<%BANNER%>

Ecohydrological Study of Watersheds within the Military Installation in Fort Benning, Georgia


PAGE 1

ECOHYDROLOGICAL STUDY OF WATE RSHEDS WITHIN THE MILITARY INSTALLATION IN FORT BENNING, GEORGIA By SHIRISH BHAT A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLOR IDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2005

PAGE 2

Copyright 2005 by Shirish Bhat

PAGE 3

This work is dedicated to my parents, Tara and Narmada Bhat, and uncle Devendra Bhat.

PAGE 4

iv ACKNOWLEDGMENTS First of all, I would like to thank Dr. Je nnifer Jacobs, for her constant guidance, encouragement, patience and continuous support over the past five years. Her enthusiasm for research and quest for excellence have left an everlasting impression in my mind. To me, she has been more than an advisor, and this research would not have been possible without her. Secondly, I would like to thank Dr. Kirk Hatfield for being on my committee and offering me guidance in the la ter part of my research. I would also like to thank Dr. Wendy Graham, Dr. Ramesh Shrestha, and Dr. Ramesh Reddy for being on my committee and their invalu able suggestions throughout th e research work. I would like to thank Dr. Richard Lowrance and Randall Williams of USDA-ARS, Georgia, whose contributions in this research have b een tremendous. I deeply benefited from all the long hours of fruitful discussions with them on a multitude of topics. I would like to thank Hugh Westbury at Fo rt Benning, Georgia, for his coordination efforts during the field trips, Dwight Dindial and Ch arles Campbell for their efforts in the collection of the field data, and Phil Harmer for his contribution in the laboratory. I wish to extend my gratitude to Lewi s and Kim Bryant, and Donna Rowland, who have been very helpful during my difficult days at UF. I also wish to extend my gratitude to Subarna and Pramila Malakar for their care and support. I wi sh to extend my thanks to Mr. Binod Palikhe for his help during my academic pursuit. I would like to extend my thank to Debra Carol, Carol Hipsley, S onja Lee, Doretha Ray, and all the Civil Engineering staff for their help during all thes e years. I would also like to thank my

PAGE 5

v friends Deepak Singh, Shashi Shrestha, Hemant Belbase, Rishi Bhattarai, Dipendra Piya, and Anand Bastola, and colleagues Nebiyu Tiruneh, Aniruddha Guha, Qing Sun, Mark Newman, Ali Sedighi, Chris Brown, Brent Wh itfield, Gerard Ripo, Haki Klammler and others for their encourag ements and moral support. I would like to thank my pare nts and all the family member s for their constant love and encouragement. They have allowed me to pursue whatever I wanted in life. Without their guidance and affection, it would have been impossible for me to advance my education.

PAGE 6

vi TABLE OF CONTENTS page ACKNOWLEDGMENTS.................................................................................................iv LIST OF TABLES.............................................................................................................ix LIST OF FIGURES.............................................................................................................x ABSTRACT....................................................................................................................... xi CHAPTER 1 GENERAL INTRODUCTION....................................................................................1 Objectives..................................................................................................................... 5 Dissertation Organization.............................................................................................6 2 ECOLOGICAL INDICATORS IN FORESTED WATERSHEDS IN FORT BENNING, GEORGIA: RELATIONS HIP BETWEEN LAND USE AND STREAM WATER QUALITY....................................................................................9 Introduction................................................................................................................... 9 Study Area..................................................................................................................12 Methods of Study........................................................................................................13 Description of Watersheds..................................................................................13 Characterization of Di sturbance Categories........................................................14 Collection and Analysis of Stream Samples.......................................................15 Statistical Analyses..............................................................................................15 Results........................................................................................................................ .16 Watershed Physical Characteristics.....................................................................16 Water Quality Parameters....................................................................................19 Effects of Disturbance Categories.......................................................................21 Relationship between Watershed Physical Characteristics and Water Quality Parameters........................................................................................................22 Discussion...................................................................................................................29 Conclusion and Recommendations.............................................................................33 3 HYDROLOGIC INDICES OF WATERSHED SCALE MILITARY IMPACTS IN FORT BENNING, GA...............................................................................................35 Introduction.................................................................................................................35

PAGE 7

vii Flow Regimes and H ydrologic Indices.......................................................................37 Effects of Flow Magnitude on Stream Ecology..................................................40 Effects of Flow Duration on Stream Ecology.....................................................40 Effects of Flow Frequency on Stream Ecology...................................................41 Effects of Flow Rate of Change on Stream Ecology...........................................42 Data Collection...........................................................................................................42 Watershed Characteristics...................................................................................42 Precipitation and Streamflow Data......................................................................45 Methods......................................................................................................................45 Annual-Based Hydrologic Indices......................................................................45 Storm-Based Hydrologic Indices........................................................................48 Statistical Analyses..............................................................................................48 Results........................................................................................................................ .48 Watershed Disturbance Characteristics...............................................................48 Annual-Based Hydrologic Indices......................................................................50 Storm-Based Hydrologic Indices........................................................................51 Relationship between Military Land Management and Storm-Based Hydrologic Indices...........................................................................................52 Discussion...................................................................................................................54 Conclusion and Recommendations.............................................................................58 4 PREDICTION OF NITROGEN LEACHING FROM FRESHLY FALLEN LEAVES: APPLICATION OF RIPARIAN ECOSYSTEM MANAGEMENT MODEL (REMM)......................................................................................................59 Introduction.................................................................................................................59 Basic Concepts of REMM..........................................................................................62 Hydrology component of REMM.......................................................................63 Nutrient component of REMM...........................................................................64 Application of REMM................................................................................................66 Study Area...........................................................................................................66 Data and Input Parameters...................................................................................68 Model Calibration................................................................................................71 Sensitivity Analysis.............................................................................................73 Results........................................................................................................................ .73 Hydrology............................................................................................................73 Nitrogen...............................................................................................................74 Sensitivity Analysis.............................................................................................77 Discussion...................................................................................................................80 Conclusion and Recommendations.............................................................................83 5 SUMMARY AND CONCLUSION...........................................................................85 APPENDIX A ANNUAL-BASED INDICES DE FINITIONS AND CALCULATION PROCEDURES..........................................................................................................88

PAGE 8

viii B STORM-BASED INDICES DEFI NITIONS AND CALCULATION PROCEDURES..........................................................................................................95 C NITROGEN TRANSFORMATIONS IN REMM.....................................................98 D PARAMETERS AND RATE CO NSTANTS USED IN REMM..............................99 LIST OF REFERENCES.................................................................................................102 BIOGRAPHICAL SKETCH...........................................................................................115

PAGE 9

ix LIST OF TABLES Table page 2-1. Physical characteristics of study wa tersheds in Fort Benning, Georgia. ................17 2-2. Pearson correlation coefficients between watershed characteristics. ......................18 2-3. t-test results for differences in mean values of watershed phys ical characteristics and water quality parameters. ..................................................................................21 2-4. Pearson correlation coefficients be tween watershed characteristics and water quality parameters....................................................................................................23 2-5. Stepwise multiple regression models for water quality parameters..........................24 3-1. Definitions of terms used to de scribe hyetograph and response hydrograph............39 3-2. Physical characteristics of study watersheds.............................................................44 3-3. Summary of the recommended hydrologic i ndices for 'perennial runoff' type of streams......................................................................................................................46 3-4. Summary of hydrologic i ndices used in the Storm-Ba sed Hydrologic Indices. ......49 3-5. Pearson correlation coe fficients between watershed physical characteristics and event based hydrologic indices................................................................................53 3-6. Stepwise multiple regression models for event based hydrologic indices. ............54 4-1. REMM model parameters values by th e riparian zone for the Bonham-South watershed in Fort Benning, Georgia........................................................................67 4-2. Soil and vegetation nutrient pools for Bonham-South watersheds riparian area.....69 4-3. Model simulated annual hydrologic budget for the riparian areas in Bonham-South watershed........................................................................................69 4-4. Sensitivity of modeled streamflow and TKN based on +/10% change in model parameters for Bonham-South watershed................................................................80

PAGE 10

x LIST OF FIGURES Figure page 2-1 Study watersheds......................................................................................................14 2-2 Box plots of water quality parameters......................................................................20 2-3 Relationships between military land and water quality parameter..........................25 2-4 Relationships between road dens ity and water quality parameter...........................26 2-5 Relationships between number of ro ads crossing streams and water quality parameter..................................................................................................................27 2-6 Relationships between disturbance index and water quality parameter...................28 3-1 Terms used to describe hy etograph and response hydrograph.................................39 3-2 Study watersheds......................................................................................................43 4-1 Study area................................................................................................................. 66 4-2. Comparison of REMM simulated daily fl ow with the observed daily flow for the Bonham-South watershed...................................................................................74 4-3. Observed canopy cover and TKN c oncentrations each month for the Bonham-South watershed........................................................................................75 4-4. Monthly simulated TKN concentrati on and simulated leaf mass in the BonhamSouths riparian area.................................................................................................75 4-5. Monthly simulated TKN concentration and simulated nitrogen present in live leaves the Bonham-Souths riparian canopy............................................................76 4-6. Observed TKN concentrations during the events......................................................78 4-7. Surface runoff comparison of the observed and si mulated TKN masses..................79 4-8. Total streamflow comparison of the observed and si mulated TKN masses..............79

PAGE 11

xi Abstract of Dissertation Pres ented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy ECOHYDROLOGICAL STUDY OF WATE RSHEDS WITHIN THE MILITARY INSTALLATION IN FORT BENNING, GEORGIA By Shirish Bhat May 2005 Chair: Kirk Hatfield Cochair: Jennifer Jacobs Major Department: Civil and Coastal Engineering Relationships among watershed physical characteristics and water quality parameters were explored for seven watersheds in Fort Benning, Geor gia, using statistical analyses to identify chemical indicators of ecological changes. Correlations were identified among the indicators and watershe d physical characteristics. Regression results suggested that pH, ch loride, total phosphorus, tota l Kjeldahl nitrogen, total organic carbon, and total suspended solids are useful indicators of watershed physical characteristics that are suscep tible to perturbations. The magnitude, frequency, duration, timi ng, and rate of change of hydrologic conditions regulate ecological processes in aquatic ecosystems. Analysis of 26 nonredundant hydrologic indices showed undistur bed watershed produced higher magnitude and more frequent lowand high-flows as compared to the disturbed watershed. Eighteen storm-based indices, grouped into four flow components, were proposed to statistically characterize hydrologic varia tion among different watersheds. Results

PAGE 12

xii showed that these storm-based indices might be used as surrogates to the indices derived from long-term data. Statistical analys is showed that the watershed physical characteristics such as military training land, road density, and the number of roads crossing streams predicted hydrologic indice s such as storm-based baseflow index, bankfull discharge, response lag, and time of rise well. Riparian vegetation has an important role in altering water quality in the forested watersheds. The leaching of organic or mi neral products on the forest floor provides potential additional effects on water quality. In these areas, nutrients are released into the fresh water systems due to the leaching a nd decomposition of vegetation litter. The release of nutrients from plan t litters prior to decomposition may be an important aspect of characterizing stream water quality. To e xplore nitrogen leaching in a riparian area during litterfall, Riparian Ecosystem Ma nagement Model developed by USDA-ARS has been applied. The simulated total Kjeldahl nitrogen masses in the study watershed were close to the observed values. The model effectively captured the trends of litter mass accumulation in the riparian area and subse quent high concentrations and masses during those periods.

PAGE 13

1 CHAPTER 1 GENERAL INTRODUCTION Ecohydrology is the interdisciplinary research field in which hydrology and ecology come together. It may be defined as the study of the functional relations between hydrology and biota at the watershed scale (Zalewski, 2000). Resource managers and land use planners seek ecohydrological knowle dge enabling them to design effective land use plans and water management strategies. Water, soil, and plant cover are funda mental components that determine the productivity of the land. The successional stag es in evolution of ecosystems depend on climatic and hydrologic conditions and on nutri ent availability. The unique composition of plants and animals determines the ability to retain water and nutrients within the system (Zalewski et al., 1997). The amount of water and its quali ty in the aquatic environment are guided mostly by climate. Water and nutrient cycling in terrestrial systems are tightly linked with each other and with human activity. The growth and activity of human populations have increased the input of nutrients to terrestrial and aquatic ecosystems (Schindler and Bayley, 1993; Vitousek et al., 1997). Various forms of environmental perturbations affect quality and quantity of inland waters by increasing runoff, erosion, sedimentation, and pollution. Modifications of land use affect the wate r quality, residence tim e, surface runoff, soil moisture, evaporation, and ground water. For example, an increase in urbanization and non-forest land increased nitrogen concen tration in the streams (Osborne and Wiley, 1988; Sponseller et al., 2001). Land-use prac tices can potentially affect and modify

PAGE 14

2 channels along the flow paths in a la ndscape and may enhance or disrupt runoff (Nakamura et al., 2000). Forest cutting can increase the frequency and volume of debris slides due to dead and decayed tree roots that contribute to soil strength on marginally stable sites (Sidle et al., 1985). Water quality is highly variable, from place to place and from time to time, even within a particular ecosystem. It is depe ndent on many factors, both natural and as a consequence of human activities. In both cases, the quality of water is further affected by the soils and vegetation over and through which it passes. Nutrients are released into the freshwater systems due to the decomposition of the vegetation litter and modify the water quality. Riparian vegetation ha s an important role in alteri ng water quality as it functions as a source or a sink of va rious organic and mineral components (Cirmo and McDonnell, 1997). Agricultural practices an d urbanization are widely recognized human impacts on water resources. A specific type of human impact is caused at military installations. Military training within a wa tershed can affect the draina ge patterns, vegetation, and soils. The impact is due to troop maneuvers a nd large tracked and wheeled vehicles that traverse thousands of hectares in a single tr aining exercise (Quist et al., 2003). The impacts of such activities range from minor soil compac tion and lodging of standing vegetation to severe compaction and complete loss of vegetation cover in areas where training is concentrated (Wilson, 1988; Milchunas et al., 1999). The resultant impacts are evident both in the stream hydrology and the stream ecosystems. Potential impacts on terrestrial and aquatic ecosys tems include disruption in so il density and water content (Helvey and Kochenderfer, 1990), addition of sediment, nutrients, and contaminants in

PAGE 15

3 aquatic ecosystems (Gjessing et al., 1984) and impairment of natural habitat development,and woody debris dynamics in forested floodplain streams (Piegay and Landon, 1997). Military roads can signi ficantly alter hillslope hydrology by redistributing soil and rock materials on slope s and increasing the ra te of debris slide initiation. Roads also directly change th e hydrology by intercepting shallow groundwater flow paths, diverting the water along the roadway and routing it to stream at road crossings. Road crossings can intercept groundwater drai nage networks and collect groundwater from upslope areas, diverting it into drainage ditches as surface water (Wemple et al., 1996). Road crossings co mmonly found in military training areas may act as barriers to the movement of fish and other aquatic ha bitats (Furniss et al., 1991). Given the nature of military land use, ma nagement or military testing and training, military land managers face the conflicting demands of balancing the primary military mission with legal requirements to protect land and water qualit y (Milchunas et al., 1999). The military impacts can result in significant disruptions to the water and nutrient cycles in the freshwater ecosystems and water resources. A few studies in the past have addressed the effects of military training on terrestrial and aquatic ecosystems. For example, Wilson (1988) found that the tank traffi c resulted in a significant loss of native species, increased abundance of in troduced species, and increased bare soil at a training site located in Manitoba. Milchunas et al. (1999) examined the effects of military vehicles on plant communities and soil char acteristics in Pinon Canyon Maneuver Site, Colorado. Whitecotton et al. (2000) examined the impact of foot traffic from military training on soil bulk density, infiltration rate and aboveground biomass. Recently, Quist

PAGE 16

4 et al. (2003) conducted a study in the Fort Riley Military Re servation, Kansas, to study the effects of military use on terrestrial and aquatic communities. Understanding human impacts in many la ndscapes needs the identification of critical landscape elements and analysis of landscape pattern change (O’Neill et al., 1997). Generalized response to military tr aining includes the reduction in native and perennial grasses, abundance of introduced species, and an in crease in bare soil. Clearly, these responses will alter biogeochemical a nd hydrological processes, which regulate nutrient and water dynamics. However, impact of landscape changes to water quality and stream is not well understood (Wand et al., 2001). In order to maintain and improve water quality, there is an increasing need to understand the relationships among watershed land use and stream ecosystems (Wang et al., 2001). In parallel, perturbation induced effects on water quantity are critical to stream ecology. One of the many approaches to study th e functional interrelations between the hydrology and the stream biota is streamflow characterization and classification. This approach develops hydrologic indices that account for characteristics of streamflow variability that are biologically relevant (Olden and Poff, 2003). Characterization of streams through development of ecologically relevant hydrologic indices is based on long-term streamflow data. However, past studies overlooked the importance of stormflow data for the development of such hydrol ogic indices. Flow characteristics are important where changes in land use are anticipated and where alterations to the flow regimes need to be assessed.

PAGE 17

5 The leaching of organic or mineral products from the forest floor provides potential additional effects on water quality. Fluxes of nitrogen through the riparian zone are intrinsically linked to water movement, bot h over and through the soil, and are also strongly influenced by biological processes occurring in that zone. Nitrogen and organic carbon dynamics in riparian zones are closely interrelated. While ma ny of the factors that can potentially influence nitrogen a nd carbon fluxes through riparian zones are broadly known, there is presently incomplete quantitative information on the relative importance of the flushing of nitrogen from freshly fallen leaves during precipitation events. Objectives The United State’s Department of De fense (DOD) policy has established ecosystem management as its approach to manage the military lands by maintaining and improving the sustainability and biological diversity of terrestrial and freshwater ecosystems while supporting human needs, including the DOD missions. In order to identify critical deficiencies and rese arch opportunities on ecosystem management problems on defense installations, the St rategic Environmental Research and Development Program (SERDP) of DOD in itiated the SERDP Ecosystem Management Program (SEMP) in December 1997. The objectiv es of SEMP were to (1) establish longterm research sites on DOD lands for military-relevant ecosystem research, (2) conduct ecosystem research and monitoring activ ities relevant to DOD requirements and opportunities, and (3) facilitate the integration of results and findings of research into DOD ecosystem management practices. The goal of this research is to illustra te how an ecohydrological approach could be used to advance our ability to predict the effects of anth ropogenic perturbations on water-

PAGE 18

6 vegetation-nutrient interactions in the military installation at Fort Benning, Georgia. The issues addressed in this research encompass many of the main scientific challenges in the military installations’ ecohydrology that in clude the effects of military related perturbations on stream wate r quality and quantity, and the value of studying nutrient dynamics in the riparian corridors of such regions. The specific objectives of this research include (1) identification and examin ation of the statistical relationships among water quality parameters and the watershed physical characteristics in low-nutrient watersheds, (2) identification and development of hydrologic i ndices that characterize the impact of military land management on watershe ds, and (3) investigati on of the effects of nitrogen leaching from freshly fallen leaves on nutrient dynamics in a riparian area. Dissertation Organization Each chapter of this dissertation, except Chapters 1 and 5, is written as a selfcontained individual paper focusing on a topi c that has not been addressed before. Contributions are in the areas of water qual ity and land use within a military installation (Chapter 2), development of storm-based hydr ologic indices (Chapter 3), and watershed scale nutrient leaching from a riparian area (Chapter 4). Chapter 5 summarizes and concludes the research work Chapter 2 is an original contribution in th e effects of military activities related land use on surface water quality. The major results from Chapter 2 are that the concentrations of total organi c carbon, total Kjeldahl nitroge n, total suspended solids, pH, and total phosphorus in the str eam show the greatest suscepti bility to direct effects of military activities. This chapte r also identifies significant statistical relationships among the water quality parameters and the military land uses. These relationships provide the

PAGE 19

7 guidance for maintaining the surface water quality within the Fort Benning military installation. Chapter 3 is an original contribution in the ecohydrol ogy that presents both annualbased and storm-based methods for determini ng hydrologic indices that are of ecological importance. Detailed descriptions of the me thods for determining th ese indices and their significance in aquatic ecosystems are descri bed in this chapter. To statistically characterize the hydrologic variation among different watersheds, 32 annual-based hydrologic indices are analyzed. Eight out of 32 annual-based indices are recommended to use for management practices within the Fort Benning military installation. As a new approach to characterize the streamflow vari ability, 18 storm-based indices are proposed. These indices are grouped into magnitude, fr equency, duration, and rate of change of flow. Storm-based indices are compared w ith annual-based indices. The storm-based methodology provides guidance for measurements of appropriate i ndices within the military installation. Chapter 4 is an original c ontribution to the water quality function of the riparian area. Nitrogen leaching from freshly falle n leaves in a riparian area during the precipitation events is quantified. The observe d nitrogen masses in the stream during the precipitation events are compared with the m odel simulated values. Riparian Ecosystem Management Model (REMM) is used to quantify the nitrogen in the ripa rian area. In this chapter, details of the nitroge n leaching from the freshly falle n leaves are explored. The results showed that the model effectively captured the trends of litter mass accumulation in the riparian area and subsequent high con centrations during those periods. Analysis

PAGE 20

8 showed that the simulated total Kjeldahl nitrogen masses during the precipitation events were close to the observed masses. Chapter 5 summarizes and concludes the re search work. This chapter contains recommendations to improve management of the water resources within the Fort Benning military installation. Future research n eeds are also outlined in this chapter.

PAGE 21

9 CHAPTER 2 ECOLOGICAL INDICATORS IN FORESTED WATERSHEDS IN FORT BENNING, GEORGIA: RELATIONSHIP BETWEE N LAND USE AND STREAM WATER QUALITY Introduction Ecological monitoring is esse ntial to protect eco logical health and integrity. As human activity alters land cover, degradati on of water resources begins in the upland areas of a watershed. The first step to ward effective ecological monitoring and assessment is to realize that the ultima te goal is to measure and evaluate the consequences of human actions on ecological systems. Human activities that alter land use eventually affect biogeochemical pro cesses that influence water quality and alter ecological processes. The National Research Council of the Unite d States recently conducted a critical evaluation of indicators used to monitor ecological changes from either natural or anthropogenic causes. During recent decades, efforts have been increasing to develop reliable and comprehensive environmental i ndicators because of growing environmental concerns (National Research Council, 2000) Indicators rapidly and effectively communicate system status. Ecological indicato rs help to elucidate both the effects of human activities and natural processes. They can also help to asse ss future implications of these factors on ecosystem integrity. Once i ndicators identify areas or elements of the environment that are under st ress, successful management of problems can be measured relative to both interim targets and long-term goals.

PAGE 22

10 Indicators that relate ke y ecological responses to hu man perturbations provide useful tools to better underst and ecological effects and thei r monitoring and management. A suite of indicators ranging from microbiol ogic to landscape metrics is necessary to capture the full spatial, temporal, and ecol ogical complexity of impacts (Dale et al., 2002). Evaluation of representative indices across major physical gradients (e.g., soils, geology, land use, water quality and quantity) can signal early environment change and help diagnose the cause of an environmental problem. Understanding human impacts in many la ndscapes needs the identification of critical landscape elements and analysis of landscape pattern change (O’Neill et al., 1997). Attention has refocused on relations hips among watershed characteristics and stream water quality (Johnson et al., 1997). In order to maintain and improve water quality, there is an increasing need to unders tand the relationships among watershed land use and stream ecosystems (Wang et al., 2001). Land use provides information about ecosy stem function and characterizes the extent and diversity of ecosystem types. National Research Council (2000) has recommended land use as one of the most eff ective indicators for ecological assessment. Hydrologists and aquatic ecologists have l ong known that the pathway by which water reaches to a stream or lake has a major effect on water quality. Early studies on the physical (Harrel and Dorris, 1968) and ch emical (Hynes, 1960) characteristics of watersheds focused on the influence of geomorph ic characteristics such as drainage area, gradient, and stream order on turbidity, disso lved oxygen concentration, and temperature. Many recent studies examine the influence of terrestrial ecosystems on stream or wetland water quality (Richards and Host, 1994; Richards et al., 1997). Many other studies have

PAGE 23

11 found relationships between land use and concen trations of nutrients in streams (e.g., Hunsaker and Levine, 1995; Johnes et al ., 1996; Bolstad and Swank, 1997). Watershed properties constrain in-s tream physicochemical and biotic f eatures. Richards et al. (1996) showed that ecosystems could be influenced by land use at regional or broad geographic scales. Osborne and Wiley (1988) found that the distance of urban land cover from the stream effectively predicts stream n itrogen and phosphorus concentrations. Within a military installation context, land managers are challenged to use the land for military training purposes in a manner that is both ecologically sound and meets military mission requirements (Garten Jr et al., 2003). Lands can suffer a slow degradation if over-utilized by long-term human activities. The heavy vehicles used in mechanized military training cause disturbance of soil structure and can change the physical properties of the soil (Iverson et al., 1981). In rangelands, tracked vehicle traffic affects the hydrological characteristics (T hurow et al., 1993). Trampled vegetation, vehicle tracks through undisturbed area, and er osion caused by the overuse of trails are some examples of the visible degradation to a landscape caused by military training exercises. Few studies have develope d predictive relationships among watershed physical characteristics and surface water chem istry specific to military land use and lownutrient systems. In the coastal plain of the Apalachicol a-Chattahoochee-Flint (ACF) river basin, cropland and silvicultural la nd in upland areas is separa ted from streams by relatively undisturbed riparian flood plai n and wetland habitats (Frick et al., 1998). This is in contrast to many intensively farmed areas of the United States where wetlands have been drained, channelized or filled, and little or no riparian buffers remain between cropland

PAGE 24

12 and streams. Frick et al. (1998) reported that the lower nutri ent concentrations in streams within the ACF river basin could partially be attributed to wetland buffer areas, and minimal use of pesticides as compared to ot her areas of the United States. Other studies in the southeastern coastal plain watershe ds (e.g., Lowrance 1984; Lowrance et al., 1992; Perry et al., 1999; Fisher et al., 2000) focused primarily on th e agricultural impacts, and urbanization on stream water quality. The contribution of areas affected by military training to nutrient discharges, specifically in Fort Benning watersheds, is yet to be quantified. This paper identifies and examines the st atistical relationships among water quality parameters and the watershed physical charact eristics in seven low-nutrient watersheds located in the Fort Benning military installati on, Georgia. It is hypot hesized that surface water quality parameters can be used as indicat ors of ecological changes in watersheds. Study Area The Fort Benning Army Installation occupies approximately 73,503 ha in Chattahoochee, Muscogee, and Marion C ounties of Georgia and Russell County of Alabama (Figure 2-1). The climate at Fort Be nning is humid and mild. Rainfall in this region occurs regularly throughout the year. July and August are the warmest months with average daily maximum and mi nimum temperatures of 37 and 15oC. An average daily maximum and minimum temperature of 15.5 and -1oC are reported in the coldest months, January and February. Annual pr ecipitation averages 1050 mm with October being the driest month (Dale et al., 2002). Mo st of the precipitation occurs in the spring and summer as a result of thunderstorms. H eavy rains are typical during the summer but can occur in any month. Snow accounts for less than 1% of the annual precipitation.

PAGE 25

13 Fort Benning is located within the sout hern Appalachian Piedmont and Coastal Plains. The northern boundary of the instal lation lies along a transi tion zone between the Piedmont and Upper Coastal Plain. The so ils in the area are dominated by loamy sand with some sandy loam. Following establis hment of the installation in 1918, with subsequent additions in 1941, we see that heavy training imp acts only selected, mostly upland, portions of the installa tion. Many areas are maintain ed as safety buffers, and have little military use. Timber mana gement includes harves ting and thinning. The loblolly and longleaf pine fore sts are subjected to regular lo w-level fires for management purposes (Dale et al., 2002). Methods of Study Description of Watersheds The study watersheds, Bonham-1 and Bonham2, Bonham, Little Pine Knot, Sally, Oswichee, and Randall (named for the creek wh ich drains the watershed), within Fort Benning represent a range of region’s so ils, topography, land use, and vegetation communities (Figure 2-1). These watersheds have a heterogeneous land cover predominantly consisting of either forested or open areas. Forest ed areas are broadly characterized as mixed pine and hardwoods or pine that are mostly 30-50 years old with the soils in A-horizon range approximately 1-10 cm in depth (Garten Jr et al., 2003). Open areas are either military, brush, or managed wildlife openings. Other cover includes upland and bottomland hardwood forest s. The military openings are clear-cut parcels of land dominated by grass and bare soil that are used as military training grounds. The brush openings consist of tall grass and immature hawthorn. The wildlife openings are natural openings in the forests that are vegetated primarily by grass. Land

PAGE 26

14 impacts due to heavy military activities (e .g., infantry, artillery, wheeled, and tracked vehicle training) occur only in sele cted portions in these watersheds. Figure 2-1. Study watersheds, Bonham-1, Bonham-2, Bonham, Little Pine Knot, Randall, Sally, and Oswichee, in the Fort Benning military installation. Also shown are stream network and sampling locations. Characterization of Disturbance Categories A disturbance index (DIN) was defined to characterize the watersheds at Fort Benning into two disturbance categories: lowimpact and high-impact. A DIN is the sum of the area of bare ground on slopes greater than 3 degrees and on roads, as a proportion of the total watershed area (Mal oney et al., in press). The road areas were estimated by multiplying their length by the measured average width of 20 m. Percentage of bare ground was determined by using TM imagery and slope was derived from digital

PAGE 27

15 elevation maps. The TM imagery and digi tal elevation maps were obtained from Strategic Environmental Research and Development Plan (SERDP)’s Ecosystem Management Project (SEMP) database. Waters heds having a disturba nce index from 0 to 11% are designated as low-impacted waters heds. High-impacted watersheds have disturbance indices greater than or equal to 11%. Collection and Analysis of Stream Samples Surface water quality data were collected at seven streams biweekly from October 2001 to November 2002, and monthly therea fter to September 2003. Water samples were collected in high-density polyethylene bo ttles. Bottles were soaked in de-ionized water and rinsed with sample water prior to collection. The filtration was conducted at the sampling sites using 0.45 m pore size polyethersulfone membranes. Filtered sample was used to determine chloride (Cl) concentr ation, whereas raw sample was used for total suspended solids (TSS) determination. Unfiltered samples for analyzing total Kjeldahl nitrogen (TKN), total phosphorus (TP), and total organic carbon (TOC) were acidified using double distilled sulfuric acid. The stream water pH conductivity, and temperature were measured at the time of sampling. A ll samples were kept cool in an icebox, transported to the Soil and Water Science Depa rtment laboratory, University of Florida, and refrigerated until analyze d. All samples were analyzed using standard methods (American Public Health Association, 1992). Statistical Analyses The Pearson’s correlation coefficients were calculated to examine the strength and significance of the rela tionships between a watershed physic al characteristic and a water quality parameter. Two-sample t-tests were pe rformed at 5% level of significance to test whether mean values of watershed physical characteristics and wate r quality parameters

PAGE 28

16 differ between lowand high-disturbance watersheds. Characteristics showing significant correlations with a water quality parameter were considered for stepwise multiple linear regression models. Only variables having less than or equal to 0.05 significance level were retained in the regression models. Results Watershed Physical Characteristics The watersheds’ physical characteristics are summarized in Table 2-1. Most of the watersheds are highly vegetated (70% or mo re) except Oswichee (38%) with the majority characterized by pine and mixed pine and hard woods. Deciduous forest typically covers only a small percentage of these watersheds However, Bonham-1 consists of 27% of deciduous forest. The study watersheds range from less than 1 to 84 km2. The topographic characteristics of study watersheds are typical of forested watersheds of southeastern coastal plain (Lowrance, 1992; Pe rry et al., 1999). Average elevations vary from 104 to 148 m above mean sea level. Ma ximum slopes vary from 4 to 6 degrees. Sandy soils are common in most of the study watersheds. However, loamy soils cover most of the Sally and Oswichee watersheds. Bottomlands comprise 6 to 20% of the watershed. The military training extent (0 to 6%) is relatively small. Total bare lands in these watersheds comprise 9 to 21% of the wa tershed area, of which 1 to 8% of the total area is unpaved roads and trails. This extent and variability of military training and bare land are typical of the entir e Fort Benning installation. Some watershed characteristic s are strongly correlated at significance level of 0.05 or lower (Table 2-2). Significant positive corr elations exist for pine with the bottomland wetlands, deciduous vegetation with stream density, number of roads crossing streams with road length and percent of loam, and disturbance index with percent bare land.

PAGE 29

17 Negative correlations were found for mixed vege tation with percentage of loamy soil and number of roads crossing streams (NRC), military areas with normalized difference vegetative index (NDVI) and stream density, and DIN, and road density with NRC. Table 2-1. Physical character istics of study watersheds in Fort Benning, Georgia. Acronyms BON-1, BON-2, BON, LPK, OSW, Ran, and SAL represent Bonham-1, Bonham-2, Bonham, Little Pine Knot, Randal, and Sally, respectively. Watersheds Physical Characteristics B ON-1BON-2BON LPK OSW RAN SAL Topography Area, km2 0.762.2112.73 18.0183.39 74.3825.31 Average Elevation, m 121.8 133.5 125.5 146.3 104.2 136.8 136.8 Average Slope, degree 5.464.895.04 5.324.48 4.575.42 Vegetation Pine, % 28 30 40 41 26 58 48 Deciduous, % 27 6 8 2 0 3 12 Mixed, % 39 50 22 34 5 9 15 Wetland, % 6 8 9 17 7 20 10 Military Land, % 0 6 5 2 5 3 2 NDVI 0.360.300.32 0.340.35 0.340.36 Soil Sandy loam, % 78 69 69 72 24 68 49 Loamy sand, % 9 9 31 28 73 32 51 Road Road Length, km 3.6 11.4 51.6 56.6 196.6 415.1 97.6 Road Density, km/km2 4.8 5.1 4.1 3.1 2.4 3.1 3.8 Stream Stream Length, km 2.6 3.9 29.1 43.3 170.6 323.5 65.2 Stream Density, km/km2 3.4 1.7 2.3 2.4 2.1 2.4 2.6 Stream Order 2 2 4 4 5 6 4 Other No. of Roads Crossing Streams 1 2 13 11 55 43 21 Bare Land, % 1 11 11 4 4 4 7 Disturbance Index, % 11 21 19 11 9 10 15

PAGE 30

18Table 2-2. Pearson correlation co efficients between watershed ch aracteristics. Characteristic s are acronymed as follows: Pine forest (PIN), Deciduous forest (DCD), Mixed forest (MXD), We tland (WET), Military land (MIL), Sandy Soil (SND), Loamy Soil (LOM), Road Length (RDL), Road Density (RDN), Stream Density (STD), Normalized Difference Vegetative Index (NDVI), No. of Roads Crossing Stream s (NRC), % Bare Land (PBL), and Di sturbance Index (DIN). *, **, and *** indicates significance at or below 0.05, 0.01, and 0.001 probability levels, respectively. PIN DCD MXD WET MIL SND LOM RDL RDN STD NDVI NRC PBL DCD -0.25 MXD -0.43 0.40 WET 0.96*** -0.33 -0.42 MIL -0.20 -0.65 0.00 -0.18 SND 0.21 0.44 0.67 0.17 -0.30 LOM 0.06 -0.50 -0.85* 0.14 0.15 -0.93*** RDL 0.63 -0.47 -0.74 0.50 0.05 -0.28 0.41 RDN -0.26 0.65 0.81* -0.34 -0.01 0.63 -0.81* -0.64 STD 0.01 0.84* 0.04 -0.04 -0.92***0.34 -0.24 -0.15 0.17 NDVI 0.10 0.41 -0.48 0.07 -0.80* -0.30 0.41 0.22 -0.39 0.74 NRC 0.23 -0.58 -0.90**0.19 0.21 -0.76* 0.83* 0.81*-0.86**-0.27 0.33 PBL 0.00 -0.28 0.25 0.09 0.72 0.04 -0.11 -0.34 0.40 -0.67 -0.79* -0.30 DIN -0.10 0.05 0.52 -0.07 0.55 0.30 -0.42 -0.52 0.71 -0.44 -0.77* -0.58*0.93***

PAGE 31

19 Water Quality Parameters Water quality parameters in the study wa tersheds varied over the sampling period and among watersheds. Variab ility of water quality parameters among watersheds is observed (Figure 2-2). Mean pH in the study watersheds ranged from 4.2 to 7.0. Mean conductivity ranged from 16.4 to 44.5 S/cm. Mean temperatures varied from 17.5 to 20.8oC. Low concentrations of TP and TKN were observed in all the watersheds under study as compared to forested watersheds in the southeastern coas tal plain watersheds (Lowrance et al., 1984), and across the Un ited States (Meader and Goldstein, 2003; Fisher et al., 2000). Concentr ations of TKN, TP, and Cl we re often below the detection limit. Mean concentrations of TP vari ed widely, ranging from 0.003 to 0.020 mg/L and TKN varied from 0.20 to 0.35 mg/L; TOC fr om 1.35 to 3.33 mg/L; Cl from 1.46 to 4.13 mg/L; and TSS from 4.15 to 10.30 mg/L. As depicted in Figure 2-2, each stream exhibited distinct water quality signatures with the exception of temperature and TSS. Seasonal variations in the wa ter quality parameters are re sponsible for much of this observed variability among watersheds. Stream pH fluctuated more during J une and July and was elevated during December through February. Conductivity values showed slight fluctuations from May to July. On multiple occasions, high conductiv ity values were observed in the Randall stream. Chloride, TKN, and TP showed distin ct seasonal patterns. The concentrations of these parameters were low from June to September, and high from March to May and from October to December. In contrast, TOC peaked from August to October and again from March to July. Higher concentrati ons of TSS were observed from July to September in all the streams.

PAGE 32

20 Figure 2-2. Box plots of water quality parameters Each plot consists of outliers, most extreme data, 75th, 50th, and 25th percentile values. Watershed IDs represent1: Bonham-1, 2: Randall, 3: Os wichee, 4: Little Pine Knot, 5: Sally, 6: Bonham, and 7: Bonham-2.

PAGE 33

21 Effects of Disturbance Categories Table 2-3 presents the t-test results of the comparison of physical characteristics and water quality parameters based on the watershed disturbance level. The only physical characteristic havi ng a significant difference ( = 0.04) between these two groups was DIN. While the low-impact wate rsheds tended to have higher chemical Table 2-3. t-test results for differences in mean values of watershed physical characteristics and water quality parame ters. indicates significance at or below 0.05 probability level. NS indi cates non-significant difference at the 0.05 probability level. Low-Impacted High-Impacted Mean SD Mean SD Watershed Characteristics Pine, % 38.3 14.8 39.3 9.0 NS Deciduous, % 8.0 12.7 8.7 3.1 NS Mixed, % 21.8 17.2 29 18.5 NS Wetland, % 14.5 9.5 16 7.2 NS Military Land, % 2.4 2.0 4.3 2.3 NS Sand, % 60.5 24.7 62.3 11.5 NS Loam, % 35.5 26.9 30.3 21.0 NS NDVI 0.35 0.01 0.32 0.03 NS Road Length, km 168 184 53.5 43.2 NS Road Density, km/km2 3.3 1.0 4.3 0.7 NS Stream Density, km/km2 2.6 0.6 2.2 0.4 NS No. of Roads Crossing Streams 27.5 25.6 12 9.5 NS Disturbance Index, % 10.2 0.9 18.3 3.1 Water Quality Parameters pH 5.6 1.3 4.5 0.3 NS Temperature, oC 19.3 0.9 18 0.3 NS Conductivity, S/cm 25.9 13.3 20.4 2.5 NS TKN, mg/L 0.3 0.03 0.2 0.05 NS TP, mg/L 0.011 0.006 0.007 0.003 NS TOC, mg/L 2.9 0.4 2.1 0.7 NS Cl, mg/L 2.4 0.5 1.8 0.3 NS TSS, mg/L 9.1 2.2 4.8 0.5

PAGE 34

22 concentrations than the high-impact waters heds, only TSS showed significant difference ( = 0.03). Even though the resu lts showed no significant sta tistical differences at a confidence level of 95%, the t-test results of all water quality parameters, except conductivity and TP, showed significant differe nces at 80% confiden ce interval between highand low-impacted watersheds. The relatively small sample size and natural variability among watersheds may have lim ited the ability to discern significance differences. Relationship between Watershed Physic al Characteristics and Water Quality Parameters Correlation and regression analyses were performed to identify relationships among the watershed physical characteristics and the water quality paramete rs. Table 2-4 shows that each water quality parameter had a significant relationship with one or more watershed physical characteristics (Table 24). The correlation results show that decreasing mixed vegetation increased pH and TP. Sandy and loamy soils had opposite effects on TP. An increase in sandy soil decr eased TP, whereas an increase in loamy soil increased TP. Increasing military land decrea sed TOC. Temperature, pH, conductivity, and Cl increased as the road length increase d. The number of roads crossing streams had positive correlations on pH and TP. Percent bare land was negatively correlated with TOC and TSS. Disturbance index was nega tively correlated with TKN and TOC. Graphical relationships provi de insight into nonlinear relationships that exist between indicators and response variables. Fi gures 2-3 to 2-6 show some of the most striking relationships between watershed characte ristics directly and/or indirectly affected by management of military lands and their e ffect on water chemistry. As the % military land increases, the TKN, TOC, and TSS decr ease in a linear fashion. However,

PAGE 35

23 Table 2-4. Pearson correlation coefficients between watershed characteristics and water quality parameters. Characteristics are acronymed as follows: Pine forest (PIN), Deciduous forest (DCD), Mixed forest (MXD), Wetland (WET), Military land (MIL), Sandy Soil (SND), Loamy Soil (LOM), Road Length (RDL), Road Density (RDN), Stream Density (STD), Normalized Difference Vegetative Index (NDVI), No. of Ro ads Crossing Streams (NRC), % Bare Land (PBL), and Disturbance Inde x (DIN). *, **, and *** indicates significance at or below 0.05, 0.01, and 0.001 probability levels, respectively. pH TemperatureConductivityTKNTP TOC Cl TSS PIN 0.36 0.78 0.54 0 -0.09 0.15 0.56 -0.14 DCD -0.47 -0.37 -0.40 -0.12 -0.33 0.39 0.13 0.42 MXD -0.79* -0.55 -0.38 -0.39 -0.83* -0.24 -0.49 -0.11 WET 0.24 0.62 0.44 0.11 -0.11 0.18 0.37 -0.29 MIL 0.09 0 -0.10 -0.50 0.10 -0.87** -0.47 -0.53 NDVI 0.32 0.11 0.10 0.60 0.46 0.80* 0.55 0.54 PBL -0.45 -0.25 -0.51 -0.72 -0.42 -0.81* -0.68 -0.87** DIN -0.65 -0.36 -0.60 -0.84*-0.63 -0.76* -0.65 -0.72 SND -0.50 -0.02 0.14 -0.19 -0.78* 0.15 0.06 0.15 LOM 0.57 0.16 0.03 0.39 0.81* 0.07 0.07 -0.14 RDL 0.94*** 0.95***0.82* 0.19 0.57 0.09 0.79*0.30 RDN -0.75* -0.42 -0.54 -0.72 -0.78* -0.35 -0.32 -0.15 NRC 0.93*** 0.59 0.51 0.36 0.90**0.06 0.47 0.19 STD -0.14 -0.11 -0.01 0.39 -0.04 0.82* 0.44 0.65 disturbance index may operate as a thres hold indicator of pH and TSS where pH decreases in response to a relatively low level DIN while the TSS threshold for DIN impact is somewhat higher. No signifi cant relationships are found between the water quality parameters and extent of pine and deciduous forest. Similarly, wetland showed no effect on these parameters. Stepwise multiple regressions identified relationships between the water quality parameters and watershed physical characteristic s that are susceptible to the disturbances (Table 2-5). A statistical ly significant regression mode l was found for every water quality parameter. Prediction of pH variab ility among watersheds is particularly well captured by pine forest and road length. The regression relationships indicate that all of

PAGE 36

24 the water quality parameters depend on at least one aspect of military management. Several water quality parameters, Cl, TP, TOC, TSS, and TKN, depend only on management aspects of the military installati on. For example, Cl depends on change in military land and road length. Total phosphorus is strongly related to the number of roads crossing streams. However, the influen ce of vegetation and soils characteristics is clearly important in pH, c onductivity, and temperature. For example, conductivity appeared to be well captured by area cove red by sandy soil and the number of roads Table 2-5. Stepwise multiple regression models for water quality parameters. pH is unitless, temperature is measured in degrees centigrade, conductivity is measured in mS/cm, TP, TKN, TOC, Cl, and TSS are measured in mg/L. Pine forest (PIN), Military land (MIL), Sandy Soil (SND), Loamy Soil (LOM), Road Length (RDL), Road Density (RDN), No. of Roads Crossing Streams (NRC), Percent Bare Land (PBL ), and Disturbance Index (DIN) are the independent variables retained in the regression analyses. *, **, and *** indicates significance at or belo w 0.05, 0.01, and 0.001 probability levels, respectively. Water Quality Independent Variables Retained and Parameters Regression Equations R2 Pine, Road Length pH 5.50 – 0.0382 PIN + 0.00924 RDL 0.98*** Military, Road Length Cl 2.19 + 0.145 MIL + 0.00344 RDL 0.90** No. of Roads Crossing Streams TP 0.00451 + 0.000236 NRS 0.81** Sandy Soil, No. of Roads Crossing Streams Conductivity 24.2 + 0.552 SND + 0.666 NRS 0.80* Loamy Soil, Road Density Temperature 26.1 – 0.051 LOM – 1.49 RDN 0.77* Military Land TOC 3.45 – 0.274 MIL 0.76** Percent of Bare Land TSS 11.2 – 0.648 PBL 0.76** Disturbance Index TKN 0.445 0.0109 DIN 0.70*

PAGE 37

25 Figure 2-3. Relationships between military la nd and water quality parameter. Vertical bars represent standard errors.

PAGE 38

26 Figure 2-4. Relationships between road densit y and water quality parameter. Vertical bars represent standard errors.

PAGE 39

27 Figure 2-5. Relationships betw een number of roads crossing streams and water quality parameter. Vertical bars represent standard errors.

PAGE 40

28 Figure 2-6. Relationships between disturba nce index and water quality parameter. Vertical bars represent standard errors.

PAGE 41

29 crossing streams, whereas temperature was cap tured by loamy sand soil and road density. Overall, the regression models s how that it is possible to qu antify the effects of watershed physical characteristics on th e water quality parameters. Discussion Variations in stream water chemistry among the study watersheds reflect differences in biogeochemical reactions occurrin g in the watersheds. The results of this study indicate that even in low impact waters heds, physical characteristics may be used to explain variations in stream water chemistry and, by infere nce, the relative watershed disturbance levels. The observed variability in many of the chemical parameters studied in these watersheds can be attributed to physic al characteristics of the watersheds or land management patterns within the watersheds as evidenced by the road network, forestry practices, and military training. Diverse human activities interact to aff ect conditions in watersheds and water bodies. Sites of interest can be grouped a nd placed on a gradient according to activities and their effects. The results of this study suggest that th e vegetation type, road length, number of roads crossing streams, and dist urbance index are impor tant predictors of water quality variability. Vegetation cover was related to stream pH and TP (mixed forest) and conductivity (pine forest). Howe ver, deciduous forest cover was not related to any of the water quality parameters sugge sting a limited effect in organic matter and nutrient production and vari ability as observed among Fort Benning watersheds. However some other subtle landscape changes resulted in relatively larger impacts on water quality parameters. Low-impact waters heds tend to produce higher concentrations of nutrients in the streams. This can be attrib uted to the availability of more soil organic

PAGE 42

30 matter and the rapid biogeochemical processes occurring in the low-impact watersheds as compared to the high-impact watersheds. It is extremely difficult to capture all asp ects of human influence in a single graph or statistical test. However, sometimes m eaningful chemical patterns can be lost by excessive dependence on the outcome of me nu-driven statistical tests (Karr and Chu, 1999). Figures 2-3 to 2-6 depict several different aspects of stream’s chemical conditions against several measures of human influences, such as military land use, disturbance index, number of roads crossing streams, and ro ad density. The distribution of circles in most of these figures illustra te that a chemical metric i ndicates little about a condition simply because it does not correlate strongly with a single surrogate of that condition. However, where the relationship between human influence and stream’s chemical response is strong, statistics and graph agree. The correlation tests identify linear relati onships between a chemical response and watershed characteristics. Weak statistica l correlations observed in these analyses may have missed important chemical patterns. For example, nonlinear patterns were observed for forest types, bare land (not shown here ), and disturbance inde x (Figure 2-6). The plots in Figure 2-6 show a step-function for TSS and pH. The scatter of this dataset shows little or no statistical si gnificance, but can be interp reted chemically. For TSS, those watersheds having a dist urbance index of 11% or lower had a higher level than those with a greater disturbance index. When a number of variable s interact to influence wa ter quality conditions, it may be difficult to explain observed variability in a single plot agai nst one dimension of human influence (Figure 2-4). Chemical responses were plotted against the road

PAGE 43

31 densities for various watersheds. The P earson correlation coefficient for TP was significant capturing human influence on this ch emical parameter. The response of TP is visibly distinguished from others. A similar discussion is true for military land with TOC (Figure 2-3); number of road s crossing streams with pH and TP (Figure 2-5); and disturbance index with TKN (Figure 2-6). The relationships between water quality pa rameters and physical characteristics indicate that disturbances in low nutrient fo rested environments decrease some chemical signatures. Watersheds with more roads, e.g., Randall and Oswichee, have relatively high pH, conductivity, and Cl compared to the wa tersheds with fewer roads. Watersheds with a small portion of military land, e.g., B onham-1, Sally, and Little Pine Knot, have relatively high TOC concentrations. In contra st, watersheds characterized by higher road densities, e.g., Bonham and Bonham-2, had low TP concentrations. Higher disturbance index, similar to the road density, showed lower TKN and TOC concentrations in the streams. Mixed vegetation, road length, percen t of bare land, DIN, and number of roads crossing streams were able to capture most of the variability in water quality parameters. In a watershed scale study conducted in Ontario, Canada, Sliva and Williams (2001) found a negative correlation of forested land cover with TSS and chloride. In contrast, Johnson et al. (1997) showed a posi tive relationship of forest with TSS in a study of landscape influence on water chemistr y in the Saginaw Bay watershed of central Michigan. Johnson et al.’s (1997) results in dicated that row crop agriculture had the highest effect on total nitrogen, nitrate, and total dissolved solids. They also observed that urban and forest areas were positively correlated good predictors of TSS, whereas row crop agriculture was positively correlated with total nitrogen. Basnyat et al. (1999)

PAGE 44

32 reported a positive association of TSS with agricultural practices in the Fish River watershed, Alabama. The Fort Benning inst allation is characterized by relatively low variability in forest cover and suggests, in co ntrast to other studies that neither TSS nor Cl may be related to forest cover under ex isting land management pr actices. Instead, Fort Benning’s road extent and percent bare land are better predictors of TSS and Cl. Most studies identified urban land use as a dominant factor caus ing elevated total nitrogen and nitrate concentrations in th e streams (Hill, 1981; Osborne and Wiley, 1988). Sponseller et al. (2001) found positive correlation of tota l inorganic nitrogen with percentage of non-forested land in southw estern Virginia watersheds. A negative correlation of TKN to DIN, in this study, is consistent with studie s (e.g., Sponseller et al., 2001; Hunsaker and Levine, 1995; Johnson et al ., 1997) that have shown stream nitrogen concentrations to be good pr edictors of non-forest area at the watershed scale. In a study of 101 watersheds in New Zealand, Close and Davis-Colley (1990) found that between 60 and 80% of the varian ce in conductivity, total nitrogen, and nitrate was accounted for by landscape factors including geology and land use. However, in that same study, landscape factors accounted for onl y 50% of the varian ce in ammonia and phosphorus species. Their result s parallel those found at Fort Benning in that no strong relationships were observed between land use and TP. The strong ne gative correlation of loamy sand soil with TP in Fort Benning is consistent with Hill’s (1981) study, conducted in a sandy loam region similar to portions of Fort Benning waters heds that reported negative correlations between phosphorous c oncentrations and abandoned farmland and forest.

PAGE 45

33 Most variations in stream water chemistry are driven by climatic and biotic factors and are therefore largely governed by the processes that are taking place in the terrestrial part of the watershed such as natural or human induced vegetation cover changes (Semkin et al., 1994). Our results show in teractions among landscape factors and water quality indicators. Results also indicate that it is possible to observe the response of these water quality parameters to physical attributes of watersheds. The importance of water quality parameters in the pres ent study appeared to be attr ibutable to the perturbations related to military training and associated parameters within a watershed as these parameters clearly captured the changes in physic al parameters that are more sensitive to such kind of influences. Conclusion and Recommendations Sometimes a single variable can capture and integrate multiple sources of influence. More often, a small number of ecological attributes pr ovide reliable signals about ecological condition. Water chemis try prediction using watershed physical characteristics in this study showed mixed re sults compared to the other investigations. However, most of the watershed physical char acteristics used in our analysis did explain the variability in water quality parameters. This study documented strong relationships between certain watershed physical characteri stics that are more susceptible to human induced perturbations, specifically military related disturbances, and water quality parameters in military installation at Fort Be nning. Watersheds with more roads crossing streams tended to produce more TP. Total Kjeldahl nitrogen and TOC variations were well captured by DIN and extent of military land, respectively. Total suspended solids’ variability, on the other hand, was captured by the percent of bare land within a watershed. Road length captured most of the variability in pH and Cl. Conductivity and

PAGE 46

34 temperature values were depende nt on soil types and road char acteristics. The variations in stream water chemistry are largely attribut able to disturbance le vels and the types of biogeochemical reactions occurring in the wa tersheds. Regression results suggest that TOC, TKN, and TSS were usef ul indicators of watershed phy sical characteristics as they are more susceptible to direct effects of military activities. Although pH, conductivity, and TP showed good correlations with the road length, these parameters indicated strong but indirect influence of military training activities on watersheds. Foreseeing a single indicator of water quality that would be sensitive to all kinds of perturbations in the watershed is extremely difficult. Ability to detect perturbations can be related to spatial and temporal scales. It is necessary to re cognize the effects of natural disturbances on ecosystem structur e and functioning. It is suggested that priorities for determining ecological indicat ors specific to water quality should include (1) development of framework to determine proper reference states within watershed against which to detect loss of ecosystem health, (2) broadening our knowledge of ecosystem sensitivity to perturbations of va rying intensity, spatio -temporal distribution, and type, and (3) development of suites of indicators necessa ry to detect the broadest spectrum of perturbations in watersheds.

PAGE 47

35 CHAPTER 3 HYDROLOGIC INDICES OF WATERSHED SCALE MILITARY IMPACTS IN FORT BENNING, GA Introduction A goal of stream flow char acterization and classificati on is to develop hydrologic indices that account for charac teristics of streamflow vari ability that are biologically relevant (Olden and Poff, 2003). Broadly, indi ces are attributes th at respond in a known way to a disturbance i.e., they relate key eco logical responses to human activities. Index identification is based on the goals and objectiv es set for a particular ecosystem or region. A good index should be sensitive to stressors, biologically a nd socially relevant, broadly applicable to many stressors and sites, diagnos tic of the particular stressor causing the problem, measurable, interpretable, and not redundant with other measured indices (Cairns et al., 1993). Ecologically relevant hydrologic indices deve loped in the past not only characterize particular regions, but also quantify flow ch aracteristics that are sensitive to various forms of human perturbations. For exampl e, early studies on hydrological indicators focused on variation of mean daily flow to study the pattern of fish in Illinois and Missouri (Horwitz, 1978). In Great Britain, Moss et al. (1987) used average flow conditions to predict macro-invertibrate fa una of unpolluted streams and Townsend et al. (1987) examined persistence of community structure for benthi c invertebrates. In arid regions of southwest United St ates, Minckley and Meffe (1987) studied effects of shortterm flood frequency in stream fish commun ities. Poff and Ward (1989) used long-term

PAGE 48

36 discharge records (17-81 years) of 78 streams from across the continental United States to develop a general quantitativ e characterization of streamflow variability. Similarly, Jowett and Duncan (1990) studied skewness in flows and peak discharges in relation to in-stream habitat and biota in New Zealand. More recent investigations have begun to focus on examining suites of hydrologic indices that are ecologically relevant to quantify hydrologic regimes. These studies report numerous such indices. For example, Poff and Allan (1995) studied stream fish assemblage for 34 sites in Wisconsin and Minnesota in conjunction with long-term stream flow variability and predictability as well as frequency and predictability of high and low flow extremes. In the process of deriving ecologically relevant hydrologic indices, Clausen and Biggs (1997) identified th irty-four hydrological variables from daily flow records at eighty-three New Zealand site s. The authors related these variables to benthic biota including periphy ton and invertibrate species richness and diversity. Wood et al. (2000) reported the importance of hydrological conditions in explaining the ecological role when the authors studied the changes in macro-inver tibrate community in response to flow variations in the Little St our River in the United Kingdom. Pettit et al. (2001) described a method for assessing seasona lity and variability of natural flows and their influence on riparian vegetation in two contrasting river systems in western Australia. To isolate core flow variables for ecological studies, it is important to know not just the ecological relevanc e of the variables, but also the interrelationships among the variables in order to avoi d redundancy in the analyses (e.g., Clausen and Biggs, 2000; Olden and Poff, 2003). Hydrologic indices have been criticized for being overly

PAGE 49

37 simplified and lacking adequate biological relevance. Stre am ecologists are now facing difficulty in choosing appropriate and relevant ones from the available suit of indices. For example, the Indicators of Hydrologic Al teration (Richter et al., 1996) approach is commonly used for characterizi ng human modification of flow regimes, yet it contains 64 statistics (32 measures of central tenden cy and 32 measures of dispersion), many of which are inter-correlated (Olden and Poff, 2003). To date, characterization of streams or regions through determination and development of ecologically relevant hydrol ogic indices are based on long-term stream flow data. However, given the multitude of methods to characterize stream flow, past studies overlooked the value of storm flow data for the development of such ecologically relevant hydrologic indices. Flow characteri stics are especially important where changes in land use are anticipated and where alterations to the flow regime need to be assessed. The primary objective of the study is to identify hydrologi c indices that characterize the impact of military land ma nagement on watersheds in Fort Benning, Georgia. Towards this end, this study inve stigates both storm and annual hydrographs. it is hypothesized that, in addition to annual-ba sed indices, storm-based hydrologic indices are indicative of alteration in stream ecology. Here, a suite of event based hydrologic indices is proposed. Storm-based and the annua l-based indices are calculated and used to compare and contrast impacted watersheds with a reference wate rshed. Additionally, specific military land management practices ar e used to predict storm-based indices. Flow Regimes and Hydrologic Indices To assess hydrologic altera tions within an ecosystem Richter et al. (1996) developed a method to compute representa tive, multi-parameter suite of hydrologic characteristics that are of ecological rele vance, commonly known as Indicators of

PAGE 50

38 Hydrologic Alteration (IHA). Olden and Poff (2003) comprehensively reviewed currently available hydrologic indices for characterizing streamflow regimes and recommended non-redundant indices for various stream types that may differ in major aspects of ecological organization. Poff ( 1996) provides a comprehensive catalog of the stream types for small to mid-size relatively undisturbed streams, classified according to variation in ecologically rele vant hydrological characterist ics, in continental United States. The assessment of IHA as well as other studies (e .g., Poff and Allan, 1995; Clausen and Biggs, 1997; Wood et al., 2000; Pet tit et al., 2001; Olden and Poff, 2003) to identify hydrologic indices is ba sed on long-term flow data. An alternative approach to identify ecol ogically relevant hydrol ogic indices is to conduct an assessment based on the storm hydrogr aph. This approach is useful when long-term data for a particular stream or re gion are not available, when significant data gaps exist, or coincident r ecords are not available. Stor m hydrographs are traditionally described by characteristics including peak flow, total volume of direct runoff, and duration as shown in Figure 3-1. The time characteristics of the hydrograph and its relationship to the precipitation event are pr esented in Table 3-1. Towards the goal of characterizing the spatial vari ations of hydrologic conditions using storm-based indices that are ecologically relevant as well as sensitive to human infl uences, the ecological function of hydrologic characteri stics that are relevant to storm-based hydrologic indices are considered. Examination of the storm hydrographs reveal s numerous potential indices. A set of 18 storm-based ecologically relevant hydrologi c indices that characterize variation in water condition in individual watershed is proposed. Included proposed indices are

PAGE 51

39 Figure 3-1. Terms used to describe hyetogr aph and response hydrograph. Refer to Table 3-1 for the definitions of the terms. Table 3-1. Definitions of terms used to describe hyetograph and response hydrograph. Time instants Time durations tp0 = beginning of precipitation Tw = tpe tp0 = duration of water input tpc = centroid of precipitation Trl = tq0 tp0 = response lag tpe = end of precipitation Tr = tpk tq0 = time of rise tq0 = beginning of hydrograph rise Tlp = tpk tp0 = lag-to-peak tpk = time of peak discharge Tlc = tqc tpc = centroid lag tqc = centroid of response hydrograph Tb = tqe tq0 = time base tqe = end of response hydrograph Tc = tqe tpe = time of concentration response factor (ratio of direct runoff depth to precipitation depth) baseflow index (ratio of baseflow volume to total volume during an event), peak discharge, dimensionless numbers related to hydrograph response lag (Trl), time of rise (Tr), and time base (Tb), bankfull discharge and the slopes of rising and falling limb of the hydrographs. A flow

PAGE 52

40 with 1.67-year return interval is often recognized as bankfu ll discharge (Poff, 1996). The eighteen ecologically relevant hydrologic indices were divide d into four components of hydrologic regimes, magnitude, frequency, durati on, and rate of change, to statistically characterize hydrologic variation. Effects of Flow Magnitude on Stream Ecology This group includes 6 parameters (mean and coefficient of variation) related to response factor, baseflow index, and the peak discharge (qpk). The magnitude of the water condition at any given time is a measure of the availability or suitability of habitat and defines such habitat attri butes as wetted area or habita t volume, or the position of a water table relative to wetland or riparian plant rooting zone s (Richter et al., 1996). High flows maintain ecosystem productivity and di versity. For example, high flows remove and transport fine sediments that would otherw ise fill the interstitial spaces in productive gravel habitats (Beschta and Jackson, 1979). Floods import woody debris into the channel (Keller and Swanson, 1979), where it cr eates new, high-quality habitat (Moore and Gregory, 1988; Wallace and Benke, 1984). Floodplains and wetlands provide important nursery grounds for fish and export organic matter and organisms back into the main channel (Junk et al., 1989; Welcomme 1992). The scouring of floodplain soils revives habitat for plant species that germin ate only on barren, wetted surfaces that are free of competition (Scott et al., 1996) or th at require access to shallow water tables (Stromberg et al., 1997). Effects of Flow Duration on Stream Ecology The 6 parameters in this group measure the duration of all the ev ents considered in the analysis. The parameters included in th is group are the mean, and coefficient of variation of dimensionless indices Trl/Tlc, Tr/Tlc, and Tb/Tlc, where Trl, Tr, Tb, and Tlc

PAGE 53

41 corresponding to the response la g, time of rise, time base, and centroid lag of a storm hydrograph, respectively. The duration is the period of time associated with a specific storm condition that determines the differe nces in tolerance to prolonged flooding in riparian plants (Chapman et al., 1982). Ch anges in duration of inundation can alter the abundance of plant cover types (Auble et al., 1994). For example, increased duration of inundation has contributed to the conversion of grassland to forest along a regulated Australian River (Bren, 1992). For aquatic i nvertebrates and fishes prolonged flows of particular levels can also be damaging (Clo ss and Lake, 1996). Whether a particular lifecycle phase can be completed or the degree to which stressful effects such as inundation of a flood plain can accumulate may be asse ssed from the duration of time over which a specific water condition exists (Richter et al., 1996). Effects of Flow Frequency on Stream Ecology The two parameters in this group measure the number of occurrences of the magnitude of the stream flow condition with respect to bankfull discharge. These numbers of occurrences of bankf ull discharges are reported as percentages of the total events under analysis. Measures of exceedance of bankfull conditions have greater ecological importance. These flows regulat e numerous ecological processes within riparian as well as flood pl ain areas. Frequent, modera tely high flows effectively transport sediment through the channel (Leopo ld et al., 1964). Movement of sediment and organic resources such as detritus a nd attached algae revive the biological community and allow many species with fast li fe cycles and good col onizing ability to reestablish (Fisher, 1983). Consequently, th e composition and relative abundance of species that are present in a stream often reflect the freque ncy and intensity of high flows (Meffe and Minckley, 1987; Schlosser, 1985).

PAGE 54

42 Effects of Flow Rate of Change on Stream Ecology This group includes 4 parameters, mean rate of change in peak discharge in rising limb and falling limb of the storm hydrograph and their variabilities. Flow conditions’ rate of change, or flashiness, refers to the rate at which flow cha nges from one magnitude to another, can influence species persistence and coexistence. At the extremes, flashy streams have rapid rates of change, whereas stable streams have slow rates of change (Poff et al., 1997). The rate of change in wa ter conditions may be tie d to the stranding of certain organisms along the water's edge or in ponded depressions, or the ability of plant roots to maintain contact with phreatic water supplies (Richter et al., 1996). Non-native fishes generally lack the be havioral adaptations to avoi d being displaced downstream by sudden floods (Minckley and Deacon, 1991). Meffe (1984) documented that a native fish, the Gila topminnow ( Poeciliopsis occidentalis ), was locally extirpated by the introduced predatory mosquitofish ( Gambusia affinis ) in locations where natural flash floods were regulated by upstream dams, but the native species pers isted in naturally flashy streams. Data Collection Watershed Characteristics The Fort Benning study watersheds (F igure 3-2), Bonham-1 and Bonham-2, Bonham, Little Pine Knot, and Sally Branch (n amed for the creek/stream that drains the watershed), represent a range of the region’ s soils, topography, land use, and vegetation communities (Table 3-2). In addition to stan dard watershed characteristics, the military training land area and a dimensionless military disturbance index (DIN) were calculated for each watershed. The military land use include fire ranges, maneuver training areas for light, amphibious, and heavy training, air fi eld, artillery and mort ar firing points,

PAGE 55

43 Figure 3-2. Study watersheds, Bonham-1, Bonham-2, Bonham, Little Pine Knot, Randall, Sally Branch, and Oswichee, in the Fort Benning military installation. airborne drop zones, speciali zed non-live-fire training asse ts, and duded impact areas. A DIN is the sum of area of bare ground on slope s greater than 3 degrees and on roads as a proportion of the total watershed area (Maloney et al., in press). Additional details regarding the study area are f ound in Bhat et al., Ecologica l indicators in forested watersheds in Fort Benning, GA: Relations hip between land use and stream water quality, submitted to Ecological Indicators; hereinafter referred to as Bhat et al., submitted manuscript). According to variations in ecologically re levant hydrological char acteristics that are based on flow variability and predictability, and lowa nd high-flow extremes, the

PAGE 56

44 Table 3-2. Physical charact eristics of study watersheds. Acronyms BON-1, BON-2, BON, LPK, and SAL represent the streams (or watersheds) Bonham-1, Bonham-2, Bonham, Little Pine Knot, and Sally Branch, respectively. Physical Characteristics BON-1 BON-2 BON LPK SAL Topography Area, km2 0.76 2.21 12.73 18.01 25.31 Average Elevation, m 121.8 133.5 125.5 146.3 136.8 Average Slope, degree 5.46 4.89 5.04 5.32 5.42 Vegetation Pine, % 28 30 40 41 48 Deciduous, % 27 6 8 2 12 Mixed, % 39 50 22 34 15 Wetland, % 6 8 9 17 10 Military Training Land, % 0 6 5 2 2 NDVI 0.36 0.30 0.32 0.34 0.36 Soil Sandy loam, % 78 69 69 72 49 Loamy sand, % 9 9 31 28 51 Road Road Length, km 3.6 11.4 51.6 56.6 97.6 Road Density, km/km2 4.8 5.1 4.1 3.1 3.8 Stream Stream Length, km 2.6 3.9 29.1 43.3 65.2 Stream Density, km/km2 3.4 1.7 2.3 2.4 2.6 Stream Order 2 2 4 4 4 Other No. of Roads Crossing Streams 1 2 13 11 21 Bare Land, % 1 11 11 4 7 Disturbance Index, % 11 21 19 11 15 streams in the study watersheds are classified as ‘perennial runoff’ (Poff, 1996). As the present study focuses on the impacts of military training within an ecosystem as well as the effects of these impacts on the stream biot a, a reference watershed was identified to contrast with watersheds impacted by military training activities. Approximately 94% of the area of Bonham-1 watershed is covered by forest. Mechan ized military activities are

PAGE 57

45 not conducted in this watershed as compared to 2-6% of the total area of other watersheds used for the same purpose. Also, the wate rshed has a small percentage of bare land, limited roads, and only one road crossing the stream. Hence, Bonham-1 was used as a reference watershed for this study. Precipitation and Streamflow Data Streamflow and precipitation were m easured from January 2000 to December 2003. Precipitation was measured by twelve tipping bucket rain gauges distributed throughout the study area. Watershed precip itation was determined by areal weighting using the Thiessen polygon method. Daily discharge values for Bonham-1 and Bonham2 were calculated from ten-minute continuous stage records using rating curves. Stream stage and velocity were measured half-hour ly for Bonham, Little Pine Knot, and Sally Branch. These data were used to calculat e daily discharges us ing the area-velocity method. Methods Annual-Based Hydrologic Indices The annual-based approach uses multi-year st reamflow records to define a series of ecologically relevant hydrologic indices. Th ese indices may be used to characterize intra-annual variation in wate r conditions, analyze tempor al variations, and compare impacts of alteration among watersheds. For this assessment, nonredundant yet biologically significant hydrol ogic indices are adapted as per the recommendations of Olden and Poff (2003). Table 3-3 lists these indices for the perennial runoff type of

PAGE 58

46 Table3-3. Summary of the recommended hydrologi c indices for 'perennial runoff' type of streams (after Olden and Poff, 2003) us ed in the Annual-Based Hydrologic Indices. BON-1, and BON-2 represent Bonham-1 and Bonham-2, respectively. BON-1 is reference (REF ) and BON-2 is (IMP) watershed. To maintain the consistency to the past studies, the symbols presenting the hydrologic indices in this table are kept as the same as those used in Olden and Poff (2003). Data types A, M, a nd D stand for annual, monthly, and daily discharge data, resp ectively. For definitions and method of calculation of the indices, refer to Appendix A. Flow components and name BON-1BON-2 Data of the hydrologic index SymbolUnits (REF) (IMP) type Magnitude of flow events Average flow conditions Variability in December flow MA26 unitless 0.70 0.47 M Mean annual runoff MA41 m3/s/km20.00397 0.00504 A Spreads in daily flows MA10 1/m3/s 81.4 43.5 D Low flow conditions Baseflow index ML17 unitless 0.12 0.57 A Mean of annual minimum flows ML14 unitless 0.17 0.58 A Median of annual minimum flows ML16 unitless 0.17 0.79 A High flow conditions High flow volume MH23 days/no. of years 9.2 0 A Mean maximum flow in May MH8 m3/s 0.016 0.017 M Median of annual maximum flows MH14 unitless 63.4 4.9 A Frequency of flow events Low flow conditions Frequency of low flow spells FL3 year –1 0.75 0 A Variability in low flood pulse count FL2 unitless 0.31 0.26 A High flow conditions High flood pulse count FH4 year –1 6 0 A Three times median flow frequency FH6 year –1 10.25 2.50 A Seven times median flow frequency FH7 year –1 6 0 A

PAGE 59

47 Table 3-3. Summary of the re commended hydrologic i ndices for 'perennial runoff' type of streams (after Olden and Poff, 2003) used in the Annual-Based Hydrologic Indices (Continued). Flow components and name BON-1 BON-2 Data of the hydrologic index SymbolUnits (REF) (IMP) type Duration of flow events Low flow conditions Variability in annual minima of 90-day means of daily discharge DL10 unitless0.78 0.32 D/M/A Variability in low flow pulse duration DL17 unitless0.09 0.07 A Variability in annual minima of 1-day means of daily discharge DL6 unitless0.60 0.28 D/M/A High flow conditions Means of 30-day maxima of daily dischargeDH13 unitless5.59 1.84 D/M/A Variability in high flow pulse duration DH16 unitless0.006 0.005 A Flood free days DH24 days 56.0 61.5 A Timing of flow events Average flow conditions Constancy TA1 unitless0.28 0.44 D Seasonal predictability of flooding TA3 unitless0.70 0.84 M High flow conditions Seasonal predictability of non-flooding TH3 unitless0.40 0.41 M Rate of change in flow events Average flow conditions Variability in reversals (Positive) RA9 unitless0.04 0.07 D Variability in reversals (Negative) RA9 unitless0.06 0.12 Change of flow (Decreasing) RA7 m3/s 0.1136 0.0324 D Change of flow (Increasing) RA6 m3/s 0.1119 0.0496 D streams found in the study region. The defi nitions and the methods used to calculate these indices are listed in Appendix A. In th is study, the annual-based indices are used to compare and contrast the Bonham-1 and Bonham-2 watersheds.

PAGE 60

48 Storm-Based Hydrologic Indices The storm-based approach uses the stor m hydrographs to determine the indices. Hydrograph separation was used to identify distinct storm events. 44-100 storm events from 2001 to 2003 were used to calculat e the response factor, baseflow index, dimensionless indices (Trl/Tlc, Tr/Tlc, and Tb/Tlc), watershed area (A) scaled peak discharge (qpk/A), bankfull discharges, and the rate of change of peak discharges in rising and falling limbs for five watersheds, where Bonham-1 is the reference watershed (Table 3-4). Appendix B lists the indices, defi nitions, and the methods of calculation. Statistical Analyses The ANOVA and the Tukey’s multiple comparison tests were performed to determine the differences between referen ce watershed and the impacted watersheds’ mean values of indices. The percent of fore st extent, military training land, road density, the number of roads crossing streams, bare land fraction, and dist urbance index were considered to assess the effects of distur bance on hydrology and th e potential impact on stream ecology. Pearson’s correlation coeffi cients were calculated to examine the strength and significance of the relationships between a wa tershed physical characteristic and storm-based hydrologic index. Stepwise multiple linear regressions were performed to identify relationships between an i ndex and management related watershed characteristics. Only variables having p-valu es less than or equal to 0.05 were retained in the regression models. Results Watershed Disturbance Characteristics The watersheds’ physical characteristics ar e summarized in Table 3-2. Most of the watersheds are highly vegetated (70% or mo re) with the majority characterized by pine

PAGE 61

49 Table 3-4. Summary of hydrologic indices used in the St orm-Based Hydrologic Indices. BON-1, BON-2, BON, LPK, and SAL repr esent the streams (or watersheds) Bonham-1, Bonham-2, Bonham, Little Pine Knot, and Sally Branch, respectively. BON-1 is reference (REF ) and other watersheds are impacted (IMP). represents the mean values of the indices of the impacted watersheds that are different at significance le vel of 0.05 from the mean values of reference watershed as confirmed by Tukey’s multiple comparison test. For definitions and method of calculation of the indices, refer to Appendix A. Flow components and name REF IMP of the hydrologic index Symbol Units BON-1BON-2BON LPK SAL Magnitude Mean response factor MMRFunitless0.0300.010 0.060* 0.001*0.020 Variability in response factor MVRFunitless1.09 1.45 1.18 0.89 1.30 Mean baseflow index MMBFunitless0.73 0.86* 0.95* 0.90*0.95* Variability in baseflow index MVBFunitless0.19 0.11 0.04 0.07 0.04 Mean peak discharge MMPDm3/s/km20.0770.025* 0.030* 0.001*0.014* Variability in peak discharge MVPDunitless2.0 2.3 1.2 1.0 1.4 Frequency Bankfull discharge F1FD % 90 47 2 29 6 Two times bankfull dischargeF2FD % 82 32 0 10 0 Duration Duration of time base DMTBunitless1.9 1.7 1.2* 1.3* 1.2* Variability in time base DVTB unitless0.4 0.6 0.6 0.5 0.5 Duration of response lag DMRLunitless0.5 0.6 1.1* 0.8 1.3* Variability in response lag DVRL unitless0.8 1.4 0.6 0.5 0.4 Duration of time of rise DMTRunitless0.5 0.5 0.5 0.5 0.5 Variability in time of rise DVTR unitless0.5 0.8 1.0 0.9 0.7 Rate of change Mean slope of rising limb RPPD m3/s/hr 0.20 0.06* 0.10 0.01*0.04* Variability in rising slopes RPVPDunitless2.2 2.4 1.4 1.4 1.6 Slope of falling limb RNPD m3/s/hr 0.0500.020 0.070 0.001*0.020 Variability in falling slopes RNVPDunitless2.4 2.0 1.5 1.2 1.2

PAGE 62

50 and mixed pine and hardwoods. Deciduous forest typically covers only a small percentage of these watersheds. Howeve r, Bonham-1 consists of 27% of deciduous forest. The study watersheds range from less than 1 to 25 km2. Average elevations and maximum slopes are relatively constant. Sa ndy and loamy soils are common in most of the study watersheds. The military training land (0 to 6%) is relatively small, but varies among watersheds. Total bare lands in these watersheds comprise 11 to 21% of the watershed area, of which 1 to 8% of the total area is unpaved roads and trails. This extent and variability of military training and bare land are typical of the entire Fort Benning installation. While road density is relativ ely comparable among watersheds, the number of roads crossing streams varies from 1 to 21. Military training land, bare land, and disturbance index are positively correlated at significance level of 0.05 or lower. Annual-Based Hydrologic Indices Table 3-3 summarizes annual-based analysis results for Bonham-1 (reference) and Bonham-2 (impacted) watersheds. The average flow conditions revealed that the mean annual flow (MA41) is higher in the impacted wate rshed. However, the reference watershed has higher flow va riability for both annual (MA10) and December month (MA26) periods. The average flow event timing is more constant (TA1) and predictable (TA3) in the impacted watershed. The impacted watershed maintained a higher magnitude of minimum flows as depicted by the baseflow index (ML17). This result is consistent across other low flow indices (ML14 and ML16) and the findings that the impacted watershed had no low flow spells (FL3) and lower variability of low flow pulse counts (FL2). Higher coefficient of variation in annual minima of 90-day (DL10) and 1-day (DL6) means of daily discharge for the reference watershed and similar va riability in low flow pulse duration (DL17)

PAGE 63

51 were found for both watersheds. While low fl ow values vary more for the reference watershed, once the flow goes low, it stays lo w for same duration in both watersheds. The reference watershed produces higher magn itude flow and thus maintains higher median flow (MH14) during events. However, in the month of May, mean of the maximum monthly flows (MH8) were similar for both the watersheds. During high flow conditions, the reference watershed crosses a threshold of seven times the median annual daily flow volume (MH23). In the impacted watershed, these floods never occurred. The high flood pulse counts (FH4) and the frequency of floods (FH6 and FH7) in the reference watershed are more than the impacted watershed. However, on a few occasions, the flow crossed a lower threshold of 3 tim es median frequency of flood (FH6) in the impacted watershed. The 30-day floods (DH13) went higher and stayed high (DH16) in the reference watershed. Periods between floods (DH24) are similar for both watersheds, that is, they have approximately two months of flood free days a year and of comparable predictability (TH3). Storm-Based Hydrologic Indices The results of storm-based hydrologic a ssessment are summarized in Table 3-4. Analysis of variance tests indi cated that the mean values of storm-based indices except the time of rise differ at th e significance level of 0.05. Fo r a number of indices, the reference watershed exhibits distinct behavior as compared to the impacted watersheds. Tukey’s multiple comparison tests indicate that the indices other than time of rise and the rate of change in falling limb in the refere nce watershed were signi ficantly different from the impacted watersheds. The reference wa tershed is characterized by a relatively low baseflow index (MMBF) with significantly higher (MMPD) and more variable (MVPD) peak discharge. During events, 90% of the total events produced greater than bankfull

PAGE 64

52 discharge (F1FD) in the reference watershed indi cating a highly connected system as compared to 2-47% in impacted watersheds. Storm flows consistently last ed longer and responded faster to rain events in the reference watershed (DMTB and DVTB). Once the stream responded, the time of rise (MMTR) was similar in all the watersheds. The reference watershed’s combination of fast response and high peak discharge resu lts in a rapidly increasing rising limb (RPPD) as compared to impacted watersheds. Relationship between Military Land Ma nagement and Storm-Based Hydrologic Indices Correlation and regression analyses were performed to determine the relationship among the watershed physical characteristics and the storm-based hydrologic indices. Table 3-5 indicates that 7 ke y storm-based hydrologic indices are significantly related to military land management. Increased military training land, bare land, and the disturbance index will increase th e time of rise as well as the variability in the time base. Increasing the road density increases the vari ability in the time base and the rate of change of rising limb. Increasing the numbe r of roads crossing streams increases the storm response lag, but decreases the time base. Results also show that an increase in the number of roads crossing streams decreases the variability in the rate of change of falling limb. No effects on hydrologic indices were identified for forestry management practices. Stepwise multiple correlations characterized the response of storm-based indices to military impacts (Table 3-6). The greatest impact of land management is found with statistically significant predictive models for indices of time base, response lag, and time

PAGE 65

53 Table 3-5. Pearson correlation coefficients between waters hed physical characteristics and event based hydrologic indices. Char acteristics are acronymed as Forest (FOR), Military Training Land (MIL), Road Density (RDN), No. of Roads Crossing Streams (NRC), % Bare Land (PBL ), and Disturbance Index (DIN). and ** indicates significance at or below 0.05, and 0.01 probability levels, respectively. Hydrologic Indices FOR MIL RDN NRC PBL DIN MMRF 0.25 0.15 0.15 0.13 0.27 0.31 MVRF 0.40 0.59 0.71 -0.05 0.67 0.81 MMBF -0.62 0.44 -0.56 0.83 0.68 0.41 MVBF 0.63 -0.39 0.61 -0.85 -0.63 -0.35 MMPD 0.83 -0.37 0.64 -0.62 -0.47 -0.19 MVPD 0.66 0.21 0.96** -0.73 0.04 0.34 F1FD 0.57 -0.40 0.57 -0.85 -0.65 -0.39 F2FD 0.69 -0.43 0.62 -0.82 -0.64 -0.36 DMTB 0.63 -0.17 0.77 -0.91* -0.42 -0.12 DVTB -0.50 0.98** 0.14 0.08 0.96** 0.90* DMRL -0.23 0.06 -0.54 0.97** 0.41 0.18 DVRL 0.21 0.59 0.84 -0.78 0.33 0.57 DMTR -0.39 0.99** 0.32 -0.17 0.89* 0.90* DVTR -0.86 0.67 -0.45 0.35 0.68 0.47 RPPD 0.79 -0.34 0.59 -0.55 -0.42 -0.16 RPVPD 0.63 0.13 0.91* -0.78 -0.06 0.24 RNPD 0.42 0.10 0.35 -0.13 0.14 0.26 RNVPD 0.70 -0.07 0.85 -0.88* -0.29 0.03 of rise. Military training land, road dens ity, and the number of road crossing streams were the three management variables that impacted storm responses.

PAGE 66

54 Table 3-6. Stepwise multiple regression models for event based hydrologic indices. Military Training Land (MIL), Road Density (RDN), and No. of Roads Crossing Streams (NRC), are the inde pendent variables retained in the regression analyses. and ** indica tes significance at or below 0.05, and 0.01 probability levels, respectively. Hydrologic indices Independent variables retained and R2 (adj) regression equations MMBF 0.729 + 1.91 MIL + 0.00968 NRC 0.94* MVPD 1.21 + 0.666 RDN 0.88** F1FD 0.929 6.97 MIL 0.039 NRC 0.94* DMTB 1.8 0.0352 NRC 0.77* DVTB 0.418 + 3.05 MIL 0.95** DMRL 0.483 + 0.0393 NRC 0.92** DMTR 0.454 + 0.837 MIL 0.97** RNVPD 2.20 – 0.0563 NRC 0.70* Discussion Results representing annual-based av erage flow conditions showed higher magnitudes of MA26, MA41, and MA10 in the reference watershed as compared to the impacted one. In addition, these flows are less constant and le ss predictable over the years as compared to the impacted waters hed. Aquatic communities can show distinct differences to changes in velocity and reducti on in bed gradient, and associated fining of bed sediments (Clausen and Biggs, 2000). Periphyton and benthic invertebrates are particularly sensitive to diffe rent velocities and bed sedime nt size/stability (Minshall, 1984; Biggs, 1996). Clausen and Biggs (1997) found that invertebrate species richness and periphyton biomass changed based on flow These changes are likely related to riparian vegetation on the amount of leaf li tter input (Vannote et al. 1980; Davies-Colley and Quinn, 1998).

PAGE 67

55 The higher magnitude of the annual-based low flow indices (ML17, ML14, and ML16) in the impacted watershed is attributed to higher groundwater input as compared to the reference watershed. This increased input likely reflects the reduction of interception characteristic of the military land uses (Bryant et al., in press). The relative magnitude of low flows is likely to have important influences on biota through the intensity of habitat destruction associated wi th drying during low flows. In this study, the lower magnitudes of low flows as de picted by lower baseflow index (ML17) and smaller low flow pulses (FL3) suggest that the long periods of low flow condition are more likely in the reference watershed. Higher variabil ity in the duration during low flow condition (DL10, DL17, and DL6) in the reference watershed supports likelihood of such long periods of low flow conditions. Long pe riods of low flow conditions and higher variability in these conditions may provide se lective pressure for specific life history characteristics such as inve rtibrate aestivation and e gg diapause, and physiological tolerance to low dissolved oxygen (Williams and Hynes, 1977). The higher magnitude of the a nnual-based high flow indices (MH23 and MH14) and peak discharges (MMPD and MVPD) in the reference watershed as compared to the impacted watershed suggests the likelihood of habitat regenerati on associated with sediment transport and floodplain inundation during high flows. For high flow events, the degree of riverbed communities’ disturba nce is strongly related to degree of bed movement (e.g., Biggs et al., 1999). Di ssolved inorganic nitrogen and phosphorus concentrations in rivers are strongly negativ ely correlated with specific yield and high flow magnitude among watersheds, and among years within watersheds (e.g., Biggs and Close, 1989; Close and Davies-Colley, 1990; Gr imm and Fisher, 1992). This reflects the

PAGE 68

56 degree of flushing of the nutrients minera lized through organic matter breakdown in the soil profile and leachate from the underlying substrata. Floods or high flow conditions are also important in influencing community structure. The results based on the annual flow in this study clearly show that the reference watershed produced more frequent high flows (FH4, FH6, and FH7). Similarly, the results from storm-based an alysis show that the frequenc y of discharges equaling or exceeding bankfull (F1FD and F2FD) is higher in the refere nce watershed. Floods are widely viewed as reset mechanisms (Resh et al., 1988), and flood-related mortality to lotic organisms can result either directly from scouring, crushing, or downstream export of individuals (Minckley and Maffe, 1987) or indirectly from food resources loss (Hanson and Waters, 1974). As flood freque ncy increases, some invertebrates in the reference watershed actively migrate either in to the substratum or to quieter backwaters to avoid sudden floods. Floods have been shown to regulate community structure by facilitating local coexistence between asymme trically competitive algal species (Power and Stewart, 1987) and invertibrate species (Hemphill and Cooper, 1983) and between an exotic fish predator and its native, rela tively flood-resistant prey (Meffe, 1984). The result in this study showed th e consistency of storm-based F1FD and F2FD with annualbased FH4, FH6, and FH7 indices suggesting F1FD and F2FD may be used as alternative indices. As indicated by the higher annual-base d mean duration of high flow condition (DH13) and the duration of storms (DMTB) in the reference as compared to the impacted watershed, it is apparent that water resides in the referen ce watershed for a longer period of time during high flow condition. However, the DMRL is smaller for the reference

PAGE 69

57 watershed suggesting a quick ri se in hydrograph. This can be attributed to a better connectivity of riparian areas to the stream. This connectivity is important as the nutrient concentrations can strongly control auto trophic production during inter-flood periods (e.g., Biggs, 2000). Also, indivi dual high flow events grea tly reduce the biomass and change the species composition of peri phyton (e.g., Biggs a nd Stokseth, 1996), and invertebrates (e.g., Cobb et al., 1992). Predictive relationships we re identified for storm-ba sed hydrologic indices based on watershed scale military management and suggest that the key variables related to hydrologic alteration are military training land, road density, and the number of roads crossing streams. Military training within a watershed can modify annual and storm generated runoff due to changes in drainage, vegetation, and soils. The impact is due to troop maneuvers and large, tracked and wheeled vehicles that traverse thousands of hectares in a single training exercise (Quist et al., 2003). These activities’ impacts, ranging from minor soil compaction and l odging of standing vegetation to severe compaction and complete loss of vegetation cove r in areas with concentrated training use (Wilson, 1988; Milchunas et al., 1999), are evident in the stream hydrographs and suggest impacts to stream ecosystems. Di sruption in soil density and water content (Helvey and Kochenderfer, 1990), addition of sediment, nutrients, and contaminants in aquatic ecosystems (Gjessing et al., 1984) and impairment of natural habitat development and woody debris dynamics in forested floodplain streams (Piegay and Landon, 1997) are potential impacts on terrestria l and aquatic ecosystems. In a recent study in the same watersheds in Fort Be nning, Houser et al. (2 004) found that gross primary productivity, respiration, and the benthi c organic matter were low in the stream

PAGE 70

58 corresponding to the impacted watershed (highe r disturbance index) as compared to the reference one (lower disturbance index). Road crossings commonly found in military training areas may act as barriers to the move ment of fishes and other aquatic animals (Furniss et al., 1991). Roads also directly change the hydrology by intercepting shallow groundwater flow paths, diverting the water along the roadway and routing it efficiently to streams at crossings (Wemple et al, 1996). This can cause or contribute to changes in timing and routing of runoff (Jones and Grant, 1996). Conclusion and Recommendations In the present study, a reference watershe d was compared to impacted watersheds using annual and storm-based hydrologic in dices within the Fort Benning military installation. The results s uggest a subset of the hydr ologic indices recommended by Olden and Poff (2003) are necessary for the Fort Benning streams. It is recommended that mean annual runoff and its spread, base flow index, high flow volume, frequency of low flow spells, high flood pulse count, variab ility in annual minima of 90-day means of daily discharges, and constancy be used in stream ecology studies in the Fort Benning streams to identify disturbances relative to a reference waters hed at a watershed scale. The indices based on storm data may be us ed to augment the annual indices or as surrogate to those indices. Storm-based ma gnitude and variability in peak discharge, baseflow index, and the bankfull discharge were consistent with the results from annualbased analysis. With respect to the potent ial influence of the frequency and duration aspects of the flow regimes on the stream ecology, duration and vari ability of time base, and duration of response lag are identified as critical hydrologic indices. The military management practices, military training la nd, road density, and the number of road crossing streams, were found to sign ificantly affect these indices.

PAGE 71

59 CHAPTER 4 PREDICTION OF NITROGEN LEACHING FROM FRESHLY FALLEN LEAVES: APPLICATION OF RIPARIAN ECOSYSTEM MANAGEMENT MODEL (REMM) Introduction Forest canopy litterfall initiat es the major pathway for recycling of nutrients from plant to soil (Bubb et al., 1998). Nutrients are returned to the soil via litterfall and throughfall fluxes in forested ecosystem (U konmanaho and Starr, 2001). Litterfall, comprised predominantly of leaf litter, is usually the most important nutrient source (Herbohn and Congdon, 1998). Litter quantity, li tter decomposition and nutrient release patterns are factors for unde rstanding nutrient cy cling in forest ecosystems (Rogers, 2002). Litterfall is primarily caused by natura l senescence of leaves Forest litterfall quantity and composition varies among tree species, stand age and development, and reflects environmental conditions, particularly water and nutrient availability (Binkley, 1986; Polglase and Attiwill, 1992). The am ount and distribution of litterfall through time are also affected by season, rainfall amount and distribution, and wind speed (Crockford and Richardson, 1998). Fluxes of dissolved organic matter are an important vector for the movement of nutrients both within and between ecosyste ms (Cleveland et al., 2004). In many forest ecosystems, more than half of the nitrogen is soil solution is in or ganic form (Qualls et al., 1991). Hydrology plays an important role in both pr oduction and mobilization of dissolved organic matter in the forest floor (Park and Matzner, 2003). A substantial amount of potentially soluble organic matter exists in an adsorbed phase (Christ and

PAGE 72

60 David, 1996). The amount of percolating wate r mobilizes these sol uble organic matters (Tipping et al., 1999). Past studies have shown that the dissolv ed organic matter flux increases when the percolating water passes thr ough the forest floor (McDowell and Likens, 1988; Qualls et al., 1991; Currie et al., 1996; Mich alzik et al., 2001). In a rece nt study in India, Singh et al. (1999) reported that the ra pid mass loss of leaf litter o ccurred during the rainy season and the rate of litter mass loss was positively corr elated with the cumulative rainfall. In an early study, Bocock et al. (1960) observed a rapid leaching from oak and ash leaf litter, and suggested that the weight loss during the first month was largely due to a physical loss by leaching. The primary sources of dissolved nutrients are considered to be leaching of substance from fresh litter and the product of decomposition of the plant litters (Qualls et al., 1991). Water-soluble nutri ents are more easily leached from the leaf litter of deciduous species, includi ng birch, than from conifer ous species, including Scots pine ( Pinus sylvestris ) and Norway spruce ( Picea abies ) (Harris and Safford, 1996; Hongve, 1999). Depending on the species, leaf litter may release from 5 to 30% of original dry weight as dissolved organic matter within 24 hours (Cummins, 1974). In a woodland stream ecosystem in Massachusetts, McDowell and Fisher (1976) found that leaves lose weight rapidly by abiotic leaching of solubl e constituents for 1-3 days and decay more slowly thereafter as a result of microbial decomposition. Gosz et al. (1973) estimated 15% mass loss for yellow birch leaves ( Betula lutea ) due to leaching within first month on the forest floor. Heath et al. (1966) re ported a 26% weight loss from birch leaves ( B.

PAGE 73

61 alba ) that remained for two months on the fo rest floor. Nykvist (1961) found an 8% weight loss in B. verrucosa after 24 hours of leaching in distilled water. Annual fluxes of nitrogen in li tterfall are greater than ot her nutrients regardless of forest type but the amounts vary among fore st types (Ukonmaanaho and Starr, 1999). Dissolved organic nitrogen is the major form of nitrogen in stream water draining from many mature forest watersheds, comprising abou t 95% of the total nitrogen (Qualls et al., 1991). Nitrogen content of a hardwood fo rest litterfall on the Hubbard Brook experimental forest comprised approximatel y 37% of the total nutrients (Gosz et al., 1972). In a low nutrient system at the Cow eeta Hydrologic Laborator y in North Carolina, where carbon (C) to nitrogen (N ) ratios in freshly fallen litter varied from 100 to 220, Qualls et al. (1991) reported that the output of dissolved or ganic nitrogen from the forest floor was 28% of the nitrogen input from the litterfall. The Riparian Ecosystem Management Mode l (REMM) was developed by the U.S. Department of Agriculture’s (USDA) Ag ricultural Research Service (ARS) to characterize the role of the riparian area on stream water quality (Lowrance et al., 2000; Altier et al., 2002). In REMM, th e riparian areas’ water qual ity functions for control of N, phosphorus (P), and sediment transport into surface waters are simulated and analyzed. This model has primarily been te sted and applied on high-nutrient riparian buffers adjacent to agricultural fields in th e past (e.g., Inamdar et al., 1999a; Inamdar et al., 1999b; Lowrance et al., 2000). The complex dynamics related to the nutrien ts in a watershed is particularly of interest in low nutrient systems as the releas e of nutrients from plant litters prior to decomposition may be an important aspect of characterizing stream water quality. A

PAGE 74

62 watershed scale study related to surface water quality was conducted in the Fort Benning military installation located in southwest Georgia. Experimental data from Fort Benning shows an increase in total Kjeldahl nitrogen concentrations following litterfall. The release of nitrogen is relatively rapid as comp ared to that released from decomposition of organic matter. It is hypothesi zed that after leaves fall, the first precipitation events can be expected to result immediate nitroge n leaching and a corresponding increase in nitrogen levels in the stream water. The goa l of this study is to determine whether freshly fallen leaves in a riparian area are a significant source of nitrogen in low nutrient system. The specific objectives of this study are to quantify nitrogen in streamflow using REMM and verify the modeled results with the observed data. Basic Concepts of REMM Originally, REMM was developed to si mulate buffer systems that meet specifications recommended by the USDA Forest Service and the USDA Natural Resource Conservation Service as a national standard (Natural Re source Conservation Service, 1995). Three zones para llel to the stream characterize the riparian system. Thus REMM simulates a three-zone buffer. The ripari an area’s soil is characterized in three layers. In these soil layers, vertical a nd lateral movement of water and associated dissolved nutrients takes place. The top two soil layers correspond with the soil horizons A and B, respectively. Depth of the root z one or a restricting soil horizon defines the third soil layer. A litter layer that interact s with surface runoff covers the soil layer 1. The model takes upland water output supplied by the user and calculates loadings of water, nutrients, sediment, and carbon based on actual area of the zones of a riparian area (Lowrance et al., 2000).

PAGE 75

63 The model input data consist of the phys ical conditions of the riparian area, external loadings of water and nutrients to the riparian area, soil, vegetation, and soil information. Altier et al. (2002) documents a detailed description of the algorithms, equations, and parameters used in the model. Hydrology component of REMM REMM measures water stores and fluxes us ing a daily water balance. The water balance includes interception, evapotranspiration, infiltration, verti cal drainage, surface runoff, subsurface lateral flow, upward flux from water table, and deep seepage. Interception losses occur in th e vegetation canopy and litter la yer. Canopy interception is an exponential function of the canopy storage capacity and th e amount of daily rainfall, and is simulated using a modified form of the Thomas and Beasley (1986) equation. Precipitation falling through the canopy (throughfall) is subjec t to litter interception. Similar to canopy interception, litter intercepti on is determined by the mass of the litter at any given time and the litter storage capacity. Infiltration at the soil surface depends on the total depth of water. This depth of water is the sum of the throughfall depth and overland flow depth from the upslope zone or field. If the sum of incoming upland r unoff and throughfall depth is less than the infiltration capacity of the ri parian soil, runoff infiltrates at the rate of application. During high intensity storms, the runoff and throughfall rate may ex ceed the infiltration rate. Surface runoff from the riparian buffe r occurs when total water depth exceeds the infiltration depth. In such conditions, model simulates the infiltra tion using a modified form of the Green-Ampt equation (Stone et al., 1994). When soil moisture exceeds the field capaci ty, vertical drainage from a soil layer occurs. This vertical drainage also depends on the available water storage capacity of the

PAGE 76

64 receiving layer. The rate of drainage is co ntrolled by the lesser of the vertical hydraulic conductivities of the draining a nd receiving layer. Vertical unsaturated conductivity is simulated as a function of the soil moisture content using Campbell's equation (Campbell, 1974). When a water table builds up over the restricting soil layer, subsurface lateral movement occurs. Lateral water movement is simulated using Darcy’s equation. Saturation excess overland flow occurs when soil profile is completely saturated. Nutrient component of REMM Simulation of nutrients in REMM is base d on the Century Model (Parton et al., 1987). The C cycle is fundamental for simu lation of all organic matter dynamics and many nutrient cycling processes in REMM. Release and immobilization of N is in proportion to transformations of C. Soil a nd litter C is divided into residue (woody debris, leaf litter, a nd roots) and humus (soil organic ma tter) pools. Each C pool has an associated mineralization rate, transformation efficiency, and a specific range of C to N ratios (Inamdar et al., 1999b). A first order rate equation with respect to C determines the mineralization of C from each pool. Litter from leaves, stems, branches, coarse roots, and fine roots are allocated into a readily decomposable (metabolic) or a resistant (structural) residue pool based on lignin to N ratio (Pastor and Post, 1986). Each pools decomposes at different rates. REMM assumes a fixed fraction of C in fr esh plant residue. Dissolved C from metabolic residue and active humus pools diffe r by litter and soil la yer (Altier et al., 2002). The sum of C in throughfall and inco ming runoff is mixed with C on the ground surface to determine dissolved and adsorbed concentrations. Dissolved C moves with surface runoff, subsurface lateral flow, and ve rtical drainage (Lowrance et al., 2000).

PAGE 77

65 The amount of dissolved C carried out of the litter or soil layer is a f unction of the water volume and the dissolved C concentration. Nitrogen is input to the riparian area through precipitation, surface and subsurface water flow, and adsorbed to incoming sediment carried by surface flow. Presence of nitrogen in different forms and the asso ciated degree of physical and chemical stabilization influences its availability for microbial transformations and plant uptake. Effective C to N ratios influencing residue d ecomposition are calculated in the model as a function of the content of N in the litter as we ll as the content of available inorganic N in the soil. As decomposition of the litter take s place, C and N are released (Lowrance et al., 2000). Inorganic nitrogen, ammonium and nitrate, are both available for immobilization into soil organic matter. Immobilization of nitrate occurs only after all available ammonium has been used. Amm onium may be in solution or adsorbed to the soil matrix. Nitrification is calculated with a first-orde r rate equation influen ced by temperature, soil moisture, and pH. Temperature and soil moisture are variables, recalculated each day but pH is a constant for each zone and laye r combination. With increasing amounts of substrate, the rate of nitrif ication is lagged to represent a delayed microbial response. The calculation of denitrificati on is a function of the interac tion of factors representing the anaerobic condition, temperatur e, nitrate, and mineralizable C. Denitrification mostly occurs as water-filled pore space gets a bove 60% (Inamdar et al., 1999b). Nitrogen transformation processes are shown in Appendix C.

PAGE 78

66 Application of REMM Study Area The study area consisted of a second-order watershed, Bonham-South (221 ha), in the Fort Benning military installation, Georgia (Figure 4-1). A detailed description of this site along with the instrumentation and the type of data collected is provided in Figure 4-1. Study area. Bonha m-South is a second-order wa tershed within the Fort Benning military installation in Georgia. Bryant et al., in press and Bh at et al., in review. The ri parian area covers approximately 2.5% of the total watershed area. The ripari an length is 3,500 m. Based on the visual inspection in the watershed, a 12 m wide riparian buffer along the stream is considered.

PAGE 79

67 This riparian buffer is divided into three e qual zones (Zone 1, Zone 2, and Zone 3) of 4 m each. Corresponding soil layers (Soil laye r 1, Soil layer 2, and Soil layer 3) are considered in each zone. A selective list of riparian area and hydrological parameters and their corresponding values are provided in Table 4-1. The riparian area consists primarily Table 4-1. REMM model parameters values by the riparian zone for the Bonham-South watershed in Fort Benning, Georgia. The dimension of riparian area from the upland to the stream perpendicular to th e stream is referred to as width, and the distance along the stream is referred to as length. The parameter values are identical for Zone 1, Zone 2, and Zone 3 except for the slope. The superscripts 1, 2, and 3 of the slope va lues represent the Zone 1, Zone 2, and Zone 3, respectively. Parameters Units Values Riparian zone length m 3,500 Riparian zone width m 4 Slope % 3.01, 3.82, 4.23 Manning's n unitless 0.06 Total soil profile thickness m 3.3 Individual soil horizon thickness m Soil layer 1 0.3 Soil layer 2 1.0 Soil layer 3 2.0 Saturated hydraulic conductivity cm/h Soil layer 1 15 Soil layer 2 13 Soil layer 3 1 Porosity cm/cm Soil layer 1 0.40 Soil layer 2 0.35 Soil layer 3 0.35 Wilting point cm/cm Soil layer 1 0.07 Soil layer 2 0.07 Soil layer 3 0.07 Riparian area ha 4.2 Field drainage area (surface) ha 216.8 Field drainage area (subsurface) ha 221.0

PAGE 80

68 of hardwoods (approximately 60%) that are mostly scrub oak ( Q. rubra ), sweetbay ( M. virginiana ), sweet gum ( L. styraciflua ), water oak ( Q. nigra ), willow oak ( Q. phellos ), red maple ( A. rubrum ), and swamp tupelo ( N. aquatica ). Mixed species including southern scrub oak ( Q. rubra ) and yellow pine ( P. ponderosa ) cover approximately 21% of the riparian area, whereas loblolly ( P. taeda ) and longleaf ( P. palustris ) pine cover approximately 19%. The riparian area soils are predominantly sandy loam. The uplands are mostly on loamy sand. Data and Input Parameters REMM operates at a hillslope scale. To a pply the model at watershed scale, it is necessary to convert the waters hed into an equivale nt hillslope by adju sting the geometry of watershed. As the riparian buffer has sp ecified length (parallel to the stream) and width (perpendicular to the stream), upland c ontributing field width of the watershed was adjusted to equate the total area of the watershed. REMM requires soil and vegetation nutrient pools for the riparian area of the watershed to initialize the model (Table 42). Required field da ta to run the model includes daily surface and subsurface runoff from the uplands to the riparian area (Table 4-3). The model also requires daily weat her data and paramete r values representing topographic, soil, and vegetative conditions within the riparian area. Strategic Environmental Research and Development Plan’s (SERDP) Ecosystem Management Project (SEMP) continuously records the atmospheric data within the military installation. Daily weather data were obtaine d from SEMP data re pository that included precipitation amount and duration, maximum and minimum air temperatures, solar radiation, wind speed, and dew poi nt temperature. Daily st reamflow for Bonham-South

PAGE 81

69 Table 4-2. Soil and vegetation nutrient pool s for Bonham-South watershed’s riparian area. The litter and soil pools of nutrient consist of metabolic, structural, active, slow, and passive humus pools. Th e values in litter layer as well as individual soil layers in each zone ar e presented. Vegetation pools consist of nutrients in leaves, stems, branches, coar se roots, fine roots, and heartwoods. Total values of nutrients are reported in the vegetation pool. The parameter values are identical for Zone 1, Zone 2, and Zone 3. Parameters DepthBulk densityCarbon Nitrogen Biomass (m) (g/cm3) (kg/ha) (kg/ha) (kg/ha) Litter and Soil Litter 0.01 1.0 18,100 1,134 Soil layer 1 0.30 1.4 29,280 655 Soil layer 2 1.00 1.5 24,640 644 Soil layer 3 2.00 1.5 11,560 252 Total 83,580 2,685 Live vegetation 31,390 753 78,477 Table 4-3. Model simulated annual hydrologic budget for the riparian areas in BonhamSouth watershed. and ** represents the surface and subsurface flow input to the zone 3 from uplands. Values are rounded off to nearest whole numbers. Year Units 2000200120022003 4-yr average Precipitation mm/yr 475 736 731 1,014 739 Surface flow in* mm/yr Subsurface flow in** mm/yr 119 134 143 253 162 Observed stream flow (total) mm/yr 121 137 147 258 166 Simulated surface flow mm/yr 2 5 4 6 4 Simulated subsurface flow mm/yr 106 118 130 244 149 Simulated stream flow (total) mm/yr 109 124 133 250 154 was calculated from 10-minute continuous stag e records using rating curves and areavelocity method. Typically, REMM uses daily surface r unoff depth generated from the runoff collected by surface runoff samplers at the upland-Zone 3 interface. REMM uses the hydraulic gradient upland wells and Zone 3 we lls, saturated thickness at Zone 1, and the saturated hydraulic conductivity to compute s ubsurface flow loading to the buffer. Due

PAGE 82

70 to the lack of measured surface and subsurface flow from the upland to the riparian area, it is assumed that the baseflow fraction in th e measured streamflow is a contribution of upland subsurface flow to the riparian area. Therefore, measured streamflow from January 2000 through December 2003 was parti tioned into baseflow and surface runoff using constant slope base flow separa tion technique (McCuen, 1998). Surface runoff from the upland to the riparian area was not considered (Table 4-3). Parameters that describe the riparian area dimensions, soil, and vegetation character istics were derived from measured data and previously published literature (Inamdar et al., 1999a, Inamdar et al., 1999b, Lowrance et al., 2000; Ga rten Jr., et al., 2003). Im portant parameters related to litter and soil layer are listed in Appendix D. Monthly canopy cover in the riparian area was determined by direct measurement with a Model-A spherical densiometer using the method outlined by Lemmon (1956) fr om June 2001 to September 2003. Measured water and soil nutrient concentra tions in the study area included the total Kjeldahl nitrogen (TKN), nitrate, and ammonium in precipitation, streamflow, soil water, and shallow groundwater. Two transects in the riparian area near the outlet of the watershed were considered for groundwater and soil water monitoring. Groundwater monitoring wells (1-3 m deep) and adjacent tension lysimeters at 20 and 60 cm depths were positioned on both sides of the stream along the riparian transects. Stream water, groundwater, and soil water samples were co llected from October 2001 to September 2003; biweekly from October 2001 to Novemb er 2002 and monthly thereafter. BonhamSouth stream was sampled during 16 st orm events between September 2002 and September 2003 using an event triggered ISCO sampler that was programmed to collect

PAGE 83

71 hourly samples based on the flow depth. Bhat et al. (unpublished manuscript), describes the sample analysis procedures for different chemical constituents. As REMM simulations are performed on a daily time step, and observed TKN concentrations are of hourly time step, co mparisons between the two were based on the TKN masses produced during the events that lasted 24 hours or longer. The model simulates the masses of dissolved and partic ulate organic nitrogen, dissolved ammonium and nitrate in surface and subsurface flows on a daily basis. The mass of simulated TKN is the sum of dissolved and particulate organic nitrogen and dissolved ammonium. Carbon dissolves from metabolic residue a nd active humus pools. In REMM, incoming C from precipitation, surface, and subsurface flow is assumed to be in dissolved form. Carbon available to be dissolved from the ac tive humus pool is a fr action of the total C present in the same pool. This fraction of set at 0.31 based on an estimate by McGill et al. (1981) for the proportion of dying bacteria in metabolic form. The total amount of dissolved C is the sum of metabolic C residue and the dissolved C fraction in the active humus pool. In REMM, stoichiometric relati onships are assumed between C and N in the organic matter. As C is transformed, corres ponding N is also transformed. A Freundlich isotherm method determines the amount of di ssolved ammonium from inorganic pool of N in REMM. Observed masses of TKN during storm events are calculated by multiplying the concentrations by the volume of water. Model Calibration REMM’s hydrology and nitrogen components we re evaluated using data collected in this study from riparian area as well as data from the literature. Data collected in nearby watersheds in Fort Benning, Georgia we re used to initialize the C and N in the soil pools (Garten Jr. et al., 2003) The litter C and N pools ar e based on the literature

PAGE 84

72 values (Silveira, M.L., Reddy, K.R. (Departmen t of Soil and Water Science, University of Florida), Comerford, N.B. (Department of Soil and Water Science, University of Florida). Litter decomposition and soluble organic carbon and nitr ogen release in a forested ecosystem. Unpublished manuscript) Soil and vegetation nutrient pools for the riparian area of the watershed are liste d in Table 4-2. Bonham-South’s soil layer thickness, soil porosity, saturate d hydraulic conductivity, clay content, carbon decay rate, and denitrification rate were calibrated. Model calibration involved the comparison of the simulated streamflow and TKN output with the measured values. While calibrating hydrology a nd nutrient components of the model, parameters such as soil and litter C and N pools, which are based on literature values, were fixed. Other fixed pa rameters included ripa rian length and width, surface and subsurface draining area. The remaining parameters including soil layer thickness, saturated hydr aulic conductivity, soil porosity, clay content, carbon decay rate, and denitrification rate in each zone and soil la yers of the riparian buffer were adjusted to match the observed and predicted results. Fox (1981) recommends mean biased error (MBE) and mean absolute error (MAE) to measure the difference between observed and model predicted values. Mean biased error is calculated as the average erro r between predicted and observed values accumulated over the total number data points. Mean absolute error considers the absolute values of the errors. In the present study, MBE and MAE are modified to calculate the percent difference of modele d streamflow and TKN mass from the observed values. Mean bias difference (MBD) calculates the average difference accumulated over the total number of events, e xpressed as a percent of the obs erved value. Mean absolute

PAGE 85

73 difference (MAD), expressed also as a percen t, considers the absolute values of the differences. Mean absolute difference takes ne gative values and repla ces them with their absolute values. Observed and modeled streamflow was also compared using the NashSutcliffe efficiency. The Nash-Sutcliffe ef ficiency criteria is based on the normalized least square objective function that evaluates the sum of the squares of flow residuals (Nash and Sutcliffe, 1970). Sensitivity Analysis A sensitivity analysis was performed to de termine the effects of key hydrological, soil, and vegetation parameters on streamflow and TKN fluxes. The parameters were canopy cover fraction, riparian zone width, soil layer thic kness, maximum carbon decay rate for litter and humus, maximum denitr ification rate, soil saturated hydraulic conductivity, soil porosity, and so il clay content. Each pa rameter was changed by +10% and –10% from the values used as the best estimates for the ca libration simulations. Results Hydrology Simulated daily flows for the study waters hed closely matched the observed flows (Figure 4-2). Over the 4-year simulati on period, MBD and MAD between observed and predicted streamflow were 2% and 12%, re spectively. The model had Nash-Sutcliffe efficiency of 80%. REMM tends to underest imate the streamflow during low flow and overestimate during the storms. Approxima tely 1% of the streamflow was the contribution from surface runoff generated in th e riparian area. Majority of streamflow output was through the subsurface (Table 4-2).

PAGE 86

74 Figure 4-2. Comparison of REMM simulated dail y flow with the observed daily flow for the Bonham-South watershed. There was considerable variability be tween and within years for measured precipitation and streamflow data during study pe riod. It is evident from Figure 4-2 that an increase in the average flow during the ye ar 2003. This increase was consistent with the variation in precipitation during the st udy period. Overall, model simulations for streamflow can be considered good for the calibrated parameters. Nitrogen Observed stream TKN concentrations during the study period showed a strong correspondence with the leaf fa ll (Figure 4-3). Approximate ly 50% of the canopy in the riparian area dropped during November-January each year. Higher TKN concentrations were observed during the same period of time. The simulated TKN concentrations also correspond to simulated leaf fall (Figure 4-4). Modeled TKN concentr ations were higher during high leaf fall months. From Oct ober 2001 to December 2003, modeled leaf fall totaled 12,200 kg/ha, which corresponds to 5,400 kg/ha/yr. As depicted in the figure,

PAGE 87

75 Figure 4-3. Observed canopy cover and T KN concentrations each month for the Bonham-South watershed. Figure 4-4. Monthly simulated TKN concen tration and simulated leaf mass in the Bonham-South’s riparian area. simulated leaf fall began as early as J une each year and completed by the end of December-January. As observed in the ripa rian area of the watershed, leaves were present in the canopy year-round. For example, during the 2001-2002 dormant season,

PAGE 88

76 approximately 45% of the canopy was still cove red with leaves. Simulated N present in the live vegetation showed an inverse relation with the simulated TKN concentrations in the stream (Figure 4-5). Figure 4-5. Monthly simulated TKN concentration and simulate d nitrogen present in live leaves the Bonham-South’s riparian canopy. A careful analysis of model simulati on results showed that a daily total precipitation of 6 mm or more produced surface r unoff in the riparian area. This surface runoff is responsible for leaching and transporting TKN from the leaf li tter in the riparian zone to the stream. Due to the irregular nature of surface runoff generation, the TKN variability during such events, the frequencie s of observed data, and REMM’s ability to simulate on a daily basis, storm events lasti ng 24 hours or longer were selected for further study. The masses of TKN were calculated for tw o different scenarios. First, total mass of TKN during an event was calculated for th e total streamflow. Second, baseflow was separated from the total flow, and TKN mass corresponding to surface runoff was determined. The observed and simulated T KN masses during indivi dual storm events

PAGE 89

77 were compared for both the scenarios. Out of the 16 storm samples collected in the study area, 6 storms between October 2002 and Ma y 2003 were suitable for the comparison with the daily output from REMM (Figure 4-6) The simulated TKN masses produced in surface runoff during the storms were comparab le to the observed values (Figure 4-7). MBD and MAD between the observed and the simulated TKN masses were 8% and 23%, respectively. As the majority of the flow in the stream is subsurface flow, the effect of subsurface flow in carrying the TKN mass was also analyzed. The result showed that MBD and MAD between observed and simula ted total TKN masses were 17% and 24%, respectively (Figure 4-8). Sensitivity Analysis The changes in streamflow and TKN out puts for the –10% to +10% parameter changes ranged from 0 to 1% and 0 to 7%, respectively (Table 4-4). Changes in soil porosity did not have significant effects on th e streamflow and TKN. A decrease in soil porosity of 10% led to a total streamflow increase of only 0.5%, wh ereas the increase in TKN was only 0.6%. Increase in clay c ontent decreased the TKN output by 0.7%. Increase in saturated hydraul ic conductivity by 10% did not affect the streamflow but increased TKN by 7.2%. Increasing soil layer thickness slightly redu ced the streamflow, but reduced TKN by 5.4%.

PAGE 90

78 Figure 4-6. Observed TKN concentrations dur ing the events. Vertical bars represent precipitation, solid lines represent th e streamflow, and the hollow circles represent the TKN concentrations.

PAGE 91

79 Figure 4-7. Comparison of the observed and simulated TKN masses during the events. These masses represent the contribution from the surface runoff only. Figure 4-8. Comparison of the observed and simulated TKN masses during the events. These masses represent the contribution from the total flow.

PAGE 92

80 Table 4-4. Sensitivity of modeled streamfl ow and TKN based on +/10% change in model parameters for Bonham-South watershed. Percentage change Parameters Stream flow TKN Canopy cover (+) -0.9 2.3 Canopy cover (-) 0.9 -0.2 Soil layer thickness (+) -0.1 -5.4 Soil layer thickness (-) 0.1 5.7 Denitrification rate (+) 0.0 0.0 Denitrification rate (-) 0.0 0.0 Carbon decay rate (+) 0.0 0.9 Carbon decay rate (-) 0.0 -0.5 Saturated hydraulic conductivity (+) 0.0 7.2 Saturated hydraulic c onductivity (-) 0.0 -0.2 Soil porosity (+) -0.4 -1.5 Soil porosity (-) 0.5 0.6 Clay content (+) 0.0 -0.7 Clay content (-) 0.0 0.4 Discussion Over the four years, the simulated average annual surface runoff that is generated in the riparian area and contribu ted to the streamflow was less than 1% of the annual observed precipitation (Table 4-3). The simulated surface runoff generated in the riparian buffer in the study watershed is sm all compared to earlier studies done in a similar watershed in Georgia. A study conducted by Shirmohammadi et al. (1984) in Little River watershed in Georgia reported approximately 4-12% contribution of annual precipitation to surface runoff. Inamdar et al. (1999a) reported the average annual surface runoff contribution to streamflow was approximately 8% of the annual precipitation for Gibbs Farm site within th e Little River watershed. Surface runoff contribution to streamflow in our study wate rshed may not be significant in terms of

PAGE 93

81 hydrology, but the modeled results suggest that it is capable of transpor ting nitrogen from the litter and top soil layer to the stream (Figure 4-7). Subsurface contribution to the streamflow for the Bonham-South watershe d was approximately 20% of the average annual precipitation (Table 4-3). The simula ted subsurface contribution in this study is close to the range of 14-22% reported by Shirmohammadi et al. (1984) for Little River watershed. The riparian forest community f ound along Bonham-South watershed is representative of the Sout heastern riparian forest (Shure and Gottaschalk, 1985; Lowrance et al., 2000). Slash pine ( P. elliotii ), longleaf pine ( P. palustris ), black gum ( N. sylvatica ), sweet gum ( L. styraciflua ), scrub oak ( Q. rubra ), sweetbay ( M. virginiana ), water oak ( Q. nigra ), willow oak ( Q. phellos ), red maple ( A. rubrum ), and swamp tupelo ( N. aquatica ) are often dominant canopy species in these riparian forests. The modeled total leaf fall in the study watershed from October 2001 to December 2003 was approximately 12,200 kg/ha, which corres ponds to 5,400 kg/ha/yr. The modeled average litter mass produced during the same period was approximately 7,000 kg/ha/yr. Approximately 77% of the litter mass were le aves. Contribution of the modeled leaf fall to the total litter pr oduction falls in the ne ighborhood of 80% reported by Meentemeyer et al. (1982), and within the range of 72-84% reported by Shure and Gottaschalk (1985). The simulated litter mass produced in th e study area is higher than the average 5,000 kg/ha for a mixed hardwood forest in the No rtheastern U.S. (Gosz et al., 1972) and is comparable with the results found in the Sout heastern U.S. Mul holland (1981) reported litter production of 6,100 kg/ha in a small str eam swamp in eastern North Carolina. A bottomland hardwood forest and a cypress-t upelo stand located in Louisiana produced

PAGE 94

82 litter masses of 5,750 kg/ha and 6,200 kg/ha, respectively (Conner and Day Jr., 1976). Brinson et al. (1980) reported a value of 6,500 kg/ha for an alluvial swamp forest in North Carolina. Simulated results showed that the nitroge n content in the li tter during the study period averaged 77 kg/ha of which approximate ly 59 kg/ha was the contribution from the leaves alone. This value is less than 83.4 kg/ha of nitrogen contribution from the deciduous leaves in the Northeastern U.S. (Gos z et al., 1972). As observed in the riparian area, 45% canopy was present even during the dormant seasons because of the presence of approximately 19% of conifers and 21% of mixed species. Modeled results also support the presence of the carbon in the leaves and the biomass in the canopy. The modeled result showed that on an average approximately 527 kg/ha of carbon was present in the canopy during the dormant season. This value of carbon suggests the presence of approximately 22 kg/ha of n itrogen in the canopy during the same time (Figure 4-5). Higher concentr ations of TKN during the leaf fall and the inverse relation with the nitrogen present in the canopy support the hypothesis of flushing of nitrogen from the leaf litter. The simulated TKN concentrations follow the trend of leaf mass accumulation in the riparian area suggesting a higher rate of release of TKN from freshly fallen litter during the fall and early spring. Brinson et al. (1980) reported higher nutrient fluxes following litterfall in an alluvial swamp forest in North Carolina. In a study of nutrient content of litterfall in the Southeastern U.S., Gosz et al. (1972) reported that approximately 56% of the total nutrien t was released during September through December when the majority of litterfall ( 50%) occurred. A study conducted in cypress

PAGE 95

83 swamp forest in Florida, Schlesinger (1978) reported that in th e month of November when 56% litterfall occurred, 45% of nitr ogen was released from the system. Higher TKN concentrations were observed in the Bonham-South stream during the precipitation events (Figure 46). The storm events presen ted in Figure 4-6 were long enough to calculate the masses of TKN during the storms to compare REMM’s simulated daily values. Comparisons of TKN masses pr oduced during the events were comparable to REMM’s simulated values. The masses of TKN during the storms were calculated for two different scenarios. When modeled TKN mass contribution of surface runoff that is produced in the riparian zone was compared with the observed value in the stream, MBD and MAD for the six storms were 8% and 23%, respectively. Mean biased difference value is less reliable as there is the risk th at large outliers cancel each other out. As MAD takes negative values and repl aces them with their absolute values, large outliers are deemphasized; hence it is less sensitive to extr eme values. A MAD of 23% in the surface runoff suggests that the model was able to produce comparable amount of TKN as observed during the storm. The comparis on between observed and simulated TKN mass from the total flow showed comparable MB D (17%) and MAD (24%). On average, 43% of the observed total TKN mass was contributed by surface runoff. This result suggests that the surface runoff during the precipitat ion events is a major source of leaching nitrogen from the forest floor and transporting it to the stream. Conclusion and Recommendations Originally, REMM was developed to operate at a hi llslope scale. The results in this study suggested that given appropriate upla nd inputs for a site, REMM can be used at different scenarios of riparian area width, length of zones, vegetation type, and soil properties. The trend and ma gnitude of the observed streamflow for the study watershed

PAGE 96

84 was effectively simulated by REMM. The model, however, overestimated the streamflow during high flow and underest imated during low flow periods. The hydrologic budget showed a good agreemen t between observed and predicted streamflow. The simulated litter and corre sponding leaf masses in the study watershed were comparable to the values reported in the lite rature. Due to the model’s longer simulation time step, a meaningful comparison of simu lated and observed TKN was possible only through the masses produced during storms that were equal to or longer than a day. Comparison of TKN masses during six different storm events showed similar values both in the surface runoff and the total flow. Th e results supported the hypothesis of nitrogen leaching from freshly fallen leaves during the precipitation events. These results provided further insight into the nutrient dynamics of the riparian area. The model simulations respond as expected to precip itation and vegetation pa tterns over the study period. Results clearly indicated that the presen ce of fresh leaf litter in the riparian area increases the TKN concentrati on, and hence mass, in the stream. The model effectively captured the trends of leaf mass accumulation in the riparian area and subsequent high concentration and mass during those periods. With the present ve rsion of REMM, the comparison between observed and simulated va lues is possible only on a daily basis. However, given the fact that the precipitation events and the sampling frequencies are often shorter, simulation results using a smalle r time scale would be useful. Therefore, it is recommended to modify REMM from its presen t version of daily time step to a smaller time step, preferably hourly, for more eff ective and meaningful interpretation and evaluation of riparian nutrient flushing.

PAGE 97

85 CHAPTER 5 SUMMARY AND CONCLUSION The National Research Council of the Unite d States has proposed that reliable and comprehensive environmental indicators be developed to monitor ecological changes from natural and anthropogenic causes. The military installation in Fort Benning, Georgia offered a unique opportunity to study military impacts on water quality and quantity. Ecohydrological approaches were used to relate the effects of anthropogenic perturbations on water-vegetationnutrient interactions in th e Fort Benning watersheds. In this research, statisti cal relationships among water quality parameters and the watershed physical characteristics in low-nutrient watersheds were identified and examined. This research also identifi ed and developed hydrologic indices that characterize the impact of military land mana gement on watershed, and investigated the riparian corridor’s role on water quality by quantifying the nitrogen leaching from freshly fallen leaves. Relationships among watershed physical characteristics and water quality parameters in the study watersheds and the re gression analysis showed that pH, chloride, total phosphorus, total Kjeldahl nitrogen, tota l organic carbon, and total suspended solids are useful indicators of wate rshed physical characteristics that are susceptible to perturbations. A comparison of the results between a reference and impacted watershed in terms of hydrologic indices that are derived from long-term daily flow as well as the storm-based data showed a clear distinct ion between the watersheds in terms of hydrological flow regimes. Storm-based ma gnitude of baseflow index, magnitude and

PAGE 98

86 variability of peak discharge, and the frequency of bankfu ll discharge were consistent with the results from annual-based analysis Results showed that these storm-based indices might be used as surrogates to the indices derived from long-term data. The analysis identified the relationships between the extent of military training land, road density, and the number of road s crossing streams with the storm-based baseflow index, bankfull discharge, response la g, and time of rise. For the low nutrient systems of this study, seasonal and storm variations in water quality were found to be strongly influenced by precipitation events which caused nitrogen leaching from recently fallen leaf litter and increased nitrogen levels in the stream water. This study showed that the signatures of military alterations to watershed landscape are detectable in the water qual ity and the flow regime. The observed alterations to these regimes suggest impacts to aquatic ecosystems The water quality related indicators and hydrologic indices presented in this research provide specific measur es of stream water quality and instream flow, respectively, that respond to military impacts. Each indicator is linked to one or more readily measured wa tershed scale factors. The indicators could be useful in predicting effects of military land management and evaluating restoration activities on the quality as well as the quantity of the water. This research has several implications fo r the Fort Benning military installation. The results of this study indi cated that baseflow sampling of water quality can be used to assess the military training impacts on stream water quality. The water quality indicators identified in this study provide measures of the current watershed conditions. The water quality indicators identified in this stu dy provide measures of the current watershed conditions. Changes to stream water quality due to military training within the Fort

PAGE 99

87 Benning military installation can be identified by collecting long-term and routine stream water quality data and comparing indicator valu es to current conditions Future activities should develop indicator thresholds beyond wh ich improved management and restoration practices should be implemented. As nutri ents are one major factor controlling the quality of the receiving aqua tic ecosystems, the thresholds should be based on the ecosystems impacts. The U.S. Army manages approximately 4.8 106 hectares of land for military training, and it has developed the Land Cond ition Trend Analysis (LCTA) program to systematically monitor terrestrial impact s from military training and to support the mitigation and remediation of severely imp acted training lands (Quist et al., 2003). Although the magnitude of the training would diffe r from one installation to the other, the nature of the impact remains similar. Military training typically increases soil bulk densities and compaction, decrea ses infiltration, diminishes pl ant growth, degrades water quality, and affects water quantit y. Therefore, the research findings in this study can be implemented for management and restoration practices to other military installations across the U.S. to minimize the impacts of the military training. A common finding between the water quality and quantity signature s is that military training impacts differ from those found due to agri culture and urbanization. In the former, water quality degradation impacts typically manifest as high nutrient loads. However, in military installations, the loss of topsoil and vegetation results in waters that have significantly lower nutrient levels. In a similar fashion, the hydrologic regimes in watersheds with significant military impacts suffer from a lo ss of variability, but exhibit higher annual discharge values.

PAGE 100

88 APPENDIX A ANNUAL-BASED INDICES DEFINITIONS AND CALCULATION PROCEDURES Symbol Definition Method MA26 Coefficient of variation in monthly flows for December Calculate mean monthly flow for December (diq ), i = 1,…, n ; n = no. of years Calculate mean of all December flows, DQ=n qn i di/1 Calculate standard deviation ( SD ) of diq Calculate coefficient of variation, CV = SD /DQ MA41 Mean annual flow divided by watershed area Calculate mean yearly flow (yiq), i = 1,…, n; n = no. of years Divide yiqby watershed area (A) Calculate n A q A Qn i yi y/ /1 i = 1,…, n; n = no. of years MA10 Ranges in daily flows divided by median daily flows (where range in daily flows is the ratio of 20th/80th percentiles in daily flows across all years) Combine daily flows of all years Calculate 20th percentile (20pQ) Calculate 80th percentile (80 pQ) Calculate median flow (mQ) Calculate R = 20pQ/80pQ Calculate R / mQ ML17 Seven-day minimum flow divided by mean annual daily flows Calculate mean yearly flow (yiq), i = 1,…, n; n = no. of years Calculate 7-day minimum (min 7q)i flow Calculate ratios Ri = (min 7q)i /yiq Calculate n Rn i i/1

PAGE 101

89 ML14 Mean of the lowest annual daily flow divided by median annual daily flow averaged across all years Extract lowest yearly flow (min ,iq), i = 1,…, n; n = no. of years Calculate n q Qn i i/1 min min Calculate median flow (mQ ) of all years Calculate mQ Q/min ML16 Median of the lowest annual daily flows divided by median annual daily flows averaged across all years Extract lowest yearly flow (min ,iq ), i = 1,…, n ; n = no. of years Calculate median (medQ) of min iq Calculate median flow (mQ ) of all years Calculate m medQ Q/ MH23 Mean of the high flow volume (calculated as the area between the hydrograph and the upper threshold during the high flow event) divided by median annual daily flow across all years. The upper threshold is defined as 7 times median annual flow Calculate yearly median flow (med iq,), i = 1,…, n ; n = no. of years Calculate 7 times med iq, Calculate the flow volume (ViQ ) above 7 times med iq, value Calculate n Q Qn i Vi V/1 Calculate median flow (mQ ) of all years Calculate m VQ Q/ MH8 Mean of the maximum monthly flows for the month of May Extract maximum flow for May (miq ), i = 1,…, n ; n = no. of years Calculate mean of all May maximum flows, MayQ= n qn i mi/1 MH14 Median of the highest annual daily flow divided by the median annual daily flow averaged across all years Calculate median flow (miQ ), i = 1,…, n ; n = no. of years Calculate n Q Qn i mi m/1 Extract highest daily flow (hiQ ) for each year Calculate median of highest daily flow (medhQ ) Calculate medhQ /mQ

PAGE 102

90 FL3 Total number of low flow spells (threshold equal to 5% of mean daily flow) divided by the record length in years Combine daily flows of all years Calculate mean daily flow (meanQ ) Determine the threshold of 0.05*meanQ Determine the numbers of low flow counts below 0.05*meanQ (lowN ) Calculate lowN /n, n = record length in years FL2 Coefficient of variation in Fl1, where Fl1 is low flood pulse count, which is the number of annual occurrences during which the magnitude of flow remains below a lower threshold. Hydrologic pulses are defined as those periods within a year in which the flow drops below the 25th percentile (low pulse) of all daily values for the time period Calculate 25th percentile value for each year (i pQ25), i = 1,…, n ; n = no. of years Determine the numbers of low flow pulses below 25th percentile for each year (i lN25) Calculate mean of low flow pulses, n N Nn i i l l/1 25 25 Calculate standard deviation ( SD ) of i lN25 Calculate CV = SD /25 lN FH4 High flood pulse count is the number of annual occurrences during which the magnitude of flow remains above an upper threshold where the upper threshold is defined as 7 times median daily flow, and the value is represented as an average instead of a tabulated count Calculate yearly median flow (med iq,), i = 1,…, n; n = no. of years Calculate 7 times med iq, Determine the number of high flood pulses above 7 times med iq, (hiN7) Calculate n N Nn i hi h/1 7 7 FH6 Mean number of high flow events per year using an upper threshold of 3 times median flow over all years Calculate yearly median flow (med iq,), i = 1,…, n; n = no. of years Calculate 3 times med iq, Determine the number of high flood pulses above 3 times med iq, (hiN3) Calculate n N Nn i hi h/1 3 3

PAGE 103

91 FH7 Mean number of high flow events per year using an upper threshold of 7 times median flow over all years Same as FH4 DL10 Coefficient of variation in annual minima of 90day means of daily discharge Determine 90-day minimum flow (iQmin 90) for each year, i = 1,…, n; n = no. of years Calculate n Q Qn i i/1 min 90 min 90 Calculate SD of iQmin 90 Calculate CV = min 90/Q SD DL17 Coefficient of variation in low flow pulse durations Calculate 25th percentile value for each year (i pQ25),i = 1,…, n; n = no. of years Determine duration between low flow pulses below 25th percentile for each year (i lD25) Calculate mean duration of low flow pulses, n D Dn i i l l/1 25 25 Calculate standard deviation (SD) of i lD25 Calculate CV = SD/25 lD DL6 Coefficient of variation in annual minima of 1day means of daily discharge Determine 1-day minimum flow (iQmin 1) for each year, i = 1,…, n; n = no. of years Calculate n Q Qn i i/1 min 1 min 1 Calculate SD of iQmin 1 Calculate CV = min 1/Q SD DH13 Mean annual 30-day maximum divided by median flow Determine 30-day maximum flow (iQmax 30) for each year, i = 1,…, n; n = no. of years Calculate n Q Qn i i/1 max 30 max 30 Calculate median flow (miQ),i = 1,…, n; n = no. of years

PAGE 104

92 Calculate n Q Qn i mi m/1 Determine mQ Q /max 30 DH16 Coefficient of variation in high flow pulse durations Calculate 75th percentile value for each year (i pQ75), i = 1,…, n; n = no. of years Determine duration between high flow pulses above 75th percentile for each year (i hD75) Calculate mean duration of high flow pulses, n D Dn i i h h/1 75 75 Calculate standard deviation (SD) of i hD75 Calculate CV = SD/75 hD DH24 Mean annual maximum number of 365 days over all water years during which no floods occurred over all years Determine flow of magnit ude exceeding a return interval of 1.67 years based on log-normal distribution (yriQ67 1), i = 1,…, n; n = no. of years Determine the maximum number of 1.67-year flow non-exceedance days each year (i yrDmax 67 1) Calculate n D Dn i i yr yr/1 max 67 1 max 67 1 TA1 Constancy [Colwell (1974)] Standardize the daily flow values by median flow, and express as natural logarithm values Divide the log-transforme d flow values into 4-6 classes so that these classes cover all the observed flow values Create a matrix of total number of days in a year (columns, t) by number of classes (rows, s) Define the column totals (jX ), row totals (iY), and the grand total (Z) as s i ij jN X1, t j ij iN Y1, and i i j j j ij iY X N Z where Nij is the number of cycles for which the flow is in class i and time j. Determine uncertainty with respect to time t j j jZ X Z X X H1log ) (

PAGE 105

93 Determine uncertainty with respect to class s i i iZ Y Z Y Y H1log ) ( Determine uncertainty with respect to the interaction of time and scale j ij ij iZ N Z N XY Hlog ) ( Determine predictability (P) with the range (0, 1) s X H XY H Plog ) ( ) ( 1 where s is the number of classes Determine the constancy ) ( C with range (0, 1) s Y H C log ) ( 1 TA3 Maximum proportion of all floods over the period of record that fall in any one of six 60-day ‘seasonal’ windows Determine flow of magnit ude exceeding a return interval of 1.67 years based on log-normal distribution (yrQ67 1) over the period of record Divide every year into six 60-day ‘seasonal’ windows Determine the maximum number of yrQ67 1exceedance days (max 67 1 yrD ) over period of record in any one of six 60-day ‘seasonal’ windows Determine the total number of yrQ67 1exceedance days over all years in one of six 60-day seasonal window (yrtotD67 1) Calculate the ratio (max 67 1 yrD/yrtotD67 1) TH3 Maximum proportion of the year (number of days/365) during which no floods have ever occurred over the period of record Determine flow of magnit ude exceeding a return interval of 1.67 years based on log-normal distribution (yrQ67 1) over the period of record Determine the maximum number of yrQ67 1nonexceedance days (yrnonD67 1) over period of record Determine the total number of yrQ67 1nonexceedance days over period of record (yrtotD67 1) Calculate the ratio (yrnonD67 1/yrtotD67 1)

PAGE 106

94 RA9 Coefficient of variation in number of negative and positive changes in water conditions from one day to the next Determine number of rises from one day to the other each year (RiN ), i = 1,…, n ; n = no. of years Determine number of falls each year (FiN ) Calculate n N Nn i Ri R/1 Calculate SDR of RiN Calculate n N Nn i Fi F/1 Calculate SDF of FiN Calculate CVR = SDR /RN Calculate CVF = SDF /FN RA7 Median of difference between natural logarithm of flows between two consecutive days with decreasing flow Transform the daily flow values for each year using natural logarithm, Qi = ln Qi, i = no. of days For all ln Qs each year, determine [ln Qi – ln Qi+1] Separate the values of decreasing flow differences for each year Determine the median of decreasing flow differences each year (dmjQ ), j = 1, …, n; n = no. of years Calculate n Q Qn j dmj dm/1 RA6 Median of difference between natural logarithm of flows between two consecutive days with increasing flow Transform the daily flow values for each year using natural logarithm, Qi = ln Qi, i = no. of days For all ln Qs each year, determine [ln Qi – ln Qi+1] Separate the values of increasing flow differences for each year Determine the median of increasing flow differences each year (imjQ ), j = 1, …, n; n = no. of years Calculate n Q Qn j imj im/1

PAGE 107

95 APPENDIX B STORM-BASED INDICES DEFINITIONS AND CALCULATION PROCEDURES Index Symbol Definition Methods MMRF Mean value of the response factor Calculate precipitation depth for each event (iP ), i = 1,…, n ; n = no. of events Create hydrograph Separate baseflow Calculate DRO depth after de ducting the baseflow portion from the hydrograph for each event (iD ) Calculate response factor for each event, i i iP D RF / Calculate the mean, n RF RFn i i/1 MVRF Coefficient of variation in response factors Calculate RF Calculate standard deviation ( SD ) of RFi Calculate _/RF SD CV MMBF Mean value of baseflow index Create hydrograph Separate baseflow Determine baseflow volume (Vb) and total volume (Vt) for each events Calculate BFi = (Vb/Vt,)i, i = 1,…,n; n = no. of events Calculate the mean, n BF BFn i i/1 MVBF Coefficient of variation in baseflow index Calculate BF Calculate (SD) of BFi Calculate _/ BF SD CV MMPD Mean value of peak discharges divided by the watershed area Determine ( qpk)i of the event, i = 1,…, n ; n = no. of events Normalize ( qpk)i by the watershed area, A qi pk/ ) ( Calculate mean, _/ A qpk = n i i pkn A q1/

PAGE 108

96 MVPD Coefficient of variation in peak discharges Calculate _/ A qpk Calculate (SD) of A qi pk/ ) ( Calculate CV = SD / _/ A qpk F1FD Percentage of peak discharge equals bankfull discharge Determine bankfull discharge Count peak discharges equal to bankfull discharge for all events and express the count as a percentage of total events F2FD Percentage of peak discharge 2 times above bankfull discharge Determine bankfull discharge Count peak discharges e qual to 2.0 times bankfull discharge for all events a nd express the count as a percentage of total events DMTB Mean value of time base divided by the watershed response time Determine time base for each hydrograph Normalize the time base by the response time (Tbi), i = 1,…,n; n = no. of events Calculate mean, n T Tn i bi b/1 DVTB Coefficient of variation in time base Calculate bT Calculate (SD) of (Tbi) Calculate CV = SD/_ bT DMRL Mean value of response lag divided by the watershed response time Determine response lag for each hydrograph Normalize the response lag by the response time (Trli), i = 1,…,n; n = no. of events Calculate mean, n T Tn i rli rl/1 DVRL Coefficient of variation in response lag Calculate rlT Calculate (SD) of (Trli) Calculate CV = SD/_ rlT DMTR Mean value of time of rise divided by the watershed response time Determine time of rise for each hydrograph Normalize the time of rise by the response time (Tri), i = 1,…,n; n = no. of events Calculate mean, n T Tn i ri r/1

PAGE 109

97 DVTR Coefficient of variation in time of rise Calculate rT Calculate (SD) of (Tri) Calculate CV = SD/_ rT RPPD Mean rate of change in peak discharge in rising limb Determine (qpk)i of the event, i = 1,…,n; n = no. of events Normalize (qpk)i by the watershed area, A qi pk/ ) ( Normalize the time of rise by the response time (Tri), i = 1,…,n; n = no. of events Calculate the ratio of A qi pk/ ) ( to (Tri) Calculate the mean of the ratios RPVPD Coefficient of variation in the rate of change in peak discharge in rising limb Calculate mean of RPPDi Calculate (SD) of RPPDi Calculate CV = SD/mean of RPPDi RNPD Mean rate of change in peak discharge in falling limb Determine (qpk)i of the event, i = 1,…,n; n = no. of events Normalize (qpk)i by the watershed area, A qi pk/ ) ( Normalize the time of rise by the response time (Tri), i = 1,…,n; n = no. of events Normalize the time base by the response time (Tbi) Calculate the difference (Tbi Tri) Calculate the ratio of A qi pk/ ) ( to (Tbi Tri) Calculate the mean of the ratios RNVPD Coefficient of variation in the rate of change in peak discharge in falling limb Calculate mean of RNPDi Calculate (SD) of RNPDi Calculate CV = SD/mean of RNPDi

PAGE 110

98 APPENDIX C NITROGEN TRANSFORMATIONS IN REMM

PAGE 111

99 APPENDIX D PARAMETERS AND RATE CONSTANTS USED IN REMM Parameters Units Values Litter layer parameters (all zones) Layer depth cm 1 Evaporation factor unitless 4 Evaporation constant unitless -0.45 Litter moisture mm 2 Litter bulk density g/cm3 1 Ammonium adsorption coefficient (a) unitless 1.102 Ammonium adsorption coefficient (b) unitless 0.956 Litter pH unitless 6.2 Litter C structural pool kg/ha 2613 Litter C metabolic pool kg/ha 436 Litter C active pool kg/ha 5 Litter C slow pool kg/ha 3000 Litter C passive pool kg/ha 1900 Litter C lignin kg/ha 9100 Litter ammonium pool kg/ha 1.34 Litter nitrate pool kg/ha 0.24 Litter N structural pool kg/ha 86 Litter N metabolic pool kg/ha 3 Litter N active pool kg/ha 1 Litter N slow pool kg/ha 273 Litter N passive pool kg/ha 173

PAGE 112

100 Parameters Units Values Layer 1 Layer 2 Layer 3 Soil layer parameters (all zones) Pore size distribution index unitless 0.38 0.38 0.38 Bubbling pressure head cm 1.8 1.8 5.6 Starting moisture content cm/cm 0.20 0.31 0.35 Saturated conductivity cm/hr 15 11 1 Sand content % 75 60 51 Silt content % 16 9 12 Clay content % 9 31 38 Bulk density g/cm3 1.4 1.5 1.5 Start carbon structural pool kg/ha 1114 135 135 Start carbon metabolic pool kg/ha 1787 215 215 Start carbon active pool kg/ha 400 400 200 Start carbon slow pool kg/ha 12000 12000 6000 Start carbon passive pool kg/ha 7600 7600 3800 Start carbon lignin pool kg/ha 3360 1680 560 Start nitrogen ammonium pool kg/ha 0.001 0.001 0.001 Start nitrogen nitrate pool kg/ha 0.001 0.001 0.001 Start nitrogen structur al pool kg/ha 1 1 1 Start nitrogen metabolic pool kg/ha 81 10 10 Start nitrogen active pool kg/ha 16 8 8 Start nitrogen slow pool kg/ha 303 157 157 Start nitrogen passive pool kg/ha 191 81 81

PAGE 113

101 Rate constants Values Carbon release rates (litter and soil layers) Metabolic residue pool 0.60 Structural residue pool 0.15 Active humus pool 0.02 Slow humus pool 0.005 Passive humus pool 0.00002 Denitrification rates Litter 0.02 Soil layer 1 0.02 Soil layer 2 0.01 Soil layer 3 0.002

PAGE 114

102 LIST OF REFERENCES Altier, L.S., Lowrance, R.R., Williams, R.G., In amdar, S.P., Bosch, D.D., Sheridan, J.M., Hubbard, R.K., Thomas, D.L. 2002. Riparian Ecosystem Management Model (REMM): simulator for ecological processe s in riparian zones. United States Department of Agriculture, Agricultural Research Service, Conservation Research Report 46. American Public Health Association. 1992. Standard Methods for the Examination of Water and Wastewater, 20th ed ition. Washington, D.C. Auble, G.T., Friedman, J.M., Scott, M.L. 1994. Relating riparian vegetation to present and future streamflows. Ecol ogical Applications 4, 544-554. Band, L.E., Tague, C.L., Groffman, P., Belt, K. 2001. Forest ecosystem processes at the watershed scale: hydrological and ecological controls of nitrogen export. Hydrological Processes 15: 2013-2028. Basnyat, P., Teeter, L., Flynn, K., Lockaby, B. 1999. Relationships between landscape characteristics and nonpoint source pollu tion inputs to coas tal estuaries. Environmental Management 23: 539-549. Berg, B., Staaf, H. 1981. Leaching, accumulation and release of nitrogen in decomposing litter. In Clark, F.E., Rossw all, T. (editors), Terrestrial Nitrogen Cycles. Ecological Bulle tins (Stockholm) 33: 163-178. Bernal, S., Butturini, A., Nin, E., Sabater, F., Sabater, S. 2003. Leaf litter dynamics and nitrous oxide emission in a mediterranean riparian forest: implications for soil nitrogen dynamics. Journal of Environmental Quality 32: 191-197. Beschta, R.L., Jackson, W.L. 1979. The intrus ion of fine sediments into a stable gravel bed. Journal of the Fisheries Re search Board of Canada 36: 207-210. Bhat S., Jacobs, J.M. (Department of Civil Engineering, University of New Hampshire), Hatfield, K. (Department of Civil and Coastal Engineering, University of Florida), Prenger, J. Ecological i ndicators in forested waters heds in Fort Benning, GA: relationship between land use and stream water quality. Ecol ogical Indicators. Accepted pending revision. Biggs, B.J.F. 1996. Hydraulic habitat of plan ts in streams. Regulated Rivers: Research and Management 12: 131-144.

PAGE 115

103 Biggs, B.J.F. 2000. Eutrophication of stream s and rivers: dissolved nutrient-chlorophyll relationships for benthic algae. Jour nal of the North American Benthological Society 19: 17-31. Biggs, B.J.F., Close, M.E. 1989. Periphyton biomass dynamics in gravel bed rivers: the relative effects of flow s and nutrients. Freshwater Biology 22: 209-231. Biggs, B.J.F., Smith, R.A., Duncan, M.J. 1999. Velocity and sediment disturbance of periphyton in headwater streams: biomass and metabolism. Journal of the North American Benthological Society 18: 144-163. Biggs, B.J.F., Stokseth, S. 1996. Hydraulic ha bitat preferences for pe riphyton in rivers. Regulated Rivers: Research and Management 12: 251-261. Binkley, D. 1986. Forest Nutrition Ma nagement. John Wiley & Sons, New York. Bocock, K.L., Gilbert, O., Capstick, C.K., Twinn, D.C., Waid, J.S., Woodman, M.J. 1960. Changes in leaf litter when placed on the surface of soils with contrasting humus types. I. Losses in dry weight of oak and ash leaf litte r. Journal of Soil Science 11:1-9. Bolstad, P.V., Swank, W.T. 1997. Cumulativ e impacts of land use on water quality in a southern Appalachian watershed. Jo urnal of American Water Resources Association 33: 519-533. Bosch, D.D., Sheridan, J.M., Lowrance, R.R. 1996. Hydraulic gradients and flow rates of a shallow coastal plain aquifer in a forest ed riparian buffer. Transactions of the American Society of Agricu ltural Engineers 39: 865-871. Bren, L.J. 1992. Tree invasion of an interm ittent wetland in relation to changes in the flooding frequency of the River Murray, Aust ralia. Australian Journal of Ecology 17: 395-408. Brinson, M.M., Bradshaw, H.D., Holmes, R. N., Elkins, Jr., J.B. 1980. Litterfall, stemflow, and throughfall nutrient fluxes in an alluvial swamp forest. Ecology 61: 827-835. Bryant, M.L., Bhat, S., Jacobs, J.M. In Press. Characterization and modeling of throughfall temporal variability for forest communities in the southeastern U.S. Journal of Hydrology. Bubb, K.A., Xu, Z.H., Simpson, J.A., Saffi gna, P.G. 1998. Some nutrient dynamics associated with litterfall and litter decomposition in hoop pine plantations of southeast Queensland, Australia. Fore st Ecology and Management 110: 343-352. Cairns, Jr., J., McCormick, P.V., Neiderlehn er, B.R. 1993. A proposed framework for developing indicators for ecosystem h ealth. Hydrobiologia 263: 1-44.

PAGE 116

104 Chapman, R.J., Hinckley, T.M., Lee, L.C., Te skey, R.O. 1982. Impact of water level changes on woody riparian and wetland commu nities. Vol. 10. Kearneysville (WV), US Fish & Wildlife Servi ce. Publication No. OBS-82/83. Campbell, G.S. 1974. A simple method for determining unsaturated conductivity from moisture retention data. Soil Science 117: 311-314. Christ, M.J., David, M.B. 1996. Dynamics of extractable orga nic carbon in Spodosol forest floors. Soil Biology and Biochemistry 28: 1171-1179. Cirmo, C.P., McDonnell, J.J. 1997. Linki ng the hydrologic and biogeochemical controls of nitrogen transport in near-stream zone s of temperate-forested catchments: a review. Journal of Hydrology 199: 88-120. Clausen, B., Biggs, B.J.F. 1997. Relations hips between benthic biota and hydrological indices in New Zealand streams. Freshwater Biology 38: 327-342. Clausen, B., Biggs, B.J.F. 2000. Flow i ndices for ecological studies in temperate streams: groupings based on covarian ce. Journal of Hydrology 237: 184-197. Cleveland, C.C., Neff, J.C., Townsend, A. R., Hood, E. 2004. Composition, dynamics, and fate of leached dissolved organic matter in terrestrial ecosystems: results from a decomposition experiment. Ecosystems 7: 275-285. Close, M.E., Davies-Colley, R.J. 1990. Base flow chemistry in New Zealand rivers 2. influence of environmental factors. New Zealand Journa l of Marine and Freshwater Research 24: 343-356. Cole, D.W., Rapp, M. 1981. Elemental cycling in forest ecosystems. In Reichle, D.E. (editor). Dynamic Properties of Forest Ecosystems. International Biological Programme 23. Cambridge University Press, London, pp. 341-409. Closs, G.P., Lake, P.S. 1996. Drought, diffe rential mortality and the coexistence of a native and an introduced fish species in a southeast Australian intermittent stream. Environmental Biology of Fishes 47: 17-26. Cobb, D.G., Galoway, T.D., Flannagan, J.F. 1992. Effects of discharge and substrate stability on density and species composition of stream insects. Canadian Journal of Fisheries and Aquatic Sciences 49: 1788-1795. Colwell, R.K. 1974. Predictability, constanc y, and contingency of periodic phenomena. Ecology 5: 1148-1153. Conner, W.H., Day, Jr., J.W. 1976. Productiv ity and composition of a baldcypress-water tupelo site and a bottomland hardwood s ite in a Louisiana swamp. American Journal of Botany 63: 1354-1364.

PAGE 117

105 Crockford, R.H., Richardson, D.P. 1998. Litte rfall, litter and associated chemistry in a dry sclerophyll eucalypt fore st and a pine plantation in southeastern Australia: 1. litterfall and litter. Hydrologi cal Processes 12: 365-384. Cummins, K.W. 1974. Structure and functi on of Stream ecosystems. BioScience 24: 631-641. Currie, W.S., Aber, J.D., McDowell, W.H., B oone, R.D., Magill, A.H. 1996. Vertical transport of dissolved organic C and N unde r long-term N amendments in pine and hardwood forests. Biogeochemistry 35: 471-505. Dale, V.H., Beyeler, S.C. 2001. Challenge s in the development and use of ecological indicators. Ecologica l Indicators 1: 3-10. Dale, V.H., Beyeler, S.C., Jackson, B. 2002. Understory vegetation indicators of anthropogenic disturbance in longleaf pine forests at Fort Benning, Georgia, USA. Ecological Indicators 1: 155-170. Davies-Colley, R.J., Quinn, J.M. 1998. Str eam lighting in five regions of North Island, New Zealand: control by channel size and riparian vegetation. New Zealand Journal of Marine and Fres hwater Research 32: 591-606. Fisher, S.G. 1983. Succession in streams. In Barnes, J.R., Minshall, G.W. (editors), Stream Ecology: Application and Testing of General Ecological Theory. Plenum Press, New York. Fisher, D.S., Steiner, J.L., Endale, D. M., Stuedemann, J.A., Schomberg, H.H., Franzluebbers, A.J., Wilkinson, S.R. 2000. The relationship of land use practices to surface water quality in the Upper Oc onee watershed of Georgia. Forest Ecology and Management 128: 39-48. Fox, D.G. 1981. Judging air quality model performance: a summary of the AMS workshop on dispersion model performance. Bulletin of Americ an Meteorological Society 62: 599-609. Frick, E.A., Hippe, D.J., Buell, G.R., C ouch, C.A., Hopkins, E.H., Wangsness, D.J., Garrett, J.W. 1998. Water quality in the Apalachicola-Chattahoochee-Flint river basin, Georgia, Alabama, and Florida, 1992–95. U.S. Geological Survey Circular1164. Furniss, M.J., Roeloffs, T.D., Yee, C.S. 1991. Road construction and maintenance. In Meehan, W.R. (editor). Influences of forest and rangeland management on salmonid fishes and their habitats. Speci al publication 19. American Fisheries Society, Bethesda, Maryland. pp. 297-323. Garten, Jr., C.T., Ashwood, T.L., Dale, V.H. 2003. Ef fect of military training on indicators of soil quality at Fort Benning, Georgia. Ecological Indicators 3: 171179.

PAGE 118

106 Gjessing, E., Lygren, E., Berglind, L., Gulb randsen, T., Skanne, R. 1984. Effect of highway runoff on lake water quality. Science of the Total Environment 33: 247257. Gosz, J.R., Likens, G.E., Bormann, F.H. 1972. Nutrient content of litter fall on the Hubbard Brook experimental forest. Ecology 53: 769-784. Gosz, J.R., Likens, G.E., Borman. F.H. 1973. Nutrient release from decomposing leaf and branch litter in the Hubbard Broo k forest, New Hampshire. Ecological Monographs 443:173-191. Grimm, N.B., Fisher, S.G. 1992. Responses of arid-land streams to changing climate. In Firth, P. and Fisher, S.G. (editors). Global Climate Change and Freshwater Ecosystems. Springer, New York, pp. 211-233. Hanson, D.L., Waters, T.F. 1974. Recovery of standing crop and production rate of a brook trout population in a flood-damaged st ream. Transactions of American Fisheries Society 103: 431-439. Harrel, R.C., Dorris, T.C. 1968. Stream order, morphometry, physio-chemical conditions, and community structure of benthic microinvertibrates in an interminnent stream system. American Naturalist 80: 221-251. Harris, M.M., Safford, L.O. 1996. Effects of season and four tree species on soluble carbon content in fresh and decomposing litter of temperate forests. Soil Science, 161: 130-135. Heath, G.W., Arnold, M.K., Edwards, C.A. 1966. Studies in leaf litter breakdown. I. Breakdown rates of leaves of differe nt species. Pedobiologia 6: 1-12. Helvey, J.D., Kochenderfer, J.N. 1990. Soil density and moisture content on two unused forest roads during first 30 months after construction. Research paper NE-629. U.S. Forest Service. Northeast Fo rest Experiment Station, Broomhall, Pennsylvania. Hemphill, N., Cooper, S.D. 1983. The eff ect of physical disturbance on the relative abundances of two filter-feed ing insects in a small str eam. Oecologia 58: 378-383. Herbohn, J.L., Congdon, R.A. 1998. Ecosys tem dynamics at disturbed and undisturbed sites in North Queensland wet tropical rain fo rest. III. nutrient returns to the forest floor through litterfall. Journa l of Tropical Ecology 14: 217-229. Hill, A.R. 1981. Streams phosphorus exports from catchments with contrasting land uses in southern Ohio. Water Resources Bulletin 17: 627-634. Hongve, D. 1999. Production of dissolved or ganic carbon in forested catchments. Journal of Hydrology 224: 91-99.

PAGE 119

107 Horwitz, R.J. 1978. Temporal variability patterns and the dist ributional patterns of stream fishes. Ecological Monographs 48: 307-321. Houser, J.N. (Oak Ridge National Laboratory, Oak Ridge, TN), Mulholland, P.J. (Oak Ridge National Laboratory, Oak Ridge, TN), Maloney, K.O. Catchment disturbance and stream metabolism: patte rns in stream respiration and gross primary production along a gradient of upl and soil and vegetation disturbance. Unpublished manuscript. Hunsaker, C.T., Levine, D.A. 1995. Hierar chical approaches to the study of water quality in rivers. BioScience 45: 193-203. Hynes, H.B.N. 1960. The Biology of Pollu ted Waters, Liverpool University Press, Liverpool. Inamdar, S.P., Sheridan, J.M., Williams, R.G., Bosch, D.D., Lowrance, R.R., Altier, L.S., Thomas, D.L. 1999a. Riparian Ecosystem Management Model (REMM): I. testing of the hydrologic com ponent for a Coastal Plain riparian system. American Society of Agricultural Engineers 42: 1679-1689. Inamdar, S.P., Lowrance, R.R., Altier, L. S., Williams, R.G., Hubbard, R.K. 1999b. Riparian Ecosystem Management Model (REM M): II. testing of the water quality and nutrient cycling component for a Coas tal Plain riparian system. American Society of Agricultural Engineers 42: 1691-1707. Iverson, R.M., Hinckley, B.S., Webb, R.M., Hallet, B. 1981. Physical effects of vehicular disturbance on arid la ndscapes. Science 212: 915-917. Johnes, P., Moss, B., Phillips, G. 1996. Th e determination of total nitrogen and total phosphorous concentrations in freshwater s from land use, stock headage and population data: testing a model for us e in conservation and water quality management. Freshwater Biology 36: 451-473. Johnson, L.B., Richards, C., Host, G.E., Art hur, J.W. 1997. Landscape influences on water chemistry in midwestern stream ecosystems. Freshw ater Biology 37: 193208. Jones, J.A., Grant, G.E. 1996. Cumulative effects of forest harvest on peak streamflow in the western cascades of Oregon. Water Resources Research 32: 959-974. Jowett, I.G., Duncan, M.J. 1990. Flow va riability in New Zealand rivers and its relationship to in-stream habitat and biot a. New Zealand Journal of Marine and Freshwater Research 24: 305-317. Junk, W.J., Bayley, P.B., Sparks, R.E. 1989. The flood pulse concept in river-floodplain systems. Canadian Special Publication of Fisheries and Aquatic Sciences 106: 110-127.

PAGE 120

108 Karr, J.R., Chu, E.W. 1999. Restori ng Life in Running Waters: Better Biological Monitoring. Island Press, Washington, D.C. Keller, E.A., Swanson, F.J. 1979. Effects of large organic material on channel form and fluvial processes. Earth Surface Processes and Landforms 4: 351-380. Lemmon, P. E. 1956. A spherical densiomete r for estimating forest overstory density. Forest Science 2: 314-320. Leopold, L.B., Wolman, M.G., Miller, J.P. 1964. Fluvial processes in geomorphology. W.H. Freeman and Sons, San Francisco. Lowrance, R. 1992. Groundwater nitrate and denitrification in a coastal plain riparian forest. Journal of Environmental Quality 21: 401-405. Lowrance, R.R., Altier, L.S., Williams, R.G., Inamdar, S.P., Sheridan, J.M., Bosch, D.D., Hubbard, R.K., Thomas, D.L. 2000. REMM: the riparian ecosystem management model. Journal of Soil and Water Conservation 55: 27-34. Lowrance, R., Todd, R.L., Asmussen, L.E. 1984. Nutrient cycli ng in an agricultural watershed: II. Streamflow and artificia l drainage. Journa l of Environmental Quality 13: 27-32. Maloney, K.O., Mulholland, P.J., Feminella, J.W. In Press. Influence of catchment-scale military land use on physicochemical condi tions in small Southeastern Plains streams (USA). Environmental Management. McCuen, R.H. 1998. Hydrologic Analysis a nd Design. Pearson Prentice Hall, Upper Saddle River, NJ. McDowell, W.H., Fisher, S.G. 1976. Autumn al processing of dissolved organic matter in a small woodland stream ecosystem. Ecology 57: 561-569. McDowell, W.H., Likens, G.E. 1988. Origin composition, and flux of dissolved organic carbon in the Hubbard Brook Valley. Ecological Monographs 58: 177-195. McGill, W.B., Hunt, H.W., Woodmansee, R.G ., Reuss, J.O. 1981. PHOENIX, a model of the dynamics of carbon and nitrogen in gr assland soils. In Clark, F.E., Rosswall, T. (editors). Terrestrial Nitrogen Cycles. Ecological Bulletin (Stockholm) 33: 49115. Meador, M.R., Goldstein, R.M. 2003. A ssessing water quality at large geographic scales: relations among land use, water phys icochemistry, riparian condition, and fish community structure. Envi ronmental Management 31: 504-517. Meentemeyer, V., Box, E.O., Thompson, R. 1982. World patterns and amounts of terrestrial plant litter pro duction. Bioscience 32:125-128.

PAGE 121

109 Meffe, G.K. 1984. Effects of abiotic dist urbance on coexistence of predator and prey fish species. Ecology 65: 1525-1534. Meffe, G.K., Minckley, W.L. 1987. Persiste nce and stability of fi sh and invertebrate assemblages in a repeatedly disturbed S onoran Desert stream. American Midland Naturalist 117: 177-191. Meyer, J.L., Wallace, J.B. Eggert, S.L. 1998. Leaf litter as a source of dissolved organic carbon in streams. Ecosystems 1: 240-249. Michalzik, B., Kalbitz, K., Park, J.H., Soli nger, S., Matzner, E. 2001. Fluxes and concentrations of dissolved organic car bon and nitrogen – a synthesis for temperate forests. Biogeochemistry 52: 173-205. Milchunas, D.G., Schultz, K.A., Shaw, R.B. 1999. Plant community response to disturbance by mechanized military maneuve rs. Journal of Environmental Quality 28: 1533-1547. Minckley, W.L., Deacon, J.E. 1991. Battle Against Extinction: Native Fish Management in the American West. University of Arizona Press, Tucson, AZ. Minckley, W.L., Meffe, G.K. 1987. Diffe rential selection by flooding in stream-fish communities of the arid American Southwes t. In Matthews, W.J., Heins, D.C. (editors). Community and Evolutionary Ecology of North American Stream Fishes. University of Oklahoma Press, Norman, OK. Minshall, G.W. 1984. Aquatic insect-substr atum relationships. In Resh, V.H. and Rosenberg, D.M. (editors). The Ecology of Aquatic Insects. Praeger Scientific, pp. 358-400. Moore, K.M.S., Gregory, S.V. 1988. Res ponse of young-of-the-year cutthroat trout to manipulations of habitat structure in a sma ll stream. Transactions of the American Fisheries Society 117: 162-170. Moss, D., Furse, M.T., Wright, J.F., Armitage, P.D. 1987. The prediction of the macroinvertebrate fauna of unplotted runni ng-water sites in Gr eat Britain using environmental data. Freshwater Biology 17: 41-52. Mulholland, P.J. 1981. Organic carbon flow in a swamp-stream ecosystem. Ecological Monographs 51: 307-322. Nakamura, F., Swanson, F.J., Wondzell, S.M. 2000. Disturbance regimes of stream and riparian systemsa disturbance cascade pe rspective. Hydrological Processes 14: 2849-2860. Nash, J.E., Sutcliffe, V. 1970. River flow forecasting through conceptual models, I. a discussion of principles. J ournal of Hydrology 10: 282-290.

PAGE 122

110 National Research Council (NRC). 2000. Ecolog ical Indicators for the Nation. National Academy Press, Washington, D.C. Natural Resources Conservation Service (NRCS). 1995. Riparian Forest Buffer, 391. Model State Standard and General Speci fications. NRCS Watershed Science Institute. Seattle, WA. Nykvist, N. 1961. Leaching and decomposition of litter. III. experime nts on leaf litter of Betula verrucosa. Oikos 12: 249-263. Ohte, N., Tokuchi, N., Katsuyama, M., Hobara, S., Asano, Y., Koba, K. 2003. Episodic increases in nitrate concentrations in st reamwater due to the partial dieback of a pine forest in Japan: runoff generation pr ocesses control seasona lity. Hydrological Processes 17: 237-249. Olden, J.D., Poff, N.L. 2003. Redundanc y and the choice of hydrologic indices for characterizing streamflow regimes. Rive r Research and Applications 19: 101-121. O’Neill, R.V., Hunsaker, C.T., Jones, K.B., Riitters, K.H., Wickham, J.D., Schwartz, P.M., Goodman, I.A., Jackson, B.L., Baillargeon, W.S. 1997. Monitoring environmental quality at the lands cape scale. BioScience 47: 513-519. Osborne, L., Wiley, M. 1988. Empirical re lationships between land use/cover patterns and stream water quality in an agricultura l watershed. Journal of Environmental Management 26: 9-27. Park, J., Matzner, E. 2003. Controls on the release of dissolv ed organic carbon and nitrogen from a deciduous forest fl oor investigated by manipulations of aboveground litter inputs and water fl ux. Biogeochemistry 66: 265-286. Parton, W.J., Schimel, D.S., Cole, C.V., Oj ima, D.S. 1987. Analysis of factors controlling soil organic matter levels in gr eat plains grasslands Soil Society of America Journal 51: 1173-1179. Pastor, J., Post, W.M. 1986. Influence of c limate, soil moisture, and succession on forest carbon and nitrogen cycles. Biogeochemistry 2: 3-27. Perkins, H.F., Hook, J.E., Barbour, N.W. 1986. Soil characteristics of selected areas of the Coastal Plain experiment stations, University of Georgia. Research Bulletin 346. Perry, C.D., Vellidis, G., Lowrance, R., Thom as, D.L. 1999. Watershed-scale water quality impacts of riparian forest management. Journal of Water Resources Planning and Management 125: 117-125. Pettit, N.E., Froend, R.H., Davies, P.M. 2001. Identifying the natural flow regime and the relationship with ripari an vegetation for two contrasting western Australian rivers. Regulated Rivers: Res earch and Management 17: 201-215.

PAGE 123

111 Petts, G.E. (editor). 1984. Impounded Rivers : Perspectives for Ecological Management. Wiley, Chichester. Piegay, H., Landon, N. 1997. Promoting ecolo gical management of riparian forest on the Drome River, France. Aquatic C onservation: Marine and Freshwater Ecosystems 7: 287-304. Poff, N.L. 1996. A hydrogeogr aphy of unregulated streams in the United States and an examination of scale-dependence in some hydrological descript ors. Freshwater Biology 36: 101-121. Poff, N.L., Allan, J.D. 1995. Functional or ganization of stream fish assemblages in relation to hydrological vari ability. Ecology 76: 606-627. Poff, N.L., Allan, J.D., Bain, M.B., Karr, J.R ., Prestegaard, K.L., Richter, B.D., Sparks, R.E., Stromberg, J.C. 1997. The natu ral flow regime: a paradigm for river conservation and restoration. BioScience 47: 769-784. Poff, N.L., Ward, J.V. 1989. Implications of streamflow variability and predictability for lotic community structure: a regional anal ysis of streamflow patterns. Canadian Journal of Fisheries and A quatic Sciences 46: 1805-1818. Polglase, P.J., Attiwill, P.M. 1992. Nitroge n and phosphorus cycling in relation to stand age of Eucalyptus regnans F. Mu ell. I. return from plant to soil in litterfall. Plant and Soil 142: 157-166. Power, M.E., Stewart, A.J. 1987. Disturba nce and recovery of an algal assemblage following flooding in an Oklahoma stream. American Midland Naturalist 117: 333-345. Qualls, R.G., Haines, B.L., Swank, W.T. 1991. Fluxes of dissolv ed organic nutrients and humic substances in a deci duous forest. Ecology 72: 254-266. Quist, M.C., Fay, P.A., Guy, C.S., Knapp, A.K., Rubenstein, B.N. 2003. Military training effects on terrestrial and aquatic communities on a grassland military installation. Ecological Applications 13: 432-442. Resh, V.H., Brown, A.V., Covich, A.P., Gurt z, M.E., Li, H.W., Mi nshall, G.W., Reice, S.R., Sheldon, A.L., Wallace, J.B., and Wissma r, R. 1988. The role of disturbance in stream ecology. Journal of the Nort h American Benthological Society 7: 433455. Richards, C., Haro, R.J., Johnston, L.B., Ho st, G.E. 1997. Watershed and reach-scale properties as indicators of macroinvertebrate species trai ts. Freshwater Biology 37: 219-230. Richards, C., Host, G.E. 1994. Examining land use influences on stream habitats and macroinvertebrates: a GIS approach. Water Resources Bulletin 30: 729-738.

PAGE 124

112 Richards, C., Johnston, L.B., Host, G.E. 1996. Assessing the influence of landscapescale watershed features on phys ical habitat and stream biot a. Canadian Journal of Fisheries and Aquatic Sciences 53: 295-311. Richter, B.D., Baumgartner, J.V., Powell, J., Braun, D.P. 1996. A method for assessing hydrologic alteration within ecosystems. Conservation Biology 10: 1163-1174. Rogers, H.M. 2002. Litterfall, decomposition and nutrient release in a lowland tropical rain forest, Morobe Province, Papua New Guinea. Journal of Tropical Ecology 18: 449-456. Schiemer, F., Zalewski, M., Thorpe, J.E. (e ditors). 1995. The Importance of AquaticTerrestrial Ecotones for Freshwater Fish. Developments in Hydrobiology. Volume 105. Kluwer, Dordrecht. Schindler, D., Bayley, S. 1993. The biosphe re as an increasing sink for atmospheric carbon: estimates from increasing nitroge n deposition. Global Biogeochemistry Cycles 7: 717-734. Schlesinger, W.H. 1978. Co mmunity structure, dynamics and nutrient cycling in the Okefenoke cypress swamp forest. Ecological Monographs 48: 43-65. Schlosser, I.J. 1985. Flow regime, juvenile abundance, and the a ssemblage structure of stream fishes. Ecology 66: 1484-1490. Scott, M.L., Friedman, J.M., Auble, G.T. 1996. Fluvial processes and the establishment of bottomland trees. Geomorphology 14: 327-339. Semkin, R.G., Jeffries, D.S., Clair, T. A. 1994. Hydrochemical methods and relationships for study of stream output from small catchments. In Molden, B. and Cerny, J. (editors). Biogeochemistry of Small Catchments: A Tool for Environmental Research. John Wiley and Sons, Chichester, pp. 163-188. Sheridan, J.M., Lowrance, R.R., Henry, H.H. 1996. Surface flow sampler for riparian studies. Applied Engineering in Agriculture 12: 183-188. Shirmohammadi, A.J., Sheridan, J.M., Knisel W.G. 1984. An approximate method for partitioning daily streamflow data. Journal of Hydrology 74: 335-354. Shure, D.J., Gottschalk, M.R. 1985. Litte r-fall patterns within a floodplain forest. American Midland Naturalist 114: 98-111. Sidle, R.C., Pearce, A.J., O’Loughlin, C.L. 1985. Soil mass movementinfluence of natural factors and land use. Americ an Geophysical Union Water Resources Monograph 11, American Geophysical Union, Washington, D.C.

PAGE 125

113 Silveira, M.L., Reddy, K.R. (Department of Soil and Water Science, University of Florida), Comerford, N.B. (Department of Soil and Water Science, University of Florida). Litter decomposition and soluble organic carbon and nitr ogen release in a forested ecosystem. Unpublished manuscript. Singh, K.P., Singh, P.K., Tripathi, S.K. 1999. Litterfall, litter decomposition and nutrient release patterns in four native tree species raised on coal mine spoil at Singrauli, India. Biology and Fertility of Soils 29: 371-378. Sliva, L., Williams, D.D. 2001. Buffer z ones versus whole catchment approaches to studying land use impact on river water quality. Water Research 35: 3462-3472. Sponseller, R.A., Benfield, E.F., Valett, H.M. 2001. Relationships between land use, spatial scale and stream microinvertibra te communities. Fr eshwater Biology 46: 1409-1424. Stone, J. J., Hawkins, R.H., Shirley, E.D. 1994. An approximate form of the Green Ampt infiltration equation. ASCE Journa l of Irrigation and Dr ainage 120: 128-137. Stromberg, J.C., Fry, J., Patten, D.T. 1997. Marsh development afte r large floods in an alluvial, arid-land rive r. Wetlands 17: 292-300. Tipping, E., Woof, C., Rigg, E., Harrison, A. F., Inneson, P., Taylor, K., Benham, D., Poskitt, J., Rowland, A.P., Bol, R., Harkness, D.D. 1999. Climatic influences on the leaching of dissolved organic ma tter from upland UK moorland soils investigated by a field mani pulation experiment. Envir onment International 25: 8395. Thomas, D.L., Beasley, D.B. 1986. A phys ically-based forest hydrology model. I: development and sensitivity of component s. Transactions of the ASAE 29: 962972. Thurow, T.L., Warren, S.D., Carlson, D.H. 199 3. Tracked vehicle tr affic effects on the hydrologic characteristics of central Te xas rangeland. Transactions of the American Society of Agricu ltural Engineers 36: 1645-1650. Ukonmaanaho, L., Starr, M. 2001. The im portance of leaching from litter collected in litterfall traps. Environmental Monitoring and Assessment 66: 129-146. Vannote, R.L., Minshall, G.W., Cummins, K.W., Sedell, J.R., Cushing, C.E. 1980. The river continuum concept. Canadian Journa l of Fisheries and Aquatic Sciences 37: 130-137. Vitousek, P.M., Aber, J.D., Howarth, R.W., Likens, G.E., Matson, P.A., Schindler, D.W., Schlesinger, W.H., Tilman, D.G. 1997. Human alteration of the global nitrogen cycle: sources and consequences. Ecological Applications 7: 737-750.

PAGE 126

114 Wallace, J.B., Benke, A.C. 1984. Quantifica tion of wood habitat in subtropical coastal plains streams. Canadian Journal of Fisheries and Aquatic Sciences 41: 1643-1652. Wang, L., Lyons, J., Kanehl, P., Bannerman R. 2001. Impacts of urbanization on stream habitat and fish across multiple spatial scales. Environmental Management 28: 255-266. Welcomme, R.L. 1992. River conservation futu re prospects. In Boon, P.J., Calow, P., and Petts, G.E. (editors). River Cons ervation and Management. John Wiley and Sons. New York. Wemple, B.C., Jones, J.A., Grant, G.E. 1996. Channel network extension by logging roads in two basins, western Cascades, Oregon. Water Resources Bulletin 32: 1195-1207. Whitecotton, R.C.A., David, M.B., Darmody, R. G., Price, D.L. 2000. Impact of foot traffic from military training on soil a nd vegetation properties. Environmental Management 26: 697-706. Wilson, S.D. 1988. The effects of military tank traffic on prairie: a management model. Environmental Management 12: 397-403. Wood, P.J., Agnew, M.D., Petts, G.E. 2000. Flow variations and macroinvertebrate community responses in a small groundwat er-dominated stream in south-east England. Hydrological Processes 14: 3133-3147. Zalewski, M. 1999. Minimising the risk a nd amplifying opportunitie s for restoration of shallow reservoirs. Hydr obiologia 395/396: 107-114. Zalewski, M. 2000. Ecohydrologythe sc ientific background to use ecosystem properties as management tools toward sustainability of water resources, Ecological Engineering 16: 1-8. Zalewski, M., Janauer, G.A., Jolankai, G. 1997. Ecohydrology: a new paradigm for the sustainable use of aquatic resources. In Conceptual Background, Working Hypothesis, Rationale and Scientific Guid elines for the Implementation of the IHPV Projects 2.3/2.4. UNESCO, Paris T echnical Documents in Hydrology No. 7.

PAGE 127

115 BIOGRAPHICAL SKETCH Shirish Bhat was born in Pokhara, Nepal, on January 12, 1968. He graduated with a bachelor’s degree in science from Tribhuvan University, Nepal, in 1989. He graduated with his second bachelor’s degree in civil engineering from Regional Engineering College, Warangal, India, in 1994. He work ed as a civil engineer for the Canadian Center for International Studies and Coope ration (CECI) in Nepal from July 1994 to August 1995. He also worked as a civil e ngineer consultant for the Department of National Parks and Wildlife Conservation, G overnment of Nepal, from September 1995 to August 1997. In Fall 1997, he began gradua te studies in the Department of Civil and Environmental Engineering at South Dakota State Universi ty, Brookings, South Dakota, USA, and obtained his master’s degree (the sis option) in Summer 1999. Shirish joined the Department of Civil and Coastal Engineerin g at the University of Florida, Florida, USA, in the Fall of 1999 and will obtain his Ph.D. degree in Spring 2005 under the supervision of Dr. Jennifer Jac obs and Dr. Kirk Hatfield. His primary research interests include land use impacts on surface water qua lity and quantity, riparian area nutrient dynamics, biogeochemistry of wetlands, and water resources management. He is a student member of the American Ge ophysical Union (AGU) and the ASCE.


xml version 1.0 encoding UTF-8
REPORT xmlns http:www.fcla.edudlsmddaitss xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.fcla.edudlsmddaitssdaitssReport.xsd
INGEST IEID E20110320_AAAADU INGEST_TIME 2011-03-20T21:18:53Z PACKAGE UFE0009283_00001
AGREEMENT_INFO ACCOUNT UF PROJECT UFDC
FILES
FILE SIZE 143955 DFID F20110320_AACFIA ORIGIN DEPOSITOR PATH bhat_s_Page_032.jpg GLOBAL false PRESERVATION BIT MESSAGE_DIGEST ALGORITHM MD5
8e0b574b248030196f1f7e29900a4c81
SHA-1
99378f43aa7c9d9c917dd6a62911e997fd5c0308
146068 F20110320_AACFHL bhat_s_Page_012.jpg
c4ca55341f716829a41df61688560361
a0f9863d2fb0c4391f01b218999f1d328536386c
2546 F20110320_AACFGX bhat_s_Page_116.txt
8a8a2387b8641c6fff5eff88f6e9233f
b6b152b93187e770f5d875644ee66ea23dbedc6c
209727 F20110320_AACFIB bhat_s_Page_034.jpg
c3171596e5bcaaaf9fa7c6accd990eb5
ba62953b7784bf3bcfe3674d2737072fff223240
184498 F20110320_AACFHM bhat_s_Page_013.jpg
a8e96a3cc4b3b3824985d3d8b7a3e8f4
b7548b77a22a10fe1a04be950bdd2338f4453858
199923 F20110320_AACFGY bhat_s_Page_072.jpg
a5e066f211507532b77a93733318d53b
0845aa7699a712f1d1de75b00e1040f6796cf8d9
248863 F20110320_AACFIC bhat_s_Page_035.jpg
aed77632ba02e620e13185984d072cd0
ae727698396d927c1077dfe83173097e5d62a3f5
205943 F20110320_AACFHN bhat_s_Page_014.jpg
f6b4493433e912825b68dba519aae153
cfe3ebda1e32642f1bebdafc738b8f73f771e927
145747 F20110320_AACFGZ UFE0009283_00001.mets FULL
676f68d933c864d0666e3913e3d89090
2f5d46f0f169028b4235db6b94f2b788a490efc7
201020 F20110320_AACFID bhat_s_Page_036.jpg
b0df3158f86e7010e790024c3127bf30
57a837fa56d5264f5cbd093abb47a5228118ae24
190033 F20110320_AACFHO bhat_s_Page_016.jpg
5bc182fb0d58de9fd4dae7a37e18f87c
40e66631b8249788abbd82f3e5524885bfdde36f
131234 F20110320_AACFIE bhat_s_Page_037.jpg
b0211177c18d6188ade29a07c6a43768
119331cdde44eee0e49b3ef9cd382bcf505a9130
206483 F20110320_AACFHP bhat_s_Page_017.jpg
6698ccd6fda34ac3f070a6f40faf9427
cb4bc7337f8e9617bd5062093b8b7e49607c51d5
132710 F20110320_AACFIF bhat_s_Page_039.jpg
4fc03b0f92f31da0c545774ced9d2b83
aae956fc1e40c572dadec835d585ef2d70adf478
198058 F20110320_AACFHQ bhat_s_Page_019.jpg
f7c09a979634e737acabb9044053458f
901616da306bfc89e82a988f8743993c697fab29
137255 F20110320_AACFIG bhat_s_Page_040.jpg
a6d8445b1a96575b55eec2b6a18c5e9c
45c72224107f9ca53834f8a3a341fdf16d031f03
46346 F20110320_AACFHR bhat_s_Page_020.jpg
7916c315c89c20d8644dd025b962d07e
1741c6708f44b1fafc8b9d16152c37450361ee25
199232 F20110320_AACFIH bhat_s_Page_041.jpg
f2d7b795293dd338a5a20c68869d90ec
f1ac20fe603fa0c28f586a66126b4bdc395a1152
176781 F20110320_AACFHS bhat_s_Page_021.jpg
14be9ceb063d1dc02572de8a995d260d
fddf1e44860e382bfc796ed2830d48b6d8729ebe
203025 F20110320_AACFII bhat_s_Page_042.jpg
e5907d8265df246820bc80755fd0768f
62eadb07ce220df96ed30eb680f32ebbe59b7f46
203428 F20110320_AACFHT bhat_s_Page_024.jpg
6ed9dc4369514a9bf9f539702869c1bd
8a82b4284103d6626ef9069bdb16459c4481424a
217051 F20110320_AACFIJ bhat_s_Page_043.jpg
3a0bf320d2b40c43c5b80c9a0553e324
403c324addaa5fa730b8e944f046d06c5d5c6568
196983 F20110320_AACFHU bhat_s_Page_025.jpg
fdc28fbeaaa96e7b8f25839d8eed1f72
079083f8c65a36a5d97a0dba8cb040be6f857d52
216826 F20110320_AACFIK bhat_s_Page_045.jpg
a3b394b45b2b297ffb486e8a73884c0e
3d430c1f169c9cfa2434cb0b3c014d52acdf77ba
218434 F20110320_AACFHV bhat_s_Page_026.jpg
dac4d59114f55ac6fbc6c68c4abfa64f
9992dda058817249f59730476215f94f54846aa0
209850 F20110320_AACFHW bhat_s_Page_027.jpg
6050acc73b093e567e4686c7489f9aab
398ee6893e07bd21b60fbe0b8247f9b49715f98a
182014 F20110320_AACFIL bhat_s_Page_047.jpg
f0daa09857958f5fc1f8f7824210864b
75ed0b4322ac269c17f1d802786a617930cc0c1a
174523 F20110320_AACFHX bhat_s_Page_029.jpg
792b87f704f455ab3d0e58dc0cc85e17
ddd61be0cd4787f1fe46c6fcb651e26e7759e674
216320 F20110320_AACFJA bhat_s_Page_070.jpg
271782219bb5c1412af1df679907b6f4
c1d2dd0fef6a7b9f956f323d4e588cdee79ecf70
206099 F20110320_AACFIM bhat_s_Page_048.jpg
eddbf00efb9b28effdf07f7e680ce2a6
7115fa307896a736d195d22ee7aee59c1ce57a13
93701 F20110320_AACFHY bhat_s_Page_030.jpg
4b81327cbbef973e917c569dc3452788
29966597c8941ddab7200877676ffa2e8555b361
209457 F20110320_AACFJB bhat_s_Page_073.jpg
1b06b716c7a47ce0106734bd4a149799
49cbdc3a8c70addfb3dfb815b8f2b05619953fed
219886 F20110320_AACFIN bhat_s_Page_049.jpg
0da416e1d72ce6179ccb58d1b95a52f7
85e4b357e5a56efa72a3896bc8db161c7034c2e9
203489 F20110320_AACFHZ bhat_s_Page_031.jpg
cc373de8fe13b7d0269efe8fc6ce35ed
e6197d51796d520b3dcfbcdc741185a23021e275
204078 F20110320_AACFJC bhat_s_Page_074.jpg
8f28601825137809ec6dd4263d5195f7
2b23aca7b5035f11e69d2386d320cc45d46aae39
208492 F20110320_AACFIO bhat_s_Page_050.jpg
622be54957519913a06bcfe3efdd5ded
f40b32dc126825b7e3f7b6cef36478b75258c927
209006 F20110320_AACFJD bhat_s_Page_075.jpg
839248c7707d290abe608e5bd6cb1a91
06e449d25e7c2ed1706d0f77ab190520c53ff9fd
175817 F20110320_AACFIP bhat_s_Page_051.jpg
9ca533c327c6b10f686ba409cd98dd70
ab31c41b8d3d3ef03cb19557bfb894caa4c3e76a
204843 F20110320_AACFJE bhat_s_Page_076.jpg
8c56bb9942ee07e622b186a1f901552a
94bfe3a1c9549ed7855135cfadf473df8b3d40c6
207646 F20110320_AACFIQ bhat_s_Page_052.jpg
dd11adcb9d5d4d3f304abc638a12638b
c9d05057ccc7c91629027afc38ddc443d2ad55c0
185334 F20110320_AACFJF bhat_s_Page_077.jpg
58016a8e405f9922d5ce2ac6a7cd8a32
24a25ca4e00535b51406fcfdf426a46ca601218f
151992 F20110320_AACFJG bhat_s_Page_078.jpg
fbbbebd513442e1ef96db3ac407be710
88344d4e4378aaa581fddf2b59fefaa53e0a9ce5
216254 F20110320_AACFIR bhat_s_Page_053.jpg
2762074ec9f8231e30dbe8d6247d5298
c9d56d6635c0f42ee2fe81988f9c3b1c29c38db1
163543 F20110320_AACFJH bhat_s_Page_079.jpg
cc5c08873e5dde39983200238cab42e0
0cc6e2d576baedd72dfa60173462e7eb5674a746
208333 F20110320_AACFIS bhat_s_Page_054.jpg
6412d1ad9e15c808b6b7e849fbac1f2e
e11bd8f740ae705ab97f6914aba68eaa444af8b5
203560 F20110320_AACFJI bhat_s_Page_080.jpg
21db7a030375260541e07b5796c7f6b8
4d689f36e0ac2a8ea62cb8c3fad724441aa022ad
160643 F20110320_AACFIT bhat_s_Page_055.jpg
bf5cc0c671cebf2928fe0cbb90af1def
e2f6db17d6c74cb194721110e10b6064849f189b
216109 F20110320_AACFJJ bhat_s_Page_081.jpg
3139a74de06455641fb8d7593902d559
8d1b8a3b0a10808b2f408b98f4129568bfa66b4a
175866 F20110320_AACFIU bhat_s_Page_056.jpg
9fdd03c0c7f29a49f31486271e9b6bb3
fe0c8e449157ebbee852bfc3cd0d436fc3594203
217497 F20110320_AACFJK bhat_s_Page_086.jpg
fa77b22cb6c69a9da7a891c4fa30fcef
4cf956466dd4ecea5b6d77d45afd44a0eab4594f
195447 F20110320_AACFIV bhat_s_Page_058.jpg
8f6a23c0ba9730cd55c2e14fe8976d3d
b0400872cb51e090ed1256620ffe93b18ded179e
206668 F20110320_AACFJL bhat_s_Page_087.jpg
d107749111305c27d26dac91c7ae13e2
d257a37717a22c2381e1f38b2959740df072a942
195421 F20110320_AACFIW bhat_s_Page_060.jpg
17590bdf94e97fdcf8302b38b7d0312a
ef88224b1a8b1233534fb1425534d879d2374dfd
94790 F20110320_AACFKA bhat_s_Page_110.jpg
37c0903d768731b714d6ba1286ed998a
3e55095e9b2cfa8e8dc7c44f90a4af13448aea9b
228556 F20110320_AACFIX bhat_s_Page_061.jpg
ece94fbee5be5a5db2cb387c8a93506f
0d40774258ca5960f245b23de033e857c8a35a72
121668 F20110320_AACFKB bhat_s_Page_112.jpg
20445b579fde9c53063f0ac340158081
5ced75becf9a75f4b2263672fa9608b774e45e44
132253 F20110320_AACFJM bhat_s_Page_091.jpg
7676b9993854e7784bde5c878b6e9cdb
7b2f4f12ad87c4e8f1360ff14a2faed169887a1c
192748 F20110320_AACFIY bhat_s_Page_064.jpg
84636ed15c0a7a784cda9b7c81d08dfd
3115aa469981f271f840358961ab9df3120c4dfd
45125 F20110320_AACFKC bhat_s_Page_113.jpg
7b14b520b1c9787e850cfdb27c8fb21d
4f99ea3b867cc76a84afd094dcf1147adb68173c
212190 F20110320_AACFJN bhat_s_Page_093.jpg
fa3e2c32752b0ee7235099008a5697de
d670d92454d44fd720233915c97f9409aa9f7975
185519 F20110320_AACFIZ bhat_s_Page_066.jpg
a71ea49e0236bf58f635335a97e049b8
477f7ee590e5a3af7ea52857569c2a2bbc38b2b0
235886 F20110320_AACFKD bhat_s_Page_114.jpg
a8047f10bb4e603d891a402679a9b4be
7e58089fb4c39ac843ccd97460e5c6c41f3f393e
205456 F20110320_AACFJO bhat_s_Page_094.jpg
69c553699ff4c66a08fff60b93bc92b1
87ebd548853ecc5a199d8e0ceeb007e6f980f8de
266810 F20110320_AACFKE bhat_s_Page_115.jpg
e37a3eb61f0e8e2b886ce713d4b51e32
f1cece2aa30efa055e1b71d75780a399a5cb53f9
222799 F20110320_AACFJP bhat_s_Page_095.jpg
a96958ccecb2498a86d2faf09a9e7823
4ac3dd1a88fe80210e7967c23ad8026a7fcaee7b
272059 F20110320_AACFKF bhat_s_Page_116.jpg
d0351fb0ed24464d855af5fd08b5963e
75afc7458c216e4709412499190c51698942b270
213468 F20110320_AACFJQ bhat_s_Page_098.jpg
97ac6ddc2b7fd9cf6f7049e7df43b9f9
3e7e9b23e39f3d50506e1ef4e3a149047e3ce988
260720 F20110320_AACFKG bhat_s_Page_118.jpg
aaf7b4aa944111b976f7e4143220b6f2
89004d073e6ffb1820a2c022490681ced098ccbc
205053 F20110320_AACFJR bhat_s_Page_099.jpg
e5bf91b7371ad98284b59c6449d019aa
751937a89500020a035a227b9267f5fd69468a7d
253110 F20110320_AACFKH bhat_s_Page_121.jpg
2385c4fae468dbd5a2484048a6e5a60c
07ad1466156820e4323bf3a19c16df5c42cdcd84
173944 F20110320_AACFJS bhat_s_Page_101.jpg
09a9748157e819724d9ce4c497a97cb6
26c8015fc53f912cbfec4efce1609ca1479fb403
266212 F20110320_AACFKI bhat_s_Page_122.jpg
b5727058be117a4935cdc0d2af840e78
25ee0a02d3dd0f941bffe0236ad77ab8267b17f0
193723 F20110320_AACFJT bhat_s_Page_102.jpg
c0530f61f0ded7c97f12d479a17a7de9
f83556d7c759e4337ee0f0e001b1dccf0b4933d8
263317 F20110320_AACFKJ bhat_s_Page_124.jpg
8f85090a0046ee5d8a36a1726e065f8d
3fa4784329d0ee7053afc96ebe3b6a31a07fe2c2
168781 F20110320_AACFJU bhat_s_Page_104.jpg
499aa51baae274c6c086a7c7c9e4d064
e95fcb9ba0c21fb10a45453819b4a74262004166
271073 F20110320_AACFKK bhat_s_Page_125.jpg
952da0f81a65620c15c3b8d0280ebffd
5e40c024fd7bbbdd02486ab61d1d3df42e01e636
162137 F20110320_AACFJV bhat_s_Page_105.jpg
199588288b97fdb61f8a8b6ad7d7a816
4b2a3437bd23793131ed1a1caff7da861ac76bc2
212092 F20110320_AACFKL bhat_s_Page_126.jpg
53c637b5dff7a115e076d70d583a2f6d
48a40a1ba464e069a31937ff9528728c208e4a0f
150527 F20110320_AACFJW bhat_s_Page_106.jpg
87218309b90793b5002b6b879fc44b70
5e254ef6fe17de5c3352d946d0874088bcea9c61
159087 F20110320_AACFKM bhat_s_Page_127.jpg
aa12279b117ef7458e2859d961108384
c6dd11f795b7807c42bdc3c68ed5b6fea8aeb3bf
138586 F20110320_AACFJX bhat_s_Page_107.jpg
3aca00d5ab4478a9c4f9cde652d94f08
26410a75e5f03261ae197f73041e0d23dab56719
108975 F20110320_AACFLA bhat_s_Page_015.jp2
31d89ead21eade0f055746f10135d29f
b30a1c8d34202ec5ad974b6832a87cb2a8e0fdeb
169203 F20110320_AACFJY bhat_s_Page_108.jpg
d481497a800dd657ff878398fdec539b
711b7d1783b40fa8fc9e45a96ac91e290870bbfa
98982 F20110320_AACFLB bhat_s_Page_016.jp2
378b3987de57922b48277b994560ec67
a9c190b4760b2998034238c198cbd22452692b55
24432 F20110320_AACFKN bhat_s_Page_001.jp2
ad2f09811cac5b796bab3cc5be37128b
d8fc2c8978ec66e94f6ed1c372d884a451ec558a
138199 F20110320_AACFJZ bhat_s_Page_109.jpg
c486fecaba2f4f6ce5fea06b6a39eaec
c4d2e2d333e66b25925d533581f3fbda9d75948a
106783 F20110320_AACFLC bhat_s_Page_017.jp2
f0ae4857e6273304b70ecca290b296d4
09f26c9fddc3ea7fc744d4ec4c4ed79f6b47fac4
5391 F20110320_AACFKO bhat_s_Page_002.jp2
86e271774fa037a0c7db23bbbc041107
9cc4e1d13d3b287a0f7729bca05b4eabb4e53625
107706 F20110320_AACFLD bhat_s_Page_018.jp2
13fbcc72c346a082ada6be36f52849b7
7ddcec4fc60025775650da1c1aba7d2746dd0a20
8803 F20110320_AACFKP bhat_s_Page_003.jp2
fa11548a2b28c45a5c2246b9adf14b36
e33e82d49be3d171ba571955511ebd75b5b93ec4
101716 F20110320_AACFLE bhat_s_Page_019.jp2
e19fd11bd4a93b73f7d769ca086c1402
846cbbc53c93d9bba7d78894a36bfb57686454d3
96978 F20110320_AACFKQ bhat_s_Page_004.jp2
6760884f5a5f71deeafe4783ead7c711
b428f33de7df0439f2fe0e06ae69f201c811553f
24743 F20110320_AACFLF bhat_s_Page_020.jp2
d6180b2eafa7d4e17e4c85922ff67444
391de0f5a8193a6cfaeaa1902cc664fd939d60bb
37082 F20110320_AACFKR bhat_s_Page_005.jp2
b71181925ee72eb12c655ef0aa204d7a
af68414fd0d59fe878215f669565030f46d669fc
89660 F20110320_AACFLG bhat_s_Page_021.jp2
ef338ed9785c8b3105ab51ecf3788be6
3be861efbe045ffd35f935bc97802b94851ca34d
1051986 F20110320_AACFKS bhat_s_Page_006.jp2
498145e0d497941a940b4aa717283489
33c48acfc11bb3d3132982c8ddf36b3865f77643
110238 F20110320_AACFLH bhat_s_Page_022.jp2
b2a544682f953635a6b2ec1fbc71376f
63eaef19a4694277a26c383b29bc6f099349a98b
1051982 F20110320_AACFKT bhat_s_Page_007.jp2
21206e2f9cc71bff72e5a9b10f5dd975
d011a9ece02856d8376f77ddd6e89882361937fa
109535 F20110320_AACFLI bhat_s_Page_023.jp2
377e3fdfd14c6ed5c50d54a36f57824a
ea34bad5c66d826d8389e1a250176192a160fe17
1051975 F20110320_AACFKU bhat_s_Page_009.jp2
2e9885bf56966b0c617c472eb1c56ad8
2821cf9e660487aac0fe2aa475adecb152c89576
105274 F20110320_AACFLJ bhat_s_Page_024.jp2
7155f537417ad0e815fe46546eb84537
5fccb9970d457939efaf7cb6e37378ee70d6e37a
1051970 F20110320_AACFKV bhat_s_Page_010.jp2
91684613725e105f1c7b2b434690ede2
902fb92506b83cac5c325612248e0d92aa87364c
102070 F20110320_AACFLK bhat_s_Page_025.jp2
c749de86970366139975c9656e16ceb7
cce24c3363fe0a3b59221891bd6b1cbf0097da84
86686 F20110320_AACFKW bhat_s_Page_011.jp2
89de8f3e0a7eeda8a4b1a0bf204b4465
8d0b646a33402ab0fba5753fc6eb88cb0cd7479c
1051976 F20110320_AACFLL bhat_s_Page_026.jp2
3ca851697e2f02c337701bdf7642b487
8fca135daee4e00f167504aac607456932951777
75298 F20110320_AACFKX bhat_s_Page_012.jp2
1dffc5b2644133e220b1a1f776b41b92
0f05e02a2697039e731fd347357fd114f1205298
83458 F20110320_AACFMA bhat_s_Page_046.jp2
fef8945a192507a394bcdf6f88785321
99679dda25ca50e7841b0cfc655e734a9185dcae
87006 F20110320_AACFLM bhat_s_Page_029.jp2
6fa7d863b49129ee2993f31bc0dd2d9d
f65d4b0671a4e14b6f4444b3842e2f174fd72f54
96146 F20110320_AACFKY bhat_s_Page_013.jp2
4f0e153569b6c27ded67933285911c4f
6802b1326c0ed540e2724647475b56f7333f1f38
93636 F20110320_AACFMB bhat_s_Page_047.jp2
b194ba06fabcce973518d0d1bac69a43
f31fcd477519d52e10daf04e1ef40ac621d20419
72925 F20110320_AACFLN bhat_s_Page_030.jp2
0f87852fb94f087a3cd43e855805bf5e
89f87bc34c427080d3942946a0e9d4043b558739
107154 F20110320_AACFKZ bhat_s_Page_014.jp2
bb34e914e7101fe722f35d9d02cfa24d
24701a9dac95de860068bd8f26dcdc245308666f
108484 F20110320_AACFMC bhat_s_Page_048.jp2
60e58bfc86b3e575851f2c1015b259ab
cf77c4230cacabe07e020324d4bcc55b21af2611
113746 F20110320_AACFMD bhat_s_Page_049.jp2
232988dd529b1c6522a6a4f23e4e19f9
a217d019ce55dafeba9cb26cb34e386a981acca8
103703 F20110320_AACFLO bhat_s_Page_031.jp2
90d3c7c071fd450fc60b3493d6fde99e
3c6fc7b2324d9523c187a8d90ba96e21867e1d1f
109547 F20110320_AACFME bhat_s_Page_050.jp2
4476d69b6aca882528c947c297bab759
db32da0580ef935ad014f4152bdc50eb395946b2
585970 F20110320_AACFLP bhat_s_Page_032.jp2
c412b19d997035b222febcc4c17dd7f2
6039183ed4f7f533c28f7d8b27ed477d37815c4b
762931 F20110320_AACFMF bhat_s_Page_051.jp2
90af9a6538e197c46db3b00589c4623b
b43307d93d6f15b18289823523ee428011ccde4b
88657 F20110320_AACFLQ bhat_s_Page_033.jp2
ad51a26176e312887b1dbb6d30c334fb
61ebc3cdebd9768601c1e94eb66c1e04ab747346
109121 F20110320_AACFMG bhat_s_Page_052.jp2
be53f12493632cc6a88ae0a704153318
e1f1ffde5f2830eefea0fd6fa077d244c3349ce0
110177 F20110320_AACFLR bhat_s_Page_034.jp2
6469688cc9ef8f3caf7da0818f8aad8a
1ec8b29779d2934ca752f1cddd22cc6499135056
112597 F20110320_AACFMH bhat_s_Page_053.jp2
d12765783998f08885e06a1e86835660
b567e04727d776788dce35a4d8402101b337f848
122356 F20110320_AACFLS bhat_s_Page_035.jp2
6c5968dbf1ca543eca1d70acb7c97485
f9e44a65c566ee574b0c98a0b7c257846c1428c3
107688 F20110320_AACFMI bhat_s_Page_054.jp2
a106540cc52ec8210a25c4926ebf37e8
5d5ace38b2cc57f1acebe4bc5a1d41ef27e7565d
103321 F20110320_AACFLT bhat_s_Page_036.jp2
63eb161b4cf3da200deef988109e4ddb
89b53cada3cad2f6eda319e2dffa6fd0d0eac66a
671814 F20110320_AACFMJ bhat_s_Page_055.jp2
3e9b0868244fcf903dae9b259267ee4b
3cf30943cd0fef20be12fe4382843bdd0af62d76
534059 F20110320_AACFLU bhat_s_Page_037.jp2
fc643312def3eaece20a07c2af31d6e5
6cfa5780a72a8a75c487c8b549c1eb52d311ee93
87493 F20110320_AACFMK bhat_s_Page_056.jp2
ea8b55aa272ba8b26d477b107c68fd2d
1aba96b5c4f2b040c033e1b927cfb02aad913143
563443 F20110320_AACFLV bhat_s_Page_038.jp2
f429b1d2db4c48f491aff4d584a2c04f
609569982d823c7b0a4e77adb90ddbbc3e5e4dee
88525 F20110320_AACFML bhat_s_Page_057.jp2
03f1299223da98e5f081679009702969
942685cdf632d8fc65c50246c82f1cd56f89b6ad
107411 F20110320_AACFLW bhat_s_Page_041.jp2
b556ea329960e326b0b4abbac8fbc713
e35bed0d63f7775a10304f88b908cabb6811fb7b
106434 F20110320_AACFNA bhat_s_Page_076.jp2
ab360b67948067cb1769901368765f4a
82e855c2d596a6a74d1b8039180f5d8f0b859a0b
95991 F20110320_AACFMM bhat_s_Page_058.jp2
26e63dda4361dd42fa1425a467d7c6e8
da520974dfbe38feaa3156df38392a87fb2f6041
106238 F20110320_AACFLX bhat_s_Page_042.jp2
edfcd783a8b683784d98f657134afcbc
4f735f9107c9661f7ff263b55fec7bd16a08b723
95349 F20110320_AACFNB bhat_s_Page_077.jp2
fcb1a8dea33df4fc6423d380bfc2f5dc
99bee5ab95ab8a4e8799f98eafa37ccfbbe98c12
93536 F20110320_AACFMN bhat_s_Page_059.jp2
d160c5d78bacf0eda9457251a19dc43f
a3270c9b56afce03154ad9823b6f3b478e1f73ff
111179 F20110320_AACFLY bhat_s_Page_043.jp2
27230a9dcafc7969f75c3e37331684fa
3b1edc897c8107a24ac61b50cce8cd8ff9babed3
632706 F20110320_AACFNC bhat_s_Page_078.jp2
e5d9ae80073974fedc41402bd1b1c5af
2599be553e9526091ba21a0cacb7f93deec7325c
108426 F20110320_AACFMO bhat_s_Page_061.jp2
a2b35290bb326a8c377db8ee745fe11b
c99fda22d78d964c2ab081e00ab4d29c18a76901
112233 F20110320_AACFLZ bhat_s_Page_045.jp2
149bb57e85ba5d492b7cb4cdc9d28ccb
2af1cc45ccc00ea731427b53b9faa3354f8024e3
83099 F20110320_AACFND bhat_s_Page_079.jp2
31495242d7ea21a17f44ca3a5d114abe
b2fdfb1861357e79cef1aa9f6cee50626b8ba3ec
108329 F20110320_AACFNE bhat_s_Page_081.jp2
b329b6bf506d7b4103455fed562d8d80
62fea89019963c38142b6ea0be00fdb4439a6e9a
111328 F20110320_AACFMP bhat_s_Page_062.jp2
39cfb2ebbcf4617eaa27cf7183d4c647
65b2b2b93cfa677d4273a023d62cd2b20198893c
108882 F20110320_AACFNF bhat_s_Page_082.jp2
74fdda63a36961fc4749792edae209b8
85b3f52fa833e2c2d6f915863ab9c8cb261c06d9
101385 F20110320_AACFMQ bhat_s_Page_064.jp2
6eab829bdb8275e0c37bdfef5a235596
3cd0a57bcbd8b680c339491aed339ad14f8dec01
110184 F20110320_AACFNG bhat_s_Page_083.jp2
83e5b60ca62df383902b887cccfcb4dd
1654c94eec4c45bb6971908592fc892bfae0defe
79745 F20110320_AACFMR bhat_s_Page_065.jp2
662f07b7695aae008ef28d1cbaf26ddd
ceb6e75ef8e10c7deecf4121a53552d87c7cf475
92503 F20110320_AACFNH bhat_s_Page_085.jp2
c6cbf93b1804ee75f92995fee71fb112
79d529c32292909cc4146f3e994d1eecfe9415b0
93437 F20110320_AACFMS bhat_s_Page_066.jp2
d768c6ed379cf561bcab8717d58b3538
c1ceffa4d9e965f379142c1d3d41dbd30a9f6f03
1051959 F20110320_AACFNI bhat_s_Page_086.jp2
3d81fa37a6e215563f92fcf913b1fd22
aeea595bd66a74bce1634d75d7d60dde67ddc38a
112135 F20110320_AACFMT bhat_s_Page_067.jp2
918c48b4a945ac9f4d6147ea2cd3189b
0a995a3e1261dfd32a6f0814123827c7eb2217ed
1002533 F20110320_AACFNJ bhat_s_Page_087.jp2
76c22265fc67243760d6a8d1f68ffc59
aaea3e77a15a8b666534810c5ce0abfa3eeb16c3
107659 F20110320_AACFMU bhat_s_Page_068.jp2
4e3ba7b3edf838c57b41ce70150c1390
fe74fe544a13607416fedf9d5ec75f04bddf092f
81693 F20110320_AACFNK bhat_s_Page_089.jp2
ad50b86c0b8b302ddfa12d765fc8a7f6
778a82b3d009ef2e62c37d714dd1da30d691f9e7
111323 F20110320_AACFMV bhat_s_Page_069.jp2
bcc6594d928a53810ad104acf06b8fde
54da8d16d4f8a59bd465e0020a66902c28a6b20f
935528 F20110320_AACFNL bhat_s_Page_090.jp2
c2fe156b759790938adea16bae79e3c7
fa332352fa89a646ee5f887e884146307fc7852b
113537 F20110320_AACFMW bhat_s_Page_070.jp2
60814db2ab203e19f27bf4ebf03802e8
8f66c71c8f5a582d079aeaf7627f3382777eae69
509636 F20110320_AACFNM bhat_s_Page_091.jp2
9235342d4baaca9ecd1f9e8946766fff
4d1d046961205d120ccd79db55fc988181e2efaa
95178 F20110320_AACFMX bhat_s_Page_071.jp2
b739042fb8255285993125ceaa4bfce8
fcbd501a1eaa32c3e605077b4816c7f6ab2abe60
85793 F20110320_AACFOA bhat_s_Page_108.jp2
915057a974a2aad53a7e40d8f9301376
9f5018d4922ff44b608de778ee1be91f589a6614
86918 F20110320_AACFNN bhat_s_Page_092.jp2
41daa82ad525026fcfcd167bcfb4cb35
9e5f9efca614b5083d2b733421ef0b6eeabe1448
103398 F20110320_AACFMY bhat_s_Page_072.jp2
fcba1aae66640ad0ae85e6aa68ec1167
a3d790bc5e14536b7a0be2a47cd577117b765c78
67641 F20110320_AACFOB bhat_s_Page_109.jp2
801efdaf42e271816f2283d204c107a4
d0c333841301494cb5bb445fd37ab1f25ab45a34
110956 F20110320_AACFNO bhat_s_Page_093.jp2
e3fb346ade0872c0ac0b34f496f02006
5e84ceae962b8bd97851566b3cbf975bab45d5c4
106963 F20110320_AACFMZ bhat_s_Page_073.jp2
a7e6fdeee282c24009a7d298981b0e14
132ea53ba10d9dce932aaec022352f334deef1a4
50102 F20110320_AACFOC bhat_s_Page_111.jp2
83e1e70d55b638b772a8c5b6391689c7
91fecb682f1bdafed9c8b8d75565eba2a60e589a
61778 F20110320_AACFOD bhat_s_Page_112.jp2
167ab23fc1823e251a5efc23d0b8a0e4
135f4c796c42b8af24717fd0e4b543a5b4577070
114797 F20110320_AACFNP bhat_s_Page_095.jp2
cb30785031c5b408dac84ed966b215d7
b198fba6780105401cda39c638efc050ec74fd1c
116085 F20110320_AACFOE bhat_s_Page_114.jp2
06f8b46122763f4f978016257abb4613
f8702522baed97f39c8f1067bf16803dfd74456c
129112 F20110320_AACFOF bhat_s_Page_115.jp2
c820f5c06531723b2c6ee4deb1f4280e
72b1456670d8d6349c35bc025f410a70aafa2d91
97808 F20110320_AACFNQ bhat_s_Page_097.jp2
1451329ba65a405603773b395f25e3eb
7762dca166447127ebc8598d9af4cd6eb5313f43
131637 F20110320_AACFOG bhat_s_Page_116.jp2
e14f683dbf5e4d0cb5a8deeadbc2fe77
4ddd89560c92a9fb7c5885c09f5837a5d8fe06de
111072 F20110320_AACFNR bhat_s_Page_098.jp2
3520b339a05ca2aeadbfaf4bb9b6c35e
4061ba726b94228c9e861ebd01f11d2575aebd6d
131058 F20110320_AACFOH bhat_s_Page_117.jp2
b3a954c6ca2a6832784f8b0eab4b7542
d3581c5040384f78cbe532f3b1862a03bbdf8648
105708 F20110320_AACFNS bhat_s_Page_099.jp2
f087ec1c8568833434041d37b101d557
2104227c8d2f9846ab36af70ecc082970699ef0d
127915 F20110320_AACFOI bhat_s_Page_118.jp2
5079aa7bfdc61d204bb5cbb5873905ad
0a998b636e8e8f3ea34a403b275d925deaf8e5f7
69186 F20110320_AACFNT bhat_s_Page_100.jp2
84326c915f9409c4e68ad0e7338f1f9c
2c6a6c968ae900d585b9109d488e590381c7962d
130657 F20110320_AACFOJ bhat_s_Page_119.jp2
357dfc0776fb714cddc36bbc7d1788cf
63032cade49aa6d48095fd8367029aba3a014a3e
96808 F20110320_AACFNU bhat_s_Page_102.jp2
31ef3e4cefc9f1655c7be6f07bce35a5
52f9a13b71e201b919c97e5319efe93b0c932d13
126269 F20110320_AACFOK bhat_s_Page_120.jp2
44f946afdd3460e13a21dbabea62adf4
92d9edfea77913909142c6968410e665cd9ae614
70649 F20110320_AACFNV bhat_s_Page_103.jp2
e368534d138bb1e56c4518ffeb992c54
5d17fb23c04396d5ef44b5636dfb20116503c7a8
125690 F20110320_AACFOL bhat_s_Page_121.jp2
c37592f21f173c4cdb48e5514c379f96
dd270d8ebbb39705084336ea71806d75b58df401
80235 F20110320_AACFNW bhat_s_Page_104.jp2
7c2784211edf665c07131c6716d794fa
ad272f8f67c045a58631e5a17a350212479c0990
1053954 F20110320_AACFPA bhat_s_Page_012.tif
bca2d569dfbf1fe23ce54c9d996706f5
097fa6ee174dd449c5d28912ea750f60614be9cf
129297 F20110320_AACFOM bhat_s_Page_123.jp2
e0dd1975b3462b73a5392898e46ab96d
8bf265119b426b4304d3eb664ef035391d03e240
82129 F20110320_AACFNX bhat_s_Page_105.jp2
28f9f3967642accb804ef18b18b32552
579fd241f0baff65ee256a3453201a4fd9110751
F20110320_AACFPB bhat_s_Page_014.tif
1919d31ea70cc7049f2f054ad06c8a37
a3946c8b64637c08fa69a91894440d5b4c6c6edf
124937 F20110320_AACFON bhat_s_Page_124.jp2
f3c9d50e2a1e73dda669a76cc7930ed9
38930b80d2fbf618103f877a6a6020cc53b8a2f5
75745 F20110320_AACFNY bhat_s_Page_106.jp2
ffd357477c68e307fc98c82aee45fc1e
912d3792537e7e42aa27e88201e5fd080314d6cb
F20110320_AACFPC bhat_s_Page_017.tif
9c6e71d0af7a58ddf0af764429536d77
93de2cc397732611d01064c535fe41ffb0157922
130897 F20110320_AACFOO bhat_s_Page_125.jp2
a3f612fdd536a9b2e4b0c16ab28f1a27
408f90d9d7ada1ac3c8e6dc59fd48ab7e753e401
70888 F20110320_AACFNZ bhat_s_Page_107.jp2
5229cb3e943ec7862bef8d9998cca163
91599bfdfc8a212697776a61b7b032c8e4346271
F20110320_AACFPD bhat_s_Page_018.tif
0916baa1d93daaf3c115b5ebfa15f1eb
bf3578bc1e8bc253ce1cb4dcc374243b20c39022
103994 F20110320_AACFOP bhat_s_Page_126.jp2
c227892ce64b9311e8eacec7f7be6c93
9e1cd99c1d860f1237e45d2830c400c21921447e
F20110320_AACFPE bhat_s_Page_019.tif
581de9246f4596f8b2e16204eb44b611
5e7f4c34b8d6e6568c1b8e1eba190e006ee7dde5
80449 F20110320_AACFOQ bhat_s_Page_127.jp2
b498da862eb68be1f93cb4f05e5ce735
420851519bf254439daf73bcea565ef3f1a5794e
F20110320_AACFPF bhat_s_Page_020.tif
cd3441238fb8459fdbb5a060299910c3
8e0deffcf6c72b8d425d4fd67558f4cda946904e
F20110320_AACFPG bhat_s_Page_021.tif
3dba7eda927852ef0fd18528ba3370c0
196e1ebe70311a720cdd72fbaa0f782ec90cfa1c
F20110320_AACFOR bhat_s_Page_001.tif
5be7e3e1312b22659b8f123d69408322
8428bf91d00e188ee98e1e4f14449d85852cd945
F20110320_AACFPH bhat_s_Page_022.tif
79e199ce242ac0dd414fcc093ce178d8
c25cfdadf8ba0d757803b721c08d0c7020885c95
F20110320_AACFOS bhat_s_Page_002.tif
6ecd8055e7bbb4ad947bce882e967772
a0d0b1c4c4008f8e4b9f57629fa3a9927f74953c
F20110320_AACFPI bhat_s_Page_023.tif
f647ac61daf55e93b76bda8c0c627132
531cc8120a6ae364995d3d44b4105688acc306fb
F20110320_AACFOT bhat_s_Page_004.tif
cf3e25c788d89d999a83551a17b570fc
fdbb8f2ddba0aae70fe4fa2282fe0dcd9bbd7d6a
F20110320_AACFPJ bhat_s_Page_024.tif
3144af575d82ef7a142fcced93442234
d69ad49d4992278fca9936a07ab3c004b0008271
F20110320_AACFOU bhat_s_Page_005.tif
1a7fbe86044850904b9b5fa242e106bb
25cc7a789428ffbc8282ec351e6939e2c033f330
F20110320_AACFPK bhat_s_Page_025.tif
46776f31a51fb60b6e19168e55532589
6de48f5d5ccdbd4968c7c2eb2f248601631e1cce
25271604 F20110320_AACFOV bhat_s_Page_006.tif
8ab2a5e145885a438fe1cc90ac46c05b
df495d1060c00e77117a71f26a4357e24f96447f
F20110320_AACFPL bhat_s_Page_026.tif
81d3192098d637ff2233c781e3522c67
e12a3c750c1213b052c48becc537d230d349a514
F20110320_AACFOW bhat_s_Page_008.tif
483ef59e3608b8aba548e32b3c5bc8f7
9097c00226826ff36934d4add7584117bde83430
F20110320_AACFPM bhat_s_Page_027.tif
7d7036dc4861c71e28bca71a76c20c38
4317f0c41f7b9fee9a1267386ae27a1a98d8bc99
F20110320_AACFOX bhat_s_Page_009.tif
ab9c65482b742513556adc366ae14f29
0e017147b25914e728f8238c4ef7ce6698ad462b
F20110320_AACFQA bhat_s_Page_044.tif
b1dade759c258dd47cc562d77955ff3e
4caeb2e29a905dd19970b4ec41cb544accc24b57
F20110320_AACFPN bhat_s_Page_028.tif
bcf0e46f675e617db2204099205d8f8b
20419736853b035a6ea7a03787f33e73d8e86d10
F20110320_AACFOY bhat_s_Page_010.tif
ebb026c6b9f504a3c956a8ae4402df62
176f718e363889d0f4365d6f5fed271c028929ac
F20110320_AACFQB bhat_s_Page_045.tif
feb2c47bcfccbd203a62ac0e90390050
ec4f65da3af1ba44a9d5f01e7ab1ce40eb5feb3a
F20110320_AACFPO bhat_s_Page_029.tif
73119968037f32ec9c4d201df7924e03
f09906f1404b60b8518e156417fa3c97d6f6e809
F20110320_AACFOZ bhat_s_Page_011.tif
7c881a8236a84af4e847a2fd31789924
ee7ab1ccaf397707a4fe30ae149ccec26ac8b16f
F20110320_AACFQC bhat_s_Page_046.tif
5e5126c0872b74286369916eddc7f8c2
77e17aeee58fdd8030cbb51124470bb3e265cf2b
1054428 F20110320_AACFPP bhat_s_Page_030.tif
ffc1a72635301963b525a1261a90fba2
42daa849f54f693c1ea1d0fe7bb21fd296b47284
F20110320_AACFQD bhat_s_Page_047.tif
bd803da46326793e94058aa36bbf66a6
5987ecc9e82f1f67770b8323bcc1a14b76e8d68d
F20110320_AACFPQ bhat_s_Page_031.tif
1b52e9a322111a36c133cdd057f4e017
b85e35d9e7edd8850d46317aa325a94094be53ca
F20110320_AACFQE bhat_s_Page_048.tif
784451ef6a6c57e9b6c2813fad75265b
1a97f171432137129383afda6bc22dfa69177465
F20110320_AACFPR bhat_s_Page_032.tif
fd9529ae862eb186be4781b40f1f3a0c
b4aed8a27fd85c00bddd4778e248b058c73a51a6
F20110320_AACFQF bhat_s_Page_049.tif
6771570c7da673aaf49355aaca1a55b6
0491720282f9f03e5ea75c5a8a71cc8cf760daae
F20110320_AACFQG bhat_s_Page_050.tif
77a47040ba5db170b0b0fa9fc6b9aca5
f5a904d7c929a8aec8e97c4561a32bb2980fe8f0
F20110320_AACFPS bhat_s_Page_033.tif
4e0926d4d968239b603ab5f9c8f22ccd
7b6e21dec715ab97173c2a4289d72a18e86cc0fd
F20110320_AACFQH bhat_s_Page_051.tif
8de64e0d07bd8373ecdb39166d8a2e4f
94ffb3a8b9a4459cd8cbad0568db56d077ca6935
F20110320_AACFPT bhat_s_Page_035.tif
e965dd282a6df77fe1319d5c49404165
d6f9d8581e8d5cb45716d232c3aebda30952644f
F20110320_AACFQI bhat_s_Page_053.tif
64e955ba1efab9c1ee120f1fcfd88dce
2aea3e40eff79ee86353e668a24218c30343901c
F20110320_AACFPU bhat_s_Page_036.tif
63d3e0083a7689fec8d0c2070a4cffa7
8f659de8776fddc9a5f940d52269e8cb8fcdb6f9
F20110320_AACFQJ bhat_s_Page_056.tif
f219491d37206b3056d2de956c3ca79e
6d765622b23f72036b51072a5f6e8aa2ad01cca9
F20110320_AACFPV bhat_s_Page_037.tif
075299fb148f2cef49075785760b24e7
031fa2900ddbee45447632fc987b1fb1493dfa4a
F20110320_AACFQK bhat_s_Page_057.tif
fe1b031e9f090f0415d6db0dd735da6a
a2e6699e7ba68301932009be9ddda4c9f8308c4a
F20110320_AACFPW bhat_s_Page_038.tif
2d00cf9b64a732a20dd864efeba5d1a0
a3c44437eca94795af70197252a5c46c870d267c
F20110320_AACFQL bhat_s_Page_058.tif
4ae048608ed9a947b296296bdfee6426
deede8ab73661f4ca5d613780f42fb4159789e59
F20110320_AACFPX bhat_s_Page_039.tif
b47a9ae3c6216d59f55fa5f04df5997b
6830e13e4620b41fc086b365e0524017ee91ee05
F20110320_AACFRA bhat_s_Page_076.tif
86ac8a913c837cc21414698ee871dad3
9ad78d93ac68bf14fc2397653b39d16d2cf42437
F20110320_AACFQM bhat_s_Page_059.tif
69ebc22697f49f472ec6f61be3bc042a
ccf7d7f4a9c92c43cb42058f7c6b9d40ffb50052
F20110320_AACFPY bhat_s_Page_041.tif
6f02b558a1b41594c5d7517e586f0438
e97ad8861b2bdf3a1b211442fe530fa32f3dace4
F20110320_AACFRB bhat_s_Page_078.tif
a03e5f5326dd247420dd03f162618a81
ff6e67ff57bd109dd0645132246581c66d9fb2e8
F20110320_AACFQN bhat_s_Page_060.tif
2e4ee86357c66b3db85599afe67c3dbd
63a638c4d2ead2d66bf34cb3305d9f61ef144de2
F20110320_AACFPZ bhat_s_Page_043.tif
9ba11a9eaac1ce0c3e36ac5d46217894
f12c3aec5b8d6f20ad6a27c83a95e57cbd1fa36d
F20110320_AACFRC bhat_s_Page_079.tif
478bf2578edacdbd43993dbd5644eeda
737d9ca37217f48f445de5159e85430a9a6c753b
F20110320_AACFQO bhat_s_Page_061.tif
2ad061518cafacc8ad9db3cfc261547b
30ed95ad06835dcb0a59870159322279130dad21
F20110320_AACFRD bhat_s_Page_081.tif
dd85d172c847c8aa65d390464200eac9
616d743950c0f24fba7ef5fe48ca020387d34a5c
F20110320_AACFQP bhat_s_Page_062.tif
fa59fdc9c9ef331988b8f58c63195e13
b04ead7289036d1cf7752d9c97ea6da9ccb64ebd
F20110320_AACFRE bhat_s_Page_082.tif
f442a0e4c3ca97722911c7deeaf50302
70e18e67c82029a0b4d3a008676a53c9194114a8
F20110320_AACFQQ bhat_s_Page_063.tif
2f72cdfa51954051274ebb3fac35c25c
78031d2efcc9e90a648057d080cee0cc17886f67
F20110320_AACFRF bhat_s_Page_084.tif
3d48994285e70500232bf1910ff8abe3
945e97d0f9b9aa08d24e36191e53b2290decdfaf
F20110320_AACFQR bhat_s_Page_064.tif
2f292ad80fc9b2800728a5bf947aa1a9
9eeb60903aef70fd5fe627b48130561986ed79a4
F20110320_AACFRG bhat_s_Page_087.tif
defdf355f21f7313f8183d819e999851
03abfe133bbdc7c7e7a63f091bd663b20be5b67c
F20110320_AACFQS bhat_s_Page_065.tif
7198da2cabdb961a5e1da28e3bbd22a2
dafcc3332b109aa75a20bc53cd6c0c797f1526a5
F20110320_AACFRH bhat_s_Page_089.tif
c2e864e8f090ccb6a33d46c31de2fe91
9cfaf1eea2f89817716e12580cfea756c7fb9cec
F20110320_AACFRI bhat_s_Page_090.tif
31285c808ec27238999abb2d65676e83
7c2549a5754b44948b27d6548b85b59b04b5b739
F20110320_AACFQT bhat_s_Page_066.tif
9581fa948095f8cb204b98ea1e32b8c8
0a3d1ffc1b09ec0fe4da9154bea0c53d6de3307c
F20110320_AACFRJ bhat_s_Page_091.tif
307fbdda3de3e29e836da51f9cf181b5
475bcfed12fce16fef89f074debb3ab7e1ea542e
F20110320_AACFQU bhat_s_Page_067.tif
e56ae5ddbbcfbce45c213ddee1a80a28
56e2c7676a85285c7fe74c999ad92d2a51d3db06
F20110320_AACFRK bhat_s_Page_092.tif
38a95e0ec30d5d6388b5f22a592f7c9f
4e33448d95e79e949af5381a49116eb3a84f6151
F20110320_AACFQV bhat_s_Page_069.tif
b4176acee43ebc50bcf6f3f2ea7fb546
a4d7151e23c8cbd4c463e8c5e6a6a23382df373b
F20110320_AACFRL bhat_s_Page_093.tif
d7bcd6bb0605f84504c6eb908a0b5f84
1ae777bd886895a62412bef82c3d5d3fbfd3151b
F20110320_AACFQW bhat_s_Page_070.tif
d4e3a7664c6dccb47d294509f5df2d6d
0849711ef21f9d30aa89cbed3a391a260b3e508a
F20110320_AACFSA bhat_s_Page_111.tif
1e5115109f440fd47514074cd5d4e80f
6316798280b0500b3d5466344c36182c2787640c
F20110320_AACFRM bhat_s_Page_094.tif
2cf31cb1334c3265c759a4d09909cab8
f6ea17e185e093fc54dc06178b68f05d9e692a38
F20110320_AACFQX bhat_s_Page_071.tif
aaf5d6fd92bbf774b3ddf274871c1a8b
d7219f094c5f43e40a2d48199411f417f49c699a
F20110320_AACFSB bhat_s_Page_112.tif
1f10e5989fc12281bf50e283d420127b
8342ef53e129d4ea43df1b5e836430f210af678c
F20110320_AACFRN bhat_s_Page_095.tif
5ba0f8373fbf602885b1feb65894367e
1d73ae58cc83d44135bd15441735ee62feefef15
F20110320_AACFQY bhat_s_Page_072.tif
d54caaed0fa332c450da73aead4ba841
004ae2ece8797a806c3b7c7d4b61c8f140c26d9f
F20110320_AACFSC bhat_s_Page_113.tif
28b2a08f4df5017900040199d22b5646
6c6e50971fd4bfc18201f3e3cff82d38a0b6021c
F20110320_AACFRO bhat_s_Page_096.tif
1547af21923bd53d1315f492acd49c68
03c7845907c895dbfaab7b5c3b54068a21dec9e8
F20110320_AACFQZ bhat_s_Page_074.tif
ec5cc954e26d5a8d4271afed7fd37fcd
dcf3f4068debb721bb2ecf7c6f8dead8a15a12c2
F20110320_AACFSD bhat_s_Page_114.tif
e509d5579448cc441d53cf08b1183a2d
f1f4bbd33c8f3281af12534725755644ab57eac8
F20110320_AACFRP bhat_s_Page_099.tif
4fac56e43832cdc40a83dba4f982cc8f
63bf90873da7079f47659864f8f99b1fe0cd453e
F20110320_AACFSE bhat_s_Page_115.tif
effb7f023cca974a8fb5e748f583c579
e318d98dfff43c140eb790b176c67e78392b5e80
F20110320_AACFRQ bhat_s_Page_100.tif
f3da5ac3b5a846bd20add72dbd43147b
6d4df85febc44018796564af69ce762c04769fe9
F20110320_AACFSF bhat_s_Page_117.tif
85f181499ec7436b2d970bdd89621840
92cb673d3f6dcdccf0a1a6b87fa8324e1e22e194
F20110320_AACFRR bhat_s_Page_101.tif
96938e6a4900bdbb8d847a8ae57fcf97
1f883704a6b350da653e9467725055372477129f
F20110320_AACFSG bhat_s_Page_118.tif
fb47cc0ece472b3a60ab4339fea7cc13
d6d3a3941b97faf9234365df987b412a96deb321
F20110320_AACFRS bhat_s_Page_102.tif
3f7274bf2d451ad438358e8c5eed2eb4
c9257dd0f108a521f37197c41717f9871c1a883a
F20110320_AACFSH bhat_s_Page_120.tif
b21e5f2caff8f7d083b7b8459bd3fd7e
b79e5914b0527c0d7070c90401fcf231dff528f4
F20110320_AACFRT bhat_s_Page_104.tif
5a88b7b2546e2c513e619da099e69600
bad0c4497148e62c08e41ecd6042c0f924872ca4
F20110320_AACFSI bhat_s_Page_121.tif
be496cc85db97a5b97c5a8c83cccd124
8ca5260d825448887c373dcb71476b759f354fba
F20110320_AACFSJ bhat_s_Page_122.tif
31e2fddf2e3f32fbdd5cd09323eeec2e
e5f58ccfb27390ca091e8ac85b73db8112d88970
F20110320_AACFRU bhat_s_Page_105.tif
c3ea18e67c0ff61d50965b4b987fa077
139d77b555c02897d467381f6f1eaa1f40c1b5b2
F20110320_AACFSK bhat_s_Page_123.tif
42c98ea757c8b6787419359b78214ef3
ca9466030d91fc9698c796430987165890703423
F20110320_AACFRV bhat_s_Page_106.tif
ede6671012ad771b5e3e84286d3546d7
5331cf08c5baf6b302ccd989ff5bde05fabe53ae
F20110320_AACFSL bhat_s_Page_124.tif
258ead6b7aac0159dcef747b06687488
99087b590d67d7542b013b69f0459aac780dea66
F20110320_AACFRW bhat_s_Page_107.tif
a7f442efd3d03ca43374acbe354b224d
70dfa19284c608dfd17614a5324e28a8e046fc30
F20110320_AACFRX bhat_s_Page_108.tif
230f529378dbc3e5b298d8a3235b30c9
ecc485675169812fc1267bd86d593b80ade4d32f
49728 F20110320_AACFTA bhat_s_Page_014.pro
f869ced7823bd340611c67758b9af0e5
54b464e310a12af10e21ec76feb5cc125880aa24
F20110320_AACFSM bhat_s_Page_125.tif
2730197643c5ca66f49b4eebfc761208
25ca692ebc8a25417af43cda3a17e1e869c224de
F20110320_AACFRY bhat_s_Page_109.tif
0885546712a290fdfbfe1e22c66cfc5e
a6dd974f53984dc26af559edadf94d46a828cbb2
50262 F20110320_AACFTB bhat_s_Page_015.pro
40297b5d52b4c76cd8e12333bd8f0cdc
900e6e8843acdabbaa61f76a448f358d8f3de9c2
F20110320_AACFSN bhat_s_Page_126.tif
b4d3ddd94e9bf9f85f9df6fae2ad9635
0b7773cd0b75696c736ff8798c25780d2762cd5f
F20110320_AACFRZ bhat_s_Page_110.tif
87ec0d7bb17a19a727cb399ed7459627
1eb4858aa4101480dec1318a9affe119beabf1d0
44863 F20110320_AACFTC bhat_s_Page_016.pro
db3de2071057dfbbfba82971a163ebe0
1eff6d0421cf4106ebff191220dfd97690070cbd
F20110320_AACFSO bhat_s_Page_127.tif
0b02ede408599ecbcbd0dbfe8d584f6f
8283dcd77beb229bef9f8f1d315de0d29c145aa9
50340 F20110320_AACFTD bhat_s_Page_018.pro
88ce1e567687930bc1d11ff4710cd2d6
8849afc256507ec23e473ccf0458b0804cdeecca
8552 F20110320_AACFSP bhat_s_Page_001.pro
907334fadd3a641d2d74c489d0fe99ff
50eda3f8edf35f3a364a0000e5fb233e65846833
46744 F20110320_AACFTE bhat_s_Page_019.pro
e9211cefca55c4688f53990c28ec4f65
e4f67898368c86ea8ea807adde94f1dc8be5f3aa
1095 F20110320_AACFSQ bhat_s_Page_002.pro
b0b8f5335440dff8ce429e439aba1573
af70bd93e9fd7e990a746e97767f368846c9cc28
9871 F20110320_AACFTF bhat_s_Page_020.pro
df9df521cdab0ab973dfbaf6e9675f73
342ebde722a406f0aad2733916ff813a51eeafbe
2495 F20110320_AACFSR bhat_s_Page_003.pro
493b5ce03c7ba393f86758b05da25f9c
3ae9cdd56ae2014cc42ee3a7566afbfce57da426
39756 F20110320_AACFTG bhat_s_Page_021.pro
3136450e36c4a2e1a7f0015b1ad29358
9d9050a9e3d7d5ec26af8c1ce4d37d0a80a970a2
45664 F20110320_AACFSS bhat_s_Page_004.pro
a0c7e770c569f3587e349a466af43d50
a0035467daae217fedd2627b15d9ed108693f6f4
50371 F20110320_AACFTH bhat_s_Page_022.pro
aeec7c9d2429a781a1fd22b03d05a0ca
85228085d9491eecd49b2528ac6d0545f2da2ec3
77746 F20110320_AACFST bhat_s_Page_006.pro
b7555136dbcf6cfef7bf9174634ccde4
d2d7145383c2c40c8160e06a3afcb052dece3260
50827 F20110320_AACFTI bhat_s_Page_023.pro
643d59492d3432e56923ee22d3cbb34e
ab0c821945c5b32109fec60693f40a83bdb9b120
104176 F20110320_AACFSU bhat_s_Page_007.pro
b4c7d64931faf451955fb6edf4b8c500
93fe1d3f544bc4e40f604de378edea9d8a247d2f
48567 F20110320_AACFTJ bhat_s_Page_024.pro
1118334735d58715f01d4f57102038ec
263e88de7995fe53b9f4a9506bc2459b9f999a21
46828 F20110320_AACFTK bhat_s_Page_025.pro
dc772f45416bf3ce23310be55442024e
0c41dfba88b47994cdcb165bd8846c29128993ee
15882 F20110320_AACFSV bhat_s_Page_008.pro
7f51186f36f239f0a6e2a108a4e3c05e
eb93428d64d69459f4a1c9e714212582f9339754
28349 F20110320_AACFTL bhat_s_Page_026.pro
3fdaac25764bb1f6912cf0cec4ea610e
4ce9af702e723127a2376dbb3986dc13291948bf
57023 F20110320_AACFSW bhat_s_Page_009.pro
eccf0ce0b652974703ff1a256f8c89e7
fbd6ae73ce090d6cf0e53e779f444a44687b5462
51475 F20110320_AACFUA bhat_s_Page_043.pro
9c1e77b573bdfd6b1f40e7bf41f69454
0893e35c9de526db5c15d80d02e58cc7689c7f89
49179 F20110320_AACFTM bhat_s_Page_027.pro
a4d322b90aa9eec9fbb56f3bb1c39002
4f76bb09d18137b322e250dfa40ef17b7c4ec756
55923 F20110320_AACFSX bhat_s_Page_010.pro
7206b91cb0336823be0cdfab8052fe89
461579871f041595d28b1cf699657537d267b532
48101 F20110320_AACFUB bhat_s_Page_044.pro
e519c0944561e90ff28ef129e6b2ae2c
ff3a2d855d86aeaf39826ea284499a0d34a703c5
49560 F20110320_AACFTN bhat_s_Page_028.pro
1b7e3da90004671e327145925abcb301
1ab1ece8449ef93ad1ac633dbfb2ebf1585a9068
38510 F20110320_AACFSY bhat_s_Page_011.pro
d6f85e2cf58bc3e3c6160fcdb868d859
b1128ff3824de1d5479dcdecfc395aebfc7662db
52186 F20110320_AACFUC bhat_s_Page_045.pro
14fa1fb106fabf032f267448acf01dde
267463ee208458ce3480099f463b84ca6740f012
39460 F20110320_AACFTO bhat_s_Page_029.pro
b0ed37f10fe5708617f3d771ee9933da
c0b4f53cc54b5a2dfce70c6dd953ea3aea320557
33609 F20110320_AACFSZ bhat_s_Page_012.pro
37abe587ea29e06a8eb98ca93f6d03d9
3832cfcc4e0ee2f6a92f3868200866745a9b0959
37446 F20110320_AACFUD bhat_s_Page_046.pro
3cd46ce7dfd13cc36c7bd6f77b2b1562
9929b7f643ad78672a6ca82c0b527e98798e54ee
46615 F20110320_AACFTP bhat_s_Page_031.pro
197758f851f8ad6b599a3d846c7c33b8
771fbb1aabb9822068af07594c89b34813eb9e62
42924 F20110320_AACFUE bhat_s_Page_047.pro
96501d490e6e7d7333b070c6694be4cb
197988ae857c6ea71ebe1e0358167d48d5ce33cc
16695 F20110320_AACFTQ bhat_s_Page_032.pro
0615ff7bfc08b006acdfe58816c2a466
68d0c813de773604d1208e86fa0189be987ab280
49741 F20110320_AACFUF bhat_s_Page_048.pro
565d2cefc247aa212947ece53d6ac7b0
b8837121ebcfde94c88aa437f7f01c05fe972871
49348 F20110320_AACFTR bhat_s_Page_033.pro
a395e6755a0dea5b4780933631ed7dff
ead2cc06a9045d1144c949a21880a2c50664a0ad
211 F20110320_AACGAA bhat_s_Page_110.txt
ca055a08a3f9eb8abe466c4e2a22f4c7
21b31acb9bf37b7e0bd9d44d87ba08040f4b7fb3
52190 F20110320_AACFUG bhat_s_Page_049.pro
ab3d8a4878c53922bbace61415f80e65
c079f2de010c91295f722997f35cc789926d3c18
49856 F20110320_AACFTS bhat_s_Page_034.pro
88fd47b7de67575c0a88f4b489c4cf7c
c0e52826c68f41ac59001c3faf4bcb744df4b0b1
1888 F20110320_AACGAB bhat_s_Page_111.txt
c76b4f5c469c70278e79903301d421bd
7ac0b31fa59b43e2fa00ef948639d642f7ec1969
49978 F20110320_AACFUH bhat_s_Page_050.pro
0835b647eb895b455755bb0a64f01e8f
d6f1ac6386eb36d1726d7d9ff94e81ef186f3448
61757 F20110320_AACFTT bhat_s_Page_035.pro
3daf8e504fcaac3e453101ea8e4eeb07
c99bad1e8f746e39179e56f65e4d66ef89ab8c66
54827 F20110320_AACFUI bhat_s_Page_051.pro
15145ac120cac0684375a8512bf235e6
f3db18bfbb240fd47ae37b7fdb7e422904fdf7c6
47990 F20110320_AACFTU bhat_s_Page_036.pro
fd9614a0d34198ad18c1bf89c0ad94b2
d5416f22644413e4ba546a5bfdec1428b8147f0d
1891 F20110320_AACGAC bhat_s_Page_112.txt
73c584e49da0e3effc66492a9198b986
f8d22553e405588371c52cf6cad0cb06d414ced4
50334 F20110320_AACFUJ bhat_s_Page_052.pro
099b2f43da4b55370c8a7483a9b60799
fde2971f15d5d63c2fa5d3ae8f55b196bf6bed83
12221 F20110320_AACFTV bhat_s_Page_038.pro
40aeb65bfef6b91e4dc3d1c27c61c30b
27e9f59e4f762084c00db695721a1b611c9bb704
2285 F20110320_AACGAD bhat_s_Page_114.txt
2eae4cc493dd3d3e69ba59a5db042ff4
997f5e91e393b174c756811e22e7fdcf80043cf2
51767 F20110320_AACFUK bhat_s_Page_053.pro
407f0f0b015ce8b140d1fc8018e92368
42f057abce78de9bc2b8b6bed5b54f281d37f5b5
2571 F20110320_AACGAE bhat_s_Page_115.txt
96fd982952369e79c04ff9074955c582
62f21f78ba4c980ca20df017a3ae17c02b98d8d0
49252 F20110320_AACFUL bhat_s_Page_054.pro
1f00c254c47d4a3b34b55cf8b1494818
67d38f6d693aba7aa3ede4b7464f0531213b2e6e
15397 F20110320_AACFTW bhat_s_Page_039.pro
dfe7a08f437655ed7e7b26327e899ac8
f38f8546b1b54d9e243c74687d5754424d24e537
2600 F20110320_AACGAF bhat_s_Page_117.txt
d76d31e959edad9c60d387196e1c2762
01da7967a509dabb38b0343f2fa5be1966b81562
48246 F20110320_AACFUM bhat_s_Page_056.pro
37a7729f480f55c8fdc41482c22aa66b
950d28296a9d29fdc4bd88bc9b79db4a59add30e
13471 F20110320_AACFTX bhat_s_Page_040.pro
e22cbb22816882badbe792f9ee98c686
1bb2aa52fe9573e956cb823052650a36738c2926
2549 F20110320_AACGAG bhat_s_Page_118.txt
69a1758180f57ef6bd016370c97378f1
3fe58ffaa6099bac4837e0454990b1d94c786513
51259 F20110320_AACFVA bhat_s_Page_075.pro
265abed49ffd33b3b2d57ebb78684766
cbed08b0bdb1e4a78fff4bb7dff522ebb4806986
38676 F20110320_AACFUN bhat_s_Page_057.pro
4e4d779eafb10a596ad78c9cc8582609
c5f572ae7235f48dd37b817dc3bd4cd6c71254da
48985 F20110320_AACFTY bhat_s_Page_041.pro
8e195845e3c7cf950e21fa2ba36f9fa2
0d6b6c6f60e2906b31bca50e3803af5fa5d26e51
2565 F20110320_AACGAH bhat_s_Page_119.txt
9ea2ff94e3bcceb3734bee93477bb625
52c719222402db0d5378b4ace006cf6ed53d473f
44329 F20110320_AACFVB bhat_s_Page_077.pro
d1ddfc69658b957f49dc6fc8fbc56f86
d913823a160f6cb70b0f77c564b4aa3eb167d3e7
42600 F20110320_AACFUO bhat_s_Page_059.pro
035957fdbd68d20230b63389744ec663
3f24ec95a5e7231a41617c88a6b5e51a3fc4f3b4
49128 F20110320_AACFTZ bhat_s_Page_042.pro
70e793852e54aee954c3af2c9d29d9d0
aa63e100e6c26b54ecb0adf20461dd8fcfe05fd0
2452 F20110320_AACGAI bhat_s_Page_121.txt
959a3a10fb302665216e14488e0a03de
de62a95d8a384f40708679920f22ae2b3256f565
19448 F20110320_AACFVC bhat_s_Page_078.pro
46f3e4cffb4e99083697a9637b67f4bb
596d9b286ebcd4befd7e19775a909cc8e7b3d415
45831 F20110320_AACFUP bhat_s_Page_060.pro
afea7fb050d7b7731cc25c895328d0aa
64283d7ce3c2f441a086ddfa6e64e2ab91608cd6
2432 F20110320_AACGAJ bhat_s_Page_124.txt
f76ba98da81805f5f7bce85138ec72cf
268592bf7610fd345f4a74f6e63bf10f07dac447
47620 F20110320_AACFVD bhat_s_Page_079.pro
1d554b3b5b6d47a508bbffe40730e1cd
1c42e220046eb4e5fbafc3bd4c8253650dc10d71
51167 F20110320_AACFUQ bhat_s_Page_061.pro
c8c8b1d057e7c625dd376e0eaee3ef76
fd2669c19ea06870b7473b4668a1c8c9e7150392
2635 F20110320_AACGAK bhat_s_Page_125.txt
097a501410922f5dd534e735874e8769
a12d271d17ab922c9cd5f620f66a9555bd7ea410
47463 F20110320_AACFVE bhat_s_Page_080.pro
cb97e37ff2878787223fd540673ab844
8ba749fcf112b1fe5360f2d761b99e0ca5e0666d
51196 F20110320_AACFUR bhat_s_Page_062.pro
f4042d7725940d75fc746ec779e7244a
fadb7bdf2366d32f7ccd52ef1748b668b47905df
23705 F20110320_AACGBA bhat_s_Page_031thm.jpg
7b2c3173f98e2ba49b23b066885f54a1
6a69b4aedb1729b983a8704c0b7a50d859917220
1990 F20110320_AACGAL bhat_s_Page_126.txt
3102d3ad5e875768c497d1f357f82c4c
7d9447ab9f69ef0559140b3a65427eb073d12d86
50839 F20110320_AACFVF bhat_s_Page_081.pro
513117c1ec59371eef163778f3aa61f3
829adce6e6624e9e8abd0878af56679e7a8b19d7
51434 F20110320_AACFUS bhat_s_Page_063.pro
e712746ca75d94cccb76214d2e0cd9b2
2a48164296c3b7f250cf66913f111425d78dc262
60822 F20110320_AACGBB bhat_s_Page_011.QC.jpg
61388ae7dc3c0e8837a85817d264b34e
a4501dc700fbbdc09f45c6e8df51b3c748374262
1457 F20110320_AACGAM bhat_s_Page_127.txt
3f19aae2c30727e823c5e1fc6391d900
f317a084e6d4e005ee0dcf3a91167294c628386a
49386 F20110320_AACFVG bhat_s_Page_082.pro
8658cfaeae20ef4987c838cf2944a5da
20a0b0e84f20458b5bbb8586f86eb6930726782e
45649 F20110320_AACFUT bhat_s_Page_064.pro
46c1975d2c62d57ed81f83133f202637
359cbfa35de06ce155cc4990089e4f66b1f1eb6b
65038 F20110320_AACGBC bhat_s_Page_038.QC.jpg
c35159a8f884750140086e8e5078644a
8aba8cb21714cba94d2cd9befacbd09702c9dbeb
1017779 F20110320_AACGAN bhat_s.pdf
5ba2823457d725184d74386573afe00b
ed7b8a9d94460edb3b24ed5adf83689896625e4a
51133 F20110320_AACFVH bhat_s_Page_083.pro
4662554ed3372a4aa6ab4c5af946b594
36692d3677fb6d7f2bf7edaae41e587d1381161b
44999 F20110320_AACFUU bhat_s_Page_066.pro
0b1177a0b24c9e6a0c283c163ec6417d
6e1aefed2505bb1272ba1088a4c4480a9af5206f
24735 F20110320_AACGAO bhat_s_Page_074thm.jpg
371e04f6b6e0fc7f9c174b293c1053f3
8c926f0b31b5142fe033a21b8856dc27ea4ca66f
49764 F20110320_AACFVI bhat_s_Page_084.pro
f7cb85efbea2e74a1f1a815ea764a99f
5b9bb86542a59f51f5ad19bcbf6411bc0d24e980
49365 F20110320_AACFUV bhat_s_Page_068.pro
5c2fcaa230d841746d0ad42fd555e2f4
2c998db8510405aa4f4a4f5a471dadc463c295de
83010 F20110320_AACGBD bhat_s_Page_067.QC.jpg
6682bed3d6661b5f9ba368f790ab4494
7f3c53e12f62b6531b181b82e32001d45fcfdcf8
25115 F20110320_AACGAP bhat_s_Page_063thm.jpg
97646d7db1a1110719ad674daae3463a
7a114a7a90c60a3784968d8beaa5da7f4b50bd05
40990 F20110320_AACFVJ bhat_s_Page_085.pro
56383e442b627f77e75c73b06b50ec71
2a01e9dcb81f9dad7026e08b0752ab10ae73fab6
52287 F20110320_AACFUW bhat_s_Page_070.pro
97c92c0c1ead622a05488113736cdf40
6996d7c2a6bc8268d134efe5edeb33db04e98026
24132 F20110320_AACGBE bhat_s_Page_028thm.jpg
80a68696e3ecb3aa0f39ff54053fb8ab
1241bb4fd7f7dc2a3a15abaeb8f0c291859cac26
81729 F20110320_AACGAQ bhat_s_Page_098.QC.jpg
a9c6dc8ba34651874a14423d9aecfa37
b029973e7215c5fa629af580342eeb531547728e
39293 F20110320_AACFVK bhat_s_Page_086.pro
8d83bc405c4d249c3b9b3430b50845ab
0364856a178abb9f61291e2c14da2b5b35abe188
61028 F20110320_AACGBF bhat_s_Page_092.QC.jpg
e4c333c7906f5bb48a4516b622b11cc7
f04627f2a9e1efe6407642456801e547edb78897
85737 F20110320_AACGAR bhat_s_Page_116.QC.jpg
52e5872b949286a7f17a00a57fe0cb5f
89feeff4a6b95979efc5b3301f678b4e432b4e47
25376 F20110320_AACFVL bhat_s_Page_087.pro
7cf11b25639c1e2c8f4d3aee4eb2730d
7d2e3b0503eaff5ad1b4d135360cb881761b73ac
43724 F20110320_AACFUX bhat_s_Page_071.pro
52968367cf3b786f709d10f69f2f2992
9c91983918baf4ac548e325188fe6e4a3a43c4fa
15937 F20110320_AACGBG bhat_s_Page_103thm.jpg
abaa6fc0b9b35823a3b1770bb40e3814
409f78c9c59a35e7ab7fff60e235a4a4831bd167
35371 F20110320_AACFWA bhat_s_Page_106.pro
e210536c4840c41572610da667fd1edc
8b984dfae23b00b8fc6d73c155a95cebb541b58c
36672 F20110320_AACGAS bhat_s_Page_111.QC.jpg
b30b03544b6a6e200068370f91ee9df5
7b9bd40e6897aa75dbd9f7e610cb8d6d7a577347
44392 F20110320_AACFVM bhat_s_Page_088.pro
84439987874c66797fa88200107bc3d1
9efecffb8ed5c49c20e11bba89440b737c34502c
49696 F20110320_AACFUY bhat_s_Page_073.pro
5d6ca28176cf8e820f23af66654443e5
2b5b692addb937dacd4ac4ed5e05a3dfdfb32b86
118277 F20110320_AACGBH bhat_s_Page_007.QC.jpg
08bf990ab979d614e9c00d1d9ce42a7b
50911478e28ebbeee9f0456a6c7807af12a6d901
39804 F20110320_AACFWB bhat_s_Page_108.pro
b45e5e43021095d6454de29e5b446d4b
4d1729b1dd41a1fc292136f75bb3007afbdac90b
45164 F20110320_AACGAT bhat_s_Page_087thm.jpg
0716df4a3c5315ef629a0192a99def10
8997616de62441f453254bfe12bb3b15dced85d3
35043 F20110320_AACFVN bhat_s_Page_089.pro
ee647e3514bdb4db01a01e6fff095a20
12f979a80e010b355869299a3d3797e037d3af0d
49059 F20110320_AACFUZ bhat_s_Page_074.pro
bb1c210eacd79e47ad614d827109d342
3d2f05cfecc3ce17c7a09c2e57da790437b7963d
84366 F20110320_AACGBI bhat_s_Page_087.QC.jpg
a9ab05f90504a9a1906c786e3399efdc
9001489bb5a9dd2767a10a3f9ddfc80cd31f1845
32029 F20110320_AACFWC bhat_s_Page_109.pro
63c275295bfc6302cd5c15e122dd7459
0ab02fd467ae533952891f39215d12c7c875f097
32882 F20110320_AACGAU bhat_s_Page_008thm.jpg
51edf7e950ee0d3fda9c43c5aad37d30
b4ffcc172da8e94993f6b16ff121d5ba27018e79
24995 F20110320_AACFVO bhat_s_Page_090.pro
812daa942c4440450bd142b5131d7140
cce28744e0c1d999218e503095e6630cf2fe5361
24243 F20110320_AACGBJ bhat_s_Page_075thm.jpg
111670b70470afbeee75b5d3dc4865c1
ed3a6d537a188080fd054a111424f1e4d028433d
4378 F20110320_AACFWD bhat_s_Page_110.pro
f02224c6c1a32bff7d5a3be9e91b616b
c8e362b7898431b072519a4d7ba41cbd2c94f5dc
15347 F20110320_AACGAV bhat_s_Page_107thm.jpg
d2957b3265eeaf3f1ae575f614e69151
3016a2a3bd3b6050368a0afe45bfe8a194a9f309
17808 F20110320_AACFVP bhat_s_Page_091.pro
c17f464d7505c8a5d63d5368cd1542a0
a75387753e2381d9b4c9968b6a8adcfb8e3e120b
42376 F20110320_AACGBK bhat_s_Page_051thm.jpg
225d196182aa7b994b843f1107c821d5
f9b3e9f8071772838d3e41b89b0a5084ab294fd9
33721 F20110320_AACFWE bhat_s_Page_111.pro
1a7e15aa5cdea914d35e8115b7b07bb0
fe933740aff8ef2de5e1e11d8aa7fc1dfc01017b
21221 F20110320_AACGAW bhat_s_Page_036thm.jpg
e8154d3ea8c2ca37a2005ff2c361c4bf
ee79f16cff180efefb8d2fdbcc043b44d590eb4a
48398 F20110320_AACFVQ bhat_s_Page_094.pro
a101374821ed7bfe8b0e04483887edab
6843f3d90ee5bf308d36181d57741863204cd53d
65426 F20110320_AACGBL bhat_s_Page_058.QC.jpg
e965b426c789bd80ac47c4d94689a2bb
fe03543fc11051828dd14d25fe93cef244695a7c
34593 F20110320_AACFWF bhat_s_Page_112.pro
bcad49e1caaba4c560c5e84b9f50975c
a8c2542dc99735427de4914998cbb3b463823eba
63095 F20110320_AACGAX bhat_s_Page_021.QC.jpg
d29b6dfdf772f4aafa8229916af87e9c
fb99637f1e2999651d649d5458f8ef9bd6188cab
52072 F20110320_AACFVR bhat_s_Page_095.pro
45ec9aea6df2345e8f779fc5b90cbb7a
d4072b03d2f8ba6ef75dafdb334b56f3aa0e175f
83064 F20110320_AACGCA bhat_s_Page_070.QC.jpg
6ff77995570638ce2e580c64172a4dbd
973318721daa65afaee2815e75e03665bdf2ce8d
24837 F20110320_AACGBM bhat_s_Page_093thm.jpg
87d7c6cd0eedf12985c2cd2327752eea
44c6db8fd11aecd4abc50ca7076aad080bba67f6
14959 F20110320_AACFWG bhat_s_Page_113.pro
e6ed43e1c5d3fd650306028223b6d5f5
21352fa4a52a25e0078cc4d764e9da2e2301f0a1
110650 F20110320_AACGAY bhat_s_Page_009.QC.jpg
065aab9fd7e80f7d051704eaf1150ecd
6df741b4242589acdd050e309244bf5f9c3233f6
48653 F20110320_AACFVS bhat_s_Page_096.pro
6ef216811182aae9e036b7bca78a842b
ca8f8650a391ad9fd0f71bf3b055043fe50be780
31485 F20110320_AACGCB bhat_s_Page_030.QC.jpg
9b3361fa4d1a6c7acff3a98b4c850640
08dff00bfbdf284511e7d71b8dcfefda9f89cb2d
24579 F20110320_AACGBN bhat_s_Page_014thm.jpg
dc34f04ad9035c35f40ffd9206ad653e
f74085f96313f875876dc8f08ed3b81360f55fb6
56233 F20110320_AACFWH bhat_s_Page_114.pro
b6e0100bf204073fda5303945643f7e0
830f65bd44e814af05f124ba1d8eebf2a0cbc843
18956 F20110320_AACGAZ bhat_s_Page_127thm.jpg
1ad4b247c17ae251ca3e32030df7b264
44f24f4c071999264bf5870f483cde9b7fcde1b2
52107 F20110320_AACFVT bhat_s_Page_098.pro
cd734f9c72744a8816278895c8792d67
ce9e746ad364932fa3bc0987ba8f5b92fbf0dca3
24868 F20110320_AACGCC bhat_s_Page_082thm.jpg
5795f718d43b441d1c614c7a4ec1b9e9
187b5bf923b740fe8f2ce67ab616833de4555bb2
25343 F20110320_AACGBO bhat_s_Page_067thm.jpg
485c0cfbc003b1cd2fd9f5c9e3e77d6b
4fd8a5f9398d0dbb0f3722604d04c2a7e444368d
63117 F20110320_AACFWI bhat_s_Page_115.pro
783befb5c2d91a6d7cd52438bf6cd078
a11956741e8b43f10253f9701daf5afc6e5e2b16
49346 F20110320_AACFVU bhat_s_Page_099.pro
f69cadfcb47f96d563591d6074e3d2ad
399e6e321e2414ce68c534303154435c48b0c441
23141 F20110320_AACGCD bhat_s_Page_072thm.jpg
30b47f62bc3dc23cde7ef5265c336d27
939c3f259f6907d3d5212b16865874e42e7edadd
21985 F20110320_AACGBP bhat_s_Page_016thm.jpg
896d51361fd2465e72fd0edf4a546221
7fb7222b880b1eefc891cfc058e677b66101edb3
62928 F20110320_AACFWJ bhat_s_Page_116.pro
e54e828871806de82d700b35a4887dc9
06a6a584a4557ce951be56e0b7e084543046a996
30941 F20110320_AACFVV bhat_s_Page_100.pro
197b9b92624bd34da58f53cfab5b3e9d
9d9dc1e46714bcc17c2ebc0fabd1bbeddab0b569
81952 F20110320_AACGBQ bhat_s_Page_083.QC.jpg
31580fae1f9ae7f94e0b9325213ae066
12fe3f2b7ed9a0231e469aa43e11363f60b8af6f
63915 F20110320_AACFWK bhat_s_Page_117.pro
23cd6c8f967887126715c9b86474fc74
fe4d4741f096606c79993b25f9bec07c56bff3bc
41522 F20110320_AACFVW bhat_s_Page_101.pro
5ff6e253b7e08da9f8405be1769195b3
673e36a2a08848e2bea03354d1ff174c602a6394
51229 F20110320_AACGCE bhat_s_Page_110.QC.jpg
cfabeff110db386c4bf2f93a16d25ae9
eb8f01b3e17c13fc3b2dd673255fbc213cd06418
73044 F20110320_AACGBR bhat_s_Page_031.QC.jpg
fbcbb8a98dfdc487bc17078376978e4f
90f10855798ee66e5c64b4229f5b5d732484801f
62454 F20110320_AACFWL bhat_s_Page_118.pro
f45574d5c455bf19ea91b815a2d1ef08
c172075006dc926fe337c461d81ded9dee2d8ae7
47318 F20110320_AACFVX bhat_s_Page_102.pro
e924a23d2a890666e6380014963e017f
1991d1352e86db7ef5199f5b66f0d591fcf3f294
69319 F20110320_AACGCF bhat_s_Page_077.QC.jpg
ed7f56b44b10748b327d59a94aed2e0d
fb47db329a5935c13942633e51dcc9d9b9196f86
1818 F20110320_AACFXA bhat_s_Page_013.txt
94ee42630e5af207868c672d721e4f91
fc0c8935a8af4688cacf02934dfd896d14da89b1
24777 F20110320_AACGBS bhat_s_Page_018thm.jpg
0118c3cb0cd32a6a3736170bd13a60ca
11191d00fee3513b69bff890656c595a13c8b6ff
63140 F20110320_AACFWM bhat_s_Page_119.pro
600179a735e6043605d87d0c26c7150b
46af4b3ff222008dd058e5068cbd7651fa65ebdd
24889 F20110320_AACGCG bhat_s_Page_062thm.jpg
338371a63ea597ad8471d0c8fedaffdf
504d052d2a9cfb1e3ef69c1e3f8603c92eee5946
1958 F20110320_AACFXB bhat_s_Page_014.txt
0a6ac09fe4e0babc1c8c930f8b10d692
f58b239609c926367d0222b4982d3b7bdb78eb43
64895 F20110320_AACGBT bhat_s_Page_039.QC.jpg
062fd5ffb546efde0eb7c93b6916586b
95bf726fbcebdcc683094c8647d0c3410b778609
60561 F20110320_AACFWN bhat_s_Page_120.pro
ef0835f13c75f5cc7fb6c2e88a3a1d44
de506799f988beb23d0a30726b3c4c9138f36570
33456 F20110320_AACFVY bhat_s_Page_103.pro
5c7824f40b59ff37bf871a168ccece9e
7ca4bdeea5133348c112f4641e90aa52eb09e00d
25159 F20110320_AACGCH bhat_s_Page_117thm.jpg
6d1d511c69ff9d98174b335b0cd0a06d
e00e001d4a2570ee08cf42afbf41157b8e7cddc4
1973 F20110320_AACFXC bhat_s_Page_015.txt
732f3b47a2c239ac8d761b27324b5fe0
27d9eb33e025b44ff85a09160cb02234a16cbb9d
22242 F20110320_AACGBU bhat_s_Page_056thm.jpg
3420e2fc9503aa28d9c22f04effe96af
8073d226154c4e4fdb34ca2e57e8d45b37562d23
60124 F20110320_AACFWO bhat_s_Page_121.pro
0f26a58a91a11bc1a1c9ec47b2439eeb
0be9f06323a55c749a025689b00f37da6c55d7d6
38972 F20110320_AACFVZ bhat_s_Page_105.pro
f1df93553b8921013eb92b34d42c72fd
40da71640b828c4f802eb7ea4b76f02682473c5e
16627 F20110320_AACGCI bhat_s_Page_105thm.jpg
1925f257770b339390c40be9e60beb0a
9b09cd953b20f7aa1a0b773da4b37b06d5feafd7
1786 F20110320_AACFXD bhat_s_Page_016.txt
44e291a9e7d96905f1ef1ef26f0608cf
6449fd88b10fd594a06dbe32217f4690c9b5c31d
18213 F20110320_AACGBV bhat_s_Page_020.QC.jpg
b26240fe73e1d30e8b484a4a23d964b1
a182b49bf5981af7427280396e62cbcbb61bbec5
60060 F20110320_AACFWP bhat_s_Page_124.pro
685226ea480f8d23df22a17b21f1a399
716a9e6298140fe6ba3fe089feedd35656b8ad50
82876 F20110320_AACGCJ bhat_s_Page_049.QC.jpg
b3daad7f786fd0babd254d4aabb21e0b
344b7264ce4fcced582b74724cbfef75ec2cf1c7
1954 F20110320_AACFXE bhat_s_Page_017.txt
a667f0eda577aec943d004a9386dcb05
8391f81076942def50b64533c5ea71a94139d635
19460 F20110320_AACGBW bhat_s_Page_102thm.jpg
bd5bb8a945b1d49a1a14d96268567a18
df6ce80fee3ef4846459ba10be0871cd5ad36610
48479 F20110320_AACFWQ bhat_s_Page_126.pro
c7607086c376134fe894a42ad8568f20
68779e96277cd904425589d2e0cb248b67a4b4e7
11922 F20110320_AACGCK bhat_s_Page_111thm.jpg
c5dee85d23f31c05e25c3799d699fe74
586913371b1fcf99104da8fbeb42fc2a5ee3e4f9
2011 F20110320_AACFXF bhat_s_Page_018.txt
3de534d8967023070b7c5772c112efc8
73ffc8086e44908287701166864d26caf081b58d
54423 F20110320_AACGBX bhat_s_Page_105.QC.jpg
e54867ebacadfa1ec13922d7f7a3bc1e
3070993fa83761bf30c9f47050dd7ff7a6b38bea
35620 F20110320_AACFWR bhat_s_Page_127.pro
aee7a6196da3fafb9dbcbd1fb216c879
5bdc18f71b951bc9e7a23d3cfaf866ed3cdd418d
38409 F20110320_AACGDA bhat_s_Page_032thm.jpg
8d75ac035cb7b425e720a4408bf8dc76
b9ef0d9b46c5b0fe712af91e0664018e59bf1461
45470 F20110320_AACGCL bhat_s_Page_112.QC.jpg
4fd3d27479becff1b5a48d4bfe5d8867
f8ecac0a3fe4a692afe587cf2e3fdb18126cd2c8
1847 F20110320_AACFXG bhat_s_Page_019.txt
de4fee4cba289272c0a38ed1b0919975
c6eeb1b3dc8f50f17f90e405f518d3ff07feda0c
62190 F20110320_AACGBY bhat_s_Page_056.QC.jpg
c3c90bac3cc8efed0c72cff4a0846d46
677ee34ac43588b89e4dd86313bd42ab2a05f005
462 F20110320_AACFWS bhat_s_Page_001.txt
64246b1997e1c32b4edc176a4092a239
c68b6d38f82e09fca756daf8c8de430083c1e93d
23868 F20110320_AACGDB bhat_s_Page_025thm.jpg
eeca6480972910904bc2145494700c67
beae72a5e7e9d1681b66767935171e39050c24ef
10611 F20110320_AACGCM bhat_s_Page_030thm.jpg
b220e4bb8fdaeeec09e41fc50c818963
d67e40d21c98d8fa7fb9f50b1d3dc828b7628712
441 F20110320_AACFXH bhat_s_Page_020.txt
658aa9184d3784ecff8715e969dabc26
d01b9351af19a3f46106ece404a9d1aaea6d6ece
82653 F20110320_AACGBZ bhat_s_Page_124.QC.jpg
4c0d329af10f99e48a6e26dd9bd58b9e
b588ac68b09a418763f50c6c13029e860c1a3124
109 F20110320_AACFWT bhat_s_Page_002.txt
ca34fa1b36af4e596b0bf80bac706749
038f3677fa543e3d36c25c7edcdbded1324982ff
167231 F20110320_AACFAA bhat_s_Page_092.jpg
238498300788758d33c3ed70eed7b68e
d19d4303a30fa6329058b6d290751cac809621b7
46095 F20110320_AACGDC bhat_s_Page_026thm.jpg
dc08f616dee82fb37fe1e0d68ca19ed4
5f4bae971553685e61e6dfdab3e3ea295c07ba9f
23936 F20110320_AACGCN bhat_s_Page_060thm.jpg
925a0a52d9e044a46581b1f6eceda173
5cb13bb053212b293831225ff92ec32756bc3429
1692 F20110320_AACFXI bhat_s_Page_021.txt
a2209e6c6015348191b37b8fed893e02
d0f236c7ed549aa1a86c8b771ae34ed97fdd0dc7
148 F20110320_AACFWU bhat_s_Page_003.txt
76f3d141ba5073cbda7dfca53808184c
a971af48c4237e2c847f40d3daa49d6abad9a264
78540 F20110320_AACFAB bhat_s_Page_027.QC.jpg
a2b6a46ce74215ab62ca06df96f7f89e
21e494e7b756de0eb83105d5da169978d81a8c84
9589 F20110320_AACGDD bhat_s_Page_005thm.jpg
c4644aac0c712f4c46d2871589fae09a
f0fd532eb42b756c9e1c7527a31de0da18026a6e
78539 F20110320_AACGCO bhat_s_Page_034.QC.jpg
8552e2b74636c3d9d1fbd47339b6cdb7
a79236abfa73edc832d595ac95d07274a8d038e5
1837 F20110320_AACFWV bhat_s_Page_004.txt
5f5e4429dc92c3c0f068e648a28db1b1
107d2181e3a29daa7aa638b3c8a4d9e343065e50
23895 F20110320_AACFAC bhat_s_Page_055.pro
8f77afd5a89169a847f8e57cddab9c9f
c48d4085bf26d303d778cca1c5e3d4f2c2edf7aa
105880 F20110320_AACGDE bhat_s_Page_010.QC.jpg
63c266a26a170f49bc0b3ae34baeea95
deec2884a88f8ea7d9302523cda55fb73bf39049
18013 F20110320_AACGCP bhat_s_Page_012thm.jpg
4c9a71366938aa2e7a512fdd5ae0af2a
3ab1fbf503690c000d11768b515d7cd0e310b94d
1882 F20110320_AACFXJ bhat_s_Page_025.txt
ddac88eb9667b4d3c1d3764cb15e1e1f
9c5b64b1f33061550c763dff1f5d040c67da3de9
3222 F20110320_AACFWW bhat_s_Page_006.txt
be555b6d03b2cd0c02270b594ddad858
5f61ee02c1000abcd06003162561ef44b8ce5c6a
77917 F20110320_AACGCQ bhat_s_Page_014.QC.jpg
414b75e0e144e4694954af1e8c107174
d9e8344a8eea017abd60759d3d4e1c96c7a07073
1150 F20110320_AACFXK bhat_s_Page_026.txt
940d37cc0988e55bbfbeb76dc545acb3
4e1646cf9c80786f72e6f16ba7d30221af4f7ece
4259 F20110320_AACFWX bhat_s_Page_007.txt
e785054e184893456394f014b3722170
699e46a6141f709c866eb019eb07f051563cb269
78517 F20110320_AACFAD bhat_s_Page_015.QC.jpg
c9c856cd795763094a60fb9f25819bfd
4819c8dbdc894c6fd0da8020b3b2f19f17dcd171
66270 F20110320_AACGDF bhat_s_Page_085.QC.jpg
46fdaa15d8fff99c7abfdd5959eb0aba
bd8d55546c5306f881a28dd07a9f36530277b914
22980 F20110320_AACGCR bhat_s_Page_061thm.jpg
b8fa869e9428181c8afc56b3462bafdc
e76eb862f95629ede4455087edf9f479c283da20
1937 F20110320_AACFXL bhat_s_Page_027.txt
d87258846c7c021bfda59a54f1b4d069
8566b254b0e3b0ba48802987a234738bb3c79d44
612 F20110320_AACFWY bhat_s_Page_008.txt
dd591484ee37656059f97d5c0cd453c6
71fa842afa3b1722e809664fa158c7b5b5cadf73
35099 F20110320_AACFAE bhat_s_Page_110thm.jpg
618a1a081147bfbf7f676291c8281773
ed8e1d633d7ebee943fffbee797d5928fdbde586
23301 F20110320_AACGDG bhat_s_Page_080thm.jpg
2259522060caea808e565bb91862048a
0dbf5d84c02fc7da557a81d7d12fe052d9ca8fee
1491 F20110320_AACFYA bhat_s_Page_046.txt
a54ebc67257400cf0f7e8cda58538bbe
d11e1c17f407415151727ff86dcd19aeb19e99fb
17121 F20110320_AACGCS bhat_s_Page_104thm.jpg
12e333b6fc0a9e5ea93d0cb91c8789b0
f5a63a6bff487fd712432e7b6b2914c0b4dbccca
1989 F20110320_AACFXM bhat_s_Page_028.txt
37299bb88fe6bdd118729793face9321
48f6df62d96d3cb7216d47277118f5ee9a0b2f8d
82263 F20110320_AACFAF bhat_s_Page_120.QC.jpg
35464932e939ac1d745091f6c86297c2
240c85d2ec5f0206093cc5c917c4a7ac435325a5
24656 F20110320_AACGDH bhat_s_Page_083thm.jpg
84f2efe45a4d7478a8c56355b1f70e0e
bdf6536cb8b0601623563c129a47ba66f8557ee0
1959 F20110320_AACFYB bhat_s_Page_048.txt
dc77a0afb1e7cfa93409fb8dd801840c
5e458234f93977448666e2d759e65b780a7ab90e
24823 F20110320_AACGCT bhat_s_Page_096thm.jpg
1138a0b2566666fed64378a1caf82abc
d36bc73a6dc8015b37e9301aedfc1e43c2a22f6b
2008 F20110320_AACFXN bhat_s_Page_029.txt
2fe67540ad0b89b5b50897826e588892
a6022573d9bb961119a56e468080e4d66cc1a134
1698 F20110320_AACFWZ bhat_s_Page_011.txt
a1b6977ebf6cd039a57cc55f5723d58e
328a86d48fe8534347ee37adbfc6cbc16f60bcd0
19667 F20110320_AACFAG bhat_s_Page_021thm.jpg
9211a3e1f7c474b7ff44d580c05355b1
f73f391488108115cdf3da58a82345e06a4dae7c
40036 F20110320_AACGDI bhat_s_Page_040thm.jpg
d092e4e472e31e4f5580ad83fe8b95f4
4ec746daa2f7a03bc39e99b93e08fd077f9d5f86
2083 F20110320_AACFYC bhat_s_Page_049.txt
aadb067123d7fb25b43cd42fd2b20982
16894441aea6b8636710f82590dc2be1eec20118
7027 F20110320_AACGCU bhat_s_Page_001thm.jpg
74b5c7b3d5c5db634f9aa56d4f8967c9
8a7fd1ad440a5ca3c07acd9bbcbbb56f84be80c3
1606 F20110320_AACFXO bhat_s_Page_030.txt
51c3b16e92c540e16cda9de48b67eb04
e50e040de4893cc3a0ad6b05c9a7b71cd26424d8
184794 F20110320_AACFAH bhat_s_Page_097.jpg
45333b6b355bccdfc0dca4debeb80697
99a362bfd8fa4b983dda46d57b2daec7ece0503c
25396 F20110320_AACGDJ bhat_s_Page_122thm.jpg
c80cb1f88ecea740b0bab89a9942ec59
f87137e45b6ce4bea661294095d9d5ec4c63a018
F20110320_AACFYD bhat_s_Page_050.txt
fe89f4502e0e05a4a764b10fa5f1639a
107275f27879070022ba8eb25660dd413d6354a3
22728 F20110320_AACGCV bhat_s_Page_004thm.jpg
55b4869eb632f7082e31c74a46ff9798
aa4828234bcdb12fa46305d69352cb8bbf8170ef
1845 F20110320_AACFXP bhat_s_Page_031.txt
394d4b3e26bbaaa3d607db91dbb89cc2
12877593b1b802653001aa539f922ae76370fd84
210472 F20110320_AACFAI bhat_s_Page_028.jpg
d99ca84cbe6effd393aad302898b7846
62809c49180aa6d7e24fd7a4c4e4867a546f7562
F20110320_AACGDK bhat_s_Page_124thm.jpg
3abf53e5d87ece7173039a7b0b9402fd
45dacfcfcfb2017b3dd864bc9d90fa9cda631934
2408 F20110320_AACFYE bhat_s_Page_051.txt
eb6f029b37e42db932d71d1c6a4e9df2
023359f504038cc0d7c2c3d793095c9ebafebd02
23663 F20110320_AACGCW bhat_s_Page_041thm.jpg
5e2aeece7e93d942dd315bfe15b1fe74
f44517f0d32ed6be86c9d089efc0d6cdfdb9f2ce
975 F20110320_AACFXQ bhat_s_Page_032.txt
27e80175b9c878aa022eb1d69cb3bc85
6ffb9a01db0c5af91e1986b77bb19b4d5f319ecd
18198 F20110320_AACGEA bhat_s_Page_079thm.jpg
b8ca94157fa0aa4c0a436368a0a67326
5c30cafd050ac064e11d2365cac568d410fdaad4
31841 F20110320_AACFAJ bhat_s_Page_107.pro
5ce1dc329a0444aa2f52a56a46fab013
15cb1d6962f43886fd045c7d44bc290c3fca732b
64564 F20110320_AACGDL bhat_s_Page_040.QC.jpg
85b0674174081d72f4a5d6699ac395d7
5602444044aa7805998fc549b1f913a629beaff1
1985 F20110320_AACFYF bhat_s_Page_052.txt
8d23f25b710647e2256709250d6c673f
1a029ff0dea75d65fe50a2d609be69b5c578e98b
76424 F20110320_AACGCX bhat_s_Page_048.QC.jpg
cc8e92f3bf9921ff0aeb93259857aa1f
7f279c22c5f37f0152aa86340b5ff8119e86192a
2732 F20110320_AACFXR bhat_s_Page_035.txt
60e2f182b6cd00ee3188054485eb6ab9
4876a61a5c3d3799c16ea516eff29fb5bb674572
23154 F20110320_AACGEB bhat_s_Page_019thm.jpg
c1a69fa39bae3f00c8208ecc9a3a6166
93967a018f5c6b369ec392fb08e50c4bc6c121bc
108676 F20110320_AACFAK bhat_s_Page_075.jp2
f887907a63e5208e63bae1cdd9945d2c
56339434fc9e4880f56b74b974e56582d47a6d1f
68485 F20110320_AACGDM bhat_s_Page_047.QC.jpg
0d8aca7e2d51626897e24dd5690d6347
397d7157cf0f380978e65b5c2c1673634b0bd0ff
2030 F20110320_AACFYG bhat_s_Page_053.txt
9863fddd69c6e7b601d284084cc520d9
77e54c495b12601d9798917aed124ac9c7621fc0
47849 F20110320_AACGCY bhat_s_Page_086thm.jpg
4f759960e2efdba5c6e141ed4f4b6e18
669649615e55b05afbce86bea5f00a54b283960e
2028 F20110320_AACFXS bhat_s_Page_036.txt
ef9f64af0ff39e699dde1f7be6d129f1
c819b61fd8505ce0714e085339501d408afd4c30
75218 F20110320_AACGEC bhat_s_Page_074.QC.jpg
0ad845de1850e0f167b23ba1f9d62ccc
65f626dca96b14cd9ed54a18b4be9a18d37cc3ad
130789 F20110320_AACFAL bhat_s_Page_038.jpg
b4cb61e6860a8658055594564ee1b0e5
4245dd4044449ffb56ceac4ec8e7e38ef163f395
19871 F20110320_AACGDN bhat_s_Page_001.QC.jpg
66e02de2dbafa33332a72e9a9b77bffc
0037256799fa7c754f4360cbf0757e71d8fba996
1005 F20110320_AACFYH bhat_s_Page_055.txt
97af89c0a54ef153e74ed1d2bec85c52
7d28c1e3d33bfb98464c6d51bf91b68e39b2b0b2
74349 F20110320_AACGCZ bhat_s_Page_081.QC.jpg
d3e2479c74cc54e05e156d17d51f0e0d
a23b6d3b53836823fc92ba143e7e6abf718414a1
541 F20110320_AACFXT bhat_s_Page_037.txt
f449e8e7808a71eb03b236e2830d337f
299aa51e0c0577b920dd6eb80d10fc126e59b9f7
177008 F20110320_AACFBA bhat_s_Page_033.jpg
90b8df31477a499bdea8983be24a6295
0137d3b97f5248bb559d3a7a3d5cb2e5ecda7872
51382 F20110320_AACGED bhat_s_Page_106.QC.jpg
c71dad3918a76e46d03cbf58e107e7ed
ba678497d2f7a6ce3492316e65f927073cb8e09e
77424 F20110320_AACFAM bhat_s_Page_096.QC.jpg
c609de3ec6efb8cc468c8299eb33fffb
e9eba7917e3a51361fd774027379b233802f73b5
24843 F20110320_AACGDO bhat_s_Page_048thm.jpg
a71a7423c2c532f7a6e187d9b195a794
35c04923731d9edcd94833fa525c5c80adfae673
2320 F20110320_AACFYI bhat_s_Page_056.txt
51829666af3b43404e16b539ce9c5d00
c99b270d5ab6e2556d27e95a4189b0ada1b15c45
846 F20110320_AACFXU bhat_s_Page_039.txt
d0c1bdbc94fd6fc22f2cfbc40c97fbb9
9eafe61b24d371089a1bc89a6a14bd38f9984f0d
75774 F20110320_AACFBB bhat_s_Page_044.QC.jpg
0879f17a44e9cadfa9bfd3133885290d
d0ea8c2949d4323828b166cfea86a6a2e9b3290e
203679 F20110320_AACGEE UFE0009283_00001.xml
86e094976f27eb67f59d2c522c6a8787
8a4bc79e60ed04b800c67d5d5460a8fcaf2e23f8
201300 F20110320_AACFAN bhat_s_Page_044.jpg
8c4a1d6a34bb36415ebb9cb087e3771f
43f26b060d2bda6897959eff63f9454fafcc8250
21351 F20110320_AACGDP bhat_s_Page_126thm.jpg
27000e87171f12410bd672c3037b5b6a
cde99ef1becaadc105d2c6c819cb45e74a153376
1581 F20110320_AACFYJ bhat_s_Page_057.txt
770d1bdc48e85dbb6781fea349c1d6bd
a65580db7b45fe4bbbb991405eb1cfa7bc804900
647 F20110320_AACFXV bhat_s_Page_040.txt
f77057c50bf95e7259ad1d4199ef4794
8bf7b5f48c39a0b78e187525f77a0d043fb78140
218999 F20110320_AACFBC bhat_s_Page_062.jpg
2fbeeda0c497b7829194eaa4adf99d1a
4af5274a6c349fe6984987409338d04dca6a86fd
6747 F20110320_AACGEF bhat_s_Page_003.QC.jpg
33ae59b75595992e15a7f997d083c0f3
8e3e92a9fe5802e6e6c40a6e248c36e3351e3679
2059 F20110320_AACFAO bhat_s_Page_066.txt
e3ca98838a1a5b61f3b7eaa647794367
924976978d93a9ecedb9638d47dc55bbe58e4dec
39175 F20110320_AACGDQ bhat_s_Page_037thm.jpg
9dc356dab4ba564fbddac653e080cb3a
fa5dbf94cd714ec5f8ba588d88f0ecc456d2d2ba
F20110320_AACFYK bhat_s_Page_058.txt
a7a64f48a1c85e8cbbf5347171cf5b59
a41b0b3ca29543d42af1659a7f5191ab0d0dfa70
1967 F20110320_AACFXW bhat_s_Page_041.txt
1b25d57fc6b94795bcb6d95fb25505cc
d89238181ac6ce53b742eeeb93de36a5b9b19cb4
2572 F20110320_AACFBD bhat_s_Page_123.txt
c0d6a3ffa6fd9c968445c806690de279
d9c70a80ccee493aad791833866e80e680e647f8
104862 F20110320_AACFAP bhat_s_Page_094.jp2
047ea55741b54b98f68547f2815a524d
840a07baeeb94b9abb6c2cc54cc4831e7c496719
20222 F20110320_AACGDR bhat_s_Page_058thm.jpg
d2111c747d6808b8a0a14d500bf12da7
9c007f1d42293aa311d5c0971164a9124b60eeaf
1770 F20110320_AACFYL bhat_s_Page_059.txt
0d059560e77bda2340a122eb520216ea
4c2b7930fcbdc8dff72024d506f7aa0017c30aea
1945 F20110320_AACFXX bhat_s_Page_042.txt
b5344a73342c6f79cce4bc7a08394977
53e291fd1201848cdc42ac17301155736032edfe
55121 F20110320_AACGEG bhat_s_Page_012.QC.jpg
0d3cc1e8bbd1201128a29eaf237def5e
ce4f2ad2e055046f611a3e373735e58c55757b85
72585 F20110320_AACFAQ bhat_s_Page_004.QC.jpg
46c1936694bc131a07164c60d20eb330
3b1e307391ee6a50ec25a1762dddf3bf1afccbd5
2702 F20110320_AACFZA bhat_s_Page_079.txt
7c675d5fbf22705570c5e97a84673c74
8c306ccb37a1777a2e4eff82fe182653274bed80
21262 F20110320_AACGDS bhat_s_Page_114thm.jpg
420665dede26bab2ef5a87ccd349a2f1
d1f6ae315c8b9be493b97296f75f5f798d4a2fda
1863 F20110320_AACFYM bhat_s_Page_060.txt
84cb51a6fba544b024a316c385936d02
a8245a6ba3a1c49308e0ec929681a88d4b5a1785
1906 F20110320_AACFXY bhat_s_Page_044.txt
09381db438d8ead04765bde599113472
4ef4b07dd4369801feccd82d52699b909d5a7511
51328 F20110320_AACFBE bhat_s_Page_067.pro
7b747eb4c422a7b957580724d9a70952
8565c3cfe32bbef0c321fa039342cda38611185b
69881 F20110320_AACGEH bhat_s_Page_013.QC.jpg
79285439181dd56b80388a5c215e74dc
e9c5106dfd3290680dce73fbc0401c440771e61a
1290 F20110320_AACFAR bhat_s_Page_090.txt
b558ac4c443ad54574e6247897d8aab7
f3e2172f1df86bc494348f6dc8e158c1f4b49874
1873 F20110320_AACFZB bhat_s_Page_080.txt
5b093f0c959952aff7e46f096e7dc330
8ccbe70c47a6b83721cd6bbba3f30e05a65c91dd
95987 F20110320_AACGDT bhat_s_Page_088.QC.jpg
b72fa7d4baaaecc567be08b4210635e6
abd7c010e2c6bca2149eca994d7327fa848f04aa
2226 F20110320_AACFYN bhat_s_Page_061.txt
d3fce08a825cbe88bd2d07f1f849e353
8228e30f0ca1844747c75966086ed83cb4b761e5
2074 F20110320_AACFXZ bhat_s_Page_045.txt
45c94e70a65cfff63412a0a615c19fb0
5602f0765bcd8e8ebaad99f13713bd193e8524fd
9306 F20110320_AACFBF bhat_s_Page_037.pro
6bcbffaa3e8b282ac89fb2b8cad52ca3
c9c1a91c95ea21574e6c9220190368a9135c3dd5
71670 F20110320_AACGEI bhat_s_Page_016.QC.jpg
608bdd49687ac121bdae4e4298c05c98
d23585ffc71eeb79c7b6f102fc0b1d8730297fb1
171845 F20110320_AACFAS bhat_s_Page_085.jpg
0cf46b1ab8fa22c96ae2b38ab6d865cd
2c78c54821c21a47aea8653fdf60b764575ddd09
F20110320_AACFZC bhat_s_Page_082.txt
193beb353b2d5b043d7e553aef432848
b7cc71adbcc265592b0c11c0a3526424ba4e3db9
24811 F20110320_AACGDU bhat_s_Page_015thm.jpg
bb367d15a4bff04104f5b598ff46b170
c1f0a811414c3c2399f4054e9ea0814d305e1083
F20110320_AACFYO bhat_s_Page_062.txt
66c8ec015265ef8e30edec164800e8b4
61deddace583fd3b103241f4a4fb3bd58e239171
105602 F20110320_AACFBG bhat_s_Page_074.jp2
410276e161e330f263bf5aeaec032622
ca43b3aba2f83c68e5b032445f04e28079e68e83
75645 F20110320_AACGEJ bhat_s_Page_017.QC.jpg
30245ce69f74043a14b547e15c28febc
d52678ebd46c3bc24e1f9ad67f23ed2ef483a1ba
49502 F20110320_AACFAT bhat_s_Page_076.pro
1e90d2a75f2eac7efb3f6a53e6cd724c
8f575e9193718898d228a5a11ae25c44993786be
2013 F20110320_AACFZD bhat_s_Page_083.txt
54be18e178d94c13ff9f8477797b377d
e1f5f8bc441a30f0a109ecb32fa97036d705c0bb
24550 F20110320_AACGDV bhat_s_Page_035thm.jpg
8618fcfe5a6395b75e78ea286a8579d7
1015cc5e201f3b504f86d44e08e2e71c6cc3cf03
1814 F20110320_AACFYP bhat_s_Page_064.txt
de2feaf9b9d4e29468167a778b7a35c6
14ba84154392a5d1e645fc6df99ef79c944d6212
F20110320_AACFBH bhat_s_Page_034.tif
af2e554d84417d253551642bb713dcec
19bb29e263616a0cd91f733b0bea8dcc1aa2fd83
78914 F20110320_AACGEK bhat_s_Page_018.QC.jpg
f37644974e62e5c98a135764b171f8e6
33e03dc8d03cd9e745ea13ab33e8423498621fcb
F20110320_AACFAU bhat_s_Page_080.tif
becc7b4385300cf5328ad84b1a058703
5e9274b811a0cb0c9ef92543b17488b325b867da
1960 F20110320_AACFZE bhat_s_Page_084.txt
75fada97995539bb35019b7eab0d83d5
41716a6b0b572612673f7d645f92d781f40733d4
68097 F20110320_AACGDW bhat_s_Page_036.QC.jpg
fb3d72a46fb255c469e61e02e24c3672
bd07718a80db075fbb18718884737d90b7ec3453
2063 F20110320_AACFYQ bhat_s_Page_065.txt
0f4b3b432c7dc72c18ebca1e1212f494
f1ce34fc59909accfb75b29fd0419379778a5b1c
F20110320_AACFBI bhat_s_Page_055.tif
5986f5e577e1c2cade45a09855ce75be
9c7264962335920618f74035689a79233d24ed79
80622 F20110320_AACGFA bhat_s_Page_051.QC.jpg
bd65db576877f8522fc9146edb37c21b
0d53ecc94381b9753c85a601b4d4d620ca08b341
79386 F20110320_AACGEL bhat_s_Page_022.QC.jpg
5afd190a3c98adde4d4c81755a665663
0a9299f2f36f71db2a09f6f3a24835ff2c432ab8
219003 F20110320_AACFAV bhat_s_Page_067.jpg
56122d89fb3171a69d94f07b59963c99
a9632ee3276989e2da70d3675f562403511104b1
1670 F20110320_AACFZF bhat_s_Page_085.txt
43add7d72296a0b3afe2b5004d159cd1
9580f55501e465ff1c47e8b47a4dd898cbdccc7d
60794 F20110320_AACGDX bhat_s_Page_102.QC.jpg
d6ef59b77d0eda9c42e5c5a5fc6b6b40
df8cc13c87b9f384b1acf20bb71dff474425dae3
2020 F20110320_AACFYR bhat_s_Page_067.txt
8fdbd98445c5884769a740cf62ee2536
6ff7379fd95c38e859ddb613ef51eaf89515300c
35553 F20110320_AACFBJ bhat_s_Page_030.pro
589243cff08348b9f5f9b1249c494fb7
39e19fdc2ad6cf35886c16ed120addd0910e423b
79602 F20110320_AACGFB bhat_s_Page_052.QC.jpg
73600f3a8776abffefb2d0c1f0da201e
16187580cdafea55e965cd0a2b908ae87d84d0c7
78818 F20110320_AACGEM bhat_s_Page_023.QC.jpg
bdee8445d022412660ca99421a1f0880
273c57a81d9e7f760a5d1ffb9a530aee1c4ae9ba
25786 F20110320_AACFAW bhat_s_Page_043thm.jpg
9cce80a2053bafcd96cbb56f8603fbb7
4fb8321bd30ca7e2578068139eda46c3c2db5ed1
1752 F20110320_AACFZG bhat_s_Page_086.txt
85ed448f48e14d82f12c34225e2122db
3bae83cb62c6b89c1cf7cca379da9f0785963321
80344 F20110320_AACGDY bhat_s_Page_045.QC.jpg
371b0be0516ba40a1d4225ad7b56a83c
a1d311827f7e7be9df65a2f410429c3c260fce42
F20110320_AACFYS bhat_s_Page_068.txt
01af360895fb59e6147ca2adfa977de0
559e4db80c6ee9d454a75060dc23509b3dd64fbf
24397 F20110320_AACFBK bhat_s_Page_027thm.jpg
fda318a34a9238c552f5448221d900df
02c7842474c2a3975f960568fe1aa834007be20e
64980 F20110320_AACGFC bhat_s_Page_059.QC.jpg
48d075064cd729e847f677332183a96b
181b660fc2f925355d485911c812bca009965ca4
77531 F20110320_AACGEN bhat_s_Page_024.QC.jpg
53152ea1b0839e6bbde52c1164e0bbda
91f1f5d9fab094b06eb5c429cb2f7b55cb2029e0
22413 F20110320_AACFAX bhat_s_Page_077thm.jpg
91c81c32e1b5d4d301de2c680f687dbf
a8c035a1202d0cb870371cf3493caec43f0a53dd
1372 F20110320_AACFZH bhat_s_Page_087.txt
4bd9ae9fbe2e4ef2df4f802b6d97363a
625afd48880d0b6ee8bee5d5bb8bc9db8c1df1c0
48105 F20110320_AACGDZ bhat_s_Page_088thm.jpg
efb8d4e13fa41a4f45e775f0dba2968a
5bbb52f46ec1a55f07e5e9da488e960699f333e5
1998 F20110320_AACFYT bhat_s_Page_069.txt
933d99214b054efd371a30ae03139a45
d9641aecb3e4e54227d9f3604fd632005187278e
F20110320_AACFCA bhat_s_Page_085.tif
7112eb88646007ac761a43a6f21452af
46f91e4fd000802f2560e04b74c7de2b4801a20e
F20110320_AACFBL bhat_s_Page_075.tif
fc12621029917435276c48f7357113c4
cd45e2ea3a4f1b0b7183a8ae8ac47aa815493978
71232 F20110320_AACGFD bhat_s_Page_060.QC.jpg
7583df708750e54df90cf658dce63dc5
e7a2336bc2e69233752ac4cb0e77ea166230349c
75785 F20110320_AACGEO bhat_s_Page_025.QC.jpg
00ce19485c50aa893bc0a0578798e08e
6cd12451a22cc8a8d7d4eff85a23015c9bb5f33b
14976 F20110320_AACFAY bhat_s_Page_005.pro
a51315495c13456909c1d0ed7742d586
0e7cecc8f92a827a262a3efa3364b101421e3290
2176 F20110320_AACFZI bhat_s_Page_088.txt
1158717b1d14418aad9bfa9b552b6975
51b0262f37916f56bce1df7affb75c0813c29a96
2079 F20110320_AACFYU bhat_s_Page_070.txt
951f4dbac69d1469779422a5ea36c8ad
e48c4d65828153cb9fc31f635e434c23f8a48e58
23209 F20110320_AACFCB bhat_s_Page_064thm.jpg
a83dca1cb6ec447576cba59d5ab196aa
7d97ef3521f0af73d1adc50e0037862f9ae60d65
64585 F20110320_AACFBM bhat_s_Page_125.pro
fd7a3bddf87fe43677dc1da2b7d43a10
a459c3e1be77cb6cfd4fb491a066bf9ce0b283c1
78297 F20110320_AACGFE bhat_s_Page_061.QC.jpg
87011dd3e7c6772052d08a6718a34908
14f315a1c450653f1a252550eabb0bd22e823b86
88505 F20110320_AACGEP bhat_s_Page_026.QC.jpg
fa508b7f7ed5ca92956a0a8ea9a29516
ac3efefa54a42213be4e66476e6dff3a9f887814
20463 F20110320_AACFAZ bhat_s_Page_057thm.jpg
01b1faabd550cb306cc67dcac9cf9268
8d37028c37956e88e77aae553a8b65db40937dad
1401 F20110320_AACFZJ bhat_s_Page_089.txt
c46f65c33bcbb76807480ea9c2745b6b
8463f460d8613a4b8518f5ee57c3eaeb6152a865
1803 F20110320_AACFYV bhat_s_Page_071.txt
f6362101e439f0aed6287213359f5f65
a2067a5a17fe62f08010e89b1ca028dfa842d98a
F20110320_AACFCC bhat_s_Page_103.txt
c25982f770a2ee02d05c83875005a74c
f0a18a22403aebd0fa1617a7cabf96801ca95dee
F20110320_AACFBN bhat_s_Page_077.tif
1ec556b0ab1e16c2efb6ee85e18f443c
dc100641e7307a184a2fe8378eb9e94ebc25827d
81429 F20110320_AACGFF bhat_s_Page_062.QC.jpg
44b5ca71e3309d3b3d8220bc7ddaa026
c4146a67648030be590a9befd2e6e888bfe82189
79469 F20110320_AACGEQ bhat_s_Page_028.QC.jpg
dc29884d667994c173bf69bf72bc57a8
13f9c1fec26ee526ffc1edb16638594d40fa113f
1250 F20110320_AACFZK bhat_s_Page_091.txt
372cb56567c5f0d3936b1f4f86096689
a01a5ee1670c07f458fff4e739bdc8f10e1f93ee
1881 F20110320_AACFYW bhat_s_Page_072.txt
ecb57a0fde92736040e8fa277aa1e17e
81f1b14efc8a9c9d1ef2140ce3dc18f673560060
68791 F20110320_AACFCD bhat_s_Page_071.QC.jpg
923b1b181ad39d536fcab1704315c408
556a20b3cf65d03bc37f572010223e0e566e6d5e
56475 F20110320_AACFBO bhat_s_Page_089.QC.jpg
ae4c2b7fed2eadf5c52f7d915b41a021
1b522a3dd837735415bc2ab70a41aea8c02a42b5
82915 F20110320_AACGFG bhat_s_Page_063.QC.jpg
401a34821569e74fd50ab413ba228b8a
16d437582cc323aa619a00b7fd71480c2a0a8ca5
61153 F20110320_AACGER bhat_s_Page_029.QC.jpg
f2400389a4e4f3d6abef338fa71d72e6
0e1ff6ff164063916c12f6bdb508c960045f3cac
1970 F20110320_AACFZL bhat_s_Page_093.txt
166125091e91d55cb61f25c4cb724773
46675d97af388b9cc7aea5572876e18c748ed00a
F20110320_AACFYX bhat_s_Page_074.txt
92f86364211e9db131e36f3162bffd94
9b997a5767fc42a573322ff16b71e67b5700bbea
1695 F20110320_AACFCE bhat_s_Page_104.txt
a172d8501960e1173482fe6df6b5e8f2
ce2164d175a8566d65d75fa77c047993fdfb1048
F20110320_AACFBP bhat_s_Page_073.txt
7b72cc2569b0d7df8575005b67b6d020
566292dfa57fbdb18cbd09b980ceec3afaa280f8
67427 F20110320_AACGES bhat_s_Page_032.QC.jpg
ac37a241e4099475c50648808315fe32
d3a239721aa894c12509d53d74917e9edf3debd8
1908 F20110320_AACFZM bhat_s_Page_094.txt
4adf40ff96d851b891c86201288fb2f9
f4f92665759e3d46a357087643dbde15c30ff070
1762 F20110320_AACFYY bhat_s_Page_077.txt
5c18a9ea4e3c222d31f10600cc873ac9
cf1f484a74c9e059c4ddd5001acfa7c1c0386041
76206 F20110320_AACFBQ bhat_s_Page_054.QC.jpg
bda80c211c6467cf70acc21004bc7077
782335254c68e195bfffa2e4190060583dc4830e
55811 F20110320_AACGFH bhat_s_Page_065.QC.jpg
2ed922f287a57da7ebacb9646d3bdea4
8cc073795ee0251c1fa7f6354695d813eaee890b
61664 F20110320_AACGET bhat_s_Page_033.QC.jpg
23b2b61b2925f4050a61a070f1294a3d
8ba376f6be1f4781a1d07c59cc71c294f2d9818a
2071 F20110320_AACFZN bhat_s_Page_095.txt
e15fadce63c1296d5880db5cc6230467
9e96e287b3c3b096c765a77a19f5cc11ea5634ef
800 F20110320_AACFYZ bhat_s_Page_078.txt
5094f341a023fd2780c99a1fc9054449
537ad67cef967acc2e9e4b8de7196a7e17f4f722
19656 F20110320_AACFCF bhat_s_Page_033thm.jpg
037d8515ba9d7bf969c4c078193ac99f
ba9cfd464e6f907e3dd13ecd7fa02b47a4041e97
334942 F20110320_AACFBR bhat_s_Page_110.jp2
3f8feda2055f0107c278f526241a1a9c
6dae89ed896c9a534902fdba7497bc73a0773553
63990 F20110320_AACGFI bhat_s_Page_066.QC.jpg
d9f241874b4dd7853b6aa924933f0cd6
2aa72d749d1d4a63189a70e4bb2fc50ba347b1f3
81727 F20110320_AACGEU bhat_s_Page_035.QC.jpg
bd321bd1f9422d10e3f0e1f9862f64f9
d410607625539bcf513b279cdc2a8c8c3ab9671f
1914 F20110320_AACFZO bhat_s_Page_096.txt
f3a1c88d63588af605636370e7c90963
3d44b08991cc1bfb68eed66bd3c5e729e0cba070
1341 F20110320_AACFCG bhat_s_Page_012.txt
ccddc3e1cd25b87821c33e7ceb863eda
b7a06c5446f81bfe153c774d385e381cbdbbb9a8
266573 F20110320_AACFBS bhat_s_Page_119.jpg
ffe9b6dc258d928444b674e747f16e67
bdd6b13f89ddce57bf1cfb4eeecc6e36e3abb3f1
76701 F20110320_AACGFJ bhat_s_Page_068.QC.jpg
ee49cc934780f0b4086787db47273d46
033b5e2dbba0d9e6717bd3956c5aa7b31b87c004
62451 F20110320_AACGEV bhat_s_Page_037.QC.jpg
23e1ef7374c107bef6748875a5468cb5
251abd86b77991fa58500ec182d6785349836608
1808 F20110320_AACFZP bhat_s_Page_097.txt
c61d47f0d881109dc629b6f516d441b7
e86c3d425140422ade75d82e6ad5234001d75e47
44245 F20110320_AACFCH bhat_s_Page_013.pro
af2555811d0bfc00d1ffdf8e931a9f43
146a8802e6186d66fb242123c68bad4ecdde3478
1051862 F20110320_AACFBT bhat_s_Page_088.jp2
d76a2dd5bfbeeb2bd4aec4711b43ab4e
a0b84cda5fd17c826df0520556eefe3b38533b34
78796 F20110320_AACGFK bhat_s_Page_069.QC.jpg
1fbdf631876398149ff7ba2e0fadac6c
3324cf1b434366551455b08bf8e598452ee1430d
77688 F20110320_AACGEW bhat_s_Page_041.QC.jpg
30649c340d40bd71bb7b4c490e7e36f1
31f2f3c30dd60f2bfc285efbc524e932f859c303
2049 F20110320_AACFZQ bhat_s_Page_098.txt
852b40a63ef89c5ac0d9dc619302c63e
98debe0b0469203c4ffc91eb4754b9345fc52af8
86461 F20110320_AACFCI bhat_s_Page_118.QC.jpg
ce0018b5f14e8a0530f0d012d66545e2
f627577a38520e401444a35a8b3c1f87fd466528
41921 F20110320_AACFBU bhat_s_Page_065.pro
5886b7fe5f2b1da2cf2b8329966050ab
8f5eeac858e18ab837f4b2a3161fca3959bce710
48581 F20110320_AACGGA bhat_s_Page_107.QC.jpg
c8803199c8345f2823d7b16450811e71
6699872010d53e1364db3677b3b108284f7621b7
78496 F20110320_AACGFL bhat_s_Page_073.QC.jpg
82a4e96deb90b5c361ac9d552bc02556
4719b635e9e71cbe812bb0a8a8ba6dcbfdd98fb5
77201 F20110320_AACGEX bhat_s_Page_042.QC.jpg
04b27d1dc752291f2375a94e9823c7fb
bf6e14cbedd8215f4c4a6fc059d161731edd0d6d
1942 F20110320_AACFZR bhat_s_Page_099.txt
165c0954be5f21e29b162eb128639451
d36ba43dcf943e7b688dc71e1d5946af3946067f
1681 F20110320_AACFCJ bhat_s_Page_092.txt
3032fcab7714cafe66dc3f2f63fb8b55
92f958fea8f2c12d13d3344a17cb528089b92346
19002 F20110320_AACFBV bhat_s_Page_101thm.jpg
43bc06c9a5604faeb51a5451c257985a
51232ec82e44c5e3b6549933da0eadff118f638b
55350 F20110320_AACGGB bhat_s_Page_108.QC.jpg
4b7ee2e6d05c3c4e46b7f56340de25d9
361833c0fc1041027024cb016c047a63130e8a5d
77758 F20110320_AACGFM bhat_s_Page_075.QC.jpg
823613757050659db0855d2162143e83
ef34191306a918ccfa19c9c7c0c8f374d4aa65f5
60009 F20110320_AACGEY bhat_s_Page_046.QC.jpg
b6a219e8eb1ece3a1f01205ddc6f1a37
b4ed32658b08e033363e0d55aa869cdff412ceb5
1425 F20110320_AACFZS bhat_s_Page_100.txt
9f339cbce7d025b59fdc2c13752b899b
959cd27bc6f9ec2c6af05fea2c2fc6c6e319808f
73730 F20110320_AACFCK bhat_s_Page_064.QC.jpg
ca28df019670999a4068a5b40a272b69
13bfd79dbb93fb68959b534ed5d16363aff4096c
883 F20110320_AACFBW bhat_s_Page_113.txt
3c381529b35b4ba3fc12dfe35aabd382
6bd6e26bd13b17ef59844a46f570a1c1115a9798
46602 F20110320_AACGGC bhat_s_Page_109.QC.jpg
36f46262bdfc6f94ad998ce29a97a005
698423ac4a31fd248294546ffe3e5902bbb159ac
77161 F20110320_AACGFN bhat_s_Page_076.QC.jpg
ed8c10ef94341a6ea381759116f30480
257409ed0212014a01978ffc1f7eb0ee53ac60a4
79043 F20110320_AACGEZ bhat_s_Page_050.QC.jpg
c48fefc0a7712f9ad1b01474c0eae781
64695d6af20d078332a334788bd109ce7e3783f9
1991 F20110320_AACFZT bhat_s_Page_101.txt
ecd0a84cf498adfb1b97266f15018bb5
0b2cdfffe7ade0c89648eb5f2934dd7e5438324d
162327 F20110320_AACFDA bhat_s_Page_065.jpg
e8d9bec684834d3e5502b0eed3cd020f
a41659ecded73abca275fe213ae0a88a5ee7b7d4
110019 F20110320_AACFCL bhat_s_Page_027.jp2
63b6ca7dfdbdf8dd2974f4ff80ec64c3
f1e8bbad0623dd0c19b7d0b92c02ba28e52ecdfa
F20110320_AACFBX bhat_s_Page_054.tif
f5b23ff62e252c46dacd1bde7d422f2c
ee2d948910bcd3bea043e93fee1e0632eb08676b
15765 F20110320_AACGGD bhat_s_Page_113.QC.jpg
81a4a8f4835716faee7b56bf3c1e1a34
9ec60276d37684950affc2497340ffa5a5cfd19c
77901 F20110320_AACGFO bhat_s_Page_080.QC.jpg
1aabc7db26bd5f7fc5312b4494840580
208b14051648416ad8aa58b337f531fb30b68427
1983 F20110320_AACFZU bhat_s_Page_102.txt
4dc54daada0aefae591826675e49e901
28edddc15548ce0addbd19e9193721b5e419edd7
44434 F20110320_AACFDB bhat_s_Page_058.pro
f2723f32015ffcac22133af0e285ca6c
a0ae08e533e1c23a448ced19bddca08aec86fe33
F20110320_AACFCM bhat_s_Page_103.tif
8716e08541c2f12a69bf3cd092fc6683
e213faa75b839d12449f8d9a58713553fb377d88
F20110320_AACFBY bhat_s_Page_003.tif
4e2fa65d55ad2a4855a9a5e37958eb30
9696ad4fccc03e74da6fb7920977bc4d4b0e44d8
75074 F20110320_AACGGE bhat_s_Page_114.QC.jpg
f4c66231d90018fe2106e2638a156b09
9931aa9d178777bd69f18fbbe73705240871a33e
77473 F20110320_AACGFP bhat_s_Page_082.QC.jpg
e8babde443a0954167729fd8fcb3854e
7b23295db727edc7818e3862a13823aac2ca7b7f
1703 F20110320_AACFZV bhat_s_Page_105.txt
19571305e7c9fbcb53c03b9b1a7dd37d
4be3483d9c6c497092df24c839fca1db9fd560d7
276682 F20110320_AACFDC bhat_s_Page_009.jpg
2ddaf35cd9098f1a42467d155f861633
f3fbb47a286b8140c6806835831aded2b3f82048
F20110320_AACFCN bhat_s_Page_042.tif
44eb5428e90ab1035493302fd3213540
fe910851f96e12dd4d9a52ccf541fc7a58b18a06
112591 F20110320_AACFBZ bhat_s_Page_063.jp2
ffcc9587601b64bb5a4bce64efdbcab1
4cbe624b60d204ed1ee2169b9d063403d378a879
84870 F20110320_AACGGF bhat_s_Page_115.QC.jpg
570faf0c487029b9c429fc87502e752f
15f41f329deb8c9df2e7257e3c281ea15f9c5eac
78099 F20110320_AACGFQ bhat_s_Page_084.QC.jpg
fe5cc60fb1927212049b0075063eb68f
7eaa1390840f63245d9c4d5ea51f43c3ac4d2f1e
1589 F20110320_AACFZW bhat_s_Page_106.txt
c34dde4701f715e86b3123c7f2c690a7
ce91bcae4f0004a2719a4a2c7da2cdf95ab0a710
48572 F20110320_AACFDD bhat_s_Page_008.QC.jpg
27fcc0ddae33ebb55e85e4458bc49a25
6f115fc7cb4c2b9fb061dd07c6611fb99aa955cd
72891 F20110320_AACFCO bhat_s_Page_078.QC.jpg
d6b0b4daea48e0f00cb5f1d1c3fc232f
bfdae46be4ec770ff1a94953a34ee7ab888b4214
86229 F20110320_AACGGG bhat_s_Page_117.QC.jpg
036db9a598c9b83ffba819da43472a29
26bd022b79aa28225d826d3d131a3ecae753fa35
76836 F20110320_AACGFR bhat_s_Page_090.QC.jpg
ef242ab19cce989adcbf25297e18f9db
8c75e2de096ab41fd4f56e6ed9ba1c7f76b3e17b
1513 F20110320_AACFZX bhat_s_Page_107.txt
af0820cea7d575172b32db62d25cd635
cb7d3e88315b671066c172d1f3a053176e1c6b74
81223 F20110320_AACFDE bhat_s_Page_053.QC.jpg
334f31d3e22d6b2e633ef1ba455f4b2b
2c292a590e41418792a57da60a5902b42b1a42d4
103169 F20110320_AACFCP bhat_s_Page_060.jp2
d663cca5803c5234eb0fb6ff21fe22d1
2bcb475df958832a0a0fa588c9c2c7673c13de05
81846 F20110320_AACGGH bhat_s_Page_121.QC.jpg
452222889894f628689a886dbcf30282
839db00b99d5541cb26bd24303ff6a4a7dbdc879
78605 F20110320_AACGFS bhat_s_Page_093.QC.jpg
338d35e641c30fd1701f55ae810a8cfd
be8b3f3edd760e1509fc8a29c3a65eb31e9fc84b
1894 F20110320_AACFZY bhat_s_Page_108.txt
14e6d9a3dd96bdf657075021445f2449
79f30d6bb80afd69890699d2e2cfc7c90b325d16
2296 F20110320_AACFDF bhat_s_Page_009.txt
b65e84b1cc0c52c9ff3686a4fcc78f26
9e514adcf875ca7863d573756fae8dd96f6610ca
22522 F20110320_AACFCQ bhat_s_Page_097thm.jpg
9a8df08151b18619d4b9f9ed549e44f4
d4b048a1cf6d50dd18fa3c48060a6da59626a462
77628 F20110320_AACGFT bhat_s_Page_094.QC.jpg
edabf76473ed27be62f8497c3c77deb9
8fb3ace77cb45f31c263b7659ef7caa52e57b278
1446 F20110320_AACFZZ bhat_s_Page_109.txt
a3e049564c633eb9e049245fa9fa344d
806098412b8f1b7aebedfd53f048757932daa5ff
50177 F20110320_AACFCR bhat_s_Page_093.pro
45cd1ef8b7f4dfe21ac4bc1d8ac44eb3
cd7e90e2b87fb965073d659dec5a8a304e80d34e
87994 F20110320_AACGGI bhat_s_Page_122.QC.jpg
ae0db2d5525dbcbd6f4d8c476ca84741
ea0db957f4fcbd9d4ff6e740bd15a090685e333e
83829 F20110320_AACGFU bhat_s_Page_095.QC.jpg
ba8a35471a9a8bafaa31aa65567fc287
54ac3e9ee3d88ee8937bca2f46dc92b5146bb2da
2645 F20110320_AACFDG bhat_s_Page_122.txt
0a8c599c6b7fa7c18363631797bb8dbb
6227b89320cc2d4c64dc0778384dadb655ba7af5
564709 F20110320_AACFCS bhat_s_Page_039.jp2
20876a99ef750ce0401c4e32c5221779
8e4e1beb50cbfb4c15c5d4d0b95e6a965e0f04c9
84227 F20110320_AACGGJ bhat_s_Page_123.QC.jpg
ad6b4223db2f49b54f912b26e440ee0d
1dcd18a28c5b6985c6ccaf5388ecd5ca65a54d4c
70107 F20110320_AACGFV bhat_s_Page_097.QC.jpg
434e70b26e2ebf340f362923b5086cbe
2b2b915a6800bcfcbe353e5de9893a203835bc92
80884 F20110320_AACFDH bhat_s_Page_043.QC.jpg
027d474c3a314050799431377f28603b
b2f094d3bfe944b035c6255535091841f29f05b7
211154 F20110320_AACFCT bhat_s_Page_084.jpg
ad954be77d9e7df9c885a125a49fdffe
70f2ac4867ddd9cfc75e3ffc46dfbe5d3a0990db
84605 F20110320_AACGGK bhat_s_Page_125.QC.jpg
aa004a0a05829258e0d222ed31067261
35fb6e1325eb9360f463482432a4dbe0f08469d5
74801 F20110320_AACGFW bhat_s_Page_099.QC.jpg
cc247e63be815cf619a0910b6d215e27
3e31a39ecb0e9394223806943bd92beca62acb49
2448 F20110320_AACFDI bhat_s_Page_120.txt
bc32880de0d6bb64747a1274a3cb9691
08e3d4f1fc1cda769f94c77269fddf97a68202e7
49172 F20110320_AACFCU bhat_s_Page_100.QC.jpg
27c5f90c2719c1570af3416e79681cdb
f1fd23399fc45a494741e3888e9aa23d224227d3
18722 F20110320_AACGHA bhat_s_Page_046thm.jpg
bbc1b3c72c012c6e565ec98eab264e8c
e7b06a55592ca7b822bc175cc4b7efffddeb67c5
66563 F20110320_AACGGL bhat_s_Page_126.QC.jpg
6cbe887b2b11f9642c91b45175196def
1fe9578e567532ba27cc158b5ee96df77b73f3d3
61110 F20110320_AACGFX bhat_s_Page_101.QC.jpg
5bdd6119fb1471020d7c2d0efa7558a5
3d660d78fc9e2d09e61d10aa8ca2204b27e9d9b3
50910 F20110320_AACFDJ bhat_s_Page_069.pro
6f37e37d600bde731b37a3b5264784b7
7ffb78e8379ac657ed53fe162d3918a780f16e73
24723 F20110320_AACFCV bhat_s_Page_053thm.jpg
6dc6d2a1edb04da04216d7c538c10cca
d03911b29040d7a0b0230bef1befae8cf662e8b8
21937 F20110320_AACGHB bhat_s_Page_047thm.jpg
3444e5c55b52709dccf1dd51a5dc25fd
b66db7a0fc228711e75ff72806df42a0dcb9f44e
59628 F20110320_AACGGM bhat_s_Page_127.QC.jpg
b2bb694b915468bd6ccaf4d29b834e84
aac2071e1a5e450d58607a174a61d7e4468276d0
48617 F20110320_AACGFY bhat_s_Page_103.QC.jpg
fe0162ab7daf8907b86bd37160e545e0
37dbfa84e3cad89024b0c621c8249b507966af92
184025 F20110320_AACFDK bhat_s_Page_059.jpg
428f1b930daa6661d98c612cc1acdd1b
8c78b63f49ea0d5e31db2782142956b44a50ae34
157810 F20110320_AACFCW bhat_s_Page_046.jpg
389b6f6b7a9b5fd150a22134deafad10
41a1d037667303c33b01baee8a63c2d7d4fe17d7
24899 F20110320_AACGHC bhat_s_Page_049thm.jpg
2cca6ce29e1c9d46eb5b896305239ac9
727628346d07d350a81a7a4cb1706a9f97845b8a
3033 F20110320_AACGGN bhat_s_Page_002thm.jpg
4f0ddbc83ac575cb0515caa0bbf3d483
b044c284d573b281a8236d4a039233e264ac5a94
55679 F20110320_AACGFZ bhat_s_Page_104.QC.jpg
25566ab34b362d7c862468989315e01c
98e163609a10b0007c55479e735bb61772b00e1f
F20110320_AACFDL bhat_s_Page_068.tif
6ae98e996cea6e1463765ec9a87d3fd6
9448d46ad1ec28effb7f5c0c8e43c8c940f7630b
215756 F20110320_AACFCX bhat_s_Page_083.jpg
c02d4213ebbc022c2cb08509d0fb0ec8
22d2d4dc3611032e5bd8b4a249976d9f749bc35b
25139 F20110320_AACFEA bhat_s_Page_120thm.jpg
8391e4548e3b3872eab8b3bd8560af28
2826643ff0db13e351fba30a5e9052d33f78a51a
24145 F20110320_AACGHD bhat_s_Page_050thm.jpg
e9ba98c5d30d6ad6a8b8f1348a34e6ca
893c8d9d4deee9963cbe9e36b6060ab2040f3195
3517 F20110320_AACGGO bhat_s_Page_003thm.jpg
9b5f1c6ccd8c1498c6707f77f25a196d
87c324f098bfdee2826a083516925b2e64ee2e56
211584 F20110320_AACFDM bhat_s_Page_023.jpg
bbed6f5c5f24a9356d1fcff564762a18
60924028526ebcb3330f5af54ede4faf30566157
F20110320_AACFCY bhat_s_Page_063.txt
87d31b4e756968f72ca6a8a20db22f1e
1cdcf8a60d06eeae647c4bd93ed66837eafd5c1d
109173 F20110320_AACFEB bhat_s_Page_028.jp2
ec0ac9d5ef1c4a127d39cd4d2be1bc82
8f5830ecdf22f7671eaa199deb2d2b9b8a30e7fc
24410 F20110320_AACGHE bhat_s_Page_052thm.jpg
c7aee4a6c8fa8088f2cb820a4e7e558c
ac355541ff6de661632b55e2ef9467badd728c2e
47162 F20110320_AACGGP bhat_s_Page_006thm.jpg
4b4b0ce34373e9ab33642398fae50430
012c1e5d7d3f2c498c167b0298611524a1903341
47619 F20110320_AACFDN bhat_s_Page_072.pro
3d9210e911aaefe17e6043e74f32569c
1cf64691c8c866332dbe3cd56fc698aab6ee1f6e
263031 F20110320_AACFCZ bhat_s_Page_123.jpg
035f435633320445ad0e1a12b84b58f5
d6d73dcd662a4ed1bfdbdea20d2f3e4233fa1d3d
64979 F20110320_AACFEC bhat_s_Page_122.pro
a6a48e1f86180e7d98c2361f13e0f057
eb51dbd3230e3e507dfafa53d12c94145c6298ca
24447 F20110320_AACGHF bhat_s_Page_054thm.jpg
fd7ceadf0ed567730e42f3eb700febf7
d6e7f918888f9582f6b2635a61c7d68eb8e1b09f
49769 F20110320_AACGGQ bhat_s_Page_009thm.jpg
6a837ffcf05bf497f905c84b313dbe49
1e9727a95870014b6ce3719d929494562d634ec9
104574 F20110320_AACFDO bhat_s_Page_044.jp2
a7a4d68882db01b56055d8367b1baeb0
54b19aa26e4c53799ddfeb1f6025d98fb1012321
F20110320_AACFED bhat_s_Page_040.tif
ddf04750b04399d51ff136661b3ee148
cdd994539326189d27f842912deeb93675885b87
18454 F20110320_AACGHG bhat_s_Page_065thm.jpg
129592a4622837c76528dd962b68f54d
5a42523df5366838893eb89436602819ea710f7d
22285 F20110320_AACGGR bhat_s_Page_013thm.jpg
2ecfa5428c1c735c18d26021c655d4ff
16381218e135144aa2e23353bacedc7044d36e3e
1805 F20110320_AACFDP bhat_s_Page_047.txt
153e5e29dfb66004908e56256dc26cc7
1c0430817bf690f5d9048ec73c3834858d01af8b
23020 F20110320_AACFEE bhat_s_Page_024thm.jpg
369bf04c47b988c7bdb6ba81b2b6771c
66f60c1222fbedbb75c43fa358ac5f7c704c265f
20188 F20110320_AACGHH bhat_s_Page_066thm.jpg
caa06dc0e5676d4b044cefad8a442d83
e22e1eeb493202dddfd37882f56c561a6c2b9a65
24477 F20110320_AACGGS bhat_s_Page_017thm.jpg
e5f4e820e7633cd08df2b72cb3c1e1e5
5d81ec41506d86d31ecd870a6595960cd0ac9dc5
180990 F20110320_AACFDQ bhat_s_Page_090.jpg
9ac3cb441f12bb7e7736e200368f0fbf
c673ce7fe77f5645816936e67f8187b78160f12e
63360 F20110320_AACFEF bhat_s_Page_057.QC.jpg
f44506ad3fc347de259546e7ce123a96
c734d8c0b98afe31ae288f73f7f76830b85ff7b4
24906 F20110320_AACGHI bhat_s_Page_069thm.jpg
ffc1248618141e555975d24dd7f0f432
9eb911c9e1593f5b850270c4a4e50d4d4dc7efcc
6497 F20110320_AACGGT bhat_s_Page_020thm.jpg
beee529c6748cdfd5cf3300fadee582e
0cf308af32553b5ac51668727826a76765329569
42342 F20110320_AACFDR bhat_s_Page_090thm.jpg
15c731c155138d6627b9f1b033fe23fa
9f3509711fcd52bce677d72c9edc8baa4d8fb918
18428 F20110320_AACFEG bhat_s_Page_011thm.jpg
cf7294943c43f49178a6da57369b3cd5
02572828a00edf9ec633466b2ba8d73f413bf091
25166 F20110320_AACGGU bhat_s_Page_022thm.jpg
514515901dad1c91726469b9d1027710
e4e35e7b767d87fec11d19effd5ee56ca388f3c4
F20110320_AACFDS bhat_s_Page_086.tif
b1b4b2a98582bd54e09a68b93731ae06
2e6de5feb6da9af1b034838c999a93576e6b8fad
24607 F20110320_AACGHJ bhat_s_Page_070thm.jpg
d68ad27a40140d6bca74265ae5f00199
6f02ce6134d25b9128beb4d3c0441775a3d38bb2
20689 F20110320_AACGGV bhat_s_Page_029thm.jpg
d349856f685f3c6da0b30e7c804c9484
6b07c88d9e8056bcbea1104cfdb7aa89bcf4edd8
F20110320_AACFDT bhat_s_Page_119.tif
dabd7c499095cf99dd29cdf812c8c692
30380697cd025584a6b2eb2835a558788d1839b7
2256 F20110320_AACFEH bhat_s_Page_010.txt
dca82a59a120d9eef4e7823485b81a5b
23ddb8cc96504b6c853136a67a807cd4f526e640
23381 F20110320_AACGHK bhat_s_Page_071thm.jpg
00b75adad3d92e9de8691835aa74cfce
601dd1ee13732409f46ec0896c0c66a7f441015d
25126 F20110320_AACGGW bhat_s_Page_034thm.jpg
1291555c0931c2f0d0bafdeb2face7d3
83fe4b38d2d1d66fb70cf9b245697e3b026c6e80
604 F20110320_AACFEI bhat_s_Page_005.txt
a04037b689ae03a8d4b4f3d2614ae4c0
3acd591f7593ecc0265f43e5ab8867b1895bdce9
82335 F20110320_AACFDU bhat_s_Page_119.QC.jpg
6b31c38d68e0316ba7bb1e280239e049
d82987e9bc96c9f4af688e0afe05a9d969afc73e
6126 F20110320_AACGIA bhat_s_Page_113thm.jpg
a9ed1a19390b402eda921922cae0a174
1de0b89e016f3578d6ce6531a202a633d6b6b892
24493 F20110320_AACGHL bhat_s_Page_073thm.jpg
3736b79ccdd042d9d5f0433e7fc3cb6c
f8a2870bd8f7c568fa0c3153813af25beef1f4dc
39555 F20110320_AACGGX bhat_s_Page_039thm.jpg
bba481da02628b4269349b0a82147da8
8698ece700b2567227caa8a66cec8fcbee543d6b
92193 F20110320_AACFEJ bhat_s_Page_086.QC.jpg
0065f5a3e50222b06b60b51d499cc56c
b41072cdca566125aeb563bca7e1c77860fb3273
24445 F20110320_AACFDV bhat_s_Page_068thm.jpg
14edf5617e1be9f0e1de7edfc6f9b825
921b78cdc1c9436df13bbd9ff88c9dac2bc60047
24876 F20110320_AACGIB bhat_s_Page_115thm.jpg
fd0f0b3774a7d2013fc128f0fef53afa
bd9cc8439ecfb17edcb7ea63c398caec3c63d715
23666 F20110320_AACGHM bhat_s_Page_076thm.jpg
b5546cda1cb6e6e37130afdf2320a84b
785d05ff2bfd1a6eab9099c7edd23411e8d43e45
23952 F20110320_AACGGY bhat_s_Page_044thm.jpg
42b41fa54fe00a014b048243e152b902
3827b447528dd9c904b485340ea009194cdbee50
105326 F20110320_AACFEK bhat_s_Page_096.jp2
fa66a99de6f64fd595912391a64c011c
72cd3bd217d04e2a7c069f35dc2d5f353eedcc25
25279 F20110320_AACFDW bhat_s_Page_121thm.jpg
590728433115024776405485629d3a4a
c31fdd1fac6a99847399d4612d541ea352e2c5e9
24737 F20110320_AACGIC bhat_s_Page_116thm.jpg
0199a27e213248c164eb839c570a5a28
13dd83bd02c5ab37b954d2b04328c2bce1408419
23392 F20110320_AACGHN bhat_s_Page_081thm.jpg
e40b8b0f90dd256fc0d098feb639257d
2af1ca4950ef361cf693f2761944d021ac8985d9
25299 F20110320_AACGGZ bhat_s_Page_045thm.jpg
3c029d7c8592a518a0fc40c45142a6fc
1457cbd29b409bebd530ec22a7008936dd2dd3a5
48305 F20110320_AACFFA bhat_s_Page_010thm.jpg
7201a0f7819a60647efeadf5071a9c89
c67ffb299698017fb530218b3a105dbb244d8ee3
24911 F20110320_AACFEL bhat_s_Page_023thm.jpg
6cb81db26bf0da755620f41706d879a7
f511c3d823adb8ead73df6fb7bdabd3d4867d451
590289 F20110320_AACFDX bhat_s_Page_040.jp2
bd7734695ebde996f8d53b9bff635c84
2442a4b78e5fbdad3d7a585a21f0f4e9aebd901e
24633 F20110320_AACGID bhat_s_Page_118thm.jpg
64c55820f41781541b20c385b8e1466a
3ea470e7408f8e37f1abf09e166be23a02f70415
24977 F20110320_AACGHO bhat_s_Page_084thm.jpg
2eaff9244d5a5b6b1daa0800523e8300
c1e8d80487fa0f0664120378b732ba5bef5749c3
F20110320_AACFFB bhat_s_Page_097.tif
611b5c103cc88d0a8c350ec2153dfa20
6eec4e1ed49b51b0507d1ddc023faf17f788d19b
F20110320_AACFEM bhat_s_Page_013.tif
46eafd627ce0af9992c71f4f7f61176b
c232d623b316e796d2785e1b344daf7b11175bf1
F20110320_AACFDY bhat_s_Page_052.tif
5a4ec804cc7657f48ba8b0df9ea03927
81f7da1dfd519a72d0d29beb63b05f8505030f8e
25216 F20110320_AACGIE bhat_s_Page_123thm.jpg
4845b2d170e6f97de1a120f3af1305f9
8d01d14adc7767974ed0be3af15999c0f645b29f
18868 F20110320_AACGHP bhat_s_Page_089thm.jpg
5e9c3a898fe3e3fe067cb155e005e2ef
e18ca9ee648493c3761f7c1d8eede5d3efe574a8
209120 F20110320_AACFFC bhat_s_Page_068.jpg
b32b08cb99cde900cfb265d561921f96
1d72e856466e89ed064a36a6b97301947db23d32
94064 F20110320_AACFEN bhat_s_Page_006.QC.jpg
327762cdcb3956ae4320f103c0f5a18a
fc4b3eab39c66efc833029333ce7b18cd6e6bcc9
41092 F20110320_AACFDZ bhat_s_Page_078thm.jpg
f837533287874b1e49cdaa32a5209f3b
9b10853cbd1752b4f78f212b14694f7cbb26c959
25070 F20110320_AACGIF bhat_s_Page_125thm.jpg
5178c18a2803d039e61a5e488e3f35b4
02f61261f730351e809c806037ccd305c1e0e191
37600 F20110320_AACGHQ bhat_s_Page_091thm.jpg
9adedb445c6a30056976f2764dec2765
e8a836b8aab38958fc4119c0b1b2ebb82f91f001
105603 F20110320_AACFFD bhat_s_Page_080.jp2
2fc6428f724e5fead1f6031715d71411
3e777af574434d5fc8831ba5a141e91257dedb2c
165908 F20110320_AACFEO bhat_s_Page_057.jpg
a06785c85e31e6038f2860643dd38b3e
470b2b78df32e3d977a7eba90bba0d74fbfa63f0
19773 F20110320_AACGHR bhat_s_Page_092thm.jpg
615410888b5d7ab1d5dc469017da1d72
bd19dfdbfd491da83de58f48fb12cc102b3f69d4
38620 F20110320_AACFFE bhat_s_Page_104.pro
ff22af0e7236d82309216fd92bf9c746
f7902d616d5a57504d2f0be74bfe3e493cec5afd
227263 F20110320_AACFEP bhat_s_Page_088.jpg
500b0c53a345a091efaa12ef33e8ac6b
73970863fd289f4a3421165aaf14f53efabb16f5
25405 F20110320_AACGHS bhat_s_Page_094thm.jpg
c0621bd14df792d794dac5fed76a676c
bd8116fc3adc905cdfaf0692b8271d99cdc0e298
140580 F20110320_AACFFF bhat_s_Page_100.jpg
471f33c3e24363a73c7b6db513c7971d
700a441c0868558245e21319000264e85ad7b958
2027 F20110320_AACFEQ bhat_s_Page_075.txt
709cb182fe8139495482c389d4a30a64
61e081b19b73330d4f08e8ce09536cbcd8b16f78
25338 F20110320_AACGHT bhat_s_Page_095thm.jpg
d70096cd869c5bc5d87a5e85eb414270
3078d1ef5f53f139d8006705198081c8ef4e4ca6
217895 F20110320_AACFFG bhat_s_Page_063.jpg
aaf1185a48fa651d3bfbecc74716dc7f
ac17a351d5b124e601e28ed21a3b7af44c3ab034
52191 F20110320_AACFER bhat_s_Page_007thm.jpg
73c64230895e69395aecd7159646ae5c
3d5dc82e733a876f9906e32d128fa23fa0b163dd
25725 F20110320_AACGHU bhat_s_Page_098thm.jpg
8691dba4dc515abdc50f112157f647ad
132ecc02a70b3ca1539a7b5981933f3834db224d
42826 F20110320_AACFFH bhat_s_Page_055thm.jpg
c0b5da081d4a36f63b7f61ba58dd365f
39a3759a357d17696fac17ab3c6ba108cc9c7562
1971 F20110320_AACFES bhat_s_Page_054.txt
395d4f207242d2558a83404f420a28db
50ce6781bf74e97b6d51b56c523f6a35fad975b2
F20110320_AACGHV bhat_s_Page_099thm.jpg
32019ef67aaae954457300e7e0e261ac
57627ca39d84c1a8a7bb44f8ec5cc2aca5295a4b
102544 F20110320_AACFET bhat_s_Page_111.jpg
b9c8fb396aa73fb667347aa8b6ab3887
40c7fac4e5cad7e8a97528d45a04cbdc7624194b
15996 F20110320_AACGHW bhat_s_Page_100thm.jpg
d0e444cd1f0cca01492b0612063a0b23
30f8055d97dc1d8816be11e0163d5a3fe81131db
1988 F20110320_AACFFI bhat_s_Page_022.txt
7455d478bdeacc17178b6bdaa2b36008
e737d2471528e8ab58032f6b9b5f0480a09dd50b
14555 F20110320_AACFEU bhat_s_Page_112thm.jpg
9a982fec0522782eae6356113ba67468
14009854473839125efeccd82ac30458ec256046
15809 F20110320_AACGHX bhat_s_Page_106thm.jpg
15a17d1aa4af9a7ac0570a727c70c2d7
7886b9d1b9566b7de684f55182d53fdad0078797
56456 F20110320_AACFFJ bhat_s_Page_079.QC.jpg
d3b883967d3be4a90fd64182b9333423
ac32f551162a885f9d36d35888e65b9741b913ef
1950 F20110320_AACFEV bhat_s_Page_076.txt
05389277ee73944e7fd859ff7c5f27c1
56fe89ea205af6fa500e8ce2e345444aed72fefe
17080 F20110320_AACGHY bhat_s_Page_108thm.jpg
125c060cb154135c51cca6c0d11120de
7499c9ccb57517089f8e31ac95e6cc6dbaa70a81
209822 F20110320_AACFFK bhat_s_Page_018.jpg
b9109f2153eac073944014a422947457
582a9593c061a2a8fa4796fd0aa4ed7fc4528fdf
24110 F20110320_AACFEW bhat_s_Page_042thm.jpg
a5d4b7ef5ef380538e5247809f14d66f
6d9f82500191b5941ae57c7dfc5e11b5cd674798
13916 F20110320_AACGHZ bhat_s_Page_109thm.jpg
891fa4356e5bcce12351569a90ca16d0
19a8ef639219fada6380fc71b9e5cddf22068ce0
133826 F20110320_AACFFL bhat_s_Page_122.jp2
dfdc1c3c0f1fc72e82028bdd9cbc92c0
5b019600a8c1559990d29f667ef87f397da5eaa4
19895 F20110320_AACFEX bhat_s_Page_059thm.jpg
204935882111a97dece72aa5a9d823d5
6058465ec054ea0ebfd09ce31a9a8b9b32be5f22
212319 F20110320_AACFGA bhat_s_Page_069.jpg
20a183009aee184e44476c348c95bd2f
ed5e998e248e60fb621f05f3c47a6aafed8c2756
2025 F20110320_AACFFM bhat_s_Page_043.txt
70c7684109109c087f85494112929a8b
fcf71ef256c240db6f72e0ef582da4cda4564ead
157241 F20110320_AACFEY bhat_s_Page_089.jpg
db0bd3c9bc0aa2b29fbb73787e559cd3
eec40956af2fda89b26fbd1ec145c1da2c339afd
72165 F20110320_AACFGB bhat_s_Page_019.QC.jpg
d11755704a916bfdf36464c5745e99ad
48bbd966b8bc19391f5282a81dabb7f722c0cff0
138176 F20110320_AACFFN bhat_s_Page_103.jpg
c1663394dfdbc58811968ad68be0032c
3231b299d8951d013b23e95ec5602e70c6384c83
814 F20110320_AACFEZ bhat_s_Page_038.txt
4c576abb8c1ee9a41c68d03348d7eb65
971bb9422c0e790c53dd0937d6a467276fd47981
210509 F20110320_AACFGC bhat_s_Page_015.jpg
1a3eb7dadc3fb6847e61cd59c19dd9db
4168d04eb6cbaca27624d47c387a6ebbb6ffd3c8
74693 F20110320_AACFFO bhat_s_Page_055.QC.jpg
9ec6f9059f49a5e1fc308f8a3930c778
ff0738f6ef9d415685f32f363da0702432549a2f
2000 F20110320_AACFGD bhat_s_Page_023.txt
6e39f36dcef6ad87debce55528452ec5
4304576040e36bd8e08258246ccda67da6591f48
72023 F20110320_AACFFP bhat_s_Page_072.QC.jpg
744dc7999179b3c470227b448a13086d
c96aef9712dd924c737c8a99fc9421fde385bdae
48508 F20110320_AACFGE bhat_s_Page_017.pro
1f1b3292f2d32c150ee7fed889270c89
1d885849984d9eb182930f19634a6836bf60f3d7
F20110320_AACFFQ bhat_s_Page_016.tif
5994f20fbfaf9c73c8c369a0fcacd7d4
04e041ff56c8a18330cb1244ea52aef1473c22de
1952 F20110320_AACEZW bhat_s_Page_024.txt
a32fab7a40a2be73128678bf834d07eb
eb43793c4715cbcf0224893abe947fa8cfc6aa5b
44181 F20110320_AACFGF bhat_s_Page_097.pro
b0099810269fb1b78452c97803e8f4af
77146c262196079d9ab6fb0ed4ce5019fbf0a26b
210027 F20110320_AACFFR bhat_s_Page_022.jpg
68731421deb804540b24e713fcba2e1c
def5ff07be5a161416568f85e609215732f9700d
21328 F20110320_AACEZX bhat_s_Page_085thm.jpg
2afb98fc25ac0940752bf5723f37c405
e1310b37c1e24ff729d72c5d9574015211c68dc3
F20110320_AACFGG bhat_s_Page_088.tif
5f5f30a7481da1ef70385c11989c4134
5257bd1f125ec84ac5e1520a7e8cd56e54c72d9a
F20110320_AACFFS bhat_s_Page_083.tif
82e5d21ad83a251ee4105fbb0ca88380
68249892cce9b63174f30dac0096130a2fb41d72
256875 F20110320_AACEZY bhat_s_Page_010.jpg
dca9341036c29271956d099d05154440
c7d2f2046f1b7b5bca97077d734458a953302985
187698 F20110320_AACFGH bhat_s_Page_071.jpg
5bcd01482f3d842dc7c8788849ab3d6f
8a1cba1287ac73e6cae02362c1738457ca8ce0ba
271338 F20110320_AACFFT bhat_s_Page_117.jpg
a377b2fdcb9a51d9516e18f2c3ac089b
65f38fd7d5823696c441537c1e32c77ffd92394b
F20110320_AACEZZ bhat_s_Page_015.tif
906df51346fdd96486b4661a6db3b700
09ecea0847cd6732c9b4d292c8651dbb94ca7bd2
2474 F20110320_AACFGI bhat_s_Page_033.txt
47b97d02a0ab9107581b61db3b03a274
672c4ee3b2ff237e7aaa4f160ac1552ddea3bf38
2390 F20110320_AACFFU bhat_s_Page_081.txt
8df7ec1f98f434a7eef69c34acd295f0
81a25ee86d8843ba47fa0a79ddddce641baca065
F20110320_AACFFV bhat_s_Page_098.tif
a082633c3d7644edc5b2c21f3c399c61
0964a5539558e96ed50fbf88f2253bdc85d11043
206159 F20110320_AACFGJ bhat_s_Page_096.jpg
4a87e1858fa3e0e07677b522211009cc
d635ad7f7984781b306d8f078e85798bda680023
510543 F20110320_AACFFW bhat_s_Page_008.jp2
2c4f56003ad42ff37fdfb872e8fe1b1d
5b869fa239f200b4797b6ad7bcbcea7a46b388d7
88326 F20110320_AACFGK bhat_s_Page_101.jp2
ff1ddac1e952d8fb1498b8bd452399d2
64569dfc579a8954c2f88c4dfcfba79d92e1d485
25009 F20110320_AACFFX bhat_s_Page_119thm.jpg
a9950bb09269a328237d878bfaa52096
7cf6cc69d62371f1ddbb10fd7358805fd3fe1f08
243186 F20110320_AACFGL bhat_s_Page_120.jpg
5a5dd93651f47bf047f4e3d9c2e9e299
098acbcb7079de81a60215f13ab71b05fa2c21fa
F20110320_AACFFY bhat_s_Page_116.tif
341c2f2973e063aa57a03891bae2e58d
0780632e36621ddc951e40705c0113d1c98b3872
108896 F20110320_AACFGM bhat_s_Page_084.jp2
3068955236ad630d74da4fc77edfe7b2
6632a33df77e7dead8a707e4ca6dde5e8075acd7
1966 F20110320_AACFFZ bhat_s_Page_034.txt
0df3f6fd6bc35b70c6fcedf8cb54bcf9
1647981f0f9511aeb006db9b1a2e74c5012d409d
58252 F20110320_AACFHC bhat_s_Page_001.jpg
6f390819e1521e38a4cd69e782e4eb85
74a99f0fb29bf392e467d478323ebcea437092cb
63043 F20110320_AACFGN bhat_s_Page_123.pro
75293420187caa077974e9a1de19efa8
304f0c7ecd3b5142c64443c3a845bae0a1f98aa8
14097 F20110320_AACFHD bhat_s_Page_002.jpg
a68dca2a78a00e79a5fcb0616852cf10
a7493be2d4b2e1ba264913e3bf6e64610e604e4c
59809 F20110320_AACFGO bhat_s_Page_091.QC.jpg
61534ac207e2e35f590f4ecc4c1065b5
3f2edd54a459cc6bd5c38122f76a5693d18dbd0c
20825 F20110320_AACFHE bhat_s_Page_003.jpg
b6e278d62d739af49dabf4e3f4c5dd13
16f9092e2ee8a13c5158b9f09cddec8c0e894532
F20110320_AACFGP bhat_s_Page_007.tif
c5a96c1fd7174a5cb3794081f2b9ace8
2716d6f3cad082f1866f14fa4f155ddbd0b6b195
193077 F20110320_AACFHF bhat_s_Page_004.jpg
5372465b59f12a486df7f0804fa9f5ff
0f69d4f1e6f17aa6d2ea0f9e0acf8d07121e85f4
212239 F20110320_AACFGQ bhat_s_Page_082.jpg
27741a9de0f1d5e13f1fe61b96cdc001
9677ee7ad2042b0338c605b5afa9dface1362f1a
71498 F20110320_AACFHG bhat_s_Page_005.jpg
312539f8ae886565c1725e7fd0e106de
c23143de08365251d40e1670f71df7f2d66b4311
F20110320_AACFGR bhat_s_Page_073.tif
f21019d1773700b88cebaa902ed0346f
18c673b3e4386a5d027aa643d5f9ecf875de5845
269981 F20110320_AACFHH bhat_s_Page_006.jpg
355fcb44f5af89cf71c65f5f28f8529c
78984b10691d91efbb04a25f94a947bb5ba56b6f
28933 F20110320_AACFGS bhat_s_Page_005.QC.jpg
a90cf58fadd03d76025a50707206876e
ee55b4f46cc880d7dd39925487f6b52a5ab02179
355842 F20110320_AACFHI bhat_s_Page_007.jpg
50ce846eb7665c003f389c9423c5394d
44fc9e232187aa76af9d6ecca05019ba87eddb84
22017 F20110320_AACFGT bhat_s_Page_113.jp2
b374619075f2728a3b64ffef4cd66819
7121bcd2c9253ade17ed5d6c9a81c01e047999d1
96979 F20110320_AACFHJ bhat_s_Page_008.jpg
e7a80c74bad6bed8a62e053baef8777d
7c4402d96bb7d842c5e4709993c7708696cb9d54
5540 F20110320_AACFGU bhat_s_Page_002.QC.jpg
fc76d58ddd627399b341a3192960d5b6
37725a604d4ea03d0763d426d02c69162b5f1ee2
40078 F20110320_AACFGV bhat_s_Page_038thm.jpg
1cf0fa49a2d5bdc1f5096f4e3323efc9
9eeb778d6ed7840f5fae0fad71bebc19bcb35a30
175481 F20110320_AACFHK bhat_s_Page_011.jpg
dfaf8b291b8fb30ec4d4d363dbe55ced
918b3f66d65d4f0501d0ab15e1b0ab094116e906
40079 F20110320_AACFGW bhat_s_Page_092.pro
afd5d65f3ce60e9bca97ecf688578db0
d5b3a7048392345658fc01b353c3650d4b564e30


Permanent Link: http://ufdc.ufl.edu/UFE0009283/00001

Material Information

Title: Ecohydrological Study of Watersheds within the Military Installation in Fort Benning, Georgia
Physical Description: Mixed Material
Copyright Date: 2008

Record Information

Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
System ID: UFE0009283:00001

Permanent Link: http://ufdc.ufl.edu/UFE0009283/00001

Material Information

Title: Ecohydrological Study of Watersheds within the Military Installation in Fort Benning, Georgia
Physical Description: Mixed Material
Copyright Date: 2008

Record Information

Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
System ID: UFE0009283:00001


This item has the following downloads:


Full Text












ECOHYDROLOGICAL STUDY OF WATERSHEDS WITHIN THE MILITARY
INSTALLATION IN FORT BENNING, GEORGIA













By

SHIRISH BHAT


A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY

UNIVERSITY OF FLORIDA


2005

































Copyright 2005

by

Shirish Bhat
































This work is dedicated to my parents, Tara and Narmada Bhat, and uncle Devendra Bhat.















ACKNOWLEDGMENTS

First of all, I would like to thank Dr. Jennifer Jacobs, for her constant guidance,

encouragement, patience and continuous support over the past five years. Her

enthusiasm for research and quest for excellence have left an everlasting impression in

my mind. To me, she has been more than an advisor, and this research would not have

been possible without her. Secondly, I would like to thank Dr. Kirk Hatfield for being on

my committee and offering me guidance in the later part of my research. I would also

like to thank Dr. Wendy Graham, Dr. Ramesh Shrestha, and Dr. Ramesh Reddy for being

on my committee and their invaluable suggestions throughout the research work. I would

like to thank Dr. Richard Lowrance and Randall Williams of USDA-ARS, Georgia,

whose contributions in this research have been tremendous. I deeply benefited from all

the long hours of fruitful discussions with them on a multitude of topics.

I would like to thank Hugh Westbury at Fort Benning, Georgia, for his coordination

efforts during the field trips, Dwight Dindial and Charles Campbell for their efforts in the

collection of the field data, and Phil Harmer for his contribution in the laboratory.

I wish to extend my gratitude to Lewis and Kim Bryant, and Donna Rowland, who

have been very helpful during my difficult days at UF. I also wish to extend my gratitude

to Subarna and Pramila Malakar for their care and support. I wish to extend my thanks to

Mr. Binod Palikhe for his help during my academic pursuit. I would like to extend my

thank to Debra Carol, Carol Hipsley, Sonja Lee, Doretha Ray, and all the Civil

Engineering staff for their help during all these years. I would also like to thank my









friends Deepak Singh, Shashi Shrestha, Hemant Belbase, Rishi Bhattarai, Dipendra Piya,

and Anand Bastola, and colleagues Nebiyu Tiruneh, Aniruddha Guha, Qing Sun, Mark

Newman, Ali Sedighi, Chris Brown, Brent Whitfield, Gerard Ripo, Haki Klammler and

others for their encouragements and moral support.

I would like to thank my parents and all the family members for their constant love

and encouragement. They have allowed me to pursue whatever I wanted in life. Without

their guidance and affection, it would have been impossible for me to advance my

education.
















TABLE OF CONTENTS

page


A C K N O W L E D G M E N T S ................................................................................................. iv

LIST OF TABLES ....................................................... ............ ....... ....... ix

LIST OF FIGURES ............................... ... ...... ... ................. .x

ABSTRACT .............. .......................................... xi

CHAPTER

1 GENERAL INTRODUCTION ...................................................... .....................

O bj ectiv es .................................................................. .................................. .5
D issertation O organization .............................................. ........ .............................

2 ECOLOGICAL INDICATORS IN FORESTED WATERSHEDS IN FORT
BENNING, GEORGIA: RELATIONSHIP BETWEEN LAND USE AND
STR EA M W A TER Q U A LITY ......................................................... .....................9

Introduction ...................................... .................................. ......... .... .. 9
Study A rea .................................. ............................ .... ...... ........ 12
M ethods of Study .................. .................................... ........... ............ .. 13
D description of W atersheds ................................................................... ............... 13
Characterization of Disturbance Categories.................................................. 14
Collection and Analysis of Stream Samples ....................................... .......... 15
Statistical A n aly ses........... ........................................................ .. ... .... ... ... 15
R results ................................................................................ ............... 16
Watershed Physical Characteristics.................... ..... ........................ 16
W ater Q quality Param eters......................................................... ............... 19
Effects of D disturbance Categories .................................................................. 21
Relationship between Watershed Physical Characteristics and Water Quality
P aram eters .............................................................................................. ........22
D discussion ....................................................................................................... ....... 29
Conclusion and Recommendations.................... ....... ........................... 33

3 HYDROLOGIC INDICES OF WATERSHED SCALE MILITARY IMPACTS IN
F O R T B E N N IN G G A ...................................................................... ...................35

In tro d u ctio n ...................................... ................................................ 3 5
vi









Flow Regim es and Hydrologic Indices.................................... ....................... 37
Effects of Flow Magnitude on Stream Ecology ................................................40
Effects of Flow Duration on Stream Ecology ............................................40
Effects of Flow Frequency on Stream Ecology................................................41
Effects of Flow Rate of Change on Stream Ecology .....................................42
D ata C collection ...................................................... ................. 42
W atershed Characteristics ............................................................................42
Precipitation and Streamflow Data ..................... .... ............... 45
M methods .............................. .....................................4 5
Annual-Based Hydrologic Indices ............................. ..... ...................... 45
Storm -Based Hydrologic Indices ............................................. ............... 48
Statistical A naly ses........... ........................................................ .. ... .... .... 48
R e su lts .......... ... ....... ........ ......... ........... ............................. 4 8
W atershed Disturbance Characteristics............... .............................................48
Annual-Based Hydrologic Indices ............................. ..... ...................... 50
Storm -B ased H ydrologic Indices ........................ .... ... ................. .... 51
Relationship between Military Land Management and Storm-Based
H ydrologic Indices ...... ....... .... ....... ...... ...... ......... .............. .. 52
Discussion .................. .......... ..................... .......... .. 54
Conclusion and Recommendations.................... ....... ........................... 58

4 PREDICTION OF NITROGEN LEACHING FROM FRESHLY FALLEN
LEAVES: APPLICATION OF RIPARIAN ECOSYSTEM MANAGEMENT
M O D E L (R E M M ) ............................................................................ ................ .. 59

In tro d u ctio n ........................................................................................................... 5 9
B asic Concepts of REM M ................................................. ........................... 62
Hydrology component of REMM .............. ......................... ....................63
Nutrient component of REM M ................................. ................ ............... 64
A application of R EM M ........................................................................... ............... 66
Study Area .................................... ................................ .........66
D ata and Input Param eters........................................................ ............... 68
M odel C alib ration .......... ..... .......................................................... .... .... ... ..7 1
Sensitivity A naly sis ............. ........................................................ ... .... .... 73
R e su lts ...........................................................................................7 3
Hydrology ...... ......... .. ........... ..................73
N itrog en ...................................................................................................74
S en sitiv ity A n aly sis ....................................................................................... 7 7
Discussion ....................... ................................ 80
Conclusion and Recommendations............................. .......... 83

5 SUMMARY AND CONCLUSION ............. ............................................ 85

APPENDIX

A ANNUAL-BASED INDICES DEFINITIONS AND CALCULATION
P R O C E D U R E S .................................................................................................... 8 8









B STORM-BASED INDICES DEFINITIONS AND CALCULATION
P R O C E D U R E S ................................................................................ ................ .. 9 5

C NITROGEN TRANSFORMATIONS IN REMM ...................................................98

D PARAMETERS AND RATE CONSTANTS USED IN REMM ............................99

L IST O F R E F E R E N C E S ...................................................................... ..................... 102

BIOGRAPHICAL SKETCH ........................................................................115
















LIST OF TABLES


Table page

2-1. Physical characteristics of study watersheds in Fort Benning, Georgia. ..............17

2-2. Pearson correlation coefficients between watershed characteristics. ....................18

2-3. t-test results for differences in mean values of watershed physical characteristics
and w ater quality param eters. .............................................................................21

2-4. Pearson correlation coefficients between watershed characteristics and water
quality param eters. ....................... ...................... .................... .. ..... 23

2-5. Stepwise multiple regression models for water quality parameters ........................24

3-1. Definitions of terms used to describe hyetograph and response hydrograph...........39

3-2. Physical characteristics of study watersheds. ...................................................44

3-3. Summary of the recommended hydrologic indices for 'perennial runoff type of
store a m s................................. ........................................................... ............... 4 6

3-4. Summary of hydrologic indices used in the Storm-Based Hydrologic Indices. ......49

3-5. Pearson correlation coefficients between watershed physical characteristics and
event based hydrologic indices.. ......................... .............................................. 53

3-6. Stepwise multiple regression models for event based hydrologic indices. ...........54

4-1. REMM model parameters values by the riparian zone for the Bonham-South
watershed in Fort Benning, Georgia. ............................................ ............... 67

4-2. Soil and vegetation nutrient pools for Bonham-South watershed's riparian area.....69

4-3. Model simulated annual hydrologic budget for the riparian areas in
B onham -South w atershed.. ............................. .... ...................................... 69

4-4. Sensitivity of modeled streamflow and TKN based on +/- 10% change in model
parameters for Bonham-South watershed. .................................... .................80
















LIST OF FIGURES


Figure p

2-1 Study w atersheds ................................................ .............. ..... 14

2-2 Box plots of water quality parameters ......................................... ...............20

2-3 Relationships between military land and water quality parameter. .......................25

2-4 Relationships between road density and water quality parameter .........................26

2-5 Relationships between number of roads crossing streams and water quality
param eter............................................................................................... 27

2-6 Relationships between disturbance index and water quality parameter .................28

3-1 Terms used to describe hyetograph and response hydrograph.............................39

3-2 Study w atersheds ................................................ .............. ..... 43

4 -1 S tu dy area .......................................................................................................... . 6 6

4-2. Comparison of REMM simulated daily flow with the observed daily flow for
the B onham -South w atershed........................................................ ............... 74

4-3. Observed canopy cover and TKN concentrations each month for the
B onham -South w atershed. ............................................. .............................. 75

4-4. Monthly simulated TKN concentration and simulated leaf mass in the Bonham-
South 's riparian area............. ...................................... ................ .. .... ..... .. 75

4-5. Monthly simulated TKN concentration and simulated nitrogen present in live
leaves the Bonham-South's riparian canopy. ........................... ................................ 76

4-6. Observed TKN concentrations during the events.....................................................78

4-7. Surface runoff comparison of the observed and simulated TKN masses..................79

4-8. Total streamflow comparison of the observed and simulated TKN masses..............79















Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy

ECOHYDROLOGICAL STUDY OF WATERSHEDS WITHIN THE MILITARY
INSTALLATION IN FORT BENNING, GEORGIA

By

Shirish Bhat

May 2005

Chair: Kirk Hatfield
Cochair: Jennifer Jacobs
Major Department: Civil and Coastal Engineering

Relationships among watershed physical characteristics and water quality

parameters were explored for seven watersheds in Fort Benning, Georgia, using statistical

analyses to identify chemical indicators of ecological changes. Correlations were

identified among the indicators and watershed physical characteristics. Regression

results suggested that pH, chloride, total phosphorus, total Kjeldahl nitrogen, total

organic carbon, and total suspended solids are useful indicators of watershed physical

characteristics that are susceptible to perturbations.

The magnitude, frequency, duration, timing, and rate of change of hydrologic

conditions regulate ecological processes in aquatic ecosystems. Analysis of 26 non-

redundant hydrologic indices showed undisturbed watershed produced higher magnitude

and more frequent low- and high-flows as compared to the disturbed watershed.

Eighteen storm-based indices, grouped into four flow components, were proposed to

statistically characterize hydrologic variation among different watersheds. Results









showed that these storm-based indices might be used as surrogates to the indices derived

from long-term data. Statistical analysis showed that the watershed physical

characteristics such as military training land, road density, and the number of roads

crossing streams predicted hydrologic indices such as storm-based baseflow index,

bankfull discharge, response lag, and time of rise well.

Riparian vegetation has an important role in altering water quality in the forested

watersheds. The leaching of organic or mineral products on the forest floor provides

potential additional effects on water quality. In these areas, nutrients are released into the

fresh water systems due to the leaching and decomposition of vegetation litter. The

release of nutrients from plant litters prior to decomposition may be an important aspect

of characterizing stream water quality. To explore nitrogen leaching in a riparian area

during litterfall, Riparian Ecosystem Management Model developed by USDA-ARS has

been applied. The simulated total Kjeldahl nitrogen masses in the study watershed were

close to the observed values. The model effectively captured the trends of litter mass

accumulation in the riparian area and subsequent high concentrations and masses during

those periods.














CHAPTER 1
GENERAL INTRODUCTION

Ecohydrology is the interdisciplinary research field in which hydrology and

ecology come together. It may be defined as the study of the functional relations between

hydrology and biota at the watershed scale (Zalewski, 2000). Resource managers and

land use planners seek ecohydrological knowledge enabling them to design effective land

use plans and water management strategies.

Water, soil, and plant cover are fundamental components that determine the

productivity of the land. The successional stages in evolution of ecosystems depend on

climatic and hydrologic conditions and on nutrient availability. The unique composition

of plants and animals determines the ability to retain water and nutrients within the

system (Zalewski et al., 1997). The amount of water and its quality in the aquatic

environment are guided mostly by climate. Water and nutrient cycling in terrestrial

systems are tightly linked with each other and with human activity. The growth and

activity of human populations have increased the input of nutrients to terrestrial and

aquatic ecosystems (Schindler and Bayley, 1993; Vitousek et al., 1997). Various forms

of environmental perturbations affect quality and quantity of inland waters by increasing

runoff, erosion, sedimentation, and pollution.

Modifications of land use affect the water quality, residence time, surface runoff,

soil moisture, evaporation, and ground water. For example, an increase in urbanization

and non-forest land increased nitrogen concentration in the streams (Osborne and Wiley,

1988; Sponseller et al., 2001). Land-use practices can potentially affect and modify









channels along the flow paths in a landscape and may enhance or disrupt runoff

(Nakamura et al., 2000). Forest cutting can increase the frequency and volume of debris

slides due to dead and decayed tree roots that contribute to soil strength on marginally

stable sites (Sidle et al., 1985).

Water quality is highly variable, from place to place and from time to time, even

within a particular ecosystem. It is dependent on many factors, both natural and as a

consequence of human activities. In both cases, the quality of water is further affected by

the soils and vegetation over and through which it passes. Nutrients are released into the

freshwater systems due to the decomposition of the vegetation litter and modify the water

quality. Riparian vegetation has an important role in altering water quality as it functions

as a source or a sink of various organic and mineral components (Cirmo and McDonnell,

1997).

Agricultural practices and urbanization are widely recognized human impacts on

water resources. A specific type of human impact is caused at military installations.

Military training within a watershed can affect the drainage patterns, vegetation, and

soils. The impact is due to troop maneuvers and large tracked and wheeled vehicles that

traverse thousands of hectares in a single training exercise (Quist et al., 2003). The

impacts of such activities range from minor soil compaction and lodging of standing

vegetation to severe compaction and complete loss of vegetation cover in areas where

training is concentrated (Wilson, 1988; Milchunas et al., 1999). The resultant impacts are

evident both in the stream hydrology and the stream ecosystems. Potential impacts on

terrestrial and aquatic ecosystems include disruption in soil density and water content

(Helvey and Kochenderfer, 1990), addition of sediment, nutrients, and contaminants in









aquatic ecosystems (Gjessing et al., 1984), and impairment of natural habitat

development,and woody debris dynamics in forested floodplain streams (Piegay and

Landon, 1997). Military roads can significantly alter hillslope hydrology by

redistributing soil and rock materials on slopes and increasing the rate of debris slide

initiation. Roads also directly change the hydrology by intercepting shallow groundwater

flow paths, diverting the water along the roadway and routing it to stream at road

crossings. Road crossings can intercept groundwater drainage networks and collect

groundwater from upslope areas, diverting it into drainage ditches as surface water

(Wemple et al., 1996). Road crossings commonly found in military training areas may

act as barriers to the movement offish and other aquatic habitats (Fumiss et al., 1991).

Given the nature of military land use, management or military testing and training,

military land managers face the conflicting demands of balancing the primary military

mission with legal requirements to protect land and water quality (Milchunas et al.,

1999).

The military impacts can result in significant disruptions to the water and nutrient

cycles in the freshwater ecosystems and water resources. A few studies in the past have

addressed the effects of military training on terrestrial and aquatic ecosystems. For

example, Wilson (1988) found that the tank traffic resulted in a significant loss of native

species, increased abundance of introduced species, and increased bare soil at a training

site located in Manitoba. Milchunas et al. (1999) examined the effects of military

vehicles on plant communities and soil characteristics in Pinon Canyon Maneuver Site,

Colorado. Whitecotton et al. (2000) examined the impact of foot traffic from military

training on soil bulk density, infiltration rate, and aboveground biomass. Recently, Quist









et al. (2003) conducted a study in the Fort Riley Military Reservation, Kansas, to study

the effects of military use on terrestrial and aquatic communities.

Understanding human impacts in many landscapes needs the identification of

critical landscape elements and analysis of landscape pattern change (O'Neill et al.,

1997). Generalized response to military training includes the reduction in native and

perennial grasses, abundance of introduced species, and an increase in bare soil. Clearly,

these responses will alter biogeochemical and hydrological processes, which regulate

nutrient and water dynamics. However, impact of landscape changes to water quality and

stream is not well understood (Wand et al., 2001).

In order to maintain and improve water quality, there is an increasing need to

understand the relationships among watershed land use and stream ecosystems (Wang et

al., 2001). In parallel, perturbation induced effects on water quantity are critical to

stream ecology.

One of the many approaches to study the functional interrelations between the

hydrology and the stream biota is streamflow characterization and classification. This

approach develops hydrologic indices that account for characteristics of streamflow

variability that are biologically relevant (Olden and Poff, 2003). Characterization of

streams through development of ecologically relevant hydrologic indices is based on

long-term streamflow data. However, past studies overlooked the importance of storm-

flow data for the development of such hydrologic indices. Flow characteristics are

important where changes in land use are anticipated and where alterations to the flow

regimes need to be assessed.









The leaching of organic or mineral products from the forest floor provides potential

additional effects on water quality. Fluxes of nitrogen through the riparian zone are

intrinsically linked to water movement, both over and through the soil, and are also

strongly influenced by biological processes occurring in that zone. Nitrogen and organic

carbon dynamics in riparian zones are closely interrelated. While many of the factors

that can potentially influence nitrogen and carbon fluxes through riparian zones are

broadly known, there is presently incomplete quantitative information on the relative

importance of the flushing of nitrogen from freshly fallen leaves during precipitation

events.

Objectives

The United State's Department of Defense (DOD) policy has established

ecosystem management as its approach to manage the military lands by maintaining and

improving the sustainability and biological diversity of terrestrial and freshwater

ecosystems while supporting human needs, including the DOD missions. In order to

identify critical deficiencies and research opportunities on ecosystem management

problems on defense installations, the Strategic Environmental Research and

Development Program (SERDP) of DOD initiated the SERDP Ecosystem Management

Program (SEMP) in December 1997. The objectives of SEMP were to (1) establish long-

term research sites on DOD lands for military-relevant ecosystem research, (2) conduct

ecosystem research and monitoring activities relevant to DOD requirements and

opportunities, and (3) facilitate the integration of results and findings of research into

DOD ecosystem management practices.

The goal of this research is to illustrate how an ecohydrological approach could be

used to advance our ability to predict the effects of anthropogenic perturbations on water-









vegetation-nutrient interactions in the military installation at Fort Benning, Georgia. The

issues addressed in this research encompass many of the main scientific challenges in the

military installations' ecohydrology that include the effects of military related

perturbations on stream water quality and quantity, and the value of studying nutrient

dynamics in the riparian corridors of such regions. The specific objectives of this

research include (1) identification and examination of the statistical relationships among

water quality parameters and the watershed physical characteristics in low-nutrient

watersheds, (2) identification and development of hydrologic indices that characterize the

impact of military land management on watersheds, and (3) investigation of the effects of

nitrogen leaching from freshly fallen leaves on nutrient dynamics in a riparian area.

Dissertation Organization

Each chapter of this dissertation, except Chapters 1 and 5, is written as a self-

contained individual paper focusing on a topic that has not been addressed before.

Contributions are in the areas of water quality and land use within a military installation

(Chapter 2), development of storm-based hydrologic indices (Chapter 3), and watershed

scale nutrient leaching from a riparian area (Chapter 4). Chapter 5 summarizes and

concludes the research work

Chapter 2 is an original contribution in the effects of military activities related land

use on surface water quality. The major results from Chapter 2 are that the

concentrations of total organic carbon, total Kjeldahl nitrogen, total suspended solids, pH,

and total phosphorus in the stream show the greatest susceptibility to direct effects of

military activities. This chapter also identifies significant statistical relationships among

the water quality parameters and the military land uses. These relationships provide the









guidance for maintaining the surface water quality within the Fort Benning military

installation.

Chapter 3 is an original contribution in the ecohydrology that presents both annual-

based and storm-based methods for determining hydrologic indices that are of ecological

importance. Detailed descriptions of the methods for determining these indices and their

significance in aquatic ecosystems are described in this chapter. To statistically

characterize the hydrologic variation among different watersheds, 32 annual-based

hydrologic indices are analyzed. Eight out of 32 annual-based indices are recommended

to use for management practices within the Fort Benning military installation. As a new

approach to characterize the streamflow variability, 18 storm-based indices are proposed.

These indices are grouped into magnitude, frequency, duration, and rate of change of

flow. Storm-based indices are compared with annual-based indices. The storm-based

methodology provides guidance for measurements of appropriate indices within the

military installation.

Chapter 4 is an original contribution to the water quality function of the riparian

area. Nitrogen leaching from freshly fallen leaves in a riparian area during the

precipitation events is quantified. The observed nitrogen masses in the stream during the

precipitation events are compared with the model simulated values. Riparian Ecosystem

Management Model (REMM) is used to quantify the nitrogen in the riparian area. In this

chapter, details of the nitrogen leaching from the freshly fallen leaves are explored. The

results showed that the model effectively captured the trends of litter mass accumulation

in the riparian area and subsequent high concentrations during those periods. Analysis






8


showed that the simulated total Kjeldahl nitrogen masses during the precipitation events

were close to the observed masses.

Chapter 5 summarizes and concludes the research work. This chapter contains

recommendations to improve management of the water resources within the Fort Benning

military installation. Future research needs are also outlined in this chapter.














CHAPTER 2
ECOLOGICAL INDICATORS IN FORESTED WATERSHEDS IN FORT BENNING,
GEORGIA: RELATIONSHIP BETWEEN LAND USE AND STREAM WATER
QUALITY

Introduction

Ecological monitoring is essential to protect ecological health and integrity. As

human activity alters land cover, degradation of water resources begins in the upland

areas of a watershed. The first step toward effective ecological monitoring and

assessment is to realize that the ultimate goal is to measure and evaluate the

consequences of human actions on ecological systems. Human activities that alter land

use eventually affect biogeochemical processes that influence water quality and alter

ecological processes.

The National Research Council of the United States recently conducted a critical

evaluation of indicators used to monitor ecological changes from either natural or

anthropogenic causes. During recent decades, efforts have been increasing to develop

reliable and comprehensive environmental indicators because of growing environmental

concerns (National Research Council, 2000). Indicators rapidly and effectively

communicate system status. Ecological indicators help to elucidate both the effects of

human activities and natural processes. They can also help to assess future implications

of these factors on ecosystem integrity. Once indicators identify areas or elements of the

environment that are under stress, successful management of problems can be measured

relative to both interim targets and long-term goals.









Indicators that relate key ecological responses to human perturbations provide

useful tools to better understand ecological effects and their monitoring and management.

A suite of indicators ranging from microbiologic to landscape metrics is necessary to

capture the full spatial, temporal, and ecological complexity of impacts (Dale et al.,

2002). Evaluation of representative indices across major physical gradients (e.g., soils,

geology, land use, water quality and quantity) can signal early environment change and

help diagnose the cause of an environmental problem.

Understanding human impacts in many landscapes needs the identification of

critical landscape elements and analysis of landscape pattern change (O'Neill et al.,

1997). Attention has refocused on relationships among watershed characteristics and

stream water quality (Johnson et al., 1997). In order to maintain and improve water

quality, there is an increasing need to understand the relationships among watershed land

use and stream ecosystems (Wang et al., 2001).

Land use provides information about ecosystem function and characterizes the

extent and diversity of ecosystem types. National Research Council (2000) has

recommended land use as one of the most effective indicators for ecological assessment.

Hydrologists and aquatic ecologists have long known that the pathway by which water

reaches to a stream or lake has a major effect on water quality. Early studies on the

physical (Harrel and Dorris, 1968) and chemical (Hynes, 1960) characteristics of

watersheds focused on the influence of geomorphic characteristics such as drainage area,

gradient, and stream order on turbidity, dissolved oxygen concentration, and temperature.

Many recent studies examine the influence of terrestrial ecosystems on stream or wetland

water quality (Richards and Host, 1994; Richards et al., 1997). Many other studies have









found relationships between land use and concentrations of nutrients in streams (e.g.,

Hunsaker and Levine, 1995; Johnes et al., 1996; Bolstad and Swank, 1997). Watershed

properties constrain in-stream physicochemical and biotic features. Richards et al. (1996)

showed that ecosystems could be influenced by land use at regional or broad geographic

scales. Osborne and Wiley (1988) found that the distance of urban land cover from the

stream effectively predicts stream nitrogen and phosphorus concentrations.

Within a military installation context, land managers are challenged to use the land

for military training purposes in a manner that is both ecologically sound and meets

military mission requirements (Garten Jr et al., 2003). Lands can suffer a slow

degradation if over-utilized by long-term human activities. The heavy vehicles used in

mechanized military training cause disturbance of soil structure and can change the

physical properties of the soil (Iverson et al., 1981). In rangelands, tracked vehicle traffic

affects the hydrological characteristics (Thurow et al., 1993). Trampled vegetation,

vehicle tracks through undisturbed area, and erosion caused by the overuse of trails are

some examples of the visible degradation to a landscape caused by military training

exercises. Few studies have developed predictive relationships among watershed

physical characteristics and surface water chemistry specific to military land use and low-

nutrient systems.

In the coastal plain of the Apalachicola-Chattahoochee-Flint (ACF) river basin,

cropland and silvicultural land in upland areas is separated from streams by relatively

undisturbed riparian flood plain and wetland habitats (Frick et al., 1998). This is in

contrast to many intensively farmed areas of the United States where wetlands have been

drained, channelized or filled, and little or no riparian buffers remain between cropland









and streams. Frick et al. (1998) reported that the lower nutrient concentrations in streams

within the ACF river basin could partially be attributed to wetland buffer areas, and

minimal use of pesticides as compared to other areas of the United States. Other studies

in the southeastern coastal plain watersheds (e.g., Lowrance 1984; Lowrance et al., 1992;

Perry et al., 1999; Fisher et al., 2000) focused primarily on the agricultural impacts, and

urbanization on stream water quality. The contribution of areas affected by military

training to nutrient discharges, specifically in Fort Benning watersheds, is yet to be

quantified.

This paper identifies and examines the statistical relationships among water quality

parameters and the watershed physical characteristics in seven low-nutrient watersheds

located in the Fort Benning military installation, Georgia. It is hypothesized that surface

water quality parameters can be used as indicators of ecological changes in watersheds.

Study Area

The Fort Benning Army Installation occupies approximately 73,503 ha in

Chattahoochee, Muscogee, and Marion Counties of Georgia and Russell County of

Alabama (Figure 2-1). The climate at Fort Benning is humid and mild. Rainfall in this

region occurs regularly throughout the year. July and August are the warmest months

with average daily maximum and minimum temperatures of 37 and 15C. An average

daily maximum and minimum temperature of 15.5 and -1IC are reported in the coldest

months, January and February. Annual precipitation averages 1050 mm with October

being the driest month (Dale et al., 2002). Most of the precipitation occurs in the spring

and summer as a result of thunderstorms. Heavy rains are typical during the summer but

can occur in any month. Snow accounts for less than 1% of the annual precipitation.









Fort Benning is located within the southern Appalachian Piedmont and Coastal

Plains. The northern boundary of the installation lies along a transition zone between the

Piedmont and Upper Coastal Plain. The soils in the area are dominated by loamy sand

with some sandy loam. Following establishment of the installation in 1918, with

subsequent additions in 1941, we see that heavy training impacts only selected, mostly

upland, portions of the installation. Many areas are maintained as safety buffers, and

have little military use. Timber management includes harvesting and thinning. The

loblolly and longleaf pine forests are subjected to regular low-level fires for management

purposes (Dale et al., 2002).

Methods of Study

Description of Watersheds

The study watersheds, Bonham-1 and Bonham-2, Bonham, Little Pine Knot, Sally,

Oswichee, and Randall (named for the creek which drains the watershed), within Fort

Benning represent a range of region's soils, topography, land use, and vegetation

communities (Figure 2-1). These watersheds have a heterogeneous land cover

predominantly consisting of either forested or open areas. Forested areas are broadly

characterized as mixed pine and hardwoods or pine that are mostly 30-50 years old with

the soils in A-horizon range approximately 1-10 cm in depth (Garten Jr et al., 2003).

Open areas are either military, brush, or managed wildlife openings. Other cover

includes upland and bottomland hardwood forests. The military openings are clear-cut

parcels of land dominated by grass and bare soil that are used as military training

grounds. The brush openings consist of tall grass and immature hawthorn. The wildlife

openings are natural openings in the forests that are vegetated primarily by grass. Land










impacts due to heavy military activities (e.g., infantry, artillery, wheeled, and tracked

vehicle training) occur only in selected portions in these watersheds.


Figure 2-1. Study watersheds, Bonham-1, Bonham-2, Bonham, Little Pine Knot,
Randall, Sally, and Oswichee, in the Fort Benning military installation. Also
shown are stream network and sampling locations.

Characterization of Disturbance Categories

A disturbance index (DIN) was defined to characterize the watersheds at Fort

Benning into two disturbance categories: low-impact and high-impact. A DIN is the sum

of the area of bare ground on slopes greater than 3 degrees and on roads, as a proportion

of the total watershed area (Maloney et al., in press). The road areas were estimated by

multiplying their length by the measured average width of 20 m. Percentage of bare

ground was determined by using TM imagery and slope was derived from digital


&Sampihi, Locafflea
z fr efgFaulns

OM anm1

MI Lfttlo Me Knot
RaRndadt
O~wmchee
rFot oonuing


6 a 6 id 16 P3113MUN49









elevation maps. The TM imagery and digital elevation maps were obtained from

Strategic Environmental Research and Development Plan (SERDP)'s Ecosystem

Management Project (SEMP) database. Watersheds having a disturbance index from 0 to

11% are designated as low-impacted watersheds. High-impacted watersheds have

disturbance indices greater than or equal to 11%.

Collection and Analysis of Stream Samples

Surface water quality data were collected at seven streams biweekly from October

2001 to November 2002, and monthly thereafter to September 2003. Water samples

were collected in high-density polyethylene bottles. Bottles were soaked in de-ionized

water and rinsed with sample water prior to collection. The filtration was conducted at

the sampling sites using 0.45 |tm pore size polyethersulfone membranes. Filtered sample

was used to determine chloride (Cl) concentration, whereas raw sample was used for total

suspended solids (TSS) determination. Unfiltered samples for analyzing total Kjeldahl

nitrogen (TKN), total phosphorus (TP), and total organic carbon (TOC) were acidified

using double distilled sulfuric acid. The stream water pH, conductivity, and temperature

were measured at the time of sampling. All samples were kept cool in an icebox,

transported to the Soil and Water Science Department laboratory, University of Florida,

and refrigerated until analyzed. All samples were analyzed using standard methods

(American Public Health Association, 1992).

Statistical Analyses

The Pearson's correlation coefficients were calculated to examine the strength and

significance of the relationships between a watershed physical characteristic and a water

quality parameter. Two-sample t-tests were performed at 5% level of significance to test

whether mean values of watershed physical characteristics and water quality parameters









differ between low- and high-disturbance watersheds. Characteristics showing

significant correlations with a water quality parameter were considered for stepwise

multiple linear regression models. Only variables having less than or equal to 0.05

significance level were retained in the regression models.

Results

Watershed Physical Characteristics

The watersheds' physical characteristics are summarized in Table 2-1. Most of the

watersheds are highly vegetated (70% or more) except Oswichee (38%) with the majority

characterized by pine and mixed pine and hardwoods. Deciduous forest typically covers

only a small percentage of these watersheds. However, Bonham-1 consists of 27% of

deciduous forest. The study watersheds range from less than 1 to 84 km2. The

topographic characteristics of study watersheds are typical of forested watersheds of

southeastern coastal plain (Lowrance, 1992; Perry et al., 1999). Average elevations vary

from 104 to 148 m above mean sea level. Maximum slopes vary from 4 to 6 degrees.

Sandy soils are common in most of the study watersheds. However, loamy soils cover

most of the Sally and Oswichee watersheds. Bottomlands comprise 6 to 20% of the

watershed. The military training extent (0 to 6%) is relatively small. Total bare lands in

these watersheds comprise 9 to 21% of the watershed area, of which 1 to 8% of the total

area is unpaved roads and trails. This extent and variability of military training and bare

land are typical of the entire Fort Benning installation.

Some watershed characteristics are strongly correlated at significance level of 0.05

or lower (Table 2-2). Significant positive correlations exist for pine with the bottomland

wetlands, deciduous vegetation with stream density, number of roads crossing streams

with road length and percent of loam, and disturbance index with percent bare land.









Negative correlations were found for mixed vegetation with percentage of loamy soil and

number of roads crossing streams (NRC), military areas with normalized difference

vegetative index (NDVI) and stream density, and DIN, and road density with NRC.

Table 2-1. Physical characteristics of study watersheds in Fort Benning, Georgia.
Acronyms BON-1, BON-2, BON, LPK, OSW, Ran, and SAL represent
Bonham-1, Bonham-2, Bonham, Little Pine Knot, Randal, and Sally,
respectively.
Watersheds


Physical Characteristics
Topography
Area, km2
Average Elevation, m
Average Slope, degree
Vegetation
Pine, %
Deciduous, %
Mixed, %
Wetland, %
Military Land, %
NDVI
Soil
Sandy loam, %
Loamy sand, %
Road
Road Length, km
Road Density, km/km2
Stream
Stream Length, km
Stream Density, km/km2
Stream Order
Other
No. of Roads Crossing
Streams
Bare Land, %
Disturbance Index, %


BON-1 BON-2 BON


0.76
121.8
5.46


28
27
39
6
0
0.36


78
9


2.21
133.5
4.89


30
6
50
8
6
0.30


69
9


12.73
125.5
5.04


40
8
22
9
5
0.32


69
31


3.6 11.4 51.6
4.8 5.1 4.1


2.6
3.4
2



1
1
11


3.9
1.7
2



2
11
21


29.1
2.3
4



13
11
19


LPK OSW RAN SAL


18.01
146.3
5.32


41
2
34
17
2
0.34


72
28


56.6
3.1


43.3
2.4
4



11
4
11


83.39
104.2
4.48


26
0
5
7
5
0.35


24
73


196.6
2.4


170.6
2.1
5



55
4
9


74.38
136.8
4.57


58
3
9
20
3
0.34


68
32


415.1
3.1


323.5
2.4
6



43
4
10


25.31
136.8
5.42


48
12
15
10
2
0.36


49
51


97.6
3.8


65.2
2.6
4



21
7
15












Table 2-2. Pearson correlation coefficients between watershed characteristics. Characteristics are acronymed as follows: Pine forest
(PIN), Deciduous forest (DCD), Mixed forest (MXD), Wetland (WET), Military land (MIL), Sandy Soil (SND), Loamy
Soil (LOM), Road Length (RDL), Road Density (RDN), Stream Density (STD), Normalized Difference Vegetative Index
(NDVI), No. of Roads Crossing Streams (NRC), % Bare Land (PBL), and Disturbance Index (DIN). *, **, and ***
indicates significance at or below 0.05, 0.01, and 0.001 probability levels, respectively.
PIN DCD MXD WET MIL SND LOM RDL RDN STD NDVI NRC PBL
DCD -0.25
MXD -0.43 0.40
WET 0.96*** -0.33 -0.42
MIL -0.20 -0.65 0.00 -0.18
SND 0.21 0.44 0.67 0.17 -0.30
LOM 0.06 -0.50 -0.85* 0.14 0.15 -0.93***
RDL 0.63 -0.47 -0.74 0.50 0.05 -0.28 0.41
RDN -0.26 0.65 0.81* -0.34 -0.01 0.63 -0.81* -0.64
STD 0.01 0.84* 0.04 -0.04 -0.92*** 0.34 -0.24 -0.15 0.17
NDVI 0.10 0.41 -0.48 0.07 -0.80* -0.30 0.41 0.22 -0.39 0.74
NRC 0.23 -0.58 -0.90** 0.19 0.21 -0.76* 0.83* 0.81* -0.86** -0.27 0.33
PBL 0.00 -0.28 0.25 0.09 0.72 0.04 -0.11 -0.34 0.40 -0.67 -0.79* -0.30
DIN -0.10 0.05 0.52 -0.07 0.55 0.30 -0.42 -0.52 0.71 -0.44 -0.77* -0.58* 0.93***









Water Quality Parameters

Water quality parameters in the study watersheds varied over the sampling period

and among watersheds. Variability of water quality parameters among watersheds is

observed (Figure 2-2). Mean pH in the study watersheds ranged from 4.2 to 7.0. Mean

conductivity ranged from 16.4 to 44.5 [[S/cm. Mean temperatures varied from 17.5 to

20.8C. Low concentrations of TP and TKN were observed in all the watersheds under

study as compared to forested watersheds in the southeastern coastal plain watersheds

(Lowrance et al., 1984), and across the United States (Meader and Goldstein, 2003;

Fisher et al., 2000). Concentrations of TKN, TP, and Cl were often below the detection

limit. Mean concentrations of TP varied widely, ranging from 0.003 to 0.020 mg/L and

TKN varied from 0.20 to 0.35 mg/L; TOC from 1.35 to 3.33 mg/L; Cl from 1.46 to 4.13

mg/L; and TSS from 4.15 to 10.30 mg/L. As depicted in Figure 2-2, each stream

exhibited distinct water quality signatures with the exception of temperature and TSS.

Seasonal variations in the water quality parameters are responsible for much of this

observed variability among watersheds.

Stream pH fluctuated more during June and July and was elevated during

December through February. Conductivity values showed slight fluctuations from May

to July. On multiple occasions, high conductivity values were observed in the Randall

stream. Chloride, TKN, and TP showed distinct seasonal patterns. The concentrations of

these parameters were low from June to September, and high from March to May and

from October to December. In contrast, TOC peaked from August to October and again

from March to July. Higher concentrations of TSS were observed from July to

September in all the streams.














10 -
9-



S6-
a

4-
3-
2-

90-






so -
S70 -




10 -

0.107






0.0B2

0.00-


g4


2

1-


1 2 h 4 15
Wasehed ID


6 7


35-


20-

15
10-
E-' 5


1.0



gS
405
H


00_


10 -





C
u





70
70 -
60-

S30-
30 -
20 -
10 -
0..


1 2 3 W 5M
Watershed I0


Figure 2-2. Box plots of water quality parameters. Each plot consists of outliers, most
extreme data, 75th, 50th, and 25th percentile values. Watershed IDs
represent- 1: Bonham-1, 2: Randall, 3: Oswichee, 4: Little Pine Knot, 5: Sally,
6: Bonham, and 7: Bonham-2.


T yT


*


4*a
*



N 4






: ^ l ll


6 7


I I I I I I I










$~$-~=B










f d e~B,









Et3 6Q









Effects of Disturbance Categories

Table 2-3 presents the t-test results of the comparison of physical characteristics

and water quality parameters based on the watershed disturbance level. The only

physical characteristic having a significant difference (c = 0.04) between these two

groups was DIN. While the low-impact watersheds tended to have higher chemical

Table 2-3. t-test results for differences in mean values of watershed physical
characteristics and water quality parameters. indicates significance at or
below 0.05 probability level. NS indicates non-significant difference at the
0.05 probability level.
Low-Impacted High-Impacted
Mean SD Mean SD
Watershed Characteristics
Pine, % 38.3 14.8 39.3 9.0 NS
Deciduous, % 8.0 12.7 8.7 3.1 NS
Mixed, % 21.8 17.2 29 18.5 NS
Wetland, % 14.5 9.5 16 7.2 NS
Military Land, % 2.4 2.0 4.3 2.3 NS
Sand, % 60.5 24.7 62.3 11.5 NS
Loam, % 35.5 26.9 30.3 21.0 NS
NDVI 0.35 0.01 0.32 0.03 NS
Road Length, km 168 184 53.5 43.2 NS
Road Density, km/km2 3.3 1.0 4.3 0.7 NS
Stream Density, km/km2 2.6 0.6 2.2 0.4 NS
No. of Roads Crossing Streams 27.5 25.6 12 9.5 NS
Disturbance Index, % 10.2 0.9 18.3 3.1 *
Water Quality Parameters
pH 5.6 1.3 4.5 0.3 NS
Temperature, C 19.3 0.9 18 0.3 NS
Conductivity, [iS/cm 25.9 13.3 20.4 2.5 NS
TKN, mg/L 0.3 0.03 0.2 0.05 NS
TP, mg/L 0.011 0.006 0.007 0.003 NS
TOC, mg/L 2.9 0.4 2.1 0.7 NS
Cl, mg/L 2.4 0.5 1.8 0.3 NS
TSS, mg/L 9.1 2.2 4.8 0.5 *









concentrations than the high-impact watersheds, only TSS showed significant difference

(a = 0.03). Even though the results showed no significant statistical differences at a

confidence level of 95%, the t-test results of all water quality parameters, except

conductivity and TP, showed significant differences at 80% confidence interval between

high- and low-impacted watersheds. The relatively small sample size and natural

variability among watersheds may have limited the ability to discern significance

differences.

Relationship between Watershed Physical Characteristics and Water Quality
Parameters

Correlation and regression analyses were performed to identify relationships among

the watershed physical characteristics and the water quality parameters. Table 2-4 shows

that each water quality parameter had a significant relationship with one or more

watershed physical characteristics (Table 2-4). The correlation results show that

decreasing mixed vegetation increased pH and TP. Sandy and loamy soils had opposite

effects on TP. An increase in sandy soil decreased TP, whereas an increase in loamy soil

increased TP. Increasing military land decreased TOC. Temperature, pH, conductivity,

and Cl increased as the road length increased. The number of roads crossing streams had

positive correlations on pH and TP. Percent bare land was negatively correlated with

TOC and TSS. Disturbance index was negatively correlated with TKN and TOC.

Graphical relationships provide insight into nonlinear relationships that exist

between indicators and response variables. Figures 2-3 to 2-6 show some of the most

striking relationships between watershed characteristics directly and/or indirectly affected

by management of military lands and their effect on water chemistry. As the % military

land increases, the TKN, TOC, and TSS decrease in a linear fashion. However,









Table 2-4. Pearson correlation coefficients between watershed characteristics and water
quality parameters. Characteristics are acronymed as follows: Pine forest
(PIN), Deciduous forest (DCD), Mixed forest (MXD), Wetland (WET),
Military land (MIL), Sandy Soil (SND), Loamy Soil (LOM), Road Length
(RDL), Road Density (RDN), Stream Density (STD), Normalized Difference
Vegetative Index (NDVI), No. of Roads Crossing Streams (NRC), % Bare
Land (PBL), and Disturbance Index (DIN). *, **, and *** indicates
significance at or below 0.05, 0.01, and 0.001 probability levels, respectively.
pH Temperature Conductivity TKN TP TOC Cl TSS
PIN 0.36 0.78 0.54 0 -0.09 0.15 0.56 -0.14
DCD -0.47 -0.37 -0.40 -0.12 -0.33 0.39 0.13 0.42
MXD -0.79* -0.55 -0.38 -0.39 -0.83* -0.24 -0.49 -0.11
WET 0.24 0.62 0.44 0.11 -0.11 0.18 0.37 -0.29
MIL 0.09 0 -0.10 -0.50 0.10 -0.87** -0.47 -0.53
NDVI 0.32 0.11 0.10 0.60 0.46 0.80* 0.55 0.54
PBL -0.45 -0.25 -0.51 -0.72 -0.42 -0.81* -0.68 -0.87**
DIN -0.65 -0.36 -0.60 -0.84* -0.63 -0.76* -0.65 -0.72
SND -0.50 -0.02 0.14 -0.19 -0.78* 0.15 0.06 0.15
LOM 0.57 0.16 0.03 0.39 0.81* 0.07 0.07 -0.14
RDL 0.94*** 0.95*** 0.82* 0.19 0.57 0.09 0.79* 0.30
RDN -0.75* -0.42 -0.54 -0.72 -0.78* -0.35 -0.32 -0.15
NRC 0.93*** 0.59 0.51 0.36 0.90** 0.06 0.47 0.19
STD -0.14 -0.11 -0.01 0.39 -0.04 0.82* 0.44 0.65

disturbance index may operate as a threshold indicator of pH and TSS where pH

decreases in response to a relatively low level DIN while the TSS threshold for DIN

impact is somewhat higher. No significant relationships are found between the water

quality parameters and extent of pine and deciduous forest. Similarly, wetland showed

no effect on these parameters.

Stepwise multiple regressions identified relationships between the water quality

parameters and watershed physical characteristics that are susceptible to the disturbances

(Table 2-5). A statistically significant regression model was found for every water

quality parameter. Prediction of pH variability among watersheds is particularly well

captured by pine forest and road length. The regression relationships indicate that all of









the water quality parameters depend on at least one aspect of military management.

Several water quality parameters, Cl, TP, TOC, TSS, and TKN, depend only on

management aspects of the military installation. For example, Cl depends on change in

military land and road length. Total phosphorus is strongly related to the number of

roads crossing streams. However, the influence of vegetation and soils characteristics is

clearly important in pH, conductivity, and temperature. For example, conductivity

appeared to be well captured by area covered by sandy soil and the number of roads


Table 2-5. Stepwise multiple regression models for water quality parameters. pH is
unitless, temperature is measured in degrees centigrade, conductivity is
measured in mS/cm, TP, TKN, TOC, Cl, and TSS are measured in mg/L.
Pine forest (PIN), Military land (MIL), Sandy Soil (SND), Loamy Soil
(LOM), Road Length (RDL), Road Density (RDN), No. of Roads Crossing
Streams (NRC), Percent Bare Land (PBL), and Disturbance Index (DIN) are
the independent variables retained in the regression analyses. *, **, and ***
indicates significance at or below 0.05, 0.01, and 0.001 probability levels,
respectively.
Water Quality Independent Variables Retained and
Parameters Regression Equations R2


pH

Cl


TP


Conductivity


Temperature

TOC


TSS


Pine, Road Length
5.50 0.0382 PIN + 0.00924 RDL
Military, Road Length
2.19 + 0.145 MIL + 0.00344 RDL
No. of Roads Crossing Streams
0.00451 + 0.000236 NRS
Sandy Soil, No. of Roads Crossing Streams
- 24.2 + 0.552 SND + 0.666 NRS
Loamy Soil, Road Density
26.1 0.051 LOM 1.49 RDN
Military Land
3.45 0.274 MIL
Percent of Bare Land
11.2 0.648 PBL
Disturbance Index
0.445 0.0109 DIN


0.98***


0.90**


0.81**


0.80*


0.77*

0.76**


0.76**


TKN


0.70*
















R = 0.00
T


R2 = 4E-06


-S





-R = 0.01






I


R =0.22





I I


R = 0.01


R2 = 0.76


2
0


60

- 40

S20

0
C


0.025
0.020
0.015
0.010
0.005
0.000


Military (%)


Figure 2-3. Relationships between military land and water quality parameter. Vertical
bars represent standard errors.


I I


-R = 0.27




I I
m--j


Military (%)
















R = 0.55


8
6
4

2
0


60

40

S20
S0
C
i o


0.025
0.020
0.015
0.010
0.005
0.000


R2 = 0.29


RO = 0.51





-I I


R2 = 0.11



C1




R = 0.02




I I

0 2 4 6

Road Density (km/km2)


Figure 2-4. Relationships between road density and water quality parameter. Vertical
bars represent standard errors.


- I I I


R2 = 0.17




-2
-^





SR = 0.61








R =0.10


tt I




0 2 4 6
Road Density ( )
Road Density (kmn/km )


-




-
















8 -
6-
4-
2-
0


60 _
S40 -
20 -
0

C0

0.030 -
S0.020 -
g
0.010
0.000


4-

S2-

S-
0


R = 0.86


R = 0.35


R = 0.81








R = 0.22



S30 -
5 20
10
0-


S0.6

0.4

0.2

0.0



6


2
0


20 -
S15-
g 10 -
S5-
0


0 20 40 60


Number of Roads
Crossing Streams


R = 0.25


R2 = 0.12


2 0.00
- R = 0.00


I-----


R2 = 0.03






0 20 40 60


Number of Roads
Crossing Streams


Figure 2-5. Relationships between number of roads crossing streams and water quality
parameter. Vertical bars represent standard errors.


-


B














R2 = 0.42
if


R = 0.13




I I


R2 = 0.40


R2 = 0.42


R = 0.36
4-7


25
20 -
15 -
10 -
5
0-

0.5
0.4
0.3
0.2 -
0.1
0.0


R = 0.57


F~k^


0 10 20 30

Disturbance Index (%)


0 10 20 30

Disturbance Index (%)


Figure 2-6. Relationships between disturbance index and water quality parameter.
Vertical bars represent standard errors.


S= 0.69


60

S40

20

0
u


0.025
0.020
0.015 -
0.010 -
0.005
0.000


-4


R2 = 0.52


20 -
15
10
5
0


+


.









crossing streams, whereas temperature was captured by loamy sand soil and road density.

Overall, the regression models show that it is possible to quantify the effects of watershed

physical characteristics on the water quality parameters.

Discussion

Variations in stream water chemistry among the study watersheds reflect

differences in biogeochemical reactions occurring in the watersheds. The results of this

study indicate that even in low impact watersheds, physical characteristics may be used to

explain variations in stream water chemistry and, by inference, the relative watershed

disturbance levels. The observed variability in many of the chemical parameters studied

in these watersheds can be attributed to physical characteristics of the watersheds or land

management patterns within the watersheds as evidenced by the road network, forestry

practices, and military training.

Diverse human activities interact to affect conditions in watersheds and water

bodies. Sites of interest can be grouped and placed on a gradient according to activities

and their effects. The results of this study suggest that the vegetation type, road length,

number of roads crossing streams, and disturbance index are important predictors of

water quality variability. Vegetation cover was related to stream pH and TP (mixed

forest) and conductivity (pine forest). However, deciduous forest cover was not related

to any of the water quality parameters suggesting a limited effect in organic matter and

nutrient production and variability as observed among Fort Benning watersheds.

However some other subtle landscape changes resulted in relatively larger impacts on

water quality parameters. Low-impact watersheds tend to produce higher concentrations

of nutrients in the streams. This can be attributed to the availability of more soil organic









matter and the rapid biogeochemical processes occurring in the low-impact watersheds as

compared to the high-impact watersheds.

It is extremely difficult to capture all aspects of human influence in a single graph

or statistical test. However, sometimes meaningful chemical patterns can be lost by

excessive dependence on the outcome of menu-driven statistical tests (Karr and Chu,

1999). Figures 2-3 to 2-6 depict several different aspects of stream's chemical conditions

against several measures of human influences, such as military land use, disturbance

index, number of roads crossing streams, and road density. The distribution of circles in

most of these figures illustrate that a chemical metric indicates little about a condition

simply because it does not correlate strongly with a single surrogate of that condition.

However, where the relationship between human influence and stream's chemical

response is strong, statistics and graph agree.

The correlation tests identify linear relationships between a chemical response and

watershed characteristics. Weak statistical correlations observed in these analyses may

have missed important chemical patterns. For example, nonlinear patterns were observed

for forest types, bare land (not shown here), and disturbance index (Figure 2-6). The

plots in Figure 2-6 show a step-function for TSS and pH. The scatter of this dataset

shows little or no statistical significance, but can be interpreted chemically. For TSS,

those watersheds having a disturbance index of 11% or lower had a higher level than

those with a greater disturbance index.

When a number of variables interact to influence water quality conditions, it may

be difficult to explain observed variability in a single plot against one dimension of

human influence (Figure 2-4). Chemical responses were plotted against the road









densities for various watersheds. The Pearson correlation coefficient for TP was

significant capturing human influence on this chemical parameter. The response of TP is

visibly distinguished from others. A similar discussion is true for military land with TOC

(Figure 2-3); number of roads crossing streams with pH and TP (Figure 2-5); and

disturbance index with TKN (Figure 2-6).

The relationships between water quality parameters and physical characteristics

indicate that disturbances in low nutrient forested environments decrease some chemical

signatures. Watersheds with more roads, e.g., Randall and Oswichee, have relatively

high pH, conductivity, and Cl compared to the watersheds with fewer roads. Watersheds

with a small portion of military land, e.g., Bonham-1, Sally, and Little Pine Knot, have

relatively high TOC concentrations. In contrast, watersheds characterized by higher road

densities, e.g., Bonham and Bonham-2, had low TP concentrations. Higher disturbance

index, similar to the road density, showed lower TKN and TOC concentrations in the

streams. Mixed vegetation, road length, percent of bare land, DIN, and number of roads

crossing streams were able to capture most of the variability in water quality parameters.

In a watershed scale study conducted in Ontario, Canada, Sliva and Williams

(2001) found a negative correlation of forested land cover with TSS and chloride. In

contrast, Johnson et al. (1997) showed a positive relationship of forest with TSS in a

study of landscape influence on water chemistry in the Saginaw Bay watershed of central

Michigan. Johnson et al.'s (1997) results indicated that row crop agriculture had the

highest effect on total nitrogen, nitrate, and total dissolved solids. They also observed

that urban and forest areas were positively correlated good predictors of TSS, whereas

row crop agriculture was positively correlated with total nitrogen. Basnyat et al. (1999)









reported a positive association of TSS with agricultural practices in the Fish River

watershed, Alabama. The Fort Benning installation is characterized by relatively low

variability in forest cover and suggests, in contrast to other studies, that neither TSS nor

Cl may be related to forest cover under existing land management practices. Instead,

Fort Benning's road extent and percent bare land are better predictors of TSS and Cl.

Most studies identified urban land use as a dominant factor causing elevated total

nitrogen and nitrate concentrations in the streams (Hill, 1981; Osborne and Wiley, 1988).

Sponseller et al. (2001) found positive correlation of total inorganic nitrogen with

percentage of non-forested land in southwestern Virginia watersheds. A negative

correlation of TKN to DIN, in this study, is consistent with studies (e.g., Sponseller et al.,

2001; Hunsaker and Levine, 1995; Johnson et al., 1997) that have shown stream nitrogen

concentrations to be good predictors of non-forest area at the watershed scale.

In a study of 101 watersheds in New Zealand, Close and Davis-Colley (1990)

found that between 60 and 80% of the variance in conductivity, total nitrogen, and nitrate

was accounted for by landscape factors including geology and land use. However, in that

same study, landscape factors accounted for only 50% of the variance in ammonia and

phosphorus species. Their results parallel those found at Fort Benning in that no strong

relationships were observed between land use and TP. The strong negative correlation of

loamy sand soil with TP in Fort Benning is consistent with Hill's (1981) study, conducted

in a sandy loam region similar to portions of Fort Benning watersheds that reported

negative correlations between phosphorous concentrations and abandoned farmland and

forest.









Most variations in stream water chemistry are driven by climatic and biotic factors

and are therefore largely governed by the processes that are taking place in the terrestrial

part of the watershed such as natural or human induced vegetation cover changes

(Semkin et al., 1994). Our results show interactions among landscape factors and water

quality indicators. Results also indicate that it is possible to observe the response of these

water quality parameters to physical attributes of watersheds. The importance of water

quality parameters in the present study appeared to be attributable to the perturbations

related to military training and associated parameters within a watershed as these

parameters clearly captured the changes in physical parameters that are more sensitive to

such kind of influences.

Conclusion and Recommendations

Sometimes a single variable can capture and integrate multiple sources of

influence. More often, a small number of ecological attributes provide reliable signals

about ecological condition. Water chemistry prediction using watershed physical

characteristics in this study showed mixed results compared to the other investigations.

However, most of the watershed physical characteristics used in our analysis did explain

the variability in water quality parameters. This study documented strong relationships

between certain watershed physical characteristics that are more susceptible to human

induced perturbations, specifically military related disturbances, and water quality

parameters in military installation at Fort Benning. Watersheds with more roads crossing

streams tended to produce more TP. Total Kjeldahl nitrogen and TOC variations were

well captured by DIN and extent of military land, respectively. Total suspended solids'

variability, on the other hand, was captured by the percent of bare land within a

watershed. Road length captured most of the variability in pH and Cl. Conductivity and









temperature values were dependent on soil types and road characteristics. The variations

in stream water chemistry are largely attributable to disturbance levels and the types of

biogeochemical reactions occurring in the watersheds. Regression results suggest that

TOC, TKN, and TSS were useful indicators of watershed physical characteristics as they

are more susceptible to direct effects of military activities. Although pH, conductivity,

and TP showed good correlations with the road length, these parameters indicated strong

but indirect influence of military training activities on watersheds.

Foreseeing a single indicator of water quality that would be sensitive to all kinds of

perturbations in the watershed is extremely difficult. Ability to detect perturbations can

be related to spatial and temporal scales. It is necessary to recognize the effects of

natural disturbances on ecosystem structure and functioning. It is suggested that

priorities for determining ecological indicators specific to water quality should include

(1) development of framework to determine proper reference states within watershed

against which to detect loss of ecosystem health, (2) broadening our knowledge of

ecosystem sensitivity to perturbations of varying intensity, spatio-temporal distribution,

and type, and (3) development of suites of indicators necessary to detect the broadest

spectrum of perturbations in watersheds.














CHAPTER 3
HYDROLOGIC INDICES OF WATERSHED SCALE MILITARY IMPACTS IN
FORT BENNING, GA

Introduction

A goal of stream flow characterization and classification is to develop hydrologic

indices that account for characteristics of streamflow variability that are biologically

relevant (Olden and Poff, 2003). Broadly, indices are attributes that respond in a known

way to a disturbance i.e., they relate key ecological responses to human activities. Index

identification is based on the goals and objectives set for a particular ecosystem or region.

A good index should be sensitive to stressors, biologically and socially relevant, broadly

applicable to many stressors and sites, diagnostic of the particular stressor causing the

problem, measurable, interpretable, and not redundant with other measured indices

(Cairns et al., 1993).

Ecologically relevant hydrologic indices developed in the past not only characterize

particular regions, but also quantify flow characteristics that are sensitive to various

forms of human perturbations. For example, early studies on hydrological indicators

focused on variation of mean daily flow to study the pattern of fish in Illinois and

Missouri (Horwitz, 1978). In Great Britain, Moss et al. (1987) used average flow

conditions to predict macro-invertibrate fauna of unpolluted streams and Townsend et al.

(1987) examined persistence of community structure for benthic invertebrates. In arid

regions of southwest United States, Minckley and Meffe (1987) studied effects of short-

term flood frequency in stream fish communities. Poff and Ward (1989) used long-term









discharge records (17-81 years) of 78 streams from across the continental United States

to develop a general quantitative characterization of streamflow variability. Similarly,

Jowett and Duncan (1990) studied skewness in flows and peak discharges in relation to

in-stream habitat and biota in New Zealand.

More recent investigations have begun to focus on examining suites of hydrologic

indices that are ecologically relevant to quantify hydrologic regimes. These studies

report numerous such indices. For example, Poff and Allan (1995) studied stream fish

assemblage for 34 sites in Wisconsin and Minnesota in conjunction with long-term

stream flow variability and predictability as well as frequency and predictability of high

and low flow extremes. In the process of deriving ecologically relevant hydrologic

indices, Clausen and Biggs (1997) identified thirty-four hydrological variables from daily

flow records at eighty-three New Zealand sites. The authors related these variables to

benthic biota including periphyton and invertibrate species richness and diversity. Wood

et al. (2000) reported the importance of hydrological conditions in explaining the

ecological role when the authors studied the changes in macro-invertibrate community in

response to flow variations in the Little Stour River in the United Kingdom. Pettit et al.

(2001) described a method for assessing seasonality and variability of natural flows and

their influence on riparian vegetation in two contrasting river systems in western

Australia.

To isolate core flow variables for ecological studies, it is important to know not just

the ecological relevance of the variables, but also the interrelationships among the

variables in order to avoid redundancy in the analyses (e.g., Clausen and Biggs, 2000;

Olden and Poff, 2003). Hydrologic indices have been criticized for being overly









simplified and lacking adequate biological relevance. Stream ecologists are now facing

difficulty in choosing appropriate and relevant ones from the available suit of indices.

For example, the Indicators of Hydrologic Alteration (Richter et al., 1996) approach is

commonly used for characterizing human modification of flow regimes, yet it contains 64

statistics (32 measures of central tendency and 32 measures of dispersion), many of

which are inter-correlated (Olden and Poff, 2003).

To date, characterization of streams or regions through determination and

development of ecologically relevant hydrologic indices are based on long-term stream

flow data. However, given the multitude of methods to characterize stream flow, past

studies overlooked the value of storm flow data for the development of such ecologically

relevant hydrologic indices. Flow characteristics are especially important where changes

in land use are anticipated and where alterations to the flow regime need to be assessed.

The primary objective of the study is to identify hydrologic indices that

characterize the impact of military land management on watersheds in Fort Benning,

Georgia. Towards this end, this study investigates both storm and annual hydrographs. it

is hypothesized that, in addition to annual-based indices, storm-based hydrologic indices

are indicative of alteration in stream ecology. Here, a suite of event based hydrologic

indices is proposed. Storm-based and the annual-based indices are calculated and used to

compare and contrast impacted watersheds with a reference watershed. Additionally,

specific military land management practices are used to predict storm-based indices.

Flow Regimes and Hydrologic Indices

To assess hydrologic alterations within an ecosystem, Richter et al. (1996)

developed a method to compute representative, multi-parameter suite of hydrologic

characteristics that are of ecological relevance, commonly known as Indicators of









Hydrologic Alteration (IHA). Olden and Poff (2003) comprehensively reviewed

currently available hydrologic indices for characterizing streamflow regimes and

recommended non-redundant indices for various stream types that may differ in major

aspects of ecological organization. Poff (1996) provides a comprehensive catalog of the

stream types for small to mid-size relatively undisturbed streams, classified according to

variation in ecologically relevant hydrological characteristics, in continental United

States. The assessment of IHA as well as other studies (e.g., Poff and Allan, 1995;

Clausen and Biggs, 1997; Wood et al., 2000; Pettit et al., 2001; Olden and Poff, 2003) to

identify hydrologic indices is based on long-term flow data.

An alternative approach to identify ecologically relevant hydrologic indices is to

conduct an assessment based on the storm hydrograph. This approach is useful when

long-term data for a particular stream or region are not available, when significant data

gaps exist, or coincident records are not available. Storm hydrographs are traditionally

described by characteristics including peak flow, total volume of direct runoff, and

duration as shown in Figure 3-1. The time characteristics of the hydrograph and its

relationship to the precipitation event are presented in Table 3-1. Towards the goal of

characterizing the spatial variations of hydrologic conditions using storm-based indices

that are ecologically relevant as well as sensitive to human influences, the ecological

function of hydrologic characteristics that are relevant to storm-based hydrologic indices

are considered.

Examination of the storm hydrographs reveals numerous potential indices. A set of

18 storm-based ecologically relevant hydrologic indices that characterize variation in

water condition in individual watershed is proposed. Included proposed indices are














Precipitation
[Li



















Figure 3-1. Terms used to describe hyetograph and response hydrograph. Refer to Table
3-1 for the definitions of the terms.


Table 3-1. Definitions of terms used to describe hyetograph and response hydrograph.
Time instants Time durations











tpo= beginning of precipitation Tw = te tpo= duration of water input
tpc = centroid of precipitation TrI= tqO tpo= response lag
tpe = end of precipitation T tpkr- tqO time of rise

tqO= beginning of hydrograph rise T1p tpk tpO = lag-to-peak
tpk = time of peak discharge Tic tqc tpc = centroid lag
tq = centroid of response hydrograph Tb tqe qO time base
t = end of response hydrograph T= te tp = time of concentration


response factor (ratio of direct runoff depth to precipitation depth), baseflow index (ratio

of baseflow volume to total volume during an event), peak discharge, dimensionless

fife, fGt tsc A












Figure 3-1. Termsed to describe hyetograph and response hydrolag (T), time of rise (Tr), and time base (Tb),
3-1bankfull dischargfor the and the slopes of rising and falling limb of the hydrographs. A flowterms.
Table 3-1. Definitions of terms used to describe hyetograph and response hydrograph.
Time instants Time durations
tpo = beginning of precipitation T, = tp, tpo = duration of water input
tp, = centroid of precipitation Tri = tqo tpo = response lag
tpe = end of precipitation Tr = tpk tqo = time of rise
tqo = beginning of hydrograph rise TIP = tpk tpo = lag-to-peak
tpk = time of peak discharge T1c = tq, tpc = centroid lag
tq, = centroid of response hydrograph Tb = tq, tqo = time base
tq, = end of response hydrograph T, = tq, tpe = time of concentration

response factor (ratio of direct runoff depth to precipitation depth), baseflow index (ratio

ofbaseflow volume to total volume during an event), peak discharge, dimensionless

numbers related to hydrograph response lag (TrI), time of rise (Tr), and time base (Tb),

bankfull discharge and the slopes of rising and falling limb of the hydrographs. A flow









with 1.67-year return interval is often recognized as bankfull discharge (Poff, 1996). The

eighteen ecologically relevant hydrologic indices were divided into four components of

hydrologic regimes, magnitude, frequency, duration, and rate of change, to statistically

characterize hydrologic variation.

Effects of Flow Magnitude on Stream Ecology

This group includes 6 parameters (mean and coefficient of variation) related to

response factor, baseflow index, and the peak discharge (qpk). The magnitude of the

water condition at any given time is a measure of the availability or suitability of habitat

and defines such habitat attributes as wetted area or habitat volume, or the position of a

water table relative to wetland or riparian plant rooting zones (Richter et al., 1996). High

flows maintain ecosystem productivity and diversity. For example, high flows remove

and transport fine sediments that would otherwise fill the interstitial spaces in productive

gravel habitats (Beschta and Jackson, 1979). Floods import woody debris into the

channel (Keller and Swanson, 1979), where it creates new, high-quality habitat (Moore

and Gregory, 1988; Wallace and Benke, 1984). Floodplains and wetlands provide

important nursery grounds for fish and export organic matter and organisms back into the

main channel (Junk et al., 1989; Welcomme, 1992). The scouring of floodplain soils

revives habitat for plant species that germinate only on barren, wetted surfaces that are

free of competition (Scott et al., 1996) or that require access to shallow water tables

(Stromberg et al., 1997).

Effects of Flow Duration on Stream Ecology

The 6 parameters in this group measure the duration of all the events considered in

the analysis. The parameters included in this group are the mean, and coefficient of

variation of dimensionless indices Tri/Tic, Tr/Tic, and Tb/Tic, where TrI, Tr, Tb, and T1i









corresponding to the response lag, time of rise, time base, and centroid lag of a storm

hydrograph, respectively. The duration is the period of time associated with a specific

storm condition that determines the differences in tolerance to prolonged flooding in

riparian plants (Chapman et al., 1982). Changes in duration of inundation can alter the

abundance of plant cover types (Auble et al., 1994). For example, increased duration of

inundation has contributed to the conversion of grassland to forest along a regulated

Australian River (Bren, 1992). For aquatic invertebrates and fishes prolonged flows of

particular levels can also be damaging (Closs and Lake, 1996). Whether a particular life-

cycle phase can be completed or the degree to which stressful effects such as inundation

of a flood plain can accumulate may be assessed from the duration of time over which a

specific water condition exists (Richter et al., 1996).

Effects of Flow Frequency on Stream Ecology

The two parameters in this group measure the number of occurrences of the

magnitude of the stream flow condition with respect to bankfull discharge. These

numbers of occurrences of bankfull discharges are reported as percentages of the total

events under analysis. Measures of exceedance of bankfull conditions have greater

ecological importance. These flows regulate numerous ecological processes within

riparian as well as flood plain areas. Frequent, moderately high flows effectively

transport sediment through the channel (Leopold et al., 1964). Movement of sediment

and organic resources such as detritus and attached algae revive the biological

community and allow many species with fast life cycles and good colonizing ability to re-

establish (Fisher, 1983). Consequently, the composition and relative abundance of

species that are present in a stream often reflect the frequency and intensity of high flows

(Meffe and Minckley, 1987; Schlosser, 1985).









Effects of Flow Rate of Change on Stream Ecology

This group includes 4 parameters, mean rate of change in peak discharge in rising

limb and falling limb of the storm hydrograph and their variabilities. Flow conditions'

rate of change, or flashiness, refers to the rate at which flow changes from one magnitude

to another, can influence species persistence and coexistence. At the extremes, flashy

streams have rapid rates of change, whereas stable streams have slow rates of change

(Poff et al., 1997). The rate of change in water conditions may be tied to the stranding of

certain organisms along the water's edge or in ponded depressions, or the ability of plant

roots to maintain contact with phreatic water supplies (Richter et al., 1996). Non-native

fishes generally lack the behavioral adaptations to avoid being displaced downstream by

sudden floods (Minckley and Deacon, 1991). Meffe (1984) documented that a native

fish, the Gila topminnow (Poeciliopsis occidentalis), was locally extirpated by the

introduced predatory mosquitofish (Gambusia affinis) in locations where natural flash

floods were regulated by upstream dams, but the native species persisted in naturally

flashy streams.

Data Collection

Watershed Characteristics

The Fort Benning study watersheds (Figure 3-2), Bonham-1 and Bonham-2,

Bonham, Little Pine Knot, and Sally Branch (named for the creek/stream that drains the

watershed), represent a range of the region's soils, topography, land use, and vegetation

communities (Table 3-2). In addition to standard watershed characteristics, the military

training land area and a dimensionless military disturbance index (DIN) were calculated

for each watershed. The military land use include fire ranges, maneuver training areas

for light, amphibious, and heavy training, air field, artillery and mortar firing points,
































15 3 12


Figure 3-2. Study watersheds, Bonham-1, Bonham-2, Bonham, Little Pine Knot,
Randall, Sally Branch, and Oswichee, in the Fort Benning military
installation.

airborne drop zones, specialized non-live-fire training assets, and duded impact areas. A

DIN is the sum of area of bare ground on slopes greater than 3 degrees and on roads as a

proportion of the total watershed area (Maloney et al., in press). Additional details

regarding the study area are found in Bhat et al., Ecological indicators in forested

watersheds in Fort Benning, GA: Relationship between land use and stream water

quality, submitted to Ecological Indicators; hereinafter referred to as Bhat et al.,

submitted manuscript).

According to variations in ecologically relevant hydrological characteristics that are

based on flow variability and predictability, and low- and high-flow extremes, the









Table 3-2. Physical characteristics of study watersheds. Acronyms BON-1, BON-2,
BON, LPK, and SAL represent the streams (or watersheds) Bonham-1,
Bonham-2, Bonham, Little Pine Knot, and Sally Branch, respectively.
Physical Characteristics BON-1 BON-2 BON LPK SAL
Topography
Area, km2 0.76 2.21 12.73 18.01 25.31
Average Elevation, m 121.8 133.5 125.5 146.3 136.8
Average Slope, degree 5.46 4.89 5.04 5.32 5.42
Vegetation
Pine, % 28 30 40 41 48
Deciduous, % 27 6 8 2 12
Mixed, % 39 50 22 34 15
Wetland, % 6 8 9 17 10
Military Training Land, % 0 6 5 2 2
NDVI 0.36 0.30 0.32 0.34 0.36
Soil
Sandy loam, % 78 69 69 72 49
Loamy sand, % 9 9 31 28 51
Road
Road Length, km 3.6 11.4 51.6 56.6 97.6
Road Density, km/km2 4.8 5.1 4.1 3.1 3.8
Stream
Stream Length, km 2.6 3.9 29.1 43.3 65.2
Stream Density, km/km2 3.4 1.7 2.3 2.4 2.6
Stream Order 2 2 4 4 4
Other
No. of Roads Crossing Streams 1 2 13 11 21
Bare Land, % 1 11 11 4 7
Disturbance Index, % 11 21 19 11 15

streams in the study watersheds are classified as 'perennial runoff (Poff, 1996). As the

present study focuses on the impacts of military training within an ecosystem as well as

the effects of these impacts on the stream biota, a reference watershed was identified to

contrast with watersheds impacted by military training activities. Approximately 94% of

the area of Bonham-1 watershed is covered by forest. Mechanized military activities are









not conducted in this watershed as compared to 2-6% of the total area of other watersheds

used for the same purpose. Also, the watershed has a small percentage of bare land,

limited roads, and only one road crossing the stream. Hence, Bonham-1 was used as a

reference watershed for this study.

Precipitation and Streamflow Data

Streamflow and precipitation were measured from January 2000 to December

2003. Precipitation was measured by twelve tipping bucket rain gauges distributed

throughout the study area. Watershed precipitation was determined by areal weighting

using the Thiessen polygon method. Daily discharge values for Bonham-1 and Bonham-

2 were calculated from ten-minute continuous stage records using rating curves. Stream

stage and velocity were measured half-hourly for Bonham, Little Pine Knot, and Sally

Branch. These data were used to calculate daily discharges using the area-velocity

method.

Methods

Annual-Based Hydrologic Indices

The annual-based approach uses multi-year streamflow records to define a series of

ecologically relevant hydrologic indices. These indices may be used to characterize

intra-annual variation in water conditions, analyze temporal variations, and compare

impacts of alteration among watersheds. For this assessment, nonredundant yet

biologically significant hydrologic indices are adapted as per the recommendations of

Olden and Poff (2003). Table 3-3 lists these indices for the perennial runoff type of






46


Table3-3. Summary of the recommended hydrologic indices for 'perennial runoff type of
streams (after Olden and Poff, 2003) used in the Annual-Based Hydrologic
Indices. BON-1, and BON-2 represent Bonham-1 and Bonham-2,
respectively. BON-1 is reference (REF) and BON-2 is (IMP) watershed. To
maintain the consistency to the past studies, the symbols presenting the
hydrologic indices in this table are kept as the same as those used in Olden
and Poff (2003). Data types A, M, and D stand for annual, monthly, and daily
discharge data, respectively. For definitions and method of calculation of the
indices, refer to Appendix A.


Flow components and name
of the hydrologic index
Magnitude of flow events
Averageflow conditions
Variability in December flow
Mean annual runoff


Symbol Units


MA26 unitless


BON-1 BON-2 Data
(REF) (IMP) type


0.70 0.47


MA41 m3/s/km2 0.00397


0.00504


Spreads in daily flows
Low flow conditions
Baseflow index


Mean of annual minimum flows
Median of annual minimum flows
High flow conditions


MAlO 1/m3/s


81.4 43.5


ML17 unitless 0.12 0.57
ML14 unitless 0.17 0.58
ML16 unitless 0.17 0.79


High flow volume


Mean maximum flow in May
Median of annual maximum flows
Frequency of flow events
Low flow conditions
Frequency of low flow spells
Variability in low flood pulse count
High flow conditions
High flood pulse count
Three times median flow frequency
Seven times median flow frequency


days/no.
MH23 of years

MH8 m3/s


0.016 0.017


MH14 unitless 63.4


FL3 year 1 0.75
FL2 unitless 0.31


FH4 year
FH6 year


FH7 year 6


0
0.26


10.25 2.50


0 A









Table 3-3. Summary of the recommended hydrologic indices for 'perennial runoff type
of streams (after Olden and Poff, 2003) used in the Annual-Based Hydrologic
Indices (Continued).


Flow components and name
of the hydrologic index
Duration of flow events
Low flow conditions
Variability in annual minima of
90-day means of daily discharge
Variability in low flow pulse duration
Variability in annual minima of
1-day means of daily discharge
High flow conditions
Means of 30-day maxima of daily discharge
Variability in high flow pulse duration
Flood free days
Timing of flow events
Average flow conditions
Constancy
Seasonal predictability of flooding
High flow conditions
Seasonal predictability of non-flooding
Rate of change in flow events
Average flow conditions
Variability in reversals (Positive)
Variability in reversals (Negative)
Change of flow (Decreasing)

Change of flow (Increasing)


BON-1
Symbol Units (REF)


DL10
DL17


unitless 0.78
unitless 0.09


BON-2
(IMP)


Data
type


0.32 D/M/A


0.07


DL6 unitless 0.60 0.28 D/M/A


DH13
DH16
DH24


unitless
unitless
days


TAl unitless
TA3 unitless


5.59
0.006
56.0


0.28
0.70


1.84
0.005
61.5


D/M/A
A
A


0.44
0.84


TH3 unitless 0.40 0.41


RA9
RA9
RA7

RA6


unitless
unitless
m3/s

m3/s


0.04
0.06


0.07
0.12


0.1136 0.0324

0.1119 0.0496


streams found in the study region. The definitions and the methods used to calculate

these indices are listed in Appendix A. In this study, the annual-based indices are used to


compare and contrast the Bonham-1 and Bonham-2 watersheds.









Storm-Based Hydrologic Indices

The storm-based approach uses the storm hydrographs to determine the indices.

Hydrograph separation was used to identify distinct storm events. 44-100 storm events

from 2001 to 2003 were used to calculate the response factor, baseflow index,

dimensionless indices (Trl/Tic, Tr/Tic, and Tb/Tic), watershed area (A) scaled peak

discharge (qpk/A), bankfull discharges, and the rate of change of peak discharges in rising

and falling limbs for five watersheds, where Bonham-1 is the reference watershed (Table

3-4). Appendix B lists the indices, definitions, and the methods of calculation.

Statistical Analyses

The ANOVA and the Tukey's multiple comparison tests were performed to

determine the differences between reference watershed and the impacted watersheds'

mean values of indices. The percent of forest extent, military training land, road density,

the number of roads crossing streams, bare land fraction, and disturbance index were

considered to assess the effects of disturbance on hydrology and the potential impact on

stream ecology. Pearson's correlation coefficients were calculated to examine the

strength and significance of the relationships between a watershed physical characteristic

and storm-based hydrologic index. Stepwise multiple linear regressions were performed

to identify relationships between an index and management related watershed

characteristics. Only variables having p-values less than or equal to 0.05 were retained in

the regression models.

Results

Watershed Disturbance Characteristics

The watersheds' physical characteristics are summarized in Table 3-2. Most of the

watersheds are highly vegetated (70% or more) with the majority characterized by pine









Table 3-4. Summary of hydrologic indices used in the Storm-Based Hydrologic Indices.
BON-1, BON-2, BON, LPK, and SAL represent the streams (or watersheds)
Bonham-1, Bonham-2, Bonham, Little Pine Knot, and Sally Branch,
respectively. BON-1 is reference (REF) and other watersheds are impacted
(IMP). represents the mean values of the indices of the impacted watersheds
that are different at significance level of 0.05 from the mean values of
reference watershed as confirmed by Tukey's multiple comparison test. For
definitions and method of calculation of the indices, refer to Appendix A.


Flow components and name
of the hydrologic index
Magnitude
Mean response factor
Variability in response factor
Mean baseflow index
Variability in baseflow index
Mean peak discharge
Variability in peak discharge
Frequency
Bankfull discharge
Two times bankfull discharge
Duration
Duration of time base
Variability in time base
Duration of response lag
Variability in response lag
Duration of time of rise
Variability in time of rise
Rate of change

Mean slope of rising limb
Variability in rising slopes

Slope of falling limb
Variability in falling slopes


REF


IMP


Symbol Units BON-1 BON-2 BON LPK SAL


MMRF
MyRF
MMBF
MvBF

MMPD
MvPD


FIFD
F2FD


DMTB
DvTB
DMRL
DvRL
DMTR
DvTR


RpPD
RpvPD

RNPD


unitless
unitless
unitless
unitless

m3/s/km2
unitless


0.030
1.09
0.73
0.19

0.077
2.0


0.010
1.45
0.86*
0.11

0.025*
2.3


% 90
% 82


unitless
unitless
unitless
unitless
unitless
unitless


m3/s/hr
unitless

m3/s/hr


0.20
2.2

0.050


0.060*
1.18
0.95*
0.04

0.030*
1.2


2
0


1.2*
0.6
1.1*
0.6
0.5
1.0


0.06*
2.4

0.020


0.10
1.4

0.070


0.001*
0.89
0.90*
0.07
0.001*
1.0


29
10


1.3*
0.5
0.8
0.5
0.5
0.9


0.01*
1.4

0.001*


0.020
1.30
0.95*
0.04
0.014*
1.4


6
0


1.2*
0.5
1.3*
0.4
0.5
0.7


0.04*
1.6

0.020


RNVPD unitless 2.4 2.0 1.5 1.2 1.2









and mixed pine and hardwoods. Deciduous forest typically covers only a small

percentage of these watersheds. However, Bonham-1 consists of 27% of deciduous

forest. The study watersheds range from less than 1 to 25 km2. Average elevations and

maximum slopes are relatively constant. Sandy and loamy soils are common in most of

the study watersheds. The military training land (0 to 6%) is relatively small, but varies

among watersheds. Total bare lands in these watersheds comprise 11 to 21% of the

watershed area, of which 1 to 8% of the total area is unpaved roads and trails. This extent

and variability of military training and bare land are typical of the entire Fort Benning

installation. While road density is relatively comparable among watersheds, the number

of roads crossing streams varies from 1 to 21. Military training land, bare land, and

disturbance index are positively correlated at significance level of 0.05 or lower.

Annual-Based Hydrologic Indices

Table 3-3 summarizes annual-based analysis results for Bonham-1 (reference) and

Bonham-2 (impacted) watersheds. The average flow conditions revealed that the mean

annual flow (MA41) is higher in the impacted watershed. However, the reference

watershed has higher flow variability for both annual (MAlO) and December month

(MA26) periods. The average flow event timing is more constant (TAl) and predictable

(TA3) in the impacted watershed.

The impacted watershed maintained a higher magnitude of minimum flows as

depicted by the baseflow index (ML17). This result is consistent across other low flow

indices (ML14 and ML16) and the findings that the impacted watershed had no low flow

spells (FL3) and lower variability of low flow pulse counts (FL2). Higher coefficient of

variation in annual minima of 90-day (DL10) and 1-day (DL6) means of daily discharge

for the reference watershed and similar variability in low flow pulse duration (DL17)









were found for both watersheds. While low flow values vary more for the reference

watershed, once the flow goes low, it stays low for same duration in both watersheds.

The reference watershed produces higher magnitude flow and thus maintains higher

median flow (MH14) during events. However, in the month of May, mean of the

maximum monthly flows (MH8) were similar for both the watersheds. During high flow

conditions, the reference watershed crosses a threshold of seven times the median annual

daily flow volume (MH23). In the impacted watershed, these floods never occurred. The

high flood pulse counts (FH4) and the frequency of floods (FH6 and FH7) in the reference

watershed are more than the impacted watershed. However, on a few occasions, the flow

crossed a lower threshold of 3 times median frequency of flood (FH6) in the impacted

watershed. The 30-day floods (DH13) went higher and stayed high (DH16) in the

reference watershed. Periods between floods (DH24) are similar for both watersheds, that

is, they have approximately two months of flood free days a year and of comparable

predictability (TH3).

Storm-Based Hydrologic Indices

The results of storm-based hydrologic assessment are summarized in Table 3-4.

Analysis of variance tests indicated that the mean values of storm-based indices except

the time of rise differ at the significance level of 0.05. For a number of indices, the

reference watershed exhibits distinct behavior as compared to the impacted watersheds.

Tukey's multiple comparison tests indicate that the indices other than time of rise and the

rate of change in falling limb in the reference watershed were significantly different from

the impacted watersheds. The reference watershed is characterized by a relatively low

baseflow index (MMBF) with significantly higher (MMPD) and more variable (MvPD)

peak discharge. During events, 90% of the total events produced greater than bankfull









discharge (F1FD) in the reference watershed indicating a highly connected system as

compared to 2-47% in impacted watersheds.

Storm flows consistently lasted longer and responded faster to rain events in the

reference watershed (DMTB and DvTB). Once the stream responded, the time of rise

(MMTR) was similar in all the watersheds. The reference watershed's combination of

fast response and high peak discharge results in a rapidly increasing rising limb (RpPD)

as compared to impacted watersheds.

Relationship between Military Land Management and Storm-Based Hydrologic
Indices

Correlation and regression analyses were performed to determine the relationship

among the watershed physical characteristics and the storm-based hydrologic indices.

Table 3-5 indicates that 7 key storm-based hydrologic indices are significantly related to

military land management. Increased military training land, bare land, and the

disturbance index will increase the time of rise as well as the variability in the time base.

Increasing the road density increases the variability in the time base and the rate of

change of rising limb. Increasing the number of roads crossing streams increases the

storm response lag, but decreases the time base. Results also show that an increase in the

number of roads crossing streams decreases the variability in the rate of change of falling

limb. No effects on hydrologic indices were identified for forestry management

practices.

Stepwise multiple correlations characterized the response of storm-based indices to

military impacts (Table 3-6). The greatest impact of land management is found with

statistically significant predictive models for indices of time base, response lag, and time









Table 3-5. Pearson correlation coefficients between watershed physical characteristics
and event based hydrologic indices. Characteristics are acronymed as Forest
(FOR), Military Training Land (MIL), Road Density (RDN), No. of Roads
Crossing Streams (NRC), % Bare Land (PBL), and Disturbance Index (DIN).
and ** indicates significance at or below 0.05, and 0.01 probability levels,
respectively.
Hydrologic Indices FOR MIL RDN NRC PBL DIN
MMRF 0.25 0.15 0.15 0.13 0.27 0.31
MvRF 0.40 0.59 0.71 -0.05 0.67 0.81
MMBF -0.62 0.44 -0.56 0.83 0.68 0.41
MyBF 0.63 -0.39 0.61 -0.85 -0.63 -0.35
MMPD 0.83 -0.37 0.64 -0.62 -0.47 -0.19
MvPD 0.66 0.21 0.96** -0.73 0.04 0.34
F1FD 0.57 -0.40 0.57 -0.85 -0.65 -0.39
F2FD 0.69 -0.43 0.62 -0.82 -0.64 -0.36
DMTB 0.63 -0.17 0.77 -0.91* -0.42 -0.12
DvTB -0.50 0.98** 0.14 0.08 0.96** 0.90*
DMRL -0.23 0.06 -0.54 0.97** 0.41 0.18
DvRL 0.21 0.59 0.84 -0.78 0.33 0.57
DMTR -0.39 0.99** 0.32 -0.17 0.89* 0.90*
DvTR -0.86 0.67 -0.45 0.35 0.68 0.47
RpPD 0.79 -0.34 0.59 -0.55 -0.42 -0.16
RpvPD 0.63 0.13 0.91* -0.78 -0.06 0.24
RNPD 0.42 0.10 0.35 -0.13 0.14 0.26
RNVPD 0.70 -0.07 0.85 -0.88* -0.29 0.03

of rise. Military training land, road density, and the number of road crossing streams


were the three management variables that impacted storm responses.









Table 3-6. Stepwise multiple regression models for event based hydrologic indices.
Military Training Land (MIL), Road Density (RDN), and No. of Roads
Crossing Streams (NRC), are the independent variables retained in the
regression analyses. and ** indicates significance at or below 0.05, and
0.01 probability levels, respectively.
Hydrologic indices Independent variables retained and R2 (adj)
regression equations
MMBF 0.729 + 1.91 MIL + 0.00968 NRC 0.94*
MvPD 1.21 + 0.666 RDN 0.88**
F1FD 0.929 6.97 MIL 0.039 NRC 0.94*
DMTB 1.8 0.0352 NRC 0.77*
DvTB 0.418 + 3.05 MIL 0.95**
DMRL 0.483 + 0.0393 NRC 0.92**
DMTR 0.454 + 0.837 MIL 0.97**
RNVPD 2.20 0.0563 NRC 0.70*

Discussion

Results representing annual-based average flow conditions showed higher

magnitudes of MA26, MA41, and MA10 in the reference watershed as compared to the

impacted one. In addition, these flows are less constant and less predictable over the

years as compared to the impacted watershed. Aquatic communities can show distinct

differences to changes in velocity and reduction in bed gradient, and associated fining of

bed sediments (Clausen and Biggs, 2000). Periphyton and benthic invertebrates are

particularly sensitive to different velocities and bed sediment size/stability (Minshall,

1984; Biggs, 1996). Clausen and Biggs (1997) found that invertebrate species richness

and periphyton biomass changed based on flow. These changes are likely related to

riparian vegetation on the amount of leaf litter input (Vannote et al. 1980; Davies-Colley

and Quinn, 1998).









The higher magnitude of the annual-based low flow indices (ML17, ML14, and

ML16) in the impacted watershed is attributed to higher groundwater input as compared

to the reference watershed. This increased input likely reflects the reduction of

interception characteristic of the military land uses (Bryant et al., in press). The relative

magnitude of low flows is likely to have important influences on biota through the

intensity of habitat destruction associated with drying during low flows. In this study, the

lower magnitudes of low flows as depicted by lower baseflow index (ML17) and smaller

low flow pulses (FL3) suggest that the long periods of low flow condition are more likely

in the reference watershed. Higher variability in the duration during low flow condition

(DL10, DL17, and DL6) in the reference watershed supports likelihood of such long

periods of low flow conditions. Long periods of low flow conditions and higher

variability in these conditions may provide selective pressure for specific life history

characteristics such as invertibrate aestivation and egg diapause, and physiological

tolerance to low dissolved oxygen (Williams and Hynes, 1977).

The higher magnitude of the annual-based high flow indices (MH23 and MH14) and

peak discharges (MMPD and MvPD) in the reference watershed as compared to the

impacted watershed suggests the likelihood of habitat regeneration associated with

sediment transport and floodplain inundation during high flows. For high flow events,

the degree of riverbed communities' disturbance is strongly related to degree of bed

movement (e.g., Biggs et al., 1999). Dissolved inorganic nitrogen and phosphorus

concentrations in rivers are strongly negatively correlated with specific yield and high

flow magnitude among watersheds, and among years within watersheds (e.g., Biggs and

Close, 1989; Close and Davies-Colley, 1990; Grimm and Fisher, 1992). This reflects the









degree of flushing of the nutrients mineralized through organic matter breakdown in the

soil profile and leachate from the underlying substrata.

Floods or high flow conditions are also important in influencing community

structure. The results based on the annual flow in this study clearly show that the

reference watershed produced more frequent high flows (FH4, FH6, and FH7). Similarly,

the results from storm-based analysis show that the frequency of discharges equaling or

exceeding bankfull (FiFD and F2FD) is higher in the reference watershed. Floods are

widely viewed as reset mechanisms (Resh et al., 1988), and flood-related mortality to

lotic organisms can result either directly from scouring, crushing, or downstream export

of individuals (Minckley and Maffe, 1987) or indirectly from food resources loss

(Hanson and Waters, 1974). As flood frequency increases, some invertebrates in the

reference watershed actively migrate either into the substratum or to quieter backwaters

to avoid sudden floods. Floods have been shown to regulate community structure by

facilitating local coexistence between asymmetrically competitive algal species (Power

and Stewart, 1987) and invertibrate species (Hemphill and Cooper, 1983) and between an

exotic fish predator and its native, relatively flood-resistant prey (Meffe, 1984). The

result in this study showed the consistency of storm-based F1FD and F2FD with annual-

based FH4, FH6, and FH7 indices suggesting F1FD and F2FD may be used as alternative

indices.

As indicated by the higher annual-based mean duration of high flow condition

(DH13) and the duration of storms (DMTB) in the reference as compared to the impacted

watershed, it is apparent that water resides in the reference watershed for a longer period

of time during high flow condition. However, the DMRL is smaller for the reference









watershed suggesting a quick rise in hydrograph. This can be attributed to a better

connectivity of riparian areas to the stream. This connectivity is important as the nutrient

concentrations can strongly control autotrophic production during inter-flood periods

(e.g., Biggs, 2000). Also, individual high flow events greatly reduce the biomass and

change the species composition of periphyton (e.g., Biggs and Stokseth, 1996), and

invertebrates (e.g., Cobb et al., 1992).

Predictive relationships were identified for storm-based hydrologic indices based

on watershed scale military management and suggest that the key variables related to

hydrologic alteration are military training land, road density, and the number of roads

crossing streams. Military training within a watershed can modify annual and storm

generated runoff due to changes in drainage, vegetation, and soils. The impact is due to

troop maneuvers and large, tracked and wheeled vehicles that traverse thousands of

hectares in a single training exercise (Quist et al., 2003). These activities' impacts,

ranging from minor soil compaction and lodging of standing vegetation to severe

compaction and complete loss of vegetation cover in areas with concentrated training use

(Wilson, 1988; Milchunas et al., 1999), are evident in the stream hydrographs and

suggest impacts to stream ecosystems. Disruption in soil density and water content

(Helvey and Kochenderfer, 1990), addition of sediment, nutrients, and contaminants in

aquatic ecosystems (Gjessing et al., 1984), and impairment of natural habitat

development and woody debris dynamics in forested floodplain streams (Piegay and

Landon, 1997) are potential impacts on terrestrial and aquatic ecosystems. In a recent

study in the same watersheds in Fort Benning, Houser et al. (2004) found that gross

primary productivity, respiration, and the benthic organic matter were low in the stream









corresponding to the impacted watershed (higher disturbance index) as compared to the

reference one (lower disturbance index). Road crossings commonly found in military

training areas may act as barriers to the movement of fishes and other aquatic animals

(Fumiss et al., 1991). Roads also directly change the hydrology by intercepting shallow

groundwater flow paths, diverting the water along the roadway and routing it efficiently

to streams at crossings (Wemple et al, 1996). This can cause or contribute to changes in

timing and routing of runoff (Jones and Grant, 1996).

Conclusion and Recommendations

In the present study, a reference watershed was compared to impacted watersheds

using annual and storm-based hydrologic indices within the Fort Benning military

installation. The results suggest a subset of the hydrologic indices recommended by

Olden and Poff (2003) are necessary for the Fort Benning streams. It is recommended

that mean annual runoff and its spread, baseflow index, high flow volume, frequency of

low flow spells, high flood pulse count, variability in annual minima of 90-day means of

daily discharges, and constancy be used in stream ecology studies in the Fort Benning

streams to identify disturbances relative to a reference watershed at a watershed scale.

The indices based on storm data may be used to augment the annual indices or as

surrogate to those indices. Storm-based magnitude and variability in peak discharge,

baseflow index, and the bankfull discharge were consistent with the results from annual-

based analysis. With respect to the potential influence of the frequency and duration

aspects of the flow regimes on the stream ecology, duration and variability of time base,

and duration of response lag are identified as critical hydrologic indices. The military

management practices, military training land, road density, and the number of road

crossing streams, were found to significantly affect these indices.














CHAPTER 4
PREDICTION OF NITROGEN LEACHING FROM FRESHLY FALLEN LEAVES:
APPLICATION OF RIPARIAN ECOSYSTEM MANAGEMENT MODEL (REMM)

Introduction

Forest canopy litterfall initiates the major pathway for recycling of nutrients from

plant to soil (Bubb et al., 1998). Nutrients are returned to the soil via litterfall and

throughfall fluxes in forested ecosystem (Ukonmanaho and Starr, 2001). Litterfall,

comprised predominantly of leaf litter, is usually the most important nutrient source

(Herbohn and Congdon, 1998). Litter quantity, litter decomposition and nutrient release

patterns are factors for understanding nutrient cycling in forest ecosystems (Rogers,

2002). Litterfall is primarily caused by natural senescence of leaves. Forest litterfall

quantity and composition varies among tree species, stand age and development, and

reflects environmental conditions, particularly water and nutrient availability (Binkley,

1986; Polglase and Attiwill, 1992). The amount and distribution of litterfall through time

are also affected by season, rainfall amount and distribution, and wind speed (Crockford

and Richardson, 1998).

Fluxes of dissolved organic matter are an important vector for the movement of

nutrients both within and between ecosystems (Cleveland et al., 2004). In many forest

ecosystems, more than half of the nitrogen is soil solution is in organic form (Qualls et

al., 1991). Hydrology plays an important role in both production and mobilization of

dissolved organic matter in the forest floor (Park and Matzner, 2003). A substantial

amount of potentially soluble organic matter exists in an adsorbed phase (Christ and









David, 1996). The amount of percolating water mobilizes these soluble organic matters

(Tipping et al., 1999).

Past studies have shown that the dissolved organic matter flux increases when the

percolating water passes through the forest floor (McDowell and Likens, 1988; Qualls et

al., 1991; Currie et al., 1996; Michalzik et al., 2001). In a recent study in India, Singh et

al. (1999) reported that the rapid mass loss of leaf litter occurred during the rainy season

and the rate of litter mass loss was positively correlated with the cumulative rainfall. In

an early study, Bocock et al. (1960) observed a rapid leaching from oak and ash leaf

litter, and suggested that the weight loss during the first month was largely due to a

physical loss by leaching. The primary sources of dissolved nutrients are considered to

be leaching of substance from fresh litter and the product of decomposition of the plant

litters (Qualls et al., 1991). Water-soluble nutrients are more easily leached from the leaf

litter of deciduous species, including birch, than from coniferous species, including Scots

pine (Pinus sylvestris) and Norway spruce (Picea abies) (Harris and Safford, 1996;

Hongve, 1999).

Depending on the species, leaf litter may release from 5 to 30% of original dry

weight as dissolved organic matter within 24 hours (Cummins, 1974). In a woodland

stream ecosystem in Massachusetts, McDowell and Fisher (1976) found that leaves lose

weight rapidly by abiotic leaching of soluble constituents for 1-3 days and decay more

slowly thereafter as a result of microbial decomposition. Gosz et al. (1973) estimated

15% mass loss for yellow birch leaves (Betula lutea) due to leaching within first month

on the forest floor. Heath et al. (1966) reported a 26% weight loss from birch leaves (B.









alba) that remained for two months on the forest floor. Nykvist (1961) found an 8%

weight loss in B. verrucosa after 24 hours of leaching in distilled water.

Annual fluxes of nitrogen in litterfall are greater than other nutrients regardless of

forest type but the amounts vary among forest types (Ukonmaanaho and Starr, 1999).

Dissolved organic nitrogen is the major form of nitrogen in stream water draining from

many mature forest watersheds, comprising about 95% of the total nitrogen (Qualls et al.,

1991). Nitrogen content of a hardwood forest litterfall on the Hubbard Brook

experimental forest comprised approximately 37% of the total nutrients (Gosz et al.,

1972). In a low nutrient system at the Coweeta Hydrologic Laboratory in North Carolina,

where carbon (C) to nitrogen (N) ratios in freshly fallen litter varied from 100 to 220,

Qualls et al. (1991) reported that the output of dissolved organic nitrogen from the forest

floor was 28% of the nitrogen input from the litterfall.

The Riparian Ecosystem Management Model (REMM) was developed by the U.S.

Department of Agriculture's (USDA) Agricultural Research Service (ARS) to

characterize the role of the riparian area on stream water quality (Lowrance et al., 2000;

Altier et al., 2002). In REMM, the riparian areas' water quality functions for control of

N, phosphorus (P), and sediment transport into surface waters are simulated and

analyzed. This model has primarily been tested and applied on high-nutrient riparian

buffers adjacent to agricultural fields in the past (e.g., Inamdar et al., 1999a; Inamdar et

al., 1999b; Lowrance et al., 2000).

The complex dynamics related to the nutrients in a watershed is particularly of

interest in low nutrient systems as the release of nutrients from plant litters prior to

decomposition may be an important aspect of characterizing stream water quality. A









watershed scale study related to surface water quality was conducted in the Fort Benning

military installation located in southwest Georgia. Experimental data from Fort Benning

shows an increase in total Kjeldahl nitrogen concentrations following litterfall. The

release of nitrogen is relatively rapid as compared to that released from decomposition of

organic matter. It is hypothesized that after leaves fall, the first precipitation events can

be expected to result immediate nitrogen leaching and a corresponding increase in

nitrogen levels in the stream water. The goal of this study is to determine whether freshly

fallen leaves in a riparian area are a significant source of nitrogen in low nutrient system.

The specific objectives of this study are to quantify nitrogen in streamflow using REMM

and verify the modeled results with the observed data.

Basic Concepts of REMM

Originally, REMM was developed to simulate buffer systems that meet

specifications recommended by the USDA Forest Service and the USDA Natural

Resource Conservation Service as a national standard (Natural Resource Conservation

Service, 1995). Three zones parallel to the stream characterize the riparian system. Thus

REMM simulates a three-zone buffer. The riparian area's soil is characterized in three

layers. In these soil layers, vertical and lateral movement of water and associated

dissolved nutrients takes place. The top two soil layers correspond with the soil horizons

A and B, respectively. Depth of the root zone or a restricting soil horizon defines the

third soil layer. A litter layer that interacts with surface runoff covers the soil layer 1.

The model takes upland water output supplied by the user and calculates loadings of

water, nutrients, sediment, and carbon based on actual area of the zones of a riparian area

(Lowrance et al., 2000).









The model input data consist of the physical conditions of the riparian area,

external loadings of water and nutrients to the riparian area, soil, vegetation, and soil

information. Altier et al. (2002) documents a detailed description of the algorithms,

equations, and parameters used in the model.

Hydrology component of REMM

REMM measures water stores and fluxes using a daily water balance. The water

balance includes interception, evapotranspiration, infiltration, vertical drainage, surface

runoff, subsurface lateral flow, upward flux from water table, and deep seepage.

Interception losses occur in the vegetation canopy and litter layer. Canopy interception is

an exponential function of the canopy storage capacity and the amount of daily rainfall,

and is simulated using a modified form of the Thomas and Beasley (1986) equation.

Precipitation falling through the canopy (throughfall) is subject to litter interception.

Similar to canopy interception, litter interception is determined by the mass of the litter at

any given time and the litter storage capacity.

Infiltration at the soil surface depends on the total depth of water. This depth of

water is the sum of the throughfall depth and overland flow depth from the upslope zone

or field. If the sum of incoming upland runoff and throughfall depth is less than the

infiltration capacity of the riparian soil, runoff infiltrates at the rate of application.

During high intensity storms, the runoff and throughfall rate may exceed the infiltration

rate. Surface runoff from the riparian buffer occurs when total water depth exceeds the

infiltration depth. In such conditions, model simulates the infiltration using a modified

form of the Green-Ampt equation (Stone et al., 1994).

When soil moisture exceeds the field capacity, vertical drainage from a soil layer

occurs. This vertical drainage also depends on the available water storage capacity of the









receiving layer. The rate of drainage is controlled by the lesser of the vertical hydraulic

conductivities of the draining and receiving layer. Vertical unsaturated conductivity is

simulated as a function of the soil moisture content using Campbell's equation (Campbell,

1974). When a water table builds up over the restricting soil layer, subsurface lateral

movement occurs. Lateral water movement is simulated using Darcy's equation.

Saturation excess overland flow occurs when soil profile is completely saturated.

Nutrient component of REMM

Simulation of nutrients in REMM is based on the Century Model (Parton et al.,

1987). The C cycle is fundamental for simulation of all organic matter dynamics and

many nutrient cycling processes in REMM. Release and immobilization of N is in

proportion to transformations of C. Soil and litter C is divided into residue (woody

debris, leaf litter, and roots) and humus (soil organic matter) pools. Each C pool has an

associated mineralization rate, transformation efficiency, and a specific range of C to N

ratios (Inamdar et al., 1999b). A first order rate equation with respect to C determines the

mineralization of C from each pool. Litter from leaves, stems, branches, coarse roots,

and fine roots are allocated into a readily decomposable (metabolic) or a resistant

(structural) residue pool based on lignin to N ratio (Pastor and Post, 1986). Each pools

decomposes at different rates.

REMM assumes a fixed fraction of C in fresh plant residue. Dissolved C from

metabolic residue and active humus pools differ by litter and soil layer (Altier et al.,

2002). The sum of C in throughfall and incoming runoff is mixed with C on the ground

surface to determine dissolved and adsorbed concentrations. Dissolved C moves with

surface runoff, subsurface lateral flow, and vertical drainage (Lowrance et al., 2000).









The amount of dissolved C carried out of the litter or soil layer is a function of the water

volume and the dissolved C concentration.

Nitrogen is input to the riparian area through precipitation, surface and subsurface

water flow, and adsorbed to incoming sediment carried by surface flow. Presence of

nitrogen in different forms and the associated degree of physical and chemical

stabilization influences its availability for microbial transformations and plant uptake.

Effective C to N ratios influencing residue decomposition are calculated in the model as a

function of the content of N in the litter as well as the content of available inorganic N in

the soil. As decomposition of the litter takes place, C and N are released (Lowrance et

al., 2000).

Inorganic nitrogen, ammonium and nitrate, are both available for immobilization

into soil organic matter. Immobilization of nitrate occurs only after all available

ammonium has been used. Ammonium may be in solution or adsorbed to the soil matrix.

Nitrification is calculated with a first-order rate equation influenced by temperature, soil

moisture, and pH. Temperature and soil moisture are variables, recalculated each day but

pH is a constant for each zone and layer combination. With increasing amounts of

substrate, the rate of nitrification is lagged to represent a delayed microbial response.

The calculation of denitrification is a function of the interaction of factors representing

the anaerobic condition, temperature, nitrate, and mineralizable C. Denitrification mostly

occurs as water-filled pore space gets above 60% (Inamdar et al., 1999b). Nitrogen

transformation processes are shown in Appendix C.










Application of REMM

Study Area

The study area consisted of a second-order watershed, Bonham-South (221 ha), in

the Fort Benning military installation, Georgia (Figure 4-1). A detailed description of

this site along with the instrumentation and the type of data collected is provided in


Figure 4-1. Study area. Bonham-South is a second-order watershed within the Fort
Benning military installation in Georgia.

Bryant et al., in press and Bhat et al., in review. The riparian area covers approximately

2.5% of the total watershed area. The riparian length is 3,500 m. Based on the visual

inspection in the watershed, a 12 m wide riparian buffer along the stream is considered.


0 0.25 0.5 1
I iMC I Ih









This riparian buffer is divided into three equal zones (Zone 1, Zone 2, and Zone 3) of 4 m

each. Corresponding soil layers (Soil layer 1, Soil layer 2, and Soil layer 3) are

considered in each zone. A selective list of riparian area and hydrological parameters and

their corresponding values are provided in Table 4-1. The riparian area consists primarily

Table 4-1. REMM model parameters values by the riparian zone for the Bonham-South
watershed in Fort Benning, Georgia. The dimension of riparian area from the
upland to the stream perpendicular to the stream is referred to as width, and
the distance along the stream is referred to as length. The parameter values
are identical for Zone 1, Zone 2, and Zone 3 except for the slope. The
superscripts 1, 2, and 3 of the slope values represent the Zone 1, Zone 2, and
Zone 3, respectively.
Parameters Units Values
Riparian zone length m 3,500
Riparian zone width m 4
Slope % 3.01, 3.82, 4.23
Manning's n unitless 0.06
Total soil profile thickness m 3.3
Individual soil horizon thickness m
Soil layer 1 0.3
Soil layer 2 1.0
Soil layer 3 2.0
Saturated hydraulic conductivity cm/h
Soil layer 1 15
Soil layer 2 13
Soil layer 3 1
Porosity cm/cm
Soil layer 1 0.40
Soil layer 2 0.35
Soil layer 3 0.35
Wilting point cm/cm
Soil layer 1 0.07
Soil layer 2 0.07
Soil layer 3 0.07
Riparian area ha 4.2
Field drainage area (surface) ha 216.8
Field drainage area (subsurface) ha 221.0









of hardwoods (approximately 60%) that are mostly scrub oak (Q. rubra), sweetbay (M.

virginiana), sweet gum (L. styraciflua), water oak (Q. nigra), willow oak (Q. phellos),

red maple (A. rubrum), and swamp tupelo (N. aquatica. Mixed species including

southern scrub oak (Q. rubra) and yellow pine (P. ponderosa) cover approximately 21%

of the riparian area, whereas loblolly (P. taeda) and longleaf (P. palustris) pine cover

approximately 19%. The riparian area soils are predominantly sandy loam. The uplands

are mostly on loamy sand.

Data and Input Parameters

REMM operates at a hillslope scale. To apply the model at watershed scale, it is

necessary to convert the watershed into an equivalent hillslope by adjusting the geometry

of watershed. As the riparian buffer has specified length (parallel to the stream) and

width (perpendicular to the stream), upland contributing field width of the watershed was

adjusted to equate the total area of the watershed.

REMM requires soil and vegetation nutrient pools for the riparian area of the

watershed to initialize the model (Table 4-2). Required field data to run the model

includes daily surface and subsurface runoff from the uplands to the riparian area (Table

4-3). The model also requires daily weather data and parameter values representing

topographic, soil, and vegetative conditions within the riparian area. Strategic

Environmental Research and Development Plan's (SERDP) Ecosystem Management

Project (SEMP) continuously records the atmospheric data within the military

installation. Daily weather data were obtained from SEMP data repository that included

precipitation amount and duration, maximum and minimum air temperatures, solar

radiation, wind speed, and dew point temperature. Daily streamflow for Bonham-South









Table 4-2. Soil and vegetation nutrient pools for Bonham-South watershed's riparian
area. The litter and soil pools of nutrient consist of metabolic, structural,
active, slow, and passive humus pools. The values in litter layer as well as
individual soil layers in each zone are presented. Vegetation pools consist of
nutrients in leaves, stems, branches, coarse roots, fine roots, and heartwoods.
Total values of nutrients are reported in the vegetation pool. The parameter
values are identical for Zone 1, Zone 2, and Zone 3.
Parameters Depth Bulk density Carbon Nitrogen Biomass
(m) (g/cm3) (kg/ha) (kg/ha) (kg/ha)
Litter and Soil Litter 0.01 1.0 18,100 1,134
Soil layer 1 0.30 1.4 29,280 655
Soil layer 2 1.00 1.5 24,640 644
Soil layer 3 2.00 1.5 11,560 252
Total 83,580 2,685
Live vegetation 31,390 753 78,477


Table 4-3. Model simulated annual hydrologic budget for the riparian areas in Bonham-
South watershed. and ** represents the surface and subsurface flow input to
the zone 3 from uplands. Values are rounded off to nearest whole numbers.
Year
Units 2000 2001 2002 2003 4-yr average
Precipitation mm/yr 475 736 731 1,014 739


Surface flow in*
Subsurface flow in**
Observed stream flow (total)
Simulated surface flow
Simulated subsurface flow
Simulated stream flow (total)


mm/yr
mm/yr
mm/yr
mm/yr
mm/yr
mm/yr


253
258
6
244
250


was calculated from 10-minute continuous stage records using rating curves and area-

velocity method.

Typically, REMM uses daily surface runoff depth generated from the runoff

collected by surface runoff samplers at the upland-Zone 3 interface. REMM uses the

hydraulic gradient upland wells and Zone 3 wells, saturated thickness at Zone 1, and the

saturated hydraulic conductivity to compute subsurface flow loading to the buffer. Due









to the lack of measured surface and subsurface flow from the upland to the riparian area,

it is assumed that the baseflow fraction in the measured streamflow is a contribution of

upland subsurface flow to the riparian area. Therefore, measured streamflow from

January 2000 through December 2003 was partitioned into baseflow and surface runoff

using constant slope base flow separation technique (McCuen, 1998). Surface runoff

from the upland to the riparian area was not considered (Table 4-3). Parameters that

describe the riparian area dimensions, soil, and vegetation characteristics were derived

from measured data and previously published literature (Inamdar et al., 1999a, Inamdar et

al., 1999b, Lowrance et al., 2000; Garten Jr., et al., 2003). Important parameters related

to litter and soil layer are listed in Appendix D. Monthly canopy cover in the riparian

area was determined by direct measurement with a Model-A spherical densiometer using

the method outlined by Lemmon (1956) from June 2001 to September 2003.

Measured water and soil nutrient concentrations in the study area included the total

Kjeldahl nitrogen (TKN), nitrate, and ammonium in precipitation, streamflow, soil water,

and shallow groundwater. Two transects in the riparian area near the outlet of the

watershed were considered for groundwater and soil water monitoring. Groundwater

monitoring wells (1-3 m deep) and adjacent tension lysimeters at 20 and 60 cm depths

were positioned on both sides of the stream along the riparian transects. Stream water,

groundwater, and soil water samples were collected from October 2001 to September

2003; biweekly from October 2001 to November 2002 and monthly thereafter. Bonham-

South stream was sampled during 16 storm events between September 2002 and

September 2003 using an event triggered ISCO sampler that was programmed to collect









hourly samples based on the flow depth. Bhat et al. (unpublished manuscript), describes

the sample analysis procedures for different chemical constituents.

As REMM simulations are performed on a daily time step, and observed TKN

concentrations are of hourly time step, comparisons between the two were based on the

TKN masses produced during the events that lasted 24 hours or longer. The model

simulates the masses of dissolved and particulate organic nitrogen, dissolved ammonium

and nitrate in surface and subsurface flows on a daily basis. The mass of simulated TKN

is the sum of dissolved and particulate organic nitrogen and dissolved ammonium.

Carbon dissolves from metabolic residue and active humus pools. In REMM, incoming

C from precipitation, surface, and subsurface flow is assumed to be in dissolved form.

Carbon available to be dissolved from the active humus pool is a fraction of the total C

present in the same pool. This fraction of set at 0.31 based on an estimate by McGill et al.

(1981) for the proportion of dying bacteria in metabolic form. The total amount of

dissolved C is the sum of metabolic C residue and the dissolved C fraction in the active

humus pool. In REMM, stoichiometric relationships are assumed between C and N in the

organic matter. As C is transformed, corresponding N is also transformed. A Freundlich

isotherm method determines the amount of dissolved ammonium from inorganic pool of

N in REMM. Observed masses of TKN during storm events are calculated by

multiplying the concentrations by the volume of water.

Model Calibration

REMM's hydrology and nitrogen components were evaluated using data collected

in this study from riparian area as well as data from the literature. Data collected in

nearby watersheds in Fort Benning, Georgia were used to initialize the C and N in the

soil pools (Garten Jr. et al., 2003). The litter C and N pools are based on the literature









values (Silveira, M.L., Reddy, K.R. (Department of Soil and Water Science, University

of Florida), Comerford, N.B. (Department of Soil and Water Science, University of

Florida). Litter decomposition and soluble organic carbon and nitrogen release in a

forested ecosystem. Unpublished manuscript). Soil and vegetation nutrient pools for the

riparian area of the watershed are listed in Table 4-2. Bonham-South's soil layer

thickness, soil porosity, saturated hydraulic conductivity, clay content, carbon decay rate,

and denitrification rate were calibrated.

Model calibration involved the comparison of the simulated streamflow and TKN

output with the measured values. While calibrating hydrology and nutrient components

of the model, parameters such as soil and litter C and N pools, which are based on

literature values, were fixed. Other fixed parameters included riparian length and width,

surface and subsurface draining area. The remaining parameters including soil layer

thickness, saturated hydraulic conductivity, soil porosity, clay content, carbon decay rate,

and denitrification rate in each zone and soil layers of the riparian buffer were adjusted to

match the observed and predicted results.

Fox (1981) recommends mean biased error (MBE) and mean absolute error (MAE)

to measure the difference between observed and model predicted values. Mean biased

error is calculated as the average error between predicted and observed values

accumulated over the total number data points. Mean absolute error considers the

absolute values of the errors. In the present study, MBE and MAE are modified to

calculate the percent difference of modeled streamflow and TKN mass from the observed

values. Mean bias difference (MBD) calculates the average difference accumulated over

the total number of events, expressed as a percent of the observed value. Mean absolute









difference (MAD), expressed also as a percent, considers the absolute values of the

differences. Mean absolute difference takes negative values and replaces them with their

absolute values. Observed and modeled streamflow was also compared using the Nash-

Sutcliffe efficiency. The Nash-Sutcliffe efficiency criteria is based on the normalized

least square objective function that evaluates the sum of the squares of flow residuals

(Nash and Sutcliffe, 1970).

Sensitivity Analysis

A sensitivity analysis was performed to determine the effects of key hydrological,

soil, and vegetation parameters on streamflow and TKN fluxes. The parameters were

canopy cover fraction, riparian zone width, soil layer thickness, maximum carbon decay

rate for litter and humus, maximum denitrification rate, soil saturated hydraulic

conductivity, soil porosity, and soil clay content. Each parameter was changed by +10%

and -10% from the values used as the best estimates for the calibration simulations.

Results

Hydrology

Simulated daily flows for the study watershed closely matched the observed flows

(Figure 4-2). Over the 4-year simulation period, MBD and MAD between observed and

predicted streamflow were 2% and 12%, respectively. The model had Nash-Sutcliffe

efficiency of 80%. REMM tends to underestimate the streamflow during low flow and

overestimate during the storms. Approximately 1% of the streamflow was the

contribution from surface runoff generated in the riparian area. Majority of streamflow

output was through the subsurface (Table 4-2).










2.0- 0
1.8 -
1.650
1.6 ------- Simulated
S1.4 Observed
1.2ate
SPrecipitation
0 6


















correspond to simulated leaf fall (Figure 4-4). Modeled TKN concentrations were higher.




















during high leaf fall months. From October 2001 to December 2003, modeled leaf fall
o l 0.41
0.2
0 .0 .......



Date

Figure 4-2. Comparison of REMM simulated daily flow with the observed daily flow for
the Bonham-South watershed.

There was considerable variability between and within years for measured

precipitation and streamflow data during study period. It is evident from Figure 4-2 that

an increase in the average flow during the year 2003. This increase was consistent with

the variation in precipitation during the study period. Overall, model simulations for

streamflow can be considered good for the calibrated parameters.

Nitrogen

Observed stream TKN concentrations during the study period showed a strong

correspondence with the leaf fall (Figure 4-3). Approximately 50% of the canopy in the

riparian area dropped during November-January each year. Higher TKN concentrations

were observed during the same period of time. The simulated TKN concentrations also

correspond to simulated leaf fall (Figure 4-4). Modeled TKN concentrations were higher

during high leaf fall months. From October 2001 to December 2003, modeled leaf fall

totaled 12,200 kg/ha, which corresponds to 5,400 kg/ha/yr. As depicted in the figure,











100 1.2
0 1.0
80 a
70 -
)- 0.8
50 CanopyCover -0.6
S40 TN
30 \ 0
S20 0.2
10
0 . 0 .0




Date

Figure 4-3. Observed canopy cover and TKN concentrations each month for the
Bonham-South watershed.


3000

2500

2000

1500

S1000

500

0


r -"CC] C"-] C-] C C']""- C10 mmC C mm i C



Date


Figure 4-4. Monthly simulated TKN concentration and simulated leaf mass in the
Bonham-South's riparian area.

simulated leaf fall began as early as June each year and completed by the end of

December-January. As observed in the riparian area of the watershed, leaves were

present in the canopy year-round. For example, during the 2001-2002 dormant season,









approximately 45% of the canopy was still covered with leaves. Simulated N present in

the live vegetation showed an inverse relation with the simulated TKN concentrations in

the stream (Figure 4-5).


80 1.6
:70 -mE Simulated leafN 1.4
70 1.4
---- Simulated TIKN
5 0 1.0
.* -- : ,*
40 .. 0.8
30 -- 0.6!
S20 0.4
10 0.2
0 -----0.0
ooOooooooooooooooooooOOoooo
II I I I I I I I I I I I I I I I I


Date

Figure 4-5. Monthly simulated TKN concentration and simulated nitrogen present in live
leaves the Bonham-South's riparian canopy.

A careful analysis of model simulation results showed that a daily total

precipitation of 6 mm or more produced surface runoff in the riparian area. This surface

runoff is responsible for leaching and transporting TKN from the leaf litter in the riparian

zone to the stream. Due to the irregular nature of surface runoff generation, the TKN

variability during such events, the frequencies of observed data, and REMM's ability to

simulate on a daily basis, storm events lasting 24 hours or longer were selected for further

study. The masses of TKN were calculated for two different scenarios. First, total mass

of TKN during an event was calculated for the total streamflow. Second, baseflow was

separated from the total flow, and TKN mass corresponding to surface runoff was

determined. The observed and simulated TKN masses during individual storm events









were compared for both the scenarios. Out of the 16 storm samples collected in the study

area, 6 storms between October 2002 and May 2003 were suitable for the comparison

with the daily output from REMM (Figure 4-6). The simulated TKN masses produced in

surface runoff during the storms were comparable to the observed values (Figure 4-7).

MBD and MAD between the observed and the simulated TKN masses were 8% and 23%,

respectively. As the majority of the flow in the stream is subsurface flow, the effect of

subsurface flow in carrying the TKN mass was also analyzed. The result showed that

MBD and MAD between observed and simulated total TKN masses were 17% and 24%,

respectively (Figure 4-8).

Sensitivity Analysis

The changes in streamflow and TKN outputs for the -10% to +10% parameter

changes ranged from 0 to 1% and 0 to 7%, respectively (Table 4-4). Changes in soil

porosity did not have significant effects on the streamflow and TKN. A decrease in soil

porosity of 10% led to a total streamflow increase of only 0.5%, whereas the increase in

TKN was only 0.6%. Increase in clay content decreased the TKN output by 0.7%.

Increase in saturated hydraulic conductivity by 10% did not affect the streamflow but

increased TKN by 7.2%. Increasing soil layer thickness slightly reduced the streamflow,

but reduced TKN by 5.4%.

























momn m oom
Storm on 10/15102 begin ng 11
Storm on 10/15102 beginning 11 AM


I 0 0



00 0 0


0 0 0 0 00 0 0 0
000002000 0

Storm beginning 02/16/2003 6:30 AM


0 0



0-0

0 c


D 0i D C I D 0 0 00 0 a


Storm on 04/24/2003 beginning 8:30 PM


0

21

4

6
'*g S


0

2

4

6 .
CI
4J


0

2

4

6
C.
=


0000000000000
0000000000000
-Storm on 11-11002 beginning 7 AM
Stonn on 11/11/2002 beginning 7 AM


2

4

6 H
' 1


Storm egiing 0222/03 4:30 AM
Stonn beginning 02/22103 4:30 AM


0

2

4

6 .,
CI
Hi


0



4


ud


0000000000000000

CeMn] en 0 -en-en]- in- en-nM
Storm on 05/22/2003 beginning 9:30 PM


Figure 4-6. Observed TKN concentrations during the events. Vertical bars represent
precipitation, solid lines represent the streamflow, and the hollow circles
represent the TKN concentrations.


20

I15

S10

0.5


00


2.5

2.0-

a|1.5

S1.0
0 .
0.5
Sn


20


15


10


0 5


0 0


rZ


U.1
















- 0.4

0.3

0.2


C".] C"-] C\1 CA"] C" CA]
-' I ,-' CA] C1 ".]
O C1 C-O .

Storm Events


Figure 4-7. Comparison of the observed and simulated TKN masses during the events.
These masses represent the contribution from the surface runoff only.



0.8
0.7 E Observed
0.6 Simulated
0.5

C 0.4


0.2
0.1
0.0
ci c] enm m m en
o o o o o o
o o o o o o
C'i C-i C\1 C-i C\1 C-i

0< C\.1 CA] n

Storm Events


Figure 4-8. Comparison of the observed and simulated TKN masses during the events.
These masses represent the contribution from the total flow.










Table 4-4. Sensitivity of modeled streamflow and TKN based on +/- 10% change in
model parameters for Bonham-South watershed.


Parameters
Canopy cover (+)
Canopy cover (-)
Soil layer thickness (+)
Soil layer thickness (-)
Denitrification rate (+)
Denitrification rate (-)
Carbon decay rate (+)
Carbon decay rate (-)
Saturated hydraulic conductivity (+)
Saturated hydraulic conductivity (-)
Soil porosity (+)
Soil porosity (-)
Clay content (+)
Clay content (-)


Discussion

Over the four years, the simulated average annual surface runoff that is generated in

the riparian area and contributed to the streamflow was less than 1% of the annual

observed precipitation (Table 4-3). The simulated surface runoff generated in the

riparian buffer in the study watershed is small compared to earlier studies done in a

similar watershed in Georgia. A study conducted by Shirmohammadi et al. (1984) in

Little River watershed in Georgia reported approximately 4-12% contribution of annual

precipitation to surface runoff Inamdar et al. (1999a) reported the average annual

surface runoff contribution to streamflow was approximately 8% of the annual

precipitation for Gibbs Farm site within the Little River watershed. Surface runoff

contribution to streamflow in our study watershed may not be significant in terms of


Percentage change
Stream flow
-0.9
0.9
-0.1
0.1
0.0
0.0
0.0
0.0
0.0
0.0
-0.4
0.5
0.0
0.0


TKN
2.3
-0.2
-5.4
5.7
0.0
0.0
0.9
-0.5
7.2
-0.2
-1.5
0.6
-0.7
0.4









hydrology, but the modeled results suggest that it is capable of transporting nitrogen from

the litter and top soil layer to the stream (Figure 4-7). Subsurface contribution to the

streamflow for the Bonham-South watershed was approximately 20% of the average

annual precipitation (Table 4-3). The simulated subsurface contribution in this study is

close to the range of 14-22% reported by Shirmohammadi et al. (1984) for Little River

watershed.

The riparian forest community found along Bonham-South watershed is

representative of the Southeastern riparian forest (Shure and Gottaschalk, 1985;

Lowrance et al., 2000). Slash pine (P. elliotii), longleaf pine (P. palustris), black gum

(N. sylvatica), sweet gum (L. styraciflua), scrub oak (Q. rubra), sweetbay (M

virginiana), water oak (Q. nigra), willow oak (Q. phellos), red maple (A. rubrum), and

swamp tupelo (N. aquatic) are often dominant canopy species in these riparian forests.

The modeled total leaf fall in the study watershed from October 2001 to December 2003

was approximately 12,200 kg/ha, which corresponds to 5,400 kg/ha/yr. The modeled

average litter mass produced during the same period was approximately 7,000 kg/ha/yr.

Approximately 77% of the litter mass were leaves. Contribution of the modeled leaf fall

to the total litter production falls in the neighborhood of 80% reported by Meentemeyer et

al. (1982), and within the range of 72-84% reported by Shure and Gottaschalk (1985).

The simulated litter mass produced in the study area is higher than the average 5,000

kg/ha for a mixed hardwood forest in the Northeastern U.S. (Gosz et al., 1972) and is

comparable with the results found in the Southeastern U.S. Mulholland (1981) reported

litter production of 6,100 kg/ha in a small stream swamp in eastern North Carolina. A

bottomland hardwood forest and a cypress-tupelo stand located in Louisiana produced









litter masses of 5,750 kg/ha and 6,200 kg/ha, respectively (Conner and Day Jr., 1976).

Brinson et al. (1980) reported a value of 6,500 kg/ha for an alluvial swamp forest in

North Carolina.

Simulated results showed that the nitrogen content in the litter during the study

period averaged 77 kg/ha of which approximately 59 kg/ha was the contribution from the

leaves alone. This value is less than 83.4 kg/ha of nitrogen contribution from the

deciduous leaves in the Northeastern U.S. (Gosz et al., 1972). As observed in the riparian

area, 45% canopy was present even during the dormant seasons because of the presence

of approximately 19% of conifers and 21% of mixed species. Modeled results also

support the presence of the carbon in the leaves and the biomass in the canopy. The

modeled result showed that on an average approximately 527 kg/ha of carbon was

present in the canopy during the dormant season. This value of carbon suggests the

presence of approximately 22 kg/ha of nitrogen in the canopy during the same time

(Figure 4-5). Higher concentrations of TKN during the leaf fall and the inverse relation

with the nitrogen present in the canopy support the hypothesis of flushing of nitrogen

from the leaf litter.

The simulated TKN concentrations follow the trend of leaf mass accumulation in

the riparian area suggesting a higher rate of release of TKN from freshly fallen litter

during the fall and early spring. Brinson et al. (1980) reported higher nutrient fluxes

following litterfall in an alluvial swamp forest in North Carolina. In a study of nutrient

content of litterfall in the Southeastern U.S., Gosz et al. (1972) reported that

approximately 56% of the total nutrient was released during September through

December when the majority of litterfall (50%) occurred. A study conducted in cypress









swamp forest in Florida, Schlesinger (1978) reported that in the month of November

when 56% litterfall occurred, 45% of nitrogen was released from the system.

Higher TKN concentrations were observed in the Bonham-South stream during the

precipitation events (Figure 4-6). The storm events presented in Figure 4-6 were long

enough to calculate the masses of TKN during the storms to compare REMM's simulated

daily values. Comparisons of TKN masses produced during the events were comparable

to REMM's simulated values. The masses of TKN during the storms were calculated for

two different scenarios. When modeled TKN mass contribution of surface runoff that is

produced in the riparian zone was compared with the observed value in the stream, MBD

and MAD for the six storms were 8% and 23%, respectively. Mean biased difference

value is less reliable as there is the risk that large outliers cancel each other out. As MAD

takes negative values and replaces them with their absolute values, large outliers are de-

emphasized; hence it is less sensitive to extreme values. A MAD of 23% in the surface

runoff suggests that the model was able to produce comparable amount of TKN as

observed during the storm. The comparison between observed and simulated TKN mass

from the total flow showed comparable MBD (17%) and MAD (24%). On average, 43%

of the observed total TKN mass was contributed by surface runoff. This result suggests

that the surface runoff during the precipitation events is a major source of leaching

nitrogen from the forest floor and transporting it to the stream.

Conclusion and Recommendations

Originally, REMM was developed to operate at a hillslope scale. The results in this

study suggested that given appropriate upland inputs for a site, REMM can be used at

different scenarios of riparian area width, length of zones, vegetation type, and soil

properties. The trend and magnitude of the observed streamflow for the study watershed









was effectively simulated by REMM. The model, however, overestimated the

streamflow during high flow and underestimated during low flow periods. The

hydrologic budget showed a good agreement between observed and predicted

streamflow.

The simulated litter and corresponding leaf masses in the study watershed were

comparable to the values reported in the literature. Due to the model's longer simulation

time step, a meaningful comparison of simulated and observed TKN was possible only

through the masses produced during storms that were equal to or longer than a day.

Comparison of TKN masses during six different storm events showed similar values both

in the surface runoff and the total flow. The results supported the hypothesis of nitrogen

leaching from freshly fallen leaves during the precipitation events. These results

provided further insight into the nutrient dynamics of the riparian area. The model

simulations respond as expected to precipitation and vegetation patterns over the study

period. Results clearly indicated that the presence of fresh leaf litter in the riparian area

increases the TKN concentration, and hence mass, in the stream. The model effectively

captured the trends of leaf mass accumulation in the riparian area and subsequent high

concentration and mass during those periods. With the present version of REMM, the

comparison between observed and simulated values is possible only on a daily basis.

However, given the fact that the precipitation events and the sampling frequencies are

often shorter, simulation results using a smaller time scale would be useful. Therefore, it

is recommended to modify REMM from its present version of daily time step to a smaller

time step, preferably hourly, for more effective and meaningful interpretation and

evaluation of riparian nutrient flushing.














CHAPTER 5
SUMMARY AND CONCLUSION

The National Research Council of the United States has proposed that reliable and

comprehensive environmental indicators be developed to monitor ecological changes

from natural and anthropogenic causes. The military installation in Fort Benning,

Georgia offered a unique opportunity to study military impacts on water quality and

quantity. Ecohydrological approaches were used to relate the effects of anthropogenic

perturbations on water-vegetation-nutrient interactions in the Fort Benning watersheds.

In this research, statistical relationships among water quality parameters and the

watershed physical characteristics in low-nutrient watersheds were identified and

examined. This research also identified and developed hydrologic indices that

characterize the impact of military land management on watershed, and investigated the

riparian corridor's role on water quality by quantifying the nitrogen leaching from freshly

fallen leaves.

Relationships among watershed physical characteristics and water quality

parameters in the study watersheds and the regression analysis showed that pH, chloride,

total phosphorus, total Kjeldahl nitrogen, total organic carbon, and total suspended solids

are useful indicators of watershed physical characteristics that are susceptible to

perturbations. A comparison of the results between a reference and impacted watershed

in terms of hydrologic indices that are derived from long-term daily flow as well as the

storm-based data showed a clear distinction between the watersheds in terms of

hydrological flow regimes. Storm-based magnitude of baseflow index, magnitude and









variability of peak discharge, and the frequency of bankfull discharge were consistent

with the results from annual-based analysis. Results showed that these storm-based

indices might be used as surrogates to the indices derived from long-term data. The

analysis identified the relationships between the extent of military training land, road

density, and the number of roads crossing streams with the storm-based baseflow index,

bankfull discharge, response lag, and time of rise. For the low nutrient systems of this

study, seasonal and storm variations in water quality were found to be strongly influenced

by precipitation events which caused nitrogen leaching from recently fallen leaf litter and

increased nitrogen levels in the stream water.

This study showed that the signatures of military alterations to watershed landscape

are detectable in the water quality and the flow regime. The observed alterations to these

regimes suggest impacts to aquatic ecosystems. The water quality related indicators and

hydrologic indices presented in this research provide specific measures of stream water

quality and instream flow, respectively, that respond to military impacts. Each indicator

is linked to one or more readily measured watershed scale factors. The indicators could

be useful in predicting effects of military land management and evaluating restoration

activities on the quality as well as the quantity of the water.

This research has several implications for the Fort Benning military installation.

The results of this study indicated that baseflow sampling of water quality can be used to

assess the military training impacts on stream water quality. The water quality indicators

identified in this study provide measures of the current watershed conditions. The water

quality indicators identified in this study provide measures of the current watershed

conditions. Changes to stream water quality due to military training within the Fort









Benning military installation can be identified by collecting long-term and routine stream

water quality data and comparing indicator values to current conditions. Future activities

should develop indicator thresholds beyond which improved management and restoration

practices should be implemented. As nutrients are one major factor controlling the

quality of the receiving aquatic ecosystems, the thresholds should be based on the

ecosystems impacts.

The U.S. Army manages approximately 4.8 x 106 hectares of land for military

training, and it has developed the Land Condition Trend Analysis (LCTA) program to

systematically monitor terrestrial impacts from military training and to support the

mitigation and remediation of severely impacted training lands (Quist et al., 2003).

Although the magnitude of the training would differ from one installation to the other, the

nature of the impact remains similar. Military training typically increases soil bulk

densities and compaction, decreases infiltration, diminishes plant growth, degrades water

quality, and affects water quantity. Therefore, the research findings in this study can be

implemented for management and restoration practices to other military installations

across the U.S. to minimize the impacts of the military training. A common finding

between the water quality and quantity signatures is that military training impacts differ

from those found due to agriculture and urbanization. In the former, water quality

degradation impacts typically manifest as high nutrient loads. However, in military

installations, the loss of topsoil and vegetation results in waters that have significantly

lower nutrient levels. In a similar fashion, the hydrologic regimes in watersheds with

significant military impacts suffer from a loss of variability, but exhibit higher annual

discharge values.















APPENDIX A
ANNUAL-BASED INDICES DEFINITIONS AND CALCULATION PROCEDURES


Symbol


Definition


Method


MA26 Coefficient of variation
in monthly flows for
December


MA41 Mean annual flow
divided by watershed
area


Calculate mean monthly flow for December (q ), i
= 1,..., n;, n =no. of years
Calculate mean of all December flows,
- n
QD qd /n
1=1
Calculate standard deviation (SD) of qd

Calculate coefficient of variation, CV= SD/QD

Calculate mean yearly flow (q), i = 1,..., n =
no. ofyears
Divide q, by watershed area (A)


MAlO Ranges in daily flows
divided by median daily
flows (where range in
daily flows is the ratio of
20th/80th percentiles in
daily flows across all
years)


ML17 Seven-day minimum
flow divided by mean
annual daily flows


Calculate Q/A = qYl /n,i

of years

Combine daily flows of all years
Calculate 20th percentile (Q,20)
Calculate 80th percentile (Qps,)
Calculate median flow (Qm)
Calculate R= Q,,/Q,,o
Calculate R/ Qm


1,..., n;, n= no.


Calculate mean yearly flow (q ), i = 1,..., n;, n
no. ofyears
Calculate 7-day minimum ( q7m, ) flow
Calculate ratios R, = (q7mn)' / q
n
Calculate R,l /n
1=1