USF Libraries
USF Digital Collections

Hydrologic controls on salinity in mangroves and lagoons

MISSING IMAGE

Material Information

Title:
Hydrologic controls on salinity in mangroves and lagoons
Physical Description:
Book
Language:
English
Creator:
Stringer, Christina
Publisher:
University of South Florida
Place of Publication:
Tampa, Fla
Publication Date:

Subjects

Subjects / Keywords:
Wetlands
Hydrology
Water sources
Conductivity
Evapotranspiration
Electromagnetic survey
Stable isotopes
Dissolved ions
Mass-balance mixing
Florida
Mexico
Dissertations, Academic -- Geology -- Masters -- USF   ( lcsh )
Genre:
non-fiction   ( marcgt )

Notes

Abstract:
ABSTRACT: This dissertation explores the hydrologic controls on salinity within mangroves and lagoons at sites in Florida and Mexico. The main objective of this research is to better understand hydrologic controls on mangrove ecosystem structure and develop ideas that will be useful to land managers attempting to regulate and conserve these critical habitats. This study was conducted at sites in Ft. Pierce, FL and Costalegre on the central Pacific coast of Mexico. We examined controls on water levels and salinity in a mangrove on a carbonate barrier island along the Indian River Lagoon, east-central Florida. Spectral analysis of water levels showed that mangrove groundwater levels are not tidally influenced. Salinities vary spatially, with values of approximately 10 in upland environments to approximately75 psu in irregularly-flushed mangroves. Water chemistry indicates that water salinities are largely controlled by enrichment due to evapotranspiration. An electrical resistivity survey showed that the freshwater lens is restricted to uplands and that hypersaline waters extend deeply below the mangrove. These results indicate that evapotranspiration lowers water levels in the mangrove, which causes Indian River Lagoon water to flow into the mangrove where it evapoconcentrates and descends, forming a thick layer of high-salinity water below the mangrove. Spatial variability of terrain conductivity in the Ft. Pierce mangrove varied under two hydrologic management regimes, breached rotational impoundment management and rotational impoundment management. The difference in coefficient of variation (DeltaCV) between the breached RIM and RIM data was calculated to examine spatial variability in both the shallow and deep layers. A null-hypothesis model was employed to examine the statistical significance of the Delta CV results. The average water levels were -0.06 m amsl and 0.49 m amsl during the breached-RIM and RIM regimes, respectively. The average shallow (EM31) layer terrain conductivity shifted slightly from 1868 mS m-1 to 1825 mS m-1 after the alteration in management regime, yet the standard deviation of these averages decrease from 656 mS m-1 to 216 mS m-1. The average deep (EM34) layer terrain conductivities were 328 mS m-1 and 255 mS m-1 during the breached-RIM and RIM regimes, respectively. The temporal Delta CVs were 0.23 and -0.04 for the shallow and deep layers, respectively. The null-hypothesis model for the shallow layer illustrates that the difference in spatial structure is statistically significant. The deep layer DeltaCV was not statistically significant. These results indicate that the transition from breached RIM to RIM resulted in changes to both the physical and chemical hydrologic character of the impoundment, especially in the shallow layer. The second study sites were three mangrove communities along the central Pacific Mexican coast. Salinities varied by water type, with values of approximately 9 in La Manzanilla, 17 in La Vena, 33 in Barra de Navidad, 0.4 in the fresh waters, and 34 in the seawater. Sodium and Chloride concentrations and isotopic signatures, as well as salinity, were used as tracers in mass-balance mixing models to quantify estimates of relative fresh-water and seawater contributions to each site. La Manzanilla, a basin mangrove, had mean fresh-water contribution estimates of 63-84 percent. La Vena, a riverine mangrove, had fresh-water estimates of 39-51 percent. Barra de Navidad, a fringe mangrove, had low fresh-water contributions of 0-5 percent. These results illustrates that the role groundwater plays in mangrove hydrodynamics is dependent on the site hydrogeomorphology.
Thesis:
Dissertation (PHD)--University of South Florida, 2010.
Bibliography:
Includes bibliographical references.
System Details:
Mode of access: World Wide Web.
System Details:
System requirements: World Wide Web browser and PDF reader.
Statement of Responsibility:
by Christina Stringer.
General Note:
Title from PDF of title page.
General Note:
Document formatted into pages; contains X pages.

Record Information

Source Institution:
University of South Florida Library
Holding Location:
University of South Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
usfldc doi - E14-SFE0004760
usfldc handle - e14.4760
System ID:
SFS0028052:00001


This item is only available as the following downloads:


Full Text

PAGE 1

Hydrologic Controls on Salinity in Mangroves and Lagoons by Christina Elaine Stringer A dissertation submitted in partial fulfillment of the requirements for the degree of Doctor of Philosophy Department of Geology College of Arts and Sciences University of South Florida Major P rofessor: Mark C. Rains Ph.D. Thomas Crisman Ph.D. Sarah Kruse Ph.D. Mark Stewart Ph.D. H.L. Vacher Ph.D. Date of A pproval: November 10, 2010 Keywords: wetlands, hydrology, water sources, conductivity, evapotranspiration, electromagnetic survey, stable isotopes dissolved ions, mass balance mixing, Florida, Mexico Copyright 2010 Christina E laine Stringer

PAGE 2

D EDICATION This dissertation is dedicated to my family especially my sisters, who have supported and encouraged me every step of the way.

PAGE 3

ACKNOWLEDGMENTS Although this is ultimately my dissertation, this work could not have been possible without the contri butions of many people, especially my committee members. First and foremost is Mark Rains who, aside from guiding all of my ies. Sarah Kruse has been a wonderful mentor and a constant source of encouragement. Mark Stewart and H.L. Vacher have been enthusiastic supporters and greatly expanded my knowledge of geological concepts. Thomas Crisman had the graciousness to step in and fill a void late in the game. My family, friends, and fellow graduate students all helped to positively shape my doctoral experience. Many students and volunteers have aided in data collection over the years; this work simply would not have happened without the benefit of their time. Beth Fratesi has worked diligently to help me create, improve, and format the majority of my figures. During the course of my doctoral research I have received financial support from the United States Geological Survey The Smithsonian Institution, The National Science Foundation, The University of South Florida Graduate School, The Great Basin Institute, and the Patel Center for Global Solutions.

PAGE 4

i TABLE OF CONTENTS LIST OF TABLES ................................ ................................ ................................ iii LIST OF FIGURES ................................ ................................ ............................... iv ABSTRACT ................................ ................................ ................................ ......... vii CHAPTER 1: INTRODUCT ION ................................ ................................ ............ 1 References ................................ ................................ ................................ 4 CHAPTER 2: CONTROLS ON WATER LEVELS AND SALINITY IN A BARRIER ISLAND MANGR OVE, INDIAN RIVER LA GOON, FLORIDA ......... 7 Abstract ................................ ................................ ................................ ...... 7 Introduction ................................ ................................ ................................ 8 Study Site ................................ ................................ ................................ 10 Methods ................................ ................................ ................................ ... 13 Physical Hydrology ................................ ................................ ........ 13 Chemical Hydrology ................................ ................................ ...... 14 Evapoconcentration and Isotope Enrichment Modeling ................ 16 Resistivity Survey ................................ ................................ .......... 18 Results ................................ ................................ ................................ ..... 19 Physical Hydrology ................................ ................................ ........ 19 Chemical Hydrology ................................ ................................ ...... 20 Evapoconcentration and Isotope Enrichment Modeling ................ 26 Resistivity ................................ ................................ ...................... 28 Discussion ................................ ................................ ............................... 29 Acknowledgments ................................ ................................ .................... 32 References ................................ ................................ .............................. 33 CHAPTER 3: CHANGES I N THE SPATIAL VARIAB ILITY OF TERRAIN CONDUCTIVITY IN A MA NGROVE, INDIAN RIVER LAGOON, FLORIDA ................................ ................................ ................................ ....... 38 Abstract ................................ ................................ ................................ .... 38 Introduction ................................ ................................ .............................. 39 Study Area ................................ ................................ ............................... 42 Methods ................................ ................................ ................................ ... 45 Water Levels and Ground Water Conductivities ........................... 45 Electromagnetic Survey Measurements ................................ ........ 45

PAGE 5

ii Ground Water vs. Terrain Conductivity ................................ ......... 47 Null Hypothesis Modeling ................................ ............................. 47 Results ................................ ................................ ................................ ..... 48 Physical Hydrology ................................ ................................ ........ 48 Relationship between Ground Water and Terrain Conductivities ................................ ................................ .......... 49 Terrain Conductivity ................................ ................................ ...... 50 Temporal Variability in Terrain Conductivity ................................ .. 51 Discussion ................................ ................................ ............................... 54 References ................................ ................................ .............................. 57 CHAPTER 4: WATER SOU RCES IN MANGROVES IN THREE HYDROGEOMORPHIC SETT IN GS ON THE COSTALEGRE JALISCO, MEXICO ................................ ................................ ....................... 60 Abstract ................................ ................................ ................................ .... 60 Introduction ................................ ................................ .............................. 61 Study Site Description ................................ ................................ .............. 64 Methods ................................ ................................ ................................ ... 68 Results ................................ ................................ ................................ ..... 70 Salinity ................................ ................................ .......................... 70 Sodium and Chloride ................................ ................................ ..... 71 Stable Isotopes ................................ ................................ ............. 72 Mass Balance Mixing ................................ ................................ .... 73 Spatial Variability ................................ ................................ ........... 74 Discussion ................................ ................................ ............................... 75 Acknowledgments ................................ ................................ .................... 80 References ................................ ................................ .............................. 80

PAGE 6

iii LIST OF TABLES Table 2 1. Average cation and anion concentration (ppm) and isotopic investigation.. ................................ ................................ .................... 24 Table 4 1 Average cation and anion concentration (ppm) a nd isotopic investigation. ................................ ................................ ..................... 71

PAGE 7

iv LIST OF FIGURES Figure 2 1. Local setting showing piezometer and instrumentation locations, as well as resistivity transect position. ............................. 11 Figure 2 2. Equivalent fresh water head (m) of tide in the Indian R iver Lagoon and ground water level in each of the piezometers. ............ 19 Figure 2 3. Fourier transform results showing frequency spectra of equivalent fresh water head da ta for the tide and piezometer ground water levels. ................................ ................................ ........ 21 Figure 2 4. (A) Representative long term water level data from November 25 26, 2007 showing daily changes due to evapotranspiration. (B) Average daily cycle of entire equivalent fresh water head (m) data set for the tide and piezometers showing lack of tidal influence on piezometer water levels. ................................ .............. 22 Figure 2 5. Average ground water salinity values for each wet and dry sampling event in each of the five habitat classifications. ................ 23 Figure 2 6. Schoeller diagram showing the average cation and anion concentrations (meq/L) of each water type collected in both the wet and dry seasons. ................................ ................................ 24 Figure 2 7. D and 18 term records at Kennedy Space Center, located ~113 km north and in the same physiographic region as the study site, theoretical sea water, and surface water and ground water collected during this study. ................................ ................................ ......................... 26 Figure 2 8. (A) Evapoconcentration modeling results showing that sodium and chloride concentrations (ppm) in each of the community types can be achieved by evapora ting precipitation and/or Indian River Lagoon water. (B) Isotope enrichment results showing that D and 18 can be achieved through the Rayleigh distillation of precipitation and/or sea water. ................................ ......................... 27

PAGE 8

v Figure 2 9. (A) Profile showing resistivity inversion results. Horizontal axis is distance along the transect (m) from the coastline. (B) General interpretation of resistivity results sho wing corresponding large scale spatial patterns in conductivity. .............. 28 Figure 2 10. Conceptual model showing importance of evapotranspiration in controlling water lev els and salinities in the barrier island mangrove. ................................ ................................ ........................ 32 Figure 3 1. Local setting showing EM31 and EM34 sampling points, piezometers and instrumentation, as well as cross section location. ................................ ................................ ........................... 42 Figure 3 2. Geologic cross section characterizing the site subsurface on a west east transect. ................................ ................................ ........... 44 Figure 3 3. Water level in the mangrove (A) under breached RIM regime in water year 2007, and (B) under RIM regime in water year 2009. The management regime change occurred in March 2009. ................................ ................................ ................................ 48 Figure 3 4. Correlation between ground water conductivity and modeled terrain conductivity. ................................ ................................ .......... 49 Figure 3 5. Contoured surfaces of sha llow (EM31) terrain conductivities under (A) breached RIM in 2007 and (B) RIM in 2009. .................... 52 Figure 3 6. Contoured surfaces of deep (EM34) terrain conductivities unde r (A) breached RIM in 2007 and (B) RIM in 2009. .................... 53 Figure 3 7. Histograms showing null hypothesis modeling results for (A) shallow (EM31) CV randomized values and (B) deep (EM34) CV randomized values. ................................ ................................ .. 54 Figure 4 1. Site map showing general location of the Costalegre along the central Pacific coast of Mexico. ................................ ........................ 65 Figure 4 2. Average ground water and surface water salinity values for each wet and dry sampling event in each of the three mangroves and in the so urce waters. ................................ .............. 72 Figure 4 3. Sodium and chloride concentrations (ppm) of fresh waters, seawaters, and mangrove waters collected from each of the three study sites during the course of the study. .............................. 73

PAGE 9

vi Figure 4 4. D and 18 O ( ) of fresh waters, seawaters, and mangrove waters collected from each of the three study sites during the course of the study. ................................ ................................ ......... 74 Figure 4 5. Fresh water fraction of mangrove waters in each of the three study sites using each of the five tracers. ................................ ........ 76 Figure 4 6. Fresh water fraction of mangrove waters at each individual sampling location. ................................ ................................ ............ 77

PAGE 10

vii Hydrologic Controls on Salinity in Mangroves and Lagoons Christina Elaine Stringer ABSTRACT This dissertation explores the hydrologic controls on salinity within mangroves and lagoons at sites in Florida and Mexico. The ma in objective of this research is to better understand hydrologic controls on mangrove ecosystem structure and develop ideas that will be useful to land managers attempting to regulate and conserve these critical habitats. This study was conducted at sites in Ft. Pierce, FL and Costalegre on the central Pacific coast of Mexico. We examined controls on water levels and salinity in a mangrove on a carbonate barrier island along the Indian River Lagoon, east central Florida. Spectral analysis of water levels showed that mangrove groundwater levels are not tidally influenced. Salinit ies vary spatially, with values of ~10 in upland environments to ~75 psu in irregularly flushed mangroves. Water chemistry indicates that water salinities are largely controlled by enrichment due to evapotranspiration. An electrical resistivity survey show ed that the freshwater lens is restricted to uplands and that hypersaline waters extend deeply below the mangrove. These results indicate that evapotranspiration lowers water levels in the mangrove, which causes Indian River Lagoon water to flow into the m angrove where it evapoconcentrates and descends, forming a thick layer of high salinity water below the mangrove.

PAGE 11

viii Spatial variability of terrain conductivity in the Ft. Pierce mangrove varied under two hydrologic management regimes, breached rotational imp oundment management and rotational impoundment management. The difference in coefficient of variation ( CV) between the breached RIM and RIM data was calculated to examine spatial variability in both the shallow and deep layers. A null hypothesis model w as employed to examine the statistical significance of the CV results. The average water levels were 0.06 m amsl and 0.49 m amsl during the breached RIM and RIM regimes, respectively. The average shallow (EM31) layer terrain conductivity shifted slight ly from 1868 mS m 1 to 1825 mS m 1 after the alteration in management regime, yet the standard deviation of these averages decrease from 656 mS m 1 to 216 mS m 1 The average deep (EM34) layer terrain conductivities were 328 mS m 1 and 255 mS m 1 during t he breached RIM and RIM regimes, respectively. The temporal CVs were 0.23 and 0.04 for the shallow and deep layers, respectively. The null hypothesis model for the shallow layer illustrates that the difference in spatial structure is statistically sign ificant. The deep layer CV was not statistically significant. These results indicate that the transition from breached RIM to RIM resulted in changes to both the physical and chemical hydrologic character of the impoundment, especially in the shallow la yer. The second study sites were three mangrove communities along the central Pacific Mexican coast. Salinities varied by water type, with values of ~9 in La Manzanilla, ~17 in La Vena, ~33 in Barra de Navidad, ~0.4 in the fresh waters, and ~34 in the sea water. Sodium and Chloride concentrations and isotopic signatures, as well as salinity, were used as tracers in mass balance mixing models to quantify estimates of relative fresh water and seawater contributions to each site. La Manzanilla, a basin mangro ve, had mean fresh water contribution estimates of 63 84%. La Vena, a riverine mangrove, had fresh water estimates of 39 51%. Barra de Navidad, a fringe mangrove,

PAGE 12

ix had low fresh water contributions of 0 5%. These results illustrates that the role groundw ater plays in mangrove hydrodynamics is dependent on the site hydrogeomorphology.

PAGE 13

1 CHAPTER 1: INTRODUCTION Mangroves are an important coastal ecosystem that provides numerous ecological functions, as well as goods and services, such as storm and eros ion buffering and fish hatchery habitat. Mangroves cover ~180,000 km 2 of sheltered subtropical and tropical coasts between latitudes 24 north and south where mean annual temperatures are >20C (Lugo 1990; Dawes 1998, Alongi 2002). Worldwide, mangrove ec osystems are just one example of a coastal wetland that is being lost quickly, with approximately 35% disappearing since the 1980s at a rate of 2.1% per year globally (Alongi 2002, Martinuzzi et al. 2009) In the Americas, this rate is as high as 3.6% per year with an approximately 38% loss in coverage area (Valiela et al. 2001, Valiela et al. 2009). These high rates of loss make mangroves t he most threatened major coastal habitat in the world (Valiela et al. 2009). As in the case of other coastal habitats, anthropogenic activities are largely responsible for the loss and degradation of mangrove ecosystems (Alongi 2002; Gunawardena and Rowan 2005; Islam and Haque 2004; Vijay et al. 2005) Hydrology is a master variable in coastal mangroves, directly or indirectly controlling e cosystem structure and function. Plant community composition can be controlled by salinity and/or water levels (Odum and McIvor 1990), while primary productivity is largely controlled by salinity, nutrients, and/or water levels

PAGE 14

2 (Feller 1995; Chen and Twill ey 1999; Suarez and Medina 2005; Cardona Olarte et al. 2006; Lovelock et al. 2007). The hydrology of coastal wetlands is complex, with climatically and geologically controlled fluxes of fresh water interacting with oceanographically and geologically con trolled fluxes of sea water. Water levels and fluxes can be controlled by tidal variations, which can be propagated rapidly through surface water and ground water environments (Hughes et al. 1998; Ataie Ashtiani et al. 2001). However, water levels and flux es also can be controlled by episodic direct precipitation and associated surface water and ground water inflows during storms (Hughes et al. 1998) and/or steady ground water inflow between storms (Drexler and De Carlo 2002). In some cases, water levels an d fluxes also can be controlled by evapotranspiration, which can draw down water levels in the mangroves and draw water in from the surrounding surface water and ground water environments (Hughes et al. 1998). Salinity and specific solute concentrations in coastal wetlands vary primarily as functions of water sources, and additionally as functions of evaporation and water rock interaction. The salinity of direct precipitation is typically ~0.01 psu; the salinities of fresh surface water and ground water typ ically are ~0.03 0.60 psu; and the salinity of sea water is ~35 psu (Maidment 1993). Therefore, the degree of sea water intrusion strongly controls salinity. However, solutes remain in solution when water evaporates, so salinity also may be greatly influen ced by the rate at which water flows through the coastal

PAGE 15

3 wetlands and the rate at which water evaporates while in the coastal wetlands (Hughes et al. 1998; Twilley and Chen 1998) This dissertation explores the hydrologic controls on salinity within mangroves and lagoons at sites in Florida and Mexico. The main objective of this research is to better understand hydrologic controls on mangrove ecosystem structure and develop ideas t hat will be useful to land managers attempting to regulate and conserve these critical habitats. This dissertation is organized into three core content chapters, each of which is an independent manuscript prepared for journal publication. Chapter Two exa mines the role of evapotranspiration on the hydrology and salinity of a mangrove mosquito impoundment in Ft. Pierce, FL. This manuscript was published in Wetlands volume 30 in September 2010. Chapter Three utilizes electromagnetic measurements to examine the change in salinity spatial variability after an adjustment in management strategy at the Ft. Pierce mangrove mosquito impoundment. Before this manuscript is suitable for publication, another round of electromagnetic measurements is most likely necess ary. This field work is scheduled for November or December 2010. After this information is added to the paper, the manuscript will be submitted to Journal of Wetlands Ecology and Management Chapter Four looks at relative source water contributions in t hree different mangrove sites on the Central Pacific coast of Mexico, each with a different hydrogeomorphic character. This manuscript will be submitted to Estuarine, Coastal, and Shelf Science or Journal of Wetlands Ecology and Management

PAGE 16

4 References Alo ngi DM (2002) Present state and future of the world's mangrove forests. Environmental Conservation 29:331 349 Ataie Ashtiani B, Volker RE, Lockington DA (2001) Tidal effects on groundwater dynamics in unconfined aquifers. Hydrological Processes 15:655 669 Cardona Olarte P, Twilley RR, Krauss KW, Rivera Monroy V (2006) Responses of neotropical mangrove seedlings grown in monoculture and mixed culture under treatments of hydroperiod and salinity. Hydrobiologia 569:325 341 Chen R, Twilley RR (1999) Patterns of mangrove forest structure and soil nutrient dynamics along the Shark River Estuary, Florida. Estuaries 22: 955 970 Dawes CJ (1998) Marine Botany John Wiley and Sons, Inc., New York, New York, USA Drexler JZ, De Carlo EW (2002) Source water partitioning as a means of characterizing hydrological function in mangroves. Wetlands Ecology and Management 10:103 113 Feller IC (1995) Effects of nutrient enrichment on growth and herbivory of dwarf red mangrove ( Rhizophora mangle ). Ecological Monograph s 65:477 505 Gunawardena M, Rowan JS (2005) Economic valuation of a mangrove ecosystem threatened by shrimp aquaculture in Sri Lanka. Environmental Management 36:535 550

PAGE 17

5 Hughes CE, Binning P, Willgoose GR (1998) Characterisation of the hydrology of an estu arine wetland. Journal of Hydrology 211:34 49 Islam MS, Haque M (2004) The mangrove based coastal and nearshore fisheries of Bangladesh: ecology, exploitation and management. Reviews in Fish Biology and Fisheries 14:153 180 Lovelock CE, Feller IC, Ellis J, Schwarz AM, Hancock N, Nichols P, Sorrell B (2007) Mangrove growth in New Zealand estuaries: the role of nutrient enrichment at sites with contrasting rates of sedimentation. Oecologia 153:633 641 Lugo AE, Snedaker SC (1974) The ecology of mangroves. A nnual Review of Ecological Systems 5: 39 64 Maidment DR (ed) (1993) Handbook of Hydrology. McGraw Hill Inc., New York, NY, USA Martinuzzi S, Gould WA, Lugo AE, Medina E (2009) Conversion and Recovery of Puerto Rican Mangroves: 200 Years of Change. Forest E cology and Management 257:75 84 Odum WE, McIvor CC (1990) Mangroves. In R. L. Myers and J. J. Ewel (eds.), Ecosystems of Florida. University of Central Florida, Orlando, FL, USA, pp 517 548 Surez N, Medina E (2005) Salinity effect on plant growth and leaf demography of the mangrove, Avicennia germinans L. Trees 19:721 727

PAGE 18

6 Twilley R R, Chen R (1998) A water budget and hydrology model of a basin mangrove forest in Rookery Bay, Florida. Australian Journal of Freshwater and Marine Research 49:309 323 Valiela I Threatened Major Tropical Environments. BioScience 51:807 815 Valiela I, Kinney E, Culbertson J, Peacock E, Smith S (2009) Global Losses of Mangroves and Salt Marshes in: Gloabal Loss of Coast al Habitats: Rates, Causes, and Consequences. Carlos M. Duarte (ed.) Fundacion BBVA. Bilbao, Spain. Vijay V, Biradar RS, Inamdar AB, Deshmukhe G, Baji S, Pikle M (2005) Mangrove mapping and change detection around Mumbai (Bombay) using remotely sensed data Indian Journal of Marine Sciences 34:310 315

PAGE 19

7 CHAPTER 2: CONTROLS ON WATER LE VELS AND SALINITY IN A BARRIER ISLAND MANGROVE, INDIAN RIV ER LAGOON, FLORIDA Abstract We examined controls on water levels and salinity in a mangrove on a carbonate barrier island along the Indian River Lagoon, east central Florida. Piezometers were installed at 19 sites throughout the area. Groundwater was sampled at 17 of these sites seas onally for three years. Head measurements were taken at the other two sites at 15 minute intervals for one year. Water levels in the mangrove are almost always lower than lagoon water levels. Spectral analysis of water levels showed that mangrove groundwat er levels are not tidally influenced. Salinities vary spatially, with values of ~10 psu in uplands, ~30 psu in regularly flushed mangroves, and ~75 psu in irregularly flushed mangroves. Cation and anion concentrations and stable isotope compositions indica te that water salinities are largely controlled by enrichment due to evapotranspiration. A shore perpendicular electrical resistivity survey showed that the freshwater lens is restricted to uplands and that hypersaline waters extend deeply below the mangro ve. These results indicate that evapotranspiration lowers water levels in the mangrove, which causes Indian River Lagoon water to flow into the

PAGE 20

8 mangrove where it evapoconcentrates and descends, forming a thick layer of high salinity water below the mangrov e. I ntroduction Mangroves, which cover ~240,000 km 2 of sheltered subtropical and tropical coasts between latitudes 24 north and south where mean annual temperatures are >20C (Lugo 1990; Dawes 1998), provide numerous ecological functions as well as goods and services. Mangroves support estuarine and near shore marine productivity, in part by providing critical habitat for juvenile fish and through the export of nutrient rich water (McKee 1995; Rivera Monroy et al. 1998) or plant, algal, or animal biomass ( Zetina Rejn et al. 2003). Mangroves also protect coastal habitats against the destructive forces of hurricanes, typhoons, and tsunamis (e.g., Kandasamy and Narayanasamy 2005; Granek and Rutenberg 2007; Alongi 2008). Hydrology is a master variable in coas tal mangroves, directly or indirectly controlling ecosystem structure and function. Plant community composition can be controlled by salinity and/or water levels (Odum and McIvor 1990), while primary productivity is largely controlled by salinity, nutrient s, and/or water levels (Feller 1995; Chen and Twilley 1999; Suarez and Medina 2005; Cardona Olarte et al. 2006; Lovelock et al. 2007). Typically, primary productivity can be greatly enhanced by additions of nutrient rich water (McKee 1995; Rivera Monroy et al. 1998).

PAGE 21

9 The hydrology of coastal wetlands is complex, with climatically and geologically controlled fluxes of fresh water interacting with oceanographically and geologically controlled fluxes of sea water. Water levels and fluxes can be controlled b y tidal variations, which can be propagated rapidly through surface water and ground water environments (Hughes et al. 1998; Ataie Ashtiani et al. 2001). However, water levels and fluxes also can be controlled by episodic direct precipitation and associate d surface water and ground water inflows during storms (Hughes et al. 1998) and/or steady ground water inflow between storms (Drexler and De Carlo 2002). In some cases, water levels and fluxes also can be controlled by evapotranspiration, which can draw do wn water levels in the mangroves and draw water in from the surrounding surface water and ground water environments (Hughes et al. 1998). In many coastal mangroves, natural hydrologic patterns have been modified for a variety of reasons. On the east coast of Florida, for example, most of the mangroves have been impounded for purposes of mosquito control or the creation of habitat for migratory and resident waterfowl (e.g., Montague et al. 1987). Salinity and specific solute concentrations in coastal wetland s vary primarily as functions of water sources, and additionally as functions of evaporation and water rock interaction. The salinity of direct precipitation is typically ~0.01 psu; the salinities of fresh surface water and ground water typically are ~0.03 0.60 psu; and the salinity of sea water is ~35 psu (Maidment 1993). Therefore, the degree of sea water intrusion strongly controls salinity. However, solutes remain in solution when water evaporates, so salinity also may

PAGE 22

10 be greatly influenced by the rate at which water flows through the coastal wetlands and the rate at which water evaporates while in the coastal wetlands (Hughes et al. 1998; Twilley and Chen 1998). This study is part of a broader series of investigations of the controls on species composit ion, primary productivity, and nutrient cycling in mangroves on a carbonate barrier island on the east central coast of Florida. Our investigation is specifically focused on identifying and quantifying the controls on water levels and salinity, because the re is little fresh water on the barrier island, infrequent tidal inundation over much of the mangrove, and greater than sea water salinities in much of the mangrove. We hypothesized that evapotranspiration plays a major role in controlling water levels and salinities, by lowering the water table in the mangrove and creating a hydraulic gradient that draws surface water and ground water from the lagoon into the mangrove. Study Site The study was conducted in SLC 24, a mosquito control impoundment located at N27 o 33', W80 o 33' on the west shore of North Hutchinson Island This barrier island is ~35 km in length and ~0.2 2 km in width, and is part of the system of barrier islands that bound the Indian River Lagoon, a 250 km estuarine system located on the east ce ntral coast of Florida ( Figure 2 1 ) The study site is ~9 km north of the Ft. Pierce Inlet the northern most of five channels that connect the Atlantic Ocean and the Indian River Lagoon

PAGE 23

11 Figure 2 1 Local setting showing piezometer and instrumentation locations, as well as resistivity transect position. Points described in legend as habitat types represent the piezometer nests located within that habitat. The impoundment was developed in 1970 when a dike was constructed around an existing mangrove dominated wetland (Rey et al. 1990; Rey and Kain 1991). The impoundment was hydrologically isolated from the lagoon between 1970 and 1985. Tidal exchange was minimally restored in 1985 when a culvert was pl aced through the dike to remove water that had accumulated during two tropical storms. The culvert was later closed and the impoundment was again hydrologically isolated until 1987 when the culvert was reopened and four additional culverts were added. The climate is subtropical (mean annual, maximum, and minimum temperatures are approximately 23, 28, and 18 o C, respectively). Annual precipitation is approximately 1340 mm and it is distributed over a November May dry season and a shorter June October rainy s eason.

PAGE 24

12 North Hutchinson Island sediments vary in size and texture, with a high concentration of shell debris and a mean CaCO 3 concentration of 65% (Wang and Horwitz 2007). Vegetation cover in the impoundment decreased from 75% to near 30% between 1970 and 1985 but began to recover following the installation of the culverts in 1987 (Rey et al. 1990). Black mangrove ( Avicennia germinans (L.) L. ) is the dominant mangrove, but red ( Rhizophora mangle L ) and white ( Laguncularia racemosa (L.) C.F. Gaertn.) mangr ove also are common. Buttonwood ( Conocarpus erectus L. ) also occurs, as do understory plants such as Batis maritima L., Salicornia virginica L., Salicornia bigelovii Torr., and Borrichia frutescens (L.) DC. The abundance, density, and sizes of mangroves va ry across the impoundment from highly saline salt pans that have no mangroves or are fringed by dwarf black mangroves, areas that have dense and relatively short black mangroves, areas that have larger and more widely spaced black mangroves, to areas assoc iated with open water where red mangroves are most abundant. Adjacent to and up gradient from the mangrove is a woody upland dominated by Sabal palmetto (Walter) Lodd. Ex Schult. & Schult. F. and Schinus terebinthifolius Raddi (Brazilian Peppertree). There fore, for purposes of this study, f ive community types based o n species composition and structure were identified : (1) salt pan, (2) sparse black mangrove, (3) dense black mangrove (4) red mangrove and (5) woody upland.

PAGE 25

13 M ethods Physical Hydrology Precipitation was measured continuously with a tipping bucket rain gage (HOBO Event Rainfall Logger Onset Computer, Pocasset, Massachusetts, USA). Stage in the Indian River Lagoon was modeled on 15 minute intervals using XTide 2 (http://www.flaterco.com/x tide/index.html ), which was developed to predict tides and currents using an algorithm developed and used by the National Oceanic and Atmospheric Administration, National Ocean Service. The modeling program uses station specific harmonic constants provided by the National Ocean Service and based on local tide gauge data, resulting in predictions that are accurate to 1 minute and 0.03 m of measured high and low tides assuming no episodic storm related surge (e.g., surge related to expansion of water duri ng periods of extremely low pressure during the passage of extratropical storms). Piezometers were installed at 19 locations. At 17 locations, 2.5 cm inside diameter stainless steel piezometers were installed in nests. These piezometers were screened over 15 cm beginning at 0.6, 1.2, and 1.8 m below the ground surface, with the exception of two sites in the woody upland that were screened over 15 cm beginning at 1.2 and 1.8 m below the ground surface. These piezometers were used for ground water sampling. At two locations, 5.0 cm inside diameter PVC piezometers were installed. These piezometers were screened over 30 cm beginning at 2.0 m below the ground surface. Heads at these piezometers were measured on 15 minute intervals with pressure

PAGE 26

14 transducers and d ata loggers ( Model 3001 Levelogger Solinst, Georgetown, Ontario, Canada). The pressure transducers had a 1.5 m range and a 1 mm accuracy. Head data were compensated for changes in barometric pressure using barometric pressure data also measured on 15 min ute intervals with a pressure transducer and data logger (Barologger, Solinst, Georgetown, Ontario, Canada). Frequencies in the long term water level data from the piezometers and the Indian River Lagoon were determined using a fast Fourier Transform (FFT) algorithm in MATLAB (Version R2008A, Mathworks Inc., Natick, Massachusetts, USA). Prior to the FFT, a Tukey window function was applied to taper the data (Gubbins 2004). A Butterworth filter with a low frequency cutoff of 33 hours was applied to the data to remove any long term trends in the data. Mean daily head and stage cycles were also calculated using MATLAB. Daily evapotranspiration was calculated by quantifying daily fluctuations in the long term water level record following the procedure outlined in White (1932). Daily evapotranspiration was computed as: where S y is the specific yield of the sediments (10 1 in this setting), r is the rate of water table rise between 00:00 and 04:00, and s is the net change of the water level durin g the daily period (White 1932). Chemical Hydrology Water samples were collected and analyzed for temperature, pH, salinity, cations (e.g., Na + K + Mg 2+ Ca 2+ ), anions (Cl and SO 4 ), and stable isotopes

PAGE 27

15 (e.g., deuterium and oxygen 18) in the wet and dry seasons in 2005 and temperature, pH, and salinity in the wet and dry seasons in 2006 and 2007. Water samples were collected from the rain collector, the Indian River Lagoon, the primary perimeter ditch inside the impoundment, and piezometers in which water was found. The rain collector consisted of 1 L bottle contained within a plastic cylinder with a funnel top. A ping pong ball was placed in the funnel to limit evaporation. Rainfall sampling was opportunistic and immediately following rainfall, with only five samples collected throughout the study period. Approximately three volumes of water were pumped from each piezometer prior to the collection of samples for chemical analyses. S amples were pumped through 0.45 m in line fil ters (Whatman, Maidstone, Kent England) directly into pre cleaned, acid washed HDPE sample bottles Ani on samples were acidi fied with 1 ml of ni tric acid, and c ation and anion samples were stored at the method required range of 4 (2) C prior to analyses. Stable isotope sample bo ttles were filled completely with negligible head space and sealed with Parafilm (American National Can, Chicago, I llinois, USA ) to prevent the sample from equilibrating with ambient air Temperature, pH, and salinity of the surface water and ground water samples were measured in the field with a YSI 556 MPS (YSI Inc., Yellow Springs, O hio, USA ). Major cation and anion analyses w ere conducted at the University o f South Florida Center for W ater Analysis. Major cation con centrations were determined by induct ively coupled plasma emission s pectroscopy following the EPA 200 method (Clesceri et al. 1998), and major

PAGE 28

16 anion con centrations were determined by ion chromatography following the EPA 300 method (Clesceri et al. 1998). Analytical precision of the laboratory analyses were better than 1%. Stable isotope analyses were conducted at the UC Davis Department o f Geology Stable Isotope Labora tory. Deuterium analyses were performed using the chromium reduction method ( Donnelly et al. 1991), while oxygen 18 analyses were performed using the carbon dioxide equilibration technique (Epstein and Mayeda 1953). Deuterium and oxygen 18 are reported in the conventional, delta notation ( ) : where R 18 16 O for deu te rium and oxygen 18, respectively (Craig 1961). The resulting sample values of D and 18 O are reported in per mil ( ) deviation relative to Vienna Standard Mean Ocean Water (VSMOW) and, by convention the D and 18 O of VSMOW are set at 0 VSMOW (Gonfi ant ini 1978). Analytical preci sions were 1.0 and 0.05 for D and 18 O, respectively. Evapoconcentration and Isotope Enrichment Modeling Evapoconcentration is the process by which solute concentrations increase as water evaporates and solutes are retained in the remaining solution. An evapoconcentration model with Na + and Cl as conservative natural tracers was used to determine if evapoconcentration could explain solute concentrations within the five habitat types. The evapoconcentration model was run

PAGE 29

17 enco mpassing both dry and wet seasons, using seasonal mean Na + and Cl concentrations of the precipitation or Indian River Lagoon Water. The evapoconcentration model was: where C is the modeled Na + or Cl concentration in mg/L, f is t he fraction of water water (e.g., evaporated ground water in the mangrove) and initial water (e.g., precipitation or Indian River Lagoon Water), respectively. Isotope enrichme nt due to evaporation is the process by which D and 18 O increase as water evaporates and heavier isotopes are retained in the remaining solution. Isotope enrichment was modeled through the application of a Rayleigh distillation model (Clark and Fritz 199 7). The isotope enrichment model was: where R is the modeled isotope composition, R 0 is the initial isotope composition, f is the fraction of water remaining as it evaporates, and is the equilibrium fractionation factor for evaporation (Ma joube 1971). Kinetic effects due to high humidity were taken into account by applying a correction factor to the fractionation factor (Gonfiantini 1986). Two separate models were completed using theoretical precipitation and sea water as source waters. The precipitation source term was taken from the mean isotopic composition of precipitation at Kennedy Space Center, located ~113 km north and in the same physiographic

PAGE 30

18 region as the study site ( http://www.uaa.alaska.edu/enri/usnip/isotope_1989 2001.cfm ), whi le the sea water source term was taken from the isotopic composition for theoretical sea water (Gonfi antini 1978). Resistivity Survey A 400 m resistivity survey was conducted on an east west transect from the Indian River Lagoon to the dune crest of North Hutchinson Island ( Figure 2 1 ) The shoreward half of the survey was completed on September 28, 2006 and the upland half on March 3, 2007. Data were collected using an Advanced Geosciences Incorporated Marine SuperSting R8 multichannel system connected to an external switching box that controlled the flow of current along the 56 electrode cable (AGI, Austin, Texas, USA). Electrodes (metal spikes) were spaced at 1.5 m in tervals. Contact resistance was minimized as far as possible by driving the electrodes as deep as practical (typically ~ 20 30 cm deep) and watering the electrode ground contact with sea water. Three individual surveys were completed, with a 20% overlap be tween surveys. Resistivity surveys measure terrain resistivity, which is a weighted average of the deposit and fluid resistivities. Therefore, resistivity is not solely a function of salinity. However, the extreme contrast in resistivity between fresh wate r (high resistivity/low conductivity) and sea water (low resistivity/high conductivity) makes the freshwater/sea water interface a strong component of measured terrain resistivities in any setting. Following acquisition, the data were inverted using the Ea rthImager 2D inversion software from Advanced Geosciences Incorporated.

PAGE 31

19 R esults Physical Hydrology Raw water level data were converted to equivalent fresh water head data to allow direct comparison of heads in fluids with varying densities. With the except ion of a few excursions, heads in the mangrove habitats were lower than stages in the Indian River Lagoon ( Figure 2 2 ). Mean annual heads for the two continuously moni tored piezometers in the impoundment were 0.09 m and 0.02 m amsl, while mean annual stage of the Indian River Lagoon was 0.21 m amsl. Figure 2 2 Equivalent fresh water head (m) of tide in the Indian Rive r Lagoon and ground water level in each of the piezometers. Measurements were recorded on 15 minute intervals with a pressure transducer and data logger.

PAGE 32

20 Spectral analysis of the piezometer and tidal data shows that the time frequencies of patterns in head s in the mangrove and Indian River Lagoon differ ( Figure 2 3 ). Heads in the mangrove have strong spectral peaks at 24.0 hours and weak spectral peaks at 12.0 hours, while stage in the Indian River Lagoon has a strong spectral peak at 12.4 hours and weak spectral peaks at 24.0 and 25.8 hours. The direct effect of evapotranspiration on water levels can be seen in the examination of a segment of the long term water level record ( Figure 2 4 A). The water level data between November 25, 2007 and November 26, 2007 show that heads decline from the early morning to a minimum at midday, and then recover through the afternoon and night to a maximum in the early morning. These water level fluctuations correspond with daily evapotranspiration rates of 2 mm and 9 mm, which are within the range of evapotranspiration values (0 10 mm) reported for the Indian River Lagoon in the month of November from a previous study (Sumner and Belaineh 2005). The same trend in water levels is found throughout the long term record and can be seen in the mean daily head cycle ( Figure 2 4 B). Chemical Hydrology Ground water salinities varied by community type ( Figure 2 5 ). Mean SD ground water salinity was 78 9 psu in the salt pan (n = 48), 55 2 psu in the sparse black mangrove (n = 59), 44 5 psu in the dense black mangrove (n = 56), 32 4 psu in the red mangrove (n = 52), and 10 2 psu in the upland forest (n = 19). Ground water salinities varied significantly between community types

PAGE 33

21 (ANOVA, p < 0.01; all Tukey post hoc comparisons p<0.01) but showed no apparent seasonal or annual variations or tren ds throughout the course of the 2.5 year study. Figure 2 3 Fourier transform results showing frequency spectra of equivalent fresh water head data for the tide and piezometer ground water levels. The y axi s shows the relative amplitude of the peaks, with the most intense peak having a value of 1. The x axis is the frequency of the water level pattern (1/hours).

PAGE 34

22 Figure 2 4 ( A) Representative long term water level data from November 25 26, 2007 showing daily changes due to evapotranspiration. ( B) Average daily cycle of entire equivalent fresh water head (m) data set for the tide and piezometers showing lack of tidal influence on piezometer water levels. Y axis is in the number of hours after midnight.

PAGE 35

23 Figure 2 5 Average ground water salinity values for each wet and dry sampling event in each of the five habitat classifications. Bars are the average ground water salinity for all depths at all sites sampled in that habitat and are arranged in ascending date from left to There we re substantial differences in th e absolute concentrations of the cations and anions in precipitation, surface water in the primary perimeter ditch inside the impoundment and the Indian River Lagoon, and ground water in each of the five community types ( Table 2 1 ; Figure 2 6 ) In all but one case (Ca 2+ ), cation and anion concentrations were greatest in salt pan g round water, followed by the sparse black mangrove ground water, dense black mangrove ground water, red mangrove ground water, surface water in the primary perimeter ditch inside the impoundment and Indian River Lagoon, woody upland ground water, and preci pitation. Although the absolute concentrations varied, the relative

PAGE 36

24 Table 2 1 each water type sampled during the investigation. St andard deviations shown are 1 Isotopic averages for precipitation are long term records at Kennedy Space Center, located ~113 km north and in the same physiographic region as the study site. Water Type Precipita tion (n=5) Surface Water (n=6) Salt Pan ( n=18) Sparse Black (n=21) Dense Black (n=19) Red (n=12) Woody Upland (n=9) Na + (ppm) 5 4 8600 900 33480 60 23000 2000 16600 500 11200 500 2500 700 K + (ppm) 0.5 0.2 320 30 1050 30 750 90 558 7 410 10 112 2 Ca 2+ (ppm) 0.5 0.3 380 30 1020 10 1000 100 920 60 600 100 210 60 Mg 2+ (ppm) 1 1 1100 100 4220 70 3000 200 2260 60 1450 10 350 30 Cl (ppm) 10 10 15000 2000 64000 1000 42000 2000 31000 1000 20500 500 4900 600 SO 4 2 (ppm) 2 1 2300 500 7800 400 5600 400 4400 700 2700 300 700 20 20.8 9 1 8 2 7 3 5 4 6 4 15 5 d 18 3.85 0.9 0.2 1.4 0.5 0.8 0.6 0.3 0.8 0.3 0.7 3.3 0.6 Figure 2 6 Schoeller diagram showing the average cation and anion concentrations (meq/L) of each water type collected in both the wet and dry seasons.

PAGE 37

25 proportions of the cations and anions were relatively constant across precipitation, surface water in the primary perimeter ditch inside the impoundment and Indian River Lagoon, and ground water in each of the five community types ( Figure 2 6 ). Precipitation anion and cation conc entration averages were comparable to those reported by the National Atmospheric Deposition Program monitoring location at Kennedy Space Center ( http://nadp.sws.uiuc.edu/sites/siteinfo.asp?id=FL99&net=NTN ). The cation and anion averages reported by the m onitoring program all fall within the standard deviation of the averages of samples collected during this study. Isotope data indicate that most of the surface water and ground water was evaporated ( Figure 2 7 ). Ground water in the woody upland had a mean isotopic composition of for D and 18 O, respectively, which plotted on the Global Meteoric Water Line (Craig 1961) near the mean isotopic compositi on for regional precipitation ( http://www.uaa.alaska.edu/enri/usnip/isotope_1989 2001.cfm ). Surface water in the primary perimeter ditch inside the impoundment /Indian River Lagoon and ground water in the four remaining community types plotted along a local evaporative trend line that intersected the Global Meteoric Water Line near the mean isotopic composition for regional precipitation and the mean isotopic composition of sea water The slope of this local evaporative trend line was 5.0 which is indicative of evaporation in an environment with a relative humidity of ~70%, which is the approximate mean annual humidity of this region (Southeast

PAGE 38

26 Figure 2 7 D and 18 long term records at Kennedy Space Center, located ~113 km north and in the same physiographic region as the study site, theoretical sea water, and surface water and ground water collected during this study. The symbol shows the mean composition and the er ror bars represent the standard deviations defined by Craig (1961). The dashed line is the local evaporative trendline (y = 5.6x + 2.7; R 2 = 0.98), determined via linear regression of the sample isotopic compositions. Regional Climate Center data for Vero Beach, Florida). Salt pan ground water for D and 18 O, respectively. Evapoconcent ration and Isotope Enrichment Modeling The evapoconcentration model results show that Na + and Cl concentrations in mangrove ground water can be produced through the evaporation of precipitation and/or Indian River Lagoon water ( Figure 2 8 A). Mean Na + and Cl concentrations in the red mangrove, dense black mangrove, sparse black mangrove, and salt pan community types can be produced by

PAGE 39

27 evaporating >99% of the precipitation or 2 1%, 46%, 61%, and 74% of the Indian River Lagoon water, respectively. Similarly, the isotope enrichment model results also show that mangrove ground water can be produced through the evaporation of precipitation and/or Indian River Lagoon water ( Figure 2 8 B). Mean isotopic compositions in the red mangrove, dense black mangrove, sparse black mangrove, and salt pan community types can be produced by evaporating 23%, 23%, 25%, and 28% of the precipitation, respectively, or 3%, 3%, 6% and 7% of the Indian Riv er Lagoon water, respectively. Figure 2 8 ( A) Evapoconcentration modeling results showing that sodium and chloride c oncentrations (ppm) in each of the community types can be achieved by evaporating precipitation and/or Indian River Lagoon water. ( B) Isotope enrichment results showing that D and 18 h distillation of precipitation and/or sea water. In panel A, the trendlines are the calculated increases in those concentrations, assuming that sodium:chloride ratio remains constant, as water is evaporated. In both plots, the points represent the mean c omposition (sodium and chloride in Figure 8A and D and 18 O in Figure 8B) and the error bars represent the

PAGE 40

28 Resistivity The resistivity results were somewhat noisy, most likely due to poor contact between electrodes and hard or resistive soils, but larger scale spatial patterns were clear ( Figure 2 9 ). There was a discernable transition between the mangroves (i.e., the salt pan, sparse black mangrove, dense black mangrove, and red mangrove) and the woody upland at a distance of approximately 220 m along the profile ( Figure 2 9 ). Bel ow the mangrove, terrain resistivities were m to depths of at least 17 m, the maximum depth of the survey. Below the mangrove a lower resistivity layer persists in the depth range of 1 4 m, thinning somewhat towards the woody uplands. Belo w the low resistivity layer lies a zone with higher resistivities between ~5 10 m depth. Any vertical Figure 2 9 (A) Profile showing resistivity inversion results. Horizontal axis is distance along the tr ansect (m) from the coastline (B) General interpretation of resistivity results showing corresponding large scale spatial patterns in conductivity. Resistivities are shown in ohm m. Black line and arrow indicate the transition from mangrove to upland vege tative communities. Dashed line represents a confining clay layer located under the mangrove that tapers out at the vegetative upland transition.

PAGE 41

29 stratification is less clear below the woody upland, where terrain resistivities are m. D iscussion Annual water levels in the hydrologically altered impoundment appear to be largely controlled by evapotranspiration. Long term water level records show that water levels in the mangrove are ge nerally lower than those in the Indian River Lagoon ( Figure 2 2 ). Evaporation is likely higher in the Indian River Lagoon (~1500 mm year 1 ) than evapotranspiration in the mangrove (~950 mm year 1 ) (Twilley and Chen 1998; Sumner and Belaineh 2005). Nevertheless, drawdown is greater in the mangrove than in the Indian River Lagoon because (a) the Indian River Lagoon receives fresh water inflows from the mainland and sea wa ter inflows from the Atlantic Ocean and (b) the Indian River Lagoon has an effective specific yield of 10 0 while the mangrove sediments have a specific yield of 10 1 ( Johnson 1967), which means that equal evapotranspiration will lead to a 10 fold greater d rawdown in the mangrove. This greater drawdown in the mangrove maintains the hydraulic gradient from the Indian River Lagoon to the mangrove. Daily water level variations in the impoundment also appear to be largely controlled by evapotranspiration. The I ndian River Lagoon Fourier spectrum shows a strong peak at 12.4 hours and weak peaks at 24.0 and 25.8 hours ( Figure 2 3 ). The 12.4 diurnal cycle, while the 24.0 and 25.8 hour peaks correspond to the solar and lunar phases,

PAGE 42

30 shows a strong peak at 24.0 hours and a weak peak at 12.0 hours ( Figure 2 3 ). The 12.4 evident. The 24.0 hour peak corresponds to the daily solar cycle, which controls evapotranspiration. These peaks can be seen in the mean daily head and stage cycles ( Figure 2 4 B). Mean stages in the Indian River Lagoon follow a sinusoidal curve, as might be expected for stages in a lagoon so close to a large inlet. Conversely, mean heads in the mangrove follow a pattern indicative of evapotranpiration control. Heads rise most st eeply between 24:00 and 06:00 when plants are not transpiring, and decline most steeply between 06:00 and 12:00 when plants are transpiring. Interestingly, heads begin to rise after 12:00, although the rate at which they rise is slower than later in the ev ening. The transition between the heads declining and rising is abrupt. This suggests that evapotranspiration slows considerably and perhaps even ceases in some species in the afternoon, which results in the rate of inflow exceeding the rate of outflow (i. e., the rate of evapotranspiration). It has been shown that mangrove evapotranspiration rates decrease when salinities are increased (Ball and Farquhar 1984; Biber 2006). However, it is unclear if high salinity results in evapotranspiration rates that are generally lower throughout the day or in xylem dysfunction and evapotranspiration rates that are lower in the afternoon when atmospheric

PAGE 43

31 demand is highest (Tyree 1997). These results suggest the latter may be occurring, although this is merely speculative and deserving of independent study. The observed Na + and Cl concentrations and D and 18 O compositions can both be modeled by theoretically evaporating precipitation and/or Indian River Lagoon water ( Figure 2 8 ). However, the amount of evaporation necessary to model the observed Na + and Cl concentrations far exceeds the amount of evaporation necessary to model the observed D and 18 O compositions. This implies that transpiration plays an important role in controlling salinities because transpiration concentrates solutes by exclusion of solutes at the roots, extrusion of solutes from glands, and/or storage of solutes in leaves that ultimately drop and decompose (Dawe s 1998), but does not result in isotope fractionation between the uptaken water and the remaining ground water (Gat 1996). Collectively, these results indicate that evapotranspiration plays an important role in controlling water levels and salinities in th e barrier island mangrove on the east central coast of Florida ( Figure 2 10 ). Evapotranspiration lowers water levels in the mangrove, which creates a hydraulic gradien t that drives water from the Indian River Lagoon to the mangrove. While infrequent tidal inundation of the mangrove does occur, the regular, daily pathway for this water movement is subsurface. Water evaporates and is transpired but solutes remain in solut ion, so mangrove water evapoconcentrates, becomes more dense, sinks, and creates the layering of high salinity waters in the subsurface 1 4 m below

PAGE 44

32 Figure 2 10 Conceptual model showing importance of evapo transpiration in controlling water levels and salinities in the barrier island mangrove. As water evaporates a head gradient is created causing Indian River Lagoon water to flow into the mangrove. The high density, hypersaline waters that result from the e vapoconcentration sink and cause layering in the subsurface below the mangrove. The solid line represents the sub marine and terrestrial land surface and the dashed line represents the clay layer that acts as a confining unit for the hypersaline waters. th e mangrove. However, the ultimate fate of the sinking, high salinity water remains unclear. Acknowledgments This study was supported by the Smithsonian Marine Sciences Network. This is contribution #811 from the Smithsonian Marine Station at Fort Pierce. Beth Fratesi, Kathryn Murphy, Pamela Stringer, Chris Reich, Jason Greenwood, and students in the USF Applied Geophysics course in Spring 2007 all helped with field data collection. We are grateful to Peter Swarzenski and the United States Geological Survey for the loan of the resistivity system and processing software. Jon Sumrall assisted by drafting Figure 10.

PAGE 45

33 References Alongi DM (2008) Mangrove forests: Resilience, protection from tsunamis and responses to global climate change. Estuarine Coastal and Shelf Science 76:1 13 Ataie Ashtiani B Volker RE, Lockington DA (2001) Tidal effects on groundwater dynamics in unconfined aquifers. Hydrological Processes 15:655 669 Ball MC, Farquhar GD (1984) Photosynthetic and stomatal responses of two mangrove species, Aegiceras corniculatum and Avicennia marina, to long term salinity and humidity conditions. Plant Physiolog y 74:16 Biber PD (2006) Measuring the effects of salinity stress in the red mangrove, Rhizophora mangle L. African Journal of Agricultural Research 1:1 4 Cardona Olarte P, Twilley RR Krauss KW Rivera Monroy V (2006) Responses of neotropical mangrove seed lings grown in monoculture and mixed culture under treatments of hydroperiod and salinity. Hydrobiologia 569:325 341 Chen R, Twilley RR (1999) Patterns of mangrove forest structure and soil nutrient dynamics along the Shark River Estuary, Florida. Estuarie s 22: 955 970 Clark ID, Fritz, P (1997) Environmental Isotopes in Hydrogeology. CRC Press LLC, Boca Raton, FL, USA Clesceri LS, Greenberg AE Eaton AD (1998) Standard Methods for the Examination of Water and Wastewater, twentieth edition. American Public H ealth Association, Washington, DC, USA

PAGE 46

34 Craig J (1961) Standard for Reporting Concentrations of Deuterium and Oxygen 18 in Natural Waters. Science 133 : 1833 1834 Cunha SR, Tognella De Rosa MMP, Costa CSB (2006) Salinity and flooding frequency as determinant of mangrove forest structure in Babitonga Bay, Santa Catarina state, southern Brazil. Journal of Coastal Research Special Issue 39:1175 1180 Dawes CJ (1998) Marine Botany first edition John Wiley & Sons, Inc, New York, NY USA Donnelly T, Waldron S,Tait A, Dougans J, Bearhop S (2001) Hydrogen isotope analysis of natural abundance and deuterium enriched waters by reduction over chromium on line to a dynamic dual inlet isotope ratio mass spectrometer. Rapid Communications in Mass Spectrometry 15 :1297 1 303 Drexler JZ, De Carlo EW (2002) Source water partitioning as a means of characterizing hydrological function in mangroves. Wetlands Ecology and Management 10: 103 113 Epstein S Mayeda T (1953) Variation of O 18 content of waters from natural sources. Geoche mica et Cosmochimica Acta 4 : 213 224 Feller IC (1995) Effects of nutrient enrichment on growth and herbivory of dwarf red mangrove ( Rhizophora mangle ). Ecological Monographs 65:477 505 Gat JR (1996) Oxygen and hydrogen isotopes in the hydrologic cycle Annual Review of Earth and Planetary Sciences 24:225 262

PAGE 47

35 Gonfiantini R (1978) Standards for stable isotope measurements in natural compounds. Nature 271 :534 536 Granek EF, Ruttenberg BI (2007) Protective capacity of mangroves during tropical storms: a ca Ecology Progress Series 343:101 105 Gubbins D (2004) Time series analysis and inverse theory for geophysicists. Cambridge University Press, Cambridge, UK Hughes CE, Binning P Willgoose GR ( 1998 ) Characte risation of the hydrology of an estuarine wetland. Journal of Hydrology 211:34 49 Johnson AI ( 1967 ) Specific Yield Compilation of Specific Yields for Various Materials. U.S. Geological Survey Water Supply Paper 1662 D. U.S. Government Printing Office, Washington, D.C Kandasamy K, Narayanasamy R (2005) Coastal mangrove forests mitigated tsunami. Estuarine Coastal and Shelf Science 65: 601 606 Lovelock CE, Feller IC, McKee KL, Thompson R (2005) Variation in mangrove forest structure and sediment character itics in Bocos del Toro, Panama, Caribbean Journal of Science 41:456 464 Lovelock CE, Feller IC Ellis J Schwarz AM H ancock N Nichols P Sorrell B (2007) Mangrove growth in New Zealand estuaries: the role of nutrient enrichment at sites with contrasting rates of sedimentation. Oecologia 153:633 641

PAGE 48

36 Lugo AE ( 1990 ) Fringe wetlands. In A. E. Lugo, M. M. Brinson, and S. Brown (eds.), Ecosystems of the World 15: Forested Wetlands. Elsevier, Amsterdam, The Netherlands p p 143 169 Maidment DR ( ed ) ( 1993 ) Handb ook of Hydrology. McGraw Hill Inc., New York, N Y, USA Majoube M (1971) Fractionnement en oxygne sa vapeur. Journal of Chemical Physics 197:1423 1436 McKee KL ( 1995 ) Interspecific variation in growth, biomass partitioning, and defensive characteristics of neotropical mangrove seedlings: Response to light and nutrient availability. American Journal of Botany 82:299 307 Melville AJ Connolly RM ( 2003 ) Spatial analysis of stable isotope data to determine primary sources of nut rition for fish. Oecologia 136:499 507 Montague CL, Alexander VZ, Percival HF (1987) Ecological effets of coastal marsh impoundments: A review. Environmental Management 11:743 756 Odum WE, McIvor CC (1990) Mangroves. In R. L. Myers and J. J. Ewel ( eds.), Ecosystems of Florida. University of Central Florida, Orlando, FL, USA, p p 517 548 Rey J, Kain T ( 1991 ) A guide to the salt marsh impoundments of Florida. University of Florida, Florida Medical Entomology Laboratory, Vero Beach, FL, USA. Rivera Monroy V H, Madden CJ Day JW Jr, Twilley RR Vera Herrera F Alvarez Guilln H ( 1998 ) Seasonal coupling of a tropical mangrove forest

PAGE 49

37 and an estuarine water column: Enhancement of aquatic primary productivity. Hydrobiologia 379:41 53 Surez N, Medina E ( 2005 ) Salini ty effect on plant growth and leaf demography of the mangrove, Avicennia germinans L. Trees 19:721 727 Sumner DM, Belaineh G ( 2005 ) Evaporation, precipitation, and associated salinity changes at a humid, subtropical estuary. Estuaries 28:844 855 Twilley R R, Chen R (1998) A water budget and hydrology model of a basin mangrove forest in Rookery Bay, Florida. Australian Journal of Freshwater and Marine Research 49:309 323 Tyree, MT ( 1997 ) The Cohesion Tension theory of sap ascent: current controversies. Journ al of Experimental Botany 48:1753 1765 Wang P, Horwitz MH ( 2007 ) Erosional and depositional characteristics of regional overwash deposits caused by multiple hurricanes. Sedimentology 54:545 564 White WN (1932) A method of estimating ground water supplies based on discharge by plants and evaporation from soil: results of investigations in Escalante Valley, Utah. US Geological Survey Water Supply Paper 659 A. US Government Printing Office, Washington, DC Zetina Rej n MJ, A rregun Snchez F Chvez EA ( 2006 ) Trophic structure and flows of energy in the Huizache Caimanero lagoon complex on the Pacific coast of Mexico. Estuarine, Coastal and Shelf Science 57:803 815

PAGE 50

38 CHAPTER 3: CHANGES IN THE SPATI AL VARIABILITY O F TERRAIN CONDUCTIV ITY IN A MANGROVE, INDIAN RIV ER LAGOON, FL ORIDA Abstract We examined the differences in spatial variability of terrain conductivity in a mangrove on a carbonate barrier island along the Indian River Lagoon under two hydrologic management regimes, breached rotational impoundment management and rotational impoundment management. Continuous water level measurements were recorded in a piezometer for two years. EM31 and EM34 measurements were taken throughout the mangrove during both regimes and then inverted to obtain terrain conductivity for the shallow and deep layers, respectively. The difference in coefficient of variation ( CV) between the breached RIM and RIM data was calculated to examine spatial variability in both the shallow and d eep layers. A null hypothesis model was employed to examine the statistical significance of the CV results, using the 90% confidence interval as a benchmark. The average water levels were 0.06 m amsl and 0.49 m amsl during the breached RIM and RIM regi mes, respectively. The average shallow (EM31) layer terrain conductivity shifted slightly from 1868 mS m 1 to 1825 mS m 1 after the alteration in management regime, yet the standard deviation of these averages decrease from 656 mS m 1 to 216 mS m 1 The average deep

PAGE 51

39 (EM34) layer terrain conductivities were 328 mS m 1 and 255 mS m 1 during the breached RIM and RIM regimes, respectively. The temporal CVs were 0.23 and 0.04 for the shallow and deep layers, respectively. The null hypothesis model for the shallow layer illustrates that the CV of 0.23 is outside the 90% confidence interval and determines that the difference in spatial structure is statistically significant. The deep layer CV was not statistically significant. These results indicate that the transition from breached RIM to RIM resulted in changes to both the physical and chemical hydrologic character of the impoundment, especially in the shallow layer. Introduction Mangroves, which cover ~240,000 km 2 of sheltered subtropical and tropical c oasts between latitudes 24 north and south where mean annual temperatures are >20C (Lugo 1990; Dawes 1998), provide numerous ecological functions as well as goods and services. Mangroves support estuarine and near shore marine productivity, in part by pr oviding critical habitat for juvenile fish and through the export of nutrient rich water (McKee 1995; Rivera Monroy et al. 1998) or plant, algal, or animal biomass (Zetina Rejn et al. 2006). Mangroves also protect coastal habitats against the destructive forces of hurricanes, typhoons, and tsunamis (e.g., Kandasamy and Narayanasamy 2005; Granek and Rutenberg 2007; Alongi 2008). Mangroves also have historically provided breeding grounds for numerous species of mosquito (Ritchie and Addison 1992). In Florid a, early accounts tell of

PAGE 52

40 mosquitoes so numerous that early European explorers were forced to sleep on the beach covered with sand (Connelly and Carlson 2009) and mosquito borne disease outbreaks were common, with the 1888 outbreak of yellow fever in Jacks onville killing 400, sickening 5,000, and causing 10,000 to flee the city, resulting in the temporary loss of ~40% of the total population ( Mulrennan 1986 ). Consequently, insect control efforts have long been undertaken, including ditching and draining, th e widespread application of pesticides, and the construction of impoundments (Connelly and Carlson 2009). Impoundments were shown to be effective in the 1930s (Hull and Dove 1939), but were not widely constructed from the 1950s to the 1970s (Rey and Kain 1 991; Connelly and Carlson 2009). By the mid 1970s, more than 16,400 ha of mangroves had been impounded on the Indian River Lagoon (Rey and Kain 1991). Most impoundments consist of perimeter levees constructed between mean lower low and mean higher high t ide. Culverts are typically placed beneath the levees to facilitate control over inflow and outflow. In some cases, gates are installed to either maintain continuous or seasonal flooding; in other cases, gates are either not installed or are installed but not closed to allow free inflow and outflow. More recently, pump systems have been increasingly installed and operated to allow better control over seasonal inflow and outflow and associated flood durations and depths (Connelly and Carlson 2009). Impoundme nt can result in numerous environmental impacts, including changes in the occurrences of certain species of fish (Harrington and Harrington 1982), birds (Provost 1959), and vegetation (Rey et al. 1990). In some cases,

PAGE 53

41 monotypic stands of red mangrove ( Rhiz ophora mangle ) replace spatially complex stands of red mangrove, black mangrove ( Avicennia germinans ), herbaceous halophytes (e.g., Batis martima ), and unvegetated salt pans (Rey et al. 1990). Porewater salinities can differ between the different habitat t ypes in these spatially complex stands ( Odum and McIvor 1990; Stringer et al. 2010 ), suggesting that impoundment both reduces overall salinities thereby favoring less salt tolerant red mangrove and reduces the spatial complexity of salinity distributions thereby favoring monotypic stands of red mangrove. This study is part of a broader series of investigations of the controls on species composition, primary productivity, and nutrient cycling in an impoundment on a carbonate barrier island on the east central coast of Florida. A previous investigation showed that evapotranspiration plays a major role in controlling salinities by lowering the water table in the mangrove and creating a hydraulic gradient that draws surface water and ground water from the lagoon into the mangrove (Stringer et al. 2010 ). Since the conclusion of that study, management has changed, with culverts without gates allowing free exchange between the impoundment and the adjacent lagoon being replaced by culverts with gates a nd pumping of water into the impoundment from the adjacent lagoon. Therefore, this subsequent investigation tests the hypothesis that flooding the impoundment both reduces overall salinities and reduces the spatial complexity of salinity distributions.

PAGE 54

42 Stu dy Area The study was conducted in SLC 24, a mosquito control impoundment located at N27 o 33', W80 o 33' on the west shore of North Hutchinson Island This barrier island is ~35 km in length and ~0.2 2 km in width, and is part of the system of barrier islands that bound the Indian River Lagoon, a 250 km estuarine system located on the east central coast of Florida ( Figure 3 1 ) The impoundment was created in 1970 when a d ike was constructed around an existing mangrove dominated wetland (Rey et al. 1990; Rey and Kain Figure 3 1 Local setting showing EM31 and EM34 sampling points, piezometer s and instrumentation, as well as cross section location

PAGE 55

43 1991). Culverts were not installed so there was no surface water exchange between the impoundment and the lagoon between 1970 and 1985. Surface water exchange was minimally restored in 1985 when one culvert was installed to facilita te the drainage of water that had accumulated in the impoundment during two tropical storms. The culvert was later closed so there was no surface water exchange between the impoundment and the lagoon again until 1987 when the existing culvert was reopened and four additional culverts were installed. From 1987 to 2009, these culverts remained open, allowing focused but otherwise free exchange of surface water between the impoundment and the lagoon, a management strategy called breached rotational impound man agement (breached RIM) (Middleton et al. 2008). In 2009, pumps and tide gates were installed and have since been operated to maintain surface water inundation during the summer months, a management strategy called rotational impoundment management (R IM) (M iddleton et al. 2008). Continuous core collected to depths of approximately 10 m in four locations across the barrier island indicate that there is ~5 m of fine sand, which is underlain by ~1.5 m of clayey sand, which is underlain shelly fine sand ( Figu re 3 2 ). The clay rich layer apparently acts as an aquitard, with surface water and ground water above the aquitard evaporating and concentrating to develop hypersaline e nvironments, particularly in areas that are infrequently flooded by high tides (Stringer et al. 2010). Vegetation cover in the impoundment decreased from 75% to near 30% between 1970 and 1985 but began to recover following the installation of the

PAGE 56

44 Figu re 3 2 Geologic cross section characterizing the site subsurface on a west east transect. The matrix is dominated by sand and sand with clay closest to the ground surface, with a confining layer of clayey s and and sandy clay occurring at 3 m below mean sea level. culverts in 1987 (Rey et al. 1990). Black mangrove ( Avicennia germinans (L.) L. ) is the dominant mangrove, but red ( Rhizophora mangle L ) and white ( Laguncularia racemosa (L.) C.F. Gaertn.) mangrove also are common. Prior to the change to RIM, the abundance, density, and sizes of mangroves varied across the impoundment from highly saline salt pans that had no mangroves or were fringed by dwarf black mangroves, areas that had dense and relatively shor t black mangroves, areas that had larger and more widely spaced black mangroves, to areas associated with open water where red mangroves were most abundant.

PAGE 57

45 Methods Water Levels and Ground Water Conductivities Piezometers were installed at 18 locations. At 17 locations, two piezometers were installed, with piezometers being constructed of stainless steel, having 2.5 cm inside diameters, and being screened over 15 cm beginning at 1.2 m and 1.8 m below the ground surface. Ground water conductivity was measured at these piezometers using a YSI 556 MPS ( YSI Inc., Yellow Springs, O hio, USA). At one location, one piezometer was installed, with the piezometer being constructed of PVC, having a 5.0 cm inside diameter, and being screened over 30 cm beginning at 2.0 m below the ground surface. Ground water head was measured at this piezo meter on 15 minute intervals with Model 3001 Levelogger (Solinst, Georgetown, Ontario, Canada). The pressure transducers had a 1.5 m range and a 1 mm accuracy. Head data were compensated for changes in barometric pressure using barometric pressure data al so measured on 15 minute intervals with a Barologger (Solinst, Georgetown, Ontario, Canada). Electromagnetic Survey Measurements Electromagnetic surveys were conducted while the impoundment was under breached RIM in April 2007 and under RIM in September 2009 ( Figure 3 1 ). Electromagnetic measurements were made throughout the site using EM 31 and EM 34 instruments (Geonics LTD., Mississauga, Ontario, Canada). For each instrument, a transmitter coil was placed at the ground surface to generate alternating currents that induced eddy currents in the subsurface. Those eddy

PAGE 58

46 currents generated secondary magnetic fields that were measured, along with the primary magnetic fiel ds, by receiver coils placed specific distances away from the transmitters. The ratios of the secondary to primary magnetic fields are proportional to the apparent conductivities (McNeill 1980). Inversion modeling software, EMIX 34 (Interpex LTD, Golden, C olorado, USA), was then used to invert the apparent conductivities to provide terrain conductivities, which are the combined conductivities of the sediments and the porewater fluids. The EMIX 34 inversion modeling software uses given values for instrument height, dipole mode, frequency, and coil spacing to invert the set of instrument responses to find the best fit values of one or more layer conductivities using a ridge regression estimation algorithm (Inman 1975). The EM 31 was operated with a fixed coil spacing of 3.7 m at a frequency of 9.8 kHz. At each location, measurements were taken in the horizontal dipole and vertical dipole modes, providing an effective penetration of 3 m and 6 m, respectively. Two measurements were taken in each mode; one on the ground surface and one at hip height, ~0.75 m above ground surface. Data were interpreted using the EMIX 34 inversion modeling software with a one layer model. The EM 34 was operated at 10 m and 20 m spacings at frequencies of 6.4 kHz and 1.6 kHz, respec tively. Measurements were only taken in the horizontal dipole mode, resulting in an effective penetration of 7.5 m and 15 m for the 10 m and 20 m spacing, respectively. No measurements were taken in the vertical dipole mode because the terrain was too cond uctive for the range of the

PAGE 59

47 instrument in this configuration. Data were interpreted using the EMIX 34 inversion modeling software with a one layer model. Ground Water vs. Terrain Conductivity Ground water conductivities were measured at the same time that EM 31 measurements were taken in both vertical and horizontal dipole modes adjacent to the piezometers. The measured ground water and modeled terrain conductivities were then compared using least squares regression in Excel (Microsoft Corporation, Redmond, Washington, USA) to ascertain the validity of using terrain conductivity as a proxy for ground water conductivity. Null Hypothesis Modeling The terrain conductivity data were used to test the hypothesis that there was greater spatial variation in terrain conductivities while the impoundment was under breached RIM in April 2007 than while the impoundment was under RIM in September 2009. First, the impoundment was stratified into shallow and deep layers, with all EM 31 data assigned to the shallow layer and all EM 34 data assigned to the deep layer. The coefficient of variation (CV), the ratio of the standard deviation to the mean, was calculated for each layer in each year. For each layer, the CV from September 2009 was subtracted from the CV from April 2007 which yielded the difference in the coefficient of determination ( CV). Subsequently, a null model was used to randomly reassign terrain conductivities within each layer, with measured values in any location in any year randomly reassigned to any locatio n in any year (Lewis et al 2006; Lewis et al. 2007). This reassignment was completed 5 x 10 5 times, with a new CV calculated for each

PAGE 60

48 of the iterations. The observed CVs were then compared to the frequency distribution of the new CVs generated by the nu ll model, and the observed CVs were determined to be significantly different than the new CVs generated by the null model if they fell outside of the 90% confidence interval. Results Physical Hydrology Continuous water level measurements from water years 2007 (i.e., October 1, 2006 to September 30, 2007) and 2009 (i.e., October 1, 2008 to September 30, 2009) show the change in water levels due to the change in impoundment management approach. In 2007, mean standard deviation head Figure 3 3 Water level in the mangrove (A) under breached RIM regime in water year 2007, and (B) under RIM regime in water year 2009. The management regime change occurred in March 2009. The arrow in panel A indi cates the first sampling period in April 2007. The arrow in panel B indicates the sampling period in September 2009. All measurements were recorded on 15 minute intervals with a pressure transducer and data logger.

PAGE 61

49 was 0.09 0.10 m amsl from October to M arch and 0.08 0.11 m amsl from April through September ( Figure 3 3 A). In 2009, mean standard deviation head was 0.06 0.19 from October to March and 0.49 0.14 m amsl from April through September ( Figure 3 3 B). Relationship between Ground Water and Terrain Conductivities Ground water conductivities ranged from 1,700 to 15,0 00 mS m 1 while the modeled terrain conductivities ranged from 330 to 3,800 mS m 1 ( Figure 3 4 ). Least squares regressions of ground water conductivity (dependent variable) against terrain conductivity (independent variable) were performed for both the 1.2 m and 1.8 m piezometers data sets. Figure 3 4 Correlation between ground water conductivity and modeled terrain conductivity. The black diamonds and open squares represent ground water conductivities taken at piezometers with a depth of 1.2 m and 1.8 m, respectivel y. The solid black and grey lines represent the least squares regression for the 1.2 m and 1.8 m conductivities, respectively. The dashed black and grey lines represent the standard deviation (SD) lines for the 1.2 m and 1.8 m conductivities, respectivel y.

PAGE 62

50 The standard deviation lines (SD lines) were calculated and plotted to illustrate the correlation that would occur with a perfect fit ( R = 1). With actual values of R = 0.79 ( p = 0.02) and R = 0.87 (p < 0.01) for the 1.2 m and 1.8 m piezometers, respec tively, the slopes of the fitted regression lines are reduced accordingly (by 21% and 13% respectively). The R 2 values mean that the fitted lines capture 62% and 76%, respectively, of the variance of the ground water conductivity, assuming a linear relati onship between ground water conductivity and terrain conductivity, and no error in the measurements of terrain conductivity. Inspection of the residuals from the regression lines gives no indication that the assumption of a linear relationship for these data is inappropriate. The most likely sources of the unexplained variability in ground water conductivity, therefore, is interpreted to be measurement error and differences in the volumes sampled during the measurement of the two conductivity variables. With regards to the latter source, ground water conductivities were measured on fluids drawn from a small volume immediately surrounding the screened intervals 1.2 m and 1.8 m below the ground surface, while terrain conductivities were integrated over a d epth of ~6 m and sample volumes of ~490 m 3 There could be heterogeneity in the ground water conductivities within that larger volume, resulting in different values. Terrain Conductivity Modeled terrain conductivities for the 2007 EM 31 survey had a mean standard deviation of 1,868 656 mS m 1 and ranged from 116 to 2,917 mS m 1 (n=38). Analysis of the spatial arrangement of these values shows an axis of

PAGE 63

51 anisotropy at 143.6 ( Figure 3 5 A). The highest conductivities occurred along this NW to SE axis, while the lowest conductivities occurred in the northeastern corner of the study area. The terrain c onductivity values of the 2009 EM 31 survey had a mean standard deviation of 1,825 216 mS m 1 and ranged from 1,489 to 2,430 mS m 1 (n=43). Forty two percent of the values were in the range of 1,800 to 1,900 mS m 1 resulting in almost no spatial variability or structure in the data ( Figure 3 5 B). The terrain conductivities for the 2007 EM 34 survey had a mean standard deviation of 328 40 mS m 1 and ranged from 246 to 404 mS m 1 (n=18). Analysis of the spatial arrangement of these values reveals an axis of anisotropy at 152.2 ( Figure 3 6 A). The lowest c onductivities occurred along this NW to SE axis, with values increasing gradually toward the edges of the study area. The terrain conductivities for the 2009 EM 34 survey had a mean standard deviation of 255 42 mS m 1 and ranged from 201 to 378 mS m 1 (n=37). Analysis of the spatial arrangement of these values shows an axis of anisotropy at 7.7 ( Figure 3 6 B). The lowest values occurred along this W to E axis, with values increasing toward the N end of the study area. Temporal Variability in Terrain Conductivity The CV values for the random reassignment of EM31 data from 2007 and 2009 ranged from 0.21 to 0.23, in a normal distribution (Kolmogorov S mirnov non normality test, p = 0.19) ( Figure 3 7 A). The CV of the actual data, 0.23, had a z score of 3.35. Because the z score exceeds the 1.96 threshold (i.e., the 95% confidence interval), we reject the null hypothesis and conclude

PAGE 64

52 that the CV was not produced by a random rearrangement of the 2007 and 2009 data. There was a statistically significant change in the spatia l structure of near surface terrain conductivity values from the 2007 to 2009 sampling periods. Figure 3 5 Contoured surfaces of shallow (EM31) terrain conductivities under ( A) breached RIM in 2007 and ( B) RIM in 2009. All units are in mS m 1 The average conductivities were 1,868 656 mS m 1 (n=38) in 2007 and 1,825 216 mS m 1 (n=43) in 2009.

PAGE 65

53 Figure 3 6 Contoured surfaces of deep (EM34) terrain conduc tivities under ( A) breached RIM in 2007 and ( B) RIM in 2009. All units are in mS m 1 The average conductivities were 328 40 mS m 1 (n=18) in 2007 and 255 42 mS m 1 (n=37) in 2009.

PAGE 66

54 Figure 3 7 Histograms showing null hypothesis modeling results for (A) shallow (EM31) CV randomized values and (B) deep (EM34) CV randomized values. Both plots illustrate the distribution of 500,000 randomizations. The arrows show the position of the observed C V values in relation to the randomized values. The observed CV for the EM31 measurements had a z score of 3.35 indicating there was a statistically significant change in spatial structure in terrain conductivity values from 2007 to 2009. The CV values f or the random reassignment of EM34 data from 2007 and 2009 ranged from 0.12 to 0.11, in a normal distribution (Kolmogorov Smirnov non normality test p = 0.38) ( Figure 3 7 B). The CV of the actual data, 0.04, had a z score of 1.56. Because the z score falls outside the 1.96 threshold (i.e., the 95% confidence interval), we do not reject our null hypothesis. There was not a statistically significant change in spatia l structure of deep terrain conductivity values from the 2007 and 2009 sampling periods. Discussion The transition from breached RIM to RIM resulted in changes to both the physical and chemical hydrologic character of the impoundment. Overall, there

PAGE 67

55 were clear changes in both water levels and the variability and spatial structure of porewater salinities. Under breached RIM, water levels generally ranged between 0.5 m to 0.1 m relative to mean sea level throughout the year ( Figure 3 3 A). Under RIM, water levels generally ranged between 0.5 m to 0.1 m relative to mean sea level when the impoundment was not flooded (i.e., October 2008 to February 2009), but greatly incr eased and generally ranged between 0.1 m to 0.6 m relative to mean sea level when the impoundment was flooded (i.e., March 2009 to the end of the study in September 2009) ( Figure 3 3 B). Mean standard deviation elevations at the 18 piezometers were 0.02 .22 m relative to mean sea level, so this relatively slight change in water level resulted in a relatively large change in water levels relative to the ground surfa ce, with water levels being generally below ground surface under breached RIM and water levels being generally above ground surface under RIM when the impoundment was flooded. Under breached RIM, there were diverse shallow salinity environments throughout the impoundment, ranging from salinities lower than sea water under the uplands to salinities higher than sea water in the infrequently flooded parts of the mangrove ( Figure 3 5 A). Stringer et al. (2010) attributed both, at least in part, to infrequent flooding by tidal waters, with the former being lower than sea water because of the effects of a fresh water lens fed by rainwater and the latter being higher than sea water because of the effects of the evapoconcentration of ground water that flowed in from the Indian River Lagoon.

PAGE 68

56 The extensive and long term flooding under RIM changed these conditions, resulting in significant changes in shallow salinity distributions throughout the impoundment. Overall, there was little change in the shallow mean terrain conductivities (i.e., 1,868 mS m 1 to 1,825 mS m 1 ). However, there was significant change in the variability and spatial structure of the shallow terrain conductiviti es. Under breached RIM, shallow terrain conductivities ranged from 116 to 2,917 mS m 1 with the highest shallow terrain conductivities being located along a NW to SE axis on the western side of the shoreline, closest to Indian River Lagoon ( Figure 3 5 A). Under RIM, shallow terrain conductivities ranged from 1,489 to 2,430 mS m 1 with little spatial structure to the shallow terrain conductivity distributions ( Figure 3 5 B). The change in structure was so significant that 500,000 random reassignments of the data could not produce the same CV that was calculated using the observed val ues ( Figure 3 7 A). Conversely, the extensive and long term flooding under RIM has not yet resulted in significant changes in deep salinity distributions throughout the impoundment. Though the spatial structure of the deep terrain conductivities visibly changed (Figure 6A and 6B), the change in the structure was not significant because 500,000 random reassignments of the data routinely produced the same CV that was calculated using the observed values ( Figure 3 7 B). However, the mean deep terrain conductivities did decrease from 328 to 255 mS m 1 suggesting that there may be a trend toward lower salinities and perhaps less spatially structured salinities in the deep subsurface of the imp oundment.

PAGE 69

57 References Alongi DM (2008) Mangrove forests: Resilience, protection from tsunamis and responses to global climate change. Estuarine Coastal and Shelf Science 76:1 13 Connelly CR and Carlson DB (eds) (2009) Florida Mosquito Control 2009: The Stat e of the Mission as Defined by Mosquito Controllers, Regulators, and Environmental Managers. University of Florida, Institute of Food and Agricultural Sciences, Florida Medical Entomology Laboratory, Vero Beach, Florida Dawes CJ (1998) Marine Botany, first edition. John Wiley & Sons, Inc, New York, N ew York USA Granek EF, Ruttenberg BI (2007) Protective capacity of mangroves during Ecology Progress Series 343:101 105 Harrington RW Jr, Harrington ES (1982) Effects on fishes and their forage organisms of impounding a Florida salt marsh to prevent breeding by salt marsh mosquitoes. Bulletin of Marine Science 32:523 531 Hull JB, and Dove WE (1939) Experimental diking for control of sand fl y and mosquito breeding in Florida saltwater marshes. Journal of Economic Entomology 32:309 312 Inman JR (1975) Resistivity inversion with ridge regression. Geophysics 40:798 817

PAGE 70

58 Kandasamy K, Narayanasamy R (2005) Coastal mangrove forests mitigated tsunami Estuarine Coastal and Shelf Science 65: 601 606 Lewis DB, Schade JD, Huth AK, Grimm NB (2006) The spatial structure of variability in a semi arid, fluvial ecosystem. Ecosystems 9:386 397 Lewis DB, Grimm NB, Harms TK, Schade JD (2007) Subsystems, flowpath s, and the spatial variability of nitrogen in a fluvial ecosystem. Landscape Ecology 22:911 924 Lugo AE (1990) Fringe wetlands in Lugo AE, Brinson MM, Brown S (eds ), Ecosystems of the World 15: Forested Wetlands. Elsevier, Amsterdam, Netherlands, pp 143 16 9 McKee KL (1995) Interspecific variation in growth, biomass partitioning, and defensive characteristics of neotropical mangrove seedlings: r esponse to light and nutrient availability. American Journal of Botany 82:299 307 Middleton B, Devlin D, Proffitt E McKee K, Cretini KF (2008) Characteristics of mangrove swamps managed for mosquito control in eastern Florida, USA. Marine Ecology Progress Seress 371:117 129 Mulrennan JA, Jr. (1986) Mosquito control its impact on the growth and development of Florid a. Journal of the Florida Medical Association 73:310 311 Odum WE, McIvor CC (1990) Mangroves. In Myers RL, Ewel JJ (eds ), Ecosystems of Florida. University of Central Florida, Orlando, F lorida USA, pp 517 548

PAGE 71

59 Provost MW (1959) Impounding salt marshes for mosquito control and its effects on bird life. Florida Naturalist 32:163 170 Rey JR, Kain T (1991) A guide to the salt marsh impoundments of Florida. University of Florida, Florida Medical Entomology Laboratory, Vero Beach, Florida Rey JR, Shaffer J, Trema in D, Crossman RA, and Kain T (1990) Effects of re establishing tidal connections in two impounded tropical marshes on fishes and physical conditions. Wetlands 10: 27 47 Ritchie SA, Addison DS (1992) Oviposition preferences of Aedes taeniorhynchus (Diptera : Culiocidae) in Florida mangrove forests. Environment and Entomology 21:737 744 Rivera Monroy V H, Madden CJ, Day JW Jr, Twilley RR, Vera Herrera F, Alvarez Guilln H (1998) Seasonal coupling of a tropical mangrove forest and an estuarine water column: Enhancement of aquatic primary productivity. Hydrobiologia 379:41 53 Stringer C E Rains MC, Kruse S Whigham D (2010) Controls on water levels and salinity in a barrie r island mangrove Indian River Lagoon, Florida. Wetlands 30:725 734 Zetina Rejn MJ, A rregun Snchez F Chvez EA ( 2006 ) Trophic structure and flows of energy in the Huizache Caimanero lagoon complex on the Pacific coast of Mexico. Estuarine, Coastal and Shelf Science 57:803 815

PAGE 72

60 CHAPTER 4: WATER SOURCES IN MAN GROVES IN THREE HYDR OGEOMORPHIC SETTINGS ON THE COST ALEGRE, JALISCO, MEX ICO Abstract We examined the fresh water and seawater contributions to three distinct mangrove communities along the Costalegre on the central Pacific coast of Mexico. Surface water, ground water, fresh water, and seawater samples were all collected over two sampling periods representing the wet and dry seasons. Salinities varied by water type, with values of ~9 in La Manzanilla, ~17 in La Vena, ~33 in Barra de Navidad, ~0.4 in the fresh waters, and ~34 in the seawater. Sodium and Chloride concentrations and isotopic signatures, as well as salinity, were used as tracers in mass balance mixing models to quantify estimat es of relative fresh water and seawater contributions to each site. La Manzanilla, a basin mangrove, had mean fresh water contribution estimates of 63 84%. La Vena, a riverine mangrove, had fresh water estimates of 39 51%. Barra de Navidad, a fringe mang rove, had low fresh water contributions of 0 5%. These results illustrate the varying role that groundwater plays in mangrove hydrodynamics, depending on the site hydrogeomorphology.

PAGE 73

61 Introduction More than one oastal areas and 2006). These large coastal populations stress the very natural resources on which they depend for the goods and services they provide (Barbier et al. 2008). Fo r example, recent estimates report up to a 30% loss of coastal wetlands worldwide, largely due to anthropogenic activities (Solomon et al. 2007). Mangroves are coastal ecosystems that cover ~180,000 km 2 of sheltered subtropical and tropical coasts between latitudes 24 north and south where mean annual temperatures are >20C (Dawes 1998). Mangroves are just one example of coastal wetlands being rapidly lost. In recent decades, global mangrove area has declined ~35% overall at a rate of ~2.1% per year (Alon gi 2002; Martinuzzi et al. 2009), while North and South American mangrove area has declined ~38% overall at a rate of ~3.6% per year (Valiela et al. 2001; Valiela et al. 2009). These high rates of loss make mangroves the most threatened major coastal habi tat in the world (Valiela et al. 2009). As in the case of other coastal habitats, anthropogenic activities are largely responsible for the loss and degradation of mangrove ecosystems (Alongi 2002; Gunawardena and Rowan 2005; Islam and Haque 2004; Vijay et al. 2005). The central Pacific coast of Mexico has ~64.5 km 2 of mangrove habitat, ~22 km 2 of which are in the state of Jalisco (RAMSAR 2010; CONABIO 2007). Much of this is on the Costalegre, a 260 km coastline from Puerto Vallarta, Jalisco in the north t o Manzanillo, Colima in the south, that is 400 km from

PAGE 74

62 Guadalajara and 950 km from Mexico City. This coastal region has traditionally been home to small agricultural and fishing communities but, in recent years, has been one of eight targeted for developm ent by the National Fund for Tourism Development and the Secretariat of Tourism, federal agencies responsible for tourism development throughout the nation (Wilson 2008). As a result, tourism has dramatically increased, with many of these small communitie s experiencing almost a complete transformation from economies based upon local agriculture and fishing to more complex economies based on tourism, real estate, and commerce (Everitt et al. 2008). Much of the related population growth is due to the additi on of European and North American expatriates and retirees, most of whom expect higher standards of living than do the traditional residents. These shifts in population and demographics have increased the stresses placed on the natural resources, includin g the aforementioned mangrove environments. These mangroves play vital roles in supporting both the regional ecology and economy, but sound management and conservation strategies are essential to ensure that they can continue to do so in spite of the rapid development and rapidly shifting demographics (Valiela et al. 2001; Alongi 2002; Martinuzzi et al. 2009). This need is amplified, because approximately one third of the mangroves in Jalisco are designated as wetlands of international importance because o f the significance of the ecological goods and services they provide (RAMSAR 2010). The flux of water into and out of mangroves controls the physical and chemical hydrological characteristics of mangroves and facilitates the exchange

PAGE 75

63 of mass, energy, and organisms between mangroves and the surrounding hydrological landscape (Twilley and Chen 1998). Mixing between fresh water and seawater plays a role in controlling salinity (Drexler and De Carlo 2007; Stringer et al. 2010). Plant community composition (O dum and McIvor 1990; Stringer et al. 2010) and primary productivity (Suarez and Medina 2005; Suarez and Medina 2006; Cardona Olarte et al. 2006) are correlated with salinity. Also, nutrient rich fresh water flowing into mangroves can enhance mangrove prim ary productivity (McKee 1995) while nutrient rich mangrove water flowing into lagoons and near shore marine environments can enhance water column primary productivity (Rivera Monroy et al. 1998). As such, understanding the water sources in mangroves is a c ritical first step in developing sound management strategies. Water sources can vary, with the relative contributions of fresh water and seawater varying depending on climate, geology, and the proximity to the coast (Drexler and De Carlo 2002; Barlow 2004 ; Stringer et al. 2010). Source water contributions also can vary tremendously between different mangrove functional environments in close proximity to one another (Drexler and De Carlo 2002). However, detailed studies are time consuming and expensive, s o there is a need for simple conceptual models that can be used to rapidly infer water sources and better inform land use planning and decision making, particularly lesser developed nations. One such simple conceptual model might be developed from the hyd rogeomorphic based mangrove classification first proposed by Lugo and Snedaker (1974) and modified thereafter, most recently by Woodroffe (2002).

PAGE 76

64 The classification is typically displayed as a ternary diagram, with the vertices representing interior, rive r dominated, and tide dominated mangroves, where interior mangroves are closed basins, river dominated mangroves are dominated by unidirectional river flows, and tide dominated mangroves are characterized by bidirectional tidal flows. The purpose of this study is to quantify the relative contributions of fresh water and seawater to each of these different mangroves classes, using three mangroves on the Costalegre as case studies. To that end, we used basic water parameters (salinity), conservative ion con centrations (sodium and chloride), and stable isotopes (deuterium and oxygen 18) as input data to mass balance mixing models to determine the relative contributions of fresh water and seawater to each mangrove setting. Study Site Description This study wa s conducted in three separate mangroves located on the Costalegre on the central Pacific coast of Mexico in the state of Jalisco ( Figure 4 1 ). This region lies at th e western edge of the Sierra Madre del Sur, which is characterized by alternating basins and ridges with few large alluvial plains. During the middle Tertiary period, the region was subject to widespread volcanic activity, resulting in massive intrusions of magma reaching the surface in batholiths sized domes and shields (Barrera and Vargas 2007). All of this material cooled and solidified, forming the rhyolites, dacites, adesites, basalts, igneous breccias, and tufts that now make up the rugged basin and ridge

PAGE 77

65 Figure 4 1 Site map showing general location of the Costalegre along the central Pacific coast of Mexico. The three insets show each of the three mangrove study sites: A) La Vena, B) La Manzanilla and C) Barra de Navidad. Each sampling point is labeled with the name used to identify it throughout the paper.

PAGE 78

66 topography of the modern Sierra Madre del Sur. The bays on the Costalegre are formed by the intersection of the alternating basins and ridges and the Pacific Ocean, with the basins providing protected depositional environments in which form coastal alluvial plains on which Quaternary stream and oceanic sediments are deposited (Barrera and Vargas 2007). All of the study sites are located in the se protected bays, with two located in the Bah a of Tenacatita and one located in the Bah a de Navidad. The regional average temperature is 25.3 C and yearly rainfall is 700 800 mm, occurring mostly in the summer months (SMN 2010). The La Manzanilla man grove is an interior mangrove typically impounded by a semi permanent beach ridge ( Figure 4 1 ). The La Manzanilla mangrove is ~2.5 km 2 in size, being ~4.3 km parallel to the shoreline and extending ~0.6 km inland (RAMSAR 2010). The Rio Purificac on and other smaller, unnamed streams flow toward the La Manzanilla mangrove. However, these streams flow only ephemerally in response to wet season (i.e., June October) stor ms, so infrequently that defined channel beds and banks do not extend across the alluvial fan to the La Manzanilla mangrove. During particularly intense wet season storms, such as tropical cyclones, the beach ridge can be breached forming an intermittent connection between the La Manzanilla mangrove and the Bah a de Tenacatita. The predominant mangrove species are the white mangrove ( Laguncularia racemosa L.) and the red mangrove ( Rhizophora mangle L.) along with the mangrove associate, buttonwood ( Conoca rpus

PAGE 79

67 erectus ) (RAMSAR 2010). The La Manzanilla mangrove is a RAMSAR site, having received this status in 2008. The La Vena mangrove is a river dominated mangrove perennially connected to the Bah a de Tenacatita ( Figure 4 1 ). The La Vena mangrove is ~0.6 km 2 in size but is long and linear, being ~5.0 km in length and ~0.1 km in width. No streams flow into the La Vena mangrove. Instead, the La Vena mangrove is fed b y ground water discharge from a headwater spring which maintains perennial flows and a perennial connection between the La Vena mangrove and the Bah a de Tenacatita. The predominant mangrove species is the red mangrove ( Rhizophora mangle L.). The Barra de Navidad mangrove is a tide dominated mangrove located on Laguna de Navidad, a small, shallow lagoon perennially connected to the Bah a de Navidad ( Figure 4 1 ). Th e lagoon is ~3.7 km 2 in area and is fringed by ~5.7 km 2 of mangroves, much of which being located on a broad delta prograding into the lagoon (CONABIO 2007; Mendez Linares et al. 2007; RAMSAR 2010). The Rio Arroyo Seco, the Canal de Marabasco and other smaller, unnamed streams flow into the Barra de Navidad mangrove, with a combined watershed area of 3.9 km 2 (RAMSAR 2010).The predominant mangrove species are the white mangrove ( Laguncularia racemosa L.) and the red mangrove ( Rhizophora mangle L.) along with the mangrove associate buttonwood ( Conocarpus erectus ) (RAMSAR 2010). The Barra de Navidad mangrove is a RAMSAR site, having received this status in 2008.

PAGE 80

68 Methods Water samples were collected and analyzed for temperature, pH, and salinity in the dry and wet seasons (i.e., January and July, respectively) in 2007; and temperature, pH, salinity, sodium and chloride concentrations, and stable isotopes (deuterium and oxygen 18) in the dry and wet seasons (i.e., January and July, respectively) in 2008. The lone exception was that no water samples were collected from the Barra de Navidad study site in the dry season of 2007. Water samples were collected from both in, and the areas surrounding, the mangroves. In the mangroves, water samples were collected from surface water bodies in all three mangroves and from shallow piezometers in the La Manzanilla mangrove (i.e., mangrove waters). Around the mangroves, water samples were collected opportunistically in a variety of irrigation and drinking wat er wells, springs and seeps, and rivers (i.e., fresh waters) and in the open waters in the Bah a de Tenacatita (i.e., seawaters) ( Figure 4 1 ). For each piezometer an d well sample, approximately three volumes of water were pumped prior to the collection of samples for chemical analyses. Samples were pumped through 0.45 m in line filters (Whatman, Maidstone, Kent, England) directly into pre cleaned, acid washed HDPE s ample bottles. Chloride samples were acidified with 1 ml of nitric acid, and chloride and sodium samples were stored at the method required range of 4 (2) C prior to analyses. Stable isotope sample bottles were filled completely with negligible head sp ace and sealed with Parafilm (American National Can, Chicago, Illinois, USA) to prevent the sample from equilibrating with ambient air.

PAGE 81

69 Temperature, pH, and salinity of the surface water and ground water samples were measured in the field with an YSI 556 M PS (YSI Inc., Yellow Springs, Ohio, USA). Sodium analysis was conducted by the University of South Florida Center for Water Analysis, with concentrations determined by atomic absorption spectrometry following EPA method 273.1 (Clesceri et al. 1998). Chlo ride analysis was conducted by Advanced Environmental Laboratories, Inc. in Tampa, Florida, with concentrations determined by the ferricyanide method following EPA method 325.2 (Clesceri et al. 1998). Based upon replicate analysis of 10% of the samples, a nalytical precision was typically better than 5 percent. All sodium and chloride concentrations are reported as ppm. Stable isotope analyses were conducted at the UC Davis Department of Geology Stable Isotope Laboratory. Deuterium analyses were performe d using the chromium reduction method (Donnelly et al. 2001), while oxygen 18 analyses were performed using the carbon dioxide equilibration technique (Epstein and Mayeda 1953). Deuterium and oxygen 18 are reported in the conventional, delta notation ( ): where R 18 16 O for deuterium and oxygen 18, respectively (Craig 1961). The resulting sample values of D and 18 O are reported in per mil convention the D and 18

PAGE 82

70 D and 18 O, respectively. The relative contributions o f fresh water and seawater to surface water samples taken within the mangrove were estimated using a series of two end, mass balance mixing models using salinity, sodium, chloride, D, and 18 O as conservative natural tracers. The mass balance mixing mode l was where C is salinity (psu), sodium or chloride concentration (ppm), or D or 18 O (per mil), and f represents the fraction of end member present. The subscripts FW SW and M refer to fresh water, seawater, and mangrove, respectively. Results Salinity The salinities of the mangrove waters varied between study sites and were generally intermediate between the fresh waters and seawaters (Table 1; Figure 4 2 ). Salinity is the ratio of the electrical conductivity of a given water sample to the electrical conductivity of a standard potassium chloride solution. Ratios have no units, but salini ty nevertheless is typically expressed in terms of practical salinity units (psu), and that convention is followed here. Mean SD salinities were 9.1 9.3 psu in the La Manzanilla mangrove (n = 50), 16.6 10.4 psu in the La Vena mangrove (n = 33), and 33.0 3.0 psu in the the Barra de Navidad mangrove (n = 19). Conversely, the salinities of the seawaters and

PAGE 83

71 Table 4 1 each wat er type sampled during the investigation. Standard deviations shown are 1 Water Type Salinity Na + (ppm) Cl (ppm) d d 18 Fresh Water 0.4 0.2 50 40 100 200 62 4 8.8 0.7 Seawater 34 2 8800 200 28500 600 1 1 0.1 0.1 La Manzanilla 9 9 3000 3000 7000 7000 40 40 5 5 La Vena 20 10 5000 3000 10000 7000 30 20 4 2 Barra de Navidad 33 3 9100 500 19000 2000 2 3 0.2 0.4 fresh waters varied little (Figure 2). Mean SD salinities were 0.4 0.2 psu for fresh waters (n = 25) and 33.5 1.7 psu for seawater (n = 10). Sodium and Chloride The sodium and chloride concentrations of the mangrove waters varied between study sites and were generally intermediate between the fresh waters and seawaters (Table 1; Figure 4 3 ). Mean SD sodium and chloride concentrations were, respectively, 2,770 2,850 ppm and 6,940 7,150 ppm in the La Manzanilla mangrove (n = 31), 4,700 3,210 ppm and 10,100 6,850 ppm in the La Vena mangrove (n = 15), and 9,040 500 ppm and 19,500 2,300 ppm in the the Barra de Navidad mangrove (n = 15). These were bracketed by the fresh waters and seawaters, with mean SD sodium and chloride concentrations being, respectively, 48 40 ppm and 123 156 ppm for fresh waters (n = 25) and 8,800 180 ppm and 20,500 580 ppm for seawater (n = 5). Linear regression of the sodium and chloride concentrations yields a local mixing line ( R 2 = 0.95; p<0.01).

PAGE 84

72 Figure 4 2 Average ground water and surface water salinity values for each wet and dry sampling event in each of the three mangroves and in the source waters Bars ar e the mean salinity for all sites sampled in that mangrove and are arranged in ascending date from left to right. Error bars are the standard deviations of those means No data were collected in the Barra de Navidad mangrove in January 2007. Stable Isotopes The D and 18 O of the mangrove waters varied between study sites and were generally intermediate between the fresh waters and seawaters (Table 1; Figure 4 4 ). Mean SD D and 18 O were, respectively, 39.33 and 5.35 in the La Manzanilla mangrove (n = 31), 16.99 and 2.43 in the La Vena mangrove (n = 15), and 0.42 in the the Barra de Navidad mangrove (n = 15). These were bracketed by the fresh waters and seawaters, with mean SD D and 18 O being 61.65 4.04 and 8.83 0.66 for fresh waters (n = 25) and 1.08 and

PAGE 85

73 Figure 4 3 Sodium and chloride concentrations (ppm) of fresh waters, seawaters, and mangrove waters collected from each of the three study sites during the cou rse of the study. The symbol shows the mean concentration and the error bars represent the standard deviations ) of those mean values. The dashed line is the local mixing line ( R 2 = 0.95; p<0.01) determined via linear regression of the sample concentr ations. 0.10 for seawaters Isotope data plot along a local mixing line ( R 2 = 0.99, p<0.01) anchored by the fresh waters and seawaters, with fresh waters plotting on the Global Meteoric Water Line (Craig 1961). Mass Balance Mixing The contribution of the fresh water end member was determined for each site using each of the five tracers: salinity, sodium concentration, chloride concentration, D, and 18 O ( Figure 4 5 ). The La Manzanilla mangrove had the highest fresh water contributions, with site averaged means ranging from 63 84%; and the La Vena mangrove had intermediate fresh water contributions, with

PAGE 86

74 Figure 4 4 D and 18 O ( ) of fresh waters, seawaters, and mangrove waters collected from each of the three study sites during the course of the study. The symbol represents the mean composition and the error bars represent the standard deviations ( 1 ) of those mean values. The solid black line is the global meteoric water line as defined by Craig (1961). The dashed line is the local mixing line ( R 2 = 0.99; p<0.01), determined via linear regression of the sample isotopic compositions. sit e averaged means ranging from 39 51%; and the Barra de Navidad mangrove had the lowest fresh water contributions, with site averaged means ranging from 0 5%. At each site, the salinity based mixing models yielded the highest mean estimates and the isotope based mixing models yielded the lowest mean estimates. Spatial Variability The mass balance mixing estimates for the La Manzanilla and La Vena mangrove surface waters show a large variability in fresh water contributions,

PAGE 87

75 while the Barra de Navidad mangro ve have a relatively narrow range ( Figure 4 6 ). In La Manzanilla mangrove, the largest fresh water contributions were to surface waters furthest from the ocean at th e toeslope of the mountainous terrain (~90%), while the largest seawater contributions were to surface waters closest to the beach ridge (~30 40%). In the La Vena mangrove, the largest fresh water contributions were to surface waters near the spring vent (~75 80%), while the largest seawater contributions were to surface waters closest to the mouth (~82 99%), with this large variation possibly being due to varying stream and tidal flows at different sampling times. In the Barra de Navidad mangrove, there was little fresh water contribution, with seawater contributions high throughout the mangrove (~99 100%). Discussion Fresh water contributions to mangroves exhibit both an intra and inter site variability, even when those sites are in close proximity to one another, as is the case with the La Manzanilla, La Vena, and Barra de Navidad mangroves ( Figure 4 5 and Figure 4 6 ). The La Manzanilla mangrove has a largest percentage of fresh water contribution; the La Vena mangrove has a slightly lower percentage of fresh water contribution, with amounts varying in relation to their posit ion on the river between the headwater spring and mouth leading to Bahia Tenacatita; and the Barra de Navidad mangrove, has a negligible percentage of fresh water contribution at any sampling location.

PAGE 88

76 Figure 4 5 Fresh water fraction of mangrove waters in each of the three study sites using each of the five tracers. Bars are the mean fresh water contribution for all sample locations Error bars are the standard deviations of those means The fresh water contributions to these mangroves can be understood in the context of hydrogeomorphic setting, as defined by Woodroffe (2002). The La Manzanilla mangrove is an interior mangrove, with a semi permanent beach ridge and only ephemeral connecti ons to the Bah a de Tenacatita. These ephemeral connections typically occur following low frequency/high magnitude storms, such as tropical cyclones, during which fresh water inflows to the mangrove accumulate and ultimately break through the beach ridge. During these connections, fresh water outflow dominates exchange between the mangrove and the Bah a de Tenacatita. For most of the year, and for all of the

PAGE 89

77 year in some years, the beach ridge remains intact and fresh water inflows, presumably from groun d water discharge, dominate mangrove water salinities. The La Vena mangrove is a river dominated mangrove, with ground water discharge from a headwater spring maintaining perennial flows and a perennial Figure 4 6 Fresh water fraction of mangrove waters at each individual sampling location. Fractions are means of fractions from all five tracers.

PAGE 90

78 connection to the Bah a de Tenacatita Fresh water dominates at the headwaters, seawater dominates at the mouth, and fresh water and seawater mix in the middle reaches of the river. The Barra de Navidad mangrove is a tide dominated mangrove, with fringing mangroves located on the margins o f a small, shallow lagoon perennially connected to the Bah a de Navidad. Though there are some inflowing streams and canals, i.e., the Rio Arroyo Seco, the Canal de Marabasco, and other smaller, unnamed streams, these have a combined watershed area of 3.9 km 2 so large surface water inflows are ephemeral, largely following low frequency/high magnitude storms such as tropical cyclones. The mouth is wide and deep, which allows a free exchange of seawater between the ocean and the mangrove so fresh water inflo ws are quickly mixed into the seawater and seawaters dominate mangrove water salinities. Therefore, hydrogeomorphic setting, as represented in the simple conceptual model most recently proposed by Woodruffe (2002), can be used as a tool to rapidly estimat e the importance of fresh water inflows to a given mangrove and therefore the potential susceptibility of a given mangrove to changes in fresh water inflows due to water resources development. Fresh water inflow to the mangrove can follow a few basic pathw ays: direct precipitation, surface water runoff, and ground water discharge. Ground water discharge almost certainly plays a large role in this region. Rainfall and runoff are limited, ephemeral, and focused in the wet season (July October). Despite the seasonality of the rainfall, the mangrove water salinities lack a clear

PAGE 91

79 seasonal trend ( Figure 4 2 ), meaning that there must be another consistent source of fresh water to the mangrove maintaining the mangrove water salinities. Mangroves are commonly found inset into high transmissivity coastal plain aquifers. The La Manzanilla mangrove, for example, is inset into a coastal plain with an unconfined aquifer with a s hallow water table and a transmissivity of 3,360 m 2 d 1 (M.C. Rains, unpublished data). Additionally, mangroves are located at the terrestrial marine interface where ground water is forced upward to ground water discharge points near the shoreline (Cady 1 941; McBride and Pfannkuch 1975; Harvey and Odum 1990; Rains et al. 2004). Both of these characteristics illustrate how geologic conditions favor ground water discharge into these mangrove environments. Additionally, ground water discharge can also be di rectly observed in these sites, such as the headwater spring in the La Vena mangrove and the many seeps and springs in the mountainous terrain surrounding the La Manzanilla mangrove. There are few perennial streams and surface water storage reservoirs in t he region. Therefore, ground water is an important source of water for irrigation and/or municipal supply. There is extensive heuristic evidence of the depletion of ground water resources in these small coastal plain aquifers. For example, salinity incr eased in two water supply wells in La Manzanilla between January 2007 and July 2008. One well is under municipal control, and ground water pumping appears to be causing salt water intrusion, with wet season salinity increasing from 1.1 in January 2007 to 2.2 in January 2008 and dry season salinity increasing from 2.5 in July 2007 to 2.9 in July 2008 (M.C. Rains,

PAGE 92

80 unpublished data). The other well is under private control and is lacking in salinity data, but ground water pumping nevertheless appears to be c ausing salt water intrusion, with the well having been abandoned and replaced with a shallower well as of January 2010. If ground water resources continue to be stressed by burgeoning populations, then heads in the coastal plain aquifers will be expected to decline potentially resulting in decreases in the ground water discharge to mangroves. This might lead to changes in salinity and related ecosystem structure and function in those mangroves in which fresh water inflows play an important role. Acknowle dgments This project was completed with the assistance of numerous people, including Jerry Keir, Sudeep Chandra, Zeb Hogan, Michael Marchetti, Kai Rains, Francisco de Ass Silva Btiz, Enrique Godnez Domnguez, Jason Bellino, Timothy Stringer, and many st udents on field courses from the University of South Florida, the University of Nevada Reno, and California State University Channel Islands. Partial funding was provided by the Patel Center for Global Solutions; the University of South Florida, Office of Undergraduate Research; and the National Science Foundation under Grant No. DBI 0620409. References Alongi DM (2002) Present state and future of the world's mangrove forests. Environmental Conservation 29:331 349

PAGE 93

81 Barbier EB, Evamaria WK, Silliman BR, Hacke r SD, Wolanski E, Primavera J, Granek EF, Polasky S, Aswani S, Cramer LA, Stoms DM, Kennedy CJ, Bael D, Kappel CV, Perillo GME, Reed DJ (2008) Coastal ecosystem based management with nonlinear ecological function and values. Science 319:321 323 Barlow PM ( 2004) Ground water in freshwater saltwater environments of the Atlantic Coast. U.S. Geological Survey Circular 1262. U.S. Government Printing Office. Washington, D.C., USA Barrera RO, Vargas FZ (2007) Geomorfologa del Estado de Jalisco. Universidad de Gua dalajara, Consultado el 30 de Octubre de 2007, de http://www.acude.udg.mx/jalisciencia/diagnostico/fisgeolsintes.pdf Brown C, Corcoran E, Herkenrath P, Thonell J (eds.) (2006) M arine and coastal ecosystems and human well being: a synthesis report based on the findings of the Millennium Ecosystem Assessment, UN Environment Programme, Nairobi, Kenya Cady RC (1941) Effect upon ground water levels of proposed surface water storage in Flathead Lake, Montana. U.S. Geological Survey Water Supply Paper 849 B. U.S. Government Printing Office, Washington, D.C., USA Cardona Olarte P, Twilley RR, Krauss KW, Rivera Monroy V (2006) Responses of neotropical mangrove seedlings grown in monoculture and mixed culture under treatments of hydroperiod and salinity. Hydrobiologia 569:325 341

PAGE 94

82 Clesceri LS, Greenberg AE, Eaton AD (1998) Standard methods for the examination of water and wastewater, twentieth edition. American Public Health Association, Washington, DC, USA CONABIO (2007) Opinin tcnica sobre el estado actual e impacto s en la Laguna Barra de Navidad, Cihuatln, Jalisco. Dra. Patricia Koleff, Directora de Anlisis y Prioridades/Comisin Nacional para el Conocimiento y Uso de la Biodiversidad. No. Oficio DTAP/255/2007. 25 de Septiembre de 2007. 6p. Craig J (1961) Standard for Reporting Concentrations of Deuterium and Oxygen 18 in Natural Waters. Science 133:1833 1834 Dawes CJ (1998) Marine botany. John Wiley and Sons, Inc., New York, New York, USA Donnelly T, Waldron S,Tait A, Dougans J, Bearhop S (2001) Hydrogen isotope a nalysis of natural abundance and deuterium enriched waters by reduction over chromium on line to a dynamic dual inlet isotope ratio mass spectrometer. Rapid Communications in Mass Spectrometry 15 :1297 1303 Drexler JZ, De Carlo EW (2002) Source water partit ioning as a means of characterizing hydrologic function in mangroves. Wetlands Ecology and Management 10:103 113 Epstein S, Mayeda T (1953) Variation of O 18 content of waters from natural sources. Geochemica et Cosmochimica Acta 4:213 224

PAGE 95

83 Everitt J, Massa m BH, Ch vez Dagostina RM, Espinosa Sanchez R, Andrade Romo E (2008) The imprints of tourism on Puerto Vallarta, Jalisco, Mexico. The Canadian Geographer 52:83 104 Gonfiantini R (1978) Standards for stable isotope measurements in natural compounds. Nature 271: 534 536 Gunawardena M, Rowan JS (2005) Economic valuation of a mangrove ecosystem threatened by shrimp aquaculture in Sri Lanka. Environmental Management 36:535 550 Harvey JW, Odum WE (1990) The influence of tidal marshes on upland groundwater discharg e to estuaries. Biogeochemistry 10:217 236 Islam MS, Haque M (2004) The mangrove based coastal and nearshore fisheries of Bangladesh: ecology, exploitation and management. Reviews in Fish Biology and Fisheries 14:153 180 Lugo, A.E., and S.C. Snedaker. 1974. The ecology of mangroves. Annual Review of Ecological Systems 5:39 64 Martinuzzi S, Gould WA, Lugo AE, Medina E (2009) Conversion and recovery of Puerto Rican mangroves: 200 years of change. Forest Ecology and Management 257:75 84 McBride MS, Pfannk uch HO (1975) The distribution of seepage within lakebeds. Journal of Research of the U.S. Geological Survey 3:505 512 McKee KL (1995) Interspecific variation in growth, biomass partitioning, and defensive characteristics of neotropical mangrove seedlings : response to light and nutrient availability. American Journal of Botany 82:299 307

PAGE 96

84 Mndez Linares, AP, Lpez Portillo J, Hernndez Santana JR, Ortiz Prez MA, Orozco OO (2007) The mangrove communities in the Arroyo Seco deltaic fan, Jalisco, Mexico, and their relation with the geomorphic and physical geographic zonation. Catena 70:127 142 Odum WE, McIvor CC (1990) Mangroves. In: Myers RL, Ewel JJ (eds). Ecosystems of Florida. University of Central Florida, Orlando, Florida, pp 517 548 Rains MC, Mount JF, Larsen EW (2004) Simulated changes in shallow groundwater and vegetation distributions under different reservoir operations scenarios. Ecological Applications 14:192 207 RAMSAR(2010) The RAMSAR List of Wetlands of International Importance. http://ramsar. wetlands.org/GISMaps/RamsarSitesinGoogleEarth/tabid/944 /language/en US/Default.aspx Accessed 1 Novem ber 2010. Rivera MonroyVH, Madden CJ, Day JW Jr, Twilley RR, Vera Herrera F, Alvarez Guilln H (1998) Seasonal coupling of a tropical mangrove forest an d an estuarine water column: Enhancement of aquatic primary productivity. Hydrobiologia 379:41 53 Solomon S, Qin D, Manning M, Chen Z, Marquis M, Averyt KB, Tignor M, and Miller HL (eds.) (2007) The physical science basis: contribution of Working Group I to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change. Cambridge University Press, Cambridge, U.K.

PAGE 97

85 SMN (2010) Servicio Meteorolgico Nacional. Comisin Nacional del Agua. http://smn2.cna.gob.mx/SMN2/Default.aspx Accessed 1 November 2010. Stringer CE, Rains NC, Kruse S, Whigham D (2010) Controls on water levels and salinity in a barrier island mangrove, Indian River Lagoon, Florida. Wetlands 30:725 734 Su rez N, Medina E (2005) Salinity effect on plant growth and leaf demogra phy of the mangrove, Avicennia germinans L. Trees 19:721 727 Su rez N, Medina E (2006) Influence of salinity on Na + and K + accumulation, and gas exchange in Avicennia germinans Photosynthetica 44:268 274 Twilley RR, Chen R (1998) A water budget and hydrol ogy model of a basin mangrove forest in Rookery Bay, Florida. Marine Freshwater Research 49:309 323 threatened major tropical environments. BioScience 51:807 815 Valiela I, Kinney E, Culbertson J, Peacock E, Smith S (2009) Global losses of mangroves and salt marshes In: Duarte CM (ed.). Global loss of coastal habitats: rates, causes, and consequences. Fundacion BBVA, Bilbao, Spain. Vijay V, Biradar RS, Inamdar AB, Deshmukhe G, Baji S, Pikle M (2005) Mangrove mapping and change detection around Mumbai (Bombay) using remotely sensed data. Indian Journal of Marine Sciences 34:310 315 Wilson, TD (2008) Economic and social impacts of tourism in Mexico. Latin American Perspectives 35:37 52

PAGE 98

86 Wo odroffe CD (2002) Coasts: form, process, evolution. Cambridge University Press, Cambridge, U.K.


xml version 1.0 encoding UTF-8 standalone no
record xmlns http:www.loc.govMARC21slim
leader nam 22 Ka 4500
controlfield tag 007 cr-bnu---uuuuu
008 s2010 flu s 000 0 eng d
datafield ind1 8 ind2 024
subfield code a E14-SFE0004760
035
(OCoLC)
040
FHM
c FHM
049
FHMM
090
XX9999 (Online)
1 100
Stringer, Christina.
0 245
Hydrologic controls on salinity in mangroves and lagoons
h [electronic resource] /
by Christina Stringer.
260
[Tampa, Fla] :
b University of South Florida,
2010.
500
Title from PDF of title page.
Document formatted into pages; contains X pages.
502
Dissertation (PHD)--University of South Florida, 2010.
504
Includes bibliographical references.
516
Text (Electronic thesis) in PDF format.
538
Mode of access: World Wide Web.
System requirements: World Wide Web browser and PDF reader.
3 520
ABSTRACT: This dissertation explores the hydrologic controls on salinity within mangroves and lagoons at sites in Florida and Mexico. The main objective of this research is to better understand hydrologic controls on mangrove ecosystem structure and develop ideas that will be useful to land managers attempting to regulate and conserve these critical habitats. This study was conducted at sites in Ft. Pierce, FL and Costalegre on the central Pacific coast of Mexico. We examined controls on water levels and salinity in a mangrove on a carbonate barrier island along the Indian River Lagoon, east-central Florida. Spectral analysis of water levels showed that mangrove groundwater levels are not tidally influenced. Salinities vary spatially, with values of approximately 10 in upland environments to approximately75 psu in irregularly-flushed mangroves. Water chemistry indicates that water salinities are largely controlled by enrichment due to evapotranspiration. An electrical resistivity survey showed that the freshwater lens is restricted to uplands and that hypersaline waters extend deeply below the mangrove. These results indicate that evapotranspiration lowers water levels in the mangrove, which causes Indian River Lagoon water to flow into the mangrove where it evapoconcentrates and descends, forming a thick layer of high-salinity water below the mangrove. Spatial variability of terrain conductivity in the Ft. Pierce mangrove varied under two hydrologic management regimes, breached rotational impoundment management and rotational impoundment management. The difference in coefficient of variation (DeltaCV) between the breached RIM and RIM data was calculated to examine spatial variability in both the shallow and deep layers. A null-hypothesis model was employed to examine the statistical significance of the Delta CV results. The average water levels were -0.06 m amsl and 0.49 m amsl during the breached-RIM and RIM regimes, respectively. The average shallow (EM31) layer terrain conductivity shifted slightly from 1868 mS m-1 to 1825 mS m-1 after the alteration in management regime, yet the standard deviation of these averages decrease from 656 mS m-1 to 216 mS m-1. The average deep (EM34) layer terrain conductivities were 328 mS m-1 and 255 mS m-1 during the breached-RIM and RIM regimes, respectively. The temporal Delta CVs were 0.23 and -0.04 for the shallow and deep layers, respectively. The null-hypothesis model for the shallow layer illustrates that the difference in spatial structure is statistically significant. The deep layer DeltaCV was not statistically significant. These results indicate that the transition from breached RIM to RIM resulted in changes to both the physical and chemical hydrologic character of the impoundment, especially in the shallow layer. The second study sites were three mangrove communities along the central Pacific Mexican coast. Salinities varied by water type, with values of approximately 9 in La Manzanilla, 17 in La Vena, 33 in Barra de Navidad, 0.4 in the fresh waters, and 34 in the seawater. Sodium and Chloride concentrations and isotopic signatures, as well as salinity, were used as tracers in mass-balance mixing models to quantify estimates of relative fresh-water and seawater contributions to each site. La Manzanilla, a basin mangrove, had mean fresh-water contribution estimates of 63-84 percent. La Vena, a riverine mangrove, had fresh-water estimates of 39-51 percent. Barra de Navidad, a fringe mangrove, had low fresh-water contributions of 0-5 percent. These results illustrates that the role groundwater plays in mangrove hydrodynamics is dependent on the site hydrogeomorphology.
590
Advisor: Mark Cable Rains, Ph.D.
653
Wetlands
Hydrology
Water sources
Conductivity
Evapotranspiration
Electromagnetic survey
Stable isotopes
Dissolved ions
Mass-balance mixing
Florida
Mexico
690
Dissertations, Academic
z USF
x Geology
Masters.
773
t USF Electronic Theses and Dissertations.
4 856
u http://digital.lib.usf.edu/?e14.4760