Citation
Interannual Lake Temperature Trends as Indicators for Climatic Change in Florida

Material Information

Title:
Interannual Lake Temperature Trends as Indicators for Climatic Change in Florida
Creator:
COENEN, DANNY
Copyright Date:
2008

Subjects

Subjects / Keywords:
Climate change ( jstor )
Climate models ( jstor )
Heating ( jstor )
Lakes ( jstor )
Latitude ( jstor )
Precipitation ( jstor )
Surface temperature ( jstor )
Surface water ( jstor )
Temperature gradients ( jstor )
Water temperature ( jstor )
City of Gainesville ( local )

Record Information

Source Institution:
University of Florida
Holding Location:
University of Florida
Rights Management:
Copyright Danny Coenen. Permission granted to the University of Florida to digitize, archive and distribute this item for non-profit research and educational purposes. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder.
Embargo Date:
8/31/2008
Resource Identifier:
495635991 ( OCLC )

Downloads

This item is only available as the following downloads:


Full Text

PAGE 1

INTERANNUAL LAKE TEMPERATURE TRENDS AS INDICATORS FOR CLIMATIC CHANGE IN FLORIDA By DANNY COENEN A THESIS PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLOR IDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE UNIVERSITY OF FLORIDA 2005

PAGE 2

Copyright 2005 by Danny Coenen

PAGE 3

To Jennifer and Julia, for th eir love, patience and support.

PAGE 4

ACKNOWLEDGMENTS This project would not have been possible without the support of my committee chair, Dr. Thomas L. Crisman, whose invaluable advice helped guide my work from the planning stages through the final manuscript revisions. Dr. Fernando R. Miralles-Wilhelm’s assistance in developing proper statistical procedures for data analysis is greatly appreciated. To Dr. Ellen E. Martin I owe much of my understanding of past and present climate change, whereas Dr. Mark Brenner’s expert knowledge on the limnology of Florida lakes was also most helpful, as were his particularly detailed suggestions for corrections and improvements to the manuscript. I thank them both. I appreciate the assistance of all individuals who provided me with data for inclusion in this project: Larry Battoe (SJRWMD), Aisa Cerik (SJRWMD), Larry Connor (Florida Fish and Wildlife Conservation Commission), David Hornsby (SJRWMD), Christy Horsburgh (Florida Lakewatch), Ted Lange (Florida Fish and Wildlife Conservation Commission), Steve Winkler (SJRWMD) and Catherine Wolden (SWFWMD). I would also like to acknowledge the following individuals who have enriched my education through their extraordinary efforts: Dr. Paul Rollinson of Southwest Missouri State University, without whom I would not have chosen to pursue studies in the environmental sciences; Bernd Flecke, Peter-Joachim Reichard and Jan Keller of the Gymnasium Zitadelle der Stadt Jlich, who prepared me exceptionally well for iv

PAGE 5

university; and Annelise Dolfus of the Gemeinschaftsgrundschule Jlich-Nord, whose instruction formed the foundation for my academic success. The love, patience and encouragement of my wife, Jennifer, and daughter, Julia Skye, are sincerely appreciated. My special thanks are extended to Brian Zahn, who was ready to support me without question when no others would. Finally, my sincere thanks and appreciation are extended to my parents, Rosemarie and Franz Peter, whose encouragement, support and love were instrumental to the realization of my academic ambitions. v

PAGE 6

TABLE OF CONTENTS page ACKNOWLEDGMENTS .................................................................................................iv LIST OF TABLES ...........................................................................................................viii LIST OF FIGURES ...........................................................................................................ix ABSTRACT .......................................................................................................................xi CHAPTER 1 INTRODUCTION........................................................................................................1 Ecological Implications of Climate Change.................................................................4 Impacts Specific to Florida...........................................................................................5 Global Climate Change during the Study Period.........................................................7 Climate of Florida.........................................................................................................9 Air Temperature..................................................................................................10 Precipitation Patterns...........................................................................................13 Precipitation Regimes..........................................................................................15 Objectives and Scientific Significance.......................................................................18 2 DATA AVAILABILITY AND SITE SELECTION..................................................20 Characteristics of Florida Lakes.................................................................................20 Data Availability.........................................................................................................27 Site Selection and Data Collection.............................................................................29 3 THERMAL CHARACTERISTICS OF STUDY LAKES.........................................35 Seasonal Patterns........................................................................................................35 Latitude-Temperature Relationships..........................................................................39 Relationships between Temperature and Other Variables..........................................42 4 ANALYSES FOR TEMPORAL CHANGE..............................................................44 Regression Analyses for Interdecadal Temperature Variability.................................44 Seasonal Temperature Trajectories.............................................................................52 Cluster Analysis..........................................................................................................57 vi

PAGE 7

5 DISCUSSION AND SUGGESTIONS FOR FUTURE RESEARCH........................63 Relationship between Air and Lake Temperatures.....................................................63 Temporal Trends.........................................................................................................64 Ecological Implications..............................................................................................66 Likely Effects during the Study Period...............................................................66 Future Scenarios..................................................................................................68 Suggestions for Future Research................................................................................71 APPENDIX A LIST OF STUDY LAKES..........................................................................................73 B LIST OF REFERENCED WEATHER STATIONS..................................................75 LIST OF REFERENCES...................................................................................................76 BIOGRAPHICAL SKETCH.............................................................................................80 vii

PAGE 8

LIST OF TABLES Table page 5-1. ENSO during the study period...................................................................................66 A-1. List of study lakes, their location and surface area...................................................73 B-1. Weather stations queried to establish the air temperature-latitude relationship in Figure 3-2.................................................................................................................75 viii

PAGE 9

LIST OF FIGURES Figure page 1-1. Life zones of Florida.....................................................................................................6 1-2. Variation in mean global air temperature in C over land from 1900 to 2000 relative to the 1961 to 1990 average.......................................................................................8 1-3. Florida climate classifications....................................................................................10 1-4. Beginning of meterological seasons in Florida..........................................................13 2-1. Geologic map of northern peninsular Florida.............................................................21 2-2. Landsat satellite view of the lake district of peninsular Florida.................................22 2-3. Distribution of named Florida lakes by surface area..................................................23 2-4. Distribution of true color in Florida lakes..................................................................25 2-5. Thermal regimes of Florida lakes...............................................................................26 2-6. Location of study lakes...............................................................................................32 2-7. Distribution of study lakes by size.............................................................................33 2-8. Peninsular lake district in relation to Thornthwaite climate zone boundaries and Dohrenwend and Harris life zone boundaries..........................................................34 3-1. Surface water temperatures of a north, central and south Florida lake compared to ambient air temperatures..........................................................................................38 3-2. Comparison of mean annual air and water temperatures by latitude.........................39 3-3. Mean annual surface water temperature range by latitude.........................................41 4-1. Location of lakes included in the analyses for temporal change................................45 4-2. Mean ten-year summer surface water temperatures by latitude.................................48 4-3. Mean winter surface water temperatures by latitude during 1975-1984....................49 ix

PAGE 10

4-4. Mean winter surface water temperatures by latitude during 1985-1994....................50 4-5. Mean winter surface water temperatures by latitude during 1995-2000....................51 4-6. Mean winter surface water temperatures by latitude during 1975-1984 with alternative model......................................................................................................52 4-7. Mean decadal winter surface temperature by latitude (overview).............................53 4-8. Winter temperature trends..........................................................................................55 4-9. Summer temperature trends.......................................................................................56 4-10. Dendrogram of winter surface water temperatures based on 22 Florida lakes between study periods 1 (1975-1984) and 2 (1985-1994).......................................58 4-11. Zonal boundaries of thermal regimes of Florida lakes based on Figure 4-10..........59 4-12. Dendrogram of winter surface water temperatures based on 22 Florida lakes between study periods 2 (1985-1994) and 3 (1995-2004).......................................60 4-13. Zonal boundaries of thermal regimes of Florida lakes based on Figure 4-12..........61 5-1. Thermal regime boundaries according to Beaver et al. (1981) relative to the study area...........................................................................................................................69 5-2. Thermal regimes zones after northward shift following warming of 3C based on extant temperature-latitude relationships.................................................................70 x

PAGE 11

Abstract of Thesis Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Master of Science INTERANNUAL LAKE TEMPERATURE TRENDS AS INDICATORS FOR CLIMATIC CHANGE IN FLORIDA By Danny Coenen August 2005 Chair: Thomas L. Crisman Major Department: Interdisciplinary Ecology Monitoring data from 50 peninsular Florida lakes collected between 1968 and 2004 were examined to determine the influence of locational, morphometric and biological variables on surface water temperatures and to evaluate detectability of long-term climatic change. Annually, latitude was the only factor to exert statistically significant control on water temperature (r 2 =0.74) and mean annual temperature range between warmest and coldest month (r 2 =0.73) at the p=0.1 level. Surface area, mean chlorophyll a and mean true color did not exert systematic influence during the study period. Potential influence of lake depth could not be conclusively determined, as bathymetry was unavailable for several study lakes. Seasonally, latitudinal control was most pronounced during winter and broke down consistently during summer. Because seasonal latitude-lake temperature patterns closely resemble air temperature gradients, lacustrine records are suggested as suitable model proxies. xi

PAGE 12

Statistical comparison of summer and winter temperature means in ten-year intervals provided no evidence for long-term change during summer, whereas winters displayed a complex pattern of interdecadal variability marked by rapid warming of 2.1.0C (SD) between study intervals 1 (1975-1984) and 2 (1985-1994), followed by moderate cooling of 1.0.8C between intervals 2 and 3 (1995-2004). Analysis of time series regression slopes produced an estimate of 1.0.9C for winter warming between 1973 and 2003. Because of the constancy of summer temperatures and non-unidirectional trajectory of winter temperatures, available data were insufficient to indicate global climate change as the causal mechanism for decadal trends in lake temperatures. Clusters of unusually cool or warm years within the margin of natural interannual variability may be an alternative explanation for the observed differences between ten-year winter lake temperature means. xii

PAGE 13

CHAPTER 1 INTRODUCTION Contemporary climate change is acknowledged by most scientists as one of the most profound environmental crises within the human experience (Kennedy 2004; Oreskes 2004). Major strides in research and modeling during the past decade have enabled the Intergovernmental Panel on Climate Change (IPCC) to conclude in its most recent assessment report that a clear signal of anthropogenic warming can now be distinguished from natural variability and remaining scientific uncertainty (IPCC 2001a). The primary causal mechanism has been identified as enhanced greenhouse forcing caused by large-scale combustion of fossil fuels and deforestation (IPCC 2001a). Because these activities are deeply rooted in the lifestyle and industrial production of all modern societies and, therefore, not readily adjustable, climate change is clearly a long-term, global problem with far-reaching ecological and socioeconomic implications. While initial efforts to curb emissions are now underway, including the Kyoto Protocol to the United Nations Framework Convention on Climate Change (UNFCCC), models project that concentrations of most major atmospheric greenhouse gases will continue to increase throughout the 21 st Century even under the most restrictive scenarios considered by the IPCC (IPCC 2001a). Tropospheric concentrations of carbon dioxide (CO 2 ), the most significant well-mixed greenhouse gas in terms of radiative forcing, are now likely at their highest levels since the Miocene/Oligocene boundary 24 million years ago, a time characterized by lack of permanent polar ice caps and vastly different distributions of biota (Pearson and Palmer 2000; Zachos et al. 2001). The current (as of 1

PAGE 14

2 2004) CO 2 concentration of approximately 380 ppm (Kennedy 2004) represents a 35% increase over the pre-industrial level of 280 ppm. Greenhouse forcing attributed to rising concentrations of other well-mixed greenhouse gases, like methane (CH 4 ), nitrogen oxide (N 2 O), tropospheric ozone (O 3 ) and chlorofluorocarbons (CFCs), is equivalent to an additional 20% increase in CO 2 (Ruddiman 2001). According to the IPCC (2001a) metastudy, mean global temperatures increased by 0.6.2C during the 20 th Century, whereas precipitation increased by 5-10%. In the United States, temperatures rose by 0.8C, with precipitation change approximating the global trend (Parmesan and Galbraith 2004). By 2100, IPCC emission scenarios project CO 2 concentrations at 1.9 to 3.5 times their pre-industrial value, resulting in an additional mean global temperature increase of 1.4 to 5.8C (IPCC 2001a). Now that the question of whether or not anthropogenic climate change is indeed occurring has been answered, evaluating likely ecological and socioeconomic impacts, with increasing focus on the regional scale, has emerged as a central research priority. Analysis at small geographic scales is crucial for risk assessment and development of site-specific mitigation efforts that minimize short-term socioeconomic costs (cf. Rosenbaum 2002). Climate change is not a globally uniform process. It exhibits substantial spatial and temporal complexity, such that the above temperature projections will not be reached in some regions, while they will be surpassed in others. Generally, rate and magnitude of warming tend to increase with increasing latitude, more so over land than the oceans due to the latter’s thermal inertia (IPCC 2001a, 2001b; Parmesan and Galbraith 2004). Subpolar Alaska, for instance, has experienced warming of 2-4C during the 20 th

PAGE 15

3 Century, exceeding the magnitude of observed temperature change in the contiguous United States 2.5 times (Parmesan and Galbraith 2004). Temporally, there has been a reduction in diurnal temperature ranges, with nighttime temperatures increasing at a rate approximately twice that of daytime highs (IPCC 2001a). These general trends, however, are insufficient to establish climate forecasts at regional and local scales, particularly because general circulation models (GCMs), despite their increasing complexity, remain insufficient to characterize small and irregularly-shaped landmasses adequately. The Climateprediction.net distributed model, for instance, has a resolution of 2.5 latitude by 3.75 longitude (Climateprediction.net 2005). Since the Florida peninsula encompasses 6.2 latitude and, at its widest, 2.4 longitude, its model representation does not at all resemble its true geography. Small-scale circulation processes occurring primarily at sub-grid scales must be parameterized and cannot be directly evaluated (Climateprediction.net 2005). This shortcoming is particularly grave as land use and surface hydrology, both of which constitute important sub-grid forcing factors (IPCC 2001a), are spatially heterogeneous in Florida. Another spatially variable factor that is thought to have a significant effect on the extent of observed warming is the production of sulfate aerosols, which produce a net cooling effect by scattering or absorbing radiation (IPCC 2001a). Data assembled by the Pew Center on Climate Change indicate that much of the Mississippi-Ohio basin experienced cooling of up to 2C during the 20 th Century, likely due to this effect, whereas much of the rest of the country, particularly the northeast and west, warmed by 1-2C (Parmesan and Galbraith 2004). Most studies of air temperature records specific to the southeastern

PAGE 16

4 U.S. did not produce conclusive evidence for a recent warming signal (Mulholland et al. 1997). Regional climate models (RCMs) have a higher resolution with horizontal grid spacings of 50-100 km (Randall 2001), but, being in early stages of development, they are generally not yet capable of producing reliable forecasts (IPCC 2001a; MacCracken et al. 2004; Schiermeier 2004). Additionally, they depend on GCM results for determination of boundary conditions, and therefore may propagate false assumptions or errors made in GCM simulations (IPCC 2001a). A RCM used as baseline for a review of climate change and its impacts on aquatic ecosystems by Mullholland et al. (1997) projects warming of 3C over peninsular Florida assuming atmospheric CO 2 concentrations of twice their pre-industrial value. The model predicts precipitation increases of 25% during summer and 5% during winter. Ecological Implications of Climate Change Climate directly controls or affects many ecological and biological processes, including nutrient cycling, growing season, timing of reproduction, environmental sex determination, animal behavior and even body size in some taxa (Parmesan and Galbraith 2004). Climate change is expected to affect physiology, phenology and geographic distributions of species, as well as induce evolutionary adaptations, particularly in r-strategists with short generation times (Hughes 2000). Sustained temperature deviations from the historical norm force stenothermic species to adjust habitat ranges to maintain a suitable environment for survival. Because climatic tolerances, as well as speed and ability to migrate, vary among taxa, non-synchronous range shifts are expected to occur, altering coevolved species interactions and substantially changing ecosystem structure and function (Walther et al. 2002).

PAGE 17

5 Successful habitat adjustment is further constrained by resource availability and community dynamics within the new geographic range (Walther et al. 2002). Even cosmopolitan species are likely to experience indirect effects, such as changes in abundance of preferred food sources and altered competition-predation dynamics (Walther et al. 2002). Numerous studies have established that climate change-related species displacement and associated community-level effects are already occurring (IPCC 2001b; Parmesan and Galbraith 2004). Because projected temperature increases for the next century greatly exceed those experienced during the 20 th Century, it is fair to assume that environmental impacts will continue at an accelerated rate (Hughes 2000). Impacts Specific to Florida Most regional-scale research on climate change and its ecological impacts have focused on temperate and sub-polar environments that are subject to the greatest and most rapid temperature changes, whereas the tropics and subtropics are generally assumed to exhibit greater stability (IPCC 2001a). Florida, however, appears to be an exception in that it is hypothesized to be sensitive to climate change despite its low latitude. This can be attributed to its geography and intense human population pressure, which is likely to have depressed ecosystem resilience and reduced the capacity to absorb additional stress exerted by impacts of global warming. Florida represents an ecotone between warm temperate taxa of the southeastern United States and subtropical/tropical biota of the Caribbean basin. The former primarily inhabit the panhandle and northern third of the peninsula, whereas most of the latter are presently restricted the southern third of the state. The narrowness of the intermediate transitional zone reflects steep climatic gradients over peninsular Florida, along which the transition from warm temperate to tropical occurs over <700 km. According to Hughes

PAGE 18

6 (2000), a 3C increase in mean temperature, as anticipated by the Mulholland et al. (1997) RCM, would likely result in a northward shift of isothermal lines by 300-400 km, a distance exceeding the width of transitional life zones in the central Florida peninsula as described by Brown (1909) and Dohrenwend and Harris (1975) (Figure 1-1). Figure 1-1. Life zones of Florida. A) Forest Regions (Brown 1909). B) Southern range limit of red oak (Quercus rubra), marking the northern extent of a transitional zone within which most temperate tree species reach their southern limit (1). Post oak (Quercus stellata) line, marking the southern extent of the transitional zone (2). Subtropical vegetation line, marking the northern extent of Caribbean flora (3) (Dohrenwend and Harris 1975). C) Life Zones according to the Holdridge (1967) prediction model after Dohrenwend and Harris (1975). WT-mf: warm temperate moist forest, T-WT/ST-mf: transitional zone between warm temperate and subtropical moist forest, ST-mf: subtropical moist forest. Even relatively small climatic changes can be expected to result in narrowing and/or shifting of ecological life zone boundaries. Regional warming would facilitate northward habitat expansion of cold-limited invasive exotic species, contributing additional stress on temperate ecosystems and species near their southern distributional limit. Loss of desirable ecosystem functions would likely result (Walther et al. 2002).

PAGE 19

7 Because the peninsula is framed by the Gulf of Mexico and the Atlantic Ocean, migration routes for most species are limited to north-south pathways. Ecosystem fragmentation resulting from activities of an exponentially increasing human population constitutes an additional impediment to synchronous community range shifts (Thomas et al. 2001). The ecological impacts discussed above apply to terrestrial as well as lacustrine ecosystems. Additionally, Florida lakes are hypothesized to experience increased inputs of dissolved organic carbon as projected warmer, more humid conditions favor expansion of cypress swamps and wetlands along shorelines (Mulholland et al. 1997). Higher biological productivity mediated by higher temperatures and deposition of fertilizers via increased runoff is likely to continue the trend towards eutrophication (Mulholland et al. 1997). Because precipitation increases are projected to occur primarily during summer in the form of stronger and more frequent thunderstorms, hydrologic changes, including more frequent summer flooding events, are likely to occur. Global Climate Change during the Study Period The global mean temperature is estimated to have increased by 0.6.2C between 1900 and 2000. In its Third Assessment Report, the IPCC (2001a) provides a detailed 20 th Century temperature record based on four comprehensive studies yielding near-identical results, clearly illustrating that warming occurred in two distinct phases (ca. 1910-1945 and post-1976) interrupted by a ca. 30-year interval of slight cooling (Figure 1-2). In 1900, the mean global temperature over land was 0.25C below the 1961-1990 benchmark average, and declined by approximately 0.13C to 0.38C below the benchmark by 1910. Temperatures then increased by approximately 0.44C between

PAGE 20

8 1910 and 1945, at which time the benchmark was exceeded by 0.06C. This rapid increase was followed by a slow cooling trend that continued through 1976. By the end of this interval, temperatures had fallen to ca. 0.08C below the benchmark. Since then, the warming trend has resumed at an accelerated rate, reaching the benchmark by 1980 and exceeding it by 0.35C in 2000. The 1990s were likely the warmest decade of the second millennium AD, and 1998 the warmest year on record (IPCC 2001a). It is notable that nighttime temperatures over land have increased at approximately twice the rate of daytime temperatures during the most recent warming period (IPCC 2001a). Figure 1-2. Variation in mean global air temperature in C over land from 1900 to 2000 relative to the 1961 to 1990 average. Annual averages are indicated in blue with black error bars (2x SE). The smoothed curve is based on a 21-point binomial filter, producing approximately decadal averages. Based on IPCC (2001a).

PAGE 21

9 The earliest lake temperature data referenced for this project are from 1968, near the end of the mid-century cooling trend. Therefore, the study period covers the most recent warming period from its beginning, corresponding to a mean global temperature increase of 0.45C. While global trends are not necessarily indicative of regional patterns, this value will serve as a benchmark against which the results of this study can be compared to determine if the pattern of change in Florida approximates the global trajectory. Climate of Florida Familiarity with the present-day climate of Florida and the processes controlling it contributes to understanding lacustrine mixing and stratification behavior and is crucial to evaluating potential impacts of climatic change. Florida spans over 6.2 degrees latitude, or approximately 700 km from the Georgia border near Boulogne in Nassau County to Key West. Over this relatively short distance, there are considerable climatic gradients. The Kppen climate classification system (Figure 1-3A), which is based on monthly and annual temperature means as well as seasonality and quantity of precipitation, splits Florida into two segments. The panhandle and northern three-quarters of the peninsula are considered humid subtropical (Cfa), whereas the southern quarter is a tropical savanna, or tropical wet-and-dry region (Aw) (Henry 1998). The Thornthwaite system (Figure 1-3B), which incorporates the balance of precipitation and potential evapotranspiration, recognizes three distinct zones: A humid mesothermal panhandle and northern peninsula, a humid tropical southern tip, and a subhumid mesothermal regime in between (Henry 1998). The zonal boundaries slope from the southwest to the northeast due to the influence of the Gulf Stream (Henry 1998).

PAGE 22

10 A B Figure 1-3. Florida climate classifications. A) Kppen system, B) Thornthwaite system. Based on Henry et al. (1998). Both classification schemes are rather coarse indices lacking sufficient resolution to display gradients within each zone, which are substantial. They also belie the fact that the Florida climate is highly variable intraand interannually, particularly for winter lows and precipitation (Henry 1998). Air Temperature As described by Winsberg (1990), the major controls on air temperature within Florida are latitude and distance from the coast. The temperature-mediating effect of the Atlantic Ocean and, to a lesser degree, the Gulf of Mexico restrict diurnal and annual temperature ranges in coastal areas, with amplitudes increasing inland. Large lakes similarly moderate the temperature of their immediate surroundings. During summer, temperatures reach average highs near 33C and lows around 22C throughout the state with very little interregional and temporal variability: mean July high temperatures in Key West and Jacksonville differ by just 0.6C. Summer sea breezes reduce highs along the east coast by up to two degrees compared to the interior of the state, but this effect does not typically penetrate far inland (Winsberg 1990). During

PAGE 23

11 winter, latitudinal temperature gradients increase sharply, with a difference in mean January temperatures between the two cities of 5.6C. This interseasonal incongruity can be explained in part by the interplay of angle and duration of insolation. During summer, the lower angle of incidence over north Florida compared to the south is compensated for by slightly longer daylight, meaning that the total amount of solar energy received as a function of duration and intensity is roughly equal throughout the peninsula. In winter, however, these effects are additive instead of compensatory, such that the north experiences both shorter days and a more acute angle of incidence than the south (Winsberg 1990). In addition, the north is strongly affected by semi-random incursions of cold continental air masses during winter that commonly depress temperatures for several days before rebounding to milder conditions mediated by the maritime influence of the Gulf of Mexico and Atlantic Ocean. Impacts of cold fronts decrease with decreasing latitude (Henry 1998). The boundary between continental and peninsular (maritime) winter climate was defined by Winsberg (1990) as a line from St. Augustine in St. John’s County west to the mouth of the Suwannee River in Dixie County. Because frequency, timing and strength of cold fronts are unpredictable, there is considerable day-to-day and interannual variability in winter temperatures. Thus, in contrast to summer, daily winter temperatures are likely to deviate substantially from the long-term average. The transitional seasons of spring and autumn lie between summer and winter relative to temperature variability (Winsberg 1990). The beginning and duration of seasons from a meteorological standpoint varies by latitude and distance from the coast. Winter, defined by Winsberg (1990) as the first

PAGE 24

12 week after July during which average maximum air temperatures fall below 23.9C, commences by the first week of November in the northern panhandle, whereas it is delayed until after the first week of January in extreme southern Florida (Figure 1-4A). Mild to moderate overnight frost is a common occurrence in north Florida associated with cold fronts, but it rarely persists during daytime. Spring, which Winsberg (1990) defined as the first week of the year when average maximum temperatures exceed 23.9C, begins by the first week of February south of Lake Okeechobee, but not until after early April in coastal areas of the western panhandle (Figure 1-4B). Summer commences when mean maximum temperatures first exceed 31.1C (Figure 1-4C). This occurs in May for almost the entire peninsula, with the exception of coastal areas of the Atlantic and western panhandle, where it does not begin until early June. A wedge-shaped area in the southwestern peninsula between Sumter and Hardee counties exceeds 31.1C prior to the first week of May (Winsberg 1990). Fall begins when average minimum temperatures drop below 15.6C for the first time after July (Figure 1-4D). This occurs by mid-October in the panhandle and not until mid-December along the southeast coast. By this definition, there is no distinct autumn in the Keys, as regional minimum temperatures never fall below 15.6C. As with most climatic parameters, there is considerable interannual variability, and departures from the stated dates by several weeks are common for any given year. The mean annual temperature range between warmest and coldest month exceeds 30C in north-central Florida, whereas it is less than 14C in the Keys. This reflects the fact that the seasons are clearly distinguished by pronounced temperature differences in

PAGE 25

13 the north, but seasonality progressively decreases southward to where diurnal fluctuations frequently exceed interseasonal variability. A B C D Figure 1-4. Beginning of meteorological seasons (month/week) in Florida. A) Winter, the first week after July during which average maximum air temperatures fall below 23.9C, B) Spring when average maximum temperatures first exceed 23.9C, C) Summer, when average maximum temperatures first exceed 31.1C, D) Fall, the first week after July during which average minimum temperatures fall below 15.6C. Based on Winsberg (1990). Precipitation Patterns While evaluation of precipitation changes over time is not subject of this study, the following summary of geospatial precipitation patterns is included in this discussion of

PAGE 26

14 Florida climate for completeness. Most landmasses of similar latitude experience arid to desert-like conditions, whereas Florida, due to its peninsular shape and position between the Gulf of Mexico and Atlantic Ocean, supplying the state with maritime tropical air (Chen and Gerber 1990), is humid. Mean annual rainfall, based on the period 1961-1990, varies from 102 cm in Key West to 175 cm in the northwestern panhandle. The southeast coast is a second area of high precipitation, reaching 160 cm near Hialeah in Miami-Dade County. The statewide, long-term average is 137 cm (Henry 1998). While winter snowfalls occur sporadically, they become exceedingly rare with decreasing latitude and contribute only trace amounts in most years. Annual precipitation receipts throughout the state are dominated by locally generated convective summer thunderstorms (Henry 1998). Onset and termination of the convective season are controlled by the position and strength of the Bermuda-Azores high pressure system. During winter, it induces subsidence of dry air over Florida, which stabilizes the atmosphere and impedes convection. In summer, subsidence is reduced, generating conditions conducive to thunderstorm formation (Winsberg 1990). Thunderstorms are typically brief, intense and exhibit a distinct afternoon peak in activity that coincides with maximum sea breeze intensity and thermal heating of the land. Convergence of the Atlantic and Gulf sea breezes over the center of the state further destabilizes the atmosphere, generating storms that tend to be larger and more intense than those associated with convection alone. Winter precipitation is lower in intensity and about equally likely to occur at any time of the day (Henry 1998). Due to the relatively flat topography of Florida, with a maximum elevation of 105 m in Walton County and a statewide mean of 30 m, orographic effects are not profound, although topography has

PAGE 27

15 been identified by Chen and Gerber (1990) as contributing to a “mosaic of microclimates” (p. 12). With the exception of the Keys, which comprise a distinct precipitation regime, seasonality of rainfall increases with decreasing latitude, resulting in very distinct wet and dry seasons in the south. Annual totals are highly variable, resulting in frequent droughts of moderate to severe intensity, as well as exceedingly wet years with localized flooding (Henry 1998). Henry’s (1998) analysis of long-term data from some stations suggests a weak 5-7 year cycle roughly correlating to the interval between successive El Nio–Southern Oscillation events. Tropical cyclonic activity is suppressed during El Nio conditions, an effect that is, however, more than compensated for by greatly increased winter rainfall of up to 300% above long-term means. A La Nia, on the contrary, is associated with particularly severe drought conditions. Atypical jet stream positions and formations of high pressure ridges originating from the Bermuda-Azores and high pressure systems over the central United States also contribute to rainfall variability. These anomalies rarely affect the entire state equally, and drought and flooding can occur simultaneously in different areas (Obeysekera et al. 1999). According to Winsberg (1990), the southeast coast from Key West to Vero Beach, as well as the Big Bend area between Tampa Bay and Pensacola, are subject to the greatest interannual variability. Precipitation Regimes Three distinct rainfall regimes are outlined by Henry (1998), differentiated primarily by the nature and timing of secondary (i.e. non-convective) moisture sources.

PAGE 28

16 Panhandle and northern peninsula. The northern zone, which encompasses the panhandle and peninsula north of Tampa Bay, is heavily influenced by frontal systems during winter. Squall lines associated with continental cold fronts tend to progress on an easterly to southeasterly trajectory before stalling and eventually dissipating, resulting in a progressive reduction in the amplitude of the winter precipitation peak with decreasing latitude and, to a lesser degree, decreasing longitude. Low-pressure systems moving northeast from the Gulf of Mexico contribute additional moisture to the area. The region between Panama City and Pensacola is second only to the southeast coast in being affected by tropical storms and hurricanes, which contribute about one-third of fall precipitation in this area on average. Seasonal differences in precipitation are low, but increase with decreasing latitude. Near the southern margin of the northern rainfall regime, dry and wet seasons are clearly distinguishable. Summer receipts range from less than 50% of annual totals in the Pensacola area to >60% near Tampa Bay. Rainfall minima occur in April and October, marking the transition between convection and frontal-dominated conditions. Maxima are reached in July for the panhandle, August in the central peninsula and Gulf Coast, and September along the Atlantic Coast (Henry 1998). The most frequent excursions from the mean are observed during autumn due to the infrequent incidence of tropical cyclones (Winsberg 1990). Southern peninsula. The southern half of the peninsula, which comprises a second precipitation regime, experiences a distinct five-month wet season lasting from late spring to early fall that is dominated by convective storms enhanced near the southeast coast by strong sea breezes carrying warm, moist air from the Gulf Stream (Henry 1998;

PAGE 29

17 Obeysekera et al. 1999). Mid-summer precipitation is slightly depressed along the eastern shore due to development of a high-pressure ridge associated with the Bermuda-Azores system, increasing subsidence and thereby temporarily stabilizing the atmosphere in the region (Henry 1998). Seasonality is pronounced, with dry season rainfall between November and April accounting for only 25% of annual receipts, which can be attributed to cold fronts that infrequently progress sufficiently far southward to affect the area (Obeysekera et al. 1999). Important additional sources of precipitation are tropical cyclones that predominantly affect the southern and southeastern coast between Florida Bay and Melbourne (Brevard County) during late summer and early fall (Winsberg 1990). Long-term averages suggest that tropical storms reach the area once annually, whereas hurricanes make landfall once to twice per decade (Obeysekera et al. 1999), although there is significant variability: Consecutive hurricanes Betsy (1965) and Andrew (1992) were separated by 27 years, whereas Charley, Frances and Jeanne occurred within a period of six weeks in 2004. Rainfall maxima vary regionally and typically occur in June, August or September. Dry season precipitation tends to be fairly equitably distributed and lacks a distinct secondary peak, although a slight minimum is recorded in April at some stations (Henry 1998). Deviations from the mean are most significant during winter (Winsberg 1990). The Keys. The Keys form a third, very distinct rainfall regime. The relatively dry conditions in this area can be explained by the small size of the islands, which is insufficient to produce significant convection. Thus, the summer precipitation peak is greatly depressed, and mean monthly totals are more equitably distributed, such that there

PAGE 30

18 is no clear distinction between wet and dry seasons (Winsberg 1990). The Keys will not be explored in detail due to an almost complete absence of lakes. In summary, the climate of Florida is characterized as highly variable both spatially and temporally. In the north, seasons are defined by clear temperature differences and equitable distribution of precipitation, whereas the south experiences distinct wet and dry seasons and a much smaller annual temperature range. Objectives and Scientific Significance This study evaluates detectability of recent climatic change in peninsular Florida via analysis of lake monitoring data collected by government agencies and universities between 1968 and 2004. It establishes a temporally and spatially differentiated record of long-term water temperature variability along the transition from warm temperate to subtropical climate regimes and ecological life zones. Robust analyses were performed to gauge the strength and trajectory of regional climatic variability in terms of temperature trends and geospatial thermal regime boundary shifts. The study null hypothesis is defined as lack of statistically significant temperature change during the study period, whereas the alternative hypothesis is defined as detectability of long-term temperature trends or cycles. Correlations between water temperature means and lake location, morphometry, chlorophyll a and true color were evaluated to identify how these variables influence lake temperatures annually and seasonally. Comparisons with known latitudinal air temperature gradients were used to evaluate whether water temperature records constitute suitable proxies for regional air temperatures. Utilization of lake temperature records provides an added benefit in that the very high specific heat of water (4186 J/kg/C) acts

PAGE 31

19 as a natural filter that sharply reduces short-term temperature variability in the water column. This study provides valuable baseline data for calibration and improvement of regional climate models to enhance the accuracy of future prognoses for the circum-Caribbean basin. It further facilitates assessment of potential long-term impacts on Florida lakes, including displacement of temperate species, rates of northward migration of exotic, invasive taxa and changes in timing and stability of lake stratification.

PAGE 32

CHAPTER 2 DATA AVAILABILITY AND SITE SELECTION Characteristics of Florida Lakes Florida has one of the largest and most diverse lake districts in the United States, with approximately 7800 lakes exceeding 0.4 ha in surface area (Brenner et al. 1990). While lakes exist throughout the state, most are concentrated along a group of late Tertiary-Quaternary sand hill ridges that bisect the northern two-thirds of the peninsula longitudinally (White 1970) (Figures 2-1, 2-2). Approximately 35% of lakes lie in the central Florida counties of Lake, Polk, Orange and Osceola (Mossa 1998), whereas they are most scarce in the Big Bend region between Bay and Levy counties, and south of Lake Okeechobee (Shafer et al. 1986). Most Florida lakes were formed or modified by karstic processes, namely chemical dissolution of Eocene Ocala limestone and Oligocene Suwannee limestone by rain and groundwater, and resulting subsidence or collapse of surface material (Shannon and Brezonik 1972; Brenner et al. 1990; Kindinger et al. 1999). The vast majority are geologically young (6000-8000 years; Brenner et al. 1990). Prior to that time, a combination of arid climate and depressed aquifers associated with lower sea levels during and following the Wisconsinan glaciation precluded shallow basins from retaining water permanently (Brenner et al. 1990). The oldest known lake is Tulane in Highlands County (>50,000 years; Grimm et al. 1993). Many have near-circular outlines typical of solution basins, and, as bathymetry reveals, the outline of irregularly-shaped lakes can often be attributed to clusters of multiple adjacent karstic depressions. 20

PAGE 33

21 Figure 2-1. Geologic map of northern peninsular Florida. Most lakes in peninsular Florida are arranged along the central Tertiary-Quaternary sand hill ridges (cf. Figure 2-2). Modified from Scott et al. (2001).

PAGE 34

22 Figure 2-2. Landsat satellite view of the lake district of peninsular Florida produced with NASA World Wind 1.3 software. Their statistical size distribution is strongly skewed (Figure 2-3), with most smaller than 20 ha (Shafer et al. 1986; Mossa 1998) and only five exceeding 100 km 2 , the largest of which is Lake Okeechobee (1770 km 2 , Brenner et al. 1990). Maximum depths of solution systems rarely exceed 10 m, whereas some collapse systems approach 30 m (Shannon and Brezonik 1972; Brenner et al. 1990; Mossa 1998).

PAGE 35

23 Distribution of Named Florida Lakes by Size0500100015002000250030003500<0.80.81 45 1011 2021 3031 4041 8081 120121 160161 200201 400401 10001001 20002001 4000>4000Surface Area (ha)Number of Lakes Figure 2-3. Distribution of named Florida lakes by surface area. Based on Shafer et al. (1986). Most are hydrologically supported by local perched water tables underlain by near-surface clay strata embedded in Hawthorn group material. These aquacludes minimize hydrologic exchange with the deep, hardwater Floridan aquifer (Kindinger et al. 1999). Surficial aquifer water tables respond sensitively to variable precipitation, resulting in appreciable changes in lake surface area seasonally and interannually, particularly in shallow systems (Shannon and Brezonik 1972). Extreme fluctuations have been observed in astatic lakes with highly porous sediments or location atop periodically active sinkholes that breach confining layers and allow lake water to drain into intermediate or deep aquifers. Such astatic lakes include Jackson (Leon County), Brooklyn and Pebble (both Clay County) (Deevey 1988). Geographic location is not inherently indicative of

PAGE 36

24 hydrologic stability, as highly astatic lakes can be found in close vicinity to others exhibiting little stage variability (Brenner et al. 1990). Interestingly, precipitation is not always the dominant factor in determining lake stage. As described by Deevey (1988), deep aquifer conditions appear to project an indirect buoying effect on lake level through the overburden, producing analogous stage fluctuations over large geographic areas that are often not tightly correlated with local precipitation records. It should also be noted that water levels of some lakes are actively managed by drainage canals or via augmentation with water pumped from the Floridan aquifer, altering hydrological characteristics (Stewart and Hughes 1974). Approximately 70% of lakes lack identifiable surface outflows and are characterized as seepage lakes (Palmer 1984). Water residence times are unusually long, with flushing rates often <20% of temperate systems with similar morphology. Deevey (1988) arrived at a mean residence time of 2.671.33 ( SE) years using a dataset of 20 lakes including both seepage and drainage systems. These very slow flushing rates imply that direct heat transfer between lakes via connecting streams is unlikely to affect thermal properties significantly. Variables hypothesized potentially to affect lake thermal characteristics include location (latitude, longitude and/or distance from the coast), morphometry (depth and surface area), as well as factors affecting light transmission (chlorophyll a and color). Florida lakes range from ultraoligotrophic systems embedded in near-sterile quartz sands of the central ridges, to hypereutrophic, associated primarily with phosphate deposits in the Hawthorn Group that outcrop along the western margin of the ridges in the northern and central peninsula (Brenner et al. 1990, Scott et al. 2001). Many Florida lakes have

PAGE 37

25 undergone cultural eutrophication during the 20 th Century as a result of exponentially increasing human encroachment. The best-known example is Lake Apopka, originally a macrophyte-dominated, clearwater system that, in 1947, shifted state to phytoplankton domination following decades of sewage discharge and muck farming along its northern perimeter (Brenner et al. 1990). Concentration of dissolved organic color varies greatly among lakes. According to Florida Lakewatch (2004a) data based on 3223 samples from 670 lakes, 79% of lakes are not or only slightly colored (<50 PCU), while others, particularly those surrounded by coniferous forests and cypress swamps, tend to be more heavily stained (Figure 2-4). Color within a single lake may fluctuate significantly, with color peaking following major precipitation events that wash organic acids produced by ecotonal vegetation into the system (Florida Lakewatch 2004a). Figure 2-4. Distribution of true color in Florida lakes. Modified from Florida Lakewatch (2004a).

PAGE 38

26 Beaver et al. (1981) defined three thermal regime zones for Florida lakes based on hierarchical cluster analyses on monthly to bimonthly temperature data obtained during a one-year study of 29 lakes in 1979: warm temperate (north of 29.5), transitional and subtropical (south of 28) (Figure 2-5). Figure 2-5. Thermal regimes of Florida lakes according to Beaver et al. (1981). These partitions closely parallel climate and ecological life zones described in Chapter 1. Analogous to air temperature patterns, water temperature gradients between regime zones broke down in late summer and were most pronounced during winter. Water temperatures of north Florida (warm temperate) lakes were significantly different from subtropical ones for all months except August (Beaver et al. 1981). Statistically significant temperature gradients between transitional and northern lakes were recorded

PAGE 39

27 during eight months, whereas transitional systems differed from southern ones during only two months. Thermal properties of transitional zone lakes were therefore more closely aligned with subtropical lakes than warm temperate systems. Because of the scarcity of natural lakes in the tropical climate zone, the subtropical-tropical regime transition was not defined by Beaver et al., and will not be evaluated in this study for the same reason. Morphometry did not play a significant role in determining regime boundaries based on statistical comparisons between mean monthly water column temperatures of deep and shallow lakes within each zone (Beaver et al. 1981). With few exceptions, all Florida lakes are thought to be polymictic or warm monomictic (Florida Lakewatch 2004b). Lakes as shallow as six meters commonly stratify between March and September due to the high density differences per degree of temperature change within the commonly observed range of summer temperatures (Beaver et al. 1981). Shallower systems frequently develop incipient stratification that is interrupted by strong winds (Brenner et al. 1990). The duration of the winter mixing period in the northern zone is four-five months, two-four months in the transitional zone and one-two months in the subtropical zone (Crisman, T. L., C. A. Chapman & L. J. Chapman, unpubl.). It is hypothesized that sustained warming will result in earlier onset of stable stratification and later breakdown thereof, diminishing the duration of winter mixing. Profundal oxygenation and nutrient cycling would thereby be altered. Data Availability While the temperate lake districts of the Midwest and northeastern United States have been systematically studied since the late 19 th Century by such renowned limnologists as Birge, Juday, Welch, Hutchinson and Deevey, little research was conducted on Florida lakes until the 1960s. Even today, only a small fraction are

PAGE 40

28 routinely and regularly monitored, as a quick survey of databases like STORET readily reveals. By 2003, the Florida Lakewatch program, a statewide, volunteer-based lake monitoring effort, had sampled 1213 lakes, representing approximately 15% of the total (Florida Lakewatch 2005). Water management districts, established in 1972, and other state agencies also maintain monitoring programs. Because this study relies on historical data, it is useful to review the history of limnological exploration in Florida briefly to provide perspective on how data limitations and availability constrained site selection. Yount (1963) emphasized in his summary of early studies on freshwater systems of the South Atlantic states that most had focused on hydrobiological rather than limnological analyses. This was also the case in Florida, where researchers initially prioritized study of highly specific aspects of individual systems deemed unusual or economically important as swimming or fishing lakes (Shannon and Brezonik 1972). The first major project to characterize the lake district of peninsular Florida holistically was launched in 1968 (Shannon and Brezonik 1972). A major focus of this and other studies conducted during the 1970s was development and application of trophic state indices (cf. Brenner et al. 1990), catalyzed by the realization that Florida lakes were increasingly affected by cultural eutrophication prompted by an exponentially increasing population–surging from 530,000 in 1900 to 6,789,443 in 1970 (U.S. Census Bureau 1995). The first compilation of limnological data from disparate sources into a central repository was the Florida Lakes Database (FLADAB), which was maintained from 1967 through 1981 and included data on 788 systems. Unfortunately, a lack of backups and improper handling of the magnetic tapes on which the database was stored resulted in the de-facto loss of this valuable resource, of which reportedly only a single paper copy

PAGE 41

29 survives (M. Brenner, pers. comm.). In addition to the work of Shannon and Brezonik, data collected by the National Eutrophication Survey, Florida Game and Fresh Water Fish Commission, U.S. Geological Survey, Florida Department of Environmental Regulation, Florida Water Management Districts and University of Florida researchers contributed to FLADAB. Huber et al. (1982) used FLADAB to develop their classification of Florida lakes for the U.S. EPA Clean Lakes Program. While other databases, most notably the U.S. EPA STORET system, contain information on Florida lakes from as early as the late 1960s, availability of regular monitoring data is largely restricted to the mid-1980s or later, even for the most economically and recreationally important systems. Site Selection and Data Collection By compiling data generated by multiple research and monitoring programs, records spanning several decades for numerous lakes in the peninsular district could be established, making it is feasible to assess trajectories of limnological variables on interannual to decadal scales. An initial list of sites considered likely to have been sampled repeatedly during the study period was generated by referencing the Canfield (1981) report on chemical and trophic state characteristics of Florida lakes, for which 165 lakes throughout the state were sampled between September 1979 and August 1980. Additionally, systems from the list of 100 largest lakes in the Gazetteer of Florida Lakes (Shafer et al. 1986) were included. Impacts of potential size biases were evaluated as part of the analysis and will be discussed in chapters 4 and 5. Temperature, chlorophyll a, dissolved color and dissolved oxygen data were requested from all five Florida Water Management Districts, the Florida Fish and Wildlife Commission, the Archbold Biological Station, the U.S. Geological Survey and the U.S. Forest Service. The

PAGE 42

30 Northwest Florida Water Management District and U.S. Geological Survey did not provide data, stating that they do not maintain lake monitoring programs. The South Florida Water Management District and U.S. Forest Service indicated that data for lakes in their jurisdiction were available via their web site or the U.S. EPA’s STORET database. The other agencies complied, and the provided data were integrated into the project database via stacked Microsoft Excel 2003 worksheets, with sample depths noted where given. Subsequently, Legacy and Modern EPA STORET databases were queried, and data from research projects and theses from the files of Dr. T. L. Crisman at the University of Florida were included. Surface areas were obtained primarily from the Gazetteer of Florida Lakes (Shafer et al. 1986) and the Florida Lakewatch web site (2005). In cases where no size record was available from these primary sources and other reviewed literature, surface areas were estimated via the DeLorme Topo USA 4.0 software’s information function. Geographic coordinates of lake centers were obtained via the same function. Mean and maximum depths were taken from literature sources, including but not limited to Deevey (1988) and Shannon and Brezonik (1972). Because different sources utilized different units of measurement, all records were converted into standard SI notation. Suspected data entry errors were corrected or discarded by outlier analysis in addition to visual examination of the sorted datasheets and distribution graphs. Lakes known to be reservoirs or highly astatic were discarded due to potential impacts of rapid, non-climate induced lake level changes on water temperatures. Lake Okeechobee was also removed from consideration because its size (1770 km 2 ) and shallowness ( z = 2.3m) favor development of pronounced horizontal temperature gradients.

PAGE 43

31 In general, the state of monitoring and reporting for water temperature data of Florida lakes was poorer than anticipated. Most of the 7800 lakes in Florida remain limnologically unexplored, and few are monitored at regular intervals. Particularly during the early years of the study period, temperature data were usually collected as part of general lake surveys, trophic state assessments or acid precipitation monitoring efforts instead of routine ambient monitoring, resulting in somewhat fragmented records with extensive gaps in temporal coverage. Data availability was particularly poor for lakes in the panhandle, which had to be excluded from the study area. Temperature profiles by depth were not available in sufficient numbers to incorporate into the analysis, which is, therefore, limited to surface water measurements (defined here as z 1m). Gaps in lake records were in many cases successfully compensated for by amalgamation of data from multiple sources as well as statistical standardization methods. Fifty lakes, ranging in surface area from 34.4 ha (Lake Annie, Highlands County) to 18,400 ha (Lake George, Volusia/Putnam County), were selected for analysis based on completeness of their temperature record ( =2877 ha) (Figures 2-6, 2-7). Latitudinal coverage extended from 30'N (Alligator Lake, Columbia County) to 26'N (Lake Trafford, Collier County), constituting a transect across multiple life and climate zones (Figure 2-8). Data subsets were chosen based on sample size and data quality requirements of the various statistical methods utilized, while attempting to minimize reduction in geographic coverage.

PAGE 44

32 Figure 2-6. Location of study lakes (dark blue markers). Map generated with Microsoft MapPoint 2004 software.

PAGE 45

33 Distribution of Study Lakes by Size03691215<0.80.81 45 1011 2021 3031 4041 8081 120121 160 161 200201 400401 10001001 20002001 4000>4000Surface Area (ha)Number of Lakes Figure 2-7. Distribution of study lakes by size.

PAGE 46

34 Figure 2-8. Peninsular lake district in relation to Thornthwaite climate zone boundaries (maroon) and Dohrenwend and Harris (1975) life zone boundaries (green). The study area is approximated by the red shaded area.

PAGE 47

CHAPTER 3 THERMAL CHARACTERISTICS OF STUDY LAKES Seasonal Patterns Thermal characteristics of study lakes were initially determined under the null hypothesis assumption that no significant warming occurred during the study period. Analysis was preceded by extensive data standardization to allow for meaningful statistical comparison of records with disparate data volumes. Specifically, standardization was aimed at ensuring equal weighting of sample dates, regardless of the number of individual samples obtained during a single day. This was achieved with a multi-step procedure that first produced average temperature values for dates on which multiple samples had been recorded. The results were again averaged to obtain monthly mean temperature values. Based on a random sample of ten lakes from the database, pairwise t-tests indicated no statistically significant differences between January and February as well as between July and August surface water temperatures, respectively (p>0.1). Based on this, data for these time periods were pooled for analyses contrasting periods of minimal and maximal heat content. Minimum surface water temperatures occurred most frequently in January (41 out of 47 lakes considered). Four lakes reached minimum temperatures in February and two in December. The coldest lake was Newnans in Alachua County, with an average January temperature of 12.7C, whereas the warmest, Lake Trafford (Collier County), averaged 20.1C. The mean winter temperature for the entire database was 15.8.53C (SD). Maximum temperatures were reached during August (25 lakes) or July (22 lakes). The 35

PAGE 48

36 warmest system during summer was Alligator Lake in Columbia County (31.2C), whereas the coolest was Lake Minneola in Lake County (29.2C). The mean high temperature for all lakes considered was 30.2.52C. The sharp contrast between pronounced latitudinal temperature gradients during winter (7.4C between lakes Newnans and Trafford) and their complete breakdown during the summer months closely mirrors seasonal air temperature patterns as discussed in Chapter 1. The correlation between air and water temperature is also evident when comparing lake records to National Climatic Data Center (NCDC) 1971-2000 air temperature normals of nearby weather stations on a monthly basis (Figure 3-1). Both temperature curves display nearly identical patterns, although surface waters were consistently warmer by approximately 2C. Regression slopes of correlations between latitude and mean annual water temperatures and NCDC normals, respectively, are nearly identical (Figure 3-2), indicating that the air-water temperature difference is geospatially consistent and does not vary substantially by location. Analysis of a record from Lake Conway (Orange County), which included detailed diurnal temperature profiles collected once monthly during the period from December 1977 to November 1978, indicated an average diurnal surface water temperature variability about the mean of approximately .6C (Coenen, D., unpubl.). If this value is representative of diurnal temperature flux of Florida lakes in general, sampling bias, i.e. the fact that water temperature measurements are almost exclusively taken during the warmer daylight hours, can explain only approximately 30% of the observed 2C offset between air and surface water temperatures.

PAGE 49

Figure 3-1. Surface water temperatures of a north (A), central (B) and south Florida lake (C) compared to ambient air temperatures. Water temperatures were obtained from the study database, air temperatures are National Climatic Data Center normals for 1971-2000 (Southeast Regional Climate Center [SERCC] 2005), whereby the station closest to the lake was queried. Crescent City is located 17 km north-northwest of the center of Lake George, Sanford 9 km northwest of the center of Lake Jessup, and Archbold Station 2.5 km south-southeast of Lake Annie.

PAGE 50

38 Lake George, Volusia/Putnam County05101520253035JanFebMarAprMayJunJulAugSepOctNovDecMonthMean Temperature (C) SurfaceWater Air (CrescentCity) A Lake Jessup, Seminole County05101520253035JanFebMarAprMayJunJulAugSepOctNovDecMonthMean Temperature (C) SurfaceWater Air(Sanford) B Lake Annie, Highlands County05101520253035JanFebMarAprMayJunJulAugSepOctNovDecMonthMean Temperature (C) SurfaceWater Air(Archbold) C

PAGE 51

39 What factors account for the balance of the difference is not certain. Possibilities include biases due to dissimilar sampling methodologies or effects associated with the substantial heat absorption and retention capacity of water bodies, owing to their high specific heat. Mean Annual Air and Water Temperature by LatitudeT = 44.69 0.73 (Lat)r2 = 0.74T = 44.01 0.78 (Lat) r2 = 0.67192021222324252627262728293031Latitude (N)Temperature (C) Surface Wate r Air Figure 3-2. Comparison of mean annual air and water temperatures by latitude. Referenced air temperature means are National Climatic Data Center normals for 1971-2000 (SERCC 2005). Water temperature means incorporate the entire period of record available for each individual lake. Latitude-Temperature Relationships Continuing under the null hypothesis assumption, simple linear, polynomial and multiple regression models were fitted to annual surface water temperature means to derive a list of variables controlling thermal characteristics of Florida lakes. Mean annual temperatures were determined by averaging monthly temperature means of 47 lakes over the entire period of record. Simple linear regression analysis showed that there was a strong, statistically significant negative relationship between latitude and annual mean surface water temperature (Figure 3-2):

PAGE 52

40 )(73.069.44LatT (r 2 =0.74, p<0.01) (T = mean annual surface water temperature in C, Lat = latitude in degrees north). Homoscedasticity and normality of residual assumptions were verified graphically and found not to be problematic. Fitting a second-order polynomial model did not improve the relationship substantially (r 2 =0.74, p 2 >0.4), and was therefore not explored further. The significance of this result is twofold. It establishes that latitudinal position alone explains 74% of variability of annual surface water temperatures over the study period. This must be considered a conservative estimate because data gaps and different lengths of record for different systems were not controlled, likely contributing to the residual error term. This would be the case particularly if the alternate hypothesis (existence of a warming trend) was found to be true. The steepness of the slope term of 0.73C temperature change per degree of latitude change (110.85 km) and the strength of latitudinal control on temperatures are a direct reflection of the climatic gradient along the peninsula. Secondly, because the relationship was linear, it can be concluded that latitudinal control over mean annual lake water temperatures was approximately constant over the study area. In other words, the expected mean annual temperature difference between two lakes separated by x units of latitude is the same whether they are situated in the north or the south of the peninsula. A similarly strong, positive relationship was identified between latitude and intraannual temperature range, defined here as the difference in mean surface water temperature between the warmest and coldest month of the year. Using the same 47 systems, the relationship was defined as: )(44.169.26LatT , (r 2 =0.62, p<0.01) (T = mean intraannual surface water temperature range in C, Lat = latitude in degrees north).

PAGE 53

41 However, two systems, Newnans (Alachua County) and Tsala Apopka (Citrus County), fit the model poorly and showed significantly higher and significantly lower than expected ranges, respectively. The reasons for this deviation are not immediately obvious, although it could be argued that Tsala Apopka represents a connected system of semi-independent pools with fluvial characteristics that is unique among Florida lakes and therefore likely has different thermal properties. Newnans Lake is moderately astatic in terms of lake stage (Crossman 1956), although it is not clear if the higher than expected temperature range can be attributed to this property. After removing these two systems from the regression, thereby also establishing normality of residuals and homoscedasticity, the relationship improved to (Figure 3-3): )(38.129.25LatT ,(r 2 =0.73, p<0.01) (T = mean intraannual water surface temperature range in C, Lat = latitude in degrees north). Mean Annual Surface Water Temperature Range by LatitudeT = 25.29 +1.38 (Lat)r2 = 0.731011121314151617181920262728293031Latitude (N)Annual Temperature Range (C) Figure 3-3. Mean annual surface water temperature range between warmest and coldest month by latitude. Removed outliers are indicated by red symbols.

PAGE 54

42 Based on this model, the difference in mean surface water temperature between the warmest and coldest month changed by approximately 1.38C per degree change in latitude, establishing that differences between periods of maximal and minimal heat storage in lakes diminish with decreasing latitude, thus, mirroring air temperature patterns. Relationships between Temperature and Other Variables Influence of longitude, surface area, mean chlorophyll a and mean true color on mean annual surface water temperatures were evaluated in addition to latitude in a multiple regression model via the best subsets and stepwise regression functions of the Minitab 14.1 statistical software. Data from 35 lakes were included. Among the newly introduced variables, only longitude improved the model in a manner approaching statistical significance (p Long =0.114), yielding the relationship: )(22.0)(84.081.29LongLatT ,(R 2 =0.80, p<0.01) (T = mean annual surface water temperature in C, Lat = latitude in degrees, Long = longitude in degrees west). It is difficult to ascertain whether the longitudinal model component was, in fact, a manifestation of lake distance from the coast, because the ridge system along which the majority of study lakes were located is nearly equidistant to the Atlantic and Gulf of Mexico. Substitution of distance from lake center to the nearest point on the coast for longitude resulted in a weaker regression (R 2 =0.78, p dist >0.6). However, wind direction, and other variables affecting atmospheric circulation and heat transport over Florida were not considered. It is plausible that measuring distance to the coast along the axis of predominant wind direction might yield improved results. Referenced literature and databases did not provide mean and maximum lake depths for most study lakes. A model based on seventeen lakes for which maximum

PAGE 55

43 depth data were available indicated that it may be a significant model variable that warrants further exploration: )(023.0)(75.032.45maxzLatT ,(R 2 =0.87, p<0.01, p Zmax <0.1) (T = mean annual surface water temperature in C, Lat = latitude in degrees north, z max = maximum lake depth). Expanding the model by substituting maximum sample depths from the study database as rough estimates for missing maximum depth values resulted in loss of statistical significance of the depth variable at the p=0.1 level: )(022.0)(77.002.46)max(ezLatT ,(R 2 =0.90, p<0.01, p Zmax =0.109) (T = mean annual surface water temperature in C, Lat = latitude in degrees north, z max(e) = estimated maximum lake depth). It must be noted, however, that sampling stations rarely coincide with the deepest point of a lake, meaning that in most cases sample depth < z max . Further, mean depth is likely to exert greater influence on thermal properties of the surface layer than maximum depth, particularly in lakes where maximum depth is reached in small, deep depressions with little associated volume. As additional bathymetric surveys of Florida lakes are completed, it will be possible to refine the above model to incorporate morphometric controls on thermal properties more accurately. In summary, latitude was the dominant factor controlling mean annual surface water temperatures of Florida lakes. Longitude and lake depth appear to have exerted additional, minor influence on temperature properties and are significant at the p=0.15 level. None of the other variables evaluated entered the regression models with significance. Further, latitudinal gradients of air and water temperature are nearly identical, indicating that lacustrine temperature records constitute a suitable proxy from which information about the climate of surrounding areas may be derived.

PAGE 56

CHAPTER 4 ANALYSES FOR TEMPORAL CHANGE Regression Analyses for Interdecadal Temperature Variability Following evaluation of the database under the null hypothesis for determination of fundamental thermal characteristics of Florida lakes over the entire study period, as described in Chapter 3, a series of analyses to detect long-term temperature variability and/or trends in support of the alternative hypothesis were initiated. Data were prepared and standardized as outlined in the previous chapter. Regression analyses were used to contrast periods of highest (July-August, henceforth referred to as “summer”) and lowest (January-February, “winter”) surface water temperatures in ten-year intervals (1975-1984, 1985-1994, 1995-2004). These somewhat unorthodox partitions were chosen to incorporate the most recent observations into the analysis, while maintaining intervals of constant duration, which would not have been possible had conventional protocol (i.e. 1971-1980, etc.) been followed. To minimize differences in record length among lakes and to ensure that calculated ten-year temperature means were representative, i.e. based on a sufficient number of samples to preclude atypical observations from exerting undue influence on the analysis, the sample size was reduced to the 22 lakes with the most complete records. This resulted in a slight reduction of latitudinal coverage to between 27’N and 29’N (Figure 4-1). 44

PAGE 57

45 Figure 4-1. Location of lakes included in the analyses for temporal change (dark blue markers). Map generated with Microsoft MapPoint 2004 software. Summer. There was no significant latitude-temperature regression (r 2 0.02, p0.6) during summer for all ten-year study intervals (“trials”), indicating that latitudinal control of water temperatures broke down consistently as lakes reached maximum heat storage, mirroring air temperature patterns discussed in Chapter 1. The finding by Beaver et al. (1981) of no significant temperature differences among warm temperate, transitional and subtropical lake zones during August 1979, therefore, appears to be representative of

PAGE 58

46 summer conditions in general. No other variables (longitude, surface area, estimated maximum depth (cf. Chapter 3), mean true color and mean chlorophyll a) correlated significantly with summer temperatures at the p=0.1 level. No interdecadal trends were detectable. Data distribution did not shift appreciably between trials (Figure 4-2), with a majority of observations consistently falling in the range between 29C and 31C (cf. Chapter 3: mean summer temperature range over the entire period of record: 29.2C-31.2C). Ten-year averages were 29.8.6C, 30.0.8C and 29.9.5C (SD), from earliest to most recent interval, respectively, which were statistically non-distinguishable based on one-way analysis of variance (ANOVA) (p>0.4). Winter. During winter, Florida lake temperatures consistently followed strong latitudinal gradients in agreement with air temperature patterns (cf. Chapter 1). Unlike summer, there were distinct changes between ten-year intervals for temperature means, variances and shape of regression fits. The first period, 1975-1984, was the coolest with mean temperatures ranging from 12.6C (Newnans, Alachua County) to 18.1C (Hatchineha, Polk/Osceola County). Averaged over all lakes, the mean temperature was 15.6.6C (SD). The corresponding regression model (Figure 4-3), )(95.122.71LatT ,(r 2 =0.73, p<0.01) (T = mean winter surface water temperature in C, Lat = latitude in degrees north) described a highly significant, negative linear relationship between temperature and latitude. No additional variables entered the model with statistical significance at the p=0.1 level.

PAGE 59

Figure 4-2. Mean ten-year summer surface water temperatures by latitude. A) 1975-1984 B) 1985-1994, and C) 1995-2004.

PAGE 60

48 Mean Summer Water Temperature by Latitude 1975-1984282930313227282930Latitude (N)Temperature (C) A Mean Summer Water Temperature by Latitude 1985-1994282930313227282930Latitude (N)Temperature (C) B Mean Summer Water Temperature by Latitude1995-2004282930313227282930Latitude (N)Temperature (C) C

PAGE 61

49 Figure 4-3. Mean winter surface water temperatures by latitude during 1975-1984 and linear regression fit. Winter water temperatures during 1985-1994 exceeded previous interval means for each individual lake. Ranging from 14.9C (Crescent, Flagler/Putnam County) to 19.4C (Cypress, Osceola County), the overall mean temperature was 17.8.2C, an increase of 2.1.0C. The latitude-temperature relationship departed from linearity and was best described by a convex second-order polynomial model that approached asymptotic characteristics near the southern margin of the study area, indicating a gradual decrease of latitudinal influence on temperatures toward the south (Figure 4-4): 2)(67.0)(54.369.478LatLatT ,(r 2 =0.71, p<0.01, p 1 <0.1, p 2 <0.1) (T = mean winter surface water temperature in C, Lat = latitude in degrees north). No other variables entered the model at the p=0.1 level.

PAGE 62

50 Figure 4-4. Mean winter surface water temperatures by latitude during 1985-1994 and second-order polynomial model fit. Mean winter temperatures during the most recent period (1995-2004) ranged from 14.2C (Newnans, Alachua County) to 18.9C (Thonotosassa, Hillsborough County), with an average of 16.8.2C. This exceeded mean temperatures for the period of 1975-1984 by 1.2.0C, but was also 1.0.8C cooler than the preceding ten-year interval. All lakes either cooled (18 cases), or displayed a reduction in the rate of warming (4 cases) relative to the prior decade. The shape and statistical significance of the corresponding regression model remained nearly unchanged from the second decade, and latitude remained the only variable significant at the p=0.1 level (Figures 4-5, 4-7): 2)(74.0)(50.403.536LatLatT ,(r 2 =0.75, p<0.01, p 1 <0.05, p 2 <0.05) (T = mean winter surface water temperature in C, Lat = latitude in degrees north).

PAGE 63

51 Figure 4-5. Mean winter surface water temperatures by latitude during 1995-2000 and second-order polynomial model fit. Because the relationship between temperature and latitude was clearly non-linear during the two most recent ten-year intervals, a polynomial model was retrospectively fitted to data from the first for comparison (Figure 4-6, 4-7): 2)(54.0)(53.281.362LatLatT , (r 2 =0.75, p<0.01, p 1 >0.1, p 2 >0.1). (T = mean winter surface water temperature in C, Lat = latitude in degrees north). While the parametric model term was statistically non-significant and the correlation coefficient improved only slightly, similarity to the consistently convex regression fits for the most recent trials implies that this alternative model is plausible and should not be disregarded by default. One-way ANOVA indicated that differences between ten-year winter means were significant (p<0.01). Multiple comparisons based on Fisher’s LSD method with 95% confidence intervals interpreted the three means as mutually distinct. The more

PAGE 64

52 conservative Tukey’s W procedure combined the two most recent periods into one group (although with minimal overlap); distinct from the first. At p0.06, Tukey’s method also recognized significant differences between all three ten-year means. Figure 4-6. Mean winter surface water temperatures by latitude during 1975-1984. Linear and alternative (second-order polynomial) model fits are indicated by the blue solid and red dotted line, respectively. In summary, seasonal regression analyses showed that none of the evaluated variables exerted significant control over summer temperatures. During winter, latitudinal control dominated. This pattern is similar to that of air temperatures as discussed in Chapter 1. Significant long-term temperature variability was identified for winter, whereas summer temperatures exhibited relative constancy over time. Seasonal Temperature Trajectories An alternative method was utilized to evaluate long-term trends based on temperature trajectories of individual lakes by season. Linear regression models were fitted to time series of mean annual summer and winter temperatures. The assemblages of

PAGE 65

53 summer and winter model slopes, respectively, were determined by Anderson-Darling (p summer >0.2, p winter >0.8), Ryan-Joiner (p summer >0.1, p winter >0.1) and Kolmogorov-Smirnov (p summer >0.15, p winter >0.15) tests to follow approximately normal distributions that could, therefore, be evaluated with one-sample t-tests. Rejection of the test null hypothesis (H 0 = “the distribution of slopes cannot be distinguished from one with mean zero”) would constitute evidence for a statistically significant temperature increase over the period of record, whereas failure to reject would indicate constancy. Figure 4-7. Mean decadal winter surface temperature by latitude (overview). Interval 1 (1975-1984) in blue (solid), interval 2 (1985-1994) in red (long dashes) and interval 3 (1995-2004) in green (short dashes). This procedure was based on the assumption that, although regression models fitted to individual lake records were in almost all cases non-significant due to substantial interannual temperature variability that by far exceeded the magnitude of any potential signal of long-term change, the distribution of slopes was nevertheless indicative of

PAGE 66

54 temperature trajectories of Florida lakes as a group. While data gaps and influential observations may have affected the quality of individual lake regressions, data collection and standardization methods utilized for this research reduced the likelihood of systematic errors or biases skewing the statistical distributions of slopes sufficiently to falsify the analysis. Error propagation was further controlled by restricting the analysis to 20 lakes with high-quality records and using a strict confidence level of =0.01. For winter, three time series slopes were negative (defined subjectively as ), five displayed no change ( )2.0(m 2.0m ), and twelve were positive () (Figure 4-8). A one-sample t-test on this distribution resulted in rejection of the null hypothesis (H 2.0m 0 : 0winterm , H a : 0winterm , =0.01, p<0.01), indicating warming. Based on the distribution mean of 0.034 and a mean period of record of 29.15 years, winter temperatures increased by 1.0.9C between 1973 and 2003 (0.034 .032C per year). By truncating this analysis to years between 1979 () and 2002 (), uncertainty due to data gaps and differences in record length were minimized (H 0 : 0winterm , H a : 0winterm , =0.01, p<0.01). During this time, winter temperatures increased by 1.4.9C (0.062.039C per year). It must be emphasized, however, that this estimate is likely liberal because the restriction to data from 1979 () to 2002 () accentuates the period of unusually rapid temperature increase between the first and second study intervals, whereas the relatively cooler first and third study intervals are only partially captured. During summer, six regression slopes were negative, nine neutral and five positive per the above definition (Figure 4-9). The t-test on the summer distribution did not support rejection of the null hypothesis (H 0 : summerm =0, H a: summerm >0, =0.01, p>0.7), indicating no significant long-term change.

PAGE 67

55 Winter Temperature Trends (Time Series)02468101214IncreaseNo ChangeDecreaseTrendNumber of Lakes A Distribution of Regression Slopes (Winter)-0.1-0.0500.050.10.150.21751091872120121614153194611813LakeSlope B Figure 4-8. Winter temperature trends. A) Overview, derived from time series regression slopes (m) based on individual, non-truncated lake records. Increase is defined as m>0.2, no change as m <0.2, and decrease as m<(-0.2). B) Distribution of time series regression slopes of individual lake records for winter temperatures. Numbers correspond to lakes as follows: 1: Annie (Highlands), 2: Istokpoga (Highlands), 3: Blue Cypress (Indian River), 4: Kissimmee (Osceola), 5: Hatchineha (Polk/Osceola), 6: Thonotosassa (Hillsborough), 7: Cypress (Osceola), 8: Tohopekaliga (Osceola), 9: East Tohopekaliga (Osceola), 10: Underhill (Orange), 11: Apopka (Lake/Orange), 12: Jessup (Seminole), 13: Beauclair (Lake/Orange), 14: Monroe (Seminole/Volusia), 15: Eustis (Lake), 16: Griffin (Lake/Marion), 17: Weir (Marion), 18: George (Volusia/Putnam), 19: Crescent (Flagler/Putnam), 20: Newnans (Alachua)

PAGE 68

56 Summer Temperature Trends (Time Series)02468101214IncreaseNo ChangeDecreaseTrendNumber of Lakes A Distribution of Regression Slopes (Summer)-0.1-0.0500.050.10.150.21918381120151714459213161761210LakeSlope B Figure 4-9. Summer temperature trends. A) Overview, derived from time series regression slopes (m) based on individual, non-truncated lake records. Increase is defined as m>0.2, no change as m <0.2, and decrease as m<(-0.2). B) Distribution of time series regression slopes of individual lake records for summer temperatures. Numbers correspond to lakes as in Figure 4-8B. While the slope analyses were subject to greater uncertainty than other methods used in this study, and trends were averaged over the entire period of record, masking complex interdecadal patterns, results were in general agreement with the previously

PAGE 69

57 derived regression models. Both methods indicated relatively constant summer temperatures and increasing winter temperatures during the study period. Cluster Analysis Data limitations precluded sequential cluster analyses based on the method by Beaver et al. (1981) to identify thermal regime boundary shifts. This was primarily because of data gaps that occurred even in the most complete lake records, preventing computation of representative ten-year averages for each month over the entire 30-year study period. It was possible, however, to establish approximate latitudinal boundary lines based on clustering of mean winter surface water temperatures of successive ten-year periods, because missing observations could be approximated by the temperature-latitude relationships derived earlier in this chapter. The Ward method of hierarchical fusion was used to establish three groups corresponding to the thermal regime zones described by Beaver et al. (1981). Subsequently, group means were statistically compared via ANOVA analyses and multiple comparisons based on Fisher’s LSD and Tukey’s W procedures. Clustering based on the first two study intervals produced associations broadly reflecting the previously described latitudinal winter temperature gradients, although two central Florida lakes failed to associate with their expected geographic groups (Figure 4-10). Apopka (Lake/Orange County) unexpectly aligned with the northern group, whereas Beauclair (Lake/Orange County), just 18 km south of Lake Apopka and hydrologically connected to it via the Apopka-Beauclair Canal, clustered with the southern group for reasons that are not inherently clear. Disregarding the two anomalous systems, approximate latitudinal boundary lines delimit the northern group to >29.1N and the

PAGE 70

58 southern group <28.4N, with a narrow, ca. 78 km wide transitional zone occupying the intermediate area (Figure 4-11). The mean temperature of the northern group, which consisted of four lakes, was 14.7.5C, that of the central group, consisting of eight lakes 16.2.4C, and that of the southern group, with eight lakes, 17.7.5C. The analysis of variance was significant at p=0.01, and both Fisher’s LSD and Tukey’s W procedures distinguished between each mean under prescribed error rates of 0.05. LakeDistance 20222112171619181311151014894637251 12.848.564.280.00 Cluster Analysis of Winter Means, Study Periods 1-2 Figure 4-10. Dendrogram of winter surface water temperatures based on 22 Florida lakes between study periods 1 (1975-1984) and 2 (1985-1994). Numbers correspond to lakes as follows: 1: Annie (Highlands), 2: Istokpoga (Highlands), 3: Blue Cypress (Indian River), 4: Kissimmee (Osceola), 5: Hatchineha (Polk/Osceola), 6: Thonotosassa (Hillsborough), 7: Cypress (Osceola), 8: Tohopekaliga (Osceola), 9: East Tohopekaliga (Osceola), 10: Underhill (Orange), 11: Fairview (Orange), 12: Apopka (Lake/Orange), 13: Jessup (Seminole), 14: Beauclair (Lake/Orange), 15: Harris (Lake), 16: Monroe (Seminole/Volusia), 17: Eustis (Lake), 18: Griffin (Lake/Marion), 19: Weir (Marion), 20: George (Volusia/Putnam), 21: Crescent (Flagler/Putnam), 22: Newnans (Alachua)

PAGE 71

59 Figure 4-11. Zonal boundaries of thermal regimes of Florida lakes based on Figure 4-10. The dots indicate location of study lakes, their color the group of association: Red: southern, green: central, blue: northern. Associated zones: >29.1N (blue shading): northern, <28.4N (red shading): southern, and central (green shading). Clusters based on the last two study intervals show a northward shift of the southern group whereas the boundary between central and northern groups has remained static. There is some overlap between southern and central groups, as lakes Beauclair and Monroe aligned with the southern assemblage despite being located north of lakes Apopka and Jessup, which associated with the central, transitional regime (Figure 4-12).

PAGE 72

60 LakeDistance 21222017161819131211610153914584271 14.039.354.680.00 Cluster Analysis of Winter Means, Study Periods 2-3 Figure 4-12. Dendrogram of winter surface water temperatures based on 22 Florida lakes between study periods 2 (1985-1994) and 3 (1995-2004). Numbers correspond to lakes as listed in Figure 4-10. Approximate boundary lines place the northern group at >29.1N and the southern at <28.7.15N (range). This indicates a northward shift of the southern zone by 33 km, reducing the width of the transitional zone by the same amount (Figure 4-13). The uncertainty in the exact boundary line location derives from the overlap between southern and central zone lakes. The mean temperature of the northern group, which consisted of three lakes, was 15.1.6C, that of the central group, consisting of six lakes 16.7.8C, and that of the southern group, with thirteen lakes, 18.10.8C (SD). The analysis of variance was significant at p=0.01, and both Fisher’s LSD and Tukey’s W procedures distinguished between each mean under prescribed error rates of 0.05.

PAGE 73

61 Figure 4-13. Zonal boundaries of thermal regimes of Florida lakes based on Figure 4-12. The dots indicate location of study lakes, their color the group of association: Red: southern, green: central, blue: northern. Associated zones: >29.1N (blue shading): northern, <28.7.15N (red shading): southern, and central (green shading). The observed pattern of change matches expectations. All zones were subject to approximately equal warming (northern: 0.45C, central: 0.53C, southern: 0.45C), yet boundary shifts were limited to the interface between southern and transitional zones. This reflects the non-linear relationships between winter water temperatures and latitude

PAGE 74

62 established via regression, which implied greater isothermal line shifts near the southern margin of the study area than in the north.

PAGE 75

CHAPTER 5 DISCUSSION AND SUGGESTIONS FOR FUTURE RESEARCH Relationship between Air and Lake Temperatures The preceding analyses have demonstrated consistently that air and lake temperatures exhibit similar geospatial and temporal patterns in Florida, with latitudinal temperature gradients most pronounced during winter and absent during summer. While Winsberg (1990) identified that distance from the coast had significant influence on air temperatures, a similar relationship could not be established for water temperatures. This is possibly due to the fact that most study lakes were located nearly equidistant between the Gulf of Mexico and Atlantic Ocean, a region where maritime influence is less evident than in the rest of the state. Among morphometric and biological variables evaluated, none exhibited systematic influence on surface water temperatures in ten-year multiple regression models, with the possible exception of depth, which could not be directly evaluated due to missing bathymetric information for several study lakes. While surface area was not significantly correlated with surface water temperatures, the relationship could not be evaluated for small lakes because the smallest study lake (Annie, A=34.4 ha) was larger than the statewide median. Results consistently indicate that, in the medium to long-term, surface water temperatures of Florida lakes are generally representative of regional air temperatures. It is suggested that, by referencing multiple lake records and interpolating between them, models of local climate can be constructed. The key advantage to aquatic temperature 63

PAGE 76

64 records is sharply reduced susceptibility to short-term variability due to the high specific heat of water, thus, reducing the influence of diurnal temperature flux and transient weather events. Lakes can be described as thermal integrators, with records providing what constitutes a smoothed, “moving average” of local temperatures. In model development, it must be taken into account that during the study period, water temperature means consistently exceeded corresponding air temperature means by approximately 2C with little geospatial variability. Accuracy of lake-based climate models would likely be reduced where steep microclimatic gradients exist, such as coastal areas where air temperature differences of several degrees over <50 km are common (Winsberg 1990). Compensation by incorporating more closely spaced reference lakes is not always possible due to their heterogeneous distribution in the Florida peninsula. Suitability of lake-based climate models would also be reduced for applications where temperature extremes and short-term variability rather than long-term trends are to be evaluated. The collected baseline data and analysis results should facilitate model development and inspire similar studies in other regions where lake districts coincide with climatic transitions. Temporal Trends Results of this study paint an ambiguous picture about detectability of climatic change in Florida between 1968 and 2004. Summer temperatures exhibited great constancy between ten-year intervals, providing no evidence for long-term trends or cycles. During winters, significant temperature differences over time were established, but the pattern and magnitude of change do not preclude random natural variability as the causal mechanism.

PAGE 77

65 The derived estimate of 1.0.9C warming between 1973 and 2003 suggests a net winter warming trend, as does the northward shift of the boundary between transitional and southern lake zones. Successive regression models, however, clearly showed that the period between 1995 and 2004, which coincided in part with the globally warmest decade on record and included the warmest year on record, was cooler relative to the preceding ten-year interval. All lakes considered in the analysis displayed cooling or a decrease in the rate of warming during winter relative to the previous study interval. As described in Chapter 1, the climate of Florida is subject to strong interannual variability. It is plausible that part of the observed temperature differences between study intervals were caused by atypically distributed clusters of warm or cool years influencing the means. There did not appear to be a relationship between ten-year means and El Nio-Southern Oscillation (ENSO) cycles. During each study interval, the number of ENSO warm and cold periods, based on the Nio Oceanic Index (NOI) for the period of January through March, were close to balanced (), and mean ten-year index values were +0.08, +0.20 and -0.02 from the first to the most recent study intervals (National Weather Service 2005). Positive index values are associated with cooler temperatures in Florida, whereas negative values indicate warm temperatures (Hansen et al. 1999). Therefore, cooler mean temperatures would have been expected during the second study interval, which does not match the observed pattern (Table 5-1). Continued monitoring and reporting of lake temperatures will generate data to build upon the baseline analyses performed in this study and allow more definitive conclusions in the future as the rate of climatic change increases. Use of study intervals longer than

PAGE 78

66 ten years is suggested to reduce the influence of atypical patterns of natural interannual variability on the analysis. Table 5-1: ENSO during the study period Study Interval No. of ENSO Warm Periods No. of ENSO Cold Periods Mean NOI Mean Lake Water Temperature 1975-1984 2 2 +0.08 15.61.6C 1985-1994 3 2 +0.20 17.81.2C 1995-2004 3 4 -0.02 16.81.2C Table 5-1. ENSO during the study period. Number of warm and cold episodes and mean Nio Oceanic Index (NOI) values apply to the period January-March (National Weather Service 2005), whereas mean lake water temperatures apply to the period January-February. ENSO warm periods are associated with cool winters in Florida, whereas cold periods produce mild winters (Hansen et al. 1999). Ecological Implications Likely Effects during the Study Period The constancy of summer temperature means during the study period implies no immediate threat of exceeding thermal tolerance limits of warm-temperate, heat-limited species near their southern distributional limit. Winters are, however, in many respects the most influential season in terms of ecology, because winter temperatures constrain habitat ranges of cold-limited exotic-invasive species (Thomas et al. 2001), and, in lentic, monomictic ecosystems, the duration and timing of winter mixing (Beaver et al. 1981), affecting nutrient cycling and hypolimnion oxygenation. Should summer temperatures remain stagnant while winter temperatures continue to rise, warm temperate species may nevertheless be displaced from habitats in the current transitional and warm temperate zones as cold-limited species expand northward, altering ecosystem composition and competition-predation dynamics. In eutrophic lakes with high rates of biomatter decay, earlier onset of stratification and later breakdown thereof increases the potential for bottom water deoxygenation, adversely impacting benthic communities.

PAGE 79

67 The polynomial temperature-latitude relationships observed during winters indicate that isothermal line shifts in response to interdecadal temperature variability were larger in the southern region of the study area than the northern. This was confirmed by the cluster analyses, which indicated boundary shifts between southern and transitional lake assemblages, whereas the northern zone was unaffected. The long-term pattern of strong warming followed by a moderate temperature reversion suggests highly dynamic ecological processes along the subtropical-transitional zonal boundary. It is likely that succession dynamics analogous to the seasonal changes in relative abundance of subtropical and temperate assemblages of Cladocera in Florida lakes, as described by Frey (1982), occur on interannual and interdecadal timescales in response to temperature variability. Specifically, subtropical species would be expected to colonize new habitats north of their present range following warm winters, with transitional and temperate species rebounding following prolonged cold (cf. Walther et al. 2002). Since the net trajectory during the study period indicated warming, it is likely that moderate range expansion of subtropical species, paralleling zonal boundary shifts, has occurred, possibly resulting in the establishment of viable populations north of historical habitats. Associated ecosystem structural adjustments and microevolutionary adaptation are likely to have increased resilience and capacity for recovery of invading species following future disturbances. The boundary between subtropical and transitional lake zones, as established in the cluster analyses for winter, passes through counties Orange and Polk, which have the highest density of lakes within the peninsular district. Transfer of aquatic species between lakes in this area is facilitated by their proximity to each other, as well as

PAGE 80

68 interconnections by streams and drainage canals. The area is also heavily populated, constituting more than 11% of Florida’s 17 million inhabitants (U.S. Census Bureau 2005). Human activities are likely to introduce exotic species and contribute to their dispersal. Future Scenarios While available evidence is insufficient to determine whether the observed pattern of variability in winter temperatures is attributable to climatic change, IPCC (2001a) predictions indicate that rates of warming will accelerate for the foreseeable future. Therefore, even if no clear signal for sustained change has been detected in Florida until now, it is likely that such a signal will manifest itself within the next few decades. Estimates for how projected warming, based on the Mulholland et al. (1997) regional climate model, will affect zonal boundaries can be made utilizing the latitude-temperature relationships derived from this study. Mulholland et al. (1997) estimates warming of 3C for Florida as CO 2 concentrations reach 560 ppm. IPCC scenarios predict these conditions will be met between 2050 and 2100. Based on the linear regression model of mean annual surface water temperatures by latitude (, Chapter 3), warming of 3C over Florida would result in a northward shift of isothermal lines by approximately 4.11 or 456 km, assuming no fundamental changes in the relationship. This value is reasonably close to Hughes’ (2000) mean global estimate of 300-400 km. Under this scenario and using the thermal regime boundary zones defined by Beaver et al. (1981) as reference (Figure 5-1), warm temperate and transitional zones would disappear from the peninsula, replaced in their )(73.069.44LatT

PAGE 81

69 entirety by the southern, subtropical zone (Figure 5-2). Due to the magnitude of the shift, it is plausible that an additional, tropical zone could become established. Figure 5-1. Thermal regime boundaries according to Beaver et al. (1981) relative to the study area (red shading). Loss of warm temperate and transitional climate zones over a period of <100 years is likely to result in severe alteration of north and central Florida ecosystems. Generalists capable of rapid expansion are likely to be at a competitive advantage over more sedentary organisms (Thomas et al. 2001), particularly specialists whose preferred foods or shelter may display different thermal tolerances or rates of dispersal, disrupting

PAGE 82

70 coevolved species interactions. Such rapid change in ecosystem structures is likely to result in undesirable consequences for ecosystem functions, including utilitarian uses. Conservation planners are urged to incorporate projections for climate-forced habitat adjustments into preserve design and management plans to ensure that conservation goals are reachable and sustainable in a time of accelerating global change. Figure 5-2. Thermal regimes zones after a 4.11 northward shift following warming of 3C based on extant temperature-latitude relationships and the Mulholland et al. (1997) RCM. Agricultural production will also be affected. While milder temperatures may permit northward expansion of cultivation areas for subtropical crops, and increased

PAGE 83

71 atmospheric CO 2 concentrations may result in higher yields, changes in rainfall patterns and range expansion of crop pests could negate these benefits (IPCC 2001b). Production methods will need to be adjusted to minimize adverse impacts. Of particular concern to lake managers will be increased incidence of deoxygenation, particularly in eutrophic lakes. In addition to the previously described mechanism for deoxygenation of hypolimnia of stratified lakes associated with shorter winter mixing periods, shallow, non-stratified lakes will likely experience more frequent oxygen stress and fish kills: Higher temperatures are associated with decreased dissolved oxygen storage capacity and increased biological productivity (and, therefore, greater biomatter decay, increasing biological oxygen demand). Periods of reduced photosynthesis due to prolonged summer cloudiness, input of high quantities of allochthonous organic matter (e.g. color) over short periods of time, or other conditions that result in respiration exceeding photosynthetic oxygen production for periods of multiple days will induce anoxia more rapidly than is presently the case. Suggestions for Future Research This study provides baseline data and assesses climate dynamics along the warm temperate-subtropical transition in the Florida peninsula, but was unable to link the observed patterns to long-term global climate change in part because data limitations restricted analyses spatially and temporally. Government agencies, universities and volunteer organizations are therefore strongly urged to report lake monitoring data to publicly accessible databases consistently. It is hoped that improvements in data collection and reporting will reduce such limitations, allowing future analyses to establish climate patterns over greater areas, as well as during spring and fall, which were not evaluated.

PAGE 84

72 Availability of depth profiles in particular would permit additional analyses of limnological and ecological interest. Lake morphometry is expected to exert greater influence on deep waters than the permanently well-mixed surface layer due to the importance of fetch in controlling vertical mixing. As such data become available, long-term studies on heat content variability, as well as changes over time in length and stability of stratification and hypolimnetic temperatures, will be possible, allowing evaluation of threats to benthic communities. The need for climate study on the regional scale was described in Chapter 1. It is hoped that the results of this study will encourage model development to improve projections of future climate change in Florida and associated ecological impacts. Assessment of ecological responses to long-term climate variability and/or change at the species, community and ecosystem levels requires field observations to augment and verify model predictions. Phenological studies of terrestrial and aquatic species are encouraged to detect changes in the timing of spring and fall activities, which constitutes one of the most accessible means of monitoring for climate-forced change on biota (Walther et al. 2002). Additionally, reassessment of distributions of species used in the past to delineate life zone boundaries is suggested to evaluate succession dynamics between warm temperate and subtropical taxa, and to track rates of dispersion of the latter.

PAGE 85

APPENDIX A LIST OF STUDY LAKES Table A-1: List of Study Lakes Lake County Latitude Longitude Area (ha) Alligator Columbia 30.411' 82.658' 135.2 Altho Alachua 29.737' 82.692' 216.0 Annie Highlands 27.451' 81.043' 34.4.0 Apopka Lake/Orange 28.55' 81.38' 12268.4 Ashby Volusia 28.745' 81.755' 412.0 Beauclair Lake/Orange 28.368' 81.695' 444.4 Blue Cypress Indian River 27.83' 80.27' 2622.0 Buffum Polk 27.879' 81.751' 617.2 Crescent Flagler/Putnam 29.99' 81.24' 6384.0 Crooked Polk 27.48' 81.80' 2215.2 Crosby Bradford 29.552' 82.453' 214.4 Cypress Osceola 28.58' 81.44' 1638.8 Disston Flagler 29.179' 81.403' 737.6 Dora Lake/Orange 28.25' 81.69' 1790.0 East Tohopekaliga Osceola 28.83' 81.00' 4787.2 Eustis Lake 2850.96' 81.54' 3122.4 Fairview Orange 28.602' 81.353' 160.4 Geneva Clay 29.02' 82.41' 652.0 George Volusia/Putnam 29.19' 81.74' 18400 Griffin Lake/Marion 28.78' 81.71' 6602.0 Hampton Bradford 29.523' 82.157' 329.2 Harris Lake 28.40' 81.07' 6610.8 Hatchineha Polk/Osceola 28.18' 81.69' 2666.0 Istokpoga Highlands 27.45' 81.78' 11076.8 Jackson Highlands 27.33' 81.80' 1364.8 Jessup Seminole 28.34' 81.04' 4004.4 Kerr Marion 29.470' 81.240' 1132.0 Kingsley Clay 29.913' 81.930' 660.8 Kissimmee Osceola 27.48' 81.70' 13979.2 Lawne Orange 28.854' 81.233' 62.4 Lochloosa Alachua 29.62' 82.09' 3338.8 Lowery Polk 28.763' 81.708' 374.4 Minneola Lake 28.456' 81.096' 755.2 Monroe Seminole/Volusia 28.29' 81.47' 3762.4 Newnans Alachua 29.81' 82.24' 2970.8 Parker Polk 28.003' 81.832' 908.8 73

PAGE 86

74 Lake County Latitude Longitude Area (ha) Rowell Bradford 29.217' 82.525' 145.6 Sampson Bradford 29.672' 82.295' 816.8 Santa Fe Alachua 29.19' 82.52' 2342.4 Seminole Pinellas 27.267' 82.880' 1025.2 Tarpon Pinellas 28.98' 82.67' 1013.6 Thonotosassa Hillsborough 28.682' 82.712' 327.6 Tohopekaliga Osceola 28.46' 81.52' 7524.0 Trafford Collier 26.372' 81.626' 597.6 Tsala Apopka Citrus 28.35' 82.79' 7644.4 Underhill Orange 28.316' 81.202' 58.8 Washington Brevard 28.59' 80.61' 1744.8 Wauberg Alachua 29.838' 82.113' 99.2 Weir Marion 29.87' 81.51' 2402.0 Winnemissett Volusia 29.394' 81.990' 65.6 Yale Lake 28.755' 81.169' 1616.8 Table A-1. List of study lakes, their location and surface area.

PAGE 87

APPENDIX B LIST OF REFERENCED WEATHER STATIONS Table B-1: Referenced Weather Stations Station Name County Latitude Longitude Archbold Biological Station Highlands 27' 81' Avon Park 2 W Highlands 27' 81' Bartow Polk 27' 81' Clermont 7 S Lake 28' 81' Crescent City Putnam 29' 81' DeLand 1 SSE Volusia 29' 81' Fort Pierce Saint Lucie 27' 80' Gainesville 3 WSW Alachua 29' 82' Gainesville Municipal Airport Alachua 29' 82' High Springs Alachua 29' 82' Immokalee 3 NNW Collier 26' 81' Inverness 3 SE Citrus 28' 82' Kissimmee 2 Osceola 28' 81' La Belle Hendry 26' 81' Lake Alfred Exp. Station Polk 28' 81' Lake City 2 E Columbia 30' 82' Lakeland Polk 28' 81' Melbourne WSO Brevard 28' 80' Moore Haven Lock 1 Glades 26' 81' Mountain Lake Polk 27' 81' Ocala Marion 29' 82' Orlando WSO McCoy Orange 28' 81' Palatka Putnam 29' 81' St. Petersburg Pinellas 27' 82' Sanford Exp. Station Seminole 28' 81' Starke Bradford 29' 82' Tampa WSCMO Airport Hissborough 2758' 82' Table B-1. Weather stations queried to establish the air temperature-latitude relationship in Figure 3-2. Data are 1971-2000 National Climatic Data Center normals provided by the Southeast Regional Climate Center (2005). 75

PAGE 88

LIST OF REFERENCES Beaver, J. R., T. L. Crisman, and J. S. Bays. 1981. Thermal regimes of Florida lakes. Hydrobiologia 83: 267-273. Brenner, M., M. W. Binford, and E. S. Deevey. 1990. Lakes, p. 364-391. In: R. L. Myers and J. J. Ewel [eds.], Ecosystems of Florida. Orlando, FL: University of Central Florida Press. Brown, N. C. 1909. Preliminary examination of the forest conditions of Florida. Report to the State of Florida. Washington, D.C.: U.S. Forest Service. Canfield, D. E. 1981. Chemical and trophic state characteristics of Florida lakes in relation to regional geology. Report to the Cooperative Fish and Wildlife Research Unit. Gainesville, FL: University of Florida. Chen, E. & J. F. Gerber. 1990. Climate, p. 11-34. In: R. L. Myers and J. J. Ewel [eds.], Ecosystems of Florida. Orlando, FL: University of Central Florida Press. Climateprediction.net. 2005. Modeling the climate. Retrieved 11 June 2005 from http://www.climateprediction.net/science/model-intro.php . Crossman, R. A. 1956. The ecology of Chaoborus in lake Newnan. Master’s Thesis, University of Florida. Deevey, E. S. 1988. Estimation of downward leakage from Florida lakes. Limnology and Oceanography 33: 1308-1320. Dohrenwend, R. E. and L. Harris. 1975. A climatic change impact analysis of peninsular Florida life zones. Impacts of climatic change on the biosphere. DOT-TST-75-55. Washington, D.C.: U.S. Department of Transportation. Florida Lakewatch. 2004a. A beginner’s guide to water management – color. Gainesville, FL: University of Florida Institute of Food and Agricultural Sciences. -----. 2004b. A beginner’s guide to water management – oxygen and temperatures. Gainesville, FL: University of Florida Institute of Food and Agricultural Sciences. -----. 2005. Florida Lakwatch. Retrieved 25 March 2005 from http://lakewatch.ifas.ufl.edu/ . 76

PAGE 89

77 Frey, D. G. 1982. Contrasting strategies of gamogenesis in northern and southern populations of Cladocera. Ecology 63: 223-241. Grimm, E. C., G. L. Jacobsen, W. A. Watts, B. C. S. Hansen, and K. A. Maasch. 1993. A 50,000-year record of climate oscillations from Florida and its temporal correlation to Heinrich events. Science 261: 198-200. Hansen, J. W., J. W. Jones, C. F. Kiker, and A. W. Hodges. 1999. El Nino-Southern Oscillation impacts on winter vegetable production in Florida. Climate 12: 92-102. Henry, J. A. 1998. Weather and climate, p. 16-37. In: E. A. Fernald and E. D. Purdum [eds.], Water resources atlas of Florida. Tallahassee, FL: Florida State University Institute of Science and Public Affairs. Huber, W. C., P. L. Brezonik, J. P. Heaney, R. E. Dickinson, S. D. Preston, D. S. Dwornik, and M. A. DeMaio. 1982. A classification of Florida lakes. Florida Water Resources Research Center Publication No. 72. Gainesville, FL: University of Florida. Hughes, L. 2000. Biological consequences of global warming: is the signal already apparent? Trends Ecol Evol 15: 56-61. Intergovernmental Panel on Climate Change [IPCC]. 2001a. Climate change 2001: the scientific basis. Cambridge, UK: Cambridge University Press. -----. 2001b. Climate change 2001: impacts, adaptation and vulnerability. Cambridge, UK: Cambridge University Press. Kennedy, D. 2004. Climate change and climate science. Science 304: 1565. Kindinger, J. L., J. B. David, and J. G. Flocks. 1999. Geology and evolution of lakes in north-central Florida. Environ Geol 38: 301-321. MacCracken, M., J. Smith, and A. C. Janetos. 2004. Reliable regional climate models not yet on horizon. Nature 429: 699. Mossa, J. 1998. Surface water, p. 64-81. In: E. A. Fernald and E. D. Purdum [eds.]: Water resources atlas of Florida. Tallahassee, FL: Florida State University Institute of Science and Public Affairs. Mulholland, P. J., G. R. Best, C. C. Coutant, G. M. Hornberger, J. L. Meyer, P. J. Robinson, J. R. Stenberg, R. E. Turner, F. Vera-Herrera, and R. G. Wetzel. 1997. Effects of climate change on freshwater ecosystems of the south-eastern United States and the gulf coast of Mexico. Hydrol Process 11: 949-970. National Weather Service. 2005. Cold and warm episodes by season. Retrieved 8 June 2005 from http://www.cpc.ncep.noaa.gov/products/analysis_monitoring/ensostuff/ ensoyears.shtml .

PAGE 90

78 Obeysekera, J., J. Browder, L. Hornung, and M. Harwell. 1999. The natural South Florida system I: climate, geology, and hydrology. Urban Ecosystems 3: 223-244. Oreskes, N. 2004. Beyond the ivory tower: the scientific consensus on climate change. Science 306: 1686. Palmer, S. L. 1984. Surface water, p. 54-67. In: E. A. Fernald and D. J. Patton [eds.]: Water resources atlas of Florida. Tallahassee, FL: Florida State University. Parmesan, C. and H. Galbraith. 2004. Observed impacts of global climate change in the U.S. Pew Center on Global Climate Change. Retrieved 22 January 2005 from http://www.pewclimate.org/docUploads/final%5FObsImpact%2Epdf . Pearson, P. N. and M. R. Palmer. 2000. Atmospheric carbon dioxide concentrations over the past 60 million years. Nature 406: 695-699. Randall, D. A. 2001. A global modeler looks at regional climate modeling. Retrieved 11 June 2005 from http://www.isse.ucar.edu/rcw/presentations/randall.pdf . Rosenbaum, W. A. 2002. Environmental politics and policy, 5th ed. Washington, D.C.: CQ Press. Ruddiman, W. F. 2001. Earth’s climate: past and future. New York: W. H. Freeman and Company. Schiermeier, Q. 2004. Modellers deplore ‘short-termism’ on climate. Nature 428: 593. Scott, T. M., K. M. Campbell, F. R. Rupert, J. D. Arthur, T. M. Missimer, J. M. Lloyd, J. W. Yon, and J. G. Duncan. 2001. Geologic map of the State of Florida. Tallahassee, FL: Florida Geological Survey. Shafer, M. D., R. E. Dickinson, J. P. Heaney, and W. D. Huber. 1986. Gazetteer of Florida lakes. Florida Water Resources Research Center Publication No. 96. Gainesville, FL: University of Florida. Shannon, E. E. and P. L. Brezonik. 1972. Limnological characteristics of north and central Florida lakes. Limnol Oceanogr 17: 97-110. Southeast Regional Climate Center [SERCC]. 2005. Historical climate summaries for Florida. Retrieved 1 July 2005 from http://www.dnr.state.sc.us/climate/sercc/climateinfo/historical/historical_fl.html . Stewart, J. W. and G. H. Hughes. 1974. Hydrologic consequences of using ground water to maintain lake levels affected by water wells near Tampa, Florida. Tallahassee, FL: Florida Department of Natural Resources.

PAGE 91

79 Thomas, C. D, E. J. Bodsworth, R. J. Wilson, A. D. Simmons, Z. G. Davies, M. Musche, and L. Conradt. 2001. Ecological and evolutionary processes at expanding range margins. Nature 411: 577-581. U.S. Census Bureau. 1995. Florida: population of counties by decennial census: 1900 to 1990. Retrieved 22 January 2005 from http://www.census.gov/population/cencounts/fl190090.txt . -----. 2005. Florida quickfacts. Retrieved 22 January 2005 from: http://quickfacts.census.gov/qfd/states/12000.html . Walther, G.-R., E. Post, P. Convey, A. Menzel, C. Parmesan, T. J. C. Beebee, J.-M. Fromentin, O. Hoegh-Guldenberg, and F. Bairlein. 2002. Ecological responses to recent climate change. Nature 416: 389-395. White, W. A. 1970. The geomorphology of the Florida peninsula. Florida Bureau of Geology Geological Bulletin 51: 1-165. Winsberg, M. D. 1990. Florida weather. Orlando, FL: University of Central Florida Press. Yount, J. L. 1963. South Atlantic states, p. 269-286. In: D. G. Frey [ed.], Limnology in North America. Madison, WI: University of Wisconsin Press. Zachos, J., M. Pagani, L. Sloan, E. Thomas, and K. Billups. 2001. Trends, rhythms and aberrations in global climate 65 ma to present. Science 292: 686-693.

PAGE 92

BIOGRAPHICAL SKETCH Danny Coenen was born in Jlich, Germany, on October 27, 1978. He resettled to the United States in December, 1997, to marry Jennifer L. Foster in January, 1998. He enrolled at Southwest Missouri State University in Springfield, Missouri, from 1998 until 2000, at which time he transferred to the School of Natural Resources and Environment at the University of Florida in Gainesville, Florida. After earning a Bachelor of Science degree in environmental science cum laude in May, 2003, he remained at the School of Natural Resources and Environment to begin graduate studies in August, 2003, earning the Master of Science in interdisciplinary ecology degree in August, 2005. Danny Coenen expects to enter the interdisciplinary ecology Ph. D. program of the School of Natural Resources and Environment as an Alumni Fellow with the support of the Department of Environmental Engineering Sciences in August, 2005. 80