Instream Swamps and their Effect on Dissolved Oxygen Dynamics on Blackwater Coastal Plain Streams

Figure 1. Map of the Little River Experimental Watershed (LREW) and the two sampling locations.

Figure 1. Map of the Little River Experimental Watershed (LREW) and the two sampling locations.

One of the hypothesized reasons for the low DO levels in blackwater streams is the slow movement or water and extended contact with the underlying sediments in the many instream swamps which characterize blackwater systems in the coastal plain.  This project investigated the role of these instream swamps on a watershed scale and specifically the influences of sediment oxygen demand (SOD), the distribution of organic sediments and the role of hydrology on water column DO concentrations.  Results support the idea of extended travel times through these instream swamps due to tortuous flow pathways and extensive transient storage.  SOD values were measured well above previously published values and were correlated with increasing organic carbon content. Results show that these organic sediments are widespread and become more prevalent in higher order streams.  While DO dynamics are a complicated mix of natural and anthropogenic factors, instream swamps play a critical role in overall watershed oxygen dynamics and lend support to these systems being naturally low in DO.

Blackwater streams are characterized by low gradients, high summertime temperatures, and extensive inundation of surrounding floodplains.  Typically lasting from winter to early spring, the long inundation period creates a multitude of instream floodplain swamps that play a vital role in overall water quality.  A key influence on DO levels within these floodplain swamps is sediment oxygen demand (SOD), a critical and dominant sink of oxygen in many river systems that is often poorly investigated or roughly estimated in oxygen budgets.  In a modeling study described on a companion page, SOD was found to be the most sensitive model parameter.

Methods – SOD
Figure 2. Measure of Discharge and Dissolved Oxygen for 2007 at Station B, the outlet of LREW showing that flow went to zero on Julian Day = 128 (7/8/2007) and did not resume for the remainder of the calendar year.

Figure 2. Measure of Discharge and Dissolved Oxygen for 2007 at Station B, the outlet of LREW showing that flow went to zero on Julian Day = 128 (7/8/2007) and did not resume for the remainder of the calendar year.

SOD was measured in two instream / floodplain swamps in the 334 km2 Little River Experimental Watershed (LREW).  While land use is primarily agricultural, riparian vegetation remains largely intact along portions of the river, with swamp hardwood communities consisting of a closed canopy and thick undergrowth.  The instream swamp selected for the majority of this experiment is a 1550 m long stretch of river located in the lower part of the LREW above the gauging site designated Station B (Figure 1).

The stream at this point is a 5th order stream and can be as wide as 350 m during periods of complete inundation.  Inundation of the floodplain usually begins in December with complete inundation until April or May.  During summertime months, flow may stop along with complete drying of the river channel (Figure 2).

Additional measurements were made in a small headwater watershed (3rd order), with a main channel 5-10 meters in width and subject to overbank flooding during high flow events and cessation of flow during summertime months.  This site was located just above the gauging site designated Station J (Figure 1).

Figure 3. Schematic of the oxygen demand chambers used in this study. Oxygen concentration within the chamber was measured using an oxygen optode (Aanderaa Instruments Oxygen Optode 3975) logging to a PDA (Dell Axim X50).

Figure 3. Schematic of the oxygen demand chambers used in this study. Oxygen concentration within the chamber was measured using an oxygen optode (Aanderaa Instruments Oxygen Optode 3975) logging to a PDA (Dell Axim X50).

SOD was measured using three experimental chambers and one control chamber originally designed by Murphy and Hicks (1986) and modified by Utley et al (2008).  Each chamber has a volume of 65.15 liters and covers a surface area of 0.27 m2 on the stream bottom (Figure 3).

The cutting flange sinks the chamber 5.08 cm into the stream sediment.  Water circulates throughout the chamber via a 12 volt DC submersible pump, powered by a 14 volt submersible, gel-cell, lead acid battery.  The pump continuously withdraws water from one diffuser and injects it back into the chamber via the second diffuser.  The diffusers force the water within the chamber to circulate around the chamber annulus, promoting continuous mixing.  The control chamber differs from the experimental chamber by having a sealed bottom which is used to measure water column respiration.  Oxygen concentration was measured within the chamber using an oxygen optode (Aanderaa Instruments Oxygen Optode 3975) logging to a handheld computer (Dell Axim X50) (Figure 4).

For each sampling event, SOD chambers were pushed into the sediment until sealed and then oxygen and temperature levels logged within chambers every

two minutes for 2-3 hours.  SOD data were collected a total of 33 times over 11 dates.

Figure 4. Typical deployment of an SOD chamber. Deployment times were typically 2 to 3 hours.

Figure 4. Typical deployment of an SOD chamber. Deployment times were typically 2 to 3 hours.

SOD is calculated as the decline in oxygen concentration over the elapsed time.  SOD was calculated using:

sod formula

 

 

 

where:

SOD = the sediment oxygen demand (g O2 m-2 day-1)
b1 = the slope from the oxygen depletion curve (mg L-1 minute-1)
b2 = the slope from the oxygen depletion curve of the control chamber (mg L-1 minute-1)
V = the volume of the chamber (L),
A = the area of bottom sediment covered by the chamber (m2)
1.44 = a units conversion constant (Caldwell and Doyle 1995)

Methods – Organic Sediments

Soil samples were taken immediately following an SOD run with one sample taken from each quadrant under a chamber.  Soil samples were taken to a depth of 15 cm with each set of four cores separated into two depth classes (0-5 cm and 5-15 cm, 5.7 cm diam.).  All four cores from each depth class were composited into a single plastic bag.  After drying, samples were rolled with a rolling pin to crush soil and sieved with a #10 sieve (<2 mm).  Material that passed the sieve was considered soil matter with that remaining above as litter/residue.  The samples were analyzed for extractable TN (mg kg-1), extractable carbon (mg kg-1), litter organic carbon (LOC) (mg g-1), soil organic carbon (SOC) (mg gtotal-1), and total organic carbon (TOC) (mg gtotal-1).

Results – SOD

Truax et al. (1995) stated that SOD rates for Southeastern United States rivers range between 0.33 – 0.77 g O2 m-2d-1.  Our SOD values ranged from 0.49–14.19 gO2m-2d-1.  All but one of our SOD measurements were higher and, in some cases, much higher (up to 18 times) than the published range.  The average SOD rate across all samples was 5.37 g O2 m-2d-1, a rate seven times higher than the upper limit of the published range.

In addition, a previous study measuring SOD within forested and agricultural catchments of the LREW which is discussed on a companion page found SOD rates between 0.6 – 1.4 g O2 m-2d-1 in the agricultural catchment and 0.9 – 2.5 g O2 m-2d-1 in the forested catchment (Crompton 2005; Utley et al. 2008).  75% of the measurements made in the instream swamp were above the highest value recorded during the previous study.  These results indicate that SOD plays an even greater role than previously believed in the LREW.  This is the first time SOD has been measured in instream swamps and our results suggest that instream swamps are areas of intense oxygen demand and are a major factor in the oxygen balance of the watershed as a whole.  Just below is an illustration of the potential effect of these SOD rates on DO.

Envision a square meter of sediment overlain by water 0.61 m deep with an initial DO concentration of 5 mg L-1.  There is a total of 3050 mg of O2 available in this cube of water.  Assuming no reaeration, photosynthesis or movement of water, available DO within the water column would be consumed in just over a half a day using the average SOD rate of 5.37 g O2 m-2d-1 calculated in this study.  Using the maximum rate of SOD measured in this study (14.19 g O2 m-2d-1), DO is depleted from the hypothetical water column in 0.215 days.  While water movement, reaeration and photosynthesis are rarely zero, this hypothetical example shows that SOD is a major sink of DO in blackwater river systems instream swamps.  If the water remained in the swamp for more than half a day, all the DO available in the water would be consumed by SOD.

Results – Organic Sediments

Multiple soil properties were significantly correlated with SOD rate with all significant relationships occurring in the 0-5 cm depth fraction.  SOC and TOC in the 0-5 cm depth fraction were the best significant predictors of SOD rate within the instream swamp system, both explaining 35 percent of the variation.  The reason for their similar relationship is likely due to the majority of the TOC being made up of SOC.  In both depth classes around 80 percent of the TOC is compromised of SOC.  Additionally, while SOC makes up a similar percentage of TOC at both depth classes, the average concentration within the 0-5 cm depth class is over four times higher as compared to 5-15 cm fraction.  This is not surprising as the freshest organic material is deposited on the surface sediments and is decomposed as it works into the soil column.


Todd, M.J., G. Vellidis, R.R. Lowrance, and C.M. Pringle. 2007. Measurement of water residence time, flowpath, and sediment oxygen demand in seasonably inundated floodplain swamps of the Georgia Coastal Plain. In A. Saleh (ed) Proceedings of the 4th Conference on Watershed Management to Meet Water Quality Standards and TMDL, ASAE, St. Joseph, Michigan, pp 140-147.  Download pdf

Todd, M. J., G. Vellidis, R.R. Lowrance, and C.M. Pringle. 2009. High sediment oxygen demand within an instream swamp in southern Georgia: Implications for low dissolved oxygen levels in coastal blackwater streams. Journal of the American Water Resources Association (JAWRA) 45(6):1493-1507.  Download pdf

Todd, M.J., R.R. Lowrance, P. Goovaerts, G. Vellidis, C.M. Pringle. 2010. Geostatistical modeling of the spatial distribution of sediment oxygen demand within a coastal plain blackwater watershed. Geoderma 59(1-2):53-62, doi:10.1016/j.geoderma.2010.06.015.  Download pdf

Download Jason Todd’s Ph.D. Dissertation

Jason Todd, Post-Doctoral Research Associate, Dept. of Environmental Engineering, Princeton University.
George Vellidis, Professor; Biological and Agricultural Engineering, University of Georgia.
Richard Lowrance, Ecologist, Southeast Watershed Research Laboratory, USDA-ARS, Tifton, Georgia.
Catherine Pringle, Professor, Odum School of Ecology, University of Georgia.