Topical Area: Nutrient Dynamics

D.T. Rudnick and S.P. Kelly, Everglades Systems Research Division, SFWMD, West Palm Beach, FL and C. Donovan, and A. Gottlieb, Center for Environmental Studies, FAU, Boca Raton, FL

Summary. Altering patterns of fresh water flow through the Everglades and into Florida Bay may also alter nutrient inputs and cycling in these ecosystems. We have measured nutrient fluxes between sediments and water seasonally since May 1996 in a pond landward of Florida Bay's mangrove fringe and in the bay (1 to 3 km seaward of the fringe). The field sites span a wide range of salinity regimes and are dominated by different seagrass species. Benthic nutrient fluxes were measured in situ using dark and light acrylic chambers, with continuous mixing. For all sites, P fluxes were very low, with mean sediment uptake or regeneration rates of < 1 umol

m^-2h^-1. Nitrogen regeneration was also surprising low, with net nitrate plus nitrite uptake in both dark and light chambers at all sites and ammonium uptake in dark chambers at one site. We calculate that this site, a mangrove pond, was a net sink for N. The resultant stoichiometry of benthic 0, N, P fluxes in the dark deviated greatly from ratios expected from the decomposition of algae or seagrass. Low N and P fluxes probably reflected the dominance of autotrophs at our sites, low P availability, and nitrification-denitrification. The effects of salinity changes on these benthic fluxes were not apparent from our results.

Introduction. A large environmental restoration project that is designed to improve the hydrological conditions of the Florida Everglades and increase water flow to Florida Bay is underway. These changes may not only alter the bay's salinity regime, but may also alter its nutrient cycles. The effects of fresh water flow on nutrient cycles may be direct, such as changing nutrient transport and exchange across the Bay-wetland ecotone. These effects may also be indirect, mediated through changing salinity, which can alter community structure and geochemical processes.

This study of benthic nutrient fluxes is part of a larger research program that is studying the exchange and cycling of nutrients across the Florida Bay-Everglades ecotone. Our field sites are in areas of Florida Bay that receive most of the fresh water that flows directly from the Everglades. The wetland's plant community is dominated by scrub mangrove trees except for a higher stature fringing forest near the Bay's shoreline. Numerous ponds and streams exist north of this fringe, but few streams flow across it and into the bay. We have focused our efforts on studying nutrient cycles in a pond immediately north of this fringe (Taylor River Pond One), a second site immediately south of one of the main mangrove streams in the region (Little Madeira Bay), and a third site about 2 km south of this coastline (southwest of Little Madeira Bay). Additional measurements have been made in two sites in north-central Florida Bay, in Terrapin Bay and in Florida Bay proper, south of Crocodile Point. Here, we present results from the first measurements of benthic nutrient fluxes that have been made in this region of Florida Bay. Our initial goal is descriptive - to quantify N and P fluxes in a region where P concentrations are low and N concentrations are high.

Methods. Nutrient and oxygen flux measurements were made using a set of 4 replicate clear chambers and 4 replicate dark chambers. On a given day, 2 clear or 2 dark chambers were incubated. [During the winter, 4 clear chambers were incubated, with day and night sampling.] Each chamber was an acrylic cylinder that covered 0.3 m² of the sediment surface. Water within the chamber was continuously mixed (without increased resuspension) by a submersible pump. Following the incubation period, final samples were taken through a sample tube. The volume of water removed from the chamber was replaced by water in an external "make-up" bag. Bethnic fluxes were calculated by subtracting pelagic nutrient and oxygen changes, which were measured in bottles that were incubated along-side the chamber, from whole chamber nutrient and oxygen changes. For each chamber, above-ground seagrass samples (from either the whole chamber or two 225 cm² quadrats) and two sediment cores (each 20 cm², with 0-2 cm and 2-6 cm horizons sliced) were also collected.

Dissolved 0₂, pH, temperature, and conductivity were measured at 15 min. intervals by a Hydrolab Recorder, which was sealed in a port in each chamber's mixing hoses. Water samples were analyzed for chlorophyll a, total suspended solids, ammonium, nitrate plus nitrite, dissolved Kjeldahl N, filterable reactive P, total P, and dissolved organic C. Sediment loss on ignition was measured using a furnace at 500° C.

Results. Salinity and vegetation were far more variable at the Taylor Pond 1 site than at the other sites. Salinity ranged from < 1 ppt to 28 ppt in the pond. Vegetation was also more variable in the pond, with a "bloom" of Ruppia maritima in late 1996. Thallasia testudinum biomass at the Little Madeira site decreased through the wet season (summer and fall).

Dissolved 0₂ uptake in dark chambers varied over about a two-fold range among the sites. No significant (p < 0.05) correlations existed between 0₂ uptake rates (all sites and times)and salinity or between these rates and plant biomass. These rates were significantly correlated with temperature, but this explained only 32% of the rate's variance.

Almost no flux of inorganic or organic P occurred at any site in the dark. Pond 1 had the highest P fluxes, with an average of < 1 umol m^-2h^-1. Nitrogen fluxes in dark chambers were far more variable among the sites than 0₂, fluxes, with no significant correlation to temperature, salinity, or plant biomass. Fluxes across the sediment (plus plant)-water interface occurred in both directions. A net loss of N0₃ plus N0₂ (NO_x) from the water column occurred at all sites in the dark chambers. In Pond 1, there was no net flux of NH₄ in the dark, despite O₂ uptake of about 50 mg m^-2h^-1. A net benthic uptake of both dissolved inorganic (DIN) and total N occurred at this site. Along the transect from Pond 1 to south of Little Madeira Bay, NH₄ and DIN fluxes increased. However, these fluxes were highly variable with time at these two open water sites. Ammonium fluxes at the two north central bay sites were much higher than at the northeastern sites.

The stoichiometry of nutrient fluxes in the dark did not reflect decomposition processes. The molar ratios of 0₂ to DIN fluxes, 0₂ to dissolved inorganic P (DIP) fluxes, and DIN to DIP fluxes all deviated greatly from ratios that would reflect plankton decomposition or seagrass detritus decomposition. An exception was O:N at western sites and the site south of the Little Madeira Bay, where flux ratios were more similar to seagrass ratios.

In light chambers, P fluxes were near zero (averaging 0.1 umol m^-2h^-1 benthic uptake). Nitrogen was removed from the water column in clear chambers during the day. Both NH₄ and NO_x had the same pattern of faster benthic uptake rates in the light than in the dark. Rates were highly variable. In Pond 1, high NO uptake rates coincided with the occurrence of low salinity water, elevated NO_x concentrations (range of 0.5 to 3 uM from 8/96 - 1/97, compared with < 0.2 uM in 5/96 and 5/97), and a period of increasing Ruppia biomass. In Little Madeira Bay, a high flux of NH₄ from the sediment occurred in 5/96. An almost identical pattern was observed at the site south of Little Madeira Bay. At both sites, the anomalously high efflux of N occurred during the only sampling period when night-time 0₂ demand exceeded daytime 0₂ production.

From integrated 24 hour estimates of net 0₂ and nutrient fluxes, it appears that Pond 1 was a net N sink from 5/96 to 5/97. The two sites south of Pond 1 also were net NO_x sinks but net DIN and total N sources. Despite the low magnitude of the P fluxes, sediments near the mangrove zone appear to be a small P source, while sediments in Florida Bay proper appear to be a small P sink.

Discussion. The stoichiometry of benthic nutrient fluxes at all of our studies sites was unusual. In our dark chambers, we measured far less N and P than would be expected from the decomposition of organic matter. There are several mechanisms that may cause the observed high, and even negative, O:N and O:P ratios, as well as the wide range of N:P ratios. Some these mechanism are as follows:

1. Autotroph interference. With their shallow depths, our sites were covered by seagrasses and benthic algal mats. These organisms may intercept P and N at the sediment-water interface during the both day and night. Furthermore, the slope of nutrient gradients in porewaters may be decreased by plant root uptake.
2. P scarcity. Water in the Everglades and in eastern Florida Bay is highly depleted in P relative to N. The Everglades has TN:TP ratios > 400 (mean TP = 0.2 uM, TN = 60 uM, DIN = 1 uM), and the eastern bay has ratios near 300 (mean TP = 0.2 uM, total N = 60 uM, and dissolved inorganic N=10 uM). Low P fluxes reflect this P scarcity, with benthic organisms (including bacteria) effectively retaining P.
3. P binding. About 90% of the sediment at our sites is carbonate. Most of this carbonate has a fine, silt-clay grain size. Inorganic P can be bound by carbonate surfaces, contributing to low mobility and availability.
4. Nitrification and denitrification. Finding low N fluxes at sites in the ecotone between two N rich ecosystems, the Everglades and Florida Bay, was quite surprising. For the given benthic metabolic rates, greater day-time uptake and night-time release would be expected. Low N fluxes may largely be a function of nitrification and denitrification. Benthic uptake of NO_x occurred at all sites and NH₄ fluxes from the sediment at our eastern sites were only found when daily O₂ production was much less than nightly 0₂ consumption. Low O₂ production may have slowed nitrification and enabled an NH₄ flux to occur. Regionally, it appears that the N cycle of eastern Florida Bay is not dominated by benthic processing, but rather pelagic processing of DON that is exported from the Everglades. We expect that rates of N cycling are strongly influenced by P availability.
5. Organic matter source. Mangrove trees are the dominant plants in the wetland adjacent to our sites. Its high C:N:P ratio (particularly in wood) probably influenced benthic flux stoichiometry at our sites. This influence is most likely at Taylor Pond 1, which is surrounded by mangrove trees. Inputs of detritus from mangrove wood may also account for N retention in the pond's sediments. Such retention occurs during detrital decomposition in forest soils. Immobilization of N by detritus may account for Pond 1 being a net sink for N. However, the Little Madeira Bay site, which probably received far less mangrove detritus, also had very low N fluxes from the sediment, indicating the importance of other mechanisms.