Seasonal and anthropogenic changes in the fresh/brackish water lens of the Torchwood Preserve, Little Torch Key, Florida

Citation
Seasonal and anthropogenic changes in the fresh/brackish water lens of the Torchwood Preserve, Little Torch Key, Florida

Material Information

Title:
Seasonal and anthropogenic changes in the fresh/brackish water lens of the Torchwood Preserve, Little Torch Key, Florida
Creator:
Meadows, Darren G.
Place of Publication:
Tampa, Florida
Publisher:
University of South Florida
Publication Date:
Language:
English
Physical Description:
vi, 68 leaves : col. ill., maps (some col.) ; 29 cm.

Subjects

Subjects / Keywords:
Groundwater flow -- Florida -- Little Torch Key ( lcsh )
Hydrogeology -- Florida -- Little Torch Key ( lcsh )
Dissertations, Academic -- Geology -- Masters -- USF ( FTS )

Notes

General Note:
Thesis (M.S.)--University of South Florida, 2001. Includes bibliographical references (leaves 67-68).

Record Information

Source Institution:
University of South Florida
Holding Location:
Universtity of South Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
028270083 ( ALEPH )
48748242 ( OCLC )
F51-00159 ( USFLDC DOI )
f51.159 ( USFLDC Handle )

Postcard Information

Format:
Book

Downloads

This item is only available as the following downloads:


Full Text

PAGE 1

SEASONAL AND ANTHROPOGENIC CHANGES IN THE FRESH/BRACKISH WATER LENS OF THE TORCHWOOD PRESERVE, LITTLE TORCH KEY, FLORIDA by DARREN G. MEADOWS v A thesis submitted in parti a l fulfillment of the requirements for the degree of Master of Science Department of Geology College of Arts and Sciences University of South Florida May 2001 Major Pro fessor: Sarah E. Kruse, Ph.D

PAGE 2

Examining Committee : Office of Graduate Studies University of South Florida Tampa, Florida CERTIFICATE OF APPROVAL This is to certify that the thesis of DARREN MEADOWS in the graduate degree program of Geo lo gy was approved on April 16 2001 for the Master of Science degree Major Professor: Sarah E. Kruse, Ph D. Leonard Vacher Ph.D. Member: Eric A. Oches, Ph.D.

PAGE 3

Acknowledgements I would first like to thank my mother and father for providing a life-long example of hard work and integrity Brian Brad and Chris, m y brother s, also deserve a sincere thank you for keeping me sane during this process. I am indebted to Sarah Kruse for her patience and excellent advice throughout this thesis. Thanks to Len Vacher for his useful comments as well as for sparking my interest in hydrogeology years ago. Rick Oches deserves an enormous thank you for constantly offering moral support and professional guidance Thank you to Mark Stewart whose door was alway s open so that I could discuss model problems with him. This thesis b e nefited greatly from conversations with Jim Schneider. For help in the field I wou ld like to thank Jame s C ulb e rt, Jim Schneider Brian Meadows and Lisa Wood. Thank s to Chris Bergh and Alison Higgins at The Nature Conservancy for their technical sup port. F inally thank s to The Nature Conservancy for fundin g this project.

PAGE 4

Table of Contents List of Figures lll Abstract Title Page VI Abstract Chapter 1: Introduction 3 Regional Geology 7 Local Geology and Vegetation 8 Island Hydrogeology 11 Chapter 2: Methods 13 EM 13 Water Conductivity Measurements 14 Resistivity 14 Ditch Water Salinity 16 Historical Aerial Photographs 1 7 Chapter 3: Results 18 Seasonal Variability in the Fresh/Brackish Water Lens 18 Wet Season Freshwater Volume 33 The Impact ofPlugging the Mosquito Ditch 36 EM data 36 Resistivity data 42 Salinity of ditch water 42 Qualitative interpretation of aerial photographs 43 Chapter 4: Discussion and Models 45 Interpreting the Absence of a High Conductivity Zone near the Ditch 45 Case 1 50 Case 2 50 1-D Analytical Model of the Impact of a Ditch 52 Numerical Model 54 Wet season 55 Dry season 61 Calculation of recharge 61

PAGE 5

Chapter 5: Conclusions References II 65 67

PAGE 6

List of Figures Figure 1 Little Torch Key location. 4 Figure 2 Aerial photograph showing location of study site 5 Figure 3 Approximate locations of natural communitites. 9 Figure 4 Circulation of freshwater and seawater on the margin of an oceanic island. 12 Figure 5 -Locations ofterrain conductivity surve y s including EM transects and a resistivity survey line, mosquito ditches and wells. 15 Figure 6 Sites of data collection. 19 Figure 7-Color contour maps showing terrain conductivity (mS / m) at the (A) 10m coil spacing and the {B) 20m coil s pacing on 2 /12/ 00 20 Figure 8-Conductivity versus depth for (A) well A and (B) well Bon 2 /12/ 00 21 Figur e 9 Color contour maps showing terrain conductivity (mS / m) at the (A) 10m coil spacing and the (B) 20m coil spacing on 4/011/ 00. 22 Figure 10Conductivity versus depth for (A) well A and (B) well Bon 4 /01100. 23 Figure 11 Color contour maps showing terrain conducti v ity (mS /m) at the (A) 10m coil spacing and the (B) 20 111 coil spacing on 4 /31//00. 24 Figure 12-Conductivity versus depth for (A) well A and (B) w ell B on 4 / 31/00. 25 Figure 13-Color contour maps showing terrain conductivity (mS /m) at the (A) 10m coil spacing and the (B) 20111 coil spacing on 7 / 10 // 00. 26 Figure 14-Conductivity versus depth for (A) well A and (B) well Bon 7/10 / 00. 27 Figur e 15Color contour maps showing terrain conductivity (mS / m) at the (A) 10m coil spacing and the (B) 20111 coil s pacing on 10/ 06//00. 2 8 Ill

PAGE 7

Figure 16 Conductivity versus depth for (A) well A and (B) well Bon 10 / 06/00. 29 Figure 17-Monthly rainfall at Key West, FL, 1999 and 2000. 31 Figure 18 Difference between 4 /01 (dry season) and 1 0/06 (wet season) raw EM data (mS/m) at 10 m and 20 m coil spacings 32 Figure 19-Conductivity versus depth curve for well A on 9 / 27 / 99 34 Figure 20 Conductivity versus depth curve for well Bon 9/27 / 99. 34 Figure 21 -Conductivity versus depth curves for (A) well A and (B) well Bon 4 / 17 / 99. 35 Figure 22-Resistivity section on 2 /12/ 00. 37 Figure 23 -Resistivity section on 4 /01/ 00. 38 Figure 24-Resistivity section on 4 /31/ 00. 39 Figure 25 Resistivity section on 7/10/00. 40 Figure 26 Resistivity section on 10 / 06/00. 41 Figure 27-Rectified, color-flooded aeria l photographs ofTorchwood Preserve, 1959, 1963, 1974, 1986 44 Figure 28 Apparent conductivity versus distance west of ditch for middle EM transect. 4 7 Figure 29 Apparent conductivity versus distance wes t of ditch for north EM transect. 48 Figure 30 Apparent conductivity versus distanc e west of ditch for south EM transect. 49 Figur e 31 Salinity of water in primary ditch versus distance from southern shore.51 Figure 32 Figure showing zo nes of recharge in numerical model. 56 Figure 33 Cross-section along column 32 of numerical model calibrated to wet season data with disp e rsivi ty. 57 IV

PAGE 8

Figure 34 Color flood contour maps of concentration in ppt from layers 1 4, 8 and 12 ofnumerical model with dispersivity calibrated to wet season data. 58 Figure 35 Cross-section along column 32 for numerical model without dispersivity calibrated to wet season data. 59 Figure 36 Color flood contour maps of concentration in ppt from layers 1, 4, 8, and 12 of numerical model with no dispersivity calibrated to wet season data 60 Figure 37-Cross-section along column 32 for dry season numerical model. 62 Figure 38 Color flood contour maps of concentration in ppt from layers 1, 4, 8, and 12 of numerical model for dry season. 63 v

PAGE 9

SEASONAL AND ANTHROPOGENIC CHANGES IN THE FRESHIBRACKJSH WATER LENS OF THE TORCHWOOD PRESERVE, LITTLE TORCH KEY, FLORIDA by DARREN G. MEADOWS An Abstract of a thesis submitted in partial fulfillment of the requirements for the degree of Master of Science Department of Geology Colle ge of Arts and Sciences University of South Florida May 2001 Major Professor : Sarah E Kruse, Ph.D. vi

PAGE 10

Abstract Approximately 40 years ago a network of mosquito control ditches was excavated to enhance surface water flow on Little Torch Key Florida. The volume of freshwater on the island has implications for the restoration of an indigenous plant species, the endangered Semaphore cactus (Opuntia spinosissima). In an attempt to restore the natural hydrology of the Torchwood Preserve on the southern portion of the island, The Nature Conservancy plugged the principal N-S-trending ditch that is directly connected to the ocean in May 2000. In this study resistivity and electromagnetic surveys and groundwater conductivity measurements are used to determine: (1) the extent of alteration of fresh/brackish water lens following the plugging of the ditch and (2) the natural seasonal variability in the fresh/brackish water lenses on the Torchwood Preserve. On the Torchwood Preserve changes in terrain conductivity with time correspond approximately to changes in groundwater salinity--higher terrain conductivities are associated with higher salinity Electromagnetic (EM) data show that the surficial fresh/brackish water lens thins and increases in salinity outward from the center of the island, as expected. Resistivity surveys across the primary ditch both before and after plugging show no clear variation in terrain conduc tivity as a function of distance within 35 meters on either side of the ditch EM surveys show no apparent change in the overall pattern of conductivities associated with the plugging of the ditch. These results suggest that in the months surveyed before plugging the ditch had little impact on the island len s. 1

PAGE 11

Both EM data and direct groundwater conductivity measurements suggest that seasonal variations in the extent of the fresh/brackish water len s are significant. The hammock supports a region of lower conductivity (fresher groundwater) throughout the year. Although the lateral dimensions ofthis zone do not change substantially, the lens does become less conductive during the wet season with the largest volume of freshwater occurring at the end of the wet season. Water throughout the hammock is brackish during the dry season. In the central hammock at the entrance to the preserve salinity increases with depth to seawater values below 7 m depth. Abstract Approved: _____ ____ Major Professor: Sarah E. Kruse, Ph.D. Professor, Department of Geology Date Approved: __ ...L..--L----2

PAGE 12

Chapter 1: Introduction This thesis is based on geophysical surveys and measurements of water conductivity on The Nature Conservancy's Torchwood Preserve, Little Torch Key, Florida (Figures 1 and 2). Approximately 40 years ago a network of mosquito control ditches was excavated to enhance surface water flow to reduce the mosquito population. The ditches may allow saltwater intrusion and may expedite runoff and discharge from the key and thereby reduce infiltration and fresh/brackish water lens thickness. Possible subsequent diminution in the volume of freshwater has implications for the restoration of an indigenous plant species, the endangered Semaphore cactus (Opuntia spinosissima) Toward the goal of restoring the natural hydrology of the preserve The Nature Conservancy plugged the principal N S-trending ditch that is directly cmmected to the ocean. The goals ofthis thesis are to (1) assess the impact of plugging the ditch on the hydrology of the island and (2) image the natural seasonal variability in groundwater salinity. As groundwater salinity affects the conductivity of the ground, measuring the terrain (ground) conductivity yields a picture of the fresh/brackish water lens underlying the preserve. Repeated surveys were used to d e termine the impact of the ditch infilling as well as the seasonal variability in the lens. Analysis of geophysical surveys is supplemented by measurements of water conductivity at two wells, interpretation of historical vegetation patterns on aerial photographs and the development of a 3

PAGE 13

26"00' 81"30' ed'aO' Bahia STUDY ARI!.A-fSir o lOMILES Figure 1 -Little Torch Key location. Study site is on southern tip of the islannd. 4

PAGE 14

Figure 2Aerial photograph showing location of study area. 5

PAGE 15

groundwater flow model for the island lens. With t his information, preserve stewards can better estimate the potential impact of future ditch alterations. 6

PAGE 16

Regional Geology The Florida Keys form an arcuate archipelago composed of Pleistocene limestone which extends from Soldier Key (15 km southeast of Miami) 240 km southwest to Key West. The chain of islands separates the Straits of Florida from Florida Bay (Figure 1). The Keys can be divided into the Upper Keys, which extend northeastward from Bahia Honda, and the Lower Keys, which extend from Big Pine Key westward to Key West. The linear, northeast-southwest-trending Upper Keys are composed primarily of the Key Largo Limestone reefal unit. At Big Pine Key the islands change to a more east-west orientation. Big Pine Key is also the site of surface contact between the Key Largo Limestone and the Miami Limestone, an oolitic facies that overlies the Key Largo unit in the Lower Keys (Hoffmeister and Multer, 1964 ) The Key Largo Limestone stretches from Miami to at least the Dry Tortugas The limestone is composed of hermatypic corals with intraand interbedded calcarenites (Hoffmeister and Multer, 1964) In Dunham's (1962) terms, the Key Largo Limestone consists of a peloid-bioclast packstone-grainstone. Its thickness is quite variable but was found to exceed 60 m in test borings on Big Pine Key (Hoffmeister and Multer 1968). The average effective porosity ofThe Key Largo Limestone is 15% but can reach as high as 40% (Coniglio and Harrison, 1983). Also, pore interconnectedness is overall very good and due to the limestone's substantial skeletal component, it is very prone to the development of secondary solution cavities (Coniglio and Harrison 1983). The hydraulic conductivity of the Key Largo Limestone beneath The Miami Limestone was estimated to be 1200-1600 m/d (Wightman, 1990) 7

PAGE 17

The Miami Limestone overlying the Key Largo Limestone is composed of two members : an upper ooid-grainstone and a lower ooid-grainstone-packstone (Hoffmeister and Multer, 1968). The oolitic facies consists ofwell-sorted ooids with variable amounts of skeletal material and some quartz sand. Hoffmeister et al. ( 1967) and Perkins ( 1977) believed this oolitic facies was deposited as a marine ooid-shoal bank. The oolitic facies ranges in thickness from 3 to 5 m reach i ng its maximum along the seaward edge of the Lower Keys. This facies displays extreme l a teral and vertical variability in cementation and grain size (Coniglio and Harrison, 1983) The degree of pore interconnectedness is relatively poor, resulting in a lower permeability in the Miami Limestone than the Key Largo Limestone (Coniglio and Harrison, 1983) The hydraulic conductivity ofthe Miami Limestone was estimated by Wightman (1990) to be 100-140 m/d Local G e ology and Vegetation The Nature Conservancy's Torchwood Preserve is the southern part of Little Torch Key (Figure 2). The island lies 2 miles west of Big Pine Key and approximately 26 miles east ofKey West. It therefore exhibits the geologic character of the Lower Keys-the less permeable Miami oolite facies overlies the more permeable basement of Key Largo Limestone. The Torchwood Preserve is a 244-acre tract consisting of five natural communities, four of which are in the study area : mangrove fringe, coastal rock barren coastal berm and rockland hammock (Figure 3) The following summary of the Preserve's vegetation communities is compiled from The Torchwood Hammock Preserve Field Guide" (The Nature Conservancy 8

PAGE 18

2727100.00 2727000.00 2726900.00 Cii 2726800.00 '-Q) 2726700.00 :c h:: 2726600.00 0 :z 2726500 1-:::> Torchwood Preserve t t Tidal Swamp + Newfound Harbor Rockland Hammock I I Coastal Rock Barren Coastal Berm -Mangrove Fringe UTM East (meters) Om 100m 200m 300m 400m Figure 3 Approximate locations of natural communities. 9

PAGE 19

(TNC) undated). The rockland hammock consists of a wide variety of salt-intolerant vegetation surrounded by marine wetlands. The hammock is essentially a tree island that forms above a lens of freshwater. The largest trees create a dense canopy that shades the ground, moderating soil moisture loss and deflecting wind thereby preventing uprooting during major storm events. The thick overstory also provides massive amounts of leaf litter. The decaying organic matter affords a thick layer of nutrient-rich humus. The spongy layer ofhurnus holds moisture near the roots of the plants making water available during drier winter months (TNC undated) Surrounding the rockland hammock is the coastal rock barren. The canopy is much more open in this community. The groundcover, which is nearly absent in the rockland hammock due to dense canopy cover, is dominated by halophytic grasses in the rock barren. Vegetation changes abruptly to much more salt-tolerant species as one enters this community. The transition zone between hammock and barren is frequently well defined along an elevational gradient. This zone of transition is also the habitat for the endangered Semaphore cactus Although very rare statewide coastal rock barren is one of the most abundant communities on the preserve (TNC undated). Landward of the shore along the southwestern coast of the preserve, a coastal berm can be seen A berm is a low ridge of shell and sand that forms parallel to the shoreline. These ridges form from storm action. The coastal berm on the preserve is up to 9 m across and lm above sea level (TNC undated) Several hammock species are present in this community due to the higher elevation Bordering the coast is the mangrove fringe community Because mangroves are poor competitors, they are typically found only where other plants cannot exist. Soft 10

PAGE 20

substrate, constant tidal fluctuations, saline conditions and anaerobic substrate make this environment inhospitable to nearly all other forms of vegetation. Mangroves are very important to coastal regions as they anchor sediment and dissipate wave energy Of the natural communities on the preserve, the coastal rock barren would be most affected by functioning mosquito ditches. Although salt-tolerant, vegetation in the barren is inundated with seawater only during atypically high tides Ditches channel saltwater into this region. This process is evidenced by the presence of red mangroves in and around the ditches. Island Hydrogeology Due to the paucity of freshwater on many smaller carbonate islands, drinking water has often been imported Hence, the hydrogeology of brackish water lenses has been the subject of relatively little research. Studies include: Vacher (1978) Hanson (1980), Budd (1984), Wightman (1990), Vacher and Quinn (1997) Freshwater lenses that develop on most islands are of meteoric origin. The density differential between the relatively small volume of fresh groundwater and seawater causes the less saline water to effectively float on top of the more saline. The idealized view of a distinct interface separating the two, however, is quite often not the case. Instead, there is a brackish transition zone of variable thickness where seawater grades to freshwater (Figure 4) The high values of hydraulic conductivity typical of carbonate islands tend to create a thick zone of gradation brought about principally by efficient tidal mixing. The location and size of freshwater lenses are controlled by the i s land s elevation above sea level, amount of recharge, properties of the substrate and e x tent of tidal influence (Vacher 1988) 11

PAGE 21

Water Table Figure 4 Circ ul atio n of freshwater and sal ine gro u ndwate r on the m argin of a n oceanic is l an d 12

PAGE 22

Chapter 2: Methods This investigation used electromagnetic (EM) surveys coupled with electrical resistivity profiles to determine season a l ch a n g es and the changes if any in the groundwater salinity patterns due to the fillin g o f the mosquito control ditch es. Stewart ( 1982) verified the effectiveness of th i s a pproach i n delineatin g z on e s of freshwater and s eawater Both geophysical methods me as ure the conduct iv ity of the ground In the relatively uniform, permeable oolitic lim e ston e of Little Torch Ke y, terrain conductivity is assumed to increase with increasing conductivity of the groundwater. Saline water, due to its higher concentration of ions, is much more conductive than freshwater. Thus the presence of brackish water or saltwater can be detected EM Terrain conductivity measurem e nt s were taken with a Geonic s Li m i ted E M 34 terrain conductivity meter operated in hori z ont a l dipole mode (McNeill 1981 ) The instrument consists of a transmitter coil a receiver coil, and control modules An alternating current is driven in the tran s mitter coil. This alternating curr e nt produces a primary magnetic field that propagates throu g h the subsurface. The prim a ry magnetic field creates secondary currents in th e g round which then g enerate secondary magnetic fields that are detected by the rec e i v e r coil. The E M-34 can be operated with coil spacin gs of 10 20, or 40 m In the horizontal dipole mode used here the depth resolution 1 3

PAGE 23

is approximately three-fourths of the coil spacing. The instrument, though, is most sensitive to shallower depths. To determine seasonal variations in the location of fresh and brackish water and any possible large-scale effects of the mosquito ditches on lens configuration terrain conductivity measurements were taken along 5 transects spanning the southern portion of the preserve (Figure 5) at the 10and 20-m coil spacings. Water Conductivity Measurements To calibrate terrain conductiv i ty values, groundwater conductivity measurements were taken at two 2-inch wells (Figure 5, wells A and B) with a YSI 30 conductivity meter. Readings were taken every 0.15 m by slowly lowering the conductivity probe down the well. Because of the high permeability of the surrounding limestone and the small well diameter, the salinity-versus-depth curve of the well water is assumed to be nearly the same as that of the surrounding groundwater. R e sistivity Repeated direct current resistivity soundings were collected on a profile crossing the mosquito ditch that was plugged in May 2000 (Figure 5). This 75 m-long profile traverses the main ditch approximately 30m landward (north) of the plug. The survey on this profile was configured as a Wenner traverse with an electrode spacing of 1 5 m. Each resistivity measurement utili z es four electrodes, which must be hammered into the ground. Direct current is driven into the g round f rom the first electrode (source) to the fourth (sink) The electric potential difference that results is then measured between two 14

PAGE 24

J: u IX 0 F- '---... e e ditches resistivity Line EM transects Plug Well A WellB Figure 5 of terrain conductivity surve ys, including EM transects and a resistivity survey line mosquito ditches, and wells 15

PAGE 25

electrodes that are set between the current source and current sink electrodes. From the potential difference, the apparent resistivity of the ground can be determined. As the distance between the electrodes increases the current travels to greater depths allowing more of the subsurface to be sampled By sequentially increasing the distance between the sink and source electrodes, a range of depths can be assessed. We used the Geopulse system by Campus Geophysical Instruments, Ltd. for all resistivity surveys For the profile crossing the ditch, we used all possible combinations of the 50 equally spaced electrodes to compute apparent resistivities along the profile These readings were then processed using the Res2dinv code (Loke, 1996) to compute a 2-D profile of resistivity as a function of distance along profile and depth. As resistivity is just the inverse of conductivity, this profile is also an image of ground conductivity Although the resistivity method yields better depth resolution than the EM-34, the invasive nature and labor intensiveness of resistivity soundings preclude their extensive use. Data from this study combined with those from a previous study (Kruse et al., 1999) were used to estimate the seasonal variability in the presence and location of fresh/brackish water lens. Ditch Water Salinity Measurements of salinity of the water in the ditch were collected monthly throughout 2000 by The Nature Conservancy u si ng a refractometer. 16

PAGE 26

Historical Aerial Photographs Rectified historical aerial photographs from 1 959 to 1986 were qualitatively examined to reveal any changes in vegetative patterns that may have been induced by the ditches. 17

PAGE 27

Chapter 3: Results Seasonal Variability in the Fresh/Brackish Water Lens As part of this study, EM, resistivity, and water conductivity data were collected on five weekends in 2000: Trip 12 / 12/00-2 / 14 / 00, dry season Trip 2 4 /01/ 00-4 / 03 / 00, dry season Trip 3 4/31/00-5 / 02/00, dry season Trip 4-7/10/00-7/12/00, wet season Trip 5 -1 0/06/00-10/08/00, wet season. In our discussion here we combine these data with the earlier data in Kruse et al. (1999). The extremely conductive environment on Little Torch Key allowed for reliable resolution with the EM instrument at only I 0and 20-m coil spacings. Instrument calibration problems in the 1999 and 2000 data were suspected upon inspection of the raw conductivity data. Therefore, each data set was normalized by fixing the point closest to the coast where conductivity values vary least the most distal point on the EW 2 EM transect (Figure 6), to the value obtained at that point on the 2/12/00 trip A constant value was added or subtracted from all the readings on a given survey. Color contour maps of normalized terrain conductivity readings made with both the I 0and 20m coil spacings (Figures 7, 9, 11, 13, 15) delimit a central zone of low conductivity in the center of the island in both wet and dry seasons. The central z one corresponds to the 18

PAGE 28

240 230 220 210 200 190 E-< ;:J 180 170 160 150 140 130 120 110 100 90 460500 460900 461000 80 Figure 6-Sites of data collection Crosses inland show locations of EM readings Crosses on coastline show points set to a value representative of EM reading on coast for the purpose of generating a more rea l istic contour map. EM data normalized so that the reading at the point shown is constant throughout the year 1 9

PAGE 29

240 230 220 210 200 190 180 170 160 150 140 130 120 110 100 90 460700 460800 461000 UTM (E) 80 460500 460500 460600 460800 460900 461000 UTM (E) Figure 7-Color contour maps showing terram conductivity (mS/m) at the (A) 10m coil spacing and the (B) 20m coil spacing on 2112/00. 20

PAGE 30

(A) 0 groundwater conductivity (mS / cm) lO 20 30 40 50 60 0.00 1.00 ,.-... 5 2 00 11) (..) ;3 3.00 Cll 0 ] 4.00 ..s:: ....... 0.. 5.00 6 00 freshwater brackish water seawater 7.00 (B) groundwater conductivity (mS / cm) 0 10 20 30 40 50 60 0 00 1.00 s '-" 2.00 11) (..) ;:::l 3 00 Cll 0 ] 4.00 t 5 00 6.00 fre hwater brackish water 7 .00 Figure 8-Conductivity versus depth curves for (A) well A and (B) well Bon 2 /12/ 00 21

PAGE 31

460500 460700 460800 460900 461000 UTM (E) 460500 460800 461000 UTM (E) Figure 9 -Color contour maps showing terrain conductivity (mS/m) at the (A) 10m coil spacing and the (B) 20m coil spacing on 4 / 01 /00. 2 2 240 230 220 210 200 190 180 170 160 150 140 130 120 110 100 90 80

PAGE 32

0 00 1.00 s 2.00 (.) ] r:ll 3.00 0 4 00 ...t:: ..... 0.. Q) '"0 5.00 6.00 (A) 0 groundwater conductivity (mS/cm) 10 20 30 40 50 60 fre hwater brackish water seawater 7.00 (B) 0 groundwater conductivity (mS /cm) 10 20 3 0 40 50 60 0.00 1.00 s ...__, 2.00 Q) (.) ;:I 3.00 r:ll 0 Q) 4.00 .0 o:S 0.. .g 5.00 6.00 7.00 :fre hwater brackish water seawa er Figure 10Conductivity versus depth curves for (A) well A and (B) well Bon 4 /01/00 23

PAGE 33

g ::s 460500 460800 460900 UTM(E) 460500 4ffi600 460800 461000 Figure 11 -Color contour maps conductivity (mS / m) at the (A) 10m coil spacing and the (B) 20m coil spacing on 4 /31/ 00 2 4 240 230 220 210 200 190 180 170 160 150 140 130 120 110 100 90 80

PAGE 34

(A) groundwater conductivity (mS / cm) 0 10 20 30 40 50 60 0.00 r---------------------, e 1.00 '--" Q) u 2.00 ::l (/) 3.00 0 -Q) ..0 t 4 .00 Q) '"0 5.00 6.00 freshwater brackish water seawater 7.00 groundwater conductivity (mS / cm) w 60 brackish water seawater Figure 12-Conductivity versus depth curves for (A) well A and (B) well Bon 4 / 31/00. 25

PAGE 35

240 230 220 210 200 190 180 170 160 150 140 130 120 110 100 90 460500 460900 80 (B) 26

PAGE 36

(A) groundwater conducti v ity (mS /cm) 0 10 20 30 40 50 60 0 .00 ,...-._ 1.00 s '-..-/ Q) u ] 2.00 Cl) 3.00 0 -Q) ..0 ..c ..... 0. 4 .00 Q) "'0 5 .00 6 .00 freshwater brackish water 7 .00 (B) groundwater conductivity (mS /cm) 0 10 20 30 40 50 60 0 .00 ,...-._ 1.00 s '-..-/ Q) u 2 .00 Cl) 3.00 0 -Q) ..0 ..c ..... 4.00 0. Q) "'0 5 .00 6 .00 freshwater brackish water seawater 7 .00 Figure 14-Conductivity versus depth curves for (A) well A and (B) well Bon 7 / 10 / 00. 27

PAGE 37

460500 460600 460700 460800 460900 461000 UTM (E) 460500 460600 460700 460800 460900 461000 UTM (E) Figure 15-Color contour maps showing terrain conductivity (mS / m) at the (A) 10m coil spacing and the (B) 20m coil spacing on 10/ 06 / 00 28 240 230 220 210 200 190 180 170 160 150 140 130 120 110 100 90 80

PAGE 38

0.00 1.00 s '-' 8 2.00 ;:l en 3.00 0 ..0 4 00
PAGE 39

higher-elevation hardwood hammock. Ground conductivities increase radially from the hammock to the coast, where they reach values corresponding to that of seawater saturated rock. (In these figures, conductivity values were set everywhere along the coast to the value measured there.) The data (Figures 7, 9, 11, 13, 15) show that the size of the low-conductivity zone does not change substantially through the year. Rather, the hammock region becomes increasingly resistive during the wet season with the most extensive zone of freshwater occurring at the end of the wet season in October. Contour maps of conductivity are compatible with monthly rainfall data (Figure 17) in that conductivities decrease in the wet season following rainfall recharge of freshwater. Based on the relatively sparse sampling in our EM profiles, the zone of maximum change in ground conductivities is in the hammock (Figure 18). Direct measurements of groundwater conductivity were taken at two wells. Well A is located in the center of the hammock, and well B is positioned at the transition from hammock to coastal rock barren (Figure 3) The groundwater-conductivity-versus-depth curves do not change significantly on the first second, and third trips (Figures 8 10, 12) on 2 / 12/00 4/01 / 00, and 4/31/00. There is little if any freshwater in either well during the dry season, although water in well A is consistently less saline. In both wells, water conductivities increase gradually with depth to seawater values at 6-7 meters depth. The absence of freshwater is reflected in the terrain conductivity maps which also delineate little change over this period (Figures 7 9, 11). 30

PAGE 40

Missing Page

PAGE 41

Difference between 4 /01 and 10/06 raw EMdata-lOmcoilspacing + 34 32 30 28 2 6 24 22 20 18 16 14 12 lO 8 6 4 2 0 -2 4 .{) -8 -10 -12 -14 Figure 18Difference between 4 /01 ( dry season) and 10 / 06 (wet season ) raw EM data (mS / m) at 10 m and 20 m coil spacings 32

PAGE 42

Wet-season data from well B (Figures 14 and 16), however are quite enigmatic. Figures 13 and 15 show a lower terrain conductivity at well A than at well B. This is contrary to what the direct groundwater conductivity measurements indicate: figures 14and 16 illustrate a fresher water column in well B than in well A. With this in mind one would expect a lower terrain conductivity near well B We are unsure of the origin of this discrepancy. A damaged well casing is one possible explanation. If rainwater is allowed to seep into the water column during wet months, the salinity would be decreased Another possibility could be a locally higher porosity near well B than near well A. There would be more void space for water to occupy near B, which would act to raise terrain conductivity values (but not water conductivity values) near well B. The groundwater in both wells becomes fresher during the wet season (Figures 14 and 16) as is shown in the terrain conductivity maps (Figures 13 and 15). Figure 18 shows that the maximum difference in EM readings in wet and dry seasons is found in the hammock as stated above Groundwater-conductivity-versus-depth plots of 1999 data (Kruse et al. 1999) reveal a similar se as onal variation: fresher water occurring in the hammock during and immediately following the wet season and more conductive groundwater in the dry season (Figures 19-21 ). Wet-Season Freshwater Volume An attempt is made to loosely quantify the volume of freshwater in the hammock following the wet season This was done by calculating from Figure 15 the area of the least conductive zone in the hammock(< 95 mS / m). There is however also a substantial region ofhammock north ofthe entrance of the Pre se rve where EM surveys were not 33

PAGE 43

Well A grolll1dwater conductivity (mS/cm) 0 10 20 30 40 50 60 0 .00 -1------1------L-..r....._ _ __.L_ __ __._ __ ___. 1.00 I 2.00 Q) (.) ;3 3.00 en 0 ] 4.00 ..s 0.. 5.00 6.00 fre hwater brackish water seawater 7 .00 .....___ ______________ ____ ------' F igure 19 Conductivity versus depth curve for well A on 9 /27/99. grotu1dwater conductivity (mS/cm) 0 10 20 30 40 50 60 0.00 +---......L...-----L---.J....._---'------L--------! ....-, 1.00 s '-' Q) (.) r.S 2.00 en 0 Q) 3.00 ..D ..c ...... 0.. Q) '"(j 4.00 5.00 fre hwater brackish water seawate r 6.00 ____ --=..::..::..:..=:..:....___J Figure 20Conductivity versus depth c urve for well Bon 9/27 / 99. 34

PAGE 44

(A) grmmdwater conductivity (mS/cm) 0 10 20 30 40 50 60 ,......_ a '-" 2 Q) (.) ::s 3 til 0 Q) 4 ..D .s 0.. Q) 5 "'0 6 e hwater brackish water se water 7 (B) groundwater conductivity (mS/cm) 0 10 20 30 40 50 60 0 8 '-" 2 Q) (.) ::s .., ..) til 0 -4 Q) ..D .s 0.. Q) 5 "'0 6 7 freshwater brackish water seawater Figure 21 -Conductivity versus depth curves for (A) well A and (B) well B on 4 / 17 /99. 35

PAGE 45

conducted To account for this, we assumed th a t a comparable thickne ss of freshwater underlies areas of similar elevation and hammock vegetation. Aerial photographs were used to demarcate areas of similar v ege tation. A "freshwater" area of 55 m x 68 m was calculated from figure 15, and the aerial photographs yielded an area of 250 m x 120 m. The sum of these two areas was then multiplied by the length of the column of freshwater in well A (1. 73 m) to yield a crude volumetric estimate of 60,000 m3 of freshwater in the hammock following trip 5. The Impa ct of Plugging th e Mosquito Control Ditch If the ditches are significantly influencing s ubsurface conditions the effects should be most noticeable near the main ditch In an earlier study Kruse et al. (1999) noticed that the zone of low terrain conductivities in the preserve is centered west of the cent e r of the southern peninsula of Little Torch Key so that the zone of high terrain conductivities on the eastern side of the key is broader than the high-conductivity zone on the western side. They inferred the broader z one of high conductivities on the eastern side could be due to the presence ofthe mosquito ditches, which were dredged further inland on the eastern side (Figure 3). They infe ned this as mosquito ditches could potentially enhance saltwater inflow, incre a se runoff and hence reduce freshwater recharge. Ifthe mosquito ditches are in fact enhancing saltwater inflow and locally reducing recharge, we would expect a l oca l increas e in groundwater salinity near the ditch prior to plugging Following plu gg ing any s uch signal s hould disappear. EM data. We tested thi s hypoth esis wit h the EM data discu s s e d above and a set of repeated resistivity surveys across the ditch along the profile shown in Figures 22-26 36

PAGE 46

E 0 ....... N fl),.._ ro r..Ll 0 0 C'l ........ C'l 0 0 ui ('") 0 N c: ti Cl> "' 0 "0 :::> Q) "' a.. >-.. u; Q) 0::: c Q) (; a. :t "0 :::> "' '" Cl> E 0 N ,.._ 0 cxi .... 0 ui CTl r; 0 N 0ml'lm..,.m..,.oll"l 0ml'1m..,.m..,.oll"l ONC'IilritDtrig::: ONC'ia.rir.cicrig::: a. a. c: 0 13 Cl> "' 0 "0 :::> Q) "' a. >-. : t:: .2!: .. u; Cl> 0::: c Q) 1i Q. a. <( .., Q) ii :; (.) (i (.) E 0 t: Q) en :::!: 0::: CTl c: 1! d I E IN ; I I..; ti Q) en ............... ..t:: dN - 0 Figure 22Resistivity section on 2 /12/ 00 Top, middle, and bottom cross sections are meas u red apparent resistivity, calculated apparent resistivity, and inverted model resistivity, respectively 37

PAGE 47

::!1 0 N w I :;o 0 en ..... en a en 0 (') a. 0 ::l 0 ::l 0 --w 0 00 0 {/) 0 0 ::n 0 N w Ql .., 0.. 0 en () ;:I '0 ct. 0 ::l Ps Z O O West 12. 0 24. 0 4/01/00 36 0 East 48 0 00 0 72. 0 m 2 3 3 8 5 4 6 9 8 4 10. 0 11. 5 Mea s ured Apparent Resistivity Pseudosection P s Z 0 0 12.0 24. 0 :36. 0 48 0 00 0 72 0 m 2 3 3 8 5 4 6 9 8 4 10 0 11. 5 Calcu late d Apparent Resis tivit y Pseudo section Depth heration 3 RMS error= 12.4% 0 0 12. 0 Ditch 0 .4j _ _ _ _________ 2 1 4 7 6 5 8 8 11. 6 Inverse Model Res istivity Section ___ _ ____ D _____ o .1oo 0 .225 0 500 1.14 2 .56 5 n 13.o 29. 2 Resistivity in ohm m

PAGE 48

'TJ .., ('1) N I :::0 ('1) [:!?. en ...... en ('1) a 0 0 0 :::3 w \.Q ...._ w -...._ 0 0 (/) ('1) ('1) :n .., ('1) N w (3> .., 0-('1) en 0 ::l. '0 0 ? West 12.0 4 / 31/00 60. 0 East Ps. Z 0.,0 . . . I I I I ::::.:: I :;;:::' I I 0 .8j ..... ._..., to 48. 0 24 0 36 0 72.0 m 2 3 3 8 5 4 6 9 8 4 10 0 11.5 Measured Apparent Resistivity Pseudosection Ps. Z 0 0 1 2.0 24 0 0 8 2 3 3 8 5 4 6 9 8 4 10. 0 11. 5 Calculated Apparent Resistivity Pseudosection Depth Iteration 3 RMS error= 3 4 % 0 0 12. 0 24 0 36.0 48. 0 60 0 72.0 m 48. 0 i ' id ' I i ; ;-.:.; I ' ' -ik ..:.::-;: I 'I 4 7 6 5 8 8 11. 6 Inverse Model Resistivity Section ___ _ ____ D ____ o .1oo o 225 o.so6 1 .14 2 56 5 .77 13. 0 29.2 ResistivitY in ohm m

PAGE 49

.p. 0 Ps. Z O O West 1 2 0 24.0 7 /10/ 00 48 0 60 0 East 72.0 m 0 8 '""\ 2 3 (1) 3 8 N Vl 5 4 10 6 9 (1) 8 4 g?. 10 0 en t:t. 11.5 < Measured Apparent Res i slivi t y Pseudosection en Ps. Z O O 12. 0 24. 0 35 0 48 0 60 0 72. 0 m (1) 0 0 8 c-. 0 2 3 ::s 3 8 0 5 4 ::s -...1 6 9 -...... 8 4 0 10 0 -0 0 11. 5 Calculated Apparent Resistivity Pseudosection 0 4 '""\ (1) N w 4 7 2 1 Q> '""\ 0. (1) en 0 ::J. '0 c-. 0 ::s 6 5 8 8 11. 6 Inverse Model Resistivity Section 0 ----0 100 0 225 0 506 1 .14 2 56 s n 13. 0 29 2 Resistivity in ohm m

PAGE 50

Ps. Z OO 12. 0West ..., 0 8 Ctl N 0\ Ctl Cll a ::;. Cll Ctl (') a 0 ::;:1 0 ::;:1 ,__ 0 ..._ 0 0\ ..._ 0 0 (/) Ctl Ctl 2 3 3 8 5 4 6 9 8 4 1 0 0 1 1 5 Mea s ured Apparent Resistivity Pseudosection P s Z O O 12 0 0 8 2 3 3 8 5 4 6 9 8 4 10. 0 1 1 5 Calculated Apparent Resis tivit y P seudose ction Depth 3 RMS error= 2.7 o/o DD 120 c 0 4 @ N w ..., 0. Ctl Cll (') ;::::!. "0 a 0 ::;:1 2 1 4 7 6 5 8 8 11. 6 Inverse Model Resistivity Sec tion --------24 0 24 0 0 .100 0 225 0 .51E 1 .14 2 56 10 / 06 / 00 35 0 Ditch 0 ----s n 1 3 0 29 2 Resis tivit y in ohm. m 48 0 60 0 East 72 .0m. 48 0 60 0 72.0 m

PAGE 51

The EM data are all west of the ditch and because of the significant seasonal variability, it is difficult from the maps in Figures 7, 9 11, 13, and 15 to discern a distinctive change associated with the plugging of the ditch. This difficulty was exacerbated as the plugging of the ditch was delayed from its original scheduling to May 2000, so that the plugging time coincided with the onset of the wet season. R esistivity data. The resistivi ty profile traverses the principal north-south ditch 30 m north (landward) ofthe plug. The ditch crosses the profile a t x = 39 m (Figures 22-26). The signal ofthe ditch itself is apparent on some ofthe lines as a resistivity high at the shallowest depths (the resistivity of the air filling the upper part of the ditch is higher than the resistivity of the surrounding rock). There is a subtle conductivity high (resistivity low) beneath the ditch on two of the three pre-plug surveys (2/12/ 00 and 4 / 01100 but not 4/31 / 00). Thus any zone of low resistivity (high conductivity, high salinity) aro und this point on the pre-plug surveys (Figures 22-24) must be well mixed and near or below the detection limit of the resist ivity method. We note also that the models for all surveys are fairly consistent at depth (5-12 m) (Figures 22-26). The primary dissimilarities are found near the surface. In particular, the model from 4 /01100 (Figure 23) shows higher resistivity values near the ground surface probably due to th e low amount of rain between 2/12/00 and 4 /01/ 00 (Figure 17) Salinity of ditch water. Salinit y of s urface water in the primary ditch was measured with a refractometer monthly throughout 2000 at seven sites between the southern and northern ends of the trench by Nature Conservancy personnel. Data in c lude salinity measurements both before and after plugging in addition to measurements on either side of the plug following its in stallatio n in May B eca u se the plug was installed at 42

PAGE 52

the beginning ofthe rainy season, any potential effect from the plug would be obfuscated by the large seasonal variability-the water in the ditch would presumably become less saline as the wet season progressed regardless of the efficacy of the plug However, if the plug is inhibiting seawater from being channeled in to the center of the island as well as allowing freshwater to mound up landward of the plug, thereby increasing recharge, then sites on either side of the plug would exhibit different trends in salinity following plugging of the ditch The sites south of the plug would continue to be tidally affected while the sites north of the plug would no longer be directly connected to the ocean and thus show evidence of more of an effect from the influx of freshwater Qualitative interpretation of aerial photographs. To assess the past effectiveness of the ditch, historical aerial photographs were examined. Four aerial photographs were acquired ranging from 1959, depicting pre-development conditions, to 1986, which shows a similar environment to the present da y (Figure 27). For comparison purposes, the images were rectified using ER Mapper 5 .5, (Earth Resource Mapping Pty Ltd.). Unfortunately, the lack of infrastructure and other fixed features on the 1959 image limited the rectification process Consequently, although very similar, the maps' scales are not exactly the same. Following rectification, the photographs were converted from the original black and white to a spectrum of colors display. This transformation increases the contrast in the images, easing identification and differentiation of vegetative regions. The dark red and magenta represent the densest vegetation and the greens, yellows, and reds indicate the sparsest. The black pixels are areas where no color is assigned. Analysis of im ages is incorporated into the discussion below. 43

PAGE 53

1959 1963 1974 Figure 27 -Rectified color-flooded aerial photographs ofTorchwood Preserve 1959, 1963, 1974, 1986. 44

PAGE 54

Chapter 4: Discussion and Models Interpreting the Absence of a High Conduc tivit y Zone near th e Ditch Several possible interpretations are proposed for the lack of a zone of high conductivity near the ditch prior to plugging. (a) The impact of the mosquito ditch ex t e nded over a z one broader than 70 meters, so it was not detected by the resistivity surveys which extended only 35 meters on either side of the ditch. (b) The mosquito ditch prior to plugging did not e nhance influx of saltwater but did affect the fresh/brackish water lens by l oca lly r e ducing recharge (water runs into the ditch where it may evaporate, instead of infiltrating the groun d). Our pre-plug" s urv eys were all conducted during the dry season, when rainfall and hence recharg e are minimal. Thus, it is possible that before plu gg in g the ditch did locally reduce recharge in the wet season, but we were not able to image this with our dry season profiles (c) The mosquito ditches shortly prior to plugging did not significantly impact the fresh/brackish water lens. This could be because they are small enough that they never had a significant impact, or that by 2000 they were "naturally plugged by mangroves and other vegetation. We can test these three scenarios agai n st our EM d a ta gro undwat e r conductivities ditch salinities and aeria l photos 45

PAGE 55

Testing Scenario a: Scenario (a): If the zone of major impact of the mosquito ditches is broader than 70 meters (35m on each side of the ditch), then we should detect it with EM surveys, which extend a greater distance from the ditch. This is a little bit tricky, as the ditches were dug throug h an area where pre-ditch conductivities certainly increased from west to east, toward the eastern shore. This conductivity gradient is inferred from vegetation patterns and apparent elevations from pre-ditch aerial photos. Access on the eastern side of the ditch was l imited, so our regularly sampled EM data are all on the western side of the ditch. If the ditch had a "broad zone of impact", then theE-W conductivity gradient west of the ditch should become less steep after plugging. To test this, we examined three E W EM profiles on the western side ofthe ditch (Figure 5) Note the northern profile is a specia l case: it also runs parallel to an EW extension of the main ditch. If scenario (a) is valid, we would expect a more dramatic pre-plug to post-plug change all along this profile than on the others Of the three EW EM transects near the ditch, none of them exhibit a sizeable disparity in the conductivity gradients between preand post plug dates (Figures 28-30). This lack of change in the conductivity gradients suggests that p l ugging the mosquito ditch did not have a "broad-scale" ( > 70 m) zone of impact. Testing Scenarios b and c: Scenario (b): We cannot test this scenario with geophysical data collected in this study. However, the ditch-water salinity data suggest that the ditch is not a steady conduit for groundwater in the dry season. If the ditch impacted the lens by increasing runoff and thereby reducing recharge but not significantly through facilitating saltwater 46

PAGE 56

Middle (A) 220 120 ';>' .... -'"" ... .... 100 (B) 80 ... .. .... -,.,..... --... 60 40 dis tance W of ditch (m) 2 / 12/00 lOrn --.-4 101100 lOrn + 4 /31/ 00 1Om 7 / 10/00 !Om l 10/ 06 / 00 1Om 1 20 0 . -21 12100 20m -.. 4 /01/ 00 20m 4/3 I / 00 2 0m -s-7/10/ 00 20m 10/ 06 / 00 20m l 200 ,......_ e Cl) e ....._., 180 .0 160 ...... u ::l "0 c:: 0 140 u ...... c:: Q) 120 0. 100 2 20 ,......_ 2 00 Cl) E ....._., 180 .0 +-' 160 u ::l "0 c:: 0 140 u ...... c:: Q)" 120 0. 100 1 2 0 100 80 60 40 2 0 0 di s tanc e W o f ditch ( m) Figure 28 -Apparent conductivity versu s di s tance west of ditch for middle EM tran s e c t. (A) i s 10m coil spacing ; (B) i s 20m. Darker lines indicate pre-plu g gin g dat e s 4 7

PAGE 57

North (A) -4/01/00 lOrn + -4 /31100 1Om -+-7/10/00 lOrn +--10/6/00 lOrn 120 100 80 (B) .... 60 40 20 meters W of ditch 0 180 0 ...... ;:. .a 160 g ] 0 140 () i:l C1) 120 [ 100 --2112/00 20m -.. 4 /01/ 00 20m - 4/31/00 20m -7110/00 20m .a..::.... ____ ......... .. ... .,..,..-/ ----10/6/00 20m .. 120 100 80 60 40 20 meters W of ditch 0 ,.-... 200 (/) 5 180 0 s: 160 g "0 c: 0 140 () E ro 120 0.. < 100 Figure 29 -Apparent conductivity versus distance west of ditch for north EM transect. (A) is 10 m coil spacing; (B) is 20 m. Darker lines indicate pre-plugging dates. 48

PAGE 58

South (A) 220 -+-2/12/00 I Om -4 / 01/00 10m ..--.. 200 CZl + -4 / 31/00 I Om 8 '-" -+-7 /10/ 00 I Om 180 0 ;; + I 0 / 06 / 00 1Om 160 '"0 c:: 0 140 u c:: . - CIJ .... 120 ell 0. -. . 0. ell 100 120 100 80 60 40 20 0 distance W of ditch (m) (B) n -. ..,. __ __ :/ .. . I -----.. 200 :: CZl I 5 180 0 I ,--2/12/ 00 20m 1 .. 4 / 0 1100 20m 4 / 3 1100 20m 9 7 /10/ 00 20m a1 0 / 06 / 00 20m I ;; ...... u 160 ::I '"0 s:: 0 u 140 t: ell 0. 0. 120 ell 120 100 80 60 40 20 0 distanc e W of ditch (m) Figure 30Apparent conductivity versus distance w es t of ditch for sout h EM transect. (A) is I 0 m coil spacing ; (B) is 20 m. Darker lin es indicate pre-plugging dates. 49

PAGE 59

influx, then we expect fresh to brackish water in the ditch (both preand post-plugging) Unfortunately, because the pre-plug data were collected only during the dry season, the impact of the ditch on the lens in the wet season, when recharge would be most significant, cannot be assessed. Scenario (c): If the mosquito ditches did not have hydrogeologic significance at the time of our pre-plug surveys, then either ( 1) they never did, or (2) they did but lost it over time as the ditches were naturally plugged by vegetation. Case (2) we can test with the historical aerial photographs. Case 1. Figure 31 consistently shows similar trends in salinity for all sites throughout the year, regardless of the ditch plugging. This supports the conclusion that the plugging had little, if any, effect on the flow dynamics. The plug lies between sites 2CS and 2C. Therefore, any changes in salinity trends should be most observable between these two locations. Sites 2CS and 2C show nearly identical patterns. Changes in salinity seem to be more affected by precipitation rates than anything else. Comparing figure 32 with monthly rainfall amounts (Figure 17), salinity is inversely related to the amount of precipitation whether the plug is in place or not. It is interesting to note the absence of a substantial north-south salinity gradient, even in the months prior to the plugging. Cas e 2. The ditches were excavated between 1959 and 1963 and can be easily seen in the 1963 image (Figure 27). The 197 4 photograph delineates a substantial decrease in vegetative cover that initially seems to be a consequence of the ditches. However, this decline is restricted to only the low-lying areas at the southern tip of the 50

PAGE 60

P-.30 0.. 0 60 P-.30 0.. 0 60 P-.30 0.. 0 60 P-.30 0.. 0 60 P-.30 0.. 0 60 P-.30 0.. 0 60 P-.30 0.. 0 ()0 P-.30 0.. 0 60 0.,30 0.. 0 60 P-.30 0.. 0 60 P-.30 0.. 0 60 P-.30 0.. 0 60 P-.30 7/26/00 10 /27/ 00 0.. 0 11/26/00 0 = f 100 200 300 400 500 distance ( 111) Figure 3 1 -Salinity of water i n pr im ary ditch versus distanc e from so uthern s h ore. Gray boxes indicate s ite s l andward of the plug. Dark gray boxes and closed circles show post-plug dat es. 51

PAGE 61

island. There is no diminution in vegetation near the ditches in the hammock. In fact there is an increase in the amount of vegetation in the upland regions-the trenches, which shortly after excavation are clearly visible, are barely discernible in the 1974 photo. There is also no significant vegetative d e cline near the ditches in the northwestern portion ofthe image which would be expected if the ditches were indeed the cause of the diminishment at the southern tip. Moreover the 1986 image shows a significant recovery ofthe vegetation in the areas that are negatively affected in the 1974 image These observations contradict the hypothesis that the ditch dredging had a long term impact on vegetation. A more likely scenario to explain the diminished vegetation in the 1974 photo is a large storm event that flooded the lower portions of the island with seawater. The storm event hypothesis is supported by the fact that there is a severe loss of mangrove fringe from 1963 to 1974 possibly du e to high winds or strong storm waves. Because mangroves thrive in saline env ironments ditch dredging would not influence the density of the mangroves at the periphery of the island Furthermore, the 1986 image shows significant recovery of the man g rove fringe. 1-D Analyti ca l Model of th e Impa c t of a Ditch The results above suggest these mosquito ditch es nev er, or only briefly acted as s teady conduits for saltwater influx. Langevin e t al. ( 1998) demonstrated that channels that serve as saltwater conduits (freshwater head at sea level) can significantly reduce freshwater lens volume Here we use a very s imple one-dimensional analytic al model to roughly quantify the reduction in len s vo lume expected for a ditch in steady contact with the ocean. In thi s model hydraulic conductivity (K) and recharge (R) are assumed to be 52

PAGE 62

homogeneous on an infinitely long island with an infinitely long ditch that parallels the shores (effectively cutting the island into two neighboring islands). The governing equa tion (Vacher 1988) is where h is the hydraulic head at a position x; a is the density of freshwater divided by the density difference between freshwater and saltwater ; Lis the width of the island and xis the position variab le originating at one edge of the island. The model of the island without the ditch produces a typical Ghyben-Herzberg l ens containing an area of 57 m2 of water in the lens. To simu lat e an island with a ditch, we essentially modeled two separate lenses and added the area of each to attain the total area of brackish water under the island The two lenses are separated by the ditch i t self, a point where head is equal to sea level effectively creating more coast line. The model with the ditch yiel ded an area of37 m2 of water under the island This is a 35% reduction in lens area due to the presence of the ditch. The decrease in lens size as predicted by the analytical model contradicts the results of this study. This suggests that the ditch si mpl y does not function as a conduit for the passage of seawater into the island. Furthermore, if the ditches are regularly bringing in seawater and thus "creating more coast lin e, the vegetat ion n e ar them would reflec t this influx: mangroves would be completely choking the ditch. While there are mangroves in the ditches they are not found all a lon g th e ditch. 53

PAGE 63

Numerical Model In order to determine whether the lens that was imaged geophysically could be replicated numerically using plausible v alues for hydraulic parameters, a numerical model was constructed using SEA W AT (Langevin 200 I) to simulate variable-density flow. SEA W AT couples MOD FLOW (McDonald and Harbaugh 1988) and MT3D (Zheng 1990) to solve the combined flow and tran s port equations. SEA W AT uses a modified finite-difference approximation of the flow equation were made. The modification is that mass rather than volumes is conserved and the flow equations include a gravity term. Additionally, instead of using hydraulic head as the primary dependent variable as MOD FLOW does SEA W AT use s equivalent freshwater head (Langevin 2001) SEAWAT discretizes time in the same manner as MODFLOW and MT3D by usin g stress periods and time steps (Lan g evin 2001) The numerical model of Little Torch K e y con s ists of 59 rows 60 columns and 12 layers. The cells around the outside of th e i s land are 40 m by 40 m and d e crease in siz e near the center of the island where the tar ge t valu e s for calibration are located. Each layer is 1m thick. The top 10 layers ofthe model repre s ent the Miami Limestone and were given a hydraulic conductivity value of 100 m / d (Wightman 1990). A hydraulic conductivity of 1000 mid was assigned to the bottom two layers which represent the Key Largo Limestone (Wightman 1990) The boundary conditions on the ea s tem, sou th em, and westem e d g es of the model were made constant head (h = 0 m) and conc e ntration ( C = 35 ppt) The northem e d g e was set as a no-flow boundary 5 4

PAGE 64

Wet season. The model was calibrated to wet season (10 / 06 / 00) groundwater salinity data. Recharge, because it is the least-known parameter and the one that varies most, was varied to yield a model that closely( 3.5 ppt) matched the target values. To do this, three zones of recharge were outlined that correspond spatially to plant communities within the preserve (Figure 32). Zone 1, which was assigned a recharge value of0.0026 mid (0.95 mlyr), encompasses the hammock where recharge is the highest. For comparison, average annual rainfall for the Keys is roughly 0 0036 mid (1.3 mlyr) Zone 2, which represents the transition from hammock to coastal rock barren, was assigned a recharge of 0.0022 mid (0.80 mlyr). The periphery of the island was given a recharge value of 0.0004 mid (0.15 mlyr). The recharge values necessary to produce a calibrated model are extremely high. The probable cause for this is that the model was calibrated to data that were collected three days following an intense rainfall The data were collected on 10/ 06 / 00, and between 10/01/00 and 10/03/ 00 Key West received 0.17 m of precipitation about 13% of the average annual rainfall. In order to replicate the fairly thick mixing zone seen in Figure 16, dispersivity values were introduced: longitudinal dispersivity = 10m, transverse dispersivity = 1 m, and vertical dispersivity = 0.2 m. This combination of recharge and dispersivity values produces a calibrated model (Figure s 33 and 34). Figures 35 and 36 show a model identical in every way except with no dispersivity. With no dispersivi ty a nd the same r e charge, the model contains too much freshwater. In attempts to calibrate the model without dispersivity, it was simply not possi ble to maintain the gradient across the fresher 55

PAGE 65

Figure 32-Figure showing zones of recharge in numerical model. Red= 0 0026 m/d (0 .95 m /y r), yellow= 0 0024 m/d (0.88 m/yr), and blue = 0.0004 m / d (0 .15 m/yr) 5 6

PAGE 66

location of well A Figure 33Cross-section along column 32 of numerical model calibrated to wet season data with dispersivity Arrows represent velocity vectors Residuals are posted in the appropriate layers in white Units for residuals and color scale in ppt. 57 40 38 36 34 32 30 28 26 24 22 20 18 16 14 12 10 8 6 4 2 0

PAGE 67

Layer 1 (0 5 m depth) Layer 4 (3. 5 m depth) Reference Vedcrs ( m/d) 0 0.35 Layer 8 (7 5 m depth) Figure 34-Color flood contour maps of concentration in ppt from layers 1 4 8 and 12 of numerical model with dispersivity calibrated to wet season data Arrows represent velocity vectors. 58

PAGE 68

Location of well A 40 38 36 34 32 30 28 26 24 22 20 18 16 14 12 10 8 6 4 2 0 Figure 35Cross-section along column 32 for numerical model without dispersivity calibrated to wet season data Arrows represent velocity vectors Residuals are posted in appropriate layers in white Units for color scale and residuals in ppt. 59

PAGE 69

.. u I l I I cross-section 1 i ) i : -,:_,. i well A t_, 1 I t 1 (( t ........ -/.. Layer 1 (0 5 m depth) Rdet"Ell'"'re VEdas (m/d) 0 0.34 Layer 8 (7 5 m depth) Figure 3 Color flood contour maps o f concentration in ppt layers 1 4 8 and 12 of numerical model with no disper s ivity calibrated to wet season data Arrows represent velocity vectors 60 40 38 36 34 32 30 28 26 24 22 20 18 16 14 12 10 8 6 4 2

PAGE 70

target values near the surface while keeping the deeper target values saline Non dispersive models consistently yield results ei ther too saline or too fresh. This exercise shows that it is possibl e to closely reproduce the geophysical and water conductivity values with a numerical s imulation incorporatin g plausible values of hydraulic parameters. Dry season. The same numerical mod e l was also calibrated to dry-season ( 4 /31/ 00) well A data The recharge zones that were used for the wet season model (Figure 32) were also used for the dry season model. In order to produce a calibrated model, zone 1 was assigned a recharge va lue of 7.0 x 10-4 rnld (0.26 rnlr); zone 2 was assigned 4.0 x 10-4 rnld (0.15 rnlyr); and zon e 3 was ass igned a value of2.0 x 10-5 rnl d (7.3 x 10-3 rnlyr). Cross-sectional and plan views of the dry season model are shown in Figures 37 and 38, respectively The dry-season data from well A (Figure 12) shows a much steeper grade to seawater than the wet-season data (Figure 16). The dispersivity that is required to fit the wet seaso n data is not necessary in the dryseason model. The numeri ca l dispersion in the dryseaso n model is sufficient to simulate the narrower mixing zone. Calculat ion of r echa r ge. With these models, it i s pos s ible to place bounds on the recharge to the southern portion of the island. An assumed upper bound is the recharge value assigned to the hammock (model zone 1) in the ca librated wet-season model (0.0026 rnld). A lower bound on wet-season recharge ca n be estimated as follows. The minimum wet-season rechar ge is that required to supply (a) the incr ease in freshwater in 6 1

PAGE 71

Figure 37Cross-section along column 32 for dry season numerical model. Arrows represent velocity vectors Residuals are posted in appropria t e layers in black. Units for color scale and residuals in ppt. 62

PAGE 72

Layer 1 (0 5 m depth) Reference Vectors (m/d) 0 0 35 Layer 8 (7 5 m depth) Layer 4 (3. 5 m depth) Layer 1 2 (11.5 m depth) 40 38 36 34 32 30 28 26 24 22 20 18 16 14 12 10 8 6 4 2 0 Figure 38-Color flood contour maps of concentration in ppt from layers 1 4, 8, and 12 of numerical model for dry season Arrows represent velocity vectors 6 3

PAGE 73

the lens that occurs from dry-to wet-season plus (b) the groundwater discharg during that time. The increase in the volume of freshwater from the dry-season model to the wet season simulation (part a) is 200 000 m3 ( I ,0 00 000 m3 minus 800 000 m\ A reasonable lower bound estimate of part (b) is simply the dry-season discharge. An approximate average dry season discharge is 3 7 x 1 o -4 m i d ( 14 m/yr). The wet season recharge estimated in this way is 1.6 x 10-3 m/d (0.57 m/yr) (to i ncrease lens volume) plus 3 7 x 10-4 m/d (to discharge), a total of2.0 x 10-3 mid (0 73 m/yr). This "lower bound" estimate of wet se ason recharge is 56% of the average annual rainfall. This result is likely too high. Possibly the overestimate is du e in part to using a non steady snapshot following a period of intense rainfall for the v olume of the wet season lens. Alternatively, or additionally, the overestimate may reflect an assumed hydraulic conductivity that is too high, an assumed porosity that is too hi g h or both 64

PAGE 74

Chapter 5: Conclusions The combined results show that the impact to date of plugging the mosquito ditch is small compared with the overall seasonal variability in the fresh/brackish water lens During the dry season, brackish groundwater in the hammock shows conductivities (salinities) increasing downward to seawater values at 6-m depth. During the wet seaso n, at the center ofthe hammock (well A) 2-m of freshwater overlies water of increasing salinity. The apparently limited influence of the mosquito ditches on groundwater salinity is probably due to their relatively small size and the growth of mangroves and other natural "plugs" into the ditches. The observation that the zo ne of lowest terrain conductivities (freshest water) is west of the center of th e peninsula is probably not due to the presence of the mosquito ditch network (as was inferre d by Kruse et al. (1999)) but is probably due to other factors such as lower elevations, increased evaporation, etc., on the eas tern margin If the Semaphore cactus is in fact being adversely affected by a loss of freshwater on the island, another pos si ble explanation for this loss is the excavation of the boat canals durin g construction of The Jolly Roger's Estates These canals, much lar ge r than the mosquito ditches may have a mor e s i gn ifi cant impact on the fresh/brack i sh water lens Canals of this size are the kinds of features modeled by Langevin et al. ( 1 998). In 65

PAGE 75

addition, the Florida Keys have been impacted by recent sea level rise, which would act to reduce the volume of freshwater on this l ow-lying island. 66

PAGE 76

References Coniglio, M., and Harrison R. S. 1983, Facies and diagenesis of Late Pleistocene carbonates from Big Pine Key Florida Bulletin of Canadian Petroleum Geology V 31, no. 3, p 135-147. Dunham, R. J., 1962, Classification of carbonate rocks according to depositional texture. in Classification of Carbonate Rocks, A Symposium ed W. D Ham, American Association of Petroleum Geologists Memoir 1, p. 108-1 21. Hanson, C F., 1980, Water resources of Big Pine Key, Monroe County Florida. U S Geological Survey Open File Report. 80-44 7, 36 pp. Hoffmeister, J. E., and Multer, H G., 1968 Geology and origin of the Florida Keys. Geological Society of America Bulletin. v 79, no. 11, p. 1487-1502. Hoffmeister, J E and Multer, H G ., 1964 Pleistocene limestones of the Florida Keys. In Geological Society of America Annual Meeting Field Trip Guidebook, Field Trip 1, p. 57-61, ed. R.N. Ginsburg. Hoffmeister, J. E., Stockman, K. W., and Multer H. G., 1967, Miami Limestone of Florida and its Recent Bahamian counterpart Geological Society of America Bulletin. v. 78, p. 175-190. Kruse S Inman, J and Liauw-a-pau H 1999, Fresh and brackish water lenses on the Torchwood Pres erve, Little Torch Key, Florida. Unpublished Report to The Nature Conservancy. Langevin, C. D ., Stewart, M. T and Beaudoin, C M., 1998, Effects of sea water canals on fresh water resources: an example from Big Pine Key, Florida Ground Water. v 36 ,no.3,p.503-513. Langevin, C. D 2001 U.S. Geological Survey Written communication Loke M H ., 19 96, RES2DINV ver 2.0: rapid 2D resistivity inversion using the least squares method (Wenner, dipole-dipole, pole-pole, pole-dipole, Schlumberger) Campus Geophysical Instruments Ltd. Birmingham, England. 67

PAGE 77

McDonald, M. G., and Harbaugh, A. W., 1988, A modular three-dimensional finite difference ground-water flow model. Techniques ofWater-Resources Investigations 06-A1, USGS 576 p McNeill J.D., 1980a, Electromagnetic terrain conductivity measurements at low induction numbers. Techn i cal Note TN 6 Geonics Ltd ., Mississisauga, Canada, 15 pp. National Oceanic and Atmospheric Administration, http :/ /ols.ncdc.noaa.gov / cgi bin/nndc/buyOL-002 cgi Perkins, R. D. 1977, Depositional framework of Pleistocene rocks in South Florida. In Quaternary Sedimentation in South Florida Memoir 147, ed P Enos and R. D. Perkins, 131-198. Boulder, Colorado : Geological Society of America. Stewart M. T., 1988, Electromagnetic mapping of fresh-water lenses on small oceanic islands. Ground Water v 26 no 2 p 187-191. Stewart, M T., 1982, Evaluation of electromagnetic methods for rapid mapping of salt water interfaces in coastal aquifers. Ground Water. v. 20, no 5, p 538 545 The Nature Conservancy of the Florida Keys Torchwood Hammock Preserve Field Guide, Little Torch Key, Florida, The Nature Conservancy P 0 Box 4958 Key West, FL, 19 p. Vacher, H L., and Quinn, T M. (eds ) 1997 Geology and hydrogeology of carbonate islands. Developments in Sedimentology 54, Elsevier Science Publishers, Amsterdam Vacher, H. L., 1988, Dupuit-Ghyben-Herzberg analysis of strip island lenses Geological Society of America Bulletin v 100 p 580-591. Vacher, H L. 1978, Hydrogeology of Bermuda -significance of an across-the-island variation in permeability. Journal of Hydrology v. 39, p 207 226. Wightman, M J., 1990, Geophysical analysis and mathematical modeling of freshwater lenses on Big Pine Key, Florida. M. S. thesis, Geology Department University of South FloridaTampa Zheng, C. 1990, MT3D a modular three d i mens i onal transport model S. S. Papadopulos & Assoc., Rockville, MD 68


printinsert_linkshareget_appmore_horiz

Download Options

close
No images are available for this item.
Cite this item close

APA

Cras ut cursus ante, a fringilla nunc. Mauris lorem nunc, cursus sit amet enim ac, vehicula vestibulum mi. Mauris viverra nisl vel enim faucibus porta. Praesent sit amet ornare diam, non finibus nulla.

MLA

Cras efficitur magna et sapien varius, luctus ullamcorper dolor convallis. Orci varius natoque penatibus et magnis dis parturient montes, nascetur ridiculus mus. Fusce sit amet justo ut erat laoreet congue sed a ante.

CHICAGO

Phasellus ornare in augue eu imperdiet. Donec malesuada sapien ante, at vehicula orci tempor molestie. Proin vitae urna elit. Pellentesque vitae nisi et diam euismod malesuada aliquet non erat.

WIKIPEDIA

Nunc fringilla dolor ut dictum placerat. Proin ac neque rutrum, consectetur ligula id, laoreet ligula. Nulla lorem massa, consectetur vitae consequat in, lobortis at dolor. Nunc sed leo odio.