Groundwater’s significance to changing hydrology, water chemistry, and biological communities of a floodplain ecosystem, Everglades, South Florida, USA
The Everglades (Florida, USA) is one of the world’s larger subtropical peatlands with biological communities adapted to waters low in total dissolved solids and nutrients. Detecting how the pre-drainage hydrological system has been altered is crucial to preserving its functional attributes. However, reliable tools for hindcasting historic conditions in the Everglades are limited. A recent synthesis demonstrates that the proportion of surface-water inflows has increased relative to precipitation, accounting for 33% of total inputs compared with 18% historically. The largest new source of water is canal drainage from areas of former wetlands converted to agriculture. Interactions between groundwater and surface water have also increased, due to increasing vertical hydraulic gradients resulting from topographic and water-level alterations on the otherwise extremely flat landscape. Environmental solute tracer data were used to determine groundwater’s changing role, from a freshwater storage reservoir that sustained the Everglades ecosystem during dry periods to a reservoir of increasingly degraded water quality. Although some of this degradation is attributable to increased discharge of deep saline groundwater, other mineral sources such as fertilizer additives and peat oxidation have made a greater contribution to water-quality changes that are altering mineral-sensitive biological communities.
KeywordsCoastal aquifers Contamination Groundwater/surface-water relations Paleohydrology USA Wetlands Everglades
Importance de l’eau souterraine dans les modifications de l’hydrologie, de la chimie des eaux et des associations biologiques d’un ecosystème de plaine inondable, les Everglades, Floride du Sud, USA
Les Everglades (Floride, U.S.A.)forment une des plus importantes étendues de tourbières subtropicales mondiale avec des associations biologiques adaptées aux eaux pauvres en sels et en éléments nutritifs dissous. La mise en évidence de la manière dont le système hydrologique antérieur au drainage a été altéré est cruciale pour la préservation des caractéristiques de son fonctionnement. Cependant, des outils fiables pour reconstituer les conditions historiques dans les Everglades ont des limites. Une synthèse récente démontre que la part des arrivées d’eau de surface a augmenté en fonction des précipitations, représentant 33% du total des apports à comparer aux 18% historiques. La nouvelle source d’eau la plus importante est le canal de drainage d’anciennes zones humides converties à l’agriculture. Les interactions entre eau souterraine et eau de surface ont aussi augmenté, du fait de l’accroissement des gradients hydrauliques verticaux résultant de modifications de la topographie et de la surface piézométrique dans un paysage extrêmement plat par ailleurs. Des données provenant de traceurs environnementaux en solution ont été utilisées pour déterminer le rôle variable de l’eau souterraine, provenant d’un réservoir de stockage d’eau douce qui a alimenté l’écosystème des Everglades pendant les périodes sèches vers un réservoir d’eau dont la qualité est de plus en plus dégradée. Bien qu’une partie de cette dégradation soit attribuable à l’augmentation de l’écoulement d’eau souterraine saline profonde, d’autres sources d’élément minéraux telles que les apports d’engrais et l’oxydation de la tourbe ont contribué de façon plus importante aux changement de qualité de l’eau ce qui altère les populations biologiques sensibles à la minéralisation.
La importancia de las aguas subterráneas en el cambio de la hidrología, la química del agua y las comunidades biológicas de un ecosistema de llanura de inundación, Everglades, Sur de Florida, EE.UU.
El Everglades (Florida, EE.UU.) es una de las turberas subtropicales más grandes del mundo con comunidades biológicas adaptadas a aguas de bajos contenidos de sólidos totales disueltos y nutrientes. Para preservar sus atributos funcionales es crucial detectar cómo el sistema hidrológico de pre-drenaje ha sido modificado. Sin embargo, las herramientas confiables para reconstruir las condiciones históricas en el Everglades son limitadas. Una reciente síntesis demuestra que proporcionalmente la afluencia de las aguas superficiales ha aumentado en relación a la precipitación, representando el 33% del total de los ingresos de agua comparados con el 18% histórico. La mayor fuente de este incremento está en el drenaje mediante canales de las zonas de los antiguos humedales que fueron convertidos en terrenos agrícolas. Las interacciones entre las aguas subterráneas y de superficie de agua también se han incrementado, debido al aumento de los gradientes hidráulicos verticales resultantes de alteraciones topográficas y del nivel de agua en un paisaje que de por sí es extremadamente llano. Los datos de trazadores de solutos ambientales fueron utilizados para determinar el papel que juega el cambio en el agua subterránea, desde un almacenamiento de agua dulce que sostenía el ecosistema de Everglades durante los períodos de sequías a un reservorio cada vez más degradado en la calidad del agua. Aunque parte de esta degradación es atribuible a un aumento de la descarga de aguas subterráneas profundas salinas, otras fuentes de minerales tales como aditivos fertilizantes y oxidación de turba han originado una mayor contribución a los cambios de la calidad del agua, que alteran a las comunidades biológicas sensibles a la composición mineral.
美国佛罗里达州Everglades地区是世界上较大的亚热带泥炭沉积地区之一, 拥有适应了在总溶解固形物和营养物质含量较低的水中生存的生物群落。查明排水后水文系统是如何发生变化的, 对于保持其生态功能非常重要。然而, 可用于重建Everglades地区历史条件的工具还很有限。最近, 一项综合的分析证明, 相对于降水而言, 流入的地表水占总补给量的比例已由过去的18%增加到了33%。新的水源主要来自渠道的渗漏。该农业区是由先前的湿地改造而成的。由于原本平坦的地形受到改变, 导致垂向水力梯度不断增加, 使得地下水与地表水的相互作用得到增强。采用环境溶质示踪剂数据确定了地下水的变化情况 : 从一个在旱季时维持Everglades地区生态系统平衡的淡水层到一个水质日益恶化的含水层。虽然部分水质恶化是由深层高矿化度地下水排泄造成的, 但其它的一些来源, 比如化肥添加剂和泥炭氧化作用, 对于水质的变化的作用更大。水质的变化正在改变当地的对矿物质敏感的生物群落。
Importância da água subterrânea na evolução da hidrologia, hidroquímica e comunidades biológicas do ecossistema de uma planície de inundação, Everglades, sul da Flórida, EUA
Everglades (Flórida, EUA) é uma das maiores zonas de turfeiras subtropicais a nível mundial, com comunidades biológicas adaptadas a águas com baixa concentração de sais dissolvidos e de nutrientes. A identificação da alteração dos pré-sistemas de drenagem hidrológica é crucial para a preservação dos seus atributos funcionais. No entanto, instrumentos fidedignos para analisar as condições históricas nos Everglades são limitados. Uma síntese recente demonstra que a proporção de entradas de água superficial aumentou relativamente à precipitação, representando 33% do total das entradas, em oposição aos 18% históricos. A maior fonte de água adicional procede da drenagem de canais de antigas zonas húmidas convertidas em zonas agrícolas. As interacções água subterrânea-água superficial também aumentaram, devido ao aumento dos gradientes hidráulicos verticais como consequência das alterações da topografia e do nível de água de uma paisagem outrora extremamente plana. Dados de traçadores ambientais foram usados para determinar o impacte da evolução da água subterrânea, a partir de um sistema natural de abastecimento de água doce que sustentou o ecossistema de Everglades durante períodos secos, a um abastecimento cuja qualidade da água se vai deteriorando gradualmente. Apesar de alguma desta degradação poder ser devida ao aumento da descarga de água salina profunda, outras fontes minerais, como os fertilizantes e a oxidação da matéria orgânica das turfeiras, têm tido um importante contributo para causas para as mudanças da qualidade da água, que estão a alterar as comunidades biológicas sensíveis à mineralização da água.
Large, low-gradient peatlands occur worldwide. Many of those, like the Everglades, are lotic peatlands with surface water flow occurring for much of the year through wetlands that are large enough to maintain low nutrient status and dominantly organic soils. Lotic peatlands dominated by emergent vegetation often exhibit a flow-related patterning; these landscapes include boreal fens (Glaser et al. 1981), the Okavango Delta in Botswana (Ellery et al. 2003; Gumbricht et al. 2004), morichal wetlands in Columbia and Venezuela (San Jose et al. 2001), buritizal wetlands in the Brazilian Pantanal, the Zapata peninsula in southern Cuba, Costa Rica’s lower Tempisque River basin, and some riparian wetlands on river floodplains. Groundwater’s influence in the hydrology of these large lotic wetlands has been revealed in a few instances (Siegel and Glaser 1987; McCarthy 2006; Harvey et al. 2006), although a comprehensive understanding of the role of groundwater in structuring these ecosystems has only begun to emerge.
The purpose of this paper is threefold: to briefly review the past decade’s advancements in understanding interactions between surface water and groundwater in the Everglades; second, to better define how groundwater’s role has changed in the overall hydrologic budget of the Everglades over the past century; and third, to synthesize knowledge about groundwater’s role in ecological function and sustainability of this low-nutrient subtropical floodplain ecosystem. There is increasing recognition that hydrologic inputs and their chemistry have changed substantially over the past century, and that even an alteration as seemingly innocuous as an increase in major ions has caused ecological changes.
Everglades hydrogeology: past and present
Prior to drainage that began in the early 1900s, surface water and groundwater interactions were a relatively small component of Everglades water budgets. Relatively small exchange fluxes between surface water and groundwater were due to the relatively weak driving forces for vertical flow on the pre-drainage landscape that was very flat and had not yet been altered by subsidence and the construction of canals and levees. Vertical flow resistance also was relatively high due to the continuous cap of undisturbed peat (ranging in thickness from 1–4 m) with a hydraulic conductivity several orders of magnitude lower than the aquifer beneath (Harvey et al. 2004). Recharge and discharge caused exchange between surface water and a relatively thin layer of “interactive” fresh groundwater near the top of the surficial aquifer (Harvey et al. 2006). Exchange fluxes varied spatially and temporally in accordance with changing patterns of precipitation and pulsed inflows of surface water from Lake Okeechobee or runoff from marginal wetlands. The primary factors influencing recharge and discharge were the patterns of rising or falling surface-water levels. For example, intense and localized rainfall, or pulsed inflow of water from the lake, created very low amplitude and long wavelength gravity waves that propagated southward through the wetlands causing fluctuations in surface water and groundwater interactions. Groundwater discharge tended to occur during average to moderately dry conditions when local surface-water levels were decreasing. Recharge tended to occur during moderately wet periods or during very dry periods just as water levels began to increase (Harvey et al. 2004).
Water management inadvertently increased the extent of surface water and groundwater interactions, beginning in the early 1900s with early efforts to drain the Everglades. Resulting oxidation and peat subsidence was most significant in the large area of wetlands south of Lake Okeechobee that was being drained for agriculture. Subsidence increased local groundwater discharge by focusing horizontal groundwater flow paths toward the area of subsidence and drainage. In the central Everglades, the direction of groundwater flow shifted from southerly to northwesterly flow (Miller 1988; Harvey et al. 2002). Dependence on canals increased throughout the second half of the twentieth century as the wetlands converted to agriculture continued to subside due to peat oxidation (Renken et al. 2005). Loss of peat and entrenchment of canals up to 5 m into the top part of the surficial aquifer in some locations brought surface waters into closer contact with groundwaters with high ionic strength (Harvey et al. 2002).
During the first half of the twentieth century, efforts to drain the Everglades eventually had some unintended effects on water-flow dynamics in the region. To serve the increasing need for fresh water in order to support human population growth in southeast Florida, construction began to enclose the remaining Everglades into the WCAs, beginning in the 1950s. Although effective in conserving water during dry periods, the levees introduced a discontinuity in the otherwise relatively flat water slope, by maintaining a relatively high water level on the up-gradient side of the levee and a relatively low water level on the down-gradient side. The resulting “stair-step” in water levels across levees locally increased vertical hydraulic gradients and caused recharge (on the up-gradient side) and discharge (on the down-gradient side) that brought surface water into contact with relatively deep groundwater (Harvey et al. 2002). It has also been observed that the high-capacity pumps used to move water through canals also have the unintended effect of increasing vertical hydraulic gradients in groundwater, which locally influences water exchange between canals and underlying groundwater and causes discharge of relatively deep groundwater (Miller 1978; Krupa et al. 2002).
Increasing importance of groundwater in Everglades hydrologic budgets
Because the Everglades often experiences prolonged dry seasons and multi-year droughts, freshwater can sometimes be a limited resource in this ecosystem. The thin lens of freshwater at the top of the aquifer (that lies above older groundwater with significantly higher ionic strength) actively exchanges with surface water on a timescale of decades to centuries (Harvey et al. 2006). This freshwater not only plays a role in sustaining the Everglades ecosystem during droughts, but the presence of different water types offers clues about the changing role of groundwater over a century of water management in the Everglades.
Appreciation of groundwater’s role in Everglades hydrologic budgets increased greatly after early efforts to achieve flood control were expanded to include water conservation. Much of what has been learned about groundwater and surface-water interactions in the Everglades has its basis in the SFWMM, i.e. South Florida Water Management Model (SFWMD 1999; 2003). The SFWMM has been a primary management tool for water-management and restoration planning (US Army Corps of Engineers 1999). The SFWMM also served as the basis of the Natural Systems Model (NSM), a daughter model that had the purpose of hindcasting a typical water budget in the pre-drainage Everglades (SFWMD 2006). The NSM does not actually attempt to simulate a specific time period, and instead operates with the standard 31-year input series of precipitation data and evapotranspiration functions used for what is referred to as the 1995 base run of the SFWMM (SFWMD 1999, 2003). The difference is that the NSM uses the best available estimates of pre-drainage topography, vegetation distribution, vegetative flow-resistance, and tidal and inflow boundaries. Therefore, NSM and SFWMM outputs can be directly compared to illustrate differences between the current managed system and the pre-drainage system under identically modeled climatic conditions.
The boundaries defined for the pre-drainage Everglades are the historic areas of contiguous emergent wetlands lying between Lake Okeechobee to the north and Everglades National Park to the south (SFWMD 2006). The boundaries for the present-day central Everglades are the same as the WCAs that collectively make up the northern and central portion of the managed Everglades system (Fig. 1). These boundaries are formally defined by the SFWMD (1999, 2006). The major difference in the pre-drainage Everglades system is the inclusion of the northernmost areas of former wetlands that were later converted to agricultural lands known as the Everglades Agricultural Area (EAA) (Fig. 1). The southern boundary of this comparison is the present-day southern boundary of WCA-3, which is located where Highway 41–Tamiami Trail crosses the Everglades. Everglades National Park water budgets are therefore not a part of the analysis.
In this study, the model results were used to identify changes in the relative importance of various water sources over the past century. Given the spatially and temporally averaged nature of the hydrologic budgets, the following results should be interpreted with caution. The relative contributions from different water sources are expected to vary considerably between specific locations in the Everglades and also among years.
Sources of water inflow and output patterns and their relative importance in the North–Central Evergladesa
Present-day managed Everglades
Model NSM v. 4.5
Model SFWMM v. 3.5
107m3 per year
% of total inputs
107m3 per year
% of total inputs
Net groundwater discharge
Net groundwater recharge
Changes over time in Everglades water budgets are summarized here. The importance of precipitation as a water source decreased (from 81 to 66%) in the managed Everglades while at the same time the surface-water inflows increased from 18 to 33%. Canal drainage from agricultural areas that were formerly wetlands (the EAA) increased the most (0 to 18%), marginal inflows stayed about the same (10 and 11% respectively), and direct inputs from Lake Okeechobee decreased slightly (from 8 to 4%). The canal drainage waters are themselves derived from a mixture of runoff of precipitation and Lake Okeechobee water used for irrigation, and possibly also from groundwater beneath the EAA that discharged to canals. The other significant change in the water budget was a substantial increase (from 1 to 16% of total inputs) in net recharge of Everglades groundwater that then flows to areas outside the Everglades. For the purpose of clarification, it is important to note that the raw precipitation and evapotranspiration (ET) values differed between the NSM and SFWMM models only because the land area involved changed.
Increasing chemical inputs to the central Everglades
Hydrologic budgets indicated a relative increase in contributions from surface-water inflows over the past century, especially from canals draining the EAA, and a relative decrease in contributions from precipitation. Surface-water inflows are inherently higher in their ionic strength than precipitation, and human activities directly and indirectly increased those mineral and nutrient inputs further with consequent effects on the biological communities of the Everglades. An evaluation of present-day chemical data in the Everglades does not necessarily reveal how hydrologic alterations affected the Everglades. For this reason, the study evaluated both groundwater and surface-water chemistry in order to take advantage of groundwater’s tendency to average chemical signatures over time due to the relatively long storage times (decades to centuries) in shallow groundwater (Harvey et al. 2006). Here, necessary background information about chemical conditions in the Everglades are briefly reviewed, based on chemical sampling and analyses by Parker et al. (1955), Harvey et al. (2002), (2005), and data obtained from the electronically accessible database of the SFWMD (SFWMD 2008). Following that, water stable isotope and ionic tracers were used to assess the relative importance of the various sources to increasing mineral inputs to the Everglades.
Mineral chemistry of peat
Historical reconstructions of the Everglades have primarily been based in paleoecological analyses of peat. These approaches are generally best at identifying century to millennial timescale trends. Everglades peat began forming about 5,000 years ago on top of limestone bedrock in a shallow trough across south Florida. Like most other peatlands, the Everglades is relatively poor in nutrients and minerals. Although limestone dissolution influenced water chemistry in the early stages of peat development, there is evidence that much of the northern and central Everglades was developing towards an increasingly mineral-poor state as peat built up in thickness during the shift towards a wetter climate in south Florida (Gleason et al. 1974; Gleason and Stone 1994). Although increasing thickness of peat is not alone enough to isolate surface water from interactions with groundwater that is high in ionic strength (Siegel and Glaser 1987), it does contribute to isolation in a landscape as flat as the Everglades. Another indication of increasing isolation of surface waters from mineral sources is the “water lily-dominated peat” found in the central and northeastern Everglades, which has only minor amounts of marl, i.e. intervening layers of calcitic mud with moderate to high mineral content, that is more commonly associated with the thin, and less frequently flooded peat found in the marl prairies in the southern Everglades (Gleason and Stone 1994; Slate and Stevenson 2000; Winkler et al. 2001).
Wetland surface waters
The northeastern apex of the historic Everglades is now a major part of the area occupied by WCA-1 and WCA-2A. Because of their position, marginal inflows of surface water may have exerted a greater influence on WCA-1 and WCA-2 compared with areas farther to the west (i.e. now occupied by the EAA and WCA-3A), where overflow from Lake Okeechobee was probably the more important surface-water inflow. Under present management, WCA-2A sometimes receives the majority of its water from canal drainage delivering mixtures of EAA farm canal runoff and Lake Okeechobee water. In contrast, WCA-1 is mostly isolated from canal waters and from marginal inflows due to complete enclosure by levees and without any flow-through pumping. For that reason, the interior of WCA-1 and WCA-2A currently represent the extremes of water chemistry of wetland surface waters in the northern and central Everglades. WCA-1 surface water is a sodium calcium–chloride bicarbonate water type and WCA-2A is a mixed cation–chloride bicarbonate water type. The chemistry of WCA-1 surface water indicates an initially significant contribution from seawater aerosols in precipitation that was further altered by addition of limestone dissolution products due to interactions with shallow groundwater. WCA-2A has a considerably higher specific conductance (with a median 946 μS/cm in WCA-2A compared to 92 μS/cm in WCA-1). WCA-2A also has greater relative contributions from Mg2+ to cations and SO42– to anions, reflecting inputs from EAA drainage canals and Lake Okeechobee (Fig. 3).
Water inputs from Lake Okeechobee
Historically, Lake Okeechobee overflowed each wet season and provided a significant source of water to the pre-drainage Everglades (Chris McVoy, SFWMD, personal communication 2006; Table 1). A dike now prevents overflows, with a system of levees and canals to control deliveries of lake water to the Everglades. EAA canals route lake water primarily through the S6 water control structure to be discharged into WCA-2A. The lake has a specific conductance of approximately 400–500 μS/cm, and relatively even proportions of Ca2+ and Na+ , with some Mg2+ , and relatively even proportion of HCO3– and Cl–, with some SO42−, resulting in a classification as a mixed cation–bicarbonate chloride water type (Fig. 3). The lake’s chemistry is primarily influenced by runoff from the Kissimmee River basin and secondarily by interactions with lakebed sediments and groundwater. Groundwater in the lake’s vicinity has higher proportional contributions from Ca2+ and Mg2+ and lower contributions from Cl− and SO42− to its dissolved solids compared to Lake surface water. Back-pumping from EAA farm canals to the lake may have influenced the lake’s chemistry during certain time periods due to agricultural influences, but is not expected to dominate Lake chemistry.
Marginal runoff from infrequently flooded wetlands and low-lying uplands
The chemical composition of drainage from infrequently flooded wetlands or low-lying uplands such as pine flatwoods to the pre-drainage Everglades is uncertain. In the northern and central areas of the Everglades these marginal areas have been altered so substantially in their topography and hydrology that finding a modern analog is difficult. It is instructive to look at the southern Everglades where there are large areas of short hydroperiod wetlands on the margins of the major sloughs in Everglades National Park. The relatively short hydroperiods in those areas promote infiltration of precipitation into the peat and bedrock (Price and Swart 2006). The wetlands also have thinner peat accumulations (Gleason et al. 1974) which offers less resistance to subsurface vertical flow. Surface-water specific conductance in the most mineral-rich freshwater areas of Everglades National Park rarely exceeds 400–600 μS/cm (Joffre Castro, Everglades National Park, personal communication, 2007). The relatively high transmissivity of the Biscayne aquifer beneath the southern Everglades (Fish and Stewart 1991; Renken et al. 2005) and the thinner peats in the southern Everglades promote significant flow through marginal wetlands into the sloughs of the southern Everglades (Price et al. 2003; Harvey et al. 2006). The chemistry of shallow groundwaters in short hydroperiod wetlands (Price and Swart 2006) is similar to the subset of eastern canal waters in the northern Everglades that have lower specific conductance, i.e. calcium sodium-bicarbonate waters. Short hydroperiod wetlands in Everglades National Park are therefore one of the potential models for chemistry of marginal runoff entering the pre-drainage Everglades.
Deeper groundwater in the western part of the Biscayne aquifer beneath WCA-2A has a specific conductance often above 10,000 μS/cm and as high as 26,000 μS/cm and is a sodium chloride water type, reflecting its origin as relict seawater trapped in the aquifer during higher sea level stands (Howie 1987; Harvey et al. 2002; Price and Swart 2006). The origin of the shallow groundwater is very different. Groundwater in the top 10 m of the western Biscayne aquifer typically is a calcium sodium-bicarbonate type water with a specific conductance of about 1,250 μS/cm and a chemical and isotopic composition indicating a source from recharge of evaporated Everglades surface water that slowly acquired a predominantly Ca2+, Mg2+, and HCO3− signature due to dissolution of shallow aquifer materials. Sulfate reduction removed SO42− from the shallow groundwater (Fig. 3). Depending on depth and location relative to water management structures and levees, the ionic strength of shallow groundwater may be elevated by Na+ and Cl− due to upward discharge of deeper groundwater. Near canals and levees, shallow groundwater, peat pore water, and even canal water and Everglades surface water can have a chemical signature that traces the upward movement and discharge of relict seawater (Miller 1988; Harvey et al. 2002; Krupa et al. 2002)
Groundwater composition beneath WCA-2A transitions from the western Biscayne aquifer, with relict seawater in its lower half, to a different (unnamed) surficial aquifer beneath the EAA and Lake Okeechobee which contains a mixed cation–bicarbonate groundwater at depth. The deeper groundwater in both aquifers is relatively high in ionic strength, but the proportions of Ca2+, Mg2+, SO42−, and Cl− differ (Fig. 3). Groundwater in the lower Biscayne aquifer beneath WCA-2A is either a sodium–chloride or sodium calcium–chloride bicarbonate type water. Deeper groundwater beneath the EAA/Lake is distinctive in having higher proportional contributions from Ca2+ and Mg2+, and lower proportional contributions from Cl −, compared to Biscayne aquifer water (Fig. 3). Whereas groundwater in the Biscayne aquifer beneath WCA-2A was derived both from freshwater recharge of Everglades surface water and trapping of relict seawater at depth, the groundwater beneath the EAA/Lake was derived only from recharge of freshwater from the Kissimmeee River basin. Fresh groundwater beneath the EAA and Lake differs in chemistry due to longer contact time with different aquifer materials.
Groundwater samples beneath Stormwater Treatment Area 1W (STA-1W) represent a transition between the Biscayne aquifer and the EAA/Lake Okeechobee surficial aquifer. STA-1W is located 40 km north of the center of WCA-2A, between WCA-1 to the east and the EAA to the west. As a result of its intermediate location, STA-1W groundwater chemical samples either represent relict seawater, sodium chloride–bicarbonate type waters characteristic of shallow groundwater in the Biscayne aquifer, or mixed cation–bicarbonate type waters found in the surficial aquifer beneath the EAA and the lake (Fig. 3).
Frazee (1982) suggested but offered no direct evidence that transitional connate groundwater with a relatively high proportional contribution from SO42− is present in the vicinity of the lake. The authors’ search of published data revealed no evidence of such a groundwater within the study boundaries, although shallow groundwater in the EAA, while similar to the deeper groundwater in most respects, does have a greater proportional contribution from SO42− compared with Cl−. Sulfur budgets in the EAA have been influenced by agricultural activities for many decades. Excess SO42− in EAA groundwater appears to have resulted from agricultural practices which produced a shallow groundwater that could be mistaken for a transitional connate groundwater (Fig. 3). This interpretation is supported in the next section and in a section further ahead that uses environmental solute tracers to distinguish water sources.
Canal inflows to the Everglades
Canal waters are an important source of water and dissolved materials to the managed Everglades ecosystem. The most significant sources are the main conveyance canals flowing into the Everglades (the West Palm Beach, Hillsboro, North New River, and Miami canals shown in Fig. 1). These canals convey runoff of rainfall, soil water, and groundwater discharge from agricultural fields in the EAA as well as flow-through waters from Lake Okeechobee into the northern and central Everglades.
The influence of canals as inputs of waters with high ionic strength is evident from spatial maps of surface-water specific conductance in the Everglades wetlands generated from the Regional Environmental Monitoring and Assessment Program (REMAP) conducted by the US Environmental Protection Agency (Stober et al. 1998). The median specific conductance of surface waters in the main canals ranges between 700–1,100 μS/cm, more than 50-fold higher than that of rainfall and twice that of current levels in Lake Okeechobee, which also is exposed to elevated mineral loads from human sources. Areas strongly influenced by canal inputs have a specific conductance near 1,000 μS/cm (e.g. WCA-2A), while those with a more rainfall-driven hydrology typically have values as low as 100 μS/cm or less (e.g. WCA-1). Specific conductance decreases from north to south across the Everglades canal network. This trend likely is caused by progressive dilution with rainwater and surface waters in the WCAs with increasing distance from the mineral sources in the north.
The chemical composition of canal water draining the EAA varies at water control structures that convey the water into the Everglades. The westernmost canals passing through the S7 and S8 structures generally contain only EAA drainage waters, while canals passing through S6 (and sometimes S5A) convey lake water in addition to farm canal drainage. The S5A structure receives the broadest mix of water types, including canal drainage from the EAA, Lake Okeechobee water, and drainage from the low-lying basins to the east of Lake Okeechobee. For this reason, the S5A canal may at certain times be the best analog of what the chemistry of low-lying marginal areas was like in the pre-drainage central Everglades. The specific conductance in S5A ranges between 400 and 1,400 μS/cm; samples with conductance less than 900 μS/cm can be classified as calcium bicarbonate or calcium–sodium bicarbonate water that is relatively poor in sulfate and possibly representative of pre-drainage marginal inflows (Harvey et al. 2002). The samples with specific conductance greater than 900 μS/cm have proportionately greater importance of magnesium and sulfate that probably reflect large contributions from EAA canals at those times.
Compared with other water sources to the Everglades, canal water has substantially higher contributions of Mg2+, SO42−, Ca2+, bicarbonate, and lower Na+ and Cl− compared to Everglades surface water or groundwater, or Lake Okeechobee surface waters (Fig. 3). EAA canal drainage water tends to have too great of a contribution from SO42−, Ca2+ and too little of a contribution from Cl− to be explained by deeper EAA/Lake Okeechobee groundwater or STA-1W groundwater. Chemical sampling in the 1940s indicates that shallow EAA groundwater also was rich in SO42− and Ca2+ and similar to the chemistry of EAA farm canals while also being very different from Lake Okeechobee surface water and all other groundwaters in the vicinity (Fig. 3). The chemical uniqueness of EAA canal drainage waters has been noted before and attributed to flushing of peat oxidation products (Schueneman 2001) and flushing of fertilizer additives (Bates et al. 2002) from farm fields. Those agricultural influences also could have affected the shallow EAA groundwaters directly beneath the farm fields through recharge of peat oxidation products and agricultural fertilizer additives through farm fields into shallow groundwater. The irrigation water and precipitation that recharges farm fields would move into shallow groundwater and eventually discharge into drainage canals that flow to the Everglades. The next section discusses the use of environmental solute tracers to distinguish the relative importance of these various sources of water to the Everglades.
Use of environmental solute tracers to delineate water sources
This section combines chemical signatures with the hydrologic budgets already presented to determine changing water sources and chemical inputs to the Everglades on a decadal to centennial timescale. Environmental solute tracers are used as a means to isolate the effects that water management has had on changing chemistry and ecology.
Water stable isotopes
Of the various surface waters, wetland surface water in the Everglades is the most variable in composition, ranging between values close to precipitation during the wet season to much more positive values during the dry season. The data illustrate the importance of direct precipitation on the Everglades with more negative water-stable isotope compositions reflecting the primary input during the wet season. Evaporation also strongly influences isotopic composition during much of the year, enabling differentiation between evaporated surface water and groundwater as the source of shallow groundwater. The relatively positive isotopic values of WCA-2A surface waters during the dry season are consistent with evaporation effects, but also may reflect mixing with other evaporated surface waters (e.g. Lake Okeechobee or canal drainage from the EAA).
Canal waters from the EAA passing through the S6 structure are similar in enrichment to Lake Okeechobee, with other canals being slightly more negative in isotopic composition (Fig. 4). The more positive isotopic composition of S6 canal waters is consistent with routing of Lake Okeechobee water primarily through the S6 structure. Lower isotopic values in the S5A and other canals probably reflect additional components of runoff generated by flushing of rainfall through EAA soils and shallow groundwater to the larger EAA canals that discharge directly into the Everglades. Unfortunately, there are no direct measurements of EAA soil water or shallow or deep groundwater under most of the EAA. Groundwater beneath STA-1W, which is located on the margin between the EAA and the WCAs, is the only water stable isotope data available that may be informative about hydrologic differences between pre- and post-drainage groundwater hydrology in the EAA.
Most STA-1W groundwater is relatively enriched, however a smaller subset of the shallow STA-1W groundwater samples located on its western side have more negative isotopic values that are much closer to precipitation (Fig. 4). Similar values of water stable isotope ratios close to precipitation values also have been observed in shallow groundwater beneath short hydroperiod wetlands of the Rocky Glades in the southern Everglades by Price and Swart (2006). The more negative isotopic composition of those waters is consistent with recharge of precipitation occurring quickly and without time for significant evaporation to occur (Harvey et al. 2002; Price and Swart 2006). In STA-1W those values reflect the history of the area as part of the EAA, where drainage and farming in the middle part of the century promoted direct infiltration of precipitation into soils until the mid 1990s when the site was re-flooded for operation as a stormwater treatment area.
The WCA-1 and WCA-2B basins were isolated from surface-water throughputs and only received water input by precipitation during the time period of sampling. The observation that both of the isolated basins vary just as broadly in their water stable isotope ratio compared with WCA-2A is significant, because WCA-2A receives isotopically enriched inputs from EAA canal drainage that includes Lake Okeechobee water. Evaporation effects on the isotopic chemistry of Everglades surface water therefore have comparable effects to mixing of source waters, i.e. mixing of precipitation with Lake Okeechobee and canal drainage waters. Discharge of deeper groundwater from the western Biscayne aquifer beneath WCA-1 and WCA-2 also has more positive water stable isotope composition, somewhat fortuitously, because seawater trapped in the Biscayne aquifer in south Florida plots on almost exactly the same line as evaporated waters (Price and Swart 2006) (Fig. 4). Because of the similarity of the mixing line along which many of the different waters plot (e.g., canals waters draining the EAA, Lake Okeechobee water, relict seawaters, and evaporated precipitation), water stable isotopes cannot delineate all potential mixing effects. Additional environmental tracers are needed to better delineate the contribution of groundwater from beneath the EAA to EAA canal drainage that flows into the Everglades.
Sulfate to chloride ratios have been used previously to help distinguish sources of Everglades waters (Bates et al. 2002; Chen et al. 2006). In particular, WCA-2A surface waters have a ratio of sulfate to chloride that is very similar to the canals draining the EAA and much higher than in the relict seawater at depth in the western Biscayne aquifer. The ratio of sulfate to chloride is also much higher than in shallow groundwater beneath WCA-2A, whose source is surface water from the Everglades that was recharged decades to centuries ago (Harvey et al. 2006). The very low ratio of sulfate to chloride in shallow groundwater beneath WCA-2A is also consistent with sulfate removal by sulfate reduction (Bates et al. 2002). Lake Okeechobee waters also have a high sulfate to chloride ratio, although Orem (2007) noted that Lake Okeechobee water is too dilute in these ions to be the major water source to WCA-2A surface water during most times of year. These results are strong indicators that EAA canal waters are a significant source of WCA-2A surface water, comparable in importance to precipitation, and the most likely source of increased sulfate and increased mineral inputs in general to the central Everglades. Left somewhat in question is the ultimate source of the increased mineral loads in canals draining the EAA.
Both the products of peat oxidation (Schueneman 2001) and fertilizer additives (Zielinski et al. 2000; Bates et al. 2002; Chen et al. 2006) have been investigated as indicators of the water and chemical sources in canals draining the EAA. The importance of fertilizer and fertilizer additives was established through use of uranium isotope ratio tracers that identified EAA fertilizer as the source of elevated P and sulfate to EAA canals as well as in Everglades surface waters (Zielinski et al. 2000). In a study using sulfur isotopes as an environmental tracer of water flow, Bates et al. (2002) and Orem (2007) found that sulfate in Everglades surface waters and EAA canal waters had a 34S/32S ratio similar to fertilizer. The sulfur isotopic ratio and sulfate concentration measured in shallow groundwater beneath WCA-2A are too high and too low, respectively, to explain the source of sulfate in EAA canal water (Bates et al. 2002). Unfortunately, there were no isotope ratio measurements made in groundwater beneath the EAA to help explain the source of excess sulfur to EAA canals. Instead, one must rely on major ion ratios to delineate the importance of groundwater discharge from directly beneath the EAA to canals entering the Everglades.
Gypsum is applied to southeastern US agricultural soils, including EAA soils, to improve structure and fertility. Schueneman (2001) recently quantified the use of sulfur additives in EAA fertilizer and suggested that current sulfur application rates on farm fields are probably low relative to oxidation of naturally occurring sulfur in EAA peat. Sulfate leaching resulting from oxidation of naturally occurring sulfur in peat soils has probably been occurring since the early 1910s, and may be a significant contributor to the high concentrations of sulfate that have been observed in EAA canals (Orem 2007). Nonetheless, Fig. 5 demonstrates a substantial source of SO42− to EAA surface water and shallow groundwater in the 1940s that cannot be explained by any source except dissolution of gypsum, and gypsum is only one possible source of SO42− in fertilizer additives. The other is oxidation of the elemental sulfur, which is another common additive in fertilizer used in the EAA (Bates et al. 2002). Schueneman (2001) states that current applications of sulfur to EAA farm fields are in quantities less than recommended amounts. Figure 5 does suggest that the effects of gypsum dissolution on EAA canal chemistry are smaller today than in the 1940s. It may be that gypsum applications were greater in the early decades of EAA farming, when it would have had its greatest value to increase soil structure, oxidation, and fertility. It is possible that more recent applications of sulfur may be in elemental sulfur form (Bates et al. 2002; Orem 2007). Additional research would be needed to distinguish the relative importance of the various sources with confidence. In particular, more chemical sampling is needed in EAA shallow groundwater to increase certainty in present-day sulfur budgets for the EAA and the Everglades.
The foregoing discussion and accompanying figures demonstrate that increased mineralization of EAA surface water was already well advanced in the 1940s. Zielinski et al. (2000) and Bates et al. (2002) found by using uranium isotopes and sulfur isotopes respectively as tracers, that a significant proportion of new mineral sources (sources of phosphorus and sulfur specifically) were derived from agricultural activities in the EAA. With drainage and conversion of wetlands to agriculture beginning early in the 1910s, prior to the availability of reliable chemical measurements, the pre-drainage water chemistry in the northern part of the Everglades now occupied by the EAA remains speculative. The best approximation that can be made is a mixed cation–bicarbonate type water whose current analogue is the shallow groundwater beneath STA-1W and also Lake Okeechobee itself (Fig. 3).
Changing groundwater and surface-water interactions and effects on chemical transport
It is important to indicate that water quality concerns are not just for the ecosystem. Municipal water budgets for the lower east coast of Florida indicate that recharge of central Everglades water into the Biscayne aquifer and eastward movement toward domestic well fields is an important source of drinking water. There is still not enough understanding of the source areas, flow paths, and travel times required for Everglades surface water to reach domestic water-supply well fields (Sonenshein 2001; Nemeth and Solo-Gabriele 2001)
Role of changing hydrologic and chemical sources on Everglades ecology
An increased proportion of surface-water inflows relative to rainfall and increased mineral inputs are some of many processes driving changes in the structure and function of the Everglades. The subtle biological and geochemical dependencies that are being altered by increased mineral inputs have not been investigated to the same extent as nutrient pollution (McCormick et al. 1998), increased variability in water levels and incidences of peat-burning fires, and processes affected by the velocity of sheetflow in the Everglades (Larsen et al. 2007). Here, some effects of increased mineral inputs on biological communities of the Everglades are examined.
There is increasing evidence that mineral and phosphorus loading from canals may directly and indirectly affect the bioavailability of many different elements in Everglades peat soils. Measurements of porewater chemistry in the 2–12 cm depth increment of WCA-1 peat showed higher pore-water pH and higher concentrations of dissolved P and N as well as Mg2+, Ca2+, K+, Cl−, S2−, and sulfate at sampling sites near the edges of WCA-1 (where mixing occurs with canals) compared to more rainfall driven sites in the interior (Susan Newman, SFWMD, personal communication, 2007). Porewater redox levels and dissolved Fe concentrations were lowest at the sites receiving canal water, apparently due to fertilization of higher metabolic activity and formation of insoluble FeS under low redox conditions.
Organic matter decomposition is a key process controlling both soil formation and nutrient cycling in peatlands. Collection and incubation of equivalent litter of several dominant plant species, sawgrass and cattail, indicated decomposition rates that were up to 30% faster at wetland sites in WCA-2A at a location that receives canal inflows with high ionic strength, compared with sites in WCA-1 where little canal water intrudes (Newman et al. 2004). Increased mineral loading may be influencing decomposition rates through: (1) increased availability of electron acceptors such as sulfate that are used in anaerobic microbial respiration; (2) increased release of bound phosphorus from peat that fuels higher decomposition rates; and (3) increased activity of minerals such as Ca2+ serving as a co-factors regulating enzyme activity. While not suggesting any specific or well-established causal mechanisms, these findings indicate empirically that intrusion of high ionic strength canal water is promoting faster rates of decomposition.
Sulfate is a significant component of canal water, and has been identified as one of the most widespread contaminants in the Everglades (Bates et al. 2002; Orem 2007). Special importance is attached to sulfate in the Everglades because of its effects on a variety of processes including cycling and bioavailability of Hg, a contaminant that enters the Everglades via atmospheric deposition and is converted to its bioavailable form (methyl-Hg) primarily through microbial pathways (Benoit et al. 2003). Elevated sulfate concentrations also can affect redox potential of Everglades water which can influence specific pathways and rates of decomposition, and can also can affect vegetation patterns due to the inhibitory effects of hydrogen sulfide—an end product of sulfate reduction—on plant growth.
Responses of Everglades vegetation
The Everglades contains many plant species with the most common being sawgrass (Cladium jamaicense), fragrant water lily (Nymphaea odorata), bladderwort (Utricularia spp.), and spikerush (Eleocharis spp.). Few studies in the Everglades have examined spatial patterns of vegetation in relation to water chemistry, although two previous studies of related species of emergent plants in other wetlands found them to be indifferent to surface-water ionic strength (Moyle 1945; Walker and Coupland 1968). In his analysis of aquatic vegetation patterns related to specific conductance, Moyle (1945) noted that “The natural separation between hard and soft waters seems to be at a total alkalinity of about 40 mg/L, 30 mg/L being the lower limit of toleration of the more typical hard-water species, and 50 mg/L the upper limit of toleration of the more characteristic soft-water species” (p. 404). While current information from the Everglades is insufficient to indicate a clear relationship between increasing mineral levels and changes in emergent plant communities, preliminary data by McCormick (SFWMD, unpublished data, 2007) are showing strong relationships between species composition and distance from canals that correlate with concentrations of total dissolved solids in WCA-1. However, the observed species distributions and dissolved chemical gradients are correlated with several other environmental changes and, thus, do not provide conclusive proof that increased mineral inputs are the cause of observed vegetation shifts. For example, canal waters had the highest P concentrations and soil P concentrations were also highest at sites closest to the canal, and this factor alone can cause pronounced shifts in Everglades vegetation (McCormick et al. 2001). Thus, soil mineral gradients caused by canal-water intrusion are partially confounded by a limiting-nutrient gradient. Water depths also tend to be more variable and greater on average near the canal perimeters of WCA-1. Deeper water depths are another key factor that is known to influence plant species composition in the Everglades.
Response of Everglades periphyton and fauna
Floating and attached periphyton mats in the Everglades (composed of algae, bacteria, and microfauna) are an important base of the food web providing a food source and habitat for aquatic invertebrates (Williams and Trexler 2006). Periphyton also is important in structuring the ecosystem by reducing phosphorus availability (Noe et al. 2001), blocking light penetration and often dominating primary productivity (Noe et al. 2001), and oxygenating the water column (McCormick and Laing 2003). Periphyton communities are extremely sensitive to changes in water chemistry, which explains their widespread use as indicators of water quality in the Everglades and other aquatic ecosystems.
The interior of WCA-1, which is relatively protected from canal water inputs, contains a characteristic periphyton community dominated by desmid and diatom species indicative of soft-water conditions. Whereas periphyton mats across the high ionic strength portions of the managed Everglades are dominated by calcium-precipitating (calcareous) cyanobacteria and have a high calcium carbonate content, those in WCA-1 are largely organic (non-calcareous) in nature. Paleoecological evidence (Slate and Stevenson 2000) indicates that the community currently found in WCA-1 was probably more widespread across the pre-drainage Everglades when rainfall was a more dominant source of water to this ecosystem. By contrast, calcareous communities historically were more abundant in the marl prairies of the southern Everglades, which support little or no peat accretion due to their short hydroperiods and thus have a water chemistry influenced more strongly by the limestone bedrock.
Summary and conclusions
Reconstructions of historic water budgets indicate that direct rainfall was the primary source of water to much of the Everglades prior to start of drainage efforts in the early twentieth century. Seasonal overflows from Lake Okeechobee and runoff from short hydroperiod wetlands and low-lying uplands along the Everglades eastern margin provided most of the additional water to the pre-drainage ecosystem. The mineral content of these other water sources was likely higher than that of rainfall but lower than that of surface flows entering the Everglades today. Mineral inputs were low in the pre-drainage Everglades with Lake Okeechobee overflows probably being the primary source with secondary sources from marginal inflows and groundwater discharge. The source of minerals from groundwater has probably increased in the present-day system due to topographic changes resulting from subsidence or removal of the hydraulically resistive peat accumulations within and near canals, and water management practices that increase vertical hydraulic gradients. However, the increased load of minerals entering the Everglades today cannot be explained solely by increases in discharge of groundwater with high ionic strength.
Historic and recent water budgets and environmental tracers were used to assess the relative importance of new sources of minerals, which suggested that canal drainage from the EAA has been the dominant source of increased mineral inputs. Surface-water inflows now account for 33% of water inputs to the Everglades compared to 18% in the pre-drainage Everglades. Canal drainage from the EAA, an area that was formerly Everglades and is now farmed, is the largest part of that increase, accounting for more than half of all present-day surface-water inputs. The canal water also has mineral concentrations more than 50-fold higher than those of rainfall and as much as 10-fold higher than those of surface waters in rainfall-fed areas of the Everglades interior. The ultimate source of water to agricultural canals appears to be predominantly irrigation water pumped from Lake Okeechobee, and secondarily a mixture of direct rainfall onto EAA agricultural fields and discharge of shallow groundwater from beneath the EAA that together flush fertilizer constituents and additives, including SO42− and Ca2+ from dissolution of gypsum, a common fertilizer additive in the EAA and elsewhere. Also important is the flushing of natural sulfur in EAA peat soils that have been managed for many decades for agricultural production, and also the chemical conversion and flushing of additional fertilizer additives such as phosphorus and elemental sulfur that are eventually transported by runoff or through shallow groundwater into EAA canals that drain to the Everglades.
Increased mineral inputs to the Everglades are correlated with changes in plant and periphyton species composition and related landscape features such as the cover of slough and sawgrass habitat (Larsen et al. 2007), although specific cause–effect linkages have only begun to be assessed. Recent experiments have demonstrated that fundamental changes occur in periphyton assemblages due to mineral enrichment, with consequences for secondary productivity in the Everglades ecosystem. Alteration of surface-water chemical conditions and its effects on periphyton species assemblages may have cascading effects at higher trophic levels by altering conditions for growth, reproduction and survival of fish and aquatic invertebrates.
Among the major ions and other dissolved constituents, elevated concentrations of phosphorus and sulfate are probably the most significant threat to Everglades ecosystem structure and function. Increased loading of these constituents elicits multiple ecological responses including shifts in macrophyte and periphyton species composition due to higher nutrient availability and production of hydrogen sulfide, which is toxic at high concentrations, as well as increased productivity and decomposition rates, lowered redox potential, increased mercury bioavailability and toxicity. Because of long storage times of contaminated surface waters that are exchanged with shallow groundwater, there are likely to be legacy effects that will prolong the influence of contamination for many decades even after the quality of canal water inputs is improved.
Groundwater’s role as a freshwater storage zone that sustains the ecosystem is not yet fully understood. What has become clear in the past several decades is that interactions between groundwater and surface water have increased as a result of water management, resulting in reduced storage of fresh, uncontaminated water in the shallow aquifer located directly beneath the Everglades and also beneath basins such as the EAA that discharge directly into the Everglades. The contamination affecting shallow Everglades groundwater comes both from above and below. Recharge from above is increasingly contaminating shallow groundwater with nutrients, sulfate, mercury, and other contaminants, while the increased vertical hydraulic gradients have contributed to upward transport of salts from the deeper aquifer. Needs for future research and modeling may include: (1) large-scale tracer experiments to investigate fate and transport of contaminants introduced with canal waters, (2) investigation of specific transport and reaction processes for various elements, including information about storage timescales in various compartments such as Everglades soils, vegetation, and groundwater; (3) studies that fill the many data gaps on the ecological effects of mineral enrichment; and (4) improved hydrologic measurements and models to identify strategies to achieve more complete mixing of canal waters at the northernmost possible points of entry to the Everglades, in order to minimize long distance transport of undiluted canal waters into the Everglades.
Hydrologic restoration of the Everglades toward the higher flows characteristic of pre-drainage conditions remains a centerpiece of restoration efforts. However, if greater reliance on canal drainage waters is needed to achieve those hydrologic goals, unintended effects may result that could deliver contaminants such as phosphorus and sulfate farther into the Everglades than ever before. This paper reviewed and synthesized a variety of research projects that collectively suggest that the increasing mineral content of water sources to the Everglades have already altered the structure and function of the Everglades ecosystem, beginning with periphyton communities and food webs, and extending to changes in the dominant species of macrophytes, with cascading effects on fish and invertebrates.
The investigation was supported by the US Geological Survey’s Greater Everglades Priority Ecosystem Sciences Initiative, and by the National Research Program of USGS. We thank Lauren McPhillips for assistance with the acquisition and interpretation of chemical data. We are grateful to colleagues Bill Orem, Greg Noe, Laurel Larsen, Ronnie Best, and Aaron Higer of the USGS, Chris McVoy of the South Florida Water Management District, Hydrogeology Journal Guest Editor Andrew Boulton, and two anonymous journal reviewers for comments on the draft, which improved the final manuscript.