generally result in greater increases in water yield than reduction in deciduous cover (Bosch and Hewlett 1981). Conifers’ ability to fixcarbon and thus transpire during the dormant season is thought to compensate for lower flux rates during the growing season (Shulze et al. 1977). However, consistent with measurements by Catovsky et al. (2002), this pattern was not observed in eastern hemlock in our simulation. Transpiration during the non-growing season months of October through April was only 67 mm in eastern hemlock compared to 29 mm in black birch. This additional transpiration does not compensate for 95 mm more transpiration in deciduous black birch during the growing season (Fig. 5.11). Air temperature, PAR, and VPD drop rapidly following leaf abscission and as a result, physiological processes in eastern hemlock also decline. The results of our simulation highlight the importance of examining timing in hydrologic flows. While species is a significant predictor of flow only during August (Table 5.1), during individual years, differences in flow were as much as 42 mm in September. Monthly increases in flow during spring time months was as much as 19 mm. The shift in the timing of streamflow is of concern for two reasons. First, there is an increased risk of drying in streams draining from deciduous forests late in the growing season which can affect benthic communities and associated fish and wildlife (Snyder et al 2002). Our modeled predicted no flow for a five day period in 2003 in the black birch dominated scenario (Fig. 5.9). Second, shifts in the timing of streamflow will alter energy and nutrient dynamics. The cycling of nutrients such as nitrogen will be altered as a result

131 an increase in flow rates in the spring and a decrease in rates late in the summer (Jenkins et al. 1999, Yorks et al. 2003). The suitability of a stream for recreation and as a habitat may decrease as the result of species replacement. Tennant (1975) proposed that minimum flows at any time of the year must be greater than 10% of mean annual flow for suitability. Flows at 30% of mean annual flow are considered fair. In our study, the low flows in August shown in Figure 5.4 were 25% of mean annual flow in eastern hemlock compared to 17% in black birch. While this is not considered poor habitat in either scenario, it does represent a decrease in the suitability. Our model successfully predicted evapotranspiration on days without precipitation (Fig. 5.6B); however, there was large scatter on days with precipitation events. There are several possibilities for this discrepancy. First, in the Brook90 model the canopy is either wet or dry and interception is based on a fixed fraction of precipitation. However, measurements of throughfall in the hemlock stand indicate that the fraction of intercepted precipitation varies depending on storm size (Hadley, personal observations). We may be overestimating interception in larger events and underestimating in smaller events. Second, evapotranspiration estimates from the eddy covariance technique may be underestimating fluxes after precipitation events. Eddy covariance systems frequently fail when the sonic anemometer is wet. Missing data was filled using regression equations based on meteorological variables such as PAR and VPD created when the eddy covariance system was operational, typically only during dry conditions. Evapotranspiration during dry conditions includes the effects of stomatal resistance.

132 However, this resistance is removed for intercepted rain and three times more energy is required for transpiration compared to intercepted evaporation (Stewart 1977). Thus, our gap filling procedure may be underestimating evapotranspiration following rain events. While annual flows may not change significantly, changes in monthly components of the water budget were predicted in this simulation (Table 5.1, Fig. 5.8). The impact of seasonality is visible in several aspects of the water budget. For example, when controlling for precipitation, species is a significant predictor in every month except May and October. These are the two months during which deciduous forests are gaining or losing leaf cover (Fig. 5.1). Similar effects of seasonality are seen in intercepted rain evaporation, intercepted snow evaporation, and soil evaporation (Table 5.1). As the impacts of hemlock woolly adelgid continue to spread across the northeast, potential shifts in flow regimes must be considered in recreation and fish and wildlife planning. The impact of shifts in components in of the water budget due to seasonality (Table 5.1, Fig. 5.7) as well as the impact of cumulative transpiration (Fig. 5.9) should be considered. Shifts in the timing and magnitude of flows potentially threaten the health of aquatic species and may limit recreational opportunities.

133

REFERENCES Baldocchi, D.D., B.B. Hicks and T.P. Meyers 1988. Measuring Biosphere-Atmosphere Exchanges of Biologically Related Gases with Micrometeorological Methods. Ecology 69:1331-1340 Bonan, G.B., S. Levis, S. Sitch, M. Vertenstein and K.W. Oleson 2003. A dynamic global vegetation model for use with climate models: concepts and description of simulated vegetation dynamics. Global Change Biology 9:1543-1566. Bond, B.J., J.A. Jones, G. Moore, N. Phillips, D. Post and J.J. McDonnell 2002. The zone of vegetation influence on baseflow revealed by diet patterns of streamflow and vegetation water use in a headwater basin. Hydrological Processes 16:1671-1677. Bosch, J. and J. Hewlett 1982. A review of catchment experiments to determine the effect of vegetation changes on water yield and evapotranspiration. Journal of Hydrology 55:3-23. Brown, A.E., L. Zhang, T.A. McMahon, A.W. Western and R.A. Vertessy 2005. A review of paired catchment studies for determining changes in water yield resulting from alterations in vegetation. Journal of Hydrology 310:28-61. Catovsky, S., N.M. Holdbrook and F.A. Bazzaz 2002. Coupling whole-tree transpiration and canopy photosynthesis in coniferous and broad-leaved tree species. Canadian Journal of Forest Research 32:295-309. Clapp, R.B. and G.M. Hornberger 1978. Empirical Equations for Some Soil HydraulicProperties. Water Resources Research 14:601-604.

134 Daley, M.J., N.G. Phillips, J.C. Pettijohn, and J.L. Hadley 2007. Water use by eastern hemlock (Tsuga canadensis) and black birch (Betula lenta): Implications of effects of the hemlock woolly adelgid. Canadian Journal of Forest Research: in publication. Dunn, T. and L.B. Leopold. 1978. Water in Environmental Planning. Freeman and Co., New York. 818 p. Federer, C.A. 2002. BROOK 90: A simulation model for evaporation, soil water, and streamflow. Federer, C.A., C. Vorosmarty and B. Fekete 2003. Sensitivity of annual evaporation to soil and root properties in two models of contrasting complexity. Journal of Hydrometeorology 4:1276-1290. Gedney, N., P.M. Cox, R.A. Betts, O. Boucher, C. Huntingford and P.A. Stott 2006. Detection of a direct carbon dioxide effect in continental river runoff records. Nature 439:835-838. Granier, A. 1985. A New Method of Sap Flow Measurement in Tree Stems. Annals of Forest Science 42:193-200. Gonzalez, A.B. 2000. Hydrologic effects of vegetative practices on ponderosa pine watersheds in Arizona. In School of Renewable Natural Resources. The University of Arizona, p. 121. Hadley, J.L. and J.L. Schedlbauer 2002. Carbon exchange of an old-growth eastern hemlock (Tsuga canadensis) forest in central New England. Tree Physiology 22:1079-1092.

135 Hornberger, G. 2001. A Plan for a New Science Initiative on the Global Water Cycle. Draft report from the Water Cycle Study Group, U.S. Global Change Research Program. Jackson, R.B., J. Canadell, J.R. Ehleringer, H.A. Mooney, O.E. Sala and E.D. Schulze 1996. A global analysis of root distributions for terrestrial biomes. Oecologia 108:389-411. Jenkins, J.C., J.D. Aber and C.D. Canham 1999. Hemlock woolly adelgid impacts on community structure and N cycling rates in eastern hemlock forests. Canadian Journal of Forest Research 29:630-645. Kizlinski, M.L., D.A. Orwig, R.C. Cobb and D.R. Foster 2002. Direct and indirect ecosystem consequences of an invasive pest on forests dominated by eastern hemlock. Journal of Biogeography 29:1489-1503. Lovett, G.M., C.D. Canham, M.A. Arthur, K.C. Weathers and R.D. Fitzhugh 2006. Forest ecosystem responses to exotic pests and pathogens in eastern North America. Bioscience 56:395-405. McClure, M.S. 1991. Density-Dependent Feedback and Population-Cycles in AdelgesTsugae (Homoptera, Adelgidae) on Tsuga-Canadensis. Environmental Entomology 20:258-264. McClaugherty, C.A., J. Pastor, J.D. Aber and J.M. Melillo 1985. Forest Litter Decomposition in Relation to Soil-Nitrogen Dynamics and Litter Quality. Ecology 66:266-275.

136 Orwig, D.A. and D.R. Foster 1998. Forest response to the introduced hemlock woolly adelgid in southern New England, USA. Journal of the Torrey Botanical Society 125:60-73. Richards L A and Weaver L R 1944 Moisture retention by some irrigated soils as related to soil moisture tension. Journal of Agricultural Research 69: 215-235. Ryan, M.G. and B.J. Yoder 1997. Hydraulic Limits to Tree Height and Tree Growth. BioScience 47:235-242. Salvucci, G.D. 1997. Soil and moisture independent estimation of stage-two evaporation from potential evaporation and albedo or surface temperature. Water Resources Research 33:111-122. Sellers, P.J., R.E. Dickinson, D.A. Randall, A.K. Betts, F.G. Hall, J.A. Berry, G.J. Collatz, A.S. Denning, H.A. Mooney, C.A. Nobre, N. Sato, C.B. Field and A. Henderson Sellers 1997. Modeling the exchanges of energy, water, and carbon between continents and the atmosphere. Science 275:502-509. Schulze, E.D., M. Fuchs and M.I. Fuchs 1977. Spacial Distribution of Photosynthetic Capacity and Performance in a Mountain Spruce Forest of Northern Germany .3. Significance of Evergreen Habit. Oecologia 30:239-248. Shuttleworth, W.J. and J.S. Wallace 1985. Evaporation from Sparse Crops - an Energy Combination Theory. Quarterly Journal of the Royal Meteorological Society 111:839-855. Snyder, C.D., J.A. Young, D.P. Lemarie and D.R. Smith 2002. Influence of eastern hemlock (Tsuga canadensis) forests on aquatic invertebrate assemblages in

137 headwater streams. Canadian Journal of Fisheries and Aquatic Sciences 59:262275. Stadler, B., T. Muller and D. Orwig 2006. The ecology of energy and nutrient fluxes in hemlock forests invaded by hemlock woolly adelgid. Ecology 87:1792-1804 Stewart, J.B. 1977. Evaporation from Wet Canopy of a Pine Forest. Water Resources Research 13:915-921. Templer, P.H., G.M. Lovett, K.C. Weathers, S.E. Findlay and T.E. Dawson 2005. Influence of tree species on forest nitrogen retention in the Catskill Mountains, New York, USA. Ecosystems 8:1-16. USDA Forest Service. 2006. List of States and Counties with Known HWA Infestations. Available from http://www.na.fs.fed.us/fhp/hwa/hwatable_web/hwatable4.pdf [accessed 31 January 2007]. USDA Forest Service. 2007. Soil Classification: Official Series Descriptions. Available from http://soils.usda.gov/technical/classification/ [accessed 17 January 2007]. Yorks, T.E., D.J. Leopold and D.J. Raynal 2003. Effects of Tsuga canadensis mortality on soil water chemistry and understory vegetation: possible consequences of an invasive insect herbivore. Canadian Journal of Forest Research 33:1525-1537.

138 Table 5.1. The effect of species on components of the ecosystem water budget including streamflow (FLOW), evapotranspiration (EVAP), intercepted rain evaporation (IRVP), intercepted snow evaporation (ISVP), snow evaporation (SNVP), soil evaporation (SLVP) and transpiration (TRAN). Species effects are were regressed by month while controlling for precipitation. Months where species is a significant predictor at the p<0.05 level are indicated with *. MONTH FLOW Jan Feb Mar Apr May Jun Jul Aug * Sep Oct Nov Aug

EVAP

* * * *

IRVP

* * * *

ISVP

* * *

SNVP

* * * * * *

* * * * * *

* * *

* *

SLVP

*

*

TRAN

* * * * * * * * * *

139 Figure 0.22 Mean daily solar radiation (A), maximum (closed) and minimum (open) air temperature (B) and vapor pressure deficit (C) by month for five years at Harvard Forest. Also shown, average monthly precipitation (D). 25

A

MJ m

-2

20 15 10

B 20 10 0

Vapor Pressure Defit (kPa)

-10

2.0

Precipitation (mm)

Air Temperature (deg C)

5

150

C

1.5 1.0 0.5

D

100 50 0

2

4

6 Month

8

10

12

140 Figure 0.23 Estimated relative leaf area index by day of year for the eastern hemlock (open) and deciduous stand (closed) used in the simulation.

Relative Leaf Area Index

1.2 1.0 0.8 0.6 0.4 0.2 0.0 0

100

200

Day of Year

300

141 Figure 0.3 Soil water release curve for four soil cores collected on Little Prospect Hill at Harvard Forest. Field capacity was determined by examining the water content at -33 kPa. The dashed line illustrates a water content at field capacity of about 0.270 at -33 kPa which is consistent with parameters from Clapp and Hornberger (1978)

0.6 Core 1 Core 2 Core 3 Core 4

Vol. Water Content (%)

0.5

0.4

0.3

0.2

0.1 0

10

20

30

40

50

60

-h (kPa)

70

80

90

100

110

120

142 Figure 0.4 Measured (closed) and observed (open) measurements of streamflow a small catchment at Harvard Forest. 50

Flow (mm/day)

40

Brook90 Model Manual Measurements

30 20 10 0 5/1/05

6/1/05

7/1/05

8/1/05

9/1/05

10/1/05

11/1/05

12/1/05

Day of Year

1/1/06

2/1/06

3/1/06

4/1/06

5/1/06

143 Figure 0.5 The relationship of field estimates (A) and model estimates (B) of black birch and eastern hemlock transpiration from 20 June to 26 July, 2005. Transpiration was measured in the field in eight trees of each species using Granier style sap flux sensors. Sap flux estimates were scaled to projected canopy area.

B

5

slope =1.60 2 R =0.89 p<0.0001

250 200

Black Birch Modeled Transpiration (mm)

Measured Black Birch Transpiration (mol m-2 day-1)

6

A

300

150 100

slope =1.68 R2=0.99 p<0.0001

4 3 2 1

50 0

0 0

50

100

150

Measured Eastern Hemlock

Transpiration (mol m-2 day-1)

200

0

1

2

Eastern Hemlock Modeled Transpiration (mm)

3

4

144 Figure 0.6 The relationship between modeled and measured evapotranspiration in eastern hemlock (A). The relationship between the model and measured ET are shown for clear days only (B) and for days with precipitation (C).

Model Evapotranspiration (mm)

5

A 4 3 2 1

Model Evapotranspiration (mm) on Days Without Precipitation

0

B 4 3 2 1

Model Evapotranspiration (mm) on Days with Precipitation

0

C 4 3 2 1 0 0

1

2

3

Eddy Covariance Evapotranspiration (mm)

4

145 Figure 0.7 A) Average monthly flow and standard error in a simulated eastern hemlock (open circles) and black birch (closed circles) stand. Monthly averages are from five years of model simulation. B). Average monthly evapotranspiration in a simulated eastern hemlock (open) and black birch (closed) stand.

140

A

Flow (mm)

120 100 80 60 40 20 0

B

Evapotranspiration (mm)

100 80 60 40 20

1

2

3

4

5

6

7

Month

8

9

10

11

12

146 Figure 0.8 Components of monthly evapotranspiration including A) transpiration, B) intercepted rain evaporation, C) soil evaporation, D) intercepted snow evaporation, and E) snow evaporation.

Transpiration (mm)

100 80 60 40 20 0 Intercepted Rain Evaporation (mm)

A

B

20 15 10 5

Soil Evaporation (mm)

0 20

C

15 10 5

Intercepted Snow Evaporation (mm)

0 8

D

6 4 2

Snow Evaporation (mm)

0 4

E

3 2 1 0 1

2

3

4

5

6

7

Month

8

9

10

11

12

147 Figure 0.9 Simulated draw down of baseflow (A) in eastern hemlock (open) and black birch (closed) from 29 June to 3 August, 2003. B) The cumulative difference in evapotranspiration between eastern hemlock (open) and black birch (closed) for this period.

1.0

Flow (mm)

0.8 0.6 0.4 0.2

Cumulative Evapotranspiration (mm)

0.0 100 80 60 40 20 0 180

185

190

195

200

Day of Year (2003)

205

210

215

148 Figure 0.10 Components of evapotranspiration in the simulated black birch and eastern hemlock stand during the non-growing season defined as 1 October to 30 April. During this period, evapotranspiration was 102 mm in black birch compared to 176 mm in eastern hemlock.

Non-Growing Season (1 Oct to 30 April) Evapotranspiration (mm)

200 180 160 140

Intercepted Rain Evaporation Intercepted Snow Evaporation Soil Evaporation Snow Evaporation Transpiration

120 100 80 60 40 20 0 Black Birch

Eastern Hemlock

Species

149 Figure 0.11 Components of evapotranspiration in the simulated black birch and eastern hemlock stand during the peak growing season defined as 1 May to 31 September. During this period, the black birch stand transpired 311 mm compared to 206 in eastern hemlock.

Peak Season (1 May to 31 Sept) Evapotranspiration (mm)

500 Intercepted Rain Evaporation Soil Evaporation Transpiration

400

300 Transpiration 301 mm

Transpiration 210 mm

200

100

0 Black Birch

Eastern Hemlock

Species

150

APPENDIX A Tables describing the major parameters used in the eastern hemlock and black birch simulations. Table A1. Location parameters used in the hemlock and black birch simulations. Parameter

Description

Units

Value

RSTEMP

Base temperature for the separation of precipitation into rain and snow A degree-day factor for a day with daylength of 0.5 and no canopy Location of field site Slope of location used for net radiation and snowmelt

°C

-0.5

MJm-2d-1K-1

1.5

°N °

42.53 0

MELFAC LATITUDE ESLOPE

Table A2. Flow parameters used in the simulation of eastern hemlock and black birch. Parameter

Description

Units

Value

FPAR

Fraction of water content between field capacity and saturation at which the quickflow reaches its maximum allowed value of 1

-

0.3

mm mm

1000 1000

-

1

-

0.01 1

m ° -

0 100 0 0 0

QDEPTH IDEPTH INFEXP IMPERV DRAIN BYPAR LENGTH DSLOPE GSC GSP

Dimensionless factor that determines distribution of infiltrated water Fraction of the soil surface that is impermeable Multiplier of vertical flow from the lowest soil layer to groundwater Allow or prevent bypass flow Hillside length from ridge to channel Hillside slope for downslope matric flow Fraction of groundwater storage Fraction of groundwater discharge that goes to deep seepage and not added to streamflow

151 Table A3. Soil parameters used in the simulation of eastern hemlock and black birch. THICK represents soil layer thickness (mm), STONEF is the stone fraction, PSIF (kPa) is the water potential at field capacity, THETAF (m3 m-3) is the water content at field capacity, THSAT (m3 m-3) is the water content at saturation, BEXP is the exponent in the Brooks and Corey (YEAR) equation, KF (mm d-1) is the conductivity at field capacity, and WETINF is the wetness at the inflection point of the Clapp and Hornberger (1978) equation. THICK 50 50 230 670

STONEF 0 0.1 0.3 0.4

PSIF -12 -12 -4.9 -4.9

THETAF 0.32 0.23 0.27 0.27

THSAT 0.9 0.61 0.43 0.43

BEXP 6 3.5 4.9 4.9

KF 2 2 5.5 5.5

WETINF 0.92 0.92 0.92 0.92

Table A4. The relative rooting density of eastern hemlock and black birch used in the model simulation. Depth (mm) 50 100 150 150 150

Relative Root Density Hemlock Black Birch 0.22 0.41 0.29 0.28 0.2 0.15 0.17 0.13 0.11 0.03

152 Table A5. The canopy parameters for eastern hemlock and black birch used in the simulation. PARAMETER ALB ALBSN KSNVP Z0G MAXHT MAXLAI MXRTLN MXKPL FXYLEM CS PSICR GLMAX LWIDTH CR TL T1 T2 TH

DESCRIPTION Surface reflectivity Albedo or surface reflectivity with snow on the ground Correction factor for snow evaporation Roughness parameter of ground surface below canopy Maximum canopy height Maximum projected leaf area index Total length of fine roots per ground area when canopy is at maximum density and seasonal height Maximum internal conductivity for water flow through the plant Fraction of internal resistance to water flow that is in the xylem Ratio of projected stem area to canopy height Critical leaf water potential at which stomates close Maximum leaf conductance when stomates are fully open Average leaf width Extinction coefficient for solar radiation and net radiation in the canopy Mean daily temperature below which stomates stay closed Lower boundary of optimal temperature for conductivity Upper boundary of optimal temperature for conductivity Mean daily temperature above which stomates stay fully closes

UNITS -

HEMLOCK 0.14

BLACK BIRCH 0.18

-

0.14

0.23

-

0.3

0.3

m

0.02

0.02

m

25

20

-

4.4

4.2

m m-2

3100

3000

mm d1 MPa-1

4

8

-

0.5

0.5

-

0.035

0.035

MPa

-2

-2

cm s-1

0.34

0.68

m

0.03

0.1

-

0.5

0.6

°C

0

0

°C

10

10

°C

30

30

°C

40

40

153 Table A6. Peak conductance data collected on 16 August, 2005 at 1400h in leaves of one black birch, one eastern hemlock, and one red maple. CO2 ref was 400, flow was 300, PARin was 1500, stomata ratio was 1, area was 6 for birch, oak, and 3.78 for hemlock, BLcond was 2.84 for birch and oak, 3.78 for hemlock.

Obs bb1 bb2 bb3 bb4 bb5 bb6 bb7 bb8 bb9 ro1 ro2 ro3 ro4 ro5 ro6 ro7 ro8 ro9 hm1 hm2 hm3 hm4 hm5 hm6 hm7 hm8 hm9

Photo 15.2 15.1 15.3 11.6 11.5 11.6 14.3 14.3 14.5 16.9 16.9 16.8 18.7 18.8 18.8 18.3 18.3 18.2 13.3 13.3 13.3 7.62 7.9 7.63 11.7 11.7 11.8

Cond 0.26 0.261 0.265 0.21 0.211 0.211 0.321 0.324 0.329 0.307 0.308 0.307 0.334 0.339 0.337 0.335 0.336 0.337 0.202 0.203 0.207 0.0762 0.0772 0.077 0.122 0.123 0.124

Cond cm/s 0.665 0.667 0.678 0.537 0.540 0.540 0.821 0.828 0.841 0.785 0.788 0.785 0.853 0.866 0.861 0.856 0.858 0.861 0.515 0.518 0.528 0.195 0.197 0.197 0.311 0.314 0.316

Ci 261 262 262 275 276 275 286 286 286 262 263 263 257 258 257 260 261 261 266 266 267 219 215 221 219 221 220

Trmmol 2.92 2.94 2.96 2.63 2.64 2.63 3.28 3.29 3.31 3.25 3.27 3.27 3.26 3.27 3.27 3.19 3.2 3.2 2.48 2.51 2.54 1.18 1.19 1.19 1.72 1.74 1.74

VpdL 1.16 1.16 1.16 1.27 1.27 1.27 1.08 1.07 1.06 1.11 1.11 1.11 1.03 1.02 1.03 1 1.01 1.01 1.22 1.23 1.23 1.5 1.5 1.49 1.38 1.38 1.38

Tair 25.48 25.48 25.48 25.44 25.44 25.43 25.42 25.42 25.42 25.37 25.37 25.37 25.37 25.38 25.37 25.34 25.34 25.34 25.46 25.45 25.46 25.46 25.46 25.46 25.42 25.43 25.42

Tleaf 26.45 26.45 26.45 26.69 26.69 26.7 26.36 26.33 26.32 26.51 26.55 26.56 26.14 26.11 26.13 26 26.02 26.03 25.55 25.62 25.61 26.09 26.09 26.05 25.78 25.83 25.79

TBlk 25.08 25.08 25.08 25.08 25.07 25.08 25.08 25.07 25.08 25.08 25.08 25.07 25.08 25.07 25.08 25.08 25.08 25.07 25.07 25.08 25.07 25.08 25.08 25.08 25.08 25.07 25.07

CO2S 368.12 368.13 367.98 375.46 375.49 375.45 369.77 369.6 369.42 364.21 364.34 364.66 360.73 360.67 360.46 361.52 361.56 361.47 382.04 381.97 381.69 389.8 389.65 389.96 384.36 384.3 384.08

154

APPENDIX B The sensitivity of flow to changes in major canopy and soil parameters are displayed in the following figures. The figures show changes in flow by month to changes in parameters. The months chosen are February, April, July, and October to represent winter, spring, summer, and fall, respectively.

Fraction Change in Flow Output

Figure B1. Sensitivity of flow in February, April, July, and October from the eastern hemlock simulation to changes in maximum leaf area index (A), maximum leaf conductance (B), and maximum height (C). Points are monthly averages calculated over five years of model simulations. Closed circles represent February, open triangles are April, open circles are July, and closed triangles are October. 0.30 0.15 0.00 -0.15 -0.30 -0.45 -0.60

A

0.30 0.20 0.10 0.00 -0.10 -0.20 -0.30

B

0.15 0.10 0.05 0.00 -0.05 -0.10 -0.15

C

-0.60

-0.40

-0.20

0.00

0.20

Fraction Change in Parameter

0.40

0.60

155 Figure B2. Sensitivity of flow in February, April, July, and October from the black birch simulation to changes in maximum leaf area index (A), maximum leaf conductance (B), and maximum height (C). Points are monthly averages calculated over five years of model simulations. Closed circles represent February, open triangles are April, open circles are July, and closed triangles are October. 0.50 0.25

A

0.00 -0.25 -0.50 Fraction Change in Flow Output

-0.75 -1.00 0.30 0.15

B

0.00 -0.15 -0.30 -0.45

0.05

C

0.00 -0.05

-0.60

-0.40

-0.20

0.00

0.20

Fraction Change in Parameter

0.40

0.60

156 Figure B3. Sensitivity of flow in February, April, July, and October from the eastern hemlock simulation (A) and black birch simulation (B) to changes in the MELFAC which is a degree-day factor for a day with daylength of 0.5 and no canopy. Points are monthly averages calculated over five years of model simulations. Closed circles represent February, open triangles are April, open circles are July, and closed triangles are October.

Fraction Change in Flow Output

0.20 0.10

A

0.00 -0.10 -0.20

0.10

B

0.00 -0.10 -0.20 -0.60

-0.40

-0.20

0.00

0.20

Fraction Change in MELFAC

0.40

0.60

157 Figure B4. Sensitivity of flow in February, April, July, and October from the eastern hemlock simulation to changes in soil and flow parameters BEXP (A), QFPAR (B), THSAT (C), and THETAF (D). Points are monthly averages calculated over five years of model simulations. Closed circles represent February, open triangles are April, open circles are July, and closed triangles are October. 0.04

A

0.02 0.00 -0.02

Fraction Change in Flow Output

-0.04 0.04

B

0.02 0.00 -0.02 -0.04 0.04

C

0.02 0.00 -0.02 -0.04 0.04

D

0.02 0.00 -0.02 -0.04

-0.60

-0.40

-0.20

0.00

0.20

Fraction Change in Parameter

0.40

0.60

158 Figure B5. Sensitivity of flow in February, April, July, and October from the black birch simulation to changes in soil and flow parameters BEXP (A), QFPAR (B), THSAT (C), and THETAF (D). Points are monthly averages calculated over five years of model simulations. Closed circles represent February, open triangles are April, open circles are July, and closed triangles are October. 0.04

A

0.02 0.00 -0.02 -0.04

Fraction Change in Flow Output

0.04

B

0.02 0.00 -0.02 -0.04 0.04

C

0.02 0.00 -0.02 -0.04 0.04

D

0.02 0.00 -0.02 -0.04

-0.60

-0.40

-0.20

0.00

0.20

Fraction Change in Parameter

0.40

0.60

159 Figure B6. Sensitivity of flow by month to changes in soil depth. The model was run at 1m soil depth and sensitivity is shown for a simulation with 2m (open circles) and 3m (closed circles). Soil depth was changed by increasing the thickness of the bottom layer.

Fraction Change in Flow

0.3 0.2 0.1 0.0 -0.1 -0.2 -0.3 0

2

4

6

8

Month

10

12

14

160

CONCLUDING REMARKS This research has demonstrated the importance of the physiology of individual species on ecosystem processes. Physiological differences can have large affects when scaled to the ecosystem level. As disturbances such as fires, hurricanes, biotic invasions, and land use change can rapidly alter the composition of vegetation cover in forested ecosystems, understanding the physiological ecology of individual species is imperative. In this research, we have used the loss and replacement of eastern hemlock from hemlock woolly adelgid as a model of physiological impacts on ecosystem processes. In Chapter 2, I showed there will be a decrease in the filtering of environmental perturbations as a result of the loss and replacement of eastern hemlock. The decrease in filtering will be due to a decrease in the time constant of the vegetation. There will be a decrease in the lag between environmental perturbations and transpiration processes. Larger lags are present in eastern hemlock due to great quantities of capacitance. A shift in the time constant of an ecosystem may have hydrological impacts as soil and stream flow dynamics are coupled with vegetation processes. The loss and replacement of eastern hemlock by black birch will also result in an increase in ecosystem transpiration during the growing season. I showed that transpiration rates will increase by 1.6 times during the peak growing season. Model simulations suggest the increase in vegetation water use during the peak growing season will impact streamflow. Streamflow will be reduced during the peak growing season months and streams risk becoming ephemeral.

161 The phenology of the ecosystem will also change as a result of the loss and replacement of eastern hemlock. As the evergreen species is replaced by a deciduous species, interception evaporation of snow and rain will be altered. During the dormant season, there will be a decrease in intercepted evaporation of both rain and snow in the replacement stand. The cumulative effect of transpiration over the growing season will significantly reduce late summer streamflow rates in the replacement forests. The disturbance from hemlock woolly adelgid is a natural experiment occurring at a regional scale. By studying ecosystem processes in healthy eastern hemlock forests, we have established a baseline of hemlock ecosystem function. Comparing processes before and after a disturbance will provide valuable insight into ecosystem function. Also, measurements from the expected replacements species allow us to predict expected ecosystem changes. To our knowledge, this research is the first to report whole-tree water use in black birch. As black birch is expected to be the dominant replacement species, this data is critical for predicting changes to ecosystem processes. Given the current trajectory of hemlock woolly adelgid, these predictions will be tested sooner rather than later in northeastern ecosystems. This work contributes to previous research on the affect of species on ecosystem processes. As a component of their studies, hydrologists have always considered the influence of ecology on movements of water. The characteristics and dynamics of vegetation play into rates of interception, evapotranspiration, and infiltration . While paired catchment studies have examined the impacts of clearing and replacing vegetation (reviewed in Hornbeck et al. 1993), hydrologic studies of the loss and replacement of

162 eastern hemlock has not been conducted. In this research, we present estimates of transpiration and evapotranspiration and models of hydrologic processes. We extrapolate our measurements to predict the impact of black birch replacement in dying hemlock stands. Species affect ecosystem processes in many ways besides in water cycling. This research contributes to studies on other ecosystem impacts of hemlock woolly adelgid, such as nitrogen cycling (Jenkins et al. 1999) and fish and wildlife dynamics (Snyder et al. 2002). Combined, this research is critical to understanding natural processes in our ecosystem and enhance our ability to forecast impacts of change.

163

REFERENCES Gurnell, A., C. Hupp and S. Gregory 2000. Linkng hydrology and ecology. Hydrological Processes 14:2813-2815. Hornbeck, J.W., M.B. Adams, E.S. Corbett, E.S. Verry and J.A. Lynch 1993. Long-Term Impacts of Forest Treatments on Water Yield - a Summary for Northeastern USA. Journal of Hydrology 150:323-344. Jenkins, J., J. Aber and C. Canham 1999. Hemlock woolly adelgid impacts on community structure and N cycling rates in eastern hemlock forests. Canadian Journal of Forest Research 29:630-645. Snyder, C.D., J.A. Young, D.P. Lemarie and D.R. Smith 2002. Influence of eastern hemlock (Tsuga canadensis) forests on aquatic invertebrate assemblages in headwater streams. Canadian Journal of Fisheries and Aquatic Sciences 59:262275.

164

References Abrams, M.D. 1990. Adaptations and response to drought in Quercus species of North America. Tree Physiology 7:227-238. Avissar, R. and R.A. Pielke 1991. The impact of plant stomatal control on mesoscale atmospheric circulations. Agricultural and Forest Meteorology 54:353-372. Avissar, R. 1992. Conceptual aspects of a statistical-dynamic approach to represent landscape subgrid-scale heterogeneities in atmospheric models. Journal of Geophysical Research-Atmospheres. 97:2729-2742. Baldocchi, D.D., B.B. Hicks and T.P. Meyers 1988. Measuring biosphere-atmosphere exchanges of biologically related gases with micrometeorological methods. Ecology 69:1331-1340 Baldocchi, D.D. and K.S. Rao 1995. Intra-field variability of scalar flux densities across a transition between a desert and an irrigated potato field. Boundary-Layer Meteorology. 76:109-136. Bazzaz, F.A. 1979. The physiological ecology of plant succession. Annual Review of Ecology And. Systematics 10: 351-271. Benyon, R.G. 1999. Nighttime water use in an irrigated Eucalyptus grandis plantation. Tree Physiology 19:853-859. Beven, K.J., (Ed.) 1997. Distributed modelling in hydrology: applications of TOPMODEL Ed. Wiley, Chichester, p. 330. Bhiry, N. and L. Filion 1996. Mid-Holocene hemlock decline in eastern North America linked with phytophagous insect activity. Quaternary Research 45:312-320.

165 Bonan, G.B., D. Pollard and S.L. Thompson 1992. Effects of Boreal Forest Vegetation on Global Climate. Nature 359:716-718. Bonan, G.B., S. Levis, S. Sitch, M. Vertenstein and K.W. Oleson 2003. A dynamic global vegetation model for use with climate models: concepts and description of simulated vegetation dynamics. Global Change Biology 9:1543-1566. Bond, B.J., J.A. Jones, G. Moore, N. Phillips, D. Post and J.J. McDonnell 2002. The zone of vegetation influence on baseflow revealed by diet patterns of streamflow and vegetation water use in a headwater basin. Hydrological Processes 16:1671-1677. Bosch, J. and J. Hewlett 1982. A review of catchment experiments to determine the effect of vegetation changes on water yield and evapotranspiration. Journal of Hydrology 55:3-23. Brown, A.E., L. Zhang, T.A. McMahon, A.W. Western and R.A. Vertessy 2005. A review of paired catchment studies for determining changes in water yield resulting from alterations in vegetation. Journal of Hydrology 310:28-61. Bucci, S.J., F.G. Scholz, J.L. Goldstein, F.C. Meinzer, J.A. Hinojosa, W.A. Hoffman and C.F. Augusto. 2004. Processes preventing nocturnal equilibrium between leaf and soil water potential in tropical savanna woody species. Tree Physiology 24:11191127. Burns, R.M. and B.H. Honkala 1990. Silvics of North American trees. 1. Conifers; 2. Hardwoods. Agriculture Handbook 654. Vol. 2. USDA Forest Service, Washington, DC, 877p.

166 Caspari, H.W., S.R. Green and W.R.N. Edwards 1993. Transpiration of well-watered and water-stressed Asian pear trees as determined by lysimetry, heat-pulse, and estimated by a Penman-Monteith model. Agricultural Forest Meteorolology 67:13-27. Castello, J.D., D.J. Leopold and P.J. Smallidge 1995. Pathogens, Patterns, and Processes in Forest Ecosystems. BioScience 45:16-24. Catovsky, S. and F. Bazzaz. 2000. Contributions of coniferous and broad-leaved species to temperate forest carbon uptake: a bottom-up approach. Canadian Journal of. Forest Research 30:100-111. Catovsky, S. and F.A. Bazzaz 2000. The role of resource interactions and seedling regeneration in maintaining a positive feedback in hemlock stands. Journal of Ecology 88:100-112. Catovsky, S., N.M. Holdbrook and F.A. Bazzaz 2002. Coupling whole-tree transpiration and canopy photosynthesis in coniferous and broad-leaved tree species. Canadian Journal of Forest Research 32:295-309. Cernusak, L.A., G.D. Farquhar, S.C. Wong and H. Stuart-Williams. 2004. Measurement and interpretation of the oxygen isotope composition of carbon dioxide respired by leaves in the dark. Plant Physiology 136:1-14. Clapp, R.B. and G.M. Hornberger 1978. Empirical Equations for Some Soil HydraulicProperties. Water Resources Research 14:601-604.

167 Clearwater, M.J., F.C. Meinzer, J.L. Andrade, G. Goldstein and N.M. Holbrook 1999. Potential errors in measurement of non-uniform sap flow using heat dissipation probes. Tree Physiology 19:681-687. Cowan, I.R. 1977. Stomatal behavior and environment. In Advances in Botanical Research Eds. R.D. Preston and H.W. Woodhouse. Academic Press, San Francisco, pp. 117-228. Cramer, W., A. Bondeau, F.I. Woodward, I.C. Prentice, R.A. Betts, V. Brovkin, P.M. Cox, V. Fisher, J.A. Foley, A.D. Friend, C. Kucharik, M.R. Lomas, N. Ramankutty, S. Sitch, B. Smith, A. White and C. Young-Molling 2001. Global response of terrestrial ecosystem structure and function to CO2 and climate change: results from six dynamic global vegetation models. Global Change Biology. 7:357-373. Czikowsky, M.J. and D.R. Fitzjarrald 2004. Evidence of seasonal changes in evapotranspiration in eastern US hydrological records. Journal of Hydrometeorology 5:974-988. Davis, M.B. 1981. Outbreaks of forest pathogens in Quaternary history. In Proceedings of the IV International Palynological Conference, Lucknow, pp. 216-227. Dunn, T. and L.B. Leopold. 1978. Water in Environmental Planning. Freeman and Co., New York. 818 p. Edwards, W., P. Jarvis, J. Landsberg and H. Talbot 1986. A dynamic model for studying flow of water in single trees. Tree Physiology. 1:309-324. Ellison, A.M., M.S. Bank, B.D. Clinton, E.A. Colburn, K. Elliott, C.R. Ford, D.R. Foster, B.D. Kloeppel, J.D. Knoepp, G.M. Lovett, J. Mohan, D.A. Orwig, N.L.

168 Rodenhouse, W.V. Sobczak, K.A. Stinson, J.K. Stone, C.M. Swan, J. Thompson, B. Von Holle and J.R. Webster 2005. Loss of foundation species: consequences for the structure and dynamics of forested ecosystems. Frontiers in Ecology and the Environment 3:479-486. Enserink, M. 1999. Biological invaders sweep in. Science 285:1834-1836. Eschtruth, A.K., N.L. Cleavitt, J.J. Battles, R.A. Evans and T.J. Fahey 2006. Vegetation dynamics in declining eastern hemlock stands: 9 years of forest response to hemlock woolly adelgid infestation. Canadian Journal of Forest Research 36:1435-1450. Evans, R.A. 2004. Hemlock woolly adelgid and the disintegration of eastern hemlock ecosystems. Park Science 22: 53-56. Fahey, T.J., T.G. Siccama, C.T. Driscoll, G.E. Likens, J. Campbell, C.E. Johnson, J.J. Battles, J.D. Aber, J.J. Cole, M.C. Fisk, P.M. Groffman, S.P. Hamburg, R.T. Holmes, P.A. Schwarz and R.D. Yanai 2005. The biogeochemistry of carbon at Hubbard Brook. Biogeochemistry 75:109-176. Federer, C. 1980. Paper birch and white oak saplings differ in responses to drought. Forest Science. 2:313-324. Federer, C.A. 2002. BROOK 90: A simulation model for evaporation, soil water, and streamflow. Federer, C.A., C. Vorosmarty and B. Fekete 2003. Sensitivity of annual evaporation to soil and root properties in two models of contrasting complexity. Journal of Hydrometeorology 4:1276-1290.

169 Finzi, A.C., C.D. Canham and N. Van Breeman 1998. Canopy tree-soil interactions within temperate forests: Species effects on pH and cations. Ecological Applications. 8:905-905. Foley, J.A., M.H. Costa, C. Delire, N. Ramankutty and P. Snyder 2003. Green surprise? How terrestrial ecosystems could affect earth's climate. Frontiers in Ecology and the Environment 1:38-44. Folland, C.K., T.R. Karl, J.R. Christy, R.A. Clarke, G.V. Gruza, J. Jouzel, M.E. Mann, O. J., S. M.J. and S.-W. Wang 2001. Observed climate variability and change. In Climate Change: The Scientific Basis Ed. J.T. Houghton, et al. Cambridge Univ. Press, Cambridge, pp. 99-181. Ford, C., M. McGuire, R. Mitchell, and R. Teskey 2004. Assessing variation in the radial profile of sap flux density in Pinus species and its effect on daily water use. Tree Physiology 24:241-249. Foster, D.R. and. T. M. Zebryk 1993. Long-term vegetation dynamics and disturbance history of a Tsuga-dominated forest in New England. Ecology 74:982-998. Foster, D.R., D.H. Knight, and J.F. Franklin 1998. Landscape patterns and legacies resulting from large, infrequent forest disturbances. Ecosystems 1:497-510. Foster, D.R., W.W. Oswald, E.K. Faison, E.D. Doughty, and B.C.S. Hansen 2006. A climatic driver for abrupt mid-holocene vegetation dynamics and the hemlock decline in New England. Ecology 87:2959–2966. Gansert, D. 2003. Xylem sap flow as a major pathway for oxygen supply to the sapwood of birch (Betula pubescens Ehr.). Plant Cell Environment 26:1803-1814.

170 Gansert, D., Burgdorf, M. & Lösch, R. 2001. A novel approach to the in situ measurement of oxygen concentrations in the sapwood of woody plants. Plant Cell Environment 24:1055-1064. Gedney, N., P.M. Cox, R.A. Betts, O. Boucher, C. Huntingford and P.A. Stott 2006. Detection of a direct carbon dioxide effect in continental river runoff records. Nature 439:835-838. Goldstein, G., J.L. Andrade, F.C. Meinzer, N.M. Holbrook, J. Cavalier, P. Jackson and A. Celis. 1998. Stem water storage and diurnal patterns of water use in tropical forest canopy trees. Plant Cell Environment 21:397-406. Gonzalez, A.B. 2000. Hydrologic effects of vegetative practices on ponderosa pine watersheds in Arizona. In School of Renewable Natural Resources. The University of Arizona, p. 121. Goulden, M.L., J.W. Munger, S.M. Fan, B.C. Daube, and S.C. Wofsy 1996. Measurements of carbon sequestration by long-term eddy covariance: Methods and a critical evaluation of accuracy. Global Change Biology 2:169-182. Granier, A. 1985. A New Method of Sap Flow Measurement in Tree Stems. Annals of Forest Science 42:193-200. Green, S.R., K.G. McNaughton and B.E. Clothier 1989. Observations of nighttime water use in kiwifruit vines and apple trees. Agricultural Forest Meteorolology 48:251261.

171 Grinsted, A., J.C. Moore and S. Jevrejeva 2004. Application of the cross wavelet transform and the wavelet coherence to geophysical time series. Nonlinear Processes in Geophysics 11:561-566. Grulke, N.E., R. Alonso, T. Nguyen, C. Cascio and W. Dobrowolski 2004. Stomata open at night in pole-sized and mature ponderosa pine: implications for O3 exposure metrics. Tree Physiology 24:1001-1010. Günthardt-Goerg, M.S., R. Matyssek, C. Scheidegger and T. Keller. 1993. Differentiation and structural decline in the leaves and bark of birch (Betula pendula) under low ozone concentration. Trees 7:104-114. Gurnell, A., C. Hupp and S. Gregory 2000. Linkng hydrology and ecology. Hydrological Processes 14:2813-2815. Hadley, J.L. 2000a. Effect of daily minimum temperature on photosynthesis in eastern hemlock (Tsuga canadensis L.) in autumn and winter. Arctic Antarctic and Alpine Research 32:368-374. Hadley, J.L. 2000b. Understory microclimate and photosynthetic response of saplings in an old-growth eastern hemlock (Tsuga canadensis L.) forest. Ecoscience 7:66-72. Hadley, J.L. and J.L. Schedlbauer 2002. Carbon exchange of an old-growth eastern hemlock (Tsuga canadensis) forest in central New England. Tree Physiology 22:1079-1092. Hatton, T.J., Catchpole, E.A. and Vertessy, R.A. 1990. Integration of sapflow velocity to estimate plant water use. Tree Physiology 6:201-209.

172 Hogg, E.H. and P.A. Hurdle 1997. Sap flow in trembling aspen: implications for stomatal responses to vapor pressure deficit. Tree Physiology 17:501-509. Hooper, D. and P. Vitousek 1997. The effects of plant composition and diversity on ecosystem processes. Science 277:1302-1305. Hooper, D., F. Chapin, J. Ewel, A. Hector, P. Inchausti, S. Lavorel, J. Lawton, D. Lodge, M. Loreau, S. Naeem, B. Schmid, H. Setala, A. Symstad, J. Vandermeer and D. Wardle 2005. Effects of biodiversity on ecosystem functioning: a consensus of current knowledge. Ecological Monographs 75:3-35. Hornbeck, J.W., M.B. Adams, E.S. Corbett, E.S. Verry and J.A. Lynch 1993. Long-Term Impacts of Forest Treatments on Water Yield - a Summary for Northeastern USA. Journal of Hydrology 150:323-344. Hornberger, G. 2001. A Plan for a New Science Initiative on the Global Water Cycle. Draft report from the Water Cycle Study Group, U.S. Global Change Research Program. Hunt, E. and P.S. Nobel 1987. Non-steady state water flow for three desert perennials with different capacitances. Australian Journal of Plant Physiology. 14:363-375. Hunt, E.R., S.W. Running and C.A. Federer 1991. Extrapolating Plant Water-Flow Resistances and Capacitances to Regional Scales. Agricultural and Forest Meteorology 54:169-195. Iritz, Z. and A. Lindroth 1994. Night-time evaporation from a short-rotation willow stand. Journal of Hydrology 157:235-245.

173 Jackson, R.B., J. Canadell, J.R. Ehleringer, H.A. Mooney, O.E. Sala and E.D. Schulze 1996. A global analysis of root distributions for terrestrial biomes. Oecologia 108:389-411. James, S., M. Clearwater, F. Meinzer, and G. Goldstein 2002. Heat dissipation sensors of variable length for the measurement of sap flow in trees with deep sapwood. Tree Physiology 22:277-283. Jarvis, P.G. 1976. The interpretation of the variations in leaf water potential and stomatal conductance found in canopies in the field. Philosophical Transactions of the Royal Society of London Series B. 273:593 510. Jarvis, P.G. and J.W. Leverenz 1983. Productivity of temperate, deciduous and evergreen forests. In Encyclopedia of Plant Physiology, vol. 12D, Physiological Plant Ecology. Springer-Verlag, Berlin, pp. 234-280. Jarvis, P.G. and K.G. McNaughton 1986. Stomatal Control of Transpiration - Scaling up from Leaf to Region. Advances in Ecological Research 15:1-49. Jeje, A.A. and M.H. Zimmermann 1979. Resistance to water flow in xylem vessels. Journal of Experimental Botany. 30:817-827. Jenkins, J., J. Aber and C. Canham 1999. Hemlock woolly adelgid impacts on community structure and N cycling rates in eastern hemlock forests. Canadian Journal of Forest Research 29:630-645. Kim, H. 2000. Transpiration of hybrid poplar in eastern Washington. In College of Forest Resources. University of Washington, Seattle, p. 107.

174 Kittredge, D.B. and P.M. Ashton 1990. Natural regeneration patterns in even-aged mixed stands in southern New England. Northern Journal of Applied Forestry 7:163-168. Kizlinski, M.L., D.A. Orwig, R.C. Cobb and D.R. Foster 2002. Direct and indirect ecosystem consequences of an invasive pest on forests dominated by eastern hemlock. Journal of Biogeography 29:1489-1503. Kramer, P.J. 1937. The relation between rate of transpiration and rate of absorption of water in plants. American Journal of Botany 24:10-15. Kueppers, L.M. and J. Harte 2005. Subalpine forest carbon cycling: Short- and long-term influence of climate and species. Ecological Applications 15:1984-1999. Lhomme, J.P. 2001. Stomatal control of transpiration: Examination of the Jarvis-type representation of canopy resistance in relation to humidity. Water Resources Research 37:689-699. Liebhold, A.M., W.L. MacDonald, D. Bergdahl and V. Mastro 1995. Invasion by exotic forest pests: a threat to forest ecosystems. Forest Science Monograph 30:1-49. Lopushinsky, W. 1986. Seasonal and diurnal trends of heat pulse velocity in Douglas-fir and ponderosa pine. Canadian Journal of Forest Research 16:814-821. Loustau, D., P. Berbigier, P. Roumagnac, C. Arruda-Pacheco, J.S. David, M.I. Ferreira, J.S. Pereira and R. Tavares. 1996. Transpiration of a 64-year-old maritime pine stand in Portugal. 1. Seasonal course of water flux through maritime pine. Oecologia 107:33–42.

175 Lovett, G.M., C.D. Canham, M.A. Arthur, K.C. Weathers and R.D. Fitzhugh 2006. Forest ecosystem responses to exotic pests and pathogens in eastern North America. Bioscience 56:395-405. Lu, P., W. Muller and E. Chacko 2000. Spatial variations in xylem sap flux density in the trunk of orchard-grown, mature mango trees under changing soil water conditions. Tree Physiology 20:683-692. Magnani, F. and M. Borghetti 1995. Interpretation of seasonal-changes of xylem embolism and plant hydraulic resistance in Fagus-sylvatica. Plant Cell and Environment. 18:689-696.

Maherali, H. and E.H. DeLucia. 2001. Influence of climate-driven shifts in biomass allocation on water transport and storage in ponderosa pine. Oecologia 129:481– 491. Mancuso, S. and A.M. Marras 2003. Different pathways of the oxygen supply in the sapwood of young Olea europaea trees. Planta online. DOI 10.1007/s00425-0020956-5. Manning, P., J.E. Newington, H.R. Robson, M. Saunders, T. Eggers, M.A. Bradford, R.D. Bardgett, M. Bonkowski, R.J. Ellis, A.C. Gange, S.J. Grayston, E. Kandeler, S. Marhan, E. Reid, D. Tscherko, H.C.J. Godfray and M. Rees 2006. Decoupling the direct and indirect effects of nitrogen deposition on ecosystem function. Ecology Letters 9:1015-1024.

176 Martin, T.A., K.J. Brown, J. Èermák, R. Ceulemans, J. Kuèera, F.C. Meinzer, J.S. Rombold, D.G. Sprugel and T.M. Hinckley. 1997. Crown conductance and tree and stand transpiration in a second- growth Abies amabilis forest. Canadian Journal of Forest Research. 27:797–808. Massie, M.R.C., E.B. Peterson, N.M. Peterson and K.A. Enns. 1994. An assessment of the stategic importance of the hardwood resource in British Columbia. BC Min. For., Victoria, BC, FRDA Rep. No. 221. Matssyek, R., M.S. Günthardt-Goerg, S. Maurer and T. Keller. 1995. Nighttime exposure to ozone reduces whole-plant production in Betula pendula. Tree Physiology 15:159-165. McClaugherty, C.A., J. Pastor, J.D. Aber and J.M. Melillo 1985. Forest Litter Decomposition in Relation to Soil-Nitrogen Dynamics and Litter Quality. Ecology 66:266-275. McClure, M.S. 1991. Density-dependent feedback and population cycles in Adelges tsugae (Homoptera: Adelgidae) on Tsuga canadensis. Environmental. Entomolology 20:258-264. McDowell, N., H. Barnard, B.J. Bond, T. Hinckley, R.M. Hubbard, H. Ishii, B. Kostner, F. Magnani, J.D. Marshall, F.C. Meinzer, N. Phillips, M.G. Ryan and D. Whitehead 2002. The relationship between tree height and leaf area: sapwood area ratio. Oecologia 132:12-20. McDowell, N.G., D.R. Bowling, A. Schauer, J. Irvine, B.J. Bond, B.E. Law and J.R. Ehleringer. 2004. Associations between carbon isotope ratios of ecosystem

177 respiration, water availability and canopy conductance. Global Change Biology 10:1767-1784. McNaughton, K.G. 1976. Evaporation and advection .2. Evaporation downwind of a boundary separating regions having different surface resistances and available energies. Quarterly Journal of the Royal Meteorological Society. 102:193-202. McWilliams, W. and T. Schmidt 2000. Composition, structure, and sustainability of hemlock ecosystems in eastern North America. In Proceedings: Sympossium on sustainable management of hemlock ecosystems in eastern North America Eds. K. McManus, K. Shields and D. Souto. USDA Forest Service, Durham, NH. Meinzer, F.C., S.A. James and G. Goldstein 2004. Dynamics of transpiration, sap flow and use of stored water in tropical forest canopy trees. Tree Physiology 24:901909. Mladenoff, D.J. 1987. Dynamics of nitrogen mineralization and nitrification in hemlock and hardwood treefall gaps. Ecology 68:1171-1180. Musselman, R.C. and T.J. Minnick. 2000. Nocturnal stomatal conductance and ambient air quality standards for ozone. Atmospheric Environment 34:719-733. Nobel, P.S. and P.W. Jordan 1983. Transpiration stream of desert species - resistances and Ccapacitances for a C-3, a C-4, and a CAM plant. Journal of Experimental Botany 34:1379-1391. Ogata K. 1990. Modern Control Engineering, 2nd Edition. Prentice Hall, New Jersey

178 Oren, R., J. Sperry, G. Katul, D. Pataki, B. Ewers, N. Phillips and S. Schaeffer 1999. Survey and synthesis of intra- and interspecific variation in stomatal sensitivity to vapour pressure deficit. Plant Cell Environment 22:1515-1526. Oren, R., J.S. Sperry, B.E. Ewers, D.E. Pataki, N. Phillips and J.P. Megonigal 2001. Sensitivity of mean canopy stomatal conductance to vapor pressure deficit in a flooded Taxodium distichum L. forest: hydraulic and non-hydraulic effects. Oecologia 126:21-29. Orwig, D., D. Foster, and D. Mausel 2002. Landscape patterns of hemlock decline in New England due to the introduced hemloc woolly adelgid. Journal of Biogeography 29:1517-1530. Orwig, D.A. 2002. Stand dynamics associated with chronic hemlock woolly adelgid infestations in southern New England. In Symposium on the Hemlock Woolly Adelgid in eastern North America proceedings. (eds. R.C. Reardon, Onken, B.P., and Lashomb, J.) pp. 36-47. New Jersey Agricultural Experiment Station Publication, New Brunswick, NJ. Orwig, D.A. and D.R. Foster 1998. Forest response to the introduced hemlock woolly adelgid in southern New England, USA. Journal of the Torrey Botanical Society 125:60-73. Pallardy, S.G., J. Èermák, F.W. Ewers, M.R. Kaufmann,W.C. Parker and J.S. Sperry. 1995. Water transport dynamics in trees and stands. In Resource Physiology of Conifers. Eds. W.K. Smith and T.M. Hinckley. Academic Press, New York, pp 301–389.

179 Pataki, D.E. and R. Oren 2003. Species differences in stomatal control of water loss at the canopy scale in a mature bottomland deciduous forest. Advances in Water Resources 26:1267-1278. Paulson, R.W., E.B. Chase, R.S. Roberts and D.W. Moody 1991. National water summary 1988-89—Hydrologic events and floods and droughts. Water-Supply Paper 2375. USGS (U.S. Geological Survey), Washington, D.C., p. 591. Phillips, N., A. Nagchaudhuri, R. Oren and G. Katul 1997. Time constant for water transport in loblolly pine trees estimated from time series of evaporative demand and stem sapflow. Trees-Structure and Function. 11:412-419.

Phillips, N., R. Oren, and R. Zimmerman 1996. Radial patterns of xylem sap flow in nondiffuse, diffuse- and ring-porous tree species. Plant Cell and Environment 19:983990. Phillips, N.G., M.G. Ryan, B.J. Bond, N.G. McDowell, T.M. Hinckley, and J. Cermak 2003. Reliance on stored water increases with tree size in three species in the Pacific Northwest. Tree Physiology 23:237-245. Phillips, N.G., R. Oren, J. Licata and S. Linder 2004. Time series diagnosis of tree hydraulic characteristics. Tree Physiology 24:879-890. Powell, D.B.B. and M.R. Thorpe 1977. Dynamic aspects of plant-water relations. In Environmental Effects of Crop Physiology Eds. J.J. Landsberg and C.V. Cutting. Academic Press, London, pp. 259-279.

180 Richards L A and Weaver L R 1944 Moisture retention by some irrigated soils as related to soil moisture tension. Journal of Agricultural Research 69: 215-235. Ross, R. M., R. M. Bennett, C. D. Snyder, J.A. Young, D.R. Smith, and D.P. Lemarie 2003. Influence of eastern hemlock (Tsuga canadensis) on fish community structure and function in headwater streams of the Delaware River basin. Ecology of Freshwater Fish 12:60–65. Running, S.W. 1980. Relating plant capacitance to the water erlations of Pinus contorta. Journal of Experimental Botany. 27:473-479. Ryan, M.G. and B.J. Yoder 1997. Hydraulic Limits to Tree Height and Tree Growth. BioScience 47:235-242. Salvucci, G.D. 1997. Soil and moisture independent estimation of stage-two evaporation from potential evaporation and albedo or surface temperature. Water Resources Research 33:111-122. Schmid, H.P. 1994. Source areas for scalars and scalar fluxes. Boundary-Layer Meteorology 67:293-318. Scholz, F.G., S.J. Bucci, G. Goldstein, F.C. Meinzer, A.C. Franco and F. MirallesWilhelm 2007. Biophysical properties and functional significance of stem water storage tissues in Neotropical savanna trees. Plant Cell and Environment. 30:236248. Schulze, E.D., J. Èermák, R. Matyssek, M. Penka, R. Zimmermann, F. Vasicek, W. Gries and J. Kuèera 1985. Canopy transpiration and water fluxes in the xylem of the

181 trunk of Larix and Picea trees—a comparison of xylem flow, porometer and cuvette measurements. Oecologia 66:475–483. Schulze, E.D., M. Fuchs and M.I. Fuchs 1977. Spacial Distribution of Photosynthetic Capacity and Performance in a Mountain Spruce Forest of Northern Germany .3. Significance of Evergreen Habit. Oecologia 30:239-248. Scholz, F.G., S.J. Bucci, G. Goldstein, F.C. Meinzer, A.C. Franco and F. MirallesWilhelm 2007. Biophysical properties and functional significance of stem water storage tissues in Neotropical savanna trees. Plant Cell and Environment. 30:236248. Sellers, P.J., R.E. Dickinson, D.A. Randall, A.K. Betts, F.G. Hall, J.A. Berry, G.J. Collatz, A.S. Denning, H.A. Mooney, C.A. Nobre, N. Sato, C.B. Field and A. Henderson Sellers 1997. Modeling the exchanges of energy, water, and carbon between continents and the atmosphere. Science 275:502-509. Shaw, E.M. 1994. Hydrology in Practice. Chapman & Hall, New York. Shuttleworth, W.J. and J.S. Wallace 1985. Evaporation from Sparse Crops - an Energy Combination Theory. Quarterly Journal of the Royal Meteorological Society 111:839-855. Smith, D.M. and P.M. Ashton, 1993. Early dominance of pioneer hardwood after clearcutting and removal of advanced regeneration. Northern Journal of Applied Forestry 10: 14-19.

182 Smith, W.K., D.R. Young, G.A. Carter, J.L. Hadley and G.M. McNaughton 1984. Autumn stomatal closure in six conifer species of the Central Rocky Mountains. Oecologia 63:237-242. Snyder, C.D., J.A. Young, D.P. Lemarie and D.R. Smith 2002. Influence of eastern hemlock (Tsuga canadensis) forests on aquatic invertebrate assemblages in headwater streams. Canadian Journal of Fisheries and Aquatic Sciences 59:262275. Snyder, K.A., J.H. Richards and L.A. Donavan. 2003. Night-time conductance in C3 and C4 species: do plants lose water at night? Journal of Experimental Botany 54:861865. Spicer, R. and N. Holbrook 2005. Within-stem oxygen concentration and sap flow in four temperate tree species: does long-lived xylem parenchyma experience hypoxia? Plant, Cell and Environment 28:192-201. Stadler, B., T. Muller and D. Orwig 2006. The ecology of energy and nutrient fluxes in hemlock forests invaded by hemlock woolly adelgid. Ecology 87:1792-1804. Steppe, K. and R. Lemeur 2007. Effects of ring-porous and diffuse-porous stem wood anatomy on the hydraulic parameters used in a water flow and storage model. Tree Physiology. 27:43-52. Stewart, J.B. 1977. Evaporation from Wet Canopy of a Pine Forest. Water Resources Research 13:915-921. Swank, W.T. and J.E. Douglas 1974. Streamflow greatly reduced by converting hardwood stands to pine. Science 185:857-859.

183 Templer, P.H., G.M. Lovett, K.C. Weathers, S.E. Findlay and T.E. Dawson 2005. Influence of tree species on forest nitrogen retention in the Catskill Mountains, New York, USA. Ecosystems 8:1-16. Tilman, D., J. Knops, D. Wedin, P. Reich, M. Ritchie and E. Siemann 1997. The influence of functional diversity and composition on ecosystem processes. Science 277:1300-1302. Tobiessen, P. 1982. Dark opening of stomata in successional trees. Oecologia 52:356-359. Tognetti, R., A. Longobucco and A. Raschi 1998. Vulnerability of xylem to embolism in relation to plant hydraulic resistance in Quercus pubescens and Quercus ilex cooccurring in a Mediterranean coppice stand in central Italy. New Phytologist. 139:437-447. Tolk, J.A., S.R. Evett and T.A. Howell 2006. Advection influences on evapotranspiration of alfalfa in a semiarid climate. Agronomy Journal. 98:1646-1654. Torrence, C. and G.P. Compo 1998. A practical guide to wavelet analysis. Bulletin of the American Meteorological Society 79:61-78. Tyree, M.T. and F.W. Ewers 1991. The Hydraulic Architecture of Trees and Other Woody-Plants. New Phytologist. 119:345-360. USDA Forest Service 2007. Hemlock woolly adelgid infestations 2005. Available from http://www.na.fs.fed.us/fhp/hwa/maps/hwa_2005.jpg [accessed 31 January 2007]. USDA Forest Service. 2006. List of States and Counties with Known HWA Infestations. Available from http://www.na.fs.fed.us/fhp/hwa/hwatable_web/hwatable4.pdf [accessed 31 January 2007].

184 USDA Forest Service. 2007. Forest inventory and analysis national program. Available from http://fia.fs.fed.us [accessed 17 January 2007]. USDA Forest Service. 2007. Soil Classification: Official Series Descriptions. Available from http://soils.usda.gov/technical/classification/ [accessed 17 January 2007]. Vitousek, P.M., C.M. Dantonio, L.L. Loope, M. Rejmanek and R. Westbrooks 1997. Introduced species: A significant component of human-caused global change. New Zealand Journal of Ecology 21:1-16. Ward, J.S. and G.R. Stephens, 1996. Influence of crown class on survival and development of Betula lenta in Connecticut, U.S.A. Canadian Journal of Forest Research 26: 277-288. Waring, R., D. Whitehead and P. Jarvis 1979. The contribution of stored water to transpiration in Scot pine. Plant Cell Environment 2:309-318. Wedin, D.A. and D. Tilman 1996. Influence of nitrogen loading and species composition on the carbon balance of grasslands. Science 274:1720-1723. Woodwell, G.M., F.T. Mackenzie, R.A. Houghton, M. Apps, E. Gorham and E. Davidson 1998. Biotic feedbacks in the warming of the earth. Climatic Change 40:495-518. Wullschleger, S. and A.W. King 2000. Radial variation in sap velocity as a function of stem diameter and sapwood thickness in yellow-poplar trees. Tree Physiology 20:511-518. Wullschleger, S.D., F. Meinzer, and R. Vertessy 1998. A review of whole-plant water use studies in trees. Tree Physiology 18:499-512.

185 Yorks, T.E., D.J. Leopold and D.J. Raynal 2003. Effects of Tsuga canadensis mortality on soil water chemistry and understory vegetation: possible consequences of an invasive insect herbivore. Canadian Journal of Forest Research 33:1525-1537. Young, R., K. Shields and G. Berlyn 1995. Hemlock woolly adelgid (Homoptera: Adelgidae): Stylet bundle insertion and feeding sites. Annals of the Entomological Society of America 88:827-835. Zimmermann, M.H. 1983. Xylem Structure and the Ascent of Sap. Springer-Verlag, New York. 143 p. Zimmermann, M.H. and C.L. Brown 1971. Trees: Structure and Function. SpringerVerlag, New York. 336 p.

186

Curriculum Vita MICHAEL J. DALEY 2649 Harvard Yard Mail Center Cambridge, MA 02138 (617) 493-8238 / [email protected] EDUCATION Boston University, Boston MA PhD Geography and Environment, May 2007 Advisor: Dr. Nathan Phillips State University of New York at Plattsburgh, Plattsburgh, NY MS Teaching of Secondary Education in Biology, May 2000 Siena College, Loudonville, NY BS Biology, Cum Laude, May 1998 AWARDS AND HONORS National Science Foundation Graduate Research Fellowship Program Area: Life Sciences – Botany

2004 - 2007

PROFESSIONAL EXPERIENCE LASELL COLLEGE, Auburndale, MA Fall 2004 Adjunct Faculty, Department of Social Sciences Fall 2005 • Developed and taught an introductory course in World Geography BOSTON UNIVERSITY, Boston, MA Department of Geography and Environment National Science Foundation Graduate Research Fellow Ecohydrology Research Assistant Teaching Fellow in Physical Geography

2004 - Present 2003- 2004 2002 - 2003

THE JASON FOUNDATION FOR EDUCATION, Needham, MA Science Curriculum Developer 2003 – 2005 MCKELVIE MIDDLE SCHOOL, Bedford, NH Teacher- 8th Grade Earth and Environmental Science

2000 - 2002

STATE UNIVERSITY OF NEW YORK AT PLATTSBURGH, Plattsburgh, NY Teaching Assistant in Ecology Fall 1999

187

188 PUBLICATIONS AND PRESENTATIONS Peer Reviewed Publications Daley, M.J. and N. Phillips. (2006) Interspecific variation in nighttime transpiration and stomatal conductance in a mixed New England deciduous forest. Tree Physiology 26:411-419. Daley, M.J., N.G. Phillips, J.C. Pettijohn, and J. Hadley. (2007) Water use by eastern hemlock (Tsuga canadensis) and black birch (Betula lenta): Implications of effects of the hemlock wooly adelgid. Canadian Journal of Forest Research. (in review). Daley, M.J., N.G. Phillips, J. Hadley, E. Boose, J.C. Pettijohn, and G. Salvucci. (2006) Modeling water flow dynamics in eastern hemlock and black birch forests: hydrological impacts of hemlock woolly adelgid. (in prep). Daley, M.J., N.G. Phillips, J.C. Pettijohn, and J. Hadley. (2006) Changes in ecosystem function due to the loss and replacement of eastern hemlock. (in prep). Ford, C.R., J.M. Vose, M.J. Daley, and N.G. Phillips. (2006) Eastern hemlock water use: implications for systemic insecticide application. Journal of Arboriculture. (in review). Fraser, D.F., J.F. Gilliam, M.J. Daley, A.N. Le, and G.T. Skalski. (2001) Explaining Leptokurtic Movement Distributions: Intrapopulation Variation in Boldness and Exploration. The American Naturalist 158:124-135. Meetings and Presentations Daley, M.J. (2006) Biotic disturbance of ecohydrologic function: the case of eastern hemlock forests. Invited seminar, Department of Biology, Boston University, Boston, MA. Daley, M.J., J. Hadley, J.C. Pettijohn, and N.G. Phillips (2006) Changes in ecohydrological function due to the loss and replacement of eastern hemlock in a New England forest. International Union of Forest Research Organizations, Canopy Processes Working Group Workshop: Regional Forest Responses to Environmental Change, Northeastern, USA. Hadley, J., P. Kuzeja, M.J. Daley, T. Mulcahy, and J Schedlbauer (2006) Differences in carbon/water cycling between early-successional deciduous forest and late-successional conifer forests: Implications for long-term effects of invasive hemlock woolly adelgid. International Union of Forest Research Organizations, Canopy Processes Working Group Workshop: Regional Forest Responses to Environmental Change, Northeastern, USA.

189

Pettijohn, J.C., G. Salvucci, N. Phillips, and M.J. Daley (2006) A comparison of long-term irrigated and non-irrigated red maple transpiration. International Union of Forest Research Organizations, Canopy Processes Working Group Workshop: Regional Forest Responses to Environmental Change, Northeastern, USA. Daley, M.J., J. Hadley, J.C. Pettijohn, and N.G. Phillips (2006) Changes in ecohydrological function due to the loss and replacement of a core species of tree in a New England ecosystem. Contributed Oral Session, Ecological Society of America, Memphis, TN. Daley, M.J., J. Hadley, J.C. Pettijohn, and N.G. Phillips (2006) A comparison of transpiration in black birch and eastern hemlock stands. Harvard Forest Seventeenth Annual Ecology Symposium, Harvard Forest, Petersham, MA. Daley, M.J. (2006). Ecosystem water balance: consequences of an invasive pest in a New England forest. Boston University Science & Engineering Symposium, Boston, MA. Pettijohn, J.C., N.G. Phillips, G.D. Salvucci, M.J. Daley (2006) An investigation of vegetative and atmospheric conductance forcing on the Bouchet-Morton Complementary Relationship Hypothesis using sap flux observations at Harvard Forest. Western Pacific Geophysics Meeting, Beijing, China. Daley, M.J. (2005) Inventory of Eastern Hemlock in the Watersheds of Massachusetts and the Threat of Hemlock Woolly Adelgid. Executive Office of Environmental Affairs. Boston, MA. Daley, M.J. and N.G. Phillips (2005) Interspecific variation in nighttime transpiration and stomatal conductance in a mixed New England deciduous forest. Invited Oral Session, Ecological Society of America, Montreal, Canada. Daley, M.J. and N.G. Phillips (2004) Interspecific variation in nighttime transpiration and stomatal conductance in a mixed New England deciduous forest. Ecological Society of America, Portland, OR. PROFESSIONAL ASSOCATIONS American Geophysical Union Ecological Society of America National Science Teachers Association Sigma Xi