JEQ Journal of Natural Resources and Life Sciences Education
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Citing Articles
Right arrow Citing Articles via ISI Web of Science (9)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Rowe, G. T.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Rowe, G. T.
Agricola
Right arrow Articles by Rowe, G. T.
Related Collections
Right arrow Water Quality
Right arrow Ecosystem Restoration
Right arrow Nutrient Management
Journal of Environmental Quality 30:281-290 (2001)
© 2001 American Society of Agronomy, Crop Science Society of America, and Soil Science Society of America

ARTICLE
SYMPOSIUM PAPER

Seasonal Hypoxia in the Bottom Water off the Mississippi River Delta

Gilbert T. Rowe

Department of Oceanography, Texas A&M University, College Station, TX 77843

Corresponding author (growe{at}ocean.tamu.edu)

Received for publication July 14, 2000.

   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 THE CONCEPTUAL MODEL: BOTTOM...
 RESULTS
 DISCUSSION
 THE EXTRINSIC FORCING FUNCTIONS...
 REFERENCES
 
Hypoxia (oxygen concentration less than 2 mg L-1 or 62.5 mmol m-3) occurs on the Louisiana continental shelf during summer when the consumption of oxygen by sediment and water column respiration exceed resupply by photosynthesis and mixing. Biological processes that consume or produce oxygen have been summarized in a budget that can be used to quantify the degree to which consumption in deep water and in the sediments exceeds net production and thus the time it takes to reach hypoxic conditions following the spring onset of stratification. The net consumption rate by the sea floor biota (sediment oxygen consumption, SOC) is inversely related to oxygen concentration and directly related to temperature. Photosynthesis is of potential importance throughout the deep water column and on the sea floor when light is adequate. A non-steady state, time-dependent numerical simulation model is used to compare biological and physical processes with shipboard measurements and continuous near-bottom records. The simulations illustrate possible variations in oxygen concentration on time scales of hours to months, and these in general match much of the variability in the direct observations at time scales of days to weeks. The frequently observed unremitting anoxia lasting weeks at some locations is not produced in the present simulations. A possible explanation is the chemical oxidation in the water column of reduced metabolic end-products produced in the sediments by anaerobic metabolism. Direct measurements of biological processes could lead to better understanding of how extrinsic forcing functions can best be managed to improve water quality.

Abbreviations: SOC, sediment oxygen consumption • WCR, water column respiration


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
THE CONCEPTUAL MODEL: BOTTOM...
 RESULTS
 DISCUSSION
 THE EXTRINSIC FORCING FUNCTIONS...
 REFERENCES
 
IT has been observed often that the bottom waters of the inner continental shelf of the northwest Gulf of Mexico become hypoxic on an annual basis during the summer months (Gaston, 1985; Rabalais et al., 1994). This phenomenon can have disastrous effects on the fishes and bottom invertebrates (Pavela et al., 1983; Renard, 1986) as far west as Texas (Harper et al., 1981). This paper attempts to explain the various intrinsic processes in the bottom water that are presumed to cause hypoxia, as opposed to extrinsic loading from nitrate, freshwater, dissolved organics, suspended clay particles, etc., that give rise to these circumstances (Turner and Rabalais, 1991; Eadie et al., 1994; Nelsen et al., 1994). The condition results from an imbalance between biological oxygen consumption and production, and physical sources and sinks, in any generic nearshore environment, but this review is based on observations and experiments made over the last several decades just west of the Mississippi River (Fig. 1) . The most recent studies involved a multidisciplinary program supported by the U. S. National Oceanic and Atmospheric Administration's Nutrient Enhanced Coastal Ocean Productivity project, or NECOP, along with regional university laboratories interested in eutrophication in the northern Gulf of Mexico. A concomitant investigation has been supported by the Minerals Management Services of the U.S. Department of the Interior. Called LATEX, this study extended from 90°30' W long. across the Texas shelf to the Mexican border. Both studies involved shipboard experiments and observations, longterm moorings, remote sensing, and mathematical modeling. Because similar problems of low oxygen have been observed elsewhere in coastal waters (Chesapeake Bay [Officer et al., 1984], Waquoit Bay, MA [D'Avanzo and Kremer, 1994], Offatts Bayou [Texas], New York Bight New York Bight [Falkowski et al., 1981], etc.), this paper will include a generic, somewhat pedagogical explanation of the processes involved within the ecosystem.



View larger version (38K):
[in this window]
[in a new window]
  		Fig. 1. Station locations in the hypoxic zone off the Mississippi River

 
The area and intensity of the hypoxia off Louisiana have been described in a series of publications resulting from NECOP and LATEX studies. The reader is referred to a dedicated issue of the journal Estuaries (1994, Vol. 17, No. 4) in which a variety of aspects of the problem are addressed. In addition, Wiseman et al. (1997) dealt with physical conditions on the shelf, in particular the density stratification. Several attempts have been made to model the physical and chemical processes to explain shelf productivity (Chen et al., 1997) and hypoxia (Bierman et al., 1994), as well as define its occurrence over long time scales (Nelsen et al., 1994). Nitrate loading in the Mississippi River discharge is acknowledged to be a major culprit (Turner and Rabalais, 1991; Bratkovich et al., 1994), but the circulation is far from predictable (Walker et al., 1994).

Two somewhat different but not mutually exclusive explanations have emerged concerning the extrinsic causes of the Louisiana hypoxia. The physicists have proposed that the freshwater plume prevents mixing of oxygen-rich surface waters into the interior water column, where respiration dominates. They demonstrate that when the vertical density gradient reaches critical levels, the bottom water can no longer be refreshed and hypoxia results in a matter of weeks to months. If the gradient is interrupted by wind events, then mixing occurs, the bottom waters are refreshed, and hypoxia is prevented or at least delayed (Wiseman et al., 1997). On the other hand, biologists and chemists have observed that nitrate concentrations in river water have increased in the last half century (Walsh et al., 1981; Turner and Rabalais, 1991), leading to an increase in algal growth (Lohrenz et al., 1990). The algal cells, cellular debris, and zooplankton feces sink by gravity into deep water (Redalje et al., 1994) where they can accumulate (Eadie et al., 1994) and ostensibly enhance respiration. In a month or so, the imbalance between supply and demand leads to hypoxia. Thus, an increase in nitrate loading in the river might explain the expansion of the affected area (Turner and Rabalais, 1991; Justic et al., 1993) over the last few decades (Nelsen et al., 1994). High concentrations of nitrate in agricultural runoff are presumed to be the most culpable forcing function. Indeed, Eadie et al. (1994) were able to define the relationship between nitrate flux and organic matter accumulation rates in the mud, based on 210Pb dating (Nelsen et al., 1994). Diminishing the nitrate loading in the river, it might be suggested, would thus slow the production of phytoplankton in the plume over the continental shelf. This would lessen the organic loading of deep water and, one might suspect, slow the net losses to respiration in deep water (Justic et al., 1993).

The present paper will not treat these extrinsic forcing functions. Rather, it will deal with the intrinsic processes that either produce or consume oxygen, often in response to the above-mentioned forcing functions. Quantitative rate data will be put into a budget for comparison with direct observations of oxygen concentration at different times. Based on the budget and the data supporting it, a numerical simulation will be created to see how closely the model results match observations. This comparison of model vs. data should give us some idea of how well the fluxes are being measured and perhaps point to new avenues for investigation.


   THE CONCEPTUAL MODEL: BOTTOM WATER OXYGEN PRODUCTION AND CONSUMPTION
TOP
 ABSTRACT
 INTRODUCTION
 THE CONCEPTUAL MODEL: BOTTOM...
RESULTS
 DISCUSSION
 THE EXTRINSIC FORCING FUNCTIONS...
 REFERENCES
 
The oxygen budget for near-bottom water can be viewed (Fig. 2) as the concentration of oxygen (the box) as it is being controlled by a suite of fluxes or processes (the arrows). At steady state the arrows entering the box should equal those leaving the box, but in conditions leading to hypoxia the arrows out will be slightly larger than those entering, that is, the system is not in steady state. This conceptual model can also be represented as a differential equation, as follows:




View larger version (21K):
[in this window]
[in a new window]
  		Fig. 2. Conceptual model of sources (arrows into box) and sinks (arrows out of box) of oxygen in the bottom water of the continental shelf off the Mississippi River

 
That is, the change in oxygen in time equals photosynthesis, plus any vertical or horizontal physical processes that would import new oxygen from oxygen-rich waters, minus respiration (bottom and water column), including chemical oxidation of reduced metabolic end-products, and nitrification.

Each process is important to the concentration of oxygen. Losses to sediment oxygen consumption (SOC), the respiratory metabolism of the entire community of sediment-living organisms, is often presumed the single most important oxygen loss (Officer et al., 1984). It is usually considered a zero-order flow across the bottom boundary in oxygen-based or food chain carbon models (Falkowski et al., 1988; Bierman et al., 1994; Kemp et al., 1994). It is measured either in situ with benthic incubation chambers (Kemp et al., 1992; Rowe et al., 1997) or aboard ship by incubating recovered cores (Pamatmat, 1971; Miller-Way et al., 1994).

Chemical oxidation of reduced metabolic end-products is also important in sediments. The principal anaerobic metabolic respiratory process is sulfate reduction, in the absence of oxygen. This produces sulfide, which is readily oxidized chemically by free oxygen and by sulfide oxidizing chemoautotrophic bacteria. This latter consumption occurs in surficial sediments when oxygen is available. It is measured as a component of the total oxygen demand of sediments (SOC). Therefore it is not added as a separate component but included with the total sediment oxygen consumption (SOC) (Morse and Rowe, 1999). The chemical oxygen demand is assumed to approximate the total sulfate reduction in the sediments (Pamatmat, 1971), but pyrite formation, although relatively small (Rowe et al., 1988), prevents this from being entirely valid.

It should be noted that some anaerobic metabolic processes might not be included as a component of the total oxygen demand (SOC) if the reduced end-products these reactions produce are not oxidized readily by free oxygen in the surficial sediments or in the bottom water included in the incubations. Examples could be denitrification and metal oxide reduction (Fe and Mn) (Aller et al., 1991; Seitzinger et al., 1993; Luther et al., 1997). It is generally presumed that they are a relatively small fraction of the total respiratory metabolism. This remains to be demonstrated. The importance of the oxidation of NH+4, Mn, and Fe may be appreciable, but such processes unfortunately have not been quantified in these sediments.

A competing sink for oxygen is water column respiration (WCR) by heterotrophic organisms, especially bacteria (Williams, 1981). Falkowski et al. (1981) presumed that the dinoflagellate Ceratium tripos was responsible for low oxygen in the New York Bight when the bulk of the population was located below its compensation depth. This created hypoxia during periods of intense stratification in late summer. Historically, WCR has been measured by observing rates of oxygen losses in dark bottle incubations both in situ and aboard ship (Chin-Leo and Benner, 1992; Biddanda et al., 1994), but this is difficult because the rates are small relative to possible error in the measurements, leading to meager precision. Alternative methods include measurements of bacterial production using labeled (tritiated) leucine or thymidine, and then presuming that respiration has some known relationship to production. One can add tracer levels of 14C-labeled organic substrate, such as mixed amino acids, and then monitor the production of the label in CO2 and fixed in the cells. Interpreting tracer results is difficult because the ambient substrate concentrations are rarely known. The electron transport system (ETS) activity has also been used to estimate respiration rates (Packard, 1985), and this technique was applied by Dortch et al. (1994) to the waters below the plume.

The two biological sources of oxygen to the deep layer are both photosynthesis: by microalgae living on the bottom and by typical phytoplankton in the water column. While this process tends to dominate shallow ecosystem dynamics (D'Avanzo and Kremer, 1994), it has generally been assumed that photosynthesis will be nearly negligible below the plume, hence the lack of pertinent data. It has been shown, however, that bottom photosynthesis on the inner continental shelf can almost equal that in the water column when light levels are adequate (Cahoon and Cooke, 1992).

The remaining sources and sinks are products of physical mixing and circulation. Oxygen will be transferred by mixing down concentration gradients, either in the horizontal direction or vertically. Vertical transfers are due mostly to mixing by turbulent diffusion and horizontal exchanges are related to advective flow, but both can be a mixture of complex processes that are not well quantified on small time and space scales. When surface water becomes supersaturated, O2 is lost to the atmosphere (Liss and Merlivat, 1986), and the loss rate is a function of wind speed (Winninkhof, 1992). The vertical transfer to depth from surface waters can be parameterized as Fickian diffusion, where the flux is equal to concentration gradient times an eddy diffusion coefficient. The vertical gradients can be steep, and thus the intensity of the mixing or lack thereof is extremely important. Horizontal exchange of water along or across the continental shelf is equal to the concentration times the current velocity. However, concentration gradients in the horizontal are often small. Changes due to horizontal exchanges can result in lower as well as higher concentrations, depending on the relative concentrations of the water masses involved. If along shore and cross-shelf (e.g., horizontal) homogeneity is assumed, then this term can be ignored in short time and space scales.


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 THE CONCEPTUAL MODEL: BOTTOM...
 RESULTS
DISCUSSION
 THE EXTRINSIC FORCING FUNCTIONS...
 REFERENCES
 
A Bottom Water Oxygen Budget below the Mississippi River Plume
The data used in this budget have been derived from field work by colleagues in NECOP or LATEX. The methods used will not be described in detail because this information can be obtained from original papers given in the references. The raw data can be obtained electronically from NOAA's Coastal Ocean Program (COP) files at the Atlantic Oceanographic and Meteorological Laboratory (AOML) in Miami, FL (Hendee, 1994).

Benthic chamber flux data are available (July 1990 and 1991, April 1992, and August 1994) from the cluster of stations pictured in Fig. 1, and have been reported by Miller-Way et al. (1994) and Rowe et al. (1995). Respiration rates were a function of temperature and oxygen concentration, as might be expected (Fig. 3 and 4) . The mean rate for August was 1.095 mmol m-2 h-1 (35 mg m-2 h-1), but this was somewhat lower in July (0.446 mmol m-2 h-1; 14.3 mg m-2 h-1). This difference appears to be related to oxygen concentration (Fig. 4), a relationship that has not previously been quantified. The mean during April was 0.644 mmol m-2 h-1 (20.6 mg m-2 h-1), but the mean temperature during that period was only 19.8°C. A typical Q10 relationship was followed in which respiration doubled for every 10°C rise in temperature, if stations with low oxygen (<50 mmol m-3 [1.6 mg L-1]) are excluded (Fig. 3).



View larger version (15K):
[in this window]
[in a new window]
  		Fig. 3. Bottom sediment community total oxygen consumption (SOC) as a function of bottom water temperature (excluding stations with bottom water oxygen less than 50 mmol m-3)

 


View larger version (16K):
[in this window]
[in a new window]
  		Fig. 4. Bottom sediment community total oxygen consumption (SOC) as a function of bottom water oxygen concentrations (excluding stations with bottom temperature less than 24°C)

 
Benthic primary production was calculated by subtracting oxygen demand in opaque (respiration alone) benthic incubation chambers from that in transparent chambers (respiration plus photosynthesis) at sites in Fig. 1. The calculated mean of all the estimates was used in the budget: 0.2 mmol O2 m-2 h-1 (Rowe et al., 1995; Dortch et al., 1994).

Bottom water column respiration (WCR) values are taken from the review by Dortch et al. (1994), who used electron transport system (ETS) activity to estimate respiration in the water column, after calibration by comparison with other methods. Their mean for the C-line stations was 4.56 mmol O2 m-2 d-1, or 0.19 mmol m-2 h-1. This value is an integration of the water column below the oxycline, or approximately 10 m. The rates increased at low oxygen concentrations, rather than decreasing, as observed for the sea floor incubations. These rates were substantially lower than those in proximate surface waters (Biddanda et al., 1994; Gardner et al., 1994; Pakulski et al., 1995), as might be expected, but it should be noted that they are only about 20% of the mean rate measured by Biddanda et al. (1994) just offshore of the stations reported here, in 50 m water depth.

Deep water photosynthesis was not measured in NECOP or LATEX. Dortch et al. (1994) estimated it indirectly from pigment concentrations: 1.87 mmol O2 m-2 d-1, or 0.16 mmol O2 m-2 h-1, for a 12-h daylight period.

Nitrification, the bacterial transformation of NH+4 to NO-2 and NO-3, also consumes oxygen. In general, it is not considered a major consumer relative to respiration, but Pakulski et al. (1995) observed heightened rates at intermediate salinities in surface waters along the course of the plume. As much as half the oxygen consumed was attributed to nitrification in locations with intermediate salinities. It has not been measured in near-bottom water below the plume. An increase in nitrate concentration to about 11 µM has been observed in LATEX data (Chen et al., 1997) in the halocline between 10 and 20 m depth along the 90°30' W long. line, but its origin is equivocal. In deep water with low light this concentration is as likely a lack of nitrate uptake by phytoplankton as it is production, based on the coupled biophysical model presented by Chen et al. (1997). Nitrification is not included in the initial oxygen budget.

The budget, based on the information summarized above, is as follows:



Note that SOC dominates the process. Mixing is ignored at this time. Time to anoxia is about 11 d. However, as oxygen declines, SOC will also decline. Therefore, at the lower rates the equation would be the following:

At this rate, anoxia will take slightly more than a month. Even so, bottom oxygen demand (SOC) is the greatest consumer of oxygen.

Vertical physical mixing into the bottom water, a potentially important source, has not yet been included, whereas nitrification, a potentially important loss, has been ignored at this stage as well.

Estimates of Vertical Turbulent Mixing
The data for 1991 provided by Dortch et al. (1994) are used for this analysis, but other years illustrate much the same pattern (e.g., for 1990, in Rabalais et al., 1994). Figure 5 illustrates the seasonal differences in conditions that are typical of the cluster of stations in Fig 1. Oxygen declined from high values at the surface to lower values at depth in all cases. During the spring phytoplankton bloom in May, surface values were supersaturated (9 mg L-1, 281 mmol m-3). In July and October, they were 25 to 30% lower at the surface. Deep water values dropped from about 5.5 mg L-1 (172 mmol m-3) in May down to slightly less than 2 mg L-1 (62.5 mmol m-3) in July. In the fall, the water was mixed and the values in deep water returned to about 6.5 mg L-1 (200 mmol m-3). Bottom water values followed more or less the same sequence. Values in May (4 mg L-1, 125 mmol m-3) dropped down to about 0.4 mg L-1 (12.5 mmol m-3) in July, or well below the nominal 2 mg L-1 (62.5 mmol m-3) upper boundary for hypoxia. At the end of the summer, mixing resulted in a value of 2.5 to 3.5 mg L-1 (80 to 109 mmol m-3) in the bottom water during October.



View larger version (18K):
[in this window]
[in a new window]
  		Fig. 5. Oxygen concentrations (top, mg O2 L-1) and density (bottom, {sigma}t) as a function of depth (m) in the water column at (Fig. 1) in May, July, and October, 1991 (modified from Dortch et al., 1994)

 
The pycnocline (sharp change in water density) has been plotted as {sigma}t ([density of sea water - density of fresh water at STP] x 1000) versus depth in the same samples (Fig. 6) (Dortch et al., 1994). This discontinuity separates surface-mixed waters from deep water and inhibits mixing between the two. The steeper the gradient, the less exchange will occur (Gargett, 1984). The rate of exchange between surface and bottom water, or any two depths in the vertical, can be parameterized as Fickian diffusion in the following manner:

where Kd has units of m-2 h-1 and the flux is in mmol m-2 h-1. The vertical exchanges will be limited to the flux across the zone in the water column with the least mixing, which we can presume is that depth interval with the steepest density gradient (Gargett, 1984). Using this relationship, we ought to be able to calculate vertical fluxes from Fig. 5 and 6, if Kd were known for these density boundaries. Using a Kd of 0.0038 m2 h-1 (Gargett, 1984), one would estimate that the vertical fluxes in the bottom meter could range from 0.11 mmol m-2 h-1 in May up to 0.34 mmol m-2 h-1 in October. The mean for the three seasons, 0.19 mmol m-2 h-1, is used in the model below.



View larger version (10K):
[in this window]
[in a new window]
  		Fig. 6. Continuous oxygen probe records made in situ 1 m above bottom at C6A. Data modified from Rabalais et al. (1994). The original data were in mg O2 L-1

 
Numerical Simulation of Variations in Deep Water Oxygen off Louisiana
The above information on the processes within the bottom water has been put into a time-dependent numerical simulation of oxygen concentrations during summer when the water is stratified. The purpose of the simulation is to compare measured values of the rates of biological oxygen production and consumption with the observed variations during the summer period within the sampling area. By comparing model results with observations, we should get some insights into how accurately the processes causing hypoxia have been measured. By adding or dropping processes, or changing their values, we can test the sensitivity of hypoxia to different processes.

The differential equation to be solved is the following, based on the conceptual model (Fig. 2):








Initial bottom water oxygen = 63 mmol m-3.

The following inputs are used:






where z is the vertical distance between levels in the water column. In this case it is 1 m (see Fig. 5). Losses are parameterized as the following:




Deep water oxygen concentration initial conditions are equal to 156 mmol m-3 (5.0 mg L-1). The diel variation in the photosynthesis term is parameterized as sin({pi} x time/12).

The initial conditions are the May deep water values in Fig. 4 converted from units of mg L-1 to mmol m-3. Units for the rates are mmol O2 m-2 h-1.

Photosynthesis and respiration are taken directly from the steady state budget above and assume that rates are integrated over the bottom 10 m of the water column. Photosynthesis, however, is parameterized as a day–night light dependent process; the rates are multiplied by a sine curve providing a day length of 12 h. Sediment oxygen consumption is directly dependent on oxygen concentration, as in the equation given in Fig. 4. The physical mixing down to the bottom water is calculated as above using a vertical eddy mixing coefficient of 0.01 cm2 s-1 (= 0.0038 m-2 h-1). Nitrification is assumed to be 30% of the water column respiration (WCR) (Pakulski et al., 1995) and advection is assumed to be nil (Rabalais et al., 1994).

The above equation was solved using the commercial software STELLA II, Version 3.0, available from High Performance Systems, Hanover, NH. Run on a Macintosh Quadra 610, the integration used Euler's method with a time step of 1/4 h.

The results of the simulation suggest that the intrinsic processes are fairly well parameterized (Fig. 7) . Oxygen concentration falls abruptly at first. It then tapers off as oxygen concentrations decline, thus preventing absolute anoxia. A small variation occurs on a daily basis, reflecting the light–dark effects on photosynthesis.



View larger version (18K):
[in this window]
[in a new window]
  		Fig. 7. Results of a numerical simulation of sediment oxygen consumption (mmol O2 m-2 h-1) and oxygen concentration for summer conditions with a vertical eddy mixing coefficient of 0.01 cm2 s-1

 
A series of experiments was run to determine how sensitive the system is to changes in different rates and the steepness of the vertical stratification. By doubling the photosynthesis both on the bottom and in the water column, the oxygen concentration still goes down, but reaches a steady state mean of 50 mmol m-3 (1.6 mg L-1). This is still by definition hypoxic. This suggests that deep water and benthic photosynthesis might be less important than previously inferred (Dortch et al., 1994; Bierman et al., 1994). When the vertical mixing coefficient is reduced by 50%, then the bottom water oxygen concentration at steady state drops from 40 mmol m-3 (1.3 mg L-1) down to about 20 mmol m-3 (0.6 mg L-1). By increasing mixing by an order of magnitude up to 0.038 m-2 h-1, bottom water actually increases up to 125 mmol m-3, or 4 mg L-1, well above the hypoxic boundary. Thus, the bottom water oxygen is very sensitive to mixing. We know little about the physics involved in this process, or how to measure it with more accuracy and precision at the necessary time and space scales.


   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 THE CONCEPTUAL MODEL: BOTTOM...
 RESULTS
 DISCUSSION
THE EXTRINSIC FORCING FUNCTIONS...
 REFERENCES
 
Model Validation: a Comparison with the Seasonal Cycle
An alternative approach to determining net rates of oxygen consumption and production is to compare the differences in oxygen over time and depth in the different seasons. These seasonal values (Fig. 6) can be used to infer approximate net rates of SOC and WCR for the deep water. Subtracting the mean value for mid-level deep water for July (2 mg L-1, 62.5 mmol m-3) from that in May (5.5 mg L-1, 172 mmol m-3) gives a difference of 3.5 mg L-1 (109 mmol m-3). Integrating this difference over the water depth in question (5 m) and dividing by the time period (approx. 60 d) gives an estimated net rate for respiratory metabolism of 12 mg O2 m-2 h-1 (0.37 mmol m-2 h-1).

Continuous recordings of oxygen concentration (ca. 1 meter above bottom) were made in 1990 on two moorings located at C6A and 4 in Fig. 1 (Rabalais et al., 1994). [Note: "4" is located near "WD32E" in Rabalais et al., 1994]. A section of the mean values at C6A in Rabalais et al.'s Fig. 6 (Fig. 8) has been modified for comparison with model results (Fig. 7). The mooring record illustrates three periods of oxygen accretion and two periods of oxygen decline over a 20-d period. The rates of decline for the two hypoxic events (between Calender Days 237 to 243 and 245 to 249) are equal to 0.33 and 0.73 mmol m-2 h-1, respectively. These rates are within the range of values expected, based on the rate measurements summarized in the budget. The mean rate of decline in the simulation during the first 90 h of the simulation is 0.28 mmol m-2 h-1, which is slightly less than the lower rates of decline calculated from the seasonal change monitored with ship-board hydrographic stations from Dortch et al. (1994). Thus, the model underestimates the rate. The causes of the unremitting anoxia (no measurable oxygen) at C6A are not explained by the model.



View larger version (10K):
[in this window]
[in a new window]
  		Fig. 8. Continuous oxygen probe records made in situ 1 m above bottom at C6A. Data modified from Rabalais et al. (1994). The original data were in mg O2 L-1

 
The disparity between the model and the records might lie in the respiration rates (SOC and WCR) used in the model. Direct measurements of oxygen demand (Biddanda et al., 1994) were higher than those used in the model (Dortch et al., 1994). Dortch et al. (1994) found evidence that WCR increases at low oxygen concentrations, but this relationship was not quantified and therefore is not in the simulation. Another explanation might be the release of reduced metabolic end-products from the sediments during low oxygen events when the redox discontinuity (RPD) migrates up to the sediment–water interface. Temperature may also be important; it increases gradually about 5°C between May and September (Fig. 3) in the deep water column. This is not incorporated into the present simulation.

Rabalais et al. (1994), and others, have questioned whether the relatively short time-scale events observed in their record are a product of biological activity or complex circulation that moves water with varying concentrations around in the narrow area of the shelf west of the plume. The general match between the time-dependent simulation model results (Fig. 7), in which lateral transfers were zero, and the record (Fig. 8), suggests that the reaction rates used were high enough to cause some of the observed declines and that horizontal physical transfers may have been of minor importance in some situations. Bierman et al. (1994), based on a three-dimensional spatial model of conditions in July 1991, concluded that reaction rates were of greater importance than horizontal exchange.

Extensive periods of anoxia occurred at C6A but not at 4, according to records reported by Rabalais et al. (1994). The intermittent fluctuations up and down at 4, they believe, resulted from physical exchanges with offshore water just a few kilometers south. They support this hypothesis by illustrating that salinity correlated well with oxygen fluctuations.

But what caused abrupt increases in oxygen, three of which were observed in the mooring record at C6A (Fig. 8)? Rabalais et al. (1994) could find little evidence that these increases were related to wind mixing. Bierman et al. (1994) suggest that light penetration was important because it enhanced deep water photosynthesis, which produced normoxic conditions. It is reasonable to suspect that both bottom and water column photosynthesis might produce net surpluses of oxygen during periods of increased light transmission in deep water, as they suggest. Indeed, Dortch et al. (1994) favored the idea that deep water and bottom photosynthesis prevented anoxia. While a good possibility, there is little experimental evidence as yet in support of significant photosynthesis. The simulation presented above demonstrated that doubling the combined photosynthesis would not elevate the bottom water above the hypoxia limit. More information is necessary on how these processes are controlled to settle this issue.


   THE EXTRINSIC FORCING FUNCTIONS VERSUS THE INTRINSIC PROCESSES: WHAT CAUSES HYPOXIA?
TOP
 ABSTRACT
 INTRODUCTION
 THE CONCEPTUAL MODEL: BOTTOM...
 RESULTS
 DISCUSSION
 THE EXTRINSIC FORCING FUNCTIONS...
REFERENCES
 
Nitrate loading (±100 mmol m-3) enhances primary production. At the same time, large volumes of freshwater increase the vertical density gradient, thus all but eliminating the transfer of dissolved oxygen from the surface to deep water. The forcing functions are therefore nitrate and freshwater.

There is a direct correlation between photosynthesis and microbial respiration in surface waters (Gardner et al., 1994; Biddanda et al., 1994; Pakulski et al., 1995). As yet, however, there is little convincing evidence that enhanced photosynthesis at the surface stimulates deep water respiration below the plume. Sediment oxygen consumption responds to temperature, but we cannot yet relate it to POC fluxes. In fact, when hypoxia hits, the SOC shuts down because respiration becomes oxygen limited.

Other factors may be important. The Mississippi River is known to contain high concentrations of dissolved organic matter (Lopez-Veneroni and Cifuentes, 1994; Trefry et al., 1994); this is consumed by bacteria (Chin-Leo and Benner, 1992; Gardner et al., 1994) that require large quantities of oxygen. Deep and bottom living organisms remineralize organic nitrogen (Gardner et al., 1994; Biddanda et al., 1994) and the inorganic nitrogen produced (Gardner et al., 1993) recharges nitrogen-limited phytoplankton photosynthesis (Bierman et al., 1994). Light penetration is inhibited by suspended clay particles in and near the river and by plankton blooms in the plume, and this diminishes the potential for photosynthesis in deep water. The relative contribution of these variables compared with the acknowledged forcing functions (nitrate and stratification) needs to be investigated.

The issue of hypoxia does not necessarily boil down to a simple standoff between enhanced productivity versus intensified density. There is little doubt that productivity is enhanced seasonally in surface water (Lohrenz et al., 1994), and that this enhances fluxes of particulates to the sea floor (Redalje et al., 1994). Although accumulation rate of sediments is high (Nelsen et al., 1994), this does not seem to accelerate bottom oxygen demand (SOC) much higher than what has been observed elsewhere along the Gulf of Mexico coastline (Rowe et al., 1995). Rapid burial does not consume oxygen; rather, it prevents it. Thus, if bottom water and bottom sediment oxygen consumption are not enhanced by particulate carbon fluxes that are the result of high nitrate loading, maybe the high nitrate loading does not matter all that much. Maybe the density stratification is what is important. This question needs an answer.

With some minor exceptions, the model uses zero-order flows for photosynthesis and respiration both on the bottom and in the water column. To improve the simulation, more information is needed on what controls deep water autotrophic and heterotrophic processes. This would lead to a better understanding of how photosynthesis in deep water responds to light, nutrients and autotrophic biomass. Likewise, information is needed on the relationship between respiration and the input of DOM and POM. The importance of nitrate-caused productivity to hypoxia in bottom water will not be convincing until the cause and effect relationship between enhanced POM input and respiration are demonstrated.


   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 THE CONCEPTUAL MODEL: BOTTOM...
 RESULTS
 DISCUSSION
 THE EXTRINSIC FORCING FUNCTIONS...
 REFERENCES
 





This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Citing Articles
Right arrow Citing Articles via ISI Web of Science (9)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Rowe, G. T.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Rowe, G. T.
Agricola
Right arrow Articles by Rowe, G. T.
Related Collections
Right arrow Water Quality
Right arrow Ecosystem Restoration
Right arrow Nutrient Management

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
The SCI Journals Agronomy Journal Crop Science
Vadose Zone Journal Journal of Plant Registrations
Journal of Natural Resources
and Life Sciences Education		 Soil Science Society of America Journal