Abstract Introduction Methods Results >Discussion Conclusions Acknowledgments Literature Cited Figures and Tables PDF Version

Our study is the first to quantify diurnal variation in community-level calcification rates in Florida Bay. Over 24-h time periods, TA (from which we calculated rates of calcification) ranged from 0.026 to 0.212 mmol kg^-1 for all data sets. Millero et al. (2001) and Boyer et al. (1999) reported seasonal nitrite, nitrate, and total phosphate concentrations for Florida Bay ranging from approximately 0 to 4.4 µmol kg^-1 with the lowest concentrations occurring in central and western Florida Bay. Average diurnal TA of 0.098 mmol kg^-1 (or 98 µmol kg^-1) observed in our study is 22 times greater than total maximum nutrient concentrations. Salinity did not vary within the incubation chamber during collection of individual 24-h data sets in basins, so it is unlikely that diurnal TA at our basin study sites resulted from variation in nutrient concentrations or salinity, and can be attributed to carbonate sediment production and dissolution.

Based on calculations of TA with varying salinity performed using the carbonate speciation program CO2SYS (Lewis and Wallace 1998), we estimate that TA changes by 0.006 mmol kg^-1 for every one unit of salinity change. TA calculations were made over a salinity range of 25.0 to 41.2 in CO2SYS using dissociation constants K1 and K2 from Merbach et al. (1973) refit by Dickson and Millero (1987), KSO₄ from Dickson (1990), TCO₂ of 1.907 mmol kg^-1, and a pH of 8.044 on the free hydrogen ion scale. We collected 41 sets of upstream and downstream TA measurements at 4-h intervals along transects across the tops of mud banks. Thirty-four (83%) of these 41 data sets showed a very small change in salinity (S) from the upstream to downstream site of less than 1.0 (average S = 0.2, equivalent to a TA of 0.001 mmol kg^-1) and an average TA from upstream to downstream sites of 0.021 mmol kg^-1 . For these 34 data sets, approximately 5% of the TA may be attributed to variation in salinity. Four (10%) of the 41 data sets showed a salinity change from the upstream to downstream sampling sites of between 1.0 and 2.0 (average S = 1.5, equivalent to a TA of 0.009 mmol kg^-1) and an average TA from upstream to downstream sites of 0.024 mmol kg^-1 indicating that approximately 38% of the TA may be attributed to salinity variation in these data sets. The remaining three data sets (7%) showed a S from upstream to downstream of between 2.0 and 4.0 (average S = 3.3, equivalent to a TA of 0.020 mmol kg^-1) and an average TA of 0.031 mmol kg^-1 indicating that approximately 65% of the TA may be attributed to salinity variation in these data sets. We are confident that 95% of the TA in 83% of our mud bank data sets can be attributed to carbonate sediment production and dissolution, while 38-65% of the TA in the remaining 17% of mud bank data sets may be due to salinity variation.

TA measurements in our study (Table 2 and Table 3) are within the range of previously measured Baywide TA (Millero et al. 2001). Lowest TA values were consistently observed during summer months, while highest values were observed during winter months. This is consistent with seasonal trends in TA observed by Millero et al. (2001) and may be related to seasonal changes in evaporation and freshwater flow from the Everglades into Florida Bay. Although winter is the dry season, salinity tends to be lower than in summer because of lower evaporation rates (Nuttle et al. 2000; Swart and Price 2002), and TA is higher as more freshwater (characterized by high TA) is retained in Florida Bay (Millero et al. 2001). Millero et al. (2001) measured changes in salinity-normalized TA (NTA), representative of seasonal changes in precipitation and dissolution of CaCO₃, for March and October 1998 ranging from approximately 4.600 to 2.800 mmol kg^-1 (NTA = 1.800) for central Florida Bay near Manatee Keys and from 3.000 to 2.600 mmol kg^-1 (NTA = 0.400) for western Florida Bay near Buchanon Keys. Our observed maximum diurnal TA at constant salinity near Manatee Keys (TA = 0.212) and in Buchanon Basin (TA 0.190) is approximately 12% and 48% of the seasonal variability observed in NTA reported for these areas from March to October 1998. The significant magnitude of diurnal TA suggests that the time of day must be carefully considered when comparing spatial or temporal TA data.

Diurnal trends in TA, DO, and pH (Fig. 5) are most notable for basin sites measured inside of the incubation chamber because bank measurements are complicated by changing current directions. Diurnal changes in DO typically reflect changes in CO₂ resulting from photosynthesis as O₂ is generated and CO₂ is consumed, and respiration as O₂ is consumed and CO₂ is generated. This relation between TA, DO, and pH suggests that diurnal calcification and dissolution trends are related to diurnal cycling of CO₂ resulting from photosynthesis and respiration. Photosynthesis during the day facilitates calcification by photosynthetic organisms and consumes CO₂, reducing production of carbonic acid and dissolution of sediments. Aerobic respiration during the night generates CO₂ (and carbonic acid), facilitating dissolution that, in some cases, exceeds nighttime calcification by calcifying organisms that are not dependent on light (e.g., mollusks). Photosynthesis is typically attenuated or inhibited during complete cloud cover and turbidity events. Our observation that dissolution occurred during daylight characterized by complete cloud cover or high levels of turbidity supports the link between diurnal calcification trends and diurnal cycling of CO₂.

Previously reported annual rates of biogenic CaCO₃ production in Florida Bay have been determined primarily from standing crop surveys of calcifying organisms and their growth rates (Nelson and Ginsburg 1986; Bosence 1989a; Frankovich and Zieman 1994). Organisms contributing to carbonate sediment production include seagrass epiphytes, small Porites and Siderastrea corals (western Florida Bay), mollusks, foraminifera, calcareous green algae, and coralline algae. Annual rates of carbonate sediment production determined from standing crop and turnover methods reflect only gross production and do not account for sediment transport or dissolution.

Bosence (1989a) measured annual sediment production for basin and bank areas near upper Cross Bank (central Florida Bay) of 128 and 331 g CaCO₃ m^-2 yr^-1, respectively, and near Buchanon Keys (western Florida Bay) of 506 and 1,081 g CaCO₃ m^-2 yr^-1, respectively, using standing crop and turnover methods. Nelson and Ginsburg (1986) reported annual carbonate sediment production rates for Thalassia epiphytes (one of the primary sources of lime mud in Florida Bay) ranging from 30 to 303 CaCO₃ m^-2 yr^-1 in the Cross Bank area. Frankovich and Zieman (1994) reported minimum annual epiphyte production ranges for 7 sites within Florida Bay of 1.9-282.7 g CaCO₃ m^-2 yr^-1. These annual rates correspond to daily production rates of 0.35-0.91 g CaCO₃ m^-2 d^-1 for basins and banks near upper Cross Bank, and 1.39 and 2.96 g CaCO₃ m^-2 d^-1 near Buchanon Keys (Bosence 1989a), 0.082-0.83 g CaCO₃ m^-2 d^-1 for Thalassia epiphytes near Cross Bank (Nelson and Ginsburg 1986), and 0.005-0.77 g CaCO₃ m^-2 d^-1 from Frankovich and Zieman (1994). Our average C_day values (which exclude the effect of net dissolution we observed during the night) for banks and basins, excluding mud bottom sites, ranged from 0.066 to 1.606 g CaCO₃ m^-2 d^-1 (Table 2 and Table 3). Our measurements fall within the range of daily gross calcification values calculated from previous investigations (Nelson and Ginsburg 1986; Bosence 1989a; Frankovich and Zieman 1994). Highest C_day and C_net rates were observed for bank and basin sites located in western Florida Bay. This observation is consistent with previous measurements (Nelson and Ginsburg 1986; Bosence 1989a) showing higher rates of sediment production for western Florida Bay banks and basins.

Bank calcification shows a strong negative correlation with both temperature and salinity; basin calcification shows a moderate correlation only with salinity (Fig. 6). Lowest rates of calcification for both banks and basins occurred on Russell and Barnes Key Banks and near Manatee Keys during hypersalinity events in September 1999 (42 psu) and July 2000 (37.6 psu). Millero et al. (2001) indicated from seasonal TA measurements that during high salinity events in western Florida Bay precipitation of CaCO₃ occurs and may be caused by inorganic precipitation due to the mixing of sediments with high salinity waters. Zieman et al. (1999) demonstrated from long-term, seasonal seagrass productivity data that higher salinities lead to reduced production by T. testudinum. Walker and Woelkerling (1988) indicated that production of some seagrass epiphytes, such as coralline algae, show reduced production during hypersalinity events. Our measurements of reduced calcification during high salinity events suggest that biogenic calcification as opposed to inorganic precipitation was the dominant contributor to carbonate sediment production at our study sites. The negative correlation between salinity and calcification indicates high salinity stress affects both calcifying organisms and seagrass production. As rates of biogenic calcification decrease and seagrass productivity (and sequestration of CO₂) decreases, rates of carbonate sediment dissolution can exceed gross carbonate sediment production. The strong correlation of bank calcification, but not basin calcification, with temperature suggests that elevated salinity, not temperature, exerts the most control on rates of calcification. During summer field expeditions, we have frequently observed that bank top temperatures are noticeably warmer than basin temperatures. The strong correlation between calcification and temperature on banks suggests evaporation primarily driven by elevated bank top temperatures may be the primary mechanism for elevation of salinity.

We report significant rates of carbonate sediment dissolution during the night at all basin sites (except hard bottom) and at 50% of all bank sites. Average C_net values ranging from -0.306 to 0.643 g CaCO₃ m^-2 24 h^-1 are significantly lower than previously reported rates of gross calcification calculated by standing crop and turnover methods (Nelson and Ginsburg 1986; Bosence 1989a; Frankovich and Zieman 1994). Nighttime dissolution rates (C_night) ranged between 8% and 71% of daytime calcification rates (C_day) on banks that showed net dissolution during the night (n = 3). C_night on basin substrates generally ranged between 23% and 122% of C_day at sites characterized by calcification during the day and dissolution at night (n = 13). Eight locations measured during complete cloud cover were characterized by dissolution during both day and night. One Manatee Keys intermediate Thalassia site (Table 2, March 17, 1999) measured during a high turbidity event showed a nighttime dissolution rate four times greater than the daytime calcification rate.

Walter and Burton (1990) characterized pore fluid chemistry of shallow carbonate sediments in Florida Bay near Captain Key, Cross Bank, and Crane Key. They reported rates of carbonate sediment dissolution ranging from 370 to 710 µmol cm^-2 yr^-1, or approximately 1.01 to 1.95 g CaCO₃ m^-2 d^-1. Walter et al. (1993) and Ku et al. (1999) confirmed carbonate sediment dissolution at these same sites at a rate of 400 µmol cm^-2 yr^-1, or 1.1 g CaCO₃ m^-2 d^-1 and suggested that over half the gross carbonate sediment production is dissolved or recrystallized. They attributed sediment dissolution to a combination of oxic respiration and sulfate reduction that generates carbonic acid for dissolution. Subsequent sulfide oxidation also generates acid for dissolution and prevents buildup of carbonate alkalinity (HCO3^-). Walter and Burton (1990) demonstrated that island flank sediments are characterized by higher total CO₂ (TCO₂), higher excess Ca²⁺ from sediment dissolution, and more bioturbation than mud bank sediments. They suggested that a greater degree of burrowing and transport of labile organic matter and O₂ into sediments along seagrass root systems facilitates organic matter decomposition, sulfide oxidation, and increased acid production for sediment dissolution.

Our measured rates of carbonate sediment dissolution observed primarily during night are similar to the range of values previously reported for porewater dissolution (Walter and Burton 1990; Walter et al. 1993; Ku et al. 1999). Basin sites, similar to island flank sites of Walter and Burton (1990), showed higher rates of dissolution (-0.035 to -1.972 g CaCO₃ m^-2 night^-1) than banks (-0.007 to -0.061 g CaCO₃ m^-2 night^-1). Within basin sites, seagrass sites showed higher rates of dissolution (-0.371 to -1.187 g CaCO₃ m^-2 night^-1) than mud bottom sites (-0.204 g CaCO₃ m^-2 night^-1). The similarity of our dissolution rates to previous work on porewater dissolution suggests that porewater dissolution may impart a chemical signature to surface waters. Corbett et al. (1999, 2000) suggest that groundwater may flow across the sediment water interface, effectively carrying a porewater signature to surface waters in Florida Bay. These studies remain controversial due to potential error associated with use of seepage meters to measure groundwater fluxes (Shinn et al. 2002). Our fluorescein dye injections showed no dilution of incubation chamber water and no leakage of significant amounts of groundwater (or pore water) into the chamber. Ku et al. (1999) suggested that the rate of oxygen supply to sediment pore water required to maintain high sediment dissolution rates facilitated by sulfide oxidation requires that pore waters in the upper 24 cm of sediment exchange with overlying seawater on the order of once per several weeks to months. Porewater advection rates, alone, are too slow to account for the magnitude of our observed changes in surface water chemistry over 24-h periods. It is likely that changes in TA measured in surface water reflect the combined effect of sediment dissolution processes resulting from porewater reactions, biogenic calcification driven by diurnal cycling of CO₂ through photosynthesis and respiration, and dissolution of surface sediments facilitated by elevated pCO₂ from respiration during the night.

With the exception of hard bottom substrate, relative rates of dissolution increase with increasing seagrass cover from mud bottom to intermediate seagrass sites (Table 3). This trend may be due to a corresponding increase in respiring benthic faunal populations (e.g., molluscs, crustaceans) with increasing seagrass density resulting in higher rates of respiration that facilitate dissolution during the night and from enhanced diffusion of inorganic carbon species (from porewater sediment dissolution) along the root and burrow systems associated with seagrass beds (Walter and Burton 1990). Lower rates of dissolution on banks as compared to basins may be due to distribution of respiring versus photosynthetic calcifying species in addition to generally higher rates of calcification observed on banks due to larger numbers of calcifying epiphytes on bank seagrass (Bosence 1989a).

Stockman et al. (1967) provided the most widely accepted long-term estimates of sediment accumulation in Florida Bay based on thickness of sediment to bedrock in basins and on mud banks, and age of sediment. This method accounts for carbonate sediment production, dissolution, and transport. They reported average accumulation rates of 5.3 cm 1000 yr^-1 for basins and 33 cm 1000 yr^-1 for banks. Weighting these two values to the relative proportion of total bank area in Florida Bay (10%) and basin area (90%; Stockman et al. 1967) yields an overall accumulation rate of 8 cm 1000 yr^-1. This long-term average is much smaller than accumulation rates derived from short-term productivity measurements.

Measurement of biogenic CaCO₃ production rates by investigators using standing crop and turnover methods typically yields a broad range of sediment production rates, some of which are much greater than rates of accumulation. Bosence (1989a) presents the most comprehensive calculation of annual carbonate sediment production rates in Florida Bay banks and basins from standing crop surveys and short-term growth measurements of a number of carbonate sediment-producing organisms including Porites, Thalassia epiphytes, mollusks, Penicillus, Soritid foraminifera, and Halimeda. He reported annual sediment production rates for banks and basins in central Florida Bay (Cross Bank) of 331 and 128 g CaCO₃ m^-2 yr^-1, respectively, and for banks and basins in western Florida Bay of 1,081 and 506 g CaCO₃ m^-2 yr^-1, respectively. Conversion of these values to accumulation rates using a density of 2.8 g cm^-3 for CaCO₃ and 60% porosity (Stockman et al. 1967) results in bank and basin accumulation rates of 28.5 and 11 cm 1000 yr^-1, respectively, for central Florida Bay, and 93 and 44 cm 1000 yr^-1 for western Florida Bay, respectively. Frankovich and Zieman (1994) estimated an overall accumulation rate of 7.2 cm 1000 yr^-1. The difference between production and accumulation of sediments in both banks and basins has generally been attributed to transport of sediment out of the Bay, and transport of sediment from basins to banks within the Bay (Stockman et al. 1967; Bosence et al. 1985; Bosence 1989a,b; Prager and Halley 1999).

The method of measuring carbonate production used in our study, known as the alkalinity anomaly technique (Smith and Key 1975), provides a measure of net carbonate sediment production defined as gross carbonate production minus dissolution of carbonate sediments. Our measurements account for carbonate sediment production and dissolution but not sediment transport. This method differs from standing crop and turnover methods (e.g., Bosence 1989a) that provide a measure only of carbonate sediment produced, and from sediment thickness methods (e.g., Stockman et al. 1967) that account for carbonate sediment production, dissolution, and sediment transport over long time periods (tens to thousands of years). Extrapolation of short-term productivity measurements to longterm sediment accumulation rates is highly speculative, and the associated error is difficult to quantify (Bosence 1989c; Frankovich and Zieman 1994). We present estimates of long-term sediment accumulation based on our short-term productivity measurements for comparison to other studies. We caution that these estimates are based on numerous assumptions and are likely to be accurate only to within an order of magnitude.

The basin substrate areas and total bank area used in our calculations were derived from Prager and Halley (1999), and include sparse seagrass, intermediate seagrass, hard bottom, mud bottom, dense seagrass, mixed bottom, open sand, and mud bottom suites. We have made the following assumptions with respect to substrate type. We collected no productivity data for open sand and have omitted it from our calculations. In our study areas, dense seagrass was located on top of banks and was included in mud bank measurements, so mud bank calcification rates were used to approximate dense seagrass values. The mixed bottom type of Prager and Halley (1999) represents a combination of the six other basin bottom type classifications and is estimated as the average of all other basin bottom type measurements included in our study. Our only hard bottom measurement is derived from western Florida Bay where hard bottom is generally characterized by the presence of coral, calcareous algae, and other calcifying organisms. Hard bottom sites in eastern and central Florida Bay contains fewer calcifying organisms, so our hard bottom accumulation estimate is likely to overestimate Bay-wide hard-bottom sediment production.

Average C_net rates for each substrate type from Table 2 and Table 3 were extrapolated to annual rates (Table 4). Sediment accumulation for each substrate type was calculated using methods of Stockman et al. (1967), a density for CaCO₃ of 2.8 g cm^-3, and porosity (P) of 60%. Annual sediment production rates (g CaCO₃ m^-2 yr^-1) were converted to sediment volume, SV (cm³ m^-2 yr^-1), using density for CaCO₃ (2.8 g cm^-3) such that SV (cm³ m^-2 yr^-1) = production (g m^-2 yr^-1) / density (2.9 g cm^-3). The volume of sediment pore space (VPS) was determined using the equation P = VPS/(VPS + SV). Addition of VPS and SV yielded the total volume of sediment and sediment pore space produced per year (cm³ CaCO₃ m^-2 yr^-1). This total volume (cm³ m^-2 yr^-1) was then converted to a sediment accumulation rate (cm 1000 yr^-1) by multiplying by the conversion factor 1 m2/10000 cm² and time (1000 yr) such that accumulation rate (cm 1000 yr^-1) = total volume (cm³ m^-2 yr^-1) X 1 m²/10000 cm² X time (1000 yr). Sediment accumulation for each basin type was weighted relative to the proportion of each basin substrate type per total area for basins in Florida Bay. Weighted sediment accumulation rates for all basin substrates were summed to derive a combined accumulation rate for basins.

Results of our long-term sediment accumulation calculations are shown in Table 4. We calculate average rates of long-term sediment accumulation in basins of 19 cm 1000 yr^-1 and on banks of 20 cm 1000 yr^-1. Our basin value is higher than the Stockman et al. (1967) value of 5.3 cm 1000 yr^-1; our bank value is lower than the Stockman et al. (1967) bank accumulation rate of 33 cm 1000 yr^-1. We assumed that enough sediment is transported to the banks from basins (Bosence 1989b,c; Bosence 1995) to generate the bank accumulation rate of 33 cm 1000 yr^-1 measured by Stockman et al. (1967), and we corrected our basin and bank accumulation values for sediment transport, accordingly, by subtracting 13 cm 1000 yr^-1 from our basin value of 19 cm 1000 yr^-1, and adding 13 cm 1000 yr^-1 to our bank value of 20 cm 1000 yr^-1. Our corrected basin and bank accumulation values then become 6 cm 1000 yr^-1 and 33 cm 1000 yr^-1, respectively. Weighting our basin and bank accumulation rates to the relative proportion of basins (90%) and banks (10%) as estimated by Stockman et al. (1967) gives an overall average rate of accumulation of 8.7 cm 1000 yr^-1. Interestingly, our sediment accumulation rate for basins (corrected for sediment transport only to banks) and our overall rate of accumulation are very similar to those of Stockman et al. (1967), suggesting that sediment dissolution may play a more important role than sediment transport out of the Bay as a cause for loss of carbonate sediment in Florida Bay.

Table 4. Long-term sediment accumulation rates for Florida Bay basins and mud banks. Substrate type, area, and % basin area data from Prager and Halley (1999). Cnet (average annual net calcification rate) is calculated from average Cnet (g CaCO3 m-2) for substrate type from Tables 1 and 2. Sediment accumulation is calculated as described in Stockman et al. (1967) for each substrate type. Weighted accumulation is the accumulation rate weighted to proportion of basin area. Combined accumulation is the sum of substrate-components (sum of weighted accumulations) for basins and banks.

[view table in full]

< Results | Conclusions >