|USFDC Home | USF Electronic Theses and Dissertations||| RSS|
This item is only available as the following downloads:
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-SFE0004793
A retrospective analysis of the potential environmental stressors responsible for the decline of the natural populations of the florida apple snail (_pomacea paludosa_) in the a.r.m. loxahatchee national wildlife refuge
h [electronic resource] /
by Shannon Ladd.
[Tampa, Fla] :
b University of South Florida,
Title from PDF of title page.
Document formatted into pages; contains X pages.
Thesis (MS)--University of South Florida, 2010.
Includes bibliographical references.
Text (Electronic thesis) in PDF format.
Mode of access: World Wide Web.
System requirements: World Wide Web browser and PDF reader.
ABSTRACT: The purpose of this thesis is to determine the factors that contributed to the decline of Florida apple snail (Pomacea paludosa) populations in the A.R.M. Loxahatchee National Wildlife Refuge with the goal of devising management recommendations to the Refuge regarding population management strategies. The factors examined that could have potentially contributed to population decline include the use of copper-based herbicides, insecticide application, the occurrence of drought, the use of other herbicides, the occurrence of fire, and non-avian predation. Annual Narrative documents produced by Refuge managers and staff members, dated from 1951 to 2007, were used to collect historical data for these factors. The quality of data reporting within the Annual Narratives was also examined. To support data on droughts documented in the Annual Narratives, surface water and rainfall data were obtained and analyzed. The methodology includes the use of conceptual ecological models and historical ecology to determine whether or not the factors examined produced an ecological effect capable of affecting the Refuge population of apple snails. Evidence from the Annual Narratives suggests that the use of copper-based herbicides, the occurrence of drought, and predation by alligators were responsible for the decline of the apple snail on the Refuge. A lack of consistently reported data regarding apple snail densities makes it difficult to determine the degree to which each factor had an effect on the apple snails or to determine if any spatio-temporal relationship existed between the Florida apple snail and Everglade snail kite (Rostrhamus sociabilis plumbeus) based on copper-based herbicide use. The overall quality of the Annual Narratives improved throughout the study period and eventually focused heavily on investigative studies. Several management recommendations were suggested to improve Florida apple snail populations on the Refuge. First, in order to monitor the health and trends of the apple snail population, a monitoring network needs to be established with results maintained in a geodatabase. Both apple snail density and egg cluster counts need to be made following an established sampling method. Second, in an attempt to sustain higher apple snail densities, stocking of the interior should be attempted. Finall, in the event that adjacent farmlands are to be restored, soil samples need to be analyzed to determine if concentrations are high enough that desorption of copper from the flooded agricultural soils could pose a serious threat to the Refuge by reintroducing toxic levels of copper.
Advisor: Kamal Alsharif, Ph.D.
Everglades snail kite
x Environmental Science & Policy
t USF Electronic Theses and Dissertations.
A Retrospective Analysis of the Potential Environmental Stressors Responsible for the Decline of the Natural Populations of the Fl orida Apple Snail ( Pomacea paludosa ) in the A.R.M. Loxahatchee National Wildlife Refug e by Shannon Ladd A thesis submitted in partial fulfillment of the requirements for the degree of Master of Science Department of Geography College of Arts and Sciences University of South Florida Major Professor: Kamal Alsharif, Ph.D. Joni Downs, Ph.D. Matthew Harwell, Ph.D. Date of Approval: November 1, 2010 Keywords: Population dynamics Toxicity Everglades rest oration Everglades snail kite Drought Copyright 2010, Shannon Ladd
ACKNOWLEDGEMENTS I wish to acknowledge and thank those people who contribu ted to this thesis: Dr. Matt Harwell for pushing me in the right directio n, donating his time to travel to Tampa multiple times to be a part of this process as w ell as showing me around the Loxahatchee Refuge, for providing valuabl e insight on how to improve my scientific writing style, and for his patience and und erstanding along the way. Dr. Joni Downs for agreeing to come on board with this project in hopes of aiding me with incorporating my findings into GIS and sticking around and providing helpful feedback even though that hope wasnt made possi ble. and most importantly Dr. Kamal Alsharif, my advisor, for reading revision after revision and helping my thesis to evolve into the qual ity paper that it is now by pushing me to strengthen areas that were weak and add d ata where needed, without whom I wouldnt be where I am now. Thank you all for helping me through this journey an d for helping me to produce a thesis I can be proud of.
i TABLE OF CONTENTS LIST OF TABLES .................................... ................................................... .......... iv LIST OF FIGURES ................................... ................................................... ......... v ABSTRACT ......................................... ................................................... ........... vii CHAPTER 1: INTRODUCTION ........................... ................................................. 1 CHAPTER 2: LITERATURE REVIEW ...................... ............................................ 7 Applied Historical Ecology .......................... ............................................... 7 Data Reporting ................................... ................................................... ..... 9 Florida Apple Snail Ecology ...................... ............................................... 10 Everglade Snail Kite Ecology ...................... ............................................. 12 Copper ........................................... ................................................... ....... 13 Insecticides ........................................ ................................................... ... 20 Drought .......................................... ................................................... ....... 22 Other Herbicides .................................. ................................................... 25 Fire ............................................ ................................................... ....... 26 Non-Avian Predation .............................. .................................................. 28 CHAPTER 3: STUDY AREA.............................. ................................................. 2 9 Climate .......................................... ................................................... ........ 30 Topography, Soils and Land Use ..................... ........................................ 30 Hydrology and Water Quality ...................... ............................................. 31 CHAPTER 4: RESEARCH DESIGN ........................ ........................................... 36 CHAPTER 5: METHODOLOGY ............................ ............................................. 31 CHAPTER 6: RESULTS ................................ ................................................... .. 49 Florida Apple Snail Trends ....................... ................................................ 49 Overview ............................................ ........................................... 49 1951-1959 .................................................. ................................... 49 1960-1969 .................................................. ................................... 51 1970-1979 .................................................. ................................... 53 1980-1989 .................................................. ................................... 58 1990-1999 .................................................. ................................... 60 2000-2007 .................................................. ................................... 62
ii Everglade Snail Kite Trends ...................... .............................................. 64 Overview ............................................ ........................................... 64 1951-1959 .................................................. ................................... 65 1960-1969 .................................................. ................................... 65 1970-1979 .................................................. ................................... 67 1980-1989 .................................................. ................................... 69 1990-1999 .................................................. ................................... 72 2000-2007 .................................................. ................................... 73 Copper ........................................... ................................................... ....... 74 Overview ............................................ ........................................... 74 1951-1959 .................................................. ................................... 75 1960-1969 .................................................. ................................... 75 1970-1979 .................................................. ................................... 75 1980-1989 .................................................. ................................... 76 1990-1999 .................................................. ................................... 77 2000-2007 .................................................. ................................... 77 Insecticides ........................................ ................................................... ... 77 Overview ............................................ ........................................... 77 1951-1959 .................................................. ................................... 78 1960-1969 .................................................. ................................... 78 1970-1979 .................................................. ................................... 79 1980-1989 .................................................. ................................... 79 1990-1999 .................................................. ................................... 80 2000-2007 .................................................. ................................... 80 Drought .......................................... ................................................... ....... 81 Overview ............................................ ........................................... 81 Wet and Dry Season Hydrologic Analysis ................ ..................... 82 1951-1959 .................................................. ................................... 87 Decadal Water Level Fluctuations ................. ..................... 92 1960-1969 .................................................. ................................... 83 Decadal Water Level Fluctuations ................. ................... 95 1970-1979 .................................................. ................................... 96 Decadal Water Level Fluctuations ................. ................... 99 1980-1989 .................................................. ................................. 100 Decadal Water Level Fluctuations ................. ................. 103 1990-1999 .................................................. ................................. 104 Decadal Water Level Fluctuations ................. ................. 107 2000-2007 .................................................. ................................. 109 Decadal Water Level Fluctuations ................. ................. 111 Water Regulation Schedule .......................... .............................. 112 Other Herbicides .................................. .................................................. 119 Overview ............................................ ......................................... 119 1951-1959 .................................................. ................................. 120 1960-1969 .................................................. ................................. 121 1970-1979 .................................................. ................................. 123 1980-1989 .................................................. ................................. 125
iii 1990-1999 .................................................. ................................. 127 2000-2007 ......................................... .......................................... 128 Fire ............................................ ................................................... ..... 129 Overview ............................................ ......................................... 129 1951-1959 .................................................. ................................. 130 1960-1969 .................................................. ................................. 131 1970-1979 .................................................. ................................. 132 1980-1989 .................................................. ................................. 132 1990-1999 .................................................. ................................. 134 2000-2007 .................................................. ................................. 135 Non-Avian Predation .............................. ................................................ 13 7 Overview ............................................ ......................................... 137 1951-1959 .................................................. ................................. 137 1960-1969 .................................................. ................................. 138 1970-1979 .................................................. ................................. 139 1980-1989 .................................................. ................................. 142 1990-1999 .................................................. ................................. 142 2000-2007 .................................................. ................................. 148 CHAPTER 7: DISCUSSION ............................. ................................................ 14 9 Data Reporting ................................... ................................................... 149 Florida Apple Snail Trends ....................... .............................................. 152 Everglade Snail Kite Trends ...................... ............................................ 155 Copper ........................................... ................................................... ..... 157 Insecticides ........................................ ................................................... 159 Drought .......................................... ................................................... ..... 162 Other Herbicides .................................. .................................................. 164 Fire ............................................ ................................................... ..... 167 Non-Avian Predation .............................. ................................................ 16 9 CHAPTER 8: CONCLUSIONS AND RECOMMENDATIONS......... .................. 172 LITERATURE CITED .................................. ................................................... .. 182
iv LIST OF TABLES TABLE 1: Example Overview of Components Examined Table............... .......... 44 TABLE 2: Overview of Components Examined Table Terminology .. ................. 45 TABLE 3: Overview of Components Examined from 1951 to 1959 ................... 51 TABLE 4: Overview of Components Examined from 1960 to 1969 ................... 53 TABLE 5: Florida Apple Snail Egg Cluster Counts ............................................ 54 TABLE 6: Overview of Components Examined from 1970 to 1979 ................... 58 TABLE 7: Overview of Components Examined from 1980 to 1989 ................... 60 TABLE 8: Overview of Components Examined from 1990 to 1999 ................... 61 TABLE 9: Overview of Components Examined from 2000 to 2007 ................... 63 TABLE 10: Mean Wet and Dry Season Surface Water Levels for 1-7 ............... 84 TABLE 11: Mean Wet and Dry Season Surface Water Levels for 1-8C ............ 85 TABLE 12: Mean Wet and Dry Season Rainfall ................................................. 87 TABLE 13: Cooperative Alligator Survey Data ................................................. 1 39 TABLE 14: Interior and Canal Alligator Nighttime Survey Data from 1998 to 2000 .................................................. .............................................. 144 TABLE 15: Interior and Canal Alligator Nighttime Survey Data from 2001 to 2003 .................................................. .............................................. 145 TABLE 16: Interior and Canal Alligator Nighttime Survey Data from 2004 to 2006 .................................................. .............................................. 146 TABLE 17: Interior and Canal Alligator Nighttime Survey Data for 2007 ......... 147
v LIST OF FIGURES FIGURE 1A: Regional Perspective of 1974 USGS Land Use ............................ 34 FIGURE 1B: Regional Perspective of 2004 SFWMD Land Use ......................... 35 FIGURE 2: Initial Florida Apple Snail Stressors Conceptual Model .................... 39 FIGURE 3: Conceptual Ecological Model Diagram ............................................ 46 FIGURE 4: Interior and Canal Transect Locations .............. ............................... 56 FIGURE 5: Location of Surface Water Gauges by Site ID ................................. 83 FIGURE 6A: Mean Monthly Surface Water Level for Site 1-7 from M ay 1954 to April 1960 .................................................. ............................ 92 FIGURE 6B: Mean Monthly Surface Water Level for Site 1-8C from May 1954 to April 1960 .................................................. ............................ 93 FIGUER 7A: Mean Monthly Surface Water Level for Site 1-7 from M ay 1960 to April 1970 .................................................. ............................ 96 FIGURE 7B: Mean Monthly Surface Water Level for Site 1-8C from May 1960 to April 1970 .................................................. ............................ 96 FIGURE 8A: Mean Monthly Surface Water Level for Site 1-7 from M ay 1970 to April 1980 .................................................. .......................... 100 FIGURE 8B: Mean Monthly Surface Water Level for Site 1-8C from May 1970 to April 1980 .................................................. .......................... 100 FIGURE 9A: Mean Monthly Surface Water Level for Site 1-7 from M ay 1980 to April 1990 .................................................. .......................... 104 FIGURE 9B: Mean Monthly Surface Water Level for Site 1-8C from May 1980 to April 1990 .................................................. .......................... 104 FIGUER 10A: Mean Monthly Surface Water Level for Site 1-7 from May 1990 to April 2000 .................................................. .......................... 108
vi FIGURE 10B: Mean Monthly Surface Water Level for Site 1-8C from May 1990 to April 2000 .................................................. .......................... 109 FIGURE 11A: Mean Monthly Surface Water Level for Site 1-7 from May 2000 to April 2007 .................................................. .......................... 112 FIGURE 11B: Mean Monthly Surface Water Level for Site 1-8C from May 2000 to April 2007 .................................................. .......................... 112 FIGURE 12A: Mean Monthly Surface Water Level for Site 1-7 Duri ng the Water Regulation Schedule from July 1960 to June 1969 ................ 113 FIGURE 12B: Mean Monthly Surface Water Level for Site 1-8C Dur ing the Water Regulation Schedule from July 1960 to June 1969 ................ 114 FIGURE 13A: Mean Monthly Surface Water Level for Site 1-7 Duri ng the Water Regulation Schedule from July 1969 to June 1975 ................ 114 FIGURE 13B: Mean Monthly Surface Water Level for Site 1-8C Dur ing the Water Regulation Schedule from July 1969 to June 1975 ................ 115 FIGURE 14A: Mean Monthly Surface Water Level for Site 1-7 Duri ng the Water Regulation Schedule from July 1975 to April 199 5 ................ 116 FIGURE 14B: Mean Monthly Surface Water Level for Site 1-8C Dur ing the Water Regulation Schedule from July 1975 to April 197 5 ................ 117 FIGURE 15A: Mean Monthly Surface Water Level for Site 1-7 Duri ng the Water Regulation Schedule from May 1995 through 2007 ............... 118 FIGURE 15B: Mean Monthly Surface Water Level for Site 1-8C Dur ing the Water Regulation Schedule from May 1995 through 2007 ............... 119 FIGURE 16: Final Florida Apple Snail Stressors Conceptual Model ............... 173
vii ABSTRACT The purpose of this thesis is to determine the factors th at contributed to the decline of Florida apple snail ( Pomacea paludosa ) populations in the A.R.M. Loxahatchee National Wildlife Refuge with the goal o f devising management recommendations to the Refuge regarding population ma nagement strategies. The factors examined that could have potentially contr ibuted to population decline include the use of copper-based herbicides, ins ecticide application, the occurrence of drought, the use of other herbicides, the o ccurrence of fire, and non-avian predation. Annual Narrative documents produ ced by Refuge managers and staff members, dated from 1951 to 2007, w ere used to collect historical data for these factors. The quality of data reporting within the Annual Narratives was also examined. To support data on drough ts documented in the Annual Narratives, surface water and rainfall data wer e obtained and analyzed. The methodology includes the use of conceptual ecological models and historical ecology to determine whether or not the factors exami ned produced an ecological effect capable of affecting the Refuge popul ation of apple snails. Evidence from the Annual Narratives suggests that the use of copper-based herbicides, the occurrence of drought, and predation by alligators were responsible for the decline of the apple snail on the Refuge. A lack of consistently reported data regarding apple snail densities makes it difficult to determine the
viii degree to which each factor had an effect on the apple snails or to determine if any spatio-temporal relationship existed between the Florida apple snail and Everglade snail kite ( Rostrhamus sociabilis plumbeus ) based on copper-based herbicide use. The overall quality of the Annual Narr atives improved throughout the study period and eventually focused heavily on in vestigative studies. Several management recommendations were suggested to improve F lorida apple snail populations on the Refuge. First, in order to monito r the health and trends of the apple snail population, a monitoring network needs to be established with results maintained in a geodatabase. Both apple snail density and egg cluster counts need to be made following an established sampling meth od. Second, in an attempt to sustain higher apple snail densities, stocking of the interior should be attempted. Finall, in the event that adjacent farml ands are to be restored, soil samples need to be analyzed to determine if concentratio ns are high enough that desorption of copper from the flooded agricultural soil s could pose a serious threat to the Refuge by reintroducing toxic levels of c opper.
1 CHAPTER 1: INTRODUCTION The Florida apple snail ( Pomacea paludosa ) has been identified as a critical prey species for wetland inhabitants and as a pot ential indicator of wetlands restoration success in South Florida (Karunaratn e et al., 2006), hence declines in apple snail populations suggest ecosystem healt h may be in danger. This thesis uses a holistic approach to examine the potent ial overlapping effects of various mechanisms of apple snail decline at the A.R. M. Loxahatchee National Wildlife Refuge (hereafter referred to as Loxahatche e Refuge or the Refuge) in the northern Everglades. The purpose of this thesis is to determine the factors that contributed to the decline of Florida apple snail populations from 1951 to 2007 with the goal of devising management recommendat ions to the Refuge regarding future Florida apple snail population man agement strategies. The Refuges Annual Narratives will serve as the source of hi storical data. In addition to examining the potential environmental stressors that could have potentially led to population decline, the quality of and detail to which the data are reported throughout the study period is also examined. While habitat loss and alteration and altered hydrol ogic regimes have been attributed to the general decline of Florida ap ple snail populations in South Florida, population changes within the Loxahatchee Re fuge have not appeared to be affected by these alterations (Winger et al., 1 984). The decline of Florida
2 apple snails in the canals around the Refuge appeared to coincide with the use of copper-based herbicides for the management of nuisan ce aquatic vegetation, such as hydrilla ( Hydrilla verticillata ) (Winger et al., 1984). However, since there is only anecdotal evidence suggesting this relationship, it is necessary to consider multiple potential mechanisms of decline as part of a longer period of record (e.g., by examining over 50 years of narrative data). While a breadth of research on the potential effects of copper and drought, and to some extent insecticides, on the apple snail exists, there are also lesse r-documented factors that could account declining populations, including the use of non-copper based herbicides, fire, and non-avian predation. Looking at historical data is an important factor for determining population trends wi thin a species, as are taking into account future factors that may affect the apple sna il. These trends could include potential negative effects from restoration, co mpetition with exotic invasive snail species, and habitat change because of clim ate change. As the main food source for the federally endangered Everglade snail kite ( Rostrhamus sociabilis plumbeus ) as well as prey for other birds (e.g., limpkins; Aramus guarauna ), fish, reptiles, and mammals, the Florida apple snai l plays an important role in the Everglades ecosystem (Sharfstein a nd Steinman, 2001; Hoang and Rand, 2009). The Everglade snail kites diet is composed almost exclusively of native Florida apple snails (Cottam and Knappen, 1939; USFWS, 2004; Frakes et al., 2008). Knowledge of the status, he alth, and trends of the Florida apple snail within the Refuge designated as critical habitat for the critically endangered Everglade snail kite is thus of g reat relevance for efforts
3 to manage the ecosystem to improve snail kite foraging su ccess and populations (Harwell, 2009). Throughout Florida, copper compounds have been used sin gly and in combination with other compounds as an algaecide, fungi cide and soil amendment (Hoang et al., 2008a; Hoang et al., 2009) Copper compounds have been used as fertilizer and fungicides for citrus crops, a s well as in algaecides and herbicides, which are permitted by the Florida Dep artment of Environmental Protection for control of nuisance planktonic and filame ntous algal and vascular plants (Hoang et al., 2008a). Repeated applications ha ve resulted in elevated copper concentrations in South Florida ecosystems (Hoang at al., 2008a; Hoang and Rand, 2009). Since the early 1970s, copper sulfate and other chemicals were applied for vegetation maintenance control purp oses within the Refuge leading to a gradient of copper concentration from the canal, where the highest concentrations occur, to the interior (Winger et al., 19 84). Since it has been discovered that copper is toxic to Florida apple snails (H oang et al., 2009; Rogevich et al., 2009), an examination of any potent ial influence between vegetation control applications and changes in the Flori da apple snail abundance and distribution in the Loxahatchee Refuge is warrant ed (Harwell, 2009). While copper-based herbicides have proven to be toxic t o Florida apple snails, various other herbicides have been used at the Re fuge for vegetation maintenance control purposes. With species such as water hya cinth ( Eichhornia crassipes ), water lettuce ( Pistia stratiotes ), Brazilian pepper ( Schinus terebinthifolius ), Australian pine ( Casuarina species), and melaleuca ( Melaleuca
4 quinquenervia ) invading the natural landscape, much time and effor t has been put into finding the most effective methods of control. Both chemical and manual efforts, along with prescribed burns, were used as contro l methods. With such a variety of nuisance vegetation, a variety of chemica l herbicides have been used including 2,4-D, diquat, Rodeo, and Arsenal according t o the Refuges Annual Narratives. If and when concern arose regarding the pot ential toxic effects of these herbicides on non-target species, such as the apple snail, arose, testing was carried out to determine whether or not there wer e any toxic effects. Organochloride insecticide residues in Everglade snail ki tes, snail kite eggs, and Florida apple snails have been low and thought to r eflect background environmental condition (Lamont and Reichel, 1970; S ykes, 1985). However, there is a need to evaluate the usage of these insecticid es, such as synthetic DDT (dichlorodiphenyltrichloroethane) and their effect s on Florida apple snails. Taking into account the known facts that some of these in secticides can magnify themselves through the food chain through bioaccumulatio n (Newsome, 1967) and their detrimental effects on avian populations, t he potential accumulation of toxic levels in snail kites could affect this sensitive specie s. Factors such as the spatial and temporal extent of drou ght affect the apple snails potential to survive (Darby et al., 2002). It has generally been assumed that apple snails have little to no tolerance to droug ht conditions (Turner, 1994; Darby et al., 2002), however this assumption is largely based on indirect evidence from snail kite observations (Karunaratne et al ., 2006), and there is some anecdotal documentation that Florida apple snail s are capable of
5 aestivation (Turner, 1994; Darby et al., 2002). Appl e snail movement tends to cease at water depths below 10 cm and both Florida appl e snails and Everglades snail kites tend to move toward refugia during drought s of limited spatial extent (Darby et al., 2002; Mooij et al., 2002). As spatial extent and severity of drought increases, not only are apple snail strandings more frequ ent, a system-wide drought greatly increases the chance for reduced reprod uction and increased mortality for the snail kite (Mooij et al., 2002). Al though drought may negatively affect populations of apple snails and snail kites, dro ughts in the Everglades are a natural and necessary process, including its role main taining vegetation communities (Darby et al., 2008). Whether or not the apple snail has the ability to survive extensive drought can have serious implications no t only for Florida apple snail populations, but also for foraging success and surviva l of the Everglades snail kite. Chapter 2, the literature review, will provide backgr ound information in order to enhance the understanding of the key species an d factors examined within this thesis. In addition, the use of applied hist orical ecology and data reporting standards will be examined. Chapter 3 intr oduces the study area and provides climate, topography, soils, land use, hydrology and water quality information pertaining to that particular area. The questions this thesis aims to answer are identified in Chapter 4 along with their corresponding hypotheses. Sets of objectives are also presented here. The methodol ogy in Chapter 5 describes the use of; the Annual Narratives for data coll ection, the South Florida Water Management Districts (SFWMD) DBHYDRO to analyze surface water
6 level, the South Florida Water Management Model (SF WMM) as a method to analyze rainfall data, conceptual ecological models, an d biological versus ecological responses. The results of analysis of the Annual Narratives, surface water levels and rainfall data will be reported in C hapter 6. Based on the results reported in Chapter 6, Chapter 7 will discuss each subje ct examined. From this, Chapter 8 will provide recommendation to the Refuge. The questions asked in Chapter 4 will also be revisited in Chapter 8 and fin al conclusions will be made. The last section will provide citations for all the lite rature used throughout the thesis.
7 CHAPTER 2: LITERATURE REVIEW Applied Historical Ecology In order to better understand and manage ecosystems an d environmental processes, historical knowledge is used (Swetnam et al., 199 9; Goforth and Minnich, 2007). This is known as applied historical ecology (Swetnam et al., 1999). According to Swetnam et al. (1999): historical ecology is a sufficiently long time sequence (ch ronology) of measurements of observations so that meaningful in formation can be gained about changes in populations, ecosystems structures, disturbance frequencies, process rates, trends, periodicities, and other dynamical behaviors. With the continuous evolution of ecological knowledge o ver time, ecological research is continuously evolving in new directions provid ing paths to overall increased understanding (Graham and Dayton, 2002). How ever, there are some limitations to using applied historical ecology data. T hese limitations include the filtering of past environmental information, limited quantity and quality of records, completeness and reliability, and the potential for hi ghly biased interpretation (Swetnam et al., 1999; Goforth and Minnich, 2007). As time passes, ecological progress is made through the ex pansion of the understanding of the functions of natural systems (Graham and Dayton, 2002).
8 The degree to which the expansion occurs can vary from g radually over time to rapid jumps into new fields. While these rapid jumps in to new directions can leave gaps in understanding, additionally ecological pr ogress can lead to ecological specialization, the erasure of history and the expansion of literature all which may present as an obstacle to future progress. H owever, enhancing historical understanding, filling these gaps in knowledge and utilizing the increased amount of literature over time will allow t hese obstacles to be overcome (Graham and Dayton, 2002). In addition to m aking ecological progress, various paradigms have emerged throughout hist ory (Graham and Dayton, 2002; Naeem, 2007). Naeem (2007) states that a paradigm describes: an unprecedented scientific achievement that is compellin g enough to convince adherents from traditional perspectiv es to shift their allegiance, regroup around the new paradigm, and tackle problems anew stemming from Thomas Kuhns (1962) definition. For exa mple, the paradigm that Naeem (2007) discusses concerns the dialect used to explain nature by studying its parts and the dialect used to explain nature by stud ying whole-system behavior. Lunt and Spooner (2005) used historical ecology to u nderstand patterns of biodiversity in fragmented agricultural landscapes. Wha t were once continuous ecosystems have become increasingly fragmented by the deve lopment of agricultural lands. In order to identify spatial and t emporal interaction and patterns across the landscape, historical ecology was used to fill the gaps.
9 Historical ecology allowed for an enhanced understandin g of why different ecosystem states and species occur where the do in highly mo dified landscapes, the use of human management practices as a variable to explain ecological patterns and the recognition of anthropogenic disturban ce as a driver of biotic patterns for temporal change (Lunt and Spooner; 2005) This study provided an example of how the use of historical ecology can fill e xisting gaps in knowledge as mentioned by Graham and Dayton (2002). Data Reporting Those who are responsible for reporting data, scientist s, are faced with taking on the role of the analyst who is objective and value free, and the advocate who is biased and value laden (Wallington and Moore, 2005). Recently the established distinction that science is objective has be gun to waiver in favor of science as an interactive activity in which critical di scussions take place in multiple forums among a diverse scientific community (Wal lington and Moore, 2005). Often study selection bias may be introduced in quantitative studies by trying to fix the data, or lack thereof, before analy sis by removing cases with missing data or substituting missing values with the vari able mean (Peugh and Enders, 2004). Additionally, scientists have to ensure th ey are not being selective in the use of statistical tests in order to prove signifi cance (Peugh and Enders, 2004). Wallington and Moore (2005) reported three differe nt dimensions of ecological reasoning concerning data gathered and hypot heses formed;
10 empirical evidence, conceptual criticism and arguments base d on experience. Empirical evidence uses evidence from the field to sup port theories. Conceptual criticism is grounded in theoretical concerns and uses accep ted theory. Finally, arguments are based on experience use to enhance the cred ibility. Regardless of the approach of ecological reasoning, criticism by the me ans of peer review is used to support the evidence, methods, assumptions, an d reasoning. In general, it is essential for scientists to make their w ork available to the masses and to report their findings through publication (Toft and Jaeger, 1998). Publication of scientific findings permanently distribute s the authors work for scrutiny from peers and the public. As previously mention ed, peer review is a mechanism to strengthen the validity of the data bein g reported. The format with which scientific findings are reported if rigidly constrain ed and highly disciplined. The style used to report scientific findings for publicat ion must be concise, accurate and unambiguous (Toft and Jaeger, 1998). Florida Apple Snail Ecology Although the importance of the Florida apple snail ( Pomacea paludosa ) has long been established, little research was conducted o n life history and ecology until the past decade, in part because a lack of a validated sampling technique until recently (Darby et al., 1999; Karunar atne et al., 2006). The Florida apple snail occurs in isolated locations in southern Geor gia and Alabama, but is widely distributed throughout the Florida peninsula, albeit sporadically in the Florida Panhandle. Its population range is limited b ecause of its inability to
11 survive the lower winter temperatures that occur to the north. The Florida apple snail inhabits shallow lentic environments including pond s, swamps, and marshes (Thompson, 1999). Apple snails are generally f ound at the soil-water interface and at least occasionally burrow under the soil surface (Hoang and Rand, 2009). The Florida apple snail exhibits amphibious characteristi cs, having the ability to obtain oxygen from the water by the use of its gill and from the air by the use of its lung (Winger et al., 1984; Darby et al., 2 008). During periods of drought, the apple snail is capable of aestivating, a state of dormancy similar to that of hibernation occurring in times of heat and dr yness, by burying itself in the mud (Darby et al., 2008). The apple snails diet consist s primarily of periphyton and submerged vascular plants (Frakes et al., 2008). As a defense mechanism, apple snails detach from the substrate they are on and simply drop to the ground (Frakes et al., 2008). Sexually mature adult Florida apple snails range in siz e from roughly 25 to 60 mm in diameter with females tending to be larger than males (Darby et al., 2008). The life span of the Florida apple snail aver ages 1.0 to 1.5 years, with their life culminating in a post reproductive die off (Darby et al., 2003). Apple snail egg clusters are found on emergent structures, generally vegetation, above the water line and contain around 20 to 30 eggs (Darby et al., 2008). Egg cluster production typically occurs from February to November. S tudies have shown a seasonal peak to occur between April and May. From this i nformation, high adult mortality rates are expected in June and July (Darby et al., 2003).
12 Darby et al. (2006) conducted limited apple snail sam pling in the Refuge interior in 2002, 2003 and 2004, collecting density i nformation from four prairieslough sites, two mixed prairie sites and four wet prair ie sites within the Refuge. Although these data are not representative enough to examine overall apple snail trends within the Refuge, it does provide some status information. Of the sites sampled, only two had apple snail densities greater tha n 0.14 snails/m 2 the estimated minimum density to support snail kite foragin g (Darby et al., 2006). Results for the sampled sites included one site with no s nails, five sites with densities less than 0.08 snails/m 2 three sites with densities between 0.12 to 0.14 snails/m 2 and one site with a density of 0.22 snails/m 2 (Darby et al., 2006). Everglade Snail Kite Ecology Primarily inhabiting freshwater marshes from Florida, Cuba, Mexico and south, individuals inhabiting Florida and Cuba make up the subspecies Rostrhamus sociabilis plumbeus, referred to as the Everglade snail kite (Sykes et al., 1995, FWS, 2004). Currently, there is no eviden ce that the snail kite moves between Florida and Cuba, and as such, the Florida pop ulation is considered to be a single population. The Everglade snail kite was li sted as an endangered species following the Endangered Species Conservation Act in 1967. The specific diet of this wide-ranging species consists almost ent irely of the Florida apple snail (Cottam and Knappen, 1939; FWS, 2004; Fr akes et al., 2008), making their survival related to those factors that driv e apple snail abundance and distribution. Additionally, the abundance and dist ribution of snail kites are
13 strongly linked to hydrology because of the negative ef fect of drought events on apple snails (Mooij et al., 2002). In order to forage for apple snails, the snail kite r equires foraging areas that are clear and open, such suitable habitat is gener ally a low profile, low density marsh with shallow, clear and calm water (Sykes, 1 987; Rodgers et al., 2001; FWS, 2004). Dense growth of herbaceous or woody vegetation and growth of exotic and invasive native plants promoted by eutr ophication limit the ability of the snail kite to forage for apple snails. Unger the Endangered Species Act, the Loxahatchee Refuge was designated as critical habitat for the Everglades snail kite in 1977. The breeding season for the snail kite varies from year to year and between areas in relation to rainfall and water leve ls. Between December and July, 98% of nesting attempts are made. The actual num ber of clutches produced per breeding season is not well documented, but clutch siz e is between 1 and 4 eggs with an incubation period of 24-30 days. Snail kit es become nomadic in response to changes in factors such as food availability, water depth and hydroperiod, yet are not considered migratory (FWS, 2 004). As a result, snail kite numbers in the Refuge are not considered individu al population estimates, rather part of the larger peninsular Florida populat ion. Copper An important goal in designing aquatic herbicides is to maximize efficacy to target species while minimizing risk to non-target species those species that
14 the pesticide or fungicide is not intended to kill (Mitr a and Raghu, 1998) nevertheless, copper-based herbicide application can have unintended consequences (Mastin and Rogers, 2000). Comparative toxici ty studies are important to establish precise margins of safety, or the magnitude of differences between toxic concentrations to target and non-target sp ecies, when copperbased products are intended for use as fertilizers, her bicides, fungicides, pesticides, etc. (de Oliveira-Filho et al., 2004). Com parative toxicity studies, such as that conducted by de Oliveira-Filho (2004), utiliz e the LC 50 concentration the median lethal concentration (Lethal Concentration, 5 0%) to determine the concentration required to kill half the individuals of the test group. de OliveiraFilho et al. (2004) compared the susceptibility of diff erent freshwater target and non-target organisms to three copper-based pesticides ( copper oxychloride, cuprous oxide and copper sulfate). By examining how to xic and non-toxic forms of copper that exist in natural waters affects the organ isms in a laboratory setting, these experiments provide a worst case scenario for deter mining risk associated with pesticide application rates in the natural environ ment. Copper toxicity depends not only on its concentration and bioavailabili ty, but also on the sensitivity of the organism. Based on results of soft wat er assay, copper sulfate is identified as being more readily bioavailable than o ther forms of copper (Mastin and Rodgers, 2000). de Oliveira-Filho et al. (2004) confirmed that increased levels of copper in water bodies are likely to negative ly affect a variety of nontarget species lethally and by growth inhibition. Since this phenomenon affects
15 organisms at the base of the food web, increased copper b ioavailability has the potential to dramatically affect freshwater ecosystems. The bioavailability of copper in water is influenced b y multiple environmental factors including pH, alkalinity, hardness, salinity, and dissolved organic carbons (DOC) (Hoang et al., 2009). Various cop per toxicity studies have shown a decrease in copper toxicity as DOC and pH increase (Rogevich et al., 2008; Hoang et al., 2008b; Hoang et al., 2009). Alt hough the influence of hardness on copper toxicity varies among different specie s, hardness appears to have no effect on copper toxicity to the Florida appl e snail (Rogevich et al., 2008). Winger et al. (1984) evaluated the effects of copper on juvenile and adult Florida apple snails, and attempted to make a prelimin ary conclusion as to whether or not these herbicides were responsible for the reported decline of Florida apple snail populations in the perimeter canal s of the Loxahatchee Refuge. Although Winger et al. (1984) established tha t Cutrine-Plus and Komeen, two chelated copper-based herbicides, were toxi c to juvenile Florida apple snails with 96-hour LC 50 values of 22 and 24 g/L respectively, they concluded, at that time, that treatment with copper wa s not responsible for the decline of apple snails in the Refuge. However, Winger et al., (1984) recognized that more information was needed on the: long-term effects of high body burdens of copper accumu lated through exposure to herbicidal applications or contami nated sediments in the absence of food on survival and reproduct ion of
16 the apple snail, susceptibility of apple snails with prior exposure to copper to repeated herbicidal application, and environ mental significance of predation on apple snails containing hig h residues of copper. Copper has the potential to be transferred to the ap ple snail through the water column, sediments, periphyton, and vascular plant s and potentially to its predators through bioaccumulation (Hoang et al., 2008a ). Composed of algae, floating plants, and associated animals, periphyton commu nities are present throughout the Everglades. Periphyton is responsible fo r a large portion of the primary production in the Everglades and as the base of the food web, periphyton is the main food source for primary consumer s (Gaiser, 2009). Recent findings in a study conducted by Frakes et al. (2008) fou nd that mean copper concentrations in Florida apple snails ranged from 23.9 mg/kg at a reference site known to receive no anthropogenic copper inputs to 732 m g/kg at a high copper site and were correlated primarily with copper concentra tions in sediments, periphyton and vascular plants (Frakes et al., 2008). I t has been demonstrated that aqueous copper uptake is strongly dependent on disso lved organic carbons (DOC), as DOC concentrations increase, copper bioavailabil ity and uptake decrease (Hoang et al., 2008a). Total organic carbon ( TOC) and copper concentrations in Everglades surface water show that aque ous copper uptake should not be of concern for the health of Florida app le snails in the current Everglades ecosystems (Hoang et al., 2008a).
17 The Florida apple snail has direct contact with all thre e potential routes of copper exposure: water, sediments, and dietary uptake. Frakes et al. (2008) showed a strong positive correlation between copper concen trations in snails and those in sediments, but no relationship between copper i n snail and surface water copper concentrations, consistent with findings of H oang and Rand (2009). Similarly, Hoang et al. (2008b) demonstrated that ad ult apple snails can accumulate copper from soils through soil ingestion and/or dermal contact. Since aquatic plants absorb and adsorb dissolved copper, diet i s also an important route of copper accumulation (Hoang et al., 2008a). F rakes et al. (2008) established a strong correlation between copper concentr ations in snails and concentrations in vascular plants and sediments. Hoang et al. (2008b) demonstrated that apple snails fed copper contaminated lettuce contained higher whole body copper concentrations than those fed copperfree lettuce. This indicated copper accumulation through the diet. Frakes e t al., (2008) identified a correlation was also made between snail and periphyton copper concentrations, however, since periphyton may not be stationary but ra ther some types of periphyton can move in response to wind and water curre nts, a periphyton sample may not be representative of copper concentration s at the area where it was collected (Frakes et al., 2008). In the Loxahatchee Refuge, periphyton is primarily stationary, growing attached to other veget ation (Gaiser, 2009). Since many Florida sediments seem to be copper poor and because copper does not degrade and accumulates in the sediments (Leslie, 1992), significant copper enrichment of the soil can have pote ntial adverse effects.
18 Samples collected throughout the Loxahatchee Refuge by Winger et al. (1984) showed an average of 34mg/kg of copper, ranging from 2 7 to 40mg/kg. A control sample collected from an area upstream and outside the Refuge not likely to have received copper from treatments or drift, had a co pper concentration of 10mg/kg (Winger et al., 1984). During collection, Frakes et al. (2008) found that Flor ida apple snail abundance appeared to be lower at locations with grea ter copper concentrations in sediments. Although Winger et al. (1984) had conclu ded that coppercontaining herbicides were not responsible for decreased F lorida apple snail populations because the acute toxicity values for dissolv ed copper were considerably greater than dissolved copper concentrations f ound in canal waters, copper concentrations in sediments from the sites sampled in Frakes et al. (2008) were significantly higher in the current study t han those of Winger at al. (1984). This, along with the fact that chronic exposure to copper can result in reduced snail survival at levels much lower than acute to xicity values (ReedJunkins et al., 1997), could explain the lack of Florida apple snail presence at high-copper locations. For all potential copper exposure routes, the distribu tion of accumulated copper in the adult apple snail was similar. The major ity of accumulated copper was found in the soft tissue, predominantly the viscera a nd the foot, with a small portion in the shell (Hoang et al., 2008b). Rogevich et al. (2009) found that copper distributions within an apple snail varied depe nding on exposure concentrations, and that higher exposure concentrations l ed to a greater amount
19 of accumulation in the viscera the internal organs o r soft tissues. Since predators, such as the Everglades snail kite, consume the sof t tissues, there is a potential for copper transfer and bioaccumulation thro ugh the food chain to higher trophic levels (Hoang et al., 2008b; Frakes et a l., 2008; Hoang and Rand, 2009). Similar to other species, it has been shown that juveni le Florida apple snails are more sensitive to acute copper toxicity than adult snails (Winger et al., 1984; Rogevich et al., 2008). Rogevich et al. (2009) concluded that as a result of copper exposure, Florida apple snails exhibit significant ly reduced clutch production and egg hatching. Reduced clutch production an d egg hatching could potentially have an effect on apple snail population growth or recruitment, affecting the foraging success of predators such as the Everg lade snail kite. Aqueous copper uptake may not be an immediate concern within the Everglades, nevertheless, there are future concerns for adjacent lands that will be restored and incorporated into the ecosystems (Hoang et al., 2008b; Hoang et al., 2009). Execution of the Comprehensive Everglades Restoration Plan (CERP) under the Water Resources Development Act of 2000 and the potential River of Grass initiative involving the purchase of extensive ag ricultural land in the Everglades Agricultural Area (EAA) (Wehle, 2009) requ ires the acquisition of agriculture, including citrus and row crops, land to be flooded for creating hydrologic buffers, storm water treatment areas, water storage reservoirs, and wetlands (Hoang et al., 2008a; Hoang and Rand, 2009; Hoang et al., 2009). Portions of these lands are currently or were formerly managed with fertilizers
20 and pesticides, including copper. Flooding will convert t hese dry aerobic lands into inundated anaerobic sediments which may promote th e release of copper from soils. If copper-enriched areas are not remediated copper-enriched runoff could then affect the surface water quality of downstre am receiving waterbodies. Toxic forms of copper desorbed from inundated soils have the potential to adversely affect the survival and growth of the Florida apple snail (Hoang et al., 2008a). To address this concern a two-tiered sediment samp ling process exists to identify potential location requiring remediatio n of copper-enriched soils (Rand and Schuler, 2009). Insecticides While a large portion of the controversy around organ ochlorine insecticides, such as DDT, lie in their effects on human he alth as a carcinogen, there are also environmental impacts of its applicatio n (Newsome, 1967, Jaga and Brosius, 1999). Not only are some insecticides effecti ve on their target insect species, but they are also toxic to a wide range of nontarget species, including aquatic life and birds (Newsome, 1967). Insecticide use ha s the potential to contaminate water by direct application to water surface s, accidental application and drift, runoff from treated areas, and waste mater ials from the production process. Insecticide residues are found in soils from applicat ion to crops and to the soil itself for control of root-feeding pests (Newso me, 1967). Even though the use of certain insecticides, such as DDT, have been banned in the United States and concentrations may be below detectable analytical levels in surface waters,
21 residues from a single application may persist in sediment s for years and accumulate to toxic levels (Newsome, 1967; Rand et al., 2004). Many avian species, yet not the only affected species, including ospreys, peregrine falcons and bald eagles, have encountered r eproductive problems as a result of DDT use (Carson, 1962; Cooke, 1973). Direct ex posure may not be highly toxic, yet repetitive use of the insecticide resul ting in bioaccumulation through the food chain has been found to be toxic. Adu lt mortality, reproductive impairment and eggshell thinning are all toxic effects on avian species (Fry, 1995). Lamont and Reichel (1970) first reported levels of org anochlorine insecticides, such as DDT (and its metabolites DDD and DDE) and Dieldrin, in dead snail kites found near the Refuge as well as in ap ple snails collected from the Refuge between 1965 and 1967. Although the auth ors did not describe the significance of the residues found, residue levels were lo w in snail kites reflecting low levels in apple snails. These levels were thought to reflect the background environmental contamination (Lamont and Reichel, 1970 ). The death of 50 snail kites following the treatment of Surinam rice fields in 1971 with sodium pentachlorophenol, an organochlorine, resulting in hig h levels of pentachlorophenol residues in 17 kites led to a second exa mination of residues in snail kites in Florida by Sykes (1985). Unhatched sna il kite eggs and dead young found in Florida including some collected from the Loxahatchee Refuge between 1970 and 1977 were analyzed for residues of organochlorines. Similar to the levels from Lamont and Reichel (1970), the low levels were considered to
22 be baseline readings of background environmental contam ination that reflected no significant accumulation. Although at the time, no problems that might be associated with insecticides were identified, there was a lso the potential for sediment accumulation because some insecticides are strongl y absorbed by soils as previously mentioned (Rand et al., 2004). Drought Wetland habitats at the Refuge routinely experience drying events. When water levels drop below ground level, as a natural pa rt of Floridas wetland hydrologic regime, anecdotal evidence from snail kite observations suggests that apple snails have little to no tolerance to dry down co nditions (Turner, 1994; Darby et al., 2002; Karunaratne et al., 2006). Using radio-telemetry to monitor the movement of 78 snails from two locations, Darby et al. (2002) examined the movements of apple snails in response to water levels and drying events. Although snail movement tended to cease when water dep ths were below 10 cm, there was a statistically significant trend for apple sn ails to move toward refugia that remained inundated during a dry down. However, the potential enhanced survival from this behavioral trait bestows may be limi ted by the spatial extent of the dry down. Darby et al. (2002) predicted that as t he spatial extent of the dry down increases, apple snail stranding would increase propo rtionally. Strandings may increase, yet it is possible that apple snails could sur vive a dry down by aestivating based on anecdotal information (Turner, 1 994; Darby et al., 2002; Darby et al., 2008).
23 A potential link between snail survival during dry dow ns and reproduction also exists. As previously mentioned, the majority of ap ple snail egg production occurs between April and June, and according to Darby et al. (2008), an analysis of several South Florida wetlands showed that most dry downs occurred during a portion of this peak time, with 70% of dry-down event s lasting less than 12 weeks. Throughout the study, apple snails in flooded co nditions exhibited a range of normal activities, including mating and egg l ying, followed by a postreproduction die-off. Apple snails in dry down conditi ons ceased normal activities, including mating and egg laying, and appeared to aest ivate once burrowed. In comparison to earlier reports, Darby et al. (2008) concl uded that 70% of prereproductive adult apple snails were able to survive a 12-week dry down, but that juvenile snails were less likely to survive. Ultimately, Darby et al. (2008) concluded that a post-reproductive die off, not hydrolo gy, was for a dominant mechanism explaining apple snail survival patterns. How ever, the ultimate affect of a dry down on apple snail populations depends on th e proportion of the area gone dry, duration, and timing. If stranding occurs, du ring breeding season, recruitment and apple snail numbers have the potential to severely decline (Darby et al., 2002). Because of the effects of dry downs can have on apple snails, the viability of the snail kite population also depends on the time i nterval between droughts and their spatial extent (Mooij et al., 2002). Mooij et al. (2002) used a model to explore the difference in affects of system wide drought s and local droughts on snail kites as well as frequency. Not only did high drou ght frequency led to
24 reduced snail kite numbers, but that the spatial extent also had an effect (Mooij et al., 2002). During droughts of low intensity and spati al extent, snail kites are likely to have a behavioral response and, provided that suit able refugia are available, move to alternative sites where conditions are more fa vorable (Sykes, 1983; Mooij et al., 2002). During a system-wide drought, as the severity and spatial extent increase increases from a local drought, the snai l kites are unable to elicit this behavioral response and find alternative sites. The occurrence of a systemwide drought greatly increases the chance for reduced rep roduction and increased mortality (Mooij et al., 2002). Even though drought has the potential to have a direct negative impact on snail kite reproduction and survival, Mooij et al. (2002) also hypothesized that habitat degradation due to extended inundation would lead to a lower number of snail kites. The periodic drying that takes place in South Florid a benefits the ecosystem and helps to maintain the habitat structure (Ka runaratne et al., 2006). Prolonged inundation can lead to habitat degradation by converting a wet prairie habitat that contains emergent vegetation, necessary for snail survival, to a slough habitat that lacks emergent vegetation (Karunara tne et al., 2006). Karunaratne et al. (2006) examined the difference in snail density between a wet prairie and slough habitat and concluded that density w as greater in wet prairie habitats, frequently by a factor of two to three. In o rder to maintain these higher densities, periods of prolonged inundation should be avoided in order to prevent habitat conversion. Darby et al. (2008) agree that in creasing water depths and longer hydroperiods have been detrimental to wetland habitat and encourage
25 periodic drying events, occurring ever two to three yea rs, to support apple snail habitat. These drying events would have minimal impa ct on snail survival and recruitment, especially if the lowest water levels do n ot overlap peak egg production and dry down duration do not exceed six to eight weeks. Other Herbicides In addition to copper, various other herbicides were use d on the Refuge for vegetation maintenance control purposes as the inva sion of nonnative plants posed a serious threat to ecosystem health. In addition, floating vegetation such as water hyacinth and water lettuce could potentially i nterfere with the foraging success of snail kites, therefore control of these invasives also benefits the species (Rodgers et al;., 2001). The most commonly used he rbicides in the area have been 2,4-dichlorophenoxy acetic acid (referred to as 2,4-D) and 6,7dihydrodipyrido (1,2-a:2,1-c) pyrazinediium dibrom ide (referred to as diquat). Since the late 1940s, 2,4-D has been used in South Fl orida for the treatment of plants such as water hyacinth and water lettuce, mimicking natural plant growth hormones and causing lethally abnormal growth (Rodgers et al., 2001). First used in South Florida in 1955, diquat is a broad spectr um, fast-acting chemical that disrupts plant membranes (Rodgers et al., 2001). Diquat binds to organic matter and fine sediments and i s quickly removed from the water column (Pratt et al., 1997). Even tho ugh diquat is quickly removed from the water column, it could still pose a threat to the apple snail similar in the way that copper was transferred to the apple snail thr ough sediments. However,
26 while evaluating the effects of copper based herbicides o n apple snails, Winger et al. (1984) also evaluated the effects of diquat, use d singly and in combination with copper-based herbicides, on the apple snail. After discovering that combinations of the copper-based herbicides and diquat were only slightly more toxic, if any difference at all, than the copper-based herbicide alone, Winger et al. (1984) concluded that copper was the toxic agent in the copper-diquat combinations (Imlay and Winger, 1980). In addition to diquat toxicity testing, Refuge scientists conducted toxicity tests for various other herbicides use within the Refuge throughout the study period. Fire A wide range of ecosystems use prescribed fire as a manag ement tool (Harris and Whitcomb, 1974). Since the drainage of the Everglades, peat and soil subsidence has decreased land elevation, altered hydroper iods and affected the formation of tree islands. One of the causes of subsidence within the Everglades is fire. Fire management is necessary to combat fire destr uctive to the balance of the Everglades ecosystem (Ross et al., 2006). Using fire as a maintenance mechanism can provide many benefits to an ecosystem; some native plant species respond to fire by increasing their growth and r eproduction rate (Towne and Owensby, 1984) and fire has the potential to lim it the invasion of woody and exotic plants into native habitats (Pauly, 1985). However, research also exists that suggest that not all f ire is beneficial (Nekola, 2002). Nekola (2002) conducted a study regardin g the effects of fire
27 management on the richness and abundance of land snails in North American grasslands in regards to research findings that fire has b een implicated in the loss and/or reduction of many native invertebrate speci es. Similar to the effects of fire burning peat in the Everglades, one direct ef fect of prairie fire is the removal of the soil mulch layer. Since almost 90% of sn ails are located within 5 cm of the soil surface (Hawkins et al., 1998), the liveli hood of snails will undoubtedly be tied to the fate of the soil mulch lay er during fire events (Nekola, 2002). While apple snails are generally found at the soil-water interface they have been known to occasionally burrow under the soil su rface (Hoang and Rand, 2009), thus potentially affected by the impact of fire. Results concluded that both species richness and abundance wer e significantly lowered in areas that received fire manag ement. Notably, at a species-level, fire most strongly impacts the rarest specie s (Nekola, 2002). A fire affecting an area composed of a relatively small number of apple snails could have the potential to strongly affect that populatio n. While the study wasnt able to identify the factors that lead directly to these imp acts, it did help to narrow down the fact that these factors must be related to the soil mulch removal (Nekola, 2002). Although natural burning fires will a lways be an issue, especially in the Everglades, this study raised questions about the intervals that should be applied for prescribing fires.
28 Non-Avian Predation It has been demonstrated that multiple predators have effects on prey that cannot be predicted by examining the effects of single p redator types (Sih et al., 1998). When examining multiple predator effects, the concern is generally on predation rates. Multiple predators have the ability to bring on lower or higher predation rates if they are producing risk-reducing or risk-enhancing effects respectively (Sih et al., 1998). Predator-predator in teractions have the ability to reduce predation especially with higher predator densit y and a tendency for mutual interference among predators. Conflicting prey responses to multiple predators, could potentially enhance the risk of predat ion, meaning that prey response to one predator could result in a greater risk from another predator (Sih et al., 1998). For example, when the apple snail rel eases from a substrate and drops to the ground (Frakes et al., 2008) to escape pred ation from the snail kite, it becomes more vulnerable to predation by a land bas ed animal such as an alligator or turtle. In one example, in response to the presence of two predators, water striders ( Aquartus remigis ) reduced their overall activity as a defense, also drastically reducing their mating activity (Sih et al., 1998). However this reaction could prove risky as decreased mating activity would red uce an already vulnerable population.
29 CHAPTER 3: STUDY AREA Under the Migratory Bird Conservation Act of 1929, t he Loxahatchee Refuge was established in 1951 through an agreement be tween the South Florida Water Management District (SFWMD) and the U.S Fish and Wildlife Service (FWS) (FWS, 2000; FWS, 2007). The 143,874-acr e Refuge, located in Palm Beach County in southern Florida (FWS, 2007), is one of three water conservation areas (WCA) surrounded by canals and levees b uilt by the U.S. Army Corps of Engineers for water storage and manageme nt (USACOE) (Winger et al., 1984). Land within the Refuge is composed of different spatial units, including the interior of the Refuge or WCA-1, the m anaged Compartments A, B, C, and D and the Cypress Swamp Unit (FWS, 2007). The interior of the Refuge consists of sloughs, wet prairies, sawgrass, brush, and tress isl and habitats (FWS, 2007). Land surrounding the Refuge has various u ses. To the north and west of the Refuge is the Everglades Agricultural Area (EAA), a large portion of the northern Everglades that was drained for agricultu ral development that includes sugar cane farms and cattle ranches. Land to the e ast of the Refuge is comprised predominantly of urban area with the excepti on of some remaining small farms. To the south and southwest of the Refuge ar e Water Conservation Areas 2 and 3, and Everglades National Park (FWS, 2000 ; FWS, 2007). Although the original mission of the Refuge was to man age habitat for migratory
30 waterfowl, invasive exotic plant infestation and water quality and quantity issues (FWS, 2007) have helped change management focus towar d the protection and management of resident species and their habitat (Winge r et al., 1984). These three issues are becoming increasingly sensitive due to e ncroaching urban development, continued population growth, intensive a griculture production, and restoration projects (FWS, 2007). Climate Characteristic of a subtropical climate, the Refuge rece ives relatively humid summers with temperature averaging in the low 8 0s F and mild winters with average lows in the upper 60s F (FWS, 2007). Ra infall during the wet season, running from May to October, is produced by loc alized thunderstorms (FWS, 2000; FWS, 2007). Between the months of June an d September, these thunderstorms produce over one-half of the rainfall f or the year (FWS, 2000). Rainfall during the dry season, running from November to April, is the product of warm maritime or cold continental air masses. Yearly ra infall averages around 55-65 inches, however rainfall can vary extremely, fr om 35 inches in drought years to 120 inches in wet years (FWS, 2007). Topography, Soils and Landuse The topography of the Refuge is low relief with elev ations above mean sea level ranging from 17 feet at the northern tip o f the Refuge, to 11 feet at the southern tip (FWS, 2007). The limestone bottom of th e Refuge is covered with a
31 layer of soil ranging in depth from 3.6 to 14 feet ( FWS, 2000). The soil is primarily Loxahatchee Peat. An indicator of a historic slough commu nity, the peat is primarily formed from the decomposed roots, rootlets, a nd rhizomes of white water lilies (FWS, 2000). The light color, fibrous and spongy nature of the peat is an indicator of high organic content. The low frequency of burns in the area, whether prescribed or natural, is indicated by the low ash content of the soil. Compared to other peats, Loxahatchee Peat is slightly m ore acidic and has a lower mineral content (FWS, 2000). Land use surrounding the Refuge is agricultural, rural and urban. The primary land use within the Refuge is recreation. Int erpretation, nature observation, and fishing comprise most of the recreation al uses. There is also a portion of visitation that is comprised of boating/canoe ing/kayaking and waterfowl hunting (FWS, 2000). Figure 1A and 1B show the boundaries of the A.R.M. Loxahatchee National Wildlife Refuge as well as surroun ding land use in 1974 and 2004. A quick glance at the two figures shows the incr ease in urban and built-up use in 2004 representing the encroachment of t he urban environment to the Refuge. Hydrology and Water Quality Rainfall (56%), the S-5A pump station (40%), and ACM E 1 and 2 pump stations (4%) historically accounted for the Refuges maj or inflow sources of water. Presently, there are no more inflows of water from the ACME 1 and 2 pump stations, and the S-5A pump station moves water i nto one of two
32 stormwater treatment areas for phosphorus cleaning befor e entering the Refuge. Approximately 91% of water pumped into the Refuge i s originates in the Everglades Agricultural Area, while the remainder is f rom agricultural and urban development, subjecting the Refuge to the risk of decrea sed water quality due to agricultural and urban runoff (FWS, 2007). A cooperat ive agreement between the USACOE, SFWMD, and FWS has established the implement ation of a Water Regulation Schedule to manage water levels in the Re fuge. The current schedule is designed to meet five criteria: maintain the health of Refuge vegetation types by f looding all wetlands during the summer and fall, enhance feeding opportunities for waterfowl and wading birds by lower ing water levels in the spring so that water is concentrated in slo ughs and shallow ponds during nesting season, maintain water stor age capacity on the Refuge during the hurricane season, store water for irrigating nearby cropland during the fall, winter, and early spring, and prevent saltwater intrusion into the Biscayne aquif er by storing water for release into coastal canal systems during the fall, winter, and spring (FWS, 2007). Most of the inflows to the Refuge are sent through const ructed wetlands, referred to as stormwater treatment areas (STAs), with a small portion of inflows bypassing these STAs untreated (Harwell et al., 2008). High nutrient runoff, specifically phosphorus, from these lands poses serious threat s to the balance of the Refuge and the abundance and distribution of flor a and fauna. The spread of
33 undesirable plant species, cattails for example, negative ly affect the ecosystem by changing the faunal composition (FWS, 2000; Harwell et al., 2008). Another potential side effect of the high influx of nutrient rich drainage and runoff is eutrophication. The Everglades was established under low nutrient conditions (FWS, 2000; Payne et al., 2009). Because a large par t of the Refuges water budget is from canal inflows, the Refuge receives high phosphorus and nitrogen input higher than that found in rainfall, thus incre asing the risk of eutrophication (Payne et al., 2009). Additionally, the Refuge is a softwater ecosystem, with lower alkalinity and pH values than other waters withi n the Everglades Protection Area (Payne et al., 2009). Other contaminants that ma y lead to decreased water quality include mercury, pesticides, and other chemicals. ( FWS, 2000; Payne et al., 2009).
34 FIGURE 1A: Regional Perspective of 1974 USGS Land Use. Florida L and Use and Cover Classification System (FLUCCS) codes as define d by the Florida Department of Transportation (FDOT) for the A.R.M. Loxahatchee National Wildlife Refuge and surrounding areas.
35 FIGURE 1B: Regional Perspective of 2004 SFWMD Land Use.
36 CHAPTER 4: RESEARCH DESIGN Historical narrative data from the Loxahatchee Refuge was analyzed and used to determine whether or not various potential me chanisms for Florida apple snail ( Pomacea paludosa ) decline affected populations on the Refuge. In doi ng so, this study attempts to establish spatio-temporal rela tionships between apple snails and these various potential mechanisms. In the pr ocess, the differences in the style of which the data were reported and the qu ality that the data were reported over the study period was also examined. The applied research for this thesis attempted to answer a number of questions: Primary Question : What are the major contributing environmental stressors to the decline of Florida apple snails ( Pomacea paludosa ) within the Loxahatchee Refuge? Primary Hypothesis: The primary hypothesis related to the proposed research is that copper-based herbicide application, drou ght, and nonavian predation are the three main contributors to the decline of Florida apple snail numbers on the Loxahatchee Refuge.
37 Additional research questions and hypotheses explored in clude: Question: Can any spatio-temporal relationship between P. paludosa and the Everglade snail kite ( Rostrhamus sociabilis plumbeus ) be determined based on available information on p ast copper-based herbicide use and subsequent P. paludosa population trends? Hypothesis: There will be a significant spatio-temporal r elationship between copper-based herbicide use and the Florida ap ple snail within the Loxahatchee Refuge, specifically that abundance will decrea sed with increased copper-based herbicide application and distribut ion patterns will be associated with areas of least application. Question: Will analysis of roughly seven decades of data yie ld any noticeable differences in the style of which the da ta is reported and the quality that the data is reported? Hypothesis: As the study period progresses from 1951 to 20 07, the quality of the data and level of detail reported with in the Annual Narratives will improve. The overall objectives of the proposed thesis include: analysis of the Refuges Annual Narratives for informa tion relating to the application of copper-based herbicides and other herbicides
38 for vegetation maintenance purposes, the use of insecti cides, the occurrence of drought and fire, and non-avian predatio n; developing an assessment of the historical abundance and distribution of Florida apple snails based on the Loxah atchee Refuges Annual Narratives for information relating t o Florida apple snails; incorporating a distillation of recent copper and drou ght related apple snail findings into the synthesis on the ecology o f the Florida apple snail; assessing the quality of historical data reporting; providing management recommendations to Refuge rega rding future population management techniques; proposing (and potential contribution to) initial est ablishment of a baseline Florida apple snail monitoring network in the Refuge interior. Figure 2 depicts initial assumptions regarding each stressors potenti al effect on apple snail population decline within the Re fuge. Based on knowledge gained from the literature review, the use of copperbased herbicides and subsequent toxic effect and drought may have strong dir ect effects on the decline of the apple snail. Although some literature exists per taining to the application of insecticides and other herbicides, not enough information is available to determine that these factors have a strong direct effe ct. Therefore it is assumed they will have a weak direct effect. Due to a current l ack of knowledge regarding
39 the effect of fire on the apple snail and the degree to which apple snails are predated by multiple predators, the effects of these str essors are unknown at the time. Additionally, it is predicted that based on the Everglade snail kites dependence on the Florida apple snail as its primary f ood source, the observed trends of the snail kite will coincide with the observed trends of the apple snail. FIGURE 2: Initial Florida Apple Snail Stressors Conceptual Model The content of this thesis contributes to the scientific lit erature regarding the decline of Florida apple snails on the Loxahatchee Refuge by taking a holistic approach and examining the potential effects of multip le factors. An important aspect of this research includes examining the influence of copper toxicity on the Florida apple snail. While much of the current publ ished scientific literature concerns itself with implications for Everglades restorati on with regards to the potential for copper desorption from flooded agricultu ral soils with a large portion of studies being conducted in a laboratory setting, this thesis discusses and evaluates the effects of past copper-based herbicide appli cations, insecticide
40 use, the occurrence of drought, the use of other herbicid es, the occurrence of fire, and non-avian predation on Florida apple snail abundance and distribution within the northern Everglades. Completion of the th esis provides information in regards to management strategies for Florida apple snai ls from a historical perspective.
41 CHAPTER 5: METHODOLOGY Copies of the Loxahatchee Refuges Annual Narratives we re obtained through contact with Refuge scientists and managers. The A nnual Narrative reports are yearly reviews of various aspects of the Refu ge including highlights, climate conditions, land acquisition, planning, administ ration, habitat management, wildlife, public use, equipment and facil ities, and any other items of importance. Authors of annual narratives include refuge managers and staff. The Annual Narratives date from 1951 to 2007. From these a historical timeline of knowledge on the abundance and distribution of Florid a apple snails and Everglade snail kites was developed. These records were use d to assess whether or not the potential mechanisms of decline are a factor for apple snails on the Refuge. The Refuges Annual Narratives were an alyzed for information relating to the application of copper (e.g., copper sul fate) and other herbicides for vegetation maintenance purposes, the application of in secticides for control of nuisance pests, and any mention of drought, fire and n on-avian predation that could have potential effects on the apple snail populat ions. Because findings may have important implications on the foraging success of th e Everglade snail kite, the Annual Narratives were analyzed for information r egarding Everglade snail kite abundance and distribution in relation to that of the Florida apple snail and other suitable habitat areas such as Lake Okeechobee and Wa ter Conservation
42 Area 3. Within the Annual Narratives are sightings and annual surveys containing observations of interest including differentiating betw een male and female snail kites, the approximate location of the sighting and nest ing activity, information pertaining to the apple snail, and various programs an d studies related to snail kites and apple snails. In addition to drought information found within the Annual Narratives, rainfall data received from Refuge Senior Hydrologist Dr. Mike Waldon and surface water level data from the South Florida Water Management Districts (SFWMD) DBHYDRO environmental database (http://www.sf wmd.gov/dbhydrop plsql/show_dbkey_info.main_menu) were analyzed. Analysis of these datasets will provide accurate drought information regarding t he fluctuations of the surface water level. The time series used for the rainfall dat a was from 1965-2005 and comes from the South Florida Water Management Model ( SFWMM). The SFWMM is a computer model that stimulates rainfall, ev apotranspiration, infiltration, flow, seepage, and groundwater pumping over 7600 mi 2 using a 2 mi by 2 mi grid. The SFWMM has been accepted by many age ncies as the best available tool for analyzing data of its kind (SFWMD, 2010). The SFWMDs DBHYDRO database includes hydrologic and water qualit y data recorded from various sites (SFWMD, 2006). As was described in the narra tives, site gauge 1-7 was the best indicator of interior marsh water level an d site gauge 1-8C was the best indicator of canal water level. Although there ar e other gauges within the Refuge, the focus was placed on these two sites for anal yzing water level. For site 1-7, time series 06713 as reported by the USACOE from 01/01/1954 to
43 12/31/1989 and P1029 as reported by the SFWMD from 0 1/01/1979 to 06/30/2007 was used. For site 1-8C, time series 06711 as reported by the USACPE from 07/25/1954 to 12/31/1990 and P1030 as re ported by the SFWMD from 01/01/1978 to 06/30/2007 was used. As described in Chapter 3, the wet season runs from May through October and the dry season r uns from November through April on the Refuge. In order to correlate w ith these water years, the annual water year is defined as from May 1 of one yea r through April 31 of the following year (e.g., Water Year 2007 ran from May 1, 2006 through April 31, 2007). In addition, to determining the mean monthly surface water level for site gauges 1-7 and 1-8C, the mean wet and dry season surface water levels were calculated. For rainfall, the mean wet and dry season r ainfall were calculated. A historical timeline was created for apple snail abun dance and distribution within the Loxahatchee Refuge summarized as best as possible for particular management units that are spatially defined by the Annual Narratives. Annual Narratives were analyzed from 1951 to 2007 to establish baseline information on the abundance and distribution of appl e snails in relation to the potential factors contributing to decline; copper-based herbicides, other herbicides, drought, insecticides, fire, and non-avian pr edation. Each factor was examined by decade. A review of the existing literatu re, or qualitative study, was used to: fill voids in existing knowledge as best possible, establish a new line of thinking, and assess an understudied issue (Furniss, 2001; Wi lson, 2007). This qualitative study provided a description, contextual un derstanding, and explanations of the theme under observation from a holistic approach (Wilson,
44 2007). Generally, traditional retrospective analytical methods involve relating the pattern of changes in abundance with the pattern of v ariation in the underlying vital rates based on available demographic data (Cooch et al., 2001). However, since no annual demographic data are available for Flo rida apple snail populations within the Loxahatchee Refuge for the per iod being examined, this thesis determined and interpreted the potential pathw ays leading to the decline of or general low numbers of Florida apple snails. Table 1 provides an example of the table used in the Florid a Apple Snail Trends section of Chapter 6 as an overview of all the components examined in order to provide a general reference point of factors that may have affected apple snail decline and indirectly snail kite numbers. For each year, a general description will be provided for each factor in regards to that factors use or presence on the Refuge that year. Table 2 provides definitions of the abbreviation and terminology used within the table. TABLE 1: Example Overview of Components Examined Table. Year Factors Apple Snail Snail Kite Copper Insecticides Drought Other Herbicides Fire Non Avian Predation
45 TABLE 2: Overview of Components Examined Table Terminology. Factors Term Definition Apple Snail Good Arbitrary term used in the Annual Narratives to refer to the number of apple snails present; not necessarily based on actual numbers or representative of the entire Refuge Low Arbitrary term used in the Annual Narratives to refer to the number of apple snails present; not necessarily based on actual numbers or representative of the entire Refuge Early In the beginning of the year Late In the later part of the year Snail Kite Total Statewide population Observed/ Observatio n Slightings ambiguous due to the nature of use in the Annual Narrative; casual sightings; not defined how these sightings are related to the number established on the Refuge Estimated An approximation of the number established on the Refuge Peak of Maximum number established on the Refuge Nesting Whether successful or not, nesting has taken place on the Refuge in some capacity Good Arbitrary term used in the Annual Narratives to refer to the number of snail kites present generally in relation to the previous year Low Arbitrary term used in the Annual Narratives to refer to the number of snail kites present generally in relation to the previous year Drought Drought A year in which drought conditions have persisted on the Refuge for any amount of time No Drought A year in which drought conditions have not persisted on the Refuge for any period of time Other Herbicides exp. Experimental use of herbicides Fire Prescribed Fire which is planned and started intentionally as part of the management plan Natural Fire which starts out of the control of man, generally as a result of lightning All Factors No Data The Annual Narrative was unavailable for analysis for that year and no data were able to be gathered from any mention of that year in other Annual Narrative documents Not noted Annual Narratives did not report any data
46 In order to convey the significance of the findings, a conceptual ecological model was developed. The conceptual ecological model is a non-quantitative expression that identifies the anthropogenic drivers and stressors on the natural system within the Loxahatchee Refuge and the biologica l attributes or indicators of these ecological responses (Ogden et al., 2005). The conceptual ecological model will essentially be a working hypothesis of how a nthropogenic and natural processes affect the ecological components of the study site. Included are descriptions of the source of the contaminant, whether n atural or anthropogenic, receiving environment, and the process by which the recep tors come to be exposed directly to the contaminants and secondarily to t he effects of the contaminants on other environmental components (Suter, 1996). Figure 3 provides an example of a basic version of the model. Co nceptual models are used as planning tools to guide and focus scientific suppo rt and to build understanding and consensus regarding the working hypot heses that explain the sources and effects of anthropogenically induced environm ental changes (Ogden et al., 2005). FIGURE 3: Conceptual Ecological Model Diagram. A simplified exam ple of the model as depicted by Ogden et al. (2005). For Example, Galiulin et al. (2005) used a conceptua l model of the ecological risk assessment to evaluate the toxicity of persi stent
47 organochlorinated compounds (POCs), which contaminate the Caspian Sea. The model consisted of sources of POCs, a comparison of concentra tions in source waters compared to maximum allowable concentrations, be havior of the toxic compounds in water, and factors responsible for increasing risk. The model produced can be used to evaluate the toxicity of POCs in the environments in which they occur. Odgen et al. (2005) produced a concep tual ecological model of the effects of anthropogenic stressors on Everglades systems as a tool to be utilized with the Everglades restoration program. Dri vers of the model include urban and agricultural expansion, industrial and agricu ltural practices, water management practices, and human influences on species compo sition. The attributes include periphyton, marsh plant communities, tree islands, marsh fish, invertebrates, reptiles and amphibians, alligators, and wading birds. The drivers could produce potential stressors, such as reduced spatial ex tent, degraded water quality, reduced water storage capacity, and compa rtmentalization, that could then produce an effect on the attributes. One exa mple includes water management practices as the driver, which produces the st ressor of reduced water storage capacity, the resulting effects include short ened hydroperiod, in turn effecting alligator nesting and causing reduced occu rrence, which in turn leads to altered distributing and reduced abundance of native aquatic animals, the leading to the attribute alligator (Ogden et al ., 2005). From the model, Odgen et al. (2005) was able to formulate research question s based on casual relationships within the model.
48 Additionally, in order to determine whether the ant hropogenic and natural processes examined are significantly affecting the populat ions of Florida apple snails within the Loxahatchee Refuge, it will be deter mined whether the apple snails show a biological response or an ecological effect. A biological response refers to an individual-level response to a stressor whil e an ecological effect refers to the collective biotic responses that are capable of affecting the various higher levels of ecological organization, the populati on level in this case (Gentile and Harwell, 1998). An individual-level response coul d potentially affect the population-level if there is a larger effect of repro duction or if so many individuals are affected that a large scale effect becomes evident ( Harwell and Gentile, 2006). In order to determine ecological significance, whether a change in [Florida apple snail numbers] is o f sufficient type, intensity, extent, and/or duration to be impor tant to the structure, function and/or health of the system and whe ther such a change exceeds natural variability or alters the natura l variability regime (Harwell and Gentile, 2006) needs to be examined.
49 CHAPTER 6: RESULTS Florida Apple Snail Trends Overview Throughout the study period, a pattern was observed t hat during periods of drought, the status of the apple snails was less known a nd drought was having a potential negative effect on snail numbers. In 1965 apple snails appeared to be concentrating in the canals and in alligator holes d uring dry periods. After drought conditions in 1956, 1662 and 2000, egg cluster s were observed showing that to some extent, apple snails were surviving drough t. A study conducted in 1974 revealed that apple snail egg cluster counts were higher in the interior than in the canals, agreeing with the observation that app le snail numbers in the canals had greatly declined. Refuge scientists and manage rs used the increased observations of the Refuges snail kite numbers to indicat e that apple snails were in abundance at least in some areas of the Refuge. In o rder to examine the trends of the apple snail in more detail, the results h ave been broken down into cohorts by decade. 1951-1959 Table 3 shows an overview of all the components examined in o rder to provide a general reference point of factors that may have affected apple snail
50 decline and indirectly snail kite numbers from 1951 to 1959. From this table, only two years within the 1950s yielded any information re garding apple snail data. One of which years the data reported good apple snail numbers coincided with a drought. For both years that there was some mention of good apple snail numbers, no snail kites were observed. Throughout the p eriod it is evident that there were generally low observations of snail kites on the Refuge. Cells labeled not noted mean that the Annual Narratives did not report any data regarding population numbers, application of insecticides or herbi cides, or that there were not documented fires. For each year within the period, the Refuge received some degree of herbicide application or experimental use ( denoted as exp. in the table). During the 1956 drought, the narrative alludes to th e fact that with the disappearance of major water supplies comes concern for the potential impact on apple snail reproduction. However, it is mentioned th at the freshwater snail is still present as eggs of the snail were commonly observed. Know ing this, Refuge staff was assured that at least a portion of this once abu ndant food was in survival stages. In 1959 freshwater snail eggs were nume rous by about March 13. From May to August, freshwater snails were abundant on the Refuge.
51 TABLE 3: Overview of Components Examined from 1950 to 1959. Year Factors 1951 1952 1953 1954 1955 Apple Snail Not noted Not noted Not noted Not noted Not noted Snail Kite Use increasing 0 kites observed 6 kites observed 0 kites observed 0 kites observed Copper Not noted Copper sulfate used Periodic use of copper occurred throughout the 1950 s and was applied to experimental cultivated crops and to far mland in Compartment B Insecticides Not noted Not noted Not noted Not noted Not noted Drought Drought No Drought No Drought No Drought Drought Other Herbicides 2,4-D 2,4-D 2,4-D; VL-600; exp. 2,4-D; exp. 2,4-D; exp. Fire Not noted Natural fire 4480 acres of natural fires Not noted 11,100 acres of natural fires Non Avian Predation The threat of non-avian predation is an ongoing fac tor; during periods of drought alligator density in the canals increases Year Factors 195 6 1957 1958 1959 Apple Snail Good Not noted Not noted Good Snail Kite 0 kites observed 3 kites observed 3 kites observed 0 kites observed Copper Periodic use of copper occurred throughout the 1950 s and was applied to experimental cultivated crops and to farmland in Co mpartment B Insecticides Not noted 120 acres treated Not noted Toxaphene use in millet fields Drought Drought Drought Drought No Drought Other Herbicides 2,4-D; exp. 2,4-D; Dalapon; exp. 2,4-D; exp. 2,4-D; Dalapon; exp. Fire Not noted Not noted Not noted Not noted Non Avian Predation The threat of non-avian predation is an ongoing fac tor; during periods of drought alligator density in the canals increases 1960-1969 Table 4 shows an overview of all the components examined in o rder to provide a general reference point of factors that may have affected apple snail decline and indirectly snail kite numbers from 1960 to 1 969. During this period, two years reported good apple snail numbers that were not associated with drought and two years reported apple snail numbers fi rst being unknown, but then returning. During these two years, drought was pr esent one, but not the other. This decade saw increased snail kite use over the l ast decade and the presence of insecticide use nearly every year. As with th e previous decade, for
52 each year within the period, the Refuge received some degree of herbicide application or experimental use. In 1962, another dry year, the status of the freshwater snail is less known than the previous decade. Trips into the interior rev ealed many empty shells, yet time alone would be able to reveal how severe of an effect the dry conditions had on the population. Later that year however, the nar ratives stated that the snail population was coming back. Improved water condition in 1964 aided in the continued comeback of apple snail populations. Low water conditions in 1965 again affected apple s nail populations. During periods of low water the apple snail appeared to be concentrated in the canals and gator holes located in the interior that wer e able to retain water and serve as refugia. Once water levels rose and were adequa te enough for transportation to the interior, a large number of em pty snail shells were present. It was difficult to determine whether these empty shells we re a result of predation by limpkins or mortality caused by drought. Although th e empty shells were present, many egg clusters were also present indication so me survival possibly by aestivation or dispersal from canals and alligator ho les by wind and water action. The drought of 1967 combined with that of 1965 led to the Refuge being unstable habitat conditions within the Refuge. Despite this, the Annual Narratives note that the apple snail can apparently survive these p eriods of low water levels. While the apple snail was commonly observed in 1968, the re was little data to base actual population numbers on. Although the abunda nce is not known,
53 Refuge scientists and managers used the increased observati ons of the Refuges snail kite numbers to indicate that apple snails are abun dant at least on the south end of the Refuge. TABLE 4: Overview of Components Examined from 1960 to 1969. Year Factors 1960 1961 1962 1963 1964 Apple Snail Not noted Not noted Early: Empty shells status unknown Late: coming back Not noted Good Snail Kite 0 kites observed 2 kites observed; low due to drought 4 kites observed; low due to drought Population of 7 kites; nesting; #s increasing Population of 15 kites; nesting Copper Not noted Not noted Not noted Not noted Not noted Insecticides Toxaphene in Compartment C; DDT Not noted Malathion in Compartment C Sevin used throughout the 1960s in Compartments C and B Drought No Drought Drought Drought No Drought No Drought Other Herbicides 2,4-D 2,4-D; diesel 2,4-D; Amitrole-T 2,4-D; AmitroleT; Dalapon 2,4-D; Amitrole-T Fire Over 5800 acres of natural fire Not noted 11,000 acres of natural fire Not noted Not noted Non Avian Predation The threat of non-avian predation is an ongoing fac tor; during periods of drought alligator density in the canals increases Year Factors 1965 1966 1967 1968 1969 Apple Snail Early: Empty shells status unknown Late: coming back Not noted Not noted Good Not noted Snail Kite 0 kites observed Use increasing Use increasing Use increasing Peak observation of 15 kites Copper Not noted Not noted Not noted Not noted Not noted Insecticides Sevin used throughout the 1960s in Compartments C a nd B Drought No Drought No Drought No Drought Abundance of water Abundance of water Other Herbicides 2,4-D 2,4-D 2,4-D 2,4-D; diquat 2,4-D; diquat Fire Not noted Not noted Prescribed burn in impoundments; natural fire in the interior Not noted Roughly 12 acre of natural fire Non Avian Predation The threat of non-avian predation is an ongoing fac tor; during periods of drought alligator density in the canals increases 1970-1979 Table 6 shows an overview of all the components examined in o rder to provide a general reference point of factors that may have affected apple snail
54 decline and indirectly snail kite numbers from 1970 to 1 979. Results from the study reported in 1974 revealed higher apple snail de nsity in the interior than in the canals. Drought was observed during both years that these data were collected revealing that while the canals may have prov ided refugia as the Refuge interior had dry conditions, densities were stil l higher in the dry interior than in the canal. Apple snail numbers varied greatly, generally decreasing with drought. However, in 1974 when drought was present o n the Refuge, there was an estimated maximum of 39 kites using the Refuge at som e point and nesting. As with the previous decade, for each year within the p eriod, the Refuge received some degree of herbicide application with the majority being only 2,4-D and diquat. It is evident that the apple snail was forced undergrou nd by 1970 spring drought conditions; however, the absence of census techn iques and qualifying procedures make it impossible to estimate apple snail num bers. Instead, it was observed that the numbers of egg clusters were significan tly less than the previous year. In 1972 an effort was made to better u nderstand the apple snail and what the egg cluster density and distribution were revealing about the population. Apple snail numbers on the Refuge were ge nerally lower than previous years and lower than in other areas of the sta te receiving snail kite use. Three apple snail collections were made at night yield ing a best of 10 snails collected per hour and a worst of five snails collected pe r hour. In the short run, a perennial aquatic situation would promote snail produ ction, but would ultimately destroy the habitat by means of loss of emergent vegeta tion due to a lack of
55 periodic drying. In order to establish stable apple snai l numbers great enough to support a snail kite population, impounded areas were u tilized to stimulate snail growth and reproduction through managed water levels and aquatic vegetation. A study conducted in 1974 showed that fertilization of aquatic habitats had potential for increasing apple snail growth and reprod uction as well as decreasing the time required by the snails to reach sex ual maturity. A 20-20-5 fertilizer was used to stimulate production under nat ural conditions. In addition to the fertilizer study, the 1974 narrative also described snail conditions from sampling. Transects were established in slough edges of th e Refuge interior, but they were not truly comparable to the impoundments wh ere vegetation is homogeneous. The counts from the slough edges represented concentrations of egg laying by snails whose home ranges extended into th e sloughs. Results of the egg cluster sampling are shown in Table 5 reveal that snail concentrations were higher in the Refuge interior than in the canal s. With earlier observations from 1963-1967 indicating that apple snail population density was once higher in the canals, it was speculated that water hyacinth spray op erations, runoff and pesticide residues from adjacent agricultural lands had ne gatively affected apple snail populations in the canals. Although no exact coordi nates were provided, Figure 4 provides the general location of the transects for the 1973 and 1974 egg cluster counts. Points 1, 4 and 5 represent the canal transects and points 2, 3 and 6 represent the interior transects.
56 TABLE 5: Florida Apple Snail Egg Cluster Counts. Canal Interior Interior Canal Canal Interior Date Transect 1 Transect 2 Transect 3 Transect 4 Transect 5 Transect 6 8/1973 0 30 37 3 2 No Data 10/1973 4 35 23 4 4 12 11/1973 1 11 8 0 3 5 12/1973 1 1 5 1 0 1 1/1974 0 2 7 0 0 No Data 2/1974 0 38 20 0 0 37 3/1974 2 43 23 0 1 43 4/1974 3 23 7 2 2 27 5/1974 0 3 13 0 1 15 6/1974 Dry Dry 13 0 3 6 7/1974 7 20 34 7 5 27 FIGURE 4: Interior and Canal Transect Locations.
57 Observations indicate that the snail kite consumes roughl y 35 to 40 snails per day according to personal communication between Ref uge staff and Paul Sykes as captured in the 1974 Annual Narrative. Althoug h apple snail numbers were low on the Refuge, kite use was high during perio ds of 1974 because of poor conditions in other traditional areas occupied by t he kites across the Florida peninsula. As management of the impoundments allowed the habitat to be maintained and persist over time, apple snail numbers w ere described in the Annual Narratives as declining as the predator-prey r elationship came into balance. In order to promote the maximum apple snai l populations, Compartments A, B and C were converted to long term w et-dry cycles. A year-long fertilizer tank study in 1977 found that growth rates were accelerated roughly 25% and the time to meet sexual ma turity went from 20 to 24 weeks to 12 to 16 weeks. This figure is important because it could potentially mean a quicker recovery time of the population after a drought. When this study was taken to the field, results were similar, but it was also noted that there was evidence of a shortened life span for the snails. In a n attempt to further understand the effects of fertilizer on apple snails, dolomite lime, agricultural lime and hydrated lime were applied to impoundments in Co mpartment C to compare the action of each in increasing productivity and alle viate the mortality rate issue. After the completion of the non-avian predator fence in 1977, 3,000 and 4,500 apple snails were stocked in Impoundment C-4 on April 20 and July 19 respectively. In 1978, after a year of observations, tw o impoundments that had
58 been treated with lime consistently produced higher sna il numbers than other impoundments. TABLE 6: Overview of Components Examined from 1970 to 1979. Year Factors 1970 1971 1972 1973 1974 Apple Snail Spring droughts forced underground Not noted Low Higher density in interior than canals Higher density in interior than canals Snail Kite Estimated 54 kites on Refuge; nesting Peak of 11 kites; loss of 80 due to drought Peak of 11 kites No Data Estimated 39 kites on Refuge; nesting; Copper Not noted Not noted Not noted Not noted Not noted Insecticides Sevin used throughout the 1970s in Compartment C Drought Abundance of water Drought Drought Drought Drought Other Herbicides 2,4-D; diquat 2,4-D; diquat 2,4-D No Data 2,4-D Fire Small natural fire Not noted Not noted No Data Not noted Non Avian Predation The threat of non-avian predation is an ongoing fac tor; during periods of drought alligator density in the canals increases Year Factors 1975 1976 1977 1978 1979 Apple Snail Not noted Not noted Stocking in C4 in April and July Not noted Not noted Snail Kite Estimated 21 kites on Refuge; nesting Peak of 10 kites Peak of 8 kites; total of 152 kites Good Good; total of over 400 kites Copper Not noted Not noted Not noted Not noted Not noted Insecticides Sevin used throughout the 1970s in Compartment C Drought No Drought No Drought No Drought No Drought No Drou ght Other Herbicides 2,4-D 2,4-D 2,4-D 2,4-D; diquat 2,4-D; diquat Fire 90 acres of natural fire in the interior Not noted Not noted Prescribed burn in Compartment C Natural fires in the interior Non Avian Predation The threat of non-avian predation is an ongoing fac tor; during periods of drought alligator density in the canals increases 1980-1989 Table 7 shows an overview of all the components examined in o rder to provide a general reference point of factors that may have affected apple snail decline and indirectly snail kite numbers. Both 1982 and 1983 saw good apple snail numbers in Impoundment C-5, possibly due to the m anagement efforts in that unit to promote apple snail reproduction and pro vide emergency habitat for
59 the snail kite. Selected impoundments were designated a s emergency snail kite habitat because they were managed to provide a stabl e water level even during times of drought. In the event of a drought, the snai l kites would be able to use this area when no other suitable habitat was available Particularly interesting is the sharp decrease in snail kites in 1987 that does not ap pear to be associated with drought. As with the previous decade, for each yea r within the period, the Refuge received application of various herbicides. The 1 980s also saw an increase in both natural and prescribed fire. During 1982 and 1983, the 20 acre Compartment C-5 s ustained high apple snail numbers. In a continuation from previous t ransects set up to monitor apple snail reproduction and obtain an index that cou ld be related to actual population numbers, 29 transects were monitored on a m onthly basis for an unspecified time period. Thirteen transects were place in the impoundments, six in the interior, five in Lake Okeechobee, and five in WCA 3. Each transect was 544 by 4 or roughly 0.05 acres and all the unhatched egg clusters were counted along each transect. With over eight years of transect data, Refuge scientists were unable to correlate egg cluster counts to snail popu lations. While the original result was not achieved, preliminary results su ggested that reproduction may have been affected by water levels. Throughout th e remainder of the 1980s, apple snail transects were maintained, but eventually discontinued without mention of results.
60 TABLE 7: Overview of Components Examined from 1980 to 1989. Year Factors 1980 1981 1982 1983 1984 Apple Snail Not noted Not noted Good numbers in C-5 Good numbers in C-5 Not noted Snail Kite Total of over 500 kites Peak of 10 kites; dispersed across state; total decreased to under 250 Peak of 12 kites; total of 302 kites Peak of 7 kites; nesting; total of 437 kites Total of 668 kites Copper Not noted Not noted Not noted Not noted Not noted Insecticides Not noted Not noted Not noted Not noted Not noted Drought Drought Drought Drought No Drought No Drought Other Herbicides Diquat Diquat; Rought-Up 2,4-D; diquat Velpar-L; diquat Rodeo; exp. Velpar-L; diquat Rodeo; exp. Fire Not noted 6640 acres of natural fire Prescribed burns in the impoundments Not noted Not noted Non Avian Predation The threat of non-avian predation is an ongoing fac tor; during periods of drought alligator density in the canals increases Year Factors 1985 1986 1987 1988 1989 Apple Snail Not noted Not noted Not noted Not noted Not noted Snail Kite Peak of 20 kites; dispersed across state; total decreased to 407 Total of 563 kites 22 reports of kites; total decreased to 326 Mediocre use 13 snail kite sightings on the Refuge Copper Not noted Not noted Not noted Not noted Not noted Insecticides Not noted Not noted Not noted Not noted Not noted Drought No Drought No Drought No Drought Drought Drought Other Herbicides Velpar-L; diquat; Arsenal; exp. Velpar-L; diquat; Arsenal; exp Velpar-L; diquat; Arsenal; exp Velpar-L; diquat; Arsenal; exp Velpar-L; diquat; Arsenal; exp Fire Some natural and prescribed burns; only small amount of acreage Natural fires and 32 acres of prescribed burns Over 3621 acres of prescribed burns Prescribed burns Prescribed fire turned wild 40,000 acres burned Non Avian Predation The threat of non-avian predation is an ongoing fac tor; during periods of drought alligator density in the canals increases 1990-1999 Table 8 shows an overview of all the components examined in o rder to provide a general reference point of factors that may have affected apple snail decline and indirectly snail kite numbers from 1990 to 1999. Apple snail numbers were mentioned only in 1997, a year not associated wit h drought, when they were reportedly low. As water levels decreased in 1997, apple snails temporarily disappeared for an unreported amount of time. Snail kite numbers seemed to be generally good and there were only two years that recei ved some degree of
61 drought, 1990 and 1991. However the 1990s narratives had a lack of snail kite reporting which had been better reported in previous decades. The 1990s also saw an increase in use of various herbicides, a trend th at appeared to begin in the mid-1980s. The increase in use of herbicides other th an 2,4-D and diquat most likely occurred in an effort to treat new invasive species that posed a threat to the Refuge and were not necessarily aquatic, such as melaleuca and Brazilian pepper. TABLE 8: Overview of Components Examined from 1990 to 1999. Year Factors 1990 1991 1992 1993 1994 Apple Snail Not noted Not noted Not noted Not noted Not noted Snail Kite No Data Total decreased to 372 Increased kite observations; nesting Nesting Peak of 17 kites; nesting Copper Not noted Not noted Not noted Not noted Not noted Insecticides Not noted Not noted Not noted Not noted Not noted Drought Drought Drought No Drought No Drought No Drought Other Herbicides Rodeo; exp. Rodeo; diquat; Garlon 4; exp. Rodeo; Round-Up; diquat; Arsenal; exp. Rodeo; diquat; diesel; ecp. Rodeo; diquat; others less effective Fire Not noted Not noted Prescribed burns mostly in the impoundments Not noted 1530 acre of natural fire Non Avian Predation The threat of non-avian predation is an ongoing fac tor; during periods of drought alligator density in the canals increases Year Factors 1995 1996 1997 1998 1999 Apple Snail Not noted Not noted Low Not noted Not noted Snail Kite Scattered sightings; nesting Kite use in interior and Compartment C Majority of observations in interior; nesting; decreased use at times Nesting Few kites observed; use of interior Copper Not noted Not noted Not noted Not noted Not noted Insecticides Not noted Not noted Not noted Not noted Not noted Drought No Drought No Drought No Drought Abundance No Droug ht Other Herbicides Rodeo; diquat; others less effective Rodeo; Reward Rodeo; Reward Rodeo; Reward; Arsenal; RoundUp; 2,4-D Rodeo; Reward; Arsenal; Round-Up; 2,4-D Fire 500 acres of natural fire Not noted Not noted Not noted Not noted Non Avian Predation The threat of non-avian predation is an ongoing fac tor; during periods of drought alligator density in the canals increases
62 2000-2007 Table 9 shows an overview of all the components examined in o rder to provide a general reference point of factors that may have affected apple snail decline and indirectly snail kite numbers from 2000 to 2007. In a 2004 assessment of the impoundment management efforts in C-6, C-7, C-8, and C-9, very few apple snails were sampled. While C-8 had prev iously received a large egg transplant, it had snail densities no higher than in the other impoundments at the time of sampling. Two years reported some apple sna il data, 2000 which was a drought year and reported low apple snail numbers, and 2004 which was also a drought year, but reported good apple snail numbers. As with the 1990s, the Refuge received application of various herbicides of th e control of invasive exotic vegetation. Additionally, the use of prescribed fires c ontinued to increase throughout the decade. Compared to other water conservation areas in 2000, ap ple snail egg clusters were much lower on the Refuge. A study was conduct ed in 2002 to estimate apple snail density in Impoundment C-7 and C8 and the headwater pond. Fourteen transects were counted using wire trap an d throw trap surveys. For the six C-7 transects, a total of three egg clusters w ere observed and only three snails were found in the ditches yielding an equi valent density of 0.05 snails/m 2 For the four C-8 transects, no egg clusters were observe d and the number of snails found was not mentioned. In the headq uarters pond, egg clusters were found in a relatively high density of 0.3 3 snails/m 2 and seven apple snails were found. Although dry conditions had initial ly prevented surveying the
63 interior, rising water levels led to the ability to sur vey the northeast portion of the interior and the finding of six egg clusters deposited within 10 days of increased water levels. While sampling may have been constraine d by low water levels, results from the interior indicated that apple snails cou ld recover from dry-down conditions and immediately produce egg clusters. TABLE 9: Overview of Components Examined from 2000 to 2007. Year Factors 2000 2001 2002 2003 Apple Snail Low Not noted Not noted Not noted Snail Kite Frequent use of interior and impoundments Sightings varied; 21 kites observed in one day Frequent use of interior and impoundments Frequent use of interior and impoundments Copper Not noted Not noted Not noted Not noted Insecticides Not noted Not noted Not noted Not noted Drought Drought Drought No Drought No Drought Other Herbicides Rodeo; Reward; Arsenal; Escort; 2,4-D Rodeo; Reward; Arsenal; Garlon 3A; Round-Up; ; 2,4-D Rodeo; Reward; Arsenal; 2,4-D Rodeo; Reward; Arsenal; 2,4-D Fire 3,000 acres of natural fire and one prescribed burn 10,454 acres of natural fire Natural fires and prescribed burns in the impoundments 2,400 acre prescribed burn in the interior Non Avian Predation The threat of non-avian predation is an ongoing fac tor; during periods of drought alligator density in the canals increases Year Factors 2004 2005 2006 2007 Apple Snail Good Not noted Not noted Not noted Snail Kite Use low especially during nesting season Use mediocre Frequent use of interior and impoundments Use mediocre Copper Not noted Not noted Not noted Not noted Inse cticides Not noted Not noted Not noted Not noted Drought Drought No Drought No Drought Drought Other Herbicides Rodeo; Reward; Arsenal; 2,4-D Rodeo; Reward; Arsenal; 2,4-D Rodeo; Reward; Arsenal; 2,4-D Rodeo; Reward; Arsenal; 2,4-D Fire 641 acres of natural fire and 6,553 acres of prescribed burns in the interior plus addition burns in the impoundments 75 acres of natural fire and prescribed burns in the impoundments 20 acres of natural fire and over 500 acres of prescribed burns 7,000 acres of natural fire and 13,000 acres or prescribed burns Non Avian Predation The threat of non-avian predation is an ongoing fac tor; during periods of drought alligator density in the canals increases
64 Everglade Snail Kite Trends Overview Statewide Everglade snail kite numbers started out at 25 in 1951 at the start of the study period and saw extreme fluctuations in health and numbers according to periods of drought throughout the Evergla des. During periods of drought the snail kite would migrate to surrounding ar eas such as Lake Okeechobee and the other WCAs. The confirmed presence o f seven kites in 1963, seen near the south end of the Refuge, and thre e nests was the first notation of the Refuge contributing to the larger sna il kite population. The Refuge saw varying nesting success with young often lost to preda tion or poor weather. The establishment of a snail kite management area in Co mpartments B and C in 1971 provided compartments that were managed to stimul ate apple snail growth and reproduction and maintain conditions suitable for snail kite nesting and feeding. By 1977, the estimated statewide snail kite po pulation was over 152 and by 1979 it was estimated to be over 400. Severe drou ght that struck in 1981 forced the snail kites to disperse across the state in search of food and what was a healthy statewide population of over 650 decreased t o less than 250 as a result of high mortality and little reproduction. Snail kite numbers began to recover and nesting took place in new areas in 1982, such as Lake Toho pekaliga. Drier conditions that occurred in 1985 again caused snail kites t o disperse and saw a 39% decrease in statewide population. For the first tim e in 15 years there was nesting on the Refuge in 1992 that produced four fled glings. A peak population of 17 snail kites was observed on the Refuge in 1994, the highest since 1975. In
65 1997, all but two of the snail kite observations occurred in the Refuge interior. Snail kite observations throughout the 2000s yielded varying numbers of observations. In order to examine trends of the number of Everglade snail kites utilizing the Refuge, the results have been broken dow n into cohorts by decade. 1951-1959 There were several observations of snail kites on the Ref uge and use of the area by the species has been increasing. Three kites w ere observed from January to April and another three kites were observed from September to December of 1953. With very few snail kites in the are a, no kites are observed again until the fall of 1957 with rare observations on three occasions. A pair of snail kites was observed within the interior and a single bird was viewed from the headquarters. It became apparent that during periods o f drought snail kites were moving to Lake Okeechobee and other areas in the St ate. A pair of snail kites was observed near the office between January and April of 1958 and one was observed between September and December of 1959. The annual narrative notes that the only hope for salvaging the snail kite and providing suitable habitat is the continued betterment in environment within th e interior. 1960-1969 The sole observation of two snail kites between January and April of 1961 led to the observation that because of the loss of feed ing areas associated with low water levels, these birds left the Refuge for unkn own destinations. The
66 following year, 1962, came to similar conclusions after only three snail kites were observed between January and April. From September to December, one snail kite was observed along C-40 and a second was seen multip le times near S-6. For the first time in Refuge narratives, the confirmed presence of seven kites in 1963, seen near the south end of the Refuge, and three nests began the establishment of the Refuge as providing habitat for ki te reproduction. At the time, Refuge scientists and managers believed that the snail kite colony from Lake Okeechobee had moved on to the Refuge. The narra tive described this group of birds as the only known colony in the continent al United States at the time. While the snail kites may have been in less danger of being shot on the Refuge, water levels would influence their nesting success and survival on the Refuge. As the year progressed, the snail kites remained on the Refuge and their numbers seemed to be increasing, if only by a little. F ive nests were observed, two of which contained eggs or young, but of which only one live young was seen. By 1964 the peak count of snail kites was 15 with birds b eing seen near the S-6 pump station and near the south end of the Re fuge where they had previously been sighted. One nest was seen and thought t o produce two fledgling that survived. The nest was located in a two acre willow head which also contained anhinga and green heron nests. However, the following year went without any snail kite sightings on the Refuge. At the time when kites normally return to the Refuge and nesting activity would typi cally occur, the Refuge had began to dry up and was then completely dry. An exten sive aerial survey found 8
67 kites in Water Conservation Area (WCA) 2A and two on L ake Okeechobee. By 1966, kites were seen on the Refuge and it became appar ent that they were moving back and forth between the Refuge and WCA 2A. Although there was kite use on the Refuge, no nests were observed. For t he first time since 1955, snail kites were observed south of the Tamiami Trail, at Everglades National Park (ENP) in 1967. Another extensive aerial survey re ported over 40 kites in four locations WCA 2A, WCA 3, ENP, and Lake Okeechobee. WCA 3, in addition to ENP, was another new area being occupied by the kit e. As the 1960s came to an end, 1968 and 1969 saw increased snail kite sightings on the Refuge. Although no nesting attempts w ere made, kites were frequently observed along the Hillsboro Canal, with a peak observation of 15 kites in this area in 1969, and north of the Refuge he adquarters. Over a threeweek period, 31 kites were observed over the Refuge. 1970-1979 For the first time in five years, the snail kite nested in the Refuge in 1970. Forty-two kites were seen and an estimated 54 were on the Refuge in early January. Although snail kite numbers decreased and remai ned low throughout the summer, by November numbers were up again with 3 1 kites recorded. That year, 11 nests yielded eight young to flight stage. N esting began during high water levels, but became stranded as water levels decrea sed, also exposing the nests to increased predator pressure.
68 While 1971 saw increased management attention for the snail kite, the population was dispersed across the state with a peak numbe r of 11 sighted near the Flying Cow Pump Station. Increased management on the Refuge led to the establishment of the snail kite management area in Comp artments B and C. Water levels in the kite management area were held be tween 15.68 and 16.19 feet above mean sea level (AMSL). It was determined t hat these depths were the best to attempt to reproduce natural Everglades slough depths to get a good mix of submergent and emergent aquatic vegetation, to stim ulate apple snail growth and reproduction and maintain conditions suitable for snail kite nesting and feeding. During 1972 a peak of 11 kites were observed u sing the Refuge. Observations occurred in the management and flats of the L-40 canal. In 1974, the Refuge had a peak of 39 snail kites with eight nests on or adjacent to the management area. Although eight nests were built, no young was produced to flight stage because of unfavorable weather conditions a nd predation. It was determined that the drought that occurred in 1971 led to the death of over 80 snail kites statewide. The year 1975 was yet again unsuccessful for nesting at tempts, although two young were produced, they were both lost to preda tion. In addition to the unsuccessful production of young on the Refuge, there w as poor nesting success throughout South Florida. The peak population dropp ed to 21 on the Refuge in 1975. While some decline in kite population was expected because of the natural aging of the management area, there was a slight incr ease in the total day use.
69 By 1976, Compartments A, B and C had been converted to long term wetdry cycle management to produce the maximum amount of apple snails, management directed toward producing a habitat with t he greatest snail kite carrying capacity. The only exception was impoundment C7 which was maintained as a moist soil area to provide additional waterfowl and wading bird habitat. After a peak population of 10 snail kites in 1976, 1977 had a peak population of 8 kites with a stable population of 5 kit es in impoundment C-5. Snail kites were also seen using a willow head in C-3 as a ro osting site, C-4, and feeding along the flats at the Refuges south end. Al though its most likely some were not counted, the estimated statewide snail kite num ber was 152, up 10 from 1976. Both 1978 and 1979 proved to be encouraging years f or the snail kites. Although no nests were found on the Refuge in 1978, mating and nesting behavior was observed as use of the management area cont inued. Additional use was also noted in the southwest corner of the Refuge. By 1979, the statewide kite population was estimated to be well over 400. While the snail kite was seen feeding in the management area, there was ye t again no nesting on the Refuge. The narratives indicated that successful nest ing and reproduction was most likely taking place at WCA 3 and Lake Okeechobee. 1980-1989 With population continuing to rise throughout South F lorida, the statewide snail kite population was estimated to be around 500 wi th the majority of the
70 population concentrated in WCA 3. Kites use continues t o occur in the management area and southwest corner of the Refuge. As severe drought struck South Florida wetlands in 1981, kites were forced to di sperse across the state in search of food. Abundant rainfall in the previous few years has provided increased wetlands for foraging, but as rainfall decrease s and water levels dropped, suitable snail kite habitat decreased as did re maining sources of food. In addition to annual kite surveys, the Everglades snail kite alert hotline supplemented the documentation of the movement of kit es to more urban areas during the drought and their use of canals, borrows pi ts and flooded farm fields. Although the snail kite posses nomadic habits by nature, t he drought took a sever toll on their population. As major habitats, su ch as WCA 3 and Lake Okeechobee, could no longer support the snail kites, they were forced to massively disperse. In addition to the drought, the lo ss of small seasonal wetlands to drainage forced the kites out of the protect ion of remote marshes and into more urban areas. Pre-drought 1981 conditions bo asted a statewide snail kite population of over 650, by the end of the year numbers fell to less than 250. In addition to high mortality, there was little repr oduction, as few as four fledglings, since most adults did not attempt to nest and for those who did failure was high. Although the water level was held in Compar tment C, roughly 300 acres in size, the overall lack of water on the Refuge lent little attraction to the Refuge. Seven kites were seen on the Refuge and the pe ak use was estimated at 10 birds.
71 With the drought over by spring 1982, the snail kite population began to recover. While little nesting took place in WCA 3 and L ake Okeechobee, the majority took place on Lake Kissimmee and Lake Tohopekal iga, the first time nesting was recorded in this area. With 10 impoundments being managed as emergency kite habitat, as previously described, the Refu ge saw increased use in 1982 and a peak bird count of 12. While the kites no rmally moved onto major breeding grounds in the fall, they did roost in a wil low head to the west of Compartment C. Snail kite numbers continued to rise in 1983 and were up to 437 from, 302 in 1982. In addition to improved numbers, 1983 also proved to be the first successful nesting season since 1980. Ninety-two percent of snail kites surveyed were located in WCA 3A, which is where the maj ority of the nesting took place. Snail kite numbers peaked at seven, down fr om the previous year. Although increased numbers meant a positive outlook for the kite, the single large concentration made the kites extremely vulnerable to environmental changes, any adverse conditions, such as another drought, could severely impact the population. Continued kite recovery in 1984 was supported by wet co nditions leading to successful nesting. Snail kite numbers continued to cl imb to 668. With the majority of snail kites populating the usual areas, inclu ding the addition of WCA 2B, most kite use on the Refuge was attributed to feedi ng birds moving through the Refuge. Yet as quickly as the snail kite population was renewed, below normal rainfall in the spring of 1985 again forced th e kites to disperse and decreased nesting success and reproduction. Little reproduct ion combined with
72 significant mortality led to a 39% decrease in populati on, down to 407, from the pervious year. The snail kite once again scattered in sea rch of canals, small marshes and impoundments, yet snail kite use on the Refu ge increased during this time and peaked at 20. In years following, kite u se on the Refuge was limited with observations of between one to four birds in the impoundments, northeastern edge and northwest interior of the Refug e. During the mid-eighties the snail kite began roosting in a single shell rock pit that was also a bird rookery and roost to many wa ding birds. This shell rock pit was located in wetlands north central Palm Beach Co unty just east of the West Palm Beach Water Catchment Area that contained sloughs, wet prairies, and tree islands. At times, more than half of the known snail kite population was found here. With this area being a potential site for kites to return to in times of drought as well as a proposed site for a landfill, a qu arter mile buffer was recommended. In 1986 the total estimated snail kite pop ulation was up to 563, but the population decreased again in 1987 to 562 snai l kites, down by 42%. 1990-1999 As the 1990s began, the majority of snail kite sighting s, 1300 of 1338, were located at the shell rock pit. While Compartment C continued to be popular for kite use, the annual kite survey once again showed a decrease in numbers, down to 372. High water levels in 1992 on the Refuge led to a high use by the kites. Snail kite observation was three times higher tha n the previous year and
73 for the first time in about 15 years there was nesting on the Refuge that produced four fledglings. Snail kites continued to take advantage of communal ro ost sites and nested in both the southeast and southwest portion of the Refuge producing four nests in 1993. A peak number of 17 snail kites was observe d in the Refuge in 1994. Although this number seemed promising, out of fo ur nests only one successfully fledged two young. It was suspected however, t hat the kites were probably just using the Refuge as a corridor for trav el. In 1997, all but two of the snail kite observations occ urred in the Refuge interior. Like previous years, nesting attempts were u nsuccessful. It was presumed that that years nesting attempts failed due t o cold fronts. The 84% drop in snail kite numbers from April to May was thoug ht to be a reflection of the uncontrolled drop in water levels of the interior. Ye t as the narratives have documented, 1998 yielded a turn around with in successf ul nests with a record high number of 13 active nests found. 2000-2007 As with the late 1990s, the new decade continued with sn ail kites frequently being observed in the Refuge interior and the impoundments. In 2001, 19 kites were observed on one day and 21 on another. I n addition to a roost being utilized by the snail kites by the boat launch, kit es were also seen making a new roost in impoundment C-8, an area being managed for apple snail reproduction and providing habitat for the kites. The mid-nineties saw scattered
74 snail kite use throughout the Refuge in various numbers with sightings frequently occurring in impoundments C-7, C-8, C-9, and C-10, the roost near the boat ramp, the interior, and along Lee Road. Surveys of the interior in 2006 for snail kites on thr ee different occasions yielded zero kites on the first trip in March and 58 a nd 57 birds flying into the roost in the evening near the boat ramp on two separa te occasions in June. Two nests located in the interior were both deemed to hav e failed. Other observations were made throughout the year in other areas of the Refuge as well including 30 kites in impoundment C-1, two kites flying over the cyp ress swamp, kites using the Loxahatchee Impounded Landscape Assessment (LILA), a nd kites using the southeast interior. In 2007, 60 kites were observed flyi ng from roosting locations in the interior to the impoundments in the Fall Migr atory Bird Count, 18 kites were counted in the Spring North American Migration Bird C ount and 27 snail kites it the Christmas Bird Count. Copper Overview Copper use was first recorded on the Refuge in 1952 and use continued periodically throughout the 1950s as copper was applied to experimental cultivated crops and farmland in Compartment B. In 19 75 it was discovered that there was very little snail productivity in the canals co mpared to the very high apple snail numbers that occurred there during the mid -1950s. A connection was made that the use of copper in the canal water for th e control of aquatic
75 vegetation could have potentially led to this decrease Testing results supported this and revealed higher copper concentrations in the ca nal than in the interior. It was also discovered that Cutrine-Plus, a copper-based he rbicide, was toxic to adult apple snails at 0.1 parts per million (ppm) and to juvenile apple snails, two to four weeks old, at 0.034ppm. Sediment samples collecte d revealed levels below the TEL guidelines of 18.7 mg/kg for all sites sam pled. In order to examine the application of copper more in depth, the results ha ve been broken down into cohorts by decade. 1951-1959 In 1952, copper was first recorded as being used on the R efuge in the Annual Narrative documents. Copper sulfate was applied to experimental plantings of domestic rice. Periodically throughout the 1950s copper was applied to experimental cultivated crops and to farmland in Compartment B. 1960-1969 During this period nothing was reported in the Annua l Narratives regarding the use of copper based herbicides on the Refuge. 1970-1979 After an informal review of several canals in 1975, it was discovered that there was very little apple snail productivity in the ca nals. Compared to these findings, a canal near Refuge headquarters yielded sever al thousand apple
76 snails from a small area, in one day in the mid-1950s. Three apple snail transects beginning in this same canal yielded only one count of an index of one with the rest being zero in several years of monthly counts. Afte r the discovery of a paper in which it was noted that aquarium water carried to t he aquarium in copper lines had toxic effects on the apple snail, the connection wa s made that large quantities of copper had been added to the canal water s to control aquatic vegetation. Following this discovery, studies of the eff ects of copper on the apple snail began to emerge. None of the Annual Narrative after the 1950s describe the application of copper in the Refuge. The first study conducted aimed to determine the concen tration of copper in sediment and vegetation at various sites on the Refu ge in relation to pesticide use and apple snail abundance and a laboratory study examining the sub-acute toxicity of copper and diquat-copper herbicides on the a pple snail. Preliminary results revealed that copper was present at a higher con centration in the canal than in the interior of the Refuge and that Cutrine -Plus, a copper-based herbicide, was toxic to adult apple snails at 0.1 parts p er million (ppm). 1980-1989 Continued studies on the effects of copper on apple sna ils revealed that Cutrine-Plus was toxic to juvenile apple snails, two to four weeks old, at 0.034ppm. When examining the effects of copper-diquat on apple snails, although mortality resulted when extremely high concen trations were applied within closed tanks, typical field applications did not cau se death.
77 1990-1999 During this period nothing was reported in the Annual Narratives regarding the use of copper based herbicides on the Refuge. 2000-2007 Sediment samples collected from 15 sites within Refuge i mpoundments A, B and C were tested for copper concentrations. Copper w as found in all 15 samples with concentrations ranging from 1.6 to 15.5 mg/ kg. However these values did not cross the Threshold Effect Concentration or Level as set by the Florida Department of Environmental Protection sedime nt quality assessment guidelines of 18.7 mg/kg (Schuler et al., 2008). Insecticides Overview In 1956 the use of insecticides began on the Refuge. To xaphene, DDT, Malathion, and Sevin were used mainly for the treatm ent of armyworms. DDT was first used in 1960 and residue analysis began in 1965 Insecticide testing continued throughout the study period in 1987, 2000 and 2002. For the most part, concentrations found in the species and samples teste d were low and thought to reflect levels of background environmental contamination. In order to examine the use of insecticides and residue analysis more i n depth, the results have been broken down into cohorts by decade.
78 1951-1959 In 1956, 120 acres of land within the Refuge were sp rayed for armyworm, however there was no mention of the type of insecticide used. Again in 1958, aerial spraying of Toxaphene was used to treat armyworm s in the millet fields. 1960-1969 Toxaphene was again used in 1960 to treat armyworms i n Compartment C, the cultivated crop area at the time, with a 5% ae rial spray. In addition to Toxaphene, 5% DDT was also used at a rate of 33lbs per acre. In 1963, Compartment C once again had issues with armyworms and an aerial spray of Malathion was used to treat the issue. Armyworms continue d to be a reoccurring problem among the cultivated crops. A ban was issued agai n the aerial application of the insecticide Sevin. Although there wa s a ban on aerial application, ground application of Sevin was used at r oughly one to one and a half pints per acre. The use of Sevin as a treatment fo r armyworms continued throughout the 1960s in Compartment C as well as in C ompartment B. Residue analysis in 1965 examined three groups of 10 a pple snails for insecticide residues. Each group was analyzed for chlorinate d hydrocarbons. Three insecticides were found, DDE, DDT and benzene he xachloride gamma isomer at an average of 0.068ppm, 0.110ppm and 0.011 ppm respectively. No mention was made as to the potential toxic effects of th ese levels of insecticides on apple snails.
79 1970-1979 The use of Sevin continued in Compartment C for army worm control throughout the 1970s. 1980-1989 A study conducted in 1987 examined the presence of or ganic residues in birds, anhingas, Anhinga anhinga, and little green herons, Egretta caerulea collected from several rookeries and fish, largemouth ba ss, Micropterus salmoides and lake shubsucker, Erimyzon sucetta collected from the perimeter canals. Preliminary evaluation indicated a slight orga nochloride insecticide contamination in birds and fish within the Refuge. T he primary contaminants were DDE, the metabolite of DDT, and Toxaphene. Whi le there was some variation among rookeries, the mean concentration of D DE was 0.15 g/g and 0.16 g/g and the mean concentration of Toxaphene was 0.11 g/g and 0.43 g/g for anhingas and little blue herons respectively. It was deemed likely that the concentrations in the birds were reflective of backgro und levels through the area. The concentration of DDT and Toxaphene found w ithin the fish sampled were below the FDA action level for these compounds, 5 g/g, yet exceeded the national means in the National Pesticides Monitoring P rogram, 0.29 g/g DDT and 0.27 g/g Toxaphene. During the time of the study, the eff ect of these mild contaminations on the Refuge was unknown, yet the level s indicated some type of agricultural contamination. Since the use of both co mpounds was banned, it was predicted that theoretically, their concentrations sh ould reduce over time.
80 1990-1999 During this period nothing was reported in the Annua l Narratives regarding the use of insectidices on the Refuge. 2000-2007 In 2000, sediments from five sites in the impoundmen ts and cypress swamp were sampled and analyzed for organochlorines. W hile Toxaphene was not detected, as was thought that it might be prior to testing due to reports that it occurred on farmlands adjacent to the Refuge, DDT meta bolites DDD and DDE were detected in the cypress swamp, the canal near Bend er Farm and in the southern portion of C-7. Compared with the FDEP Sedi ment Quality Assessment Guidelines, the highest concentration of DDD detecte d, 39 g/g, exceeded the Probable Effects Level (PEL). Above the PEL, contamina nt concentrations could have an adverse effect on aquatic organisms. Remaining concentrations were above the Threshold Effects Level (TEL), the level bel ow which concentrations would not be expected to elicit adverse biological effe cts, but below the PEL meaning that concentrations between these two levels cou ld potentially lead to adverse biological effects. While it was ruled out that Toxaphene was not a contaminant of concern and that the low contaminant le vels in the canal near Bender Farm and C-7 probably represented background le vels for the area, further testing of the cypress swamp for DDD would be needed to determine the extent of contamination.
81 Further sampling in 2002 of the Strazulla March, cyp ress swamp and impoundments all yielded some level of organochlorine concentrations. Sediments samples from Strazulla Marsh found all seven o f the sites sampled to contain DDD, DDE or both with a DDD range of 4.4 to 32.4 ppb and a DDE range of 3.0 to 16.2 ppb. Similar results were found for th e seven sites sampled in the cypress swamp, DDD ranging from 1.5 to 8.6 ppb and DDE ranging from 4.5 to 10.5 ppb, and for the 19 sites sampled within the impo undments, DDD was detected a one site and DDE at 10. However, for all t hree locations, the levels found were not expected to adversely impact the enviro nment. Drought Overview Drought conditions persisted on the Refuge in 1951, fr om the end of 1955 to early 1957, from the end of 1961 to early 1962, mid-1963, 1971 to early 1972, from 1973 to early 1974, from June 1980 to March 198 2, from the end of 1988 to early 1991, from late 2000 to early 2001, 2004, and 2007. Higher than average water levels occurred throughout parts of 1968, 1969 an d 1970 and during the dry season at the beginning of 1998. The Refuge strugg led with meeting the minimum and maximum water requirements set by the wa ter schedule due to urban and agricultural demands and at times poor timin g of water releases in addition to the natural occurrences of drought. In ord er to examine the fluctuation and deviations in surface water level and determine when drought
82 conditions were persistent on the Refuge, the data have been broken down into cohorts by decade. Wet and Dry Season Hydrologic Analysis Figure 5 shows the locations of the main surface water gauges on t he Refuge. Site 1-7 is most representative of interior sur face water levels and site 18C is more representative of canal surface water levels. For each site, the mean wet and dry surface water level was calculated from data acquired from SFWMDs DBHYDRO. As seen in Table 10 and Table 11 for both sites the dry season for the period of record examined has a higher m ean surface water level than the wet season. This is most likely attributed to the fact that in anticipation of low rainfall during the dry season, water levels are h eld higher. Whereas during the wet season, when the majority of the rainfall is e xpected, the water levels are held lower. For the most part, the periods of drought that occurred on the Refuge arent shown as readily in the mean wet and dry season surface water levels. The drought that began during the dry season in 1988 with a surface water level of 14.923 ft AMSL is the most obvious for site 1-7. Alt hough there was some initial trouble maintaining water level in the canal s before the completion of the levels, as is seen in the initial years of Table 11 the droughts that occurred on the Refuge are much more evident for site 1-8C.
83 FIGURE 5: Location of Surface Water Gauges by Site ID. Within the Refuge, site 1-7 provides the best representation of overall inter ior surface water levels and site 1-8C provides the best representation of canal surf ace water levels.
84 TABLE 10: Mean Wet and Dry Season Surface Water Levels for 1-7. Wet Season Mean Surface Water Level (ft) Dry Season Mean Surface Water Level (ft) May 1954 October 1954 16.292 November 1954 Apri l 1955 15.970 May 1955 October 1955 15.772 November 1955 April 1956 15. 643 May 1956 October 1956 15.272 November 1956 April 1957 15.767 May 1957 October 1957 16.123 November 1957 Apri l 1958 16.436 May 1958 October 1958 15.982 November 1958 Apri l 1959 15.983 May 1959 October 1959 16.147 November 1959 Apri l 1960 16.143 May 1960 October 1960 16.397 November 1960 April 1961 16.216 May 1961 October 1961 15.157 November 1961 April 1962 No Data May 1962 October 1962 16.277 November 1962 April 1963 15.995 May 1963 October 1963 15.107 November 1963 April 1964 15.283 May 1964 October 1964 15.852 November 1964 April 1965 16.348 May 1965 October 1965 15.463 November 1965 April 1966 16.120 May 1966 October 1966 16.157 November 1966 April 1967 16.125 May 1967 October 1967 15.490 November 1967 April 1968 15.721 May 1968 October 1968 16.123 November 1968 April 1969 16.420 May 1969 October 1969 16.140 November 1969 April 1970 16.703 May 1970 October 1970 15.897 November 1970 April 1971 15.09 0 May 1971 October 1971 15.2 77 November 1971 April 1972 16.318 May 1972 October 1972 16.153 November 1972 April 1973 15.365 May 1973 October 1973 15.332 November 1973 April 1974 16.165 May 1974 October 1974 15.758 November 1974 April 1975 15.935 May 1975 October 1975 15.455 November 1975 April 1976 16.048 May 1976 October 1976 15.988 November 1976 April 1977 15.885 May 1977 October 1977 15.788 November 1977 April 1978 16.317 May 1978 October 1978 15.935 November 1978 April 1979 16.267 May 1979 October 1979 15.798 November 1979 April 1980 16.248 May 1980 October 1980 15.808 November 1980 April 1981 15.627 May 1981 October 1981 15.823 November 1981 April 1982 15.952 May 1982 October 1982 16.272 November 1982 April 1983 16.615 May 1983 October 1983 16.330 November 1983 April 1984 16.612 May 1984 October 1984 16.128 November 1984 April 1985 16.065 May 1985 October 1985 16.070 November 1985 April 1986 16.535 May 1986 October 1986 16.348 November 1986 April 1987 16.461 May 1987 October 1987 15.900 November 1987 April 1988 16.328 May 1988 October 1988 16.048 November 1988 April 1989 14.923 May 1989 October 1989 14.960 November 1989 April 1990 15.487 May 1990 October 1990 15.956 November 1990 April 1 991 16.065 May 1991 October 1991 16.217 November 1991 April 1992 16.202 May 1992 October 1992 16.117 November 1992 April 1993 16.678 May 1993 October 1993 16.113 November 1993 April 1994 16.405 May 1994 October 1994 16.265 November 1994 April 1995 16.730 May 1995 October 1995 16.585 November 1995 April 1996 16.737 May 1996 October 1996 16.570 November 1996 April 1997 16.482 May 1997 October 1997 16.563 November 1997 April 1998 17.103 May 1998 October 1998 16.318 November 1998 April 1999 16.628 May 1999 October 1999 16.292 November 1999 April 2000 16.530 May 2000 October 2000 16.042 November 2000 April 2001 16.057 May 2001 October 2001 16.243 November 2001 April 2002 16.615 May 2002 October 2002 16.390 November 2002 April 2003 16.613 May 2003 October 2003 16.600 November 2003 April 2004 16.505 May 2004 October 2004 15.823 November 2004 April 2005 16.165 May 2005 October 2005 16.278 November 2005 April 2006 16.373 May 2006 October 2006 16.325 November 2006 April 2007 16.260 Wet Season Period of Record 15.991 Dry Season Period of Record 16.179 The bolded wet and dry seasons represent periods wh en drought conditions occurred. Blue shading represents the water regulation schedule oc curring from July 1960 to June 1969, green shading represents the schedule occurring from July 1969 to June 1975, red shading represents the schedule occurring from July 1975 to April 1995 and purple shading represents the schedule occurring from May 1995 to 2007.
85 TABLE 11: Mean Wet and Dry Season Surface Water Levels for 1-8C. Wet Season Mean Surface Water Level (ft) Dry Season Mean Surface Water Level (ft) May 1953 October 1953 13.450 November 1953 Apri l 1954 12.387 May 1954 October 1954 13.143 November 1954 Apri l 1955 12.203 May 1955 October 1955 12.505 November 1955 April 1956 10.908 May 1956 October 1956 11.200 November 1956 April 1957 11.760 May 1957 October 1957 13.927 November 1957 Apri l 1958 13.527 May 1958 October 1958 12.727 November 1958 Apri l 1959 12.918 May 1959 October 1959 14.932 November 1959 Apri l 1960 12.745 May 1960 October 1960 14.872 November 1960 April 1961 15.048 May 1961 October 1961 13.652 November 1961 April 1962 11.920 May 1962 October 1962 14.298 November 1962 April 1963 15.715 May 1963 October 1963 14.652 November 1963 April 1964 15.648 May 1964 October 1964 15.678 November 1964 April 1965 16.278 May 1965 October 1965 14.995 November 1965 April 1966 16.225 May 1966 October 1966 15.892 November 1966 April 1967 15.970 May 1967 October 1967 14.713 November 1967 April 1968 15.835 May 1968 October 1968 16.112 November 1968 April 1969 16.615 May 1969 October 1969 15.835 November 1969 April 1970 16.862 May 1970 October 1970 15.290 November 1970 April 1971 14.153 May 1971 October 1971 14.775 November 1971 April 1972 16.655 May 1972 October 1972 15.760 November 1972 April 1973 15.230 May 1973 October 1973 15.425 November 1973 April 1974 1 6.298 May 1974 October 1974 15.63 November 1974 April 1975 15.885 May 1975 October 1975 15.582 November 1975 April 1976 16.265 May 1976 October 1976 15.795 November 1976 April 1977 15.557 May 1977 October 1977 15.668 November 1977 April 1978 15.937 May 1978 October 1978 15.732 November 1978 April 1979 15.752 May 1979 October 1979 15.660 November 1979 April 1980 16.350 May 1980 October 1980 15.383 November 1980 April 1981 14.422 May 1981 October 1981 13.918 November 1981 April 1982 15.328 May 1982 October 1982 15.890 November 1982 April 1983 16.217 May 1983 October 1983 15.508 November 1983 April 1984 15.962 May 1984 October 1984 14.983 November 1984 April 1985 15.417 May 1985 October 1985 15.262 November 1985 April 1986 15.760 May 1986 October 1986 14.950 November 1986 April 1987 15.960 May 1987 October 1987 14.850 November 1987 April 1988 15.945 May 1988 October 1988 14.985 November 1988 April 1989 13.007 May 1989 October 1989 13.6 70 November 1989 April 1990 14.133 May 1990 October 1990 15.080 November 1990 April 1991 15.563 May 1991 October 1991 15.878 November 1991 April 1992 16.068 May 1992 October 1992 15.690 November 1992 April 1993 16.462 May 1993 October 1993 15.630 November 1993 April 1994 16.323 May 1994 October 1994 15.843 November 1994 April 1995 16.803 May 1995 October 1995 16.372 November 1995 April 1996 16.680 May 1996 October 1996 16.208 November 1996 April 1997 16.507 May 1997 October 1997 16.497 November 1997 April 1998 16.965 May 1998 October 1998 15.977 November 1998 April 1999 16.402 May 1999 October 1999 16.207 November 1999 April 2000 16.713 May 2000 October 2000 16.065 November 2000 April 2001 15.665 May 2001 October 2001 15.830 November 2001 April 2002 16.637 May 2002 October 2002 15.983 November 2002 April 2003 16.622 May 2003 October 2003 16.528 November 2003 April 2004 16.598 May 2004 October 2004 15.552 November 2004 April 2005 16.250 May 2005 October 2005 15.987 November 2005 April 2006 16.440 May 2006 October 2006 16.137 November 2006 April 2007 15.922 Wet Season Period of Record 15.162 Dry Season Period of Record 15.397 The bolded wet and dry seasons represent periods wh en drought conditions occurred. Blue shading represents the water regulation schedule oc curring from July 1960 to June 1969, green shading represents the schedule occurring from July 1969 to June 1975, red shading represents the schedule occurring from July 1975 to April 1995 and purple shading represents the schedule occurring from May 1995 to 2007.
86 In addition to examining surface water level, mean w et and dry season rainfall was also examined as seen in Table 12 While as expected, the dry season has a lower mean rainfall. As with Table 10 the drought events arent always apparent from the mean season rainfall. Yet t he droughts that began in the 1970 and 2000 dry seasons are with rainfall fallin g below a mean of 1.00 in. Droughts are also apparent during the wet season with the occurrence of abnormally low rainfall such as in 1972, 1987, 2000, and 2002.
87 TABLE 12: Mean Wet and Dry Season Rainfall. Wet Season Mean Wet Season Rainfall (in) Dry Season Mean Dry Season Rainfall (in) May 1965 October 1965 5.917 November 1965 April 1966 2.388 May 1966 October 1966 9.003 November 1966 April 1967 1.078 May 1967 October 1967 7.551 November 1967 April 1968 1.634 May 1968 October 1968 8.278 November 1968 April 1969 2.816 May 1969 October 1969 7.686 November 1969 April 1970 3.912 May 1970 October 1970 5.344 November 1970 April 1971 0.744 May 1971 October 1971 6.229 November 1971 April 1972 3.266 May 1972 October 1972 4.672 November 1972 April 1973 2.109 May 1973 October 197 3 6.266 November 1973 April 1974 1.536 May 1974 October 1974 7.061 November 1974 April 1975 1.372 May 1975 October 1975 7.787 November 1975 April 1976 1.868 May 1976 October 1976 5.420 November 1976 April 1977 1.920 May 1977 October 1977 5.797 November 1977 April 1978 2.589 May 1978 October 1978 7.082 November 1978 April 1979 2.324 May 1979 October 1979 5.965 November 1979 April 1980 3.554 May 1980 October 1980 5.641 November 1980 April 1981 1.466 May 1981 October 198 1 6.299 November 1981 April 1982 2.762 May 1982 October 1982 7.257 November 1982 April 1983 3.768 May 1983 October 1983 6.228 November 1983 April 1984 2.437 May 1984 October 1984 5.638 November 1984 April 1985 1.720 May 1985 October 1985 6.982 November 1985 April 1986 2.550 May 1986 October 1986 6.639 November 1986 April 1987 2.718 May 1987 October 1987 4.613 November 1987 April 1988 5.154 May 1988 October 1988 5.448 November 1988 April 1989 1.527 May 1989 October 198 9 5.373 November 1989 April 1990 1.534 May 1990 October 1990 5.974 November 1990 April 1991 3.888 May 1991 October 1991 5.509 November 1991 April 1992 2.150 May 1992 October 1992 6.823 November 1992 April 1993 3.801 May 1993 October 1993 5.768 November 1993 April 1994 2.180 May 1994 October 1994 8.872 November 1994 April 1995 3.930 May 1995 October 1995 6.042 November 1995 April 1996 2.135 May 1996 October 1996 6.754 November 1996 April 1997 2.392 May 1997 October 1997 5.492 November 1997 April 1998 4.064 May 1998 October 1998 5.211 November 1998 April 1999 2.479 May 1999 October 1999 6.648 November 1999 April 2000 1.174 May 2000 October 2000 4.603 November 2000 April 2001 0.760 May 2001 October 2001 6.858 November 2001 April 2002 1.910 May 2002 October 2002 4.690 November 2002 April 2003 2.360 May 2003 October 2003 3.973 November 2003 April 2004 1.705 May 2004 October 2004 5.603 November 2004 April 2005 1.376 May 2005 October 2005 6.512 Wet Season Period of Record 6.232 Dry Season Period of Record 2.376 The bolded wet and dry seasons represent periods wh en drought conditions occurred. Blue shading represents the water regulation schedule oc curring from July 1960 to June 1969, green shading represents the schedule occurring from July 1969 to June 1975, red shading represents the schedule occurring from July 1975 to April 1995 and purple shading represents the schedule occurring from May 1995 to 2007. 1951-1959 With the first narrative of 1951, the Refuge faced a bnormally low water levels with parts of the Refuge inaccessible. January to April of 1952 received normal rainfall, but again faced low water levels af ter March. In January, water
88 levels were 15.7 ft AMSL and in April water levels we re 15.0 ft AMSL. As with the previous months, May to August also received normal rai nfall yet saw lower water levels until late July. By August, water levels w ere roughly 15.64 ft AMSL In the central portion of the Refuge, however, water levels were significantly lower in the eastern and southern areas of the Refuge. One gauge near the Hillsboro Canal in the southern portion of the Refuge was read at 11.8 ft AMSL, this low reading was due in part to the low levels at which the canal was maintained. Drainage also had an effect on the Refug e by lowering the water table and ultimately changing the vegetation from ar eas that were ponds and marsh in the 1930s to sawgrass, willow and myrtle. This ch ange in cover had depreciated the wildlife habitat value; however with completion of the levees there was a hope to restore previous environments. Agai n, September to December also had normal rainfall. Water levels were h eld lower in order to facilitate the construction of the Hillsboro Canal and L -40 Canal. Compared to 1952, January to April of 1953 received higher rainfall and a better distribution as well as higher water levels. In terior marsh levels ranged from 15.0 to 15.5 ft AMSL and levels on the perimete r of the Refuge, Hillsboro Canal and L-40 ranged from 10.8 to 12.5 ft AMSL. Wh ile May to August also received heavy and well distributed rainfall, water l evels were lower than normal for these months. Interior marsh levels ranged from 10 .0 to 15.0 ft AMSL and reached a maximum of 16.5 ft above sea level in August as a result of rain and inflow from Lake Okeechobee. Water levels on the perim eter of the Refuge ranged from 8.4 to 13.2 ft AMSL. Abnormally high ra infall was recorded from
89 September to December and as a result water levels wi thin the interior holding pool were held 2.0 to 4.0 ft higher than in the bor row pit and the Hillsboro Canal. January to April of 1954 received normal rainfall, except April which received rainfall considerably above average. January h eld a high water level of 15.68 ft AMSL and April held the low of 14.75 ft AM SL .The Hillsboro Canal and L-40 borrow pit, which received some additional water through drainage from adjacent land, had water levels that fluctuated betwee n 11.0 and 13.0 ft AMSL. The drainage flowing into the L-40 borrow pit came p rimarily from the eastern portion of the Refuge and to a smaller degree from the interior holding pool. The only compensation for the loss of water in these two are as was the periods of heavy rain. Completion of the levee construction would allow higher water levels to be retained in these areas. May to August received abnormally high rainfall and had water levels higher in the borrow pit and pe rimeter than in the interior. By September to December, rainfall was back to the exp ected normal levels. Water levels in the interior were slightly higher thi s year than the same time period the previous year. A lack of rainfall from January to April of 1955 resul ted in lower interior water levels and the marsh nearly drying out. By the beginning of May, only 500 acres of the Refuge was holding water. By mid-May, howe ver, rains began to relieve the drought. However, by the September to D ecember time period rainfall was again far below normal and by December, most of th e Refuge was without surface water. It may also be of importance to note tha t at the time of writing, the
90 annual evaporation rate of surface water in the Ever glades was approximately 50 in. For the first four months of 1956, January to April, rainfall and water condition in South Florida set a 65 year low with onl y about 3.05 in of rain. As a result, water levels on the Refuge were very low. In addition, modification of the historical flow of water by means of drainage and div ersion reduced the once sponge-like terrain to that of a fibrous mass that ofte n times becomes consumed by fire. In addition, increased temperatures facilitate d increased transpiration and evaporation. The deficit in rainfall carried on thro ugh April to August and water levels remained low on the Refuge and causing parts of the Everglades to be completely devoid of surface water. With water level requirements for operation being between 14.0 and 17.0 ft AMSL, September to D ecember continued to see low rainfall and water levels far below those requir ed. With the loss of water came an invasion of plant growth that would anchor itse lf and bind to the surface matter. With the return of water, these plant masses wo uld rise to the surface and become permanent restrictive mats of growth covering what was once an open pool. The drought continued on the Refuge into the first period of 1957 with canal waters ranging from 10.5 to 11.0 ft AMSL with t he interior containing only shallow pools. Although heavy rains eventually came to provide some relief, the waters remained low on the Refuge. As more than norma l rainfall was received from May to August, the two year drought condition wa s finally broken and free water was in abundant supply. Despite this fact, canal wa ters did not exceed the
91 minimum proposed 14 ft stage. September through Decemb er saw similar issues as the previous period, average to good rainfall, but water levels falling below the minimum requirement for the majority of the period. Early September saw a high water level of 15.4 ft AMSL while late December recei ved a low of 11.20 ft AMSL. Water levels for the L-40 Canal represented a fair i ndication of water levels for the interior. In addition, the variance in elevation throughout the interior, ranging from 13.0 to 16.0 ft, can cause water coverage to vary from complete inundation of all areas to partial inundation of over one-fourt h of the areas. For approximately 30 days from January to April of 1 958, water requirements were met for the interior with an avera ge of 15.5 ft AMSL. However as water levels were brought to their desired level el sewhere in the Everglades, interior water levels on the Refuge declined to 12.4 ft AMSL. From September to December, water levels ranged from 12.5 to 13.66 ft A MSL in the L-40 Canal. By the May through August Period of 1959, large amounts of rainfall on the Refuge produced water levels higher in the interior than in the canals, of which the recommended water levels were rarely met. With a fear that higher water levels may be capable of producing wave action strong enough to damage the levees, the Central and South Florida Flood Control District (FCD) (In 1972, the FCD became the South Florida Water Management District (SF WMD)) was draining water out of the Refuge faster than inputs from raini ng and water flowing into the Refuge. This had been a reoccurring problem, over-dra ining in anticipation of rains that did not arrive. Through the remainder of the year canal waters
92 continued to not meet the recommended levels while the interior water were at least as high as recommended. Decadal Water Level Fluctuations. From 1954 to 1960 there was consistently a spike in interior marsh surface water level during the month of November as seen in Figure 6A Drought conditions that persisted on the Refuge from the end of 1955 to the early part of 19 57 can be seen with low mean surface water levels below 15.00 ft AMSL occurring in Ma rch, April, May, and June of 1956, spanning into parts of both the wet and dry seasons. Although the mean monthly surface water levels for site 1-8C dont have as consistent a pattern as that observed for 1-7, as shown in Figure 6B distinctive drops in canal water level is visible for the same months noted f or the interior. From February to August of 1956, canal water levels remaine d below 11.00 ft AMSL. FIGURE 6A: Mean Monthly Surface Water Level for Site 1-7 from May 1954 to April 1960.
93 FIGURE 6B: Mean Monthly Surface Water Level for Site 1-8C fro m May 1954 to April 1960. 1960-1969 The beginning of 1960 came with water levels similar to those of the previous year. As with years previous, the completion of another construction project, Levee 39, was anticipated in order to furthe r attempts to meet minimum water requirements. For the remainder of the year, w ater levels fluctuated above and below the required levels. By mid-year 1961, wat er levels were low, in certain areas revealing dead fish in canals. Although th e water level was at times 3 ft below schedule according to the Refuge calculations, USACOE formulas showed the water level being near the scheduled height By the end of the year, about 97% of the interior was devoid of surface water and throughout the Everglades water was only occurring in a few of the dee per sloughs. These conditions continued into 1962 with over 99% of the Re fuge being extremely dry. Despite this fact, water continued to be withdrawn from the Refuge at a rapid rate
94 to meet local agricultural needs. By May, rainfall be gan to bring up the water level to the scheduled 14.7 ft AMSL. At the beginning of the period, waters were 10.69 ft AMSL, or 3.74 ft below schedule. By August, w ater levels were above the scheduled levels and remained near schedule for th e remainder of the year. Although water was more plentiful at the beginning of 1963 compared to 1962, levels remained below their scheduled minimum. A s heavy rains in May increased the water level from 13.84 to 15.44 ft AMSL the USACOE increased discharges as this was above the scheduled requirement. As a result, the USACOE over discharged water and brought the water le vel down to 12.94 ft AMSL, resulting in fish stranding. From September to D ecember water levels averaged approximately 18 inches below the scheduled le vel. With Levee 39, or the Hillsboro Levee, having been completed in June of 1960, the Refuge saw for the first time water conditio ns near schedule for an entire year, 1964. In 1965, an extreme high of 17.6 0 and an extreme low of 10.98 ft AMSL were recorded. By June, water levels stabilized and remained at or above schedule for the remainder or the year. Fluctuat ions in water levels during 1966 were not as extreme as those observed in 1965. W ater levels ranged from 17.16 to 14.58 ft AMSL. With the water levels being low and generally below schedule for most of 1967, the Refuge saw water levels b etween 10.88 ft AMSL at the beginning of the year to 16.8 ft AMSL at the end of the year. For the majority of the year water levels were generally belo w schedule leading to a dry interior from May to June resulting in fish kills and a blue-green algae bloom.
95 With plentiful water throughout most of 1968, it w as speculated that there may have been as too much and that it may have been d rained wastefully into the ocean. Generally, summer drought conditions would o ccur on the Refuge from late April through June. These dry conditions woul d lead to a difficult situation for the Refuge as the northern half would become virtually dry during this time period. As was the case in 1968, waters on th e Refuge in 1969 were plentiful and possibly over abundant at times. As a re sult of this excess, the USACOE engineered an updated water schedule without co nsulting the Refuge staff. Decadal Water Level Fluctuations. With more than 90% of the Refuge interior being dry from the end of 1961 into early 1962, this could attribute to the lack of data reported from September 1961 through July 1962, which is visible in Figure 7A Low water levels are then observed beginning in May 1963. Similar trends for this dry period can also be seen in Figure 7B for site 1-8C. Spikes in interior marsh surface water level in November are al so observed again this year for 1960 and 1962.
96 FIGURE 7A: Mean Monthly Surface Water Level for Site 1-7 from M ay 1960 to April 1970. FIGURE 7B: Mean Monthly Surface Water Level for Site 1-8C fro m May 1960 to April 1970. 1970-1979 Issues with an high water levels carried on into the 1970s. Although this wasnt likely to be causing any damage, the FCD increase d drainage out of the
97 Refuge. As a result of over pumping, water levels fell below schedule and the Refuge entered a 50-day drought from early April to mid-May with water levels reaching 12.37 ft AMSL, or 2.63 feet below schedule. H eavy rains that relieved the drought brought on an increase in water level of 1.5 ft in one day, more than 2.0 ft in two days and 3.0 ft in less than a week. Thi s dramatic fluctuation is water level drastically affected wildlife populations within the Refuge. Although 1971 began with the worst drought on record the rains that came in mid-June helped to relieve the stress of the dro ught and by July and bring the water level back up to schedule. The narrativ e notes that dry is a relative term when concerning the Everglades. Although an area may be lacking visible water, water is retained in the peat inches be low the surface. While the drought was over before the end of 1971, 1972 saw wat er levels higher in the spring and summer months as a result of heavy rains and a decrease during the fall with a lack of rain coming from thunderstorms and h urricanes causing the water level to drop when it should have been rising Water levels in the snail kite management area were held between 15.68 and 16.19 f t AMSL in an attempt to recreate natural slough depths and maintain a suitable mix of submergent and emergent vegetation. A new water schedule was developed by the USACOE for the Refuge in 1974 and was adopted in 1975. The new schedule had a m inimum water level of 14.00 ft AMSL from May through July with a gradual i ncrease to a maximum of 17.00 ft by October. The 14.00 ft minimum was to be h eld for at least 30 days. The scheduled 17.00 ft above sea level was to be held u ntil January, when water
98 levels would gradually decrease to 14.00 ft through Ap ril. Although the minimum was changed from 15.00 ft AMSL to 14.00 ft AMSL, the main change in the new schedule was the use of several gauges as opposed to the c anal gauge to monitor and more realistically represent water levels o ver the entire interior pool. Water levels for 1975 reached a maximum of 16.67 ft A MSL and a minimum of 12.7 ft AMSL, both not reaching the maximum and exce eding the minimum levels. Two significant deviations from the water schedule occurr ed in 1976. The first occurred from May to June when water levels were h eld above the desired 14 ft minimum as a result of spring rains. The lowest wa ter level, 14.44 ft AMSL, was reached on May 1 and the peak water level, 16.64 f t AMSL, was reached in December. The second occurred in the fall when the sche dule fell short of meeting the maximum by nearly a foot. However, depa rtures such as these arent always detrimental. The departures during the spring and summer months were beneficial because apple snail populations were no t subjected to a complete dry-down and the spread of melaleuca was slowed by the higher water level. The lowest water level on the Refuge for 1977, 13.14 ft AMSL, was reached on May 3 and 4 in the canal and the peak water level, 17.20 ft AMSL, was reached in late December. When heavy rain in Septe mber brought the water levels to 17.08 ft AMSL, the USACOE opened control st ructures along the Hillsboro Canal in order to bring water levels back to schedule. Release of too much water however led to water levels being below sch edule for a short time
99 period. For 1978, the lowest water level on, 13.96 f t AMSL, was reached on May 2 in the canal and the peak water level, 17.02 ft AMS L, was reached on November 15 and 16, 1978. Waters in the L-40 canal reached the lowest levels reco rded in several years with a height of 12.2 ft AMSL in April. Heavy rains quickly relieved this deficit bringing the water level up to 15.1 ft AMSL. However, below normal rainfall from June to August brought interior water levels dow n to 14.97 ft AMSL in July. As rainfall came in September, water levels rose and re mained near schedule for the remainder of the year. Water levels in Compartme nt C were held between 16.0 and 16.4 ft AMSL in 1979. Decadal Water Level Fluctuations. Low water levels that were persistent throughout the dry season periods beginning in 1971 an d 1973 and into the beginning of the wet season can be seen in Figure 8A and Figure 8B for sites 17 and 1-8C respectively. Although a water schedule is set and some trends can be observed, such as the generally consistent mean surface water level occurring at site 1-7 for September, October and Novemb er, there is monthly variation from year to year.
100 FIGURE 8A: Mean Monthly Surface Water Level for Site 1-7 from May 1970 to April 1980. FIGURE 8B: Mean Monthly Surface Water Level for Site 1-8C fro m May 1970 to April 1980. 1980-1989 In 1980, L-40 canal levels reached a maximum of 16.92 ft AMSL in early January. Although water levels did decline, they remai ned higher than usual and
101 reached the lowest reading of 13.48 ft AMSL mid-June. Fluctuation in the canal waters brought levels below the minimum for about fi ve days and below the maximum in the later months of the year. Water level s in the interior remained between 15.16 and 16.75 ft AMSL throughout the year Although the water schedule now called for water to be held at 14.0 ft AM SL for 60-days during May and June, only nine days in those months met that requ irement in the previous five years. Instead, the alternate low of 15.0 ft AM SL was maintained in order to retain early rains on the Refuge. Low summer rains beginning in June of 1980 led to the driest conditions on the Refuge in at least 10 years and a 1 in 7 year drought for all of South Florida. The southern portion of the Refuge and the majority of the interior were dry from March to August. Water levels in the C-40 cana l were below schedule all year except for a brief time during hurricane seas on. On average, water levels were below schedule by about 1.4 ft. With the heavy r ains that began in March of 1982 came the end of the drought. As heavy rains contin ued throughout most of the summer, water levels rose almost 2.0 ft above norma l. As a result of the water shortage throughout South Florida, the alterna te of scheduled minimum of 15 ft AMSL was adopted for 1982 and water levels rema ined within one or two feet above schedule for the majority of the year. Throughout 1983, water levels remained close to schedul e. On May 25 a low of 12.2 ft AMSL was reached at the main canal gaug e, 1-8C, while interior gauge readings were producing much higher values. Howev er, the 15 ft alternate minimum was again adopted as a result of early spring ra ins. In order to meet the
102 scheduled 14 ft minimum, Refuge staff worked to modify the schedule. As a result, scheduled lows were moved from May through July to March through April in an attempt to precede early rains. The Refuge had proposed a revised schedule the previous year however, in 1984 the SFWMD decided to bring water levels to a low of 11.0 ft AMSL on May 20. By June 1 water levels had risen to nearly 16 f t AMSL, and by July water levels were back to schedule. During 1985 there were no major increases or decrease in water level and the water levels on the R efuge remained near schedule the duration of the year. The beginning of 1 985 was characterized by drought during the first half of the year due to lack of rainfall despite the fact that there were no major fluctuations in water level. Deviations from what was considered normal rainfall in 1986 led slight fluctuations above the water schedule in January. With the lack of rain in February, the water level was able to drop back down t o schedule where it remained near for the rest of the year. As with 1986 1987 saw the first few months in the beginning of the year switching between receiving above and below average rainfall. No serious affect was seen on th e water level as it remained near schedule for the majority of the year w ith higher water levels recorded for March. While 1988 began with water levels at or above schedul e, accept for the month of April when levels fell 1.2 ft below schedule for a short period, below average rainfall beginning in September caused wate r levels to drop. Throughout the month of September water levels dropped as a resu lt of the lack of rainfall
103 and were 0.9 ft below schedule by the end of the mont h. These abnormally dry conditions continued through the month of December. By the end of the year, waters were 3.62 ft below schedule, a level of 13.38 f t AMSL. As drought conditions continued into 1989, water levels remained b elow schedule on the Refuge. In the L-40 canal a record low water level of 9.7 ft AMSL was recorded in June. June, however, was also the wettest month that yea r and brought water levels in the interior of the Refuge up to 11.52 ft AMSL, 3.63 ft below schedule. By the end of the year water levels were up to 14.92 ft AMSL. Decadal Water Level Fluctuations. By looking at Figure 9A it is evident that on average, the 14.00 ft AMSL minimum that was to be maintained for 60days during May and June was not reached. The only tim e water levels came close to this minimum was during the drought that began at the end of 1988 and continued into 1989. Modifications to the water schedul e to use the alternate minimum water level of 15.00 ft AMSL and to change t he period when the minimum was to be held, from May and June to March an d April, still did not allow for minimum waters to be reached regularly. Exa mination of Figure 9B reveals that water levels in the canal approach and dro p significantly below the minimum at times during the dry season. In addition to the drought conditions that persisted at the end of 1988 into 1989, the drou ght mentioned in the Annual Narratives that occurred from June 1980 to March 1982 ar e much more apparent for site 1-8C than 1-7.
104 FIGURE 9A: Mean Monthly Surface Water Level for Site 1-7 from May 1980 to April 1990. FIGURE 9B: Mean Monthly Surface Water Level for Site 1-8C fro m May 1980 to April 1990. 1990-1999 The three year drought that began in 1988 was broken by mid-January of 1991 with heavy rains. These rains brought the water l evel from 14.48 ft AMSL at
105 the beginning of January to 16.80 ft AMSL by the en d of the month. In order to attempt to stimulate productive healthy marsh condition s in the northern quarter of the Refuge, a new water schedule was proposed by th e Refuge, SFWMD and the USACOE. Changes to the water schedule include changi ng the minimum from 14.0 ft AMSL to 15.75 ft AMSL, allowing water to rise as high as 17.5 ft AMSL during some months of the year and finally not a llowing water to be withdrawn from the Refuge if levels are below 14.0 f t AMSL, the pervious minimum had been 11.0 ft AMSL. Specific anticipated be nefits included; larger populations of aquatic organisms, increased protection aga inst drought, prevention of brush invasion into sawgrass, prevention o f sawgrass invasion into wet prairies, and an extended hydroperiod on the Ref uge (Brandt, 2006). The water schedule had been proposed the previous year was not yet in place in until 1992. Before officially implementing t he schedule, the USACOE wanted to first model the changes proposed in the new schedule. Water levels in the L-40 canal fluctuated throughout the majority of the year. Levels in the canal began at 16.24 ft AMSL and declined throughout the b eginning of the year to reach a low of 13.12 ft AMSL on June 2. Early June rai ns that brought up the water level required the gates to be open and water to be released, bringing the water level in the canal back down. Levels continued to fluctuate throughout the latter half of the year and remained between 16.5 a nd 17.22 ft AMSL in November and December. By 1993, the new water schedule had still not been implemented. Less fluctuation occurred in the canals this y ear. Water levels in the canal began the year at 16.86 ft AMSL, just sligh tly under the maximum. The
106 low for the year of 13.06 ft AMSL was reached on May 28 after a gradual declining from the beginning of the year. Until the end of August, water levels remained fairly level, and stayed near the maximum l evel until the end of the year. At the beginning of 1994, the new water schedule had still not been implemented. With high water levels at the beginning of the year, biologists asked that the control gates at the south end of the Ref uge be closed in order to maintain conditions in support of snail kites exhibitin g nesting behavior on the Refuge. By November, water levels in the interior ha d reached record levels with a reading of 18.34 ft AMSL. Although water levels de clined slightly throughout the end of the year, record levels continued to be reco rded. By May 11 of 1995 the new water schedule was approved and implemented. Water levels for the interior were gene rally at or above schedule for January through April and June through August, as w ere at or below schedule May and September through December. The high for the year, 17.95 ft AMSL, was recorded on October 20 and the low, 15.31 ft AMSL, was recorded on June 2. Except for the tree islands, the marsh interior did not dry out and remained wet the entire year. Water levels remained above schedule from January through April of 1996 until the flood gates at the south end of the Refuge were opened, for an unknown reason, and water levels dropped. Wate r levels dropped in such a short amount of time one foot in six days, that a s the marsh turned to muck, snail kites left the Refuge in search of more suitable ha bitat. Although water levels stabilized and were above schedule during May an d June, drought-like
107 conditions caused water levels to drop gradually through the rest of the year with a noticeable drop after November. During 1997, water levels were at or near schedule fo r the duration of the year. A high water level of 18.11 ft AMSL was recorded in December and a low of 15.42 ft AMSL was recorded in May, both at the 1-8 C canal gauge. On average, interior water levels were 0.25 ft higher t han the water levels observed in 1996. Hunting snail kites again left the Refuge whe n it turned to muck as a result of flood gates opening and water levels droppin g. Water levels rapidly recovered. The usually dry tree islands became inundate d as interior water levels neared record highs With the El Nio Southern Oscillation (ENSO) event th at occurred in 1998, dry season rains helped lead to water levels that were near or above the scheduled maximum for the majority of the y ear. Two major declines occurred, one at the end of March-early April and the second in June. These declines were the result of a lack of rain and the openi ng of the flood gates. Water levels in 1999 started off on schedule and began to decline in early spring. No rainfall and water released caused water to continue to decline until heavy rains fell in late June and relieved drought con ditions. Water levels fluctuated for the remained of the year with the he avy rainfall brought on by two hurricanes. At the 1-8C canal gauge, a high of 18.22 f t AMSL and a low of 14.12 ft AMSL were recorded for 1999. Decadal Water Level Fluctuations. For Figure 10A and Figure 10B two deviations are apparent that correlate to seasonal fluct uations mentioned in the Annual Narratives. The first is the continuation of dr ought conditions from 1989
108 apparent throughout 1990 with a sharp drop in mean water level in December for both site gauges. The second event noticeable is the abov e average mean water level occurring during the dry season at the beginning of 1998, roughly from January to March. These dry season rains can be attributed to the ENSO event that occurred that year. Another trend that can be se en in Figure 10A as in all figures for site 1-7 throughout the decades, is the me an surface water remaining the same for September, October and November, while v arying from year to year. FIGURE 10A: Mean Monthly Surface Water Level for Site 1-7 from May 1990 to April 2000.
109 FIGURE 10B: Mean Monthly Surface Water Level for Site 1-8C fro m May 1990 to April 2000. 2000-2007 The year 2000 began with water levels near schedule. W ater levels began fluctuating in February with the decrease and increase of rainfall until June. By the last week of June, the upper two-thirds of the Refu ge were completely dry except for the refugia provided by alligator holes. W ith hurricane season came rains and elevated water levels prompting the release of water. November and December brought little rain prompting concerns of a dr ought and low water reserves. Low water levels continued into January 2001 w ith water levels in the L-40 canal averaging 0.91 ft AMSL below the scheduled levels. Water levels continued to drop 2.5 ft throughout February as littl e rainfall was compounded with water releases. Water releases continued and caused water levels to drop to 13.36 fl AMSL by March. Rain brought water levels back up to 15.2 ft AMSL according to a reading in the L-40 canal. By the end of May, however, water
110 levels were back down to 12.06 ft AMSL in the canal. M ay rains brought water levels back up and inundated dry areas. A reading of 16 .88 ft AMSL at 1-8C ended the year, as did water levels remaining near sche dule from August on. While the water levels of 2002 began near schedule, r ain in February brought water levels to rise above schedule where they fluctuated, but remained high for the rest of the month. Water levels began to drop in April, 15.97 to 14.88 ft AMSL, and continued to drop through mid-June where some regions of the Refuge were showing water readings that represented dr ought conditions. These conditions did not last long and water levels reached a high of 16.8 ft AMSL in July. Water levels remained relatively stable until mo derate ENSO activity brought rain in the normally drier season. Stable wat er levels between 16.8 and 17.05 ft AMSL remained throughout the end of the ye ar. Stable water levels just under the scheduled maximum continued into 2003 until unusually high rainfall brought water levels up in March. Interior water lev els began to fall throughout May and the lowest level of the year, 15.23 ft AMSL, was recorded on June 3 from gauge 1-8C. A combination of rainfall and the opening of the flood gates caused water levels to fluctuate throughout June and lar gely throughout the remainder of the year. Beginning in January 2004, water levels decreased due to low rainfall and began rapidly decreasing in mid-April until drought conditions were reached in May. Although two hurricanes were able to temporarily spike water levels back up in September, levels began to decrease from mid-Octo ber through the end of the year. The low water levels of 2004 carried into 2005 and started off the year
111 with water levels 1.0 to 1.25 ft below water levels o f the past years. Water levels increased in March, but then again dropped off in Apri l. The lowest water level, 15.19 ft AMSL, on the Refuge occurred on July 22. Flu ctuations continued with the onset of the wet season as water levels increased. T he highest water level, 16.9 ft AMSL, occurred on November 2 and from this po int, levels remained near 16.75 ft AMSL throughout the end of the year. As seen in the previous year, 2006 saw water levels de crease through the end of February as a result of dry season low rainfall. Seasonal rainfall in July caused water levels to begin to rise. Average water lev els on the Refuge ranged from 17.09 to 15.33 ft AMSL. For the first time since 2004, water levels in May 2007 dropped into the drought condition zone as a resu lt of dry season decrease. For the first time since 2003, water levels were held a bove 17.00 ft AMSL throughout September and November. Average water le vels on the Refuge ranged from 17.45 to 14.17 ft AMSL. Decadal Water Level Fluctuations. As is consistent with the Annual Narratives, declines in water levels from 2000 to 2001, 2004 and 2007 can be seen in Figure 11A and Figure 11B
112 FIGURE 11A: Mean Monthly Surface Water Level for Site 1-7 from May 2000 to April 2007. FIGURE 11B: Mean Monthly Surface Water Level for Site 1-8C from May 2000 to April 2007. Water Regulation Schedule Figures 12A through 15B depict surface water fluctuations for sites 1-7 and 1-8C grouped according to the four water regulat ion schedules that persisted
113 on the Refuge throughout the study period. For each site, the mean surface water level was calculated each month from data acquir ed from SFWMDs DBHYDRO. The four water regulation schedules dated fro m July 1960 to June 1969, July 1969 to June 1975, July 1975 to April 1995 and May 1995 through 2007. From these figures it is evident that although wa ter regulation schedules were set in place, deviation from the scheduled water levels occurred. FIGURE 12A: Mean Monthly Surface Water Level for Site 1-7 Durin g the Water Regulation Schedule from July 1960 to June 1969.
114 FIGURE 12B: Mean Monthly Surface Water Level for Site 1-8C Duri ng the Water Regulation Schedule from July 1960 to June 1969. FIGURE 13A: Mean Monthly Surface Water Level for Site 1-7 Durin g the Water Regulation Schedule from July 1969 to June 1975.
115 FIGURE 13B: Mean Monthly Surface Water Level for Site 1-8C Duri ng the Water Regulation Schedule from July 1969 to June 1975.
116 FIGURE 14A: Mean Monthly Surface Water Level for Site 1-7 Durin g the Water Regulation Schedule from July 1975 to April 1995.
117 FIGURE 14B: Mean Monthly Surface Water Level for Site 1-8C Duri ng the Water Regulation Schedule from July 1975 to April 1995.
118 FIGURE 15A: Mean Monthly Surface Water Level for Site 1-7 Durin g the Water Regulation Schedule from May 1995 through 2007.
119 FIGURE 15B: Mean Monthly Surface Water Level for Site 1-8C Duri ng the Water Regulation Schedule from May 1995 through 2007. Other Herbicides Overview Prior to the observation that apple snail numbers had significantly declined in the canals, which occurred in 1975, 2,4-D and diquat were the two most heavily utilized herbicides on the Refuge for the tre atment of aquatic invasives. Even before the realization that apple snail numbers had decreased in the canals, an area that had received heavy use, testing beg an on the affects of 2,4D on apple snails in 1965. After the realization that numbers had decreased between the mid-1960s and 1975, testing began on the affects of diquat on apple snails. In order to track the use of these herbicides and others that could
120 potentially become a threat to apple snail health, t he use and studies conducted related to the use of herbicides on the Refuge has bee n examined in cohorts broken down by decade. 1951-1959 In the early 1950s spraying began in the Southern po rtion of the Refuge for the treatment of water hyacinth and alligatorwee d. Initially, 2,4-D was used to treat scattered areas. Cultivated crops in Compartment C and sometimes B were also treated with 2,4-D, VL-600, and Dalapon. In add ition to efforts made by the Refuge to control nuisance vegetation, the FCD, USACO E and FWS contributed to control efforts. Other areas treated included the le vee, dikes and canals. Experimental testing began for the control of undesir able vegetation. Experimental herbicides included 2,4-D, VL-600, 5TC06, Chloro 16, X T B, and C U M. Each herbicide was applied to various types of veg etation, at various rates per acre, and with either water or fuel oil as the di latants. Throughout the 1950s similar experimental testing was conducted with other he rbicides including Tween-20, Spreader-Sticker, Dalapon, aminotriazole, C hlorax-40, HC1281, Silvex, Kuron, Aquaherb, and Baron. However, these e xperimental tests were concerned with effective control of nuisance vegetation a s opposed to the effects of these herbicides on non-target organisms. An import ant finding was that a lesser kill was obtained in areas where there was water movement, a natural expectation as the materials drifted.
121 In 1957 2,4-D spraying began in the interior of the Refuge as water hyacinth moved into the area and the L-40 Canal becam e overrun. Although spraying may have served as a temporary fix, pumping w ould reintroduce water hyacinth into areas previously sprayed as water finds its w ay into the interior as levels increased. 1960-1969 In addition to the reintroduction of nuisance vegeta tion from pumping, water level rise from hurricanes also led to a resurgence In order to prevent areas from being entirely overtaken, aerial applicatio n of 2,4-D, combined frequently with diesel fuel, began. Although 2,4-D h ad good kill rates for water hyacinth, there was no permanent damage to alligator weed and water lettuce. In areas such as the Hillsboro Canal, water lettuce began m oving into areas where water hyacinth had been at faster than anticipated rat es. Two forms of 2,4-D were being used on the Refuge, amine solution and isoo ctyl ester. In addition to the regular use of 2,4-D for nonexperimental control of water hyacinth, Amitrole-T and Dalapon were also used. While frequent areas of v egetation control include the periphery canals and contiguous fringe areas, herbi cide treatment was also taking place in Compartment C. There was a ban on the aerial application of herbicid es in 1964, however spraying continued by boat. After the settlement of a lawsuit against the Refuge by crop farmers, a damage suit regarding agricultural cr ops as a result of aerial spraying, a strict program for aerial and ground contro l was approved to reduce
122 the effect of herbicide drift to agricultural lands. A s the use of 2,4-D for hyacinth control continues, the Refuge acknowledges that hyacinth will most likely never be eliminated entirely from the Refuge, but that an nual treatments were necessary to protect wildlife habitat and wildlife speci es themselves. In 1965 the Refuge conducted its first test on the app le snail regarding residues of 2,4-D. The objectives of the testing were t o determine if under Refuge conditions whether detectable 2,4-D residues occurr ed in snails prior to and after application of 2,4-D at a 4lbs per acre rate and if 2,4-D residues were determined in measurable amounts, attempt to measure t he residue level before and after application. In addition to sampling the ap ple snails; water, vegetation and soils were also sampled. At first, snails rapidly fed on the lettuce provided, but then decreased their feeding after the first four days. By the end of the week there was no apparent feeding and the mortality rate was rapidly increasing. Two batches of eggs were laid and hatched, those snails that did hatch died within a few hours. No 2,4-D residues were detected however. For the first time, diquat was used for the treatmen t of hyacinth on the Refuge in 1968. As the decade ends, 2,4-D and diquat are being used to treat all the canals around the perimeter, around 55 miles, wh ile the FCD continued with a spraying program of their own, in addition to the Refuges efforts, focusing on the Hillsboro Canal and water lettuce control.
123 1970-1979 As the early 1970s see continued use of 2,4-D and diq uat for water hyacinth control on the perimeter canals and L-40, conce rns over the safety of 2,4-D in the control of aquatic vegetation on public l ands arose. As a result of these concerns, extensive sampling was conducted during ae rial spraying operations. Mud, water and fish samples were collected an d observations were made on the grackle colonies that were in all stages of b reeding, in the sprayed areas. Preliminary results on both the samples analyzed and observations, in which no adverse results were noted, allowed for and e xtension of the aerial spraying program. While under ideal conditions, water hyacinth would be capable of doubling in numbers every two weeks. On the Refuge, these condit ions exist nearly year round. In addition, since seeds are dried and then inun dated, germination and reinfestation of formerly clear areas is encouraged. In addition, pumping by the FCD contained run-off from neighboring pasturelands a nd sugarcane field. The water being pumped in was laden with nutrients and of poor quality, which provides an ideal medium for growth of hyacinth. Some control was needed along the interior dikes of th e kite management area to reduce water hyacinth and water lettuce in th e feeder canals and to reduce woody vegetation on internal dikes. In additio n to the infestation of aquatic vegetation, Melaleuca starts to become an issue o n the Refuge in which 2,4-D is used.
124 As a result of the extreme difference observed in app le snail population in or near the canals between 1965 and 1976, the effects o f diquat on the apple snail was examined. Longevity, egg production and egg hatching rates were observed. No statistical differences in any factors between individual tanks or between treatment groups of tanks were observed. Of th e three snails that died during the study, only one was in a diquat treatment tank. From this, it was concluded that there was not direct short-term molluscacid al effects produced by diquat. At times, it was necessary to use herbicides to control veg etation that was much too dense for kite feeding in areas such as Compartm ent B and C. In the continued effort to control water hyacinth and water lettuce, diquat began to receive increased use over 2,4-D as it was less volatile. C are was also taken to eliminate drifting problems. In addition to spray efforts made by the Refuge, th e South Florida Water Management District (SFWMD), formerly the FCD, continu ed to assist in the spraying effort. However, the future of SFWMD vegetat ion control was in doubt because of their use of diesel fuel in their spray form ulation and would depend on the outcome of a diesel fuel toxicity study. Spawning from concerns about diesel fuel toxicity, the Refuge conducted a study on the effe cts of diesel, 2,4-D and diquat on snail hatch rate and juvenile mortality. I n past studies, petroleum products had been found to be extremely toxic to vario us animal life forms. Although immediately after application of diquat, no rmally opaque eggs became
125 translucent and did not hatch, no definitive conclusions h ad been drawn yet by then end of the decade. 1980-1989 At the beginning of the 1980s approximately 30% of canal waters were covered with water hyacinth. The SFWMD was primarily re sponsible for vegetation control in the canals. Extensive spraying led to large amounts of dead vegetation which may have depleted the water of oxyg en and led to fish kills. The Refuge continued to treat the impoundments in Compart ment C, the managed snail kite habitat, with herbicides including diquat to treat invasive aquatic plants such as cattail, water hyacinth and water lettuce as th ey moved into the managed compartments. In addition some of the already mentioned vegetatio n being treated on the Refuge, melaleuca, Melaleuca quinquenervia Brazillian Pepper, Schinus terebinthifolius and Australian Pine, Casuarinas pp became prevalent pests treated with familiar herbicides, such as 2,4-D, as well as newer herbicides, such as Round-Up. With non-native plants rapidly invading areas of the Refuge, scientists and managers tried a wide array of herbicides t o control the plants, not all of these being approved for use in water, like Arse nal and Velpar-L for example. Rodeo was used for the first time in 1983 by the SFWMD to treat para grass, Urochloa mutica in L-40 and willow in Compartment A. Rodeo, also u sed to treat melaleuca, was approved for use over water.
126 As melaleuca began to invade the interior, increasing ly being found in native sawgrass communities and tree islands, control studie s were conducted to determine the most effective herbicides and application methods for control. Application of some herbicides, like Garlon 3A and Arsen al, led to the concern of the effects of leaching chemicals on surrounding vegetat ion. It was assumed that the herbicides chemicals leached through the soil, killing plants whose root system came into contact with the contaminated soil. In ad dition to concerns about leaching, non-target damage was encountered wi th Velpar-L. Non-target damage may have been a result of applying the herbi cides at rates higher than those recommended on the label. Herbicide falling on t he vegetation surrounding the target plant could accidentally be dosed with the h erbicide. In addition to receiving assistance from the SFWMD, the University of Floridas Institute of Food and Agricultural Sciences (IF AS) also became involved with the Refuge. IFAS contributed to the melaleuca con trol studies by testing Spike, Escort glyphosphate, and Garlon 4 in the inte rior of the Refuge. IFAS also used DMA 4 on willow and Brazillian pepper arou nd Compartments A and B. With water hyacinth and water lettuce choking certain w aterways, the SFWMD continues to aid in vegetation control with diqu at and the mechanical control of floating islands of vegetation. The 1984 an nual narrative noted that the SFWMD was observed spraying a chemical not approved for use by the Refuge, however the name of that chemical was not revealed. In the same year, fifteen herbicides that contained ethylene dibromide (EDB), in cluding diquat, were banned in South Florida. After it was determined tha t the EDB levels did not pose
127 a significant environmental danger or hazard to human health when used in nonpotable waters, the herbicides were removed. In conjunction with copper toxicity studies, copper-diqua t was applied to apple snails in closed tanks and in field settings. Althoug h extremely high concentrations applied within closed tanks led to apple sna il mortality, death did not result from typical field applications. The results o f this study were reported by Winger et al. (1984). In 1988, a lawsuit was filed against the SFWMD and th e Florida Department of Environmental Regulations (DER). The l awsuit alleged that the two agencies violated state law in failing to regulate water pollution entering the Refuge. Poor water quality in the canals surrounding the Refuge led to vegetation changes most noticeable along the edge of th e Refuge. While the interior of the Refuge remained in fairly pristine co ndition, it was a mix of an oligotrophic ecosystem at the center of the Refuge with a gradual change to a eutrophic ecosystem along the outer edges. 1990-1999 As the melaleuca control studies continued with the ap plication of various herbicides at varying rates, Rodeo, which had already b een approved for use over the water, was used in the management compartment s to treat cattail, water lettuce, primrose willow, willow, carious grass species, an d pennywort. Diquat was use in the compartments as well, especially in Compart ment C to treat
128 melaleuca. Diesel fuel was also used in conjunction with herbicides for melaleuca treatment. 2000-2007 With Old World climbing fern, Lygodium microphyllum impacting roughly 84% of the interior, or 120,000 acres, of the Refuge, Rodeo, Escort, Round-Up, Garlon 3A, Reward, Arsenal, and 2,4-D were used for v egetation control at label rates with other rates being tested in experimental co ntrol studies. In 2002, samples were collected from the Strazulla Ma rsh and cypress swamp and analyzed for various contaminants and nutrien ts. In six of the seven marsh sites sampled from within the Strazulla Marsh, diq uat and/or paraquat herbicides were found. These herbicides likely drifted t o this area from the nearby canals being treated for vegetation control. D iquat levels ranged from 1.7 to 14 ppb and paraquat ranged from 1.3 to 7 ppb. It was determined that these levels would not have adverse affects on the ecosystem. Di quat and/or paraquat were also found in six of the seven sites sampled from wi thin the cypress swamp. As with the Strazulla Marsh, these levels of contaminati on were attributed to drift. Diquat levels ranged from 5 to 10 ppb and paraquat r anged from 1 to 7 ppb, also levels too low to be expected to adversely affect the e cosystem. Diquat was monitored to track the potential movement into the marsh interior and into preferred habitats used by threate ned and endangered species, such as the snail kite. A total of four sites were sampled, two in L-40 and two in the interior. For all four of the samples, no diquat was detected. Had diquat been
129 present, the water samples collected should have been abl e to track any movement of diquat from the canal into the marsh int erior. It was determined that the level of spray applied in the canal was adequate enough to kill the target vegetation, and that with proper application, residua l diquat levels would not be high enough to be of concern to wildlife in adjacent areas. Fire Overview Natural fires of varying sizes occurred on the Refuge a nd at times led to damaging peat burns. The mention of prescribed fires f irst appeared in the Annual Narratives in 1957 and was first put into use in the early 1960s as a means to control invasive vegetation in addition to t he use of herbicides. In addition to the vegetation control benefits of fire, fire has the potential to be more beneficial than detrimental to wildlife. In 1962 a l arge fire burned for weeks along L-7 affecting roughly 11,000 acres and fire was often asso ciated with dry conditions, such as in 1981 when five wildfires affected a total of 6,640 acres. With the completion of the comprehensive fire manageme nt plan in 1984, the interior of the Refuge had yet to be treated with a prescribed burn. The use of prescribed burns increased throughout the 1980s, but was o ften limited by conditions that were too wet or too dry. In 1989, a p rescribed fire along L-7 burned out of control and affected roughly 40,000 acre s from May 11 to July 27. The occurrence of natural and prescribed fires on the Ref uge are examined in corhots by decade.
130 1951-1959 During construction of one of the levees, two fires bur ned during 1952. While the natural sawgrass area was affected, the fire d id not burn deep enough to damage the peat. The loose organic peat within th e Refuge made it nearly impossible to walk upon or use ordinary equipment for f ire control. Because of this, there was no way to control fire on the Refuge w ithout investing a lot of money. It was predicted that once the levees were comple ted and water level rose, there would no longer be a fire hazard on the Refuge. In 1953 two fires burned on the Refuge, one approxi mately 2560 acres and the other 1920 acres. Saw grass, willow and myrtle were burned as a result of these fires. Although the fires may have caused some damage to the top layer of peat, they were considered beneficial for preservati on of waterfowl habitat. Lack of rainfall in 1955 created quite the fire hazar d on the Refuge. With frequent fires on and near the Refuge, some burning for days, about 1000 acres of Refuge land east of the levee and about 10,000 acr es within the impoundment area were burnt from fire. However, most fires occurred before the peat was dry enough to burn, resulting in less than 300 acres of pe at burns. Although undesirable species have the potential to quickly invad e a burn area that is followed by a dry period, peat fires followed by floo ding could result in improved waterfowl habitat. During the dry season and periods o f drought, an extremely low water table permits for a deeper burn. No planned burns occurred, however a new section appea red in the 1957 annual narrative for planned burns, which would describ e the general condition
131 prior to burning and the general conditions following burning. The edges of the Refuge, covered mainly with maidencane, sawgrass and wil low, nearest the levees are usually in a dangerous condition for burnin g. Under normal conditions, fires are unable to spread far into the interior beca use of the presence of water in the marsh area. The main concern regarding fire was wit h retaining the present aquatic habitat and the potential changes that could occu r to the land as a result of fire. 1960-1969 The early 1960s saw some fire, but in general water l evels were high enough that no serious peat burn occurred. Two wildfire s burned about 5800 acres in 1960 and contributed to the control of water hyacinth. As a result of the benefit of naturally occurring fires on the control of the Refuges invasive vegetation, experimental burnings were conducted on wa ter hyacinth that had been chemically treated between Canal 7 and Levee 7. In addition to the vegetation control benefits of fire, fire has the pote ntial to be more beneficial than detrimental to wildlife. One fire that burned for w eeks along L-7 in 1962 and burned roughly 11,000 acres was one such fire that could be more beneficial to wildlife. Exceedingly heavy willow growth was cleared and although a peat fire occurred in restricted areas, soil loss was minimal. In 1967 the management area was burned as the result of a prescribed burn and a lightning fire burned a couple of acres i n the interior during the drought. There was mention of one fire in 1969 that occurred in a dense
132 sawgrass stand and burned roughly 12 acres, however there was no peat damage. 1970-1979 During the drought of 1971 only one small fire was mentioned. It was noted that small fires of this nature often do more go od than harm in that they are essential for maintaining the Everglades in their natu ral state. Wildfires become threatening when they develop into wide, fast-moving fronts. Along the same line, a fire burned approximately 90 acres of Compartment A in 1975. Several wildfires in the interior in 1979, thought to have o ccurred due to the extremely dry summer, were also allowed to go uncontrolled. In addit ion to using herbicides to treat undesirable vegetation, burning is at times used in addition to get a more desirable result and to help clear vegetation that is m uch too dense for kite feeding. For example, in 1978 vegetation control in Compartment B and impoundments C-2 and C-3 was carried out to encourage a pple snail production and snail kite use. 1980-1989 The drought conditions that struck South Florida in 19 81 provided good conditions for wildfire and produced five fires that af fected a total of 6,640 acres. Although a heavy fuel buildup was allowed to occur due to a lack of large fires in the previous years, moisture held by the peat prevente d an extensive spread of the fire. The annual narratives note that the Refuge plans to develop and
133 implements a prescribed burn management program to red uce fuel buildup before wildfires can do serious damage to Refuge habita ts. The following year, impoundments C-1, C-6, C-7 and C-8 were burned in an effort to manage the snail kite habitat and contr ol cattails. In addition to the prescribed burns of the impoundments, one wildfire occurr ed in the interior. In 1983 Compartments A and B and impoundments C-3, and C -7 were prescribed fire treatment and a single one acre fire occurred else where on the Refuge. The Refuges comprehensive fire management plan was complete d in 1984. Although fire had been previously used to manage the impoundme nts, the 143,000 acre interior had never been treated with prescribed fire There were two major objectives of the fire management plan: to perpetuate northern Everglades habitat by interrupting plant succession and maintain co mmunity characteristics; and to reduce fuel loads to decrease the potential for disastrous fires. Although prescribed fires could lessen the impact of fire of tree i slands, soils and wildlife, the use of fire could accelerate the spread of melaleu ca, which is fire resistant. One major constraint on the Refuge to the use of prescr ibed fires was water availability, an excess of water could make ignition possi ble while a lack of water could restrict accessing the interior. The only area prescri bed a burn in 1984 was impoundment C-9. The first year following implementation of the new fire management plan saw only a total of one acre burned by prescribed fire a nd roughly five acres burned by wildfire. In some instances, rain extinguishe d the fires and in others natural barriers prevented the spread of fire. In 198 6, four known wildfires burned
134 roughly 0.55 acres and one prescribed burn in impoundme nt C-8 burned 32 acres in an attempt control cattails. In addition to fou r wildfires that were recorded that year, many other small fires occur on areas of the Refuge that were not easily accessible and went unreported. Of the three burns prescribed for 1987, two were desig ned to control cattail and willow growth in conjunction with other ma nagement techniques and the third was designed to kill melaleuca seedlings. Two of the three fires were described, one being a 21 acre burn in impoundment C-5 and the other being a 3600 acre burn of melaleuca trees. In 1998, parts of im poundments C-2, C-4, C6, C-7 were burned as were the lower part of Compart ment A and impoundments B-2 and B-3. Low water levels in the month of May al lowed for 257 acres to be burned. In 1989, a prescribed burn along L-7 that beg an on May 11 th spotted across a canal and turned into a major fire. The fire was declared controlled by May 15 th but not declared out until July 27 th sure to acres of muck fires in areas where the peat had been disturbed as a result of constru ction. The total acreage affected was roughly 40,000. Although out of control a t times, the Refuge saw good use after the fire and diversity was much more ap parent. Three lightning fires were also recorded in 1987, each totaling 100 acre s or less. 1990-1999 With no documented fires at the beginning of the 199 0s, 1992 saw several of its prescribed burns carried out. Impoundments C-3, C -4, C-8, and C-9 were successfully treated while wet weather prevented treatm ent of Compartment A
135 and led to a partially successful burn of Compartment D Although 11 burns were prescribed for 1993 totaling roughly 6,607 acres, none of the burns were conducted. With the burn season running from December 1 st through the end of August, adverse weather conditions prevented the burns. At the beginning of the 1993 burn season, conditions were too wet and water le vels were too high. Towards the latter part of the burn season, starting r oughly in May, the KeetchByram Drought Index, on a scale of 0 to 800, climbed a bove 500, too high to safely execute any prescribed burns. The following year saw the same issues; the 11 burns were once again unable to be conducted. T here were five wildfires in 1994 burning roughly 1530 acres. One wildfire burn ed approximately 500 acres in 1995. No prescribed fires were conducted in 1996, 1997, 1998, or 1999. Revisions to the old fire management plan began in 19 99 in an effort to being active prescribed burns in 2000. 2000-2007 After eight years of no prescribed fires, the first p rescribed burn was carried out in 2000 when eight acres were burned in im poundment C-6. In addition to the one prescribed burn there were also tw o wildfires that burned 1,200 and 1,800 acres in the northeaster and northweste r Refuge interior respectively. In 2001, no prescribed burns were conducted yet 11 wildfires were ignited by lightning and burned a total of 10,454 a cres in various acres across the Refuge in the month of June.
136 For the first time in 10 years, prescribed fires were car ried out in the 32 acre C-8 impoundments and the two acre C-6 impoundment to clear undesirable vegetation in 2002. In addition to the prescribed bur ns, two lightning fires were recorded north of the Hillsboro boat ramp. Then in 20 03, the Refuge conducted its first prescribed burn for the interior in nearly 18 years. The burn was conducted in the northeastern portion of the interior and covered approximately 2,400 acres. In 2004, three wildfires ignited by light ning affected the Refuge. The three fires included one fire that burned 441 acres in the northwest interior and two other fires that burned 50 and 150 acres. Prescribe d burns were also carried out over approximately 6,553 acres in the southwest co rner and impoundments B-2 and C-10. In 2005 there were 75 acres of wildfire s and burns were prescribed for impoundments C-1, C-7 and C-10. In 2006 two natural wildfires burnt approximately 20 acres, an unsuccessful prescribed b urn affected only 50 acres in the southeast corner of the interior, a 488 acre prescribed fire burned along the L-7 canal, and the LILA impoundments were b urned after being pretreated with an herbicide. In 2007, over 20,000 acres were burned by wildfire an d prescribed burns. The wildfires that burned the Refuge in 2007 that we re ignited by lightning included a 5,308 acre burn on the western side of the R efuge, a 1,250 burn in the northern portion of the Refuge, a 900 acre burn in Compartment D, and a fourth fire on the eastern side of the Refuge. The remaining 13,000 acres of burned land were a result of prescribed fires, the most that ha d ever been prescribed and carried out in a single year on the Refuge. The t hree prescribed burns were
137 located near the Lee Road boat ramp and the south cent ral portion of the Refuge. Non-Avian Predation Overview The Annual Narratives began with generally low obser vations of alligators and high incidences of poaching. Throughout the perio d examined, a trend was noticed that with decreased waters in the interior duri ng times of drought alligators would move into the canals to seek refuge i n the deeper waters. The health and movement trends of alligators were monitor ed through the Cooperative Alligator Survey, which began in 1971, a nd the nighttime surveys which began in 1998. The threat of heavy aquatic pred ation threatening apple snails began in 1972 and led to the eventual use of no n-avian predator exclusions fences in some of the managed impoundments. The health of alligators and attempts to decrease aquatic predation o f apple snails is examined in cohorts by decade. 1951-1959 Within the Refuge, observations of alligators were ge nerally low as in the past the area had been heavily hunted for alligators. Alligators larger than six feet were rarely observed. Although Refuge scientists initial ly predicted that it would take between two and three years for alligator popula tions to return to normal, numbers remained abnormally low throughout the major ity of the 1950s. In
138 addition to the high price of alligator hides and the unrestricted night use of the Refuge, as the Everglades became increasingly dry in th e mid-1950s it was assumed that alligators were moving to the canal proper The importance of the presence of alligators on the Refuge did not go overl ooked. Not only are alligators responsible for keeping channels free of vege tation and controlling rough fish such as the garfish, the deep caves or holes t hat they keep provide refugia during dry periods. Towards the end of the 19 50s alligators began returning to the Refuge interior with good success. 1960-1969 As the early 1960s began to see an increase in the pri ce of alligator hides and illegal poaching activity continued, a sudden seriou s drop in population occurred. Refuge scientists were also hopeful that more al ligators would continue to return to the Refuge as water conditions improved. Yet drought conditions continued at times and reports continued to convey that t hey were largely confined to canals. In 1963 alligators were seen in canal s and ditches in Compartment B and C, areas which would later be manag ed for apple snails. Although poaching remained a problem, alligator num bers appeared to hold their own throughout the 1960s as they were frequently seen sunning in the canals. Before the beginning of the Cooperative Alligator S urvey in 1971, most of the surveying took place during the wet season and densities w ere variable from year to year. Table 13 reports the data collected from the Cooperative Allig ator Surveys and some additional data in 1969 and 1970.
139 TABLE 13: Cooperative Alligator Survey Data. Date Location Total Number of Alligators Density (alligator/mile) June 11, 1969 16 mi from the S-5A Pump Station to HQ landing 706 44.1 May 1, 1970 56 mi perimeter canal 1881 33.6 June/July 1971 56 mi perimeter canal 2699 48.2 June/July 1971 56 mi perimeter canal 756 13.5 June 28, 1972 7 mi L-40 275 39.3 June 30, 1972 7 mi L-40 188 26.8 August 1, 1972 13 mi Hillsboro 445 34.2 August 2, 1972 13 mi Hillsboro 526 40.5 1976 2 transects combined 426 No Data 1977 2 transects combined 415 No Data Summer 1978 2 transects combined 182 No Data 1979 2 transects combined 335 No Data 1980 No Data 96 13.7 1981 No Data 546 78 1981 No Data 1309 100.7 1982 No Data 67 9.6 June 10, 1983 7 mi from HQ to Acme Pump Station 1 9 4 13.4 June 12, 1983 5.6 mi L-39 Hillsboro Recreation Area to S-6 47 8.4 May 22, 1985 13 mi L-39 Hillsboro Recreation Area t o S-6 449 34.5 May 24, 1985 7 mi from HQ to Acme Pump Station 1 16 4 23.4 June 26, 1986 13 mi L-39 Hillsboro Recreation Area to S-6 101 7.8 1987 9 mi from HQ to Acme Pump Station 1 95 10.6 1987 13 mi L-39 Hillsboro Recreation Area to S-6 22 6 17.4 1970-1979 The observation of apple snail egg laying activities following the introduction of a large amount of snails into the Refu ge in 1972 led scientists and managers to believe that aquatic predation was a poten tially controlling factor in snail population densities. Egg cluster production was mon itored following the release of 2,783 apple snails for stocking purposes in Impo undment C-1. Data collected showed that all snails released were either dea d or not reproducing. To ensure that the handling and transportation of the s nails was not the cause for the lack of reproduction, 100 snails were placed in eigh t cages in C-1 and subjected to the same conditions as those released into th e impoundment. The caged snails were able to reproduce, showing the handli ng and transportation was not the cause for the lack of reproduction in the un caged snails. From this,
140 little doubt was left that predation had completely d estroyed the transplanted population. Although snail populations initially flou rished during the initiation of the snail kite habitat management project, when aquati c predator populations were low, as the habitat aged the predator populatio n increased with the increase in food supply leading to a reduction in prey populat ion. In addition to the aging of the habitat, roughly 70% of the emergent vegetation was mechanically removed from C-1 and while this condition provided for maximu m snail kite use and apple snail production, it also exposed the snails to predation With restrictions on poaching and a depressed alligator hide market, the 19 70s finally began to see increased numbers of alligators. As the Everglades kite management area aged, snail num bers declined as the apple snail-predation relationship came into balan ce. The examination of the contents of one alligators stomach yielded 40 snail ope rculum and the flesh of four snails. Further research into alligator food habit s let to a suspected average of over 85% of stomach containing snail remains. The ex amination of the contents of 100 alligators stomachs yielded a total of 1 ,696 food items, of which apple snails made up 72.9%. Although alligators also p reyed on multiple species, since only snails were present in that area, their extre me preference was shown. The first instance of a predator exclusion study occurred in 1975 when a 2,000 foot long fence was erected in the eastern half o f impoundment C-5 to determine whether a fence would preclude some aquatic predator groups, such as alligators and turtles, from entering the compartmen t. Scientists aimed to determine, if predators could be restrained, would th e snail population remain at
141 a higher level compared to a non-fenced area in the s ame compartment section. Once the compartment section was drained, disked and preda tors removed, including alligators, turtles, fish and crayfish, the fou r-foot high fence was built. The fenced portion of impoundment C-5 had a 10 fold higher apple snail population than the unfenced portion. Two alligators and at least two turtles were able to enter the fenced section, by making their way u nder the fence in the soft peat, the following year, erection of a more permane nt fence would eliminate this problem. Prior to the intrusion by the predators, the apple snail population index reached 327, the number of egg clusters, a number far h igher than any index found in the area. Due to the studys success impoundment C-4 was fenced in 1977 and stocked with 3,000 adult snails. To combat youn g alligators that were climbing the fence to enter the section, an outward ove rhang was constructed. Both fenced in areas saw steady kite use, with at one po int an estimated 90% of the Refuges kite use occurring in C-4 attributed to th e high apple snail density, and stable populations. Further investigation of the effects of elimination of alligator and turtle predation on apple snail popula tions was conducted in 1979. In order to attempt to obtain an index of snail pop ulations, 545 foot long transects were established in the fenced sections. All apple snail egg clusters were counted along each transect within a four foot swathe on a monthly basis. The index peaked at 200 in impoundment C-4 shortly afte r a new fence was built. After that, similar densities were not repeated, possibl y because it was nearly impossible to eliminate all non-avian predators. With a high number of apple
142 snails in the kite management area, limpkins began taking advantage of the high numbers of snails. 1980-1989 Use of the fences was able to eliminate some nonavian predators; however by 1980 there were no observed differences i n apple snail reproductive activity between similar fenced and unfenced compartmen ts. Once again, a drought event that brought on lower than average int erior water levels in 1981 forced some alligators to seek refuge in the deeper w aters of the canals, as did other drought events that occurred throughout the 1980 s. The status of the American alligator was reviewed in 1982 at the request of the State. Heavy pressure to open a hunting season was expected due to th e healthy populations present on the Refuge. 1990-1999 The result of alligator surveys in the Refuge suggested that the alligator population in the Refuge was one of the healthiest in South Florida. Throughout the 1990s, droughts forced alligators to seek refuge in the deeper waters of the canals. Table 14 Table 15 Table 16 and Table 17 document the night time alligator surveying that began in 1998 and continued through the end of this thesis study period. From 1998 to 2002 the L-40 transect extended 10 km north of the Refuge headquarters boat ramp and the interio r transect followed a trail that ran 25.6 km east to west out to the west bucket stat ion. For some surveys,
143 the entire 25.6 km was not counted. From 2003 on, 10 km transects were counted for two interior transects, the east to west trai l out to the west bucket station from the west (Interior 1) and from the east ( Interior 2), L-40 which extended north of the Lee Road boat ramp, L-39 (1) which ran west of the Hillsboro boat ramp, and L-39 (2) which ran southeast f rom the S-6 pump station. A consistent trend throughout the survey period was alli gators moving to and concentrating in the canals during times of drought. Th e highest densities were observed in the canals. Overall, the alligator popula tion on the Refuge was very healthy and was estimated to be around 18,000.
144 TABLE 14: Interior and Canal Alligator Nighttime Survey Data from 1998 to 2000. Date Location Total Number of Alligators Number of Non Hatchling Alligators Density (non hatchling/km) August 1998 L-40 200 200 10 September 1998 Interior 69 55 5.1 November 1998 L-40 31 31 3.1 November 1998 Interior 255 119 4.5 March 1999 L-40 246 233 23.3 March 1999 Interior 164 123 4.6 August 1999 L-40 55 44 4.4 September 1999 Interior 69 55 2.1 November 1999 Interior 173 113 4.3 December 1999 L-40 238 158 15.8 March 2000 L-40 217 180 18 April 2000 Interior 124 92 9 August 2000 L-40 79 63 6.3 August 2000 Interior 74 61 5 Green represents interior and blue represents canal alli gator surveys conducted during the wet season running from May through October Purple represents interior and red represents canal alligator surveys cond ucted in the dry months running from November through April.
145 TABLE 15: Interior and Canal Alligator Nighttime Survey Data from 2001 to 2003. Date Location Total Number of Alligators Number of Non Hatchling Alligators Density (non hatchling/km) January 18, 2001 L-40 192 190 19 January 31, 2001 Interior 447 138 4.7 March 31, 2001 Interior 268 119 7.9 April 1, 2001 L-40 122 195 19.5 May 6, 2001 Interior 381 142 8.4 May 31, 2001 L-40 304 303 30.3 June 2, 2001 Interior 320 131 5.5 July 2, 2001 Interior 235 115 5.8 July 3, 2001 L-40 133 133 13.3 September 24, 2001 Interior 127 67 3.4 September 29, 2001 L-40 5 5 0.5 December 19, 2001 Interior 168 81 3.1 December 22, 2001 L-40 18 15 1.5 March 27, 2002 L-40 74 72 7.2 March 28, 2002 Interior 160 132 5.1 April 10, 2002 L-40 59 55 5.5 May 3, 2002 L-40 88 77 7.7 May 4, 2002 Interior 131 117 13 September 30, 2002 L-40 112 44 4.4 October 2, 2002 Interior 168 104 5.2 October 21, 2002 Interior 185 107 5.35 October 23, 2002 L-40 83 37 3.7 March 17, 2003 L-40 213 174 17.4 March 19, 2003 L-39 (1) 48 44 4.4 March 24, 2003 Interior (1) 87 58 5.8 March 24, 2003 Interior (2) 56 47 4.7 April 7, 2003 L-40 151 130 13 April 11, 2003 L-39 (1) 50 47 4.7 April 11, 2003 L-39 (2) 69 64 7.4 April 12, 2003 Interior (1) 88 62 6.2 April 12, 2003 Interior (2) 90 79 7.9 September 15, 2003 L-40 70 68 6.8 September 17, 2003 L-39 (1) 28 27 2.7 September 17, 2003 L-39 (2) 211 209 20.9 September 18, 2003 Marsh (1) 154 71 7.1 September 18, 2003 Marsh (2) 110 64 6.4 September 30, 2003 L-40 46 45 4.5 October 1, 2003 L-39 (1) 17 16 0.16 October 1, 2003 L-39 (2) 7 7 0.7 October 2, 2003 Interior (1) 95 58 5.8 October 2, 2003 Interior (2) 94 48 4.8 Green represents interior and blue represents canal alli gator surveys conducted during the wet season running from May through October Purple represents interior and red represents canal alligator surveys cond ucted in the dry months running from November through April.
146 TABLE 16: Interior and Canal Alligator Nighttime Survey Data from 2001 to 2003. Date Location Total Number of Alligators Number of Non Hatchling Alligators Density (non hatchling/km) March 13,2004 L-40 95 95 9.5 March 15,2004 L-39 (1) 37 37 3.7 March 15,2004 L-39 (2) 84 76 7.6 March 18,2004 Interior (1) 103 65 6.5 March 18,2004 Interior (2) 113 57 5.7 March 29,2004 L-40 102 102 10.2 March 31, 2004 Interior (1) 101 57 5.7 March 31, 2004 Interior (2) 134 86 8.6 April 1, 2004 L-39 (1) 36 36 3.6 April 1, 2004 L-39 (2) 134 86 8.6 September 20, 2004 L-40 54 51 5.1 September 23, 2004 Interior (1) 77 56 5.6 September 23, 2004 Interior (2) 40 35 3.5 September 24, 2004 L-39 (1) 31 28 2.8 September 24, 2004 L-39 (2) 39 36 3.6 October 4, 2004 L-40 37 36 3.6 October 10, 2004 L-39 (1) 23 23 2.3 October 7, 2004 Interior (1) 55 45 4.5 October 7, 2004 Interior (2) 47 37 3.7 March 7, 2005 L-40 102 100 10 March 13, 2005 Interior (1) 105 67 6.7 March 13, 2005 Interior (2) 87 66 6.6 March 14, 2005 L-39 (1) 45 42 4.2 March 14, 2005 L-39 (2) 85 85 8.5 March 22, 2005 L-40 17 17 1.7 March 31, 2005 L-39 (1) 47 47 4.7 March 31, 2005 L-39 (2) 42 42 4.2 April 1, 2005 Interior (1) 66 46 4.6 April 1, 2005 Interior (2) 62 52 5.2 September 14, 2005 L-40 38 38 3.8 September 15, 2005 L-39 (1) 31 28 2.8 September 15, 2005 L-39 (2) 65 57 5.7 September 16, 2005 Interior (1) 91 39 3.9 September 16, 2005 Interior (2) 56 43 4.3 September 29, 2005 L-39 (1) 27 26 2.6 October 3, 2005 Interior (1) 120 62 6.2 October 3, 2005 Interior (2) 46 36 3.6 October 12, 2005 L-40 34 28 2.8 March 18, 2006 Interior (1) 102 58 5.8 March 18, 2006 Interior (2) 96 92 9.2 March 19, 2006 L-40 94 93 9.3 March 20, 2006 L-39 (1) 38 38 3.8 March 20, 2006 L-39 (2) 109 94 9.4 April 1, 2006 Interior (1) 64 45 4.5 April 1, 2006 Interior (2) 79 71 7.1 April 2, 2006 L-40 101 101 10.1 April 3, 2006 L-39 (1) 44 44 4.4 April 3, 2006 L-39 (2) 114 113 11.3 September 25, 2006 Interior (1) 53 52 5.2 September 25, 2006 Interior (2) 33 33 3.3 October 9, 2006 Interior (1) 75 60 6.0 October 9, 2006 Interior (2) 54 42 4.2 Green represents interior and blue represents canal alli gator surveys conducted during the wet season running from May through October Purple represents interior and red represents canal alligator surveys cond ucted in the dry months running from November through April.
147 TABLE 17: Interior and Canal Alligator Nighttime Survey Data for 2007. Date Location Total Number of Alligators Number of Non Hatchling Alligators Density (non hatchling/km) March 13,2004 L-40 95 95 9.5 March 15,2004 L-39 (1) 37 37 3.7 March 15,2004 L-39 (2) 84 76 7.6 March 18,2004 Interior (1) 103 65 6.5 March 18,2004 Interior (2) 113 57 5.7 March 29,2004 L-40 102 102 10.2 March 31, 2004 Interior (1) 101 57 5.7 March 31, 2004 Interior (2) 134 86 8.6 April 1, 2004 L-39 (1) 36 36 3.6 April 1, 2004 L-39 (2) 134 86 8.6 September 20, 2004 L-40 54 51 5.1 September 23, 2004 Interior (1) 77 56 5.6 September 23, 2004 Interior (2) 40 35 3.5 September 24, 2004 L-39 (1) 31 28 2.8 September 24, 2004 L-39 (2) 39 36 3.6 October 4, 2004 L-40 37 36 3.6 October 10, 2004 L-39 (1) 23 23 2.3 October 7, 2004 Interior (1) 55 45 4.5 October 7, 2004 Interior (2) 47 37 3.7 March 7, 2005 L-40 102 100 10 March 13, 2005 Interior (1) 105 67 6.7 March 13, 2005 Interior (2) 87 66 6.6 March 14, 2005 L-39 (1) 45 42 4.2 March 14, 2005 L-39 (2) 85 85 8.5 March 22, 2005 L-40 17 17 1.7 March 31, 2005 L-39 (1) 47 47 4.7 March 31, 2005 L-39 (2) 42 42 4.2 April 1, 2005 Interior (1) 66 46 4.6 April 1, 2005 Interior (2) 62 52 5.2 September 14, 2005 L-40 38 38 3.8 September 15, 2005 L-39 (1) 31 28 2.8 September 15, 2005 L-39 (2) 65 57 5.7 September 16, 2005 Interior (1) 91 39 3.9 September 16, 2005 Interior (2) 56 43 4.3 September 29, 2005 L-39 (1) 27 26 2.6 October 3, 2005 Interior (1) 120 62 6.2 October 3, 2005 Interior (2) 46 36 3.6 October 12, 2005 L-40 34 28 2.8 March 18, 2006 Interior (1) 102 58 5.8 March 18, 2006 Interior (2) 96 92 9.2 March 19, 2006 L-40 94 93 9.3 March 20, 2006 L-39 (1) 38 38 3.8 March 20, 2006 L-39 (2) 109 94 9.4 April 1, 2006 Interior (1) 64 45 4.5 April 1, 2006 Interior (2) 79 71 7.1 April 2, 2006 L-40 101 101 10.1 April 3, 2006 L-39 (1) 44 44 4.4 April 3, 2006 L-39 (2) 114 113 11.3 September 25, 2006 Interior (1) 53 52 5.2 September 25, 2006 Interior (2) 33 33 3.3 October 9, 2006 Interior (1) 75 60 6.0 October 9, 2006 Interior (2) 54 42 4.2 Green represents interior and blue represents canal alli gator surveys conducted during the wet season running from May through October Purple represents interior and red represents canal alligator surveys cond ucted in the dry months running from November through April.
148 2000-2007 From 2000 to 2007 droughts forced alligators to seek re fuge in the deeper waters of the canals. The alligator population remain ed healthy throughout the time period.
149 CHAPTER 7: DISCUSSION Data Reporting Throughout the time period examined, there are def initive differences between the qualities of reporting for the factors ex amined. Apple snail and copper data is extremely underreported while snail kite insecticide, drought, other herbicides, fire, and non-avian predation are r epeatedly covered in detail throughout the Annual Narratives. From the first year the Annual Narratives were produced, 1951, the importance of the apple snail as the main food source for the endangered Everglade snail kite is recognized. Yet throughout the remainder of the study period, only one reliable set of transect data were reported. This could be due in part to a lack of established sampling methods. However, multiple times the Annual Narratives mention that tra nsects were established and data collected, but no data was reported. Additionall y, scientists were unsuccessfully attempting to correlate apple snail egg cl uster counts with apple snail density and perhaps were not reporting findings b ecause of this. Regardless of whether or not they were able to make t his correlation, the inclusion of this would provide a gauge of reproductiv e health and a general sense of whether apple snail numbers were high or low j udging by the amount of egg clusters counted. The application of copper also went significantly underreported in the Annual Narratives. This could possi bly be due to the fact
150 that another agency was responsible for the application. Hence, the data were never available for the researcher to analyze. Includi ng information in the Annual Narratives in regards to the SFWMD and their vegetati on control efforts would be helpful. Although the Refuge staff may not have perso nally been responsible for applying the copper-based herbicides, or any other herb icide for that matter, inclusion of this information would be helpful. If the Annual Narratives are a documentation of all of the happenings on and relate d to the Refuge, the authors of the narratives should be responsible for gathering d ata and information from these cooperating agencies in order to provide sufficient coverage. The occurrence of drought, other herbicides use, fire, non-avian predation, and snail kite data are all reported more in depth in the Annual Narratives. Insecticides were mentioned early on in regards to the cu ltivated crops that were on the Refuge. The amount of insecticide used or acreag e treated was not always included in the narratives. Drought and rainfal l data were consistently reported from year to year as were the trials and tr ibulations of trying to manage water levels to the schedule that had been set. Potent ial side effects that may be incurred by individual species or the ecosystem as a result of deviations to the water schedule were reported as well. The application of herbicides other than those that were copper-based was well documented through out the narratives for both experimental and in practice use. Additionally, a ny problems associated with non-target effects or leaching were also brought forwa rd as were any studies conducted that attempted to determine the effects of ap plication on individual species or the ecosystem. One issue arose similar to that wh ich occurred with
151 copper-based herbicides in that the exact herbicides being used by cooperating agencies, such as the SFWMD, were not always reported. F or example, in 1984 it was mentioned that the SFWMD was found applying an herbicide not approved for use by the Refuge, but the name of that herbicide was never revealed. This type of data could have been crucial in identifying ad ditional factors that may have contributed to the decline of the apple snail on the Refuge. The occurrence of fire on the Refuge was well documente d. However there is always the chance that some smaller fires, likel y initiated by lightning strike, that occurred were going unreported due to the sheer size of the Refuge. Non-avian predation was monitored indirectly by moni toring the health of the alligator population on the Refuge. Additional info rmation was reported regarding the use of the predator exclusions fences and transect da ta for the study would have been helpful as well. Although data was reported regarding the amount of apple snails ingested by alligators, data regarding consu mption by other predators such as turtles or limpkins. Snail kite observat ions were reported both through established surveys and what appeared to be casu al or opportunistic sightings. An increased use of tables, such as monthly wate r levels, for all of the factors examined would have increased the quality of th e data reporting. As the study period progressed from 1951 to 2007 there appeared to be a shift in the overall quality and focus of the Annual Narratives. With the first Annual Narratives came a lighter tone, a heavy focus on ducks, and a heavy focus on hunting. By the 1990s there was an increased fo cus on investigative reports and research studies. The authors of the Annual N arratives need to be
152 sure to follow up with each study to obtain whatever d ata they have collected and make it available to the public through the narrative s. Summary: Analysis of roughly seven decades of data yielded noticeable differences in the style and quality of the data that are reported. Florida Apple Snail Trends General observations were made periodically throughou t the study period; however the Annual Narratives lack any consistent data re presenting the spatial and temporal changes in apple snail populations. Per the Annual Narrative, an informal review in 1975 led to the discovery that appl e snail populations within the canals had markedly declined. Whether this was a gra dual process happening over many years or a process that occurred mor e rapidly is unknown without data. With difficulty sampling apple snails, a lack of established sampling methods and difficulty correlating egg cluster numbers t o apple snail density, very little data were reported. Low apple snail numbers were observed in 1962 and 197 0 and a disappearance or unavailability was associated with droug ht years while healthier apple snail numbers were observed in 1959, 1968, and 1982 and 1983 in Compartment C. Egg cluster counts observed in 1956, afte r a drought year, hinted that some snails were capable of surviving dry periods. During periods of low water the apple snail appeared to be concentrate d in the canals and gator
153 holes located in the interior that were able to reta in water and serve as refugia. Similar to 1956, egg clusters were observed after dry co nditions in 1962. Rising water levels following dry condition in 2000 allowed for egg cluster counts to be made post drought. Ten days after water levels rose, si x egg clusters were observed within the Refuge. The data in the Annual N arratives regarding the presence of egg clusters following significant drought ev ents lead to the belief that apple snails are able to survive. However, with a lack of either egg cluster counts or density counts, there is no information to comp are the post drought egg count to years leading up to the drought or during th e drought. Without these numbers it is impossible to determine what degree of lo ss occurred because of the drought or if reproduction hence egg cluster lyin g was temporarily suspended during the drought while the snails sought ref uge, again providing evidence of aestivation. Apple snails were commonly obser ved during 1968, when water on the Refuge was plentiful. The narrativ es state that apple snails were forced underground during spring droughts in 1970 and that they temporarily disappeared in 1997 as water levels decrea sed and the mudflats became exposed, no data were available to support thes e findings which correlate with other anecdotal evidence of apple snail aestivation in the literature review. Even though an effort was made in 1972 to establish som e snail density and distribution numbers, the data do not provide much use without a defined spatial extent from which the snails were collected and without data from prior or future years to compare it to. In general however, it is stated that apple snail
154 numbers were lower this year, a year in which drought conditions persisted, than previous year and were significantly lower than in oth er areas of the state. Similar to not having data regarding apple snail numbers with in the Refuge, a lack of data for surrounding areas, such as Lake Okeechobee and t he other WCAs, makes it not possible to determine how apple snail popul ations on the Refuge are faring in comparison to those surrounding areas. Po ssibly the most helpful piece of data provided was that of 1974 revealing tha t apple snail egg cluster counts were higher in the interior than in the canals. Previous observations and possibly data collections from 1963 to 1967 revealed ver y high densities of apple snails in the canals, yet no mention of previous numbers within the interior. Sampling in the managed impounded areas in 2000 resu lted in the reported densities of 0.05 snails/m 2 in Impoundment C-7 and 0.33 snails/m 2 in the headquarters pond. Considering that these data come from the managed areas, concern should be taken with the low densities re ported in C-7 as the estimated minimum density required to support snail kit e foraging is 0.14 snails/m 2 (Darby et al., 2006). In addition to this data, samp ling of the same impoundments with the addition of C-6 yielded very f ew apple snails sampled. This should cause some concern, especially considering Imp oundment C-8 had recently received a large apple snail egg transplant. If there is difficulty in regards to maintaining a stable apple snail population with in the managed impoundments, what does this say about the potential fo r increasing apple snail numbers in other areas of the Refuge that dont receiv e the level of management the impoundments do? This also raises question regarding the effectiveness of
155 using apple snail stocking as a method to boost and maint ain population numbers. Summary: There is limited historical information to understand whether and when snail densities in the Refu ge interior and impoundments are greater than the estima ted minimum density (0.14 snails/m 2 ) required to support snail kite foraging. Everglade Snail Kite Trends However uncertain the fate of apple snails may be as a result of drought, drought has a direct effect on snail kite populations and their ability to survive. The decrease of or absence of snail kite use on the Refug e cannot be used as a determinant of apple snail loss, but rather apple snai l unavailability. Regardless of whether or not apple snails are dying or surviving droughts by aestivating, they are unavailable to foraging snail kites. Sykes et al. (1 995) concluded that snail kites dont forage for apple snails in dry down conditi ons, so although the impoundments were managed as emergency, or suitable hab itat available for use during drought conditions when typically used habitat is not available, snail kite habitat, perhaps the dry conditions taking effect on the rest of the Refuge deter snail kites from optimally using the impounded areas. Additionally, the surrounding areas, such as the WCAs and Lake Okeechobee, m ay be more appealing to the snail kites and thus receive more use th an the Refuge for
156 reasons other than apple snail unavailability. One ex ample could be because of the high rate of nesting failure due to predation a nd poor weather conditions. These other areas may provide habitat more suitable fo r producing young to flight stage. The presence of snail kites on the Refuge could be inter preted that apple snails are present as well, however at times the snail ki tes have used the Refuge primarily as a corridor for travel. Thus it is importan t to identify snail kites establishing themselves on the Refuge, such as by the pr esence of nesting and roosting, as opposed to those that may be observed simp ly passing over the Refuge. Also increased management attention toward the snail kite in the 1970s allowed the population to increase outside of times of drought and for Refuge use to increase. Snail kite data should not be used as a gaug e of apple snail population health since many factors appear to influen ce snail kite movement. Summary: Snail kite status is not a direct indicator of apple snail abundance as snail kites dont forage for apple snails under dry down conditions, Refuge narrative snail kite observations are not consistently tied to successful foraging, and snail kites have a large geographic range relative to the size of the Loxahatchee Refuge.
157 Copper Prior to the observation that apple snail numbers had significantly declined in the canals in 1975, no concern was shown regarding th e use of copper on the Refuge. Nonetheless, the use of copper mentioned in th e Annual Narratives during the 1950s appears to have been confined to the impounded areas. With no documented use of copper in the canals in the Annua l Narratives or in a handful of recovered vegetation maintenance records, th e exact origin of the canal contamination is unknown. The fact that the canal s did receive copper in some manner is supported by testing in the late 1970s re vealing higher copper concentrations in the canal than in the interior of th e Refuge. Copper could have entered the Refuge in runoff from neighboring farms or was possibly applied by agencies working in conjunction with the Refuge, such as the SFWMD. There is also the chance that copper use by Refuge staff went undo cumented on the Refuge. Attempts to contact SFWMD to gain access to any i nformation regarding copper use were unsuccessful. Since the most notable decline of apple snails occurred in the canals, it may be necessary to contact the h istoric owners of adjacent farmlands and see if they have records of the u se of copper-based herbicides and are willing to provide them in order to better estimate the level of contamination that may have occurred within the canals. Testing results from the Annual Narratives indicated t hat copper was toxic to adult apple snail at 0.1 ppm and to two to four w eek old juvenile snails at 0.034 ppm. Testing by Winger et al. (1984) yielded 96-hour LC 50 values of 0.022 ppm and 0.024 ppm (converted from g/L for consistency) for two copper-based
158 herbicides for juvenile apple snails. Although there i s some discrepancy between these results, that can be expected from study to study a nd the values obtained from the Annual Narratives didnt include what method was used to calculate them. Samples collected throughout the Loxahatchee Ref uge by Winger et al. (1984) showed an average of 34mg/kg of copper, rangin g from 27 to 40mg/kg, however no mention of the origin of these samples wa s made. Copper has the potential to be transferred to the apple snail throu gh the water column, sediments, periphyton, and vascular plants and potential ly to its predators through bioaccumulation (Hoang et al., 2008a). However, the known use of copper on the Refuge and th e potential for contaminated waters to enter the Refuge from neighbor ing farmlands especially before the construction of the STAs make copper-based herbicides a likely factor in the decline of apple snails on the Refuge. I n addition to a mass die off that may have occurred as copper concentrations increased within the Refuge, reduced clutch production and egg hatching, as reported b y Rogevich et al. (2009), could have prevented those apple snails that di d survive from replenishing the population. The degree to which coppe r was reportedly applied in the canals, the potential for drift to occur and the effects of copper on the apple snail make an ecological effect that could devastate the e ntire population as a result likely. The potential conversion of these neighb oring farmlands to wetlands during future restoration processes could also lead to t he desorption of copper from those soils and a potential new source of copper contamination on the Refuge. If it is decided that such restoration should occu r, adjacent farm
159 sediments should be sampled and analyzed to determine i f copper concentrations existing in the sediments are higher than those of sediments on the Refuge. Higher concentrations of copper found on t he farmlands should be considered a potential threat to apple snail survival and investigated into more depth before conversion. Summary: The known use of copper on the Refuge makes copper-based herbicides a likely factor in the declin e of apple snails in the Refuge canals. Insecticides Throughout the period studied, the Annual Narrative s make mention of the use of four different insecticides on the Refuge; Toxap hene, DDT, Malathion, and Sevin. These insecticides were generally used to treat ar myworms within the impounded management areas. With the application of D DT only being mentioned once, and with direct exposure not being hig hly toxic, it needs to be determined whether or not any other repetitive use o ccurred. Repetitive use could be in the form of application by other agencies l ike the SFWMD that were not reported in the Annual Narratives. DDT becomes most toxic to non-target species with repetitive use resulting in bioaccumulation through the food chain. The literature review revealed that collection and a nalysis of apple snails and snail kites from the Refuge and surrounding areas betwee n 1965 and 1967 and of snail kite young and snail kite eggs in 1970 and 1971 resulted in low levels of
160 DDT. The low levels were thought to have represented levels of background environmental contamination. Analysis in 1965 per the Annual Narratives yielded residues in apple snails of 0.068 ppm of DDE and 0.110p pm of DDT. Similar testing occurred in 1987, 2000 and 2005. Anal ysis in 1987 yielded concentrations of DDE and Toxaphene in anhingas and lit tle blue herons and DDT and Toxaphene in fish sampled. The concentrations fo und in the avian species were low and again reflective of levels of backgro und environmental contamination. Slightly more concern was given to the co ncentrations found in fish. Although the levels indicated some type of agricul tural contamination, most likely from runoff from neighboring farms, the bans on DDT and Toxaphene means theoretically concentrations should reduce over tim e, reducing their risk. Although these values werent for apple snails or snail kite specifically, they are still significant in determining levels of contamination within the Refuge. Sediment sampling in 2000 and 2002 from the impoundments and cy press swamp with the addition of Strazulla Marsh in 2002 yielded concentrat ions of DDD and DDE, but not Toxaphene. Concentrations in 2002 for all locatio ns were not expected to adversely impact the environment. Only one concentratio n in 2000 was over the PEL, although the origin of this sample was not mentio ned. Overall, concentrations of these insecticides on the Refuge appear to be low and not of concern to Refuge managers. Since the majority of insecticides were applied within the impoundments and to adjacent farms leading to contamination withi n the canals, it needs to be determined whether or not this could have led to signi ficant declines in overall
161 apple snail populations on the Refuge. Since the use o f DDT was only mentioned once, Toxaphene twice with no obvious significant accum ulation at the time of testing and no concern shown for the remaining insecticid es, application by the Refuge should not have been the source of some of the higher concentrations found. A lack of older contamination data is to be expe cted as the effects of these insecticides were unknown at the time, without dat a from the 1950s and early 1960s, concentrations from before the ban was issued will never be known. Current data suggests that the use of these insecticides on the Refuge should not have been a factor for the decline of apple snail populations. However, since the most notable decline of apple snail occurred in the canals it may be necessary to contact the historic owners of adjacent farmla nds and see if they have records of the use of insecticides and are willing t o provide them in order to better estimate the level of contamination that may h ave occurred within the canals. Even if such data are obtained, the degree to which these insecticides are toxic to apple snails, if at all, is unknown and rat her the concern may be in the effect of these pesticides on the snail kite through b ioaccumulation. With both DDT and Toxapnhene use banned, levels of contamination throughout the Refuge should decline naturally and any increase bioaccum ulation should not be a future concern.
162 Summary: The content of the Annual Narratives suggests that the use of these insecticides on the Refuge may not have been a primary factor for the decline of apple sna il populations. Drought Often times throughout the Annual Narratives, the acti ons of the snail kite during times of drought were used as indicators of a la ck of apple snail presence. However, Sykes et al. (1995) concluded that snail kites do not forage for apple snails in dry down conditions. Therefore the conclusion ca nt be drawn that because no snail kites are present on the Refuge during times of drought that apple snails are not surviving. While some studies exist, such as Darby et al. (2008), that support the ability of apple snails to aestivate and survive droughts of certain magnitude, additional studies should be conduct ed to verify this fact. Without reliable apple snail population data, it is p ractically impossible to determine to what effect each drought that occurred ha d on the apple snails on the Refuge. One supporting observation that apple snails on the R efuge are surviving droughts is the presence of apple snail egg clusters shortly after a drought conditions let up. This was noted in 1956, 1962 and 200 0. One important piece of information to remember when looking at the surface water level data is that elevation ranges from 17 ft to 11 ft above AMSL decre asing from north to south. Using one value as a gauge of surface water level over the entire Refuge can be
163 misleading because of the changes in elevation. If the su rface water level is 15.5 ft AMSL, that would mean 2.5 ft of standing water in an area where the elevation is 13 ft and 0.5 ft of standing water in an area wher e the elevation is 15 ft. Additionally, although water levels in the canals may be lower than those in the interior, their ground elevation is going to be lowe r than that of the interior and are capable of providing deeper waters. The occurrence of drought on the Refuge can be influe nced by multiple factors including rainfall, the release of water from t he north and the release of water to the south, specifically the timing of these re leases. The analysis of rainfall data by wet and dry season is somewhat limiti ng because drought often extends from one season into another. Additionally, t he release of water from the Refuge to the south in anticipation of wet season rain fall can exacerbate drought effects when the wet season rainfall is below normal. As the severity of the drought increases, for both durat ion and extent, the chances of the apple snail surviving decrease. With droug hts generally affecting the entire Refuge, drought will elicit an ecological e ffect that will significantly affect the entire population of apple snails on the R efuge. In support of aestivation, it has been estimated that the apple snail can survive several weeks to months in drought conditions (Darby et al., 2008). There is quite a large difference, however, between surviving weeks and months. In future aestivation studies it will also be important to determine a more exact estimate of the duration for which apple snails can survive a drought. I n addition the moisture
164 retained by peat providing refuge for the apple snai l during times of drought, alligator holes are also providing refugia. Summary: Droughts do affect apple snail populations; however, without reliable apple snail population dat a, it is practically impossible to determine to what effect each drought that occurred had on the apple snails on the Re fuge. Other Herbicides Throughout the period of study, various different h erbicides were used both experimentally and in practice. Considering that the observation was made in 1975 that apple snail numbers had significantly decli ned in the canals, any herbicides that could be responsible for the decline shoul d be those utilized on the Refuge before 1975.Those herbicides used prior to 1975 should be the ones under consideration for being responsible for the declin e of apple snails. Prior to 1975, 2,4-D, VL-600, Dalapon, Amitrole-T, diquat, a nd diesel fuel were used for the management of nuisance vegetation and 2,4-D, VL-6 00, 5TC06, Chloro 16, X T B, C U M, Tween-20, Spreader-Sticker, Dalapon, amin otriazole, Chlorax-40, HC1281, Silvex, Kuron, Aquaherb, and Baron were used in experimental testing. The herbicides most heavily used were 2,4-D and diquat in both the interior of the Refuge and canals for the treatment of aquatic veg etation such as water hyacinth and water lettuce.
165 However, extensive testing conducted on the toxicity o f these herbicides towards apple snails can rule these two herbicides out as being responsible for the decline of apple snails. Once the extreme difference in apple snail populations was noticed in the canals, testing immediatel y began to determine the effects of these two herbicides. Early testing on the effect of application of 2,4-D on apple snails in 1965 lead to decreased feedin g, increased mortality, decreased egg hatching, and the death of the hatchlin gs from the eggs that did hatch. No 2,4-D residues were detected however. Additio nal 2,4-D studies were carried out in the early 1970s. Preliminary results not ed no adverse results. In addition to sampling mud, water and fish, grackles were observed as well. Diquat testing on the longevity, egg production and egg hatch ing rates led to no statistical difference between the individual tanks or t reatment groups of tanks within the study. Of the three snails that died during the study, only one was in a diquat treatment tank. From this, it was concluded that there was not direct shortterm molluscacidal effects produced by diquat. Another te st conducted on the effects on 2,4-D, diquat and diesel on apple snail hatc h rate and juvenile mortality resulted in diquat causing eggs to become tran slucent and not hatch. However no conclusion was ever reported. In conjunction w ith copper toxicity studies, copper-diquat was applied to apple snails in clos ed tanks and in field settings. Although extremely high concentrations applied within closed tanks led to apple snail mortality, death did not result from t ypical field applications. Further testing concluded that copper was the toxic agent in copp er-diquat combinations (Imlay and Winger 1980; Winger et al., 1984).
166 One 2,4-D study resulted in apple snail death, but wi th no residues of 2,4D found in the apple snails, some other factor could ha ve caused the mortality. Regardless of any results, 2,4-D and diquat use continued on the Refuge. Samples collected from across the Refuge, the Strazulla M arsh, cypress swamp, L-40, and the interior, and analyzed for contaminatio n from herbicides. Concentrations could be expected in L-40 as the herbicid es were heavily used in the canals and any concentrations in the remaining sampl ed areas would represent drift. Levels of diquat and paraquat found within the Strazulla Marsh and cypress canal were low and were determined to not a dversely affect the Refuge. Additional herbicides were used on the Refuge, especially as the invasion of melaleuca, Brazillian pepper and Old Wor ld climbing fern, and sometimes those that were not approved for use in wate r were used over water. Refuge scientists and managers appeared to investigate an y herbicides or insecticides throughout the study period that were thoug ht to be toxic to or harm apple snail populations. Since no other herbicides used before 1975 were investigated, or any other herbicides after 1975 for t hat matter, the use of herbicides other than those that are copper-based are no t responsible for significant declines in the apple snail populations on t he Refuge. Summary: The application of non-copper based herbicides 2,4-D and diquat likely did not contribute as primary factors for the decline of apple snail populations.
167 Fire While there is no question that fires, both natural a nd prescribed, occurred on the Refuge throughout the period studied, without reliable data on apple snails it is nearly impossible to correlate the affects of fire on apple snail population and to determine whether or not fire is a significant facto r in apple snail decline. Throughout the period studied fires occurred within th e interior, along the canals and within the managed impoundments with varying deg rees of magnitude. More concern was generally given to those fires with the poten tial for long deep burns that could affect the peat. Threats that could induce peat burns include lack of rainfall and decreased water levels. Low water table l evels during the dry season and periods of drought allow for a deeper burn. How ever, assuming decreased rainfall and water levels do not persist for long peri ods of time, serious peat burns are not likely to occur, as the water retained by the p eat does not provide the fire with fuel for a deep burn. When normal water levels persist on the Refuge, water in the interior marsh prevents the fire from spreadin g too far into the interior. The degree to which fire affects the Refuge in a given year can vary anywhere from a couple of acres to 40,000 acres, as was s een with the prescribed burn that spotted across a canal on May 11, 199 8 and was not declared out until July 27. In addition to the varyin g size, the proportion of land that receives a damaging peat burn may not reflect t he total amount of land burned, as seen in 1955 when a lack of rainfall resulte d in roughly 11,000 acres of fire with less than 300 acres of peat burns. The occur rence of fire on the Refuge appears to be well documented throughout the A nnual Narratives,
168 however mention of resulting acreage of peat burns is n ot as well documented. Regardless, the fact that fire has the potential to spr ead quickly and burn deep means these fires are reaching areas inhabited by apple snails. As with all living organisms, fire has the potential to inflict serious injury and potentially lead to death. However, to manage f ire in a way that would not affect apple snails is not a tactic that would be benefi cial to the Refuge as a whole. Burns are prescribed with the intent to control exotic vegetation, improve habitat and prevent deep burning peat fires by decre asing the amount of fuel available to these naturally occurring fires. While con ditions on the Refuge, such as too low or too high water levels, may prevent the interior from receiving prescribed burn treatment, ensuring that the interior receives treatment is important to prevent the build-up of fuel. The eigh t years leading up to 2000 went without any prescribed burns. Refuge managers and scienti sts need to follow the fire management plan and make all efforts to ensure th at prescribed burns are carried out in order to prevent fuel build-up. Altho ugh deep burning peat fires have the potential to inflict serious damage to the R efuge ecosystem, the benefits of naturally burning and managed fires far o ut weight any negative effects that may be incurred upon the apple snail. Whil e naturally occurring fires cant be predicted, they can be monitored and managed to prevent from spreading and intensifying to the degree that peat bu rns occur.
169 Summary: It is nearly impossible to determine whether or not fire was a significant factor in historical apple snai l population declines. Non-Avian Predation Aside from avian predators, the most serious predator t hreat to apple snail survival on the Refuge appeared to be the American al ligator. With heavy poaching and a high price on alligator hides, numbers on the Refuge were generally low during the beginning of the Annual Na rratives, but began to increase slowly throughout the 1960s and finally see hea lthy numbers in the 1970s. By the end of the study period, the alligator population on the Refuge was estimated to be roughly 18,000 and one of the health iest populations in the Everglades. Research conducted in 1972 was the first major hint that aquatic predation of apple snails by alligators was severely decr easing numbers on the Refuge. Comparing the release of 2,783 uncaged apple snails in Impoundment C-1 to 100 snails placed in cages within the impoundmen t, the uncaged snails appeared to be dead or not reproducing, while the ca ged snails were able to reproduce. Predation by aquatic predators was likely re sponsible for this outcome. Once it was presumed that alligators were responsible for decreased numbers in the managed impoundments, their stomach cont ents were examined to determine how much of their diet consisted of apple snails. Examination of 100 alligators stomach contents yielded an average of 72.9% apple snail
170 composition. In addition, the examination of one alli gators stomach yielded 40 apple snail operculum and the flesh of four snails. E ven when alligator numbers were lower in the early 1950s, alligators were moving from the interior to the canals during times of drought. While there is no data for the 1950s or early 1960s, the common trend of higher densities in the cana ls during dry periods in data in Table 6 would probably hold true for earlier years, likely w ith lower densities across the board due to lower alligator numbe rs on the Refuge. In addition to copper use in the canals, the high density of alligators during periods of drought most likely contributed to the decreased numb ers of apple snails present in the canals and had a significant effect on app le snail numbers on the Refuge. In an attempt to exclude aquatic predators, mostly all igators and turtles, predator exclusion fences were erected around some of t he managed impoundments beginning in 1975. Although initial resu lts lead to increased apple snail egg clusters, such as an index value of 327 and 200 as the habitats aged, such high densities were not repeated. This could be due to the fact that some aquatic predators were able to enter the fenced areas through the peat and inflict some predation pressure on the population of apple sna ils. Another factor could have been that with few to no aquatic predators, othe r avian species, such as limpkins, could take advantage of the high number of ap ple snails. Regardless of the cause, by 1980 there were no observed differences i n apple snail reproductive activity between similar fenced and unfe nced compartments.
171 Personal observation on the Refuge in 2010 revealed t hat no predator exclusion fences were currently in use. The increase in alligator numbers on the Refuge coinci ded with the decrease of apple snails in the canals. Early observatio ns from 1963-1967 indicated that apple snail density was once higher in t he canals. Data collected from transects in 1973 and 1974 indicated that apple sn ail densities were then higher in the canals. The degree of increased density co ncentration in the canals during times of drought and the amount sheer amount of apple snails consumed by the alligators was able to elicit an ecological effec t and severely decrease apple snail population numbers. Summary: The threat of non-avian predation was identi fied as an ongoing stressor to apple snail populations as all igator densities increase in the canals during drought condition s.
172 CHAPTER 8: CONCLUSIONS AND RECOMMENDATIONS A lack of consistently reported apple snail population da ta, whether density or egg cluster counts, make determining any signif icant relationships between apple snail decline in the Loxahatchee Refuge and the factors examined extremely difficult. However, the primary question addressed the major contributing environmental stressors to the decline of Florida apple snails, and with the data available, results in the Annual Na rratives suggest that the three factors hypothesized are likely strong drivers of a pple snail decline. Copper-based herbicide use, droughts of large magnitud e and duration, and heavy localized predation by American alligators in th e canals led to the decline of natural populations of Florida apple snails on the Loxahatchee Refuge therefore the hypothesis outlined in this study is likel y true based on supporting evidence. Each of these factors was capable of causing an e cological effect of affecting the entire population of apple snails in the Refuge. Figure 16 summarizes the stressors examined and identifies their ef fects on the Refuge apple snail population decline based on the results and conclusions drawn.
173 FIGURE 16: Final Florida Apple Snail Stressors Conceptual Model. Lack of reliable sampling methods and consistent reporting made correlating changes in apple snail population to the v arious anthropogenic and natural processes investigated almost impossible. In order to gauge the health and status of apple snail populations on the Refuge, a monitoring network needs to be established and maintained. Without population data, there is no way to know to what extent apple snail numbers are changing on a monthly or yearly basis and over what type of spatial extent these chang es are occurring. By accumulating a data set, apple snail numbers can be corr elated more precisely to events affecting the Refuge, such as declines in surface wa ter level due to drought, and decisions can be made to determine what p rocesses do and do not have effects on the apple snails and how severe the proce ss needs to be before a change in population occurs. The monitoring network needs to encompass not only the various habitats maintained across the Refuge, but also be compatible wi th monitoring in areas
174 outside of the Refuge landscape, including Lake Okeechob ee and WCA 2 and 3. A cooperative effort between the Refuge and other in terested agencies, such as the SFWMD or academic institutions, would aide in the systematic collection of data across the region. By also monitoring the populati ons of apple snails in these areas, Refuge scientists and managers can compare tren ds occurring on the Refuge to those of the surrounding South Florida habitats. An important factor in maintaining the monitoring network is to coll ect data on a monthly basis for multiple years in order to gain representative da ta of the trends occurring on throughout the region. Data collection locations need to be established in the managed impoundments, within the interior and along the edge of the canals. In order to monitor the apple snail population, egg cluster counts m ust be made to monitor the reproductive status of the snails in addition to est imations of density. The throw-trapping methodology established by Darby et al (1999) should be used for data collection. In order to receive a representati ve count of egg clusters, a 5 or 10m 2 quadrat should be flipped end over end 20 times alon g established transects. Each 5 or 10m 2 quadrat can be considered a sampling unit. A 1m 2 throw trap should be used in order to estimate apple snail density. Once thrown, the trap is to be pushed into the substratum to preven t any apple snails from escaping. Stand in one place and throw. Vegetation upr ooted, rinsed and examined for snails (Darby et al., 1999). Per previou s sampling efforts, a dip net should be used to clear the throw trap of apple snails. After removal of vegetation from the throw trap, the dip net should be used to r emove the apple snails until
175 20 consecutive sweeps yield no snails. At least 50 throw t rap samples should be collected per site in order to obtain some precisions in r esults. According to Darby et al. (1999), the collected density estimation needs to be adjusted by dividing by the capture probability (CP). A known am ount of snails are marked, placed within the throw trap before the vegetation i s removed, and the percentage of those snails recovered represents the CP. S ince no correlation has been able to be made between egg cluster counts and estimates of snail density in Loxahatchee Refuge, both egg clusters and ap ple snails need to be counted in an effort to use both data sets to gauge po pulation health. Since it may not always be possible to obtain a reliable egg cl uster count, due to low water levels or sampling occurring outside of the months of reproduction, more weight should be placed on the data collected from th e density data. In regards to apple snail counts, concern should be taken when densit ies drop below 0.14 snails/m 2 the estimated minimum density required to support snail kite foraging (Darby et al., 2006). In order to maintain and manage data being collected during apple snail monitoring, including both egg cluster counts and densit y estimates, Geographic Information Systems (GIS) and Global Positioning System s (GPS) should be used to establish a database in which to store the data. With the vegetation damage that comes with using a throw trap for sampling by uprooting the vegetation with each throw, it may be necessary to defi ne various spatial extents for which the throw traps may be deployed. Similar to establishing transects for egg cluster counts, areas of similar habitat cover should be established in which
176 throw traps can be used to sample from a month-to-month basis without compromising sampling due to vegetation damage (e.g., establishing 10 m by 10 m areas from which to sample from each month). These sa mpling units should be established within the impoundments, along the canal and in the interior, just as with the transects. A GPS can be used to record the la titude and longitude of the points of the sampling area and the two endpoints of each transect. This data can then be imported into a geodatabase, such as that wi thin ESRIs ArcGIS, and then digitized to represent the mapped polygon sa mpling units and transect lines, each their own feature class. In regards to collect ing density data, a GPS location can be collected each time the throw trap is d eployed within its sampling unit to be imported into the geodatabase and represe nted as a point location feature class. For each point, attribute data can be in cluded in the form of what sampling area the data was collected from, the date the data was collected, who collected the data, how many snails were counted per th row, CP, density, any pertinent observations while in the field, reasons why samples were unable to be collected, etc. For each transect, a similar approach can be taken. Since the transect will remain the same from month-to-month, it will not be necessary to collect endpoint data during every count. Rather an asso ciated table can be created for each transect that contains attribute data si milar to that found within the throw trap point location layer. Storing data in this manner will allow its users to visualize, query and analyze the data as well as di sseminate the data to other agencies and the public. Survey results should be presente d in future Annual Narratives to improve the utility of those reporting vehicles.
177 Despite the fact that copper-based herbicide use seems part ly responsible for decline of apple snail numbers, especially in the can als, there does not appear to be enough information available to establi sh whether any historical spatio-temporal relationship between snail kites and app le snails existed in relation to copper-based herbicide use. Therefore, t he evidence supports that a significant relationship cannot be determined. A lack of data, both for copper use and apple snail abundance and distribution, prevents a ny connection from being made. From what little data were provided, the majo rity of the copper use that took place on the Refuge would have affected the canal s. Decrease use by snail kites seemed to correlate with drought data more than a ny other data with generally low Refuge use as a whole, and not necessaril y for any specific part of the Refuge. In order to determine to what degree cop per contaminated the canals, it may be necessary to contact the historic owners of adjacent farmlands and see if they have records of the use of copper-based herbicides and are willing to provide them. If no data can be provided or if enough data are provided to show that enough copper-based herbicide was applied w ithin the canal to cause toxic levels to be available to apple snails, sample s should be collected and analyzed to determine if any toxic levels are still present in sediments, periphyton, vascular plants, and water and if any signi ficant bioaccumulation has occurred within the apple snails. From each site sampled, apple snails should be collected and tested to determine to what degree, if a ny, bioaccumulation has occurred within the Refuge. In addition, chronic exposu re to copper can result in
178 reduced snail survival at levels much lower than acute to xicity values (ReedJunkins et al., 1997). As with other aspects of this research, a lack of apple snail abundance and distribution data makes it impossible to determine a fter what extent and what historical droughts actually diminished populations of a pple snails of the Refuge. Yet even without this data it is evident that whether apple snails are dying or surviving droughts by aestivating, they are unavailabl e to snail kites during times of drought. Heavy snail kite use could be interpreted a s an indicator of health apple snail numbers, however the snail kites have been kn own to use the Refuge primarily as a travel corridor at times. Documenting re curring nesting and roosting could be useful indicators that the Refuge ac tually has sufficient established populations of apple snails for snail kites be yond than just transient use. If any restoration is to take place on farmland adjacen t to the Refuge, desorption of copper from flooded agricultural soils coul d pose a serious threat to the Refuge by reintroducing toxic levels of copper. If it is decided that such restoration should occur, adjacent farmland sediments sh ould be sampled and analyzed to determine if copper concentrations existing in the sediments are higher than those of sediments on the Refuge and at l evels that could be potentially lethal to apple snails. Higher concentratio ns of copper found on the farmlands should be considered a potential threat to a pple snail survival and the potential effects should be investigated in more depth before conversion. The use of additional treatment or buffer wetlands on the east and west sides of the
179 Refuge could aid in reducing the amount of copper to b e received by the Refuge from both runoff from these farmlands and from waters used to flood these lands in an effort to convert them. Enough information was available from the Annual Na rratives and associated literature to provide recommendations to the Refuge. As already mentioned, establishment of a consistently collected and reported monitoring network will be the first step towards improving manage ment techniques. Since stocking the managed impoundment areas have not produce d sustained increased apple snail densities, attempts to stock other are as of the Refuge, such as the interior, may provide the desired results. A ccording to the Annual Narratives, apple snails were generally stocked in the m anaged impoundments and although this may have initially resulted in high reproduction and increased apple snail numbers, these results were not long lasting. Instead of transplanting apple snails to the impoundments, an attempt should be made to stock apple snails in the interior. Doing so may provide longer la sting result in regard to providing increased apple snail numbers. Although preda tor exclusion was initially successful in the managed impoundments, this isnt something that could be reproduced in the interior. The alligator holes d ug by alligators in the interior provide refugia for various organisms during times of d rought. When stocking the interior with apple snails, in addition to the allig ator holes, artificial holes might be dug thus ensuring that during longer periods of droug ht there are adequate refugia. Transects and throw trap surveys should be estab lished in the areas of the interior where apple snails are stocked in order to monitor the success. By
180 utilizing the interior in addition to the managed i mpoundments, apple snails may have a better chance of repopulating areas where they once were and provide snail kites with additional feeding grounds as the inter ior was once utilized more frequently by the snail kites. Increased use of the int erior by the snail kites is a hopefully sign that stocking in the interior will prov ide additional sources of food. With the increased presence of research and investigati ve studies being covered in the Annual Narratives beginning in the mid -1990s, it can be observed that a significant change has occurred from 1951 when the first narrative was produced and there was a heavy focus on ducks and hunting Evidence from the Annual Narrative support the hypothesis that the quali ty of the data would improve as the time period progressed. The authors of t he Annual Narratives need to be sure to follow up with each study mentione d in order to obtain whatever data have been collected and make it availabl e to the public through the narratives. The overall quality of the Annual Na rratives was good, however as already mentioned there was a complete lack of data fo r copper-based herbicide use and apple snail population data. In addition to f ollowing up with any studies mentioned in the narratives, the authors need to make sure to acquire any information that cooperating agencies may have regard ing their work on the Refuge. Additionally, certain objectives to be accomplished wer e set. These objectives included analysis of the Annual Narratives, asse ssing the abundance and distribution of apple snails, incorporating recent co pper and drought related apple snail findings into the synthesis on the ecology o f the apple snails,
181 assessing the quality of historical data reporting, and p roposing initial establishment of an apple snail monitoring network in t he interior should be used for data collection as it was applied in the most recent apple snails surveys in the Refuge in 2002-2004 (Darby et al., 2006). Each object ive was accomplished with the exception of assessing the historical abundance and di stribution of the apple snail. As already reported, the Annual Narratives were lacking in consistently reported and reliable apple snail data. Without thes e data, it is difficult to assess exactly how each factor examined affected the apple sna ils abundance and distribution, only that it did or did not have an a ffect.
182 LITERATURE CITED Brandt LA (2006) Benefits Anticipated from the 1995 W ater Regulation Schedule for Water Conservation Area 1: Review and Analysis. Re port number: LOX06-006. U.S. Fish and Wildlife Service, Boynton Be ach, Florida. 52 pp. Available at: http://sofia.usgs.gov/publications/reports/wca1_review/W CA1_rev2006_Ac c5.pdf Carson R (1962) Silent Spring. Houghton Mifflin, Bost on. Cooch E, Rockwell RF, and Brault S (2001) Retrospective Analysis of Demographic Responses to Environmental Change: A Lesser Snow Goose Example. Ecological Monographs 71:377-400 Cooke AS (1973) Shell thinning in avian eggs by envir onmental pollutants. Environ Pollut 4:85-152 Cottam C and Knappen P (1939). Food of some uncommon North American birds. Auk 56:138-169. Darby PC, Bennetts RE, Croop JD, Valentine-Darby, PL, and Kitchens WM (1999) A Comparison Of Sampling Techniques For Quantif ying Abundance Of The Florida Apple Snail ( Pomacea Paludosa Say). J Moll Stud 65:195-208 Darby PC, Benetts RE, Miller SJ, and Percival HF (200 2) Movements of Florida Apple Snails in Relation to Water Levels and Drying E vents. Wetlands 22:489-498 Darby PC, Valentine-Darby PL, and Percival HF (2003) Dry season survival in a Florida apple snail ( Pomacea paludosa Say) population. Malacologia 45:179-184
183 Darby PC, Bennetts RE, and Karunaratne LB (2006) App le snail densities in habitats used by foraging snail kites. Florida Field Na turalist 34: 37-47 Darby PC, Bennetts RE, and Percival HF (2008) Dry Dow n Impacts on Apple Snail ( Pomacea paludosa ) Demography: Implications for Wetland Water Management. Wetlands 28:204-214 de Oliveira-Filho EC, Lopes RM, and Paumgartten FJR ( 2004) Comparative study on the susceptibility of freshwater species to coppe r-based pesticides. Chemosphere 56:369-374 Frakes RA, Bargar TA, Bauer EA (2008) Sediment copper bioavailability to freshwater snails in south Florida: risk implications for the Everglades snail kite ( Rostrhanus sociabilis plumbeus ). Ecotoxicology 17:598-604 Fry DM (1995) Reporductive effects in birds exposed to pesticides and industrial chemicals. Environ Health Perspect 103:165-171 Furniss E (2001) Timeline History and the Anzac Myth: S ettler Narratives of Local History in a North Australian Town. Oceania 71:2 79-297 Gaiser E (2009) Periphyton as an indicator of restorat ion in the Florida Everglades. Ecological Indicators 9:S37-S45 Galiulin RV, Bashkin VN, and Galiulina RA (2005) Eco logical Risk Assessment of Riverine Contamination in the Caspian Sea Basin: A Co nceptual Model for Persistent Organochlorinated Compounds. Water, Air, & S oil Pollution 63:33-51. Gentile JH and Harwell MA (1998) The issue of significan ce in ecological risk assessments: A scientific perspective. J Hazard Mater 35:241 -253 Goforth BR and Minnich RA (2007) Evidence, Exaggerati on, and Error in Historical Accounts of Chaparral Wildfires in California. Ecological Applications 17:779-790
184 Graham MH and Dayton PK (2002) On the Evolution of Ecological Ideas: Paradigms and Scientific Progress. Ecology 83: 1481-1489. Harris DL and Witcomb WH (1974) Effects of Fire on Popu lations of Certain Species of Groung Beetles (Coleoptera: Carabidae). The Florida Entomologist 57: 97-103 Hawkins JW, Lankester MW, and Nelson RRA (1998) Sampli ng terrestrial gastropods using cardboard sheets. Malacologia 39:1-9 Harwell MA and Gentile JH (2006) Ecological Significanc e of Residual Exposures and Effects from the Exxon Valdez Oil Spill. Integrated Environmental Assessment and Management 2:204-246 Harwell MC, Surrat DD, Barone DM, and Aumen, NG (20 08) Conductivity as a tracer of agricultural and urban runoff to delineate water quality impacts in the northern Everglades. Environ Monit Assess 147:445-4 62 Harwell MC (2009) Personal communication. Senior Ecolog ist, U.S. Fish and Wildlife Service, Boynton Beach, FL Hoang TC, Rogevich EC, Rand GM, Gardinali PR, Frakes RA, and Bargar TA (2008a) Copper desorption in flooded agricultural soil s and toxicity of the Florida apple snail ( Pomacea paludosa ): Implications in Everglades restoration. Environmental Pollution 154: 338-347 Hoang TC, Rogevich EC, Rand GM, and Frakes RA (2008b) Copper uptake and depuration by juvenile and adult Florida apple snail s ( Pomacea paludosa ). Ecotoxicology 17:605-615 Hoang TC and Rand GM (2009) Exposure routes of coppe r: Short term effects on survival, weight, and uptake in Florida apple sn ails ( Pomacea paludosa ). Chemosphere 76:407-414 Hoang TC, Schuler LJ, Rogevich EC, Bachman PM, Rand GM, and Frakes RA (2009) Copper Release, Speciation, and Toxicity Follow ing Multiple Floodings of Copper Enriched Agriculture Soils: Implicat ions in Everglades Restoration. Water Air Soil Pollut 199:79-93
185 Imlay MJ and Winger PV (1980) Loxahatchee Apple Snail Copper Problem. USFWS report on file. Available at: http://mwaldon.com/Loxahatchee/GrayLiterature/Imlay-W inger-19xx.doc Jaga K and Brosius D (1999) Pesticide exposure: human can cers on the horizon. Rev Environ Health 14:39-50 Karunaratne LB, Darby PC, and Bennetts RE (2006) The Effects of Wetland Habitat Structure on Florida Apple Snail Density. We tlands 26:1142-1150 Kuhn TS (1962) The structure of scientific revolutions. University of Chicago Press, Chicago, Illinois, USA Lamont T and Reichel W (1970) Organochlorine Pesticide Residues in Whooping Cranes and Everglade Kites. The Auk 87:158-159 Leslie AJ (1992) Copper Herbicide Use-Patterns in Flori da Waters. In Proc. Of the Bioavailability of Copper. Gainesville, FL, pp. 73-88 Lunt ID and Spooner PG (2005) Using historical ecology to understand patterns of biodiversity in fragmented agricultural landscapes. J ournal of Biogeography 32: 1859-1873 Mastin BJ and Rodgers JH (2000) Toxicity and Bioavailabi lity of Copper Herbicides (Clearigate, Cutrine-Plus, and Copper Sulfa te) to Freshwater Animals. Arch Environ Contam Toxicol 39:445-451 Mitra J and Raghu K (1998) Pesticides Non Target Plan ts Interactions: An Overview. Arch Acker-Pft Bodern 43:445-500 Mooij WM, Bennetts RE, Kitchens WM, and DeAngelis DL (2 002) Exploring the effect of drought extent and interval on the Florida snail kite: interplay between spatial and temporal scales. Ecological Modelling 149:25-39
186 Naeem S (2002) Ecosystem Consequences of Biodiversity Lo ss: The Evolution of a Paradigm. Ecology 83:1537-1552 Nekola JC (2002) Effects of fire management on the richn ess and abundance of central North American grassland snail faunas. Animal Bi odiversity and Convervation 25:53-66 Newsom LD (1967) Consequences of Insecticide use on Nontarg et Organisms. Annual Review of Entomology 12:257-286 Ogden JC (2005) Everglades Ridge and Slough Conceptual Ecological Model. Wetlands 25:810-820 Ogden JC, Davis SM, Jacobs KJ, Barnes T, and Fling HF (200 5) The Use of Conceptual Ecological Models to Guide Ecosystem Restoration in South Florida. Wetlands 25:795-809 Pauly WE (1985) How to manage small prairies fires. D ance County Environmental Council, Madison, Wisconsin Payne GG, Xue SK, and Weaver KE (2009) Chapter 3A: Status of Water Quality in the Everglades Protection Area. South Florida Envi ronmental Report Peugh JL and Enders CK (2004) Missing Data in Educationa l Research: A Review of Reporting Practices and Suggestions for Improv ement. Review of Educational Research 74:525-556 Pratt JR, Melendez AE, Barreiro R, and Bowers NJ (1997 ) Predicting the Ecological Effects Of Herbicides. Ecological Applications 7:1117-1124 Rand GM, Carriger JF, Lee TA, and Pfeuffer RJ (2004) Sediment Toxicity in the St. Lucie River Watershed and Everglades Agricultural Area. Ecotoxicology 13:261-274
187 Rand GM and Schuler LJ (2009) Aquatic Risk Assessment of Met als in Sediments from South Florida Canals. Soil and Sedimen t Contamination: An International Journal 18:155-172 Rodgers Jr. JA, Smith HT, and Thayer DD (2001) Integr ating Nonindigenous Aquatic Plant Control with Protection of Snail Kite N ests in Florida. Environmental Management 28:31-37 Rogevich EC, Hoang TC, and Rand GM (2008) The Effects of Water Quality and Age on the Acute Toxicity of Copper on the Florida Ap ple Snail, Pomacea paludosa Arch Environ Contam Toxicol 54:690-696 Rogevich EC, Hoang TC, and Rand GM (2009) Effects of S ublethal Chronic Copper Exposure on the Growth and Reproductive Success of the Florida Apple Snail ( Pomacea paludosa ). Arch Environ Contam Toxicol 56:450458 Ross MS, Mitchell-Bruker S, Sah JP, Stothoff S, Ruiz PL, Reed DL, Jayachandran K, and Coultas CL (2006) Interaction of Hy drology and Nutrient Limitation in the ridge and Slough Landscape of the Southern Everglades. Hydrobiologia 569: 37-59 Schneider MF (2001) Habitat Loss, Fragmentation and Pr edator Impact: Spatial Implications for Prey Conservation. Journal of Applied Ecology 38:720-735 Schuler LJ, Hoang TC, and Rand GM (2008) Aquatic risk ass essment of copper in freshwater and saltwater ecosystems of South Florida. Ecotoxicology 17:642-659 Sharfstein B and Steinman AD (2001) Growth and surviv al of the Florida apple snail ( Pomacea paludosa ) fed 3 naturally occurring macrophyte assemblages. Journal of the North American Benthological Society 20:8495 Sih A, Englund G, and Wooster D (1998) Emergent imp acts of multiple predators on prey. Trends in Ecology and Evolution 13:350-355
188 South Florida Water Management District (SFWMD) (2006 ) DBHYDRO Browser User Documentation. Available at: http://www.sfwmd.gov/portal/page/portal/xrepository/sf wmd_repository_pd f/dbhydrobrowseruserdocumentation.pdf South Florida Water Management District (SFWMD) (2010 ) South Florida Water Management Model (SFWMM). Available at: http://my.sfwmd.gov/sfwmd/ hesm/hpm/frame1/sfwmm.htm Swetnam TW, Allen CD, and Betancourt JL (1999) Appl ied Historical Ecology: Using the Past to Manage for the Future. Ecological Ap plications 9:11891206 Sykes Jr. PW (1983) Snail Kite use of freshwater marshes of South Florida. Florida Field Naturalist 11:73-88 Sykes Jr. PW (1985) Pesticide Concentrations in Snail Kite Eggs and Nestlings in Florida. The Condor 87:438 Sykes Jr. PW (1987) The Feeding Habits of the Snail Ki te in Florida, USA. Colonial Waterbirds 10:84-92 Sykes Jr. PW, Rodgers Jr. JA, and Bennetts RE (1995) Snai l Kite ( Rostrhamus sociabilis ). In: Poole, A., Gill, F. (Eds.) The Birds of North America. The Academy of Natural Sciences, Philadelphia, and the Amer ican Ornithologists Union, Washington, DC, No. 171 Thompson FG (2004) An Identification Manual for the F reshwater Snails of Florida. Florida Museum of Natural History, Universit y of Florida Toft CA and Jaeger RG (1998) Writing for Scientific Jo urnals I: The Manuscript. Herpetologica 54:5-42-S54 Towne G and Owensby C (1984) Long-term effects of annu al burning at different dates in ungrazed Kansas tallgrass praries. Journal of Ra nge Management 37:392-397
189 Turner RL (1994) The effects of hydrology on the popul ation dynamic of the Florida apple snail (Pomacea paludosa). St. Johns Wat er Management District, Palatka, FL, USA. Technical Publication SJ9-SP 3 U.S. Fish and Wildlife Service (USFWS) (2000) Arthur R. Marshall Loxahatchee National Wildlife Refuge Comprehensive Conservation P lan. U.S. Fish and Wildlife Service, Atlanta, GA U.S. Fish and Wildlife Service (USFWS) (2004) Multi-S pecies Recovery Plan for South Florida: Snail Kite. Available at: http://www.fws.gov/verobeach/images/pdflibrary/snki.pdf U.S. Fish and Wildlife Service (USFWS) (2007) Arthur R. Marshall National Wildlife Refuge Annual Narrative Report: Calendar Y ear 2007. U.S. Fish and Wildlife Service, Boynton Beach, FL Vermeer K, Spaans AL, and Reynolds LM (1974) Pesticide effects on fishes and birds in rice fields Surinam, South America. Environ. Pollut. 7:217-236 Wallington TJ and Moore SA (2005) Ecology, Values, an d Objectivity: Advancing the Debate. Bioscience 55:873-878 Wehle C (2009) Reviving the River of Grass: Everglade s Land Acquisition Project. South Florida Water Management District, Execu tive Director. Available at: https://my.sfwmd.gov/portal/page/portal/common/newsr/ rog_pres_commu nity_briefing_0908.pdf Wilson RA (2007) Combining Historical Research and Nar rative Inquiry to Create Chronicles and Narratives. The Qualitative Report 12:2 0-39 Winger PV, Imlay MJ, McMillan WE, Matrin TW, Takekawa J, and Johnson WW (1984) Field and Laboratory Evaluation of the influ ence of Copper-Diquat on Apple Snails in Southern Florida. Environ Toxicol and Chem 3:409424