You are on page 1of 14


DOI 10.1007/s10750-007-0643-4


Water column oxygen demand and sediment oxygen flux:
patterns of oxygen depletion in tidal creeks
Tara A. MacPherson Æ Lawrence B. Cahoon Æ
Michael A. Mallin

Received: 21 July 2006 / Revised: 24 July 2006 / Accepted: 26 January 2007 
Springer Science+Business Media B.V. 2007

Abstract Low dissolved oxygen (DO) levels from 0.0 to 9.3 g O2 m–2 d–1, and were positively
often occur during summer in tidal creeks along correlated with temperature, chlorophyll a, and
the southeastern coast of the USA. We analyzed total suspended solids, but negatively with dis-
rates of oxygen loss as water-column biochemical solved oxygen. Both forms of oxygen uptake were
oxygen demand (BOD5) and sediment oxygen flux seasonally dependent, with BOD5 elevated in
(SOF) at selected tidal creek sites monthly over a 1- spring and summer and SOF elevated in summer
year period. Ancillary physical, chemical and and fall. Average oxygen loss to sediments was
biological data were collected to identify factors greater and more variable than oxygen loss in the
related to oxygen loss. BOD5 rates ranged from water column. Oxygen deficits at three of five
0.0 mg l–1 to 7.6 mg l–1 and were correlated posi- locations were significantly related to BOD5 and
tively with organic suspended solids, total sus- SOF, but not at two sites where ground water
pended solids, chlorophyll a concentrations, discharges were observed. Correlation and princi-
temperature, and dissolved oxygen, and negatively pal component analyses suggested that BOD5 and
with pH and nitrate + nitrite. SOF rates ranged SOF responded to somewhat different suites of
environmental variables. BOD5 was driven by a set
Handling editor: K. Martens of parameters linked to warm season storm water
inputs that stimulated organic seston loads, espe-
T. A. MacPherson
Division of Water Quality: Aquifer Protection cially chlorophyll a, while SOF behaved less
Section, NC Department of the Environment and strongly so. Runoff processes that increase loads
Natural Resources, 127 Cardinal Drive Extension, of organic material and nutrients and ground water
Wilmington, NC 28405, USA discharges low in dissolved oxygen contribute to
occurrences of low dissolved oxygen in tidal creeks.
L. B. Cahoon (&)
Department of Biology and Marine Biology, Keywords Dissolved oxygen  BOD 
University of North Carolina Wilmington, 601 South Sediments  Tidal creeks
College Road, Wilmington, NC 28403, USA

M. A. Mallin Introduction
Center for Marine Science, University of North
Carolina Wilmington, 5600 Marvin K. Moss Lane,
Wilmington, NC 28409, USA Tidal creeks along the coast of the southeastern
e-mail: United States periodically experience low,


. 2004). Site description 1997). 2000. research were to determine the relative magnitudes tion in the sediments is commonly measured as of oxygen consumption by water column processes sediment oxygen demand (SOD). Dissolved oxygen concentrations integrate pro.. 1995).. chemical.. SOD is comprised (measured as BOD5) and substrate processes of biological sediment oxygen demand (BSOD) and (measured as SOF).. 1987). These mand (BOD).. Nutrient loading thos can drive significant oxygen utilization by often drives an increased demand for oxygen substrate communities (Hopkinson et al. (Middelburg et al. Bioturbation and irrigation of the substrata can also 2006). which support important biological and ecosystems. For these reasons sediment oxy- rates. more periodic low DO levels as a concern in these generally. a measure of the molecular oxygen observations indicated that primary production and (mg l–1) utilized during a specific incubation period aeration were insufficient to maintain oxygen con- (5 or 20 days to give BOD5 or BOD20. for the centrations at saturating levels throughout time and biochemical degradation of organic material and the space in these creeks and that oxygen consuming oxygen used to oxidize reduced forms of nitrogen. Water quality monitoring of tidal creeks in New duction and consumption processes. (Mallin et al. Extreme oxygen consumption leading to gen flux (SOF) is a more appropriate term for the hypoxia (<2 mg l–1 of oxygen) and anoxia can effects of substrate processes on water column DO reduce available habitat for fish and other aquatic concentrations than sediment oxygen demand life and lead to mortality of sessile organisms. Oxygen consumption in the water dard for these waters. 2005). and biological processes and. CSOD occurs farther from the sediment–water interface in a hypoxic– anoxic region where anaerobic bacteria degrade Materials and methods organic matter releasing reduced compounds that react with molecular oxygen (Rounds & Doyle. including as primary nursery Processes that reduce oxygen concentrations in areas for many species of fishes. the North Carolina stan- with the substrata. to compare the measured rates chemical sediment oxygen demand (CSOD). Oxygen consump.. and tends to Carolina Wilmington/New Hanover County Tidal occur close to the sediment surface or proximal to Creeks Program. (SOD).. DO levels in these tidal creek ecosystems may be partitioned into those creeks. Mallin et al. were fre- occurring in the water column and those associated quently below 5. located in New 123 . USA. 1991). 2004) has identified biochemical. possibly diffusive oxygen fluxes. The objectives of this respectively) (Eaton et al. 1998. 2005). (National Research Council.0 mg l–1. Hydrobiologia sub-saturating dissolved oxygen (DO) levels et al. and to identify factors controlling macrofauna burrows owing to diffusivity constraints oxygen consumption in these ecosystems. for the observed temperatures and salinities.. (Christenson et al. and below saturating values column is measured as Biochemical Oxygen De. Doyle. 1992) and greater threat when DO levels are low. Respiration by macroben- related to anthropogenic inputs. However. Rounds & Doyle. 1992). Some discharges of groundwater can also contribute to creeks also host periodic algal blooms (Mallin oxygen fluxes across the sediment–water interface et al. Water pumping by tran- 1980). 1997. SOD can be a significant percentage of the total oxygen uptake in aquatic systems (Caldwell & Five study sites were chosen in Futch Creek. 2000. recreational functions. Mallin et al. North Carolina. 2004). which likely affect oxygen demand (Simmons. providing Hanover County. Middelburg Hewletts Creek and Pages Creek. Nutrient cycling and interactions among drive significantly enhanced exchange rates of metals and sediments are heavily influenced by solutes. the ecological health of tidal creek ecosystems. since 1993 information about the balances of chemical. other processes complement (Lerberg et al.. and biological data is dominated by aerobic heterotrophs that utilize collected concurrently by the University of North organic material as an energy source. including oxygen and reduced compounds oxygen consumption or uptake (Medine et al. BSOD with other physical. Toxic trace elements may also become a sient pressure gradients (Huettel & Gust. processes were important. 1995. 2000. particularly the upper reaches..

North Carolina. Portions of the creek run along amplitude. 2000). Study sites in this creek were coliform levels (Mallin et al. adjacent to a natural springs. All sites dissolved oxygen levels in warmer months (Mallin experienced semi-diurnal tides with up to ~1. 2000) and numerous HC-3 (N 34.. (Mallin et al.19023 W 77. was in the upper southern branch.. Following the dredging lower portions of (Mallin et al. with a golf course nearby. chlorophyll a concen. Its drainage was mostly Study sites in Pages Creek (generally not highly rural. Study site FC-17 (N 34. of dissolved oxygen levels. and 77.Hydrobiologia Hanover County.30378 W private dock in the main channel of the creek. USA (Fig. USA 123 . based on data collected between 1993 experiences periodic algal blooms and has low and 2001 (Mallin et al. Hewletts Creek drains a heavily developed trations and nutrient (nitrogen and phosphorus) watershed that receives high nutrient loading and loading.85083). North Carolina.. 1996).1 m et al. roadways and receive run-off from golf courses Futch Creek has few algal blooms. areas that experienced the highest levels of fecal mote flushing and improve microbiological water coliforms. 2004). The mouth of impacted by nutrient loading) were located in Futch Creek was dredged in 1995–1996 to pro. 1998.. 1 Tidal creek study sites in coastal New Hanover County. 1).19025 W 77.. HC-SBPGR (N 34. 2004). chlorophyll a and nutrients in that creek quality. low fecal and suburban areas. the creek were re-opened to shell fishing (Mallin Sites were chosen to reflect a range in the values et al. 1998). Both of the sites chosen Fig. an anthro- which has shallow surface feeder creeks (<30 cm pogenically impacted site on the creek’s southern deep at low tide) and a small upstream spring tributary).86472).76422)..

allowing for continual mixing and thus borhood. was approximately ±0. Data from these the beginning and end of the incubation period. limited visibility some- BOD test described by Eaton et al. Initial and final samples were Dissolved oxygen fluxes extracted from each chamber into glass 60 ml BOD bottles using large syringes. 1972) in order to establish a rate of Sediment oxygen flux measurements were oxygen flux. & Parsons.. insertion into the sediment to a depth of 5 cm and stream site.14 g O2 m–2 d–1 mized disturbance artifacts and permitted opera. precision of SOF measurements sampling and incubation techniques as it mini. (1995). The upper branches of Pages Creek have uniform DO concentration within the chambers. A 2-h time sured in duplicate water samples collected in one. Individual rates were calculated for made using opaque benthic chambers deployed each of the three chambers at each study site as g sub-tidally on the rising tide at each study site each O2 m–2 d–1. Laboratory analyses of triplicate with a YSI Model 57 dissolved oxygen meter water samples from SOF chambers were con- (precision of ±0. Whirling cup rotors was adjacent to a private dock.80153) was stirring (<15 rpm) in the chamber in proportion to adjacent to a boat dock. as described by periodic hypoxia in the summer months and rare Cahoon (1996). Sediment visible macrofauna and groundwater discharges. kept on Collection bottles were iced in darkened coolers ice and brought back to the laboratory.28143 W 77.1 mg l–1). collecting water samples and deploying benthic Chambers were made of 16 cm diam PVC pipe chambers at each sampling time and site. Chamber placements deliberately avoided combustion procedures. Suspended sediments were analyzed in dupli- tion of natural processes that might affect SOF. experienced high sedimentation due to runoff and Clear 0.79417). Collected (local) on the incoming tide at each site once each samples were fixed in the field with Winkler month between July. samples were not used in subsequent calculations Dissolved oxygen levels were measured in mg l–1 or analyses.06 mg/l) and calculation oxygen consumed liter–1 d–1. near a natural spring. 1988). Final samples Biochemical oxygen demand (BOD5) was mea. external flows as slow as 0. This titration technique and the dimensions of the in situ approach was used in preference to core sampling devices. withdrawal of initial samples. The down. 2004). Hydrobiologia (PC-BDDS and PC-BDUS) have experienced water discharges around the rim. Cahoon et al.. and the upstream used the natural flow of the creek to drive slow site PC-BDUS (N 34. 2002. Based on the precision of the Winkler month between July.02 m s–1 (Cahoon.04 mg l–1 d–1 (Laws.27732 W 77. For the purposes of this study 123 . 2001 and August. respectively. ducted within 48 h of collection using Winkler tions made as needed. 1997). 1992. reagents (Strickland & Parsons. 1993). with a correspond. (TSS and OSS) were determined in duplicate by al habitats (Cahoon & Cooke. The opening was stoppered after algal blooms (Mallin et al. Although and brought to the laboratory for analysis within efforts were made to avoid contamination of DO six hours. 1998). Sediment cores and a 2 cm diam opening on the side to permit (10 cm and ‘‘fluff layer’’) were taken using 8 cm chamber placement without compression-driven diam PVC pipe. cate one-liter water samples collected at all study and followed techniques used to measure benthic sites monthly.. 2001 and August. taking sections 26 cm long with beveled bottom edges care not to disturb the chambers. 2002. samples were collected in the creek beds after but not less obvious infauna and slower seepages. procedures for non-standard volumes (Strickland ing precision of ±0. 1972). removed ~1. BOD5 was reported as mg titrations (precision = ±0. with salinity correc.5% of chamber volume. span was allowed between extraction of initial and liter NalgeneTM bottles between 6 and 9 AM final samples (Rounds & Doyle. filtration and gravimetry following drying and 1993).3 cm diam PVC tubing inserted into the development (Mallin et al. and along a tributary draining a residential neigh. Total and organic suspended solids primary production and respiration in other coast. times allowed sample withdrawal errors and air utilizing DO readings in 300 ml BOD bottles at bubbles in the final samples. Samples were processed using the 5-day samples by air bubbles. side of the chamber allowed for extraction of water samples. PC-BDDS (N 34.

ber 2001. Principal October (2001) and January and May (2002). with ~10 km from the study area (NOAA). Mean 0.0). ANOVA tests (SAS) were in color and texture than the remaining cored conducted using log-normalized BOD5 and SOF sediment and were more easily resuspended.uncwil. the lowest concentrations of ammo- normally distributed were log transformed using a nium occurred at HC-SBPGR (22.5 lg N l–1).01 m s–1. In general flow speeds Weather Service Eastern Regional Headquarters during sampling periods were greatest at the HC- website for the Wilmington International Airport.4 mg l–1) and the lowest at FC–17 (3.05.15 to 6 m s–1.4 Statistical analysis lg N l–1) and lowest at HC-3 (mean = 4. Grain Results size distributions were then determined with an LS 230 Beckman Coulter Particle Sizer. variables that could not the lowest concentration of orthophosphate oc- 123 . were regressed against indi- combustion procedures. using a YSI Model 85 meter prior to placing Mean chlorophyll a was highest at the HC- benthic chambers in the water column. Variables that were not BDUS site. and function of log10(X + 1). component analysis of all independent variables Percent organic content of sediments in well. be transformed to normality were analyzed by ters of substrate that. 1998). Site characteristics Water quality parameters Site characteristics for physical parameters. pH. using PROC UNIVARIATE or 58. Dis. temperature (C). Algal blooms using a Marsh-McBirney Flo-Mate Model 2000 (>25 lg l–1 of chlorophyll a) occurred at the flow meter with precision of ± 2% and a range of – HC-SBPGR site in June and July of 2002. DO) occurred at the FC-17 study site in Septem- pling and measurements were conducted. Significance levels also conducted using sediment cores that were for these analyses were set at p < 0. when cored. Pair-wise Sediment core samples were frozen for later correlations among parameters were identified analysis.Hydrobiologia ‘‘fluff layer’’ was defined as the top few centime.0 lg N l–1 for ammonium and 17.1 lg N l–1).12 m s–1 and 0. (5.3 lg l–1) and lowest at the velocity measurements were taken at each site HC-3 study site (2.1) 0700 local) were obtained from the National and lowest at FC-17 (9. BDUS in October 2001. transformed to nor- determined by gravimetry following drying and mality as necessary. SBPGR study site and lowest at well mixed. were different non-parametric tests.5 mg l–1). nutrients and suspended solids varied Nutrient and chlorophyll a data were obtained by seasonally and among sites (Table 1).6 lg P l–1 the Shapiro-Wilk test. Ammonium and orthophos- variables using JMP and SAS statistical software phate concentrations were highest at the PC- (SAS Institute. Mean DO the New Hanover County Tidal Creeks program concentrations were highest at the HC-3 study site (website: http://www. 2). chlo- rophyll a. 1980). Flow SBPGR study site (12. data to identify differences among sites. respectively. March. cology/TidalCreeks/Index. thawed. July and August 2002 and at PC- solved oxygen (mg l–1). Nitrate concentrations at the FC–17 study site were high compared to other sites (mean = 64. and the response mixed sub-samples of thawed sediment was variables BOD5 and SOF. heated and treated with a 30% hydrogen peroxide solution to oxidize organic matter before analysis (Folk. means of 0. were conducted for all for orthophosphate. DO concentrations were salinity and conductivity (mS/cm) were measured generally highest in the winter months (Fig.htm) for all sites at high Hypoxic dissolved oxygen levels (<2 mg l–1 of tide on the same day that oxygen demand sam. Overall mean concentrations were Normality tests.5 lg l–1). Sediment organic content and grain size using Spearman’s rank correlation procedure as were analyzed for samples taken in July and appropriate for non-normal data sets. Daily rainfall data (24 h total @ salinity levels were highest at PC-BDDS (32. was conducted using SAS. Grain size analyses were vidual principal components.

0 0.5) 25.3) 4. as mean (s.4) 15. The Tidal Creeks (Table 1).0) 16.0 (48.0) 2.5 (1.1) 25.7 (13.) Parameter FC-17 HC-3 HC-SBPGR PC-BDDS PC-BDUS Temp.6) 175. 19.0 (0.9 lg P l–1). with the highest levels occur.0 DO (mg l ) -1 6. mg l–1 50.03) 0.2) N/A 22. July 2001–August 2002.2) 11. .3 (2.2 (6.3) 3.4 (2. Chl a = chlorophyll a.2) 11. 10.6) 18.0) 6.5) DO.9) 10.1 (0. with means of 15.1) 5.7 (0. mg l–1 21.4 (1.5) 2.4) Sed.2) 10. g. period.9 (0.9 (11.0) Nitrate. Modal grain size of bottom and lowest at PC-BDDS (7.1) 58.5 (0.9) 2.0) 164.6 (9.3 (5.7 (7.3 (1.2 (0.4 (10. 123 . mg l–1 14.8 (1.02 (0.9) 21.0 2.4 (2.0 g.9) 12.0 mg l–1 for OSS. Hydrobiologia Table 1 Characteristics of tidal creek study sites. 2 Dissolved oxygen Seasonal Dissolved Oxygen Concentrations concentrations for each FC-17 HC-3 HC-SBPGR PC-BDDS PC-BDUS study site. pt ly Au ov ec l ct ar ri Au Fe Ju Ju Se Ja M O D N Ap M 2001 Date 2002 curred at HC-3 (6.7) 5.01) 0.06) 0. OSS = organic suspended solids Fig. mg l–1 5.5) 21. Organic HC-3 and PC-BDUS study sites were coarser suspended solids comprised 15% of the total than modal grain sizes at other sites. available.4 mg l–1.5 (6.2 (4. occurred at HC-SBPGR.9) 10.4 (58.8) 21. . TSS and OSS sediments changed very little over the sampling changed seasonally.2) 21.5 mg l–1 and 3.0 (6.8 (10.7 (17. ring in the summer and spring months. mg l–1 64.07 (0.4) 7.0 (6.3 (0.0 (5.5) 7.2 (1. The coarsest suspended solids measured in these tidal creek material.6 (5. Suspended solids were the lowest at Monitoring Program did not analyze ammonium HC-3.9) Salinity 9.8 (1.0 (7. Modal sediment grain sizes at the FC-17.4 (1.8) 4.0) 2.0 (8.4) 20. July 2001– August 2002 10. ne ay .1 (2.2) 8.2).0 (2.9) 4.6 mg l–1 for TSS and at the HC-3 study site.7 (1.2) 32. ch .02) Chl a. respectively percentages varied little over the sampling period. TSS and OSS concentrations PC-BDUS.9 (12.8 (0. n. 175 lm fine sand.5 (21.7) DO = dissolved oxygen.7 (8.52) 3.d.5 (6.5 (4.5 (6.5 (1.4) pH 7.6 (5. lm 144.05) 0.6) 2.4) 6.8) 5. lg l–1 2. TSS = total suspended solids.6 (0.3) Ortho-PO4.7 (78.2) 30. where mean TSS and silty mud. m s–1 0.9 (5.9 (9.0 4.4) 17.0 (9.06 (0.01 (0. mg l–1 3.9) TSS.0 8. Organic OSS were 30.5) Ammonium.3) 7.9 (2. so these data were not 2. Average DIN:DIP (dissolved inorganic Sediment characteristics also differed among N:ortho-P) ratios were highest at FC-17 (17. The finest modal grain size. b.5 (61.13 (0.6) 13.0 lm were highest at PC-BDUS.1 (1.0) OSS.2) Modal grain size. occurred at sites on average.1 (0. % organic 4.3 (24.9) 10.1) 21.3) sites (Table 1).3) 15.4) Flow.5 (10.2) 10.

the minimum rate of 0. Multiple outliers for BOD5 data for each site. df = 1. p = 0. total suspended solids. and were positively correlated with occurred in Futch Creek (FC-17) in July 2001 and temperature. sinks.59). df = 1.33.017) and SOF (F = 10. respectively.34. F = 0. df = 4. 3).3%) and lowest at PC- BDUS (2. although there were no significant Fig.8%).0 g O2 m–2 d–1 occurred a organic suspended solids and negatively with total of five times.26. outliers for SOF data for each site.60. and PC-BDDS sites combined. 4). and were positively correlated with tem- Mean SOF rates were significantly greater than perature.003) on dissolved oxygen deficits at the HC-3. SOF. but negatively mean BOD5 rates at all study sites when con.0 mg l–1 was found at the mid-creek station verted to common units (g O2 m–2 d–1) (Table 2). BOD5 rates were highest in spring and p = 0. correlated with dissolved oxygen (Table 3a).6 mg l–1 (Fig. Oxygen flux Rates of five-day biochemical oxygen demand (BOD5) for all sites over the entire sampling period varied from 0. p = 0.184).21.61 and 0.Hydrobiologia Mean organic content of bottom sediments was highest at HC-SBPGR (18. and TSS. and regressions for all five sites together. chlorophyll a. 5). July 2001–August 2002 regression of dissolved oxygen deficits against 123 . Concurrent measures of highest at PC-BDDS (2.16 g O2 m–2 d–1).3 g O2 m–2 d–1 summer.0 was a larger but more variable sink for oxygen to 9. HC-SBPGR.47). F = 0.89. (HC-3) in the lower part of Hewletts Creek in Coefficients of variation for BOD5 and SOF rates October 2001. (Fig. chlorophyll a. on dissolved oxygen deficits (mg l–1).0 mg 1–1 (Table 1). confidence intervals. 5). p = 0. Rates of SOF for all study sites varied from 0. were examined by plotting resulting values for each site and time (Fig.59. df = 1. July 2001–August 2002 SBPGR) in June 2002 and the minimum rate of 0. Mean rates were than BOD5 (Fig.005. and south branch study site in Hewletts Creek (HC. suggest- means were significantly different (1-way ANO. 6 a.92. The effects of BOD5 and SOF (g O2 m–2 d–1). were 0.65.67. Station mean rates ranged from 2. between November 2001 and April rates were higher in the summer and fall months 2002. ing that different factors controlled these oxygen VA on transformed SOF data.15 g O2 m–2 d–1) and BOD5 and SOF were not significantly correlated lowest at FC-17 (1. 4 Box plots of median.0 mg l–1 to 3. Overall mean BOD5 rates did not differ significantly among study sites (1-way ANOVA. Subsequent regression analyses showed that there were sig- nificant effects of BOD5 (F = 6. The maximum rate of 7.0 mg l–1 to 7. 1972) and mea- sured DO concentrations at the tidal creek sites. p = 0. df = 1.6 mg l–1 occurred at the Fig. F = 1.3 g O2 m–2 d–1 (Fig. at least once at each site except dissolved oxygen (Table 3a). but no site (r2 = 0. b). therefore. The maximum rate of 9. 3 Box plots of median. defined as the difference between the 100% DO saturation level for the corresponding temperature and salinity (Strickland & Parsons. confidence intervals. In contrast SOF PC-BDUS.31.

a. chlorophyll a. and PC-BDDS. HC-3. HC.001.15 ± 0.43 ± 2. 2. and dis- solved oxygen.54 F = 61.0001 Correlation analysis for all sites combined Pair-wise correlation analysis revealed many sig- nificant associations among the independent and response variables in the data set pooled for all sites and the entire 14-month sampling period (Table 3a).23 1. 123 . unpublished data). but negatively correlated with equation: temperature. Dissolved BOD5 and SOF for the three sites. and cumulative rainfall over 24 h. There Fig.16 F = 56. ware. Only the first four principal components BDUS. 6a.0001 HC-SBPGR 0.11 and 13%. data are means ± standard deviation and results of 1-way ANOVA on log-transformed data Study site BOD5 SOD ANOVA FC-17 0. FC-17 and PC.05). Significant groundwater discharges from were significant (P < 0. OSS.65 F = 42. total suspended solids and rainfall (Table 3b). df = 1.06 2.8.10 1. df = 1. and BOD5.0001 PC-BDDS 0. Chlorophyll a in turn was posi- tively correlated with TSS.77 and 30%. P < 0. Principal component 2 was most strongly Wilmington.07 1.1.93 F = 28. Principal Components Analysis (PCA) of deficits at those sites.27 ± 0. 5 Oxygen demand rates (BOD5 and SOF) in com. 2. and negatively correlated with pH. Hydrobiologia Table 2 Comparisons of mean BOD5 and SOF rates (g O2 m–2 d–1) over the sampling period July 2001–August 2002. In contrast. OSS. 48 h and 72 h prior to sam- pling. dissolved oxygen. pH.26. yielded the following a.14 ± 0. b) and tively. so it was inappropriate to try to identify cause-and- The intercept and coefficients were all signifi.0001 PC-BDUS 0.15 1. NC 28409. respec- values >4 mg l–1 at these two sites (Fig. P < 0. but a significant negative relationship between SOF and dissolved oxygen.19 ± 0. and TSS.31 ± 1.16 F = 43. UNC Wilmington Center for Marine Science.26. water temperature.0001 HC-3 0. and 2.15 ± 1. df = 1. TSS. P < 0.90 and occurrence of all the dissolved oxygen deficit 18%.2. df = 1. The correlation matrix demonstrated that DODeficit¼0:93þ2:30ðBOD5 Þþ0:34ðSOFÞ ð1Þ many variables co-varied in this data set. P < 0. TSS.5. as evidenced by components 1 to 4 were 4.25.80 ± 1. and SOF. there was no the 16 independent environmental variables ana- significant effect of BOD5 or SOF on oxygen lyzed was conducted using SAS statistical soft- deficits at the other two sites. and they collectively springs in the vicinity of these two sites likely had a explained 74% of the variability in the data set. July 2001–August 2002 SOF rates and temperature.05 and 13%. oxygen was positively correlated with chlorophyll SBPGR. Principal component 1 was most strongly hypoxic oxygen concentrations in groundwater driven by the variables temperature. df = 1.26 ± 0. BOD5 was positively correlated with temperature. indicating that both BOD5 and ses of the independent variables and BOD5 and SOF were important in creating dissolved oxygen SOF.24.3.16 ± 1. OSS. was a significant positive relationship between mon units for all sites.23. effect relationships from simple pair-wise analy- cant at p < 0. P < 0. chlorophyll a. chlorophyll discharges from nearby springs (Douglas Parsons. associated with nutrients and (negative) salinity. greater effect than BOD5 or SOF on water column Eigenvalues and percent variance for principal DO concentrations and deficits.

TSS = total suspended solids.19 –0.00 NIT 0.00 DO –0.04 0.17 –0.18 0.23 –0. NH4 = ammonium.44* 0.08 0.218 0.04 0.066 0.278 0.48* 1. revealed significant relationships between BOD5 nent 4 was most strongly driven by ammonium.147 0.18 0.30* –0.34* –0.40* 0.08 –0.296 0.18 0.204 –0.59* 0.00 OrgC 0. and principal components 1 (P < 0.00 OSS 0.107 0. Rain48.040 (L)Phos –0.30 –0.12 0.252 Cond 0.071 0.034 pH –0.197 0.005 0. Values sets of independent variables controlled the two of BOD5 decreased with increasing values of 123 .360 0.73* 0.154 Temp 0.41* 1.352 0. suspended solids and (negative) (P < 0. Chl a = chlorophyll a.40* 1.08 1.069 0.00 Rain24 0.149 (L)TSS 0. OrgC = organic carbon.324 (L)N:P 0.15 0.36* –0.35* 0.00 NH4 0.336 0.07 0.20 1.33* 0.0001): BOD5 = 2.31* Rain48 0.12 –0.30* –0.05 SOF = sediment oxygen flux.140 0.14 0.68* Parameter PC1 PC2 PC3 PC4 (b) Depth 0.04 –0.66* –0. Principal compo.217 0.03 0.77 1.259 –0.419 (L)Nit –0.14 0.21 –0.067 (L)NH4 0.06 0.343 (L)OSS 0. BOD5 = 5-day biochemical oxygen demand.320 0.016 –0.25* 0.10 0.20 0.158 –0.29* 0.21 0.36* 0.068 0.01 –0.51* Temp 0.34* –0.067 (L)DO –0.395 0. Sal = salinity.129 0.10 0. (b) Weightings of independent variables in principal components axes 1–4 SOF BOD5 Chl a NH4 NIT PO4 TSS OSS pH Sal OrgC Temp.12 –0.226 (L)Chl a 0.00 Chl a 0.063 0.06 0.02 1.07 0. Rain24.404 0.310 –0.27 1.22 0.12 0.21 0.61* 0.27* 0.35 –0.14 0. (negative) chlorophyll a and (nega.00 PO4 0.109 –0.069 –0.11 –0.11 –0.06 0.482 0. Regression analyses tive) organic suspended solids.15 0.06 –0.15 –0.09 1.21 –0.19 –0. OSS = organic suspended solids.20 0.18 0.215 –0.37* –0.308 Asterisks denote correlations significant at p < 0.06 0.033 –0. (a) BOD5 –0.03 0.28* 0.302 (L)Rain24 0.364 0.58* 0.342 –0.389 0.360 (L)Rain72 0.09 1.23 –0.310 0.16 0.390 0.34* Rain72 0. which were not by salinity.21 –0.00 0.0001) and 3 phosphorus.60* 0.00 pH –0.049 Sal 0. Values of BOD5 increased with positive BOD5 and SOF were regressed against values of values of principal component 1.57 + 0. 48 and 72 hours prior to sampling.59* –0.351 0.31 PC1 – 0.146 0.12 –0.56* 0. DO = dissolved oxygen.09 –0.340 0.02 –0. correlated with each other.253 –0.26 –0. NIT = nitrate + nitrite. with a strong each principal component to determine which seasonal increase from winter to summer.334 (L)Rain48 0.221 0.219 0.318 0.04 –0.00 Sal 0.57* –0.187 –0.09 0.05 –0.06 0.09 0.008 –0.25* 0.44* 0.56 cumulative rainfall totals.06 –0.50* 1.187 0.07 0.08 –0.26* 0.Hydrobiologia Table 3 (a) Spearman’s rho correlation matrix for all study sites using un-transformed data.69* –0. PO4 = orthophosphate. respectively Temp = temperature. and N:P = dissolved inorganic nitrogen:orthophosphate Principal component 3 was most strongly driven oxygen consumption variables.33* 0.09 –0.01 0.00 TSS 0.21 –0.10 0. Measured values of PC3.38* 0. Cond = conductivity.36* –0.206 –0.11 0. and Rain72 = cumulative rainfall in 24.18 0.105 0.

chlorophyll a. pers.21 g O2 m–2 d–1. Thus. Hydrobiologia chemical oxygen demand in the water column (BOD5). have been attributed to localized oxygen depletion (Richard Carpenter. SOF frequently exceeded BOD5 as a sink for dissolved oxygen. comm. prediction and management of oxygen depletion must consider sediment oxygen consumption as well as other factors in these very shallow and well-flushed estuarine ecosystems. 2006). creating some situations in which BOD5 was the dominant oxygen sink. Regression anal. respectively). e. TSS.. Principal components analysis indicated that BOD5 and SOF were controlled by different suites of environmental factors in these tidal creek Fig. These patterns suggested that BOD5 rose immediately following runoff events and that SOF Oxygen depletion in tidal creeks ecosystems can subsequently rose when plankton blooms depleted be attributed to a variable combination of bio. SOF increased with a variety of fresh and brackish water bodies (Mallin higher temperatures and was related to warm et al. Division of Marine Fisheries. with higher values of principal component 1. replacement of oxygenated surface waters by oxygen-depleted groundwater. nutrient levels and their organic seston loads 123 . 2005). and OSS that may be viewed collectively as warm season characteristics that support higher organic principal component 3. such as respiration by nekton.g. (b) Dissolved oxygen deficits for each study site versus SOF bination of temperature. which sion as chlorophyll a (Table 3a). but variability in SOF was substantially higher as well. and were lower in winter seston loads. representing a positive relationship oxygen demand are to be expected. Tidal creeks in this area typically when chlorophyll a levels and organic suspended exhibited higher densities of organic suspended solid concentrations were lower. but not at all sampling times and locations. BOD5. and advection of oxygen-depleted water from outside the tidal creeks. deaths of large schools of menhaden in adjacent ICW waters during December. matter during the warm season. but in a different implied a temperature effect similar to that driving manner than BOD5. 6 (a) Dissolved oxygen deficits for each study site ecosystems.038): SOF increased phytoplankton production (Mallin et al. so increases in water column 0.) and macrobenthos. Measured values of SOF increased weather algal bloom formation and their expres. but SOF was also positively correlated with some of the seston indicators that were important Discussion in PC 1. Average SOF for all locations and times was approximately 8 times higher than BOD5 (1.. partly as a result of ysis for SOF revealed a significant relationship rainfall-driven runoff and partly as a result of with principal component 1 only (P = 0.57 – 2004. Johnson. N. Chlorophyll a between untransformed values of SOF and values biomass makes strong contributions to the BOD of of principal component 1. 2005. and other factors not addressed in this study. oxygen uptake across the sediment– water interface (SOF).58 vs. 0.01 PC1.C. BOD5 responded positively to a com- versus BOD5.. (inverse log-transformed to normality) = 0.

1 3. a discharges are widespread in nearshore zones in variable frequently identified as a major determi. presumably. source of labile organic material (i. (1990) mean values or ranges.4 – Barranguet et al. shallow tidal variability in attempting to predict oxygen deficits. (1991) ‘‘E’’—estuary.64 – Bertuzzi et al.g. e. (1991) ‘‘C’’—coastal Onslow Bay. Cahoon & Cooke (1992). OR/USA/E 0.52–1. oxygen Roberts. Submarine groundwater summer (Mallin et al.5–2.85 – Hargrave et al. CA/USA/E 0. CA/USA/C 0.16–0. creating substantial chlorophyll a sinks and yielded generally lower dissolved oxy- concentrations.5 3. The ranges and means of values we complex set of driving parameters and temporal report here for relatively small. so effects of rainfall may (Mallin et al.07–1.01 – Cahoon & Cooke (1992) Chesapeake Bay/USA/E 0–1.15 – Dollar et al.2 Hopkinson (1985) of 1.64 – Kemp et al. nant of respiration rates and. did drive water deficits when groundwater discharges are signif- column and sediment oxygen fluxes in this study. creeks of southeastern North Carolina. Nutrient addition experiments have BDUS suggested that inputs of oxygen-depleted demonstrated that phytoplankton biomass groundwater by springs observed in those responds rapidly (one day) to nutrient inputs in locations obscured the effects of these two oxygen these tidal creeks. NC/USA/C 0. NC/USA/E 1.57 – Hofman et al.85 – Reay et al. particularly at those two sites. suggested that groundwater inputs may have also where sediment organic loads are likely to be reduced dissolved oxygen concentrations in tidal lower than in estuarine habitats.06–0. Netarts Bay.. which are in turn generally some- vicinity of two of the study sites noted above what higher than values from coastal ecosystems. Tomales Bay. The (1985). site- but with different combined effects when other specific factor in addition to BOD5 and SOF. 2004. Thus temperature.36 Herndl et al. (1992) Albufera.Hydrobiologia settled to the bottom to become an organic source of BOD5 and SOF values at sites FC-17 and PC- for SOF. (1983) in carbon units converted Chesapeake Bay/USA/E 1. are The low dissolved oxygen content of ground. 2005). Hopkinson creek water.0–3. similar to those reported from most other estua- water discharged from natural springs in the rine systems.1–4. aquatic ecosystems generally (Simmons.2–19. and thus may lag correlated with water temperature in these creeks surface runoff events. 2004). icant will have to consider this physical. NC/USA/E 0–9.68 This study Tidal creeks.e. (1995) Gulf of Fos/France/C 0. particularly during late spring and gen values at these sites. Models of oxygen demand in these shallow coastal habitats highlight several additional points ecosystems must therefore take into account a (Table 4). (1989) estuarine and coastal Southwest lagoon/New Caledonia/C 0.58 0.7 – van Es (1982) (WC) (g O2 m–2 d–1.93–1.0) reported from Gulf of Trieste/Yugoslavia/C 1. Chlorophyll a biomass Groundwater discharges are likely driven by and phytoplankton production are both positively longer-term rainfall patterns.. Concurrent failure of a multiple regression model to predict measures of SOF and water column oxygen dissolved oxygen deficits based on a combination consumption in other ecosystems frequently show Table 4 Values of Location/Country/Habitat SED WC Reference sediment oxygen demand (SED) and water column La Jolla. providing a vary considerably among basins.08–1. Oosterschelde/Netherlands/E 1.55 – Lo´pez et al.3 0–0.21 This study 123 .4 0. 1992. (1996) Gulf of Trieste/Italy/C 0. organic Comparisons of the results of this study with suspended solids) to stimulate water column those of studies in other estuarine and shallow BOD. (1995) Chesapeake Bay/USA/E 0.33–0. (1997) Tidal creeks..1–0.51 – Davis & McIntire (1983) values originally reported Bay of Fundy/Canada/E 0. Johnson. USA. variables are considered.92 – Boucher & Clavier (1990) marine habitats as overall Roskjilde Fjord/Denmark/E 2.02 3.5–3. 2002) so efforts to predict oxygen demand in tidal creek ecosystems.8 Sand-Jensen et al.1 Boynton & Kemp (1985) using respiratory quotient Georgia Bight/USA/C 1.70–1.17–0..84 – Hartwig (1978) oxygen consumption Ems-Dollard/Netherlands/E 0. Majorca/Spain/E 0.

1996. Contribution of benthic ically driven respiration as a major control of biomass to overall metabolism in New Caledonia water column oxygen levels. which contributes to organic suites of environmental variables. tion of parameters. e. Dargan ation. Marine Ecology Progress Series 64: how important sediment oxygen fluxes. M. J. in 1996. and Mrs. and oxygen in the Gulf of Trieste (Northern Adriatic). 1998. as with consumption processes. The relatively frequent occur.. Many studies demonstrated the importance Conclusions of temperature as a controlling factor for sedi- ment and water column oxygen consumption. as noted by Mallin and Corbett (2006) in the French Mediterranean coast. ecosystem respiration. Sand-Jensen et al. Microphytobenthos production in the Gulf of Fos. Brambati. 2004). inflows of oxygen-saturated rain water. A. highlights the importance of biolog. SOF is thus particularly need to be taken into account. and 1999. highlighting the role of shallow depth in the relative importance of SOF in this study. Aer- Dr. Alivon. more complex. are driven by Gina Root. Lynn Leonard for laboratory guidance. & J. 1997.and seasonal-scale variability. and time Acknowledgements Funding for this project came from the Glaxo Wellcome Fellowship in Oceans and Human lags between runoff events and SOF responses Health. Dr. always. G. and. particularly in suspended solids. 123 . Herndl et al. Boynton & Kemp consequences of rain events. Boucher. Virginia Johnson.S. Paul Hosier and Mr. Bryan Bishop for assistance in the field and laboratory. Oxygen depletion in these and less responsive ecosystems. which struck Benthic fluxes of dissolved inorganic carbon. covaries with other (SOF) in small tidal estuaries demonstrated factors whose effects are more difficult to identify strong variability in response to complex interac- than temperature alone in larger. Water. Factors driving these oxygen sinks include Results from this study suggest that models for seasonal variation in primary production (mostly predicting BOD5 and SOF must incorporate phytoplankton). however. ecosystems (Mallin et al. nutrients the southeastern U. such as large important ecosystems is driven by both sources. the Graduate School ecosystems also reflect oxygen inputs as well as at the University of North Carolina Wilmington and the New Hanover County Tidal Creeks Program. estuaries or coastal ocean waters. access to sampling sites through their property. tudes of these competing processes determine the extent in space and time of hypoxic events in tidal creek ecosystems. water column and benthos.g. References rence of hypoxia in warmer seasons and after major rain events observed in these tidal creek Barranguet. Bonnie. (1989). Scott Ensign. a gen Demand (BOD) and Sediment Oxygen Flux strongly seasonal variable. 2000. of Science. Groundwater inputs It is especially difficult to create simple models to can alter or obscure the SOF signal in these land– predict SOF rates because numerous variables water interface habitats. Doug Parsons. C. there is high hard to predict using single effects models. Dawn Carroll York.. We thank the consumption processes discussed above. Hydrobiologia 333: 181– 193. but not and highly responsive tidal creek ecosystems. The magni. and. C. 1998. respectively. and Alex Croft. a Bryden grant from the North Carolina Academy may obscure causal relationships. Air and Soil Pollution 99: 305–314. (1990). Davie for allowing producers all contribute to oxygen content. the Patterns of oxygen availability in tidal creek Department of Biological Sciences. and Floyd. which 271–280. Cape Fear River Estuary following flooding by Bertuzzi. and are highly related to tem- small tidal creek ecosystems that are tightly perature and run-off driven responses in the coupled to both terrestrial and marine influences. and oxygenation by planktonic and benthic primary Mrs. variability for all water quality parameters mea- sured in dynamic tidal creek systems. the UNCW Center for Marine Science. Simultaneous measurements of Biochemical Oxy- Analysis of our data showed that temperature. 1990.. Faganeli. This study shows lagoon sediments. Dr. Hurricanes Fran. can be in total (1985). R. Plante-Cuny & E. and Frierson for assistance with statistical analysis.. Clavier. than in small usually more so by SOF than by BOD. Hydrobiologia relatively higher areal oxygen consumption in respond significantly to temperature and the deeper water columns. Heather CoVan Wells and event. Welter & A.

X. Mallin. Baltimore. 2000. 137–152.. M. 1985. J. L. E. chambers and flume flows. Johnson. Benthic R. Merritt. E. Meritt. Wilmington. change processes. E. R. Jr. B. United States Geological Survey Water-Re. Marine Ecology Progress Series 85: sources Investigations Report 95–4196: 1–14. Alphin. In nutrient cycling at Gray’s Reef. 1992. & C. Hydrobiologia 160: 193–198. F. 73 pp. 1987. T. 1991. Benthic microalgal production at versity of North Carolina Wilmington.. associated algae.. M. Austin. Fanuko. H.C. 1983. Oregon. SW Netherlands). and contribution of the Mallin. Sanger. production in Onslow Bay. 1983... Sizemore. Clesceri & A. L. CMS Report 98–03: 1–76. F. & G. & C. Lowe. B. Marine Ecology Progress Series 23: Hopkinson. 1980. L. New Hanover County Tidal Creeks. F. V. The Uni- McDonald. 1996 Huettel. sediments has major impact on the oceanic nitrogen Lerberg. B. Herndl. 1995.. Williams & E. A. L. H. Estuaries and Coasts 29: 1046–1061. J. M. A. Merritt. W. M. G. S. J. L. Prouse. E. 182 pp. J. benthic chambers. R. S. C. D. Parsons. P. L.. N. M. Hollibaugh. A. W. J.. Corbett. J. Seasonal depletion of oxygen Caldwell. D. K. and different coastal systems: How hurricane attri- ogy Progress Series 79: 115–125. Texas. L. Hubertz.. Tomales Bay. A. Marine Ecology Progress Cahoon. rates measured with microelectrodes and bell jars: Mallin. University of North Hargrave. Parsons & T. Marine dredging. A. 1–67. Use of whirling cup rotors to stir shore Georgia Bight. L. Primary productivity by phytoplank- Ecology Progress Series 84: 185–196.. 1993. Manock. North Carolina. A. E.. Annual cycle of benthic nutrient fluxes in and Freshwater Research 46: 45–53. S. B.. Spain). J. Holland & D. G. 2006.. M. H. M. 1991. C. D. Cen- Hemphill Publishing. A. Jansson & Cahoon. B. Sediment 103–114. C. Water quality in Folk. & C. 1985. J. 1995. M. Restoration community metabolism and microbial dynamics in the of shellfishing waters in a tidal creek following limited Gulf of Trieste (Northern Adriatic Sea). Ensign & T. M. L. Fallon. B.3–5. D. Petrology of Sedimentary Rocks. A. J. & M. 5. Applications to oxygen budgets in estuarine inter. Eaton. Obreski & J. Schubauer. Alphin. Marine Ecology Progress Series Techniques of Underwater Research. 1995–1996. C. V. Journal of Exper- coefficients for oxygen and oxygen consumption imental Marine Biology and Ecology 298: 211–231. P. C. North Carolina tidal creeks. Apparent sediment diffusion blooms in urbanizing tidal creeks. P. Johnson. and New Hanover County watersheds. 838–853. 1): 229–243. Cahoon. Vidal. 1992. USA. Massachusetts Bay. L. 2006. 1996. butes determine the extent of environmental impacts. J. Greenberg. Lluch & J. Marine Ecology Progress Series 13: Lo´pez. Cooke. 1996. Smith. Vink. S. Neame. M. and benthic communities at two intertidal sites in the M. J. oxygen consumption by sediments along an estuarine Marine Ecology Progress Series 69: 261–272. O. Wastewater. Stellwagen Bank. An instrumented lander for mea. Wagenvoort & & H. upper Bay of Fundy.S. A. V. 2004. Marine Biology 46: 283–293. Nutrient regeneration and tidal sediments (Ossterschelde.. Gust. 1997–1998. Aquatic Pollution.-O. P. C. budget. 1989. siliceous sediment. B. spatial and tidal variability in two Cahoon. D. Proceedings of 73: 105–120. Sampou.. Macpherson. Massachusetts.C. P. M. Nutrient limitation and algal A. Posey. Carolina at Wilmington. Global Biogeochemical Cycles 1: 91–116. 2000. R. metabolism: evidence of heterotrophy in the near- Cahoon. N. W. Factors affecting respiration and Center for Marine Science. Shallow-water benthic and pelagic 45–55. J. Effects of physical to the effects of watershed development. Morgui. Christenson. A. R. Devol & L. & W. E. Phillips & P. Laws. J. Mallin. A. Hartwig.. 45–52. 611 pp.. John Wiley and Sons Codispoti. R. Manock.. Sizemore & K. McIver Hofman. J.. M. Williams. B. 2005. J. G. A. Denitrification in continental shelf Inc. M. Marine Ecol. the American Academy of Underwater Sciences. L. benthic and planktonic respiration and physical ex- 1994. E. Tuttle & Ecology (Progress Series) 102: 179–185. Murray.. American Academy of from confined sediment cores in stirred benthic Underwater Sciences. S. Thesis. Marine Biology 87: 19–32. United Book Press Inc. a hard bottom habitat Lang. CoVan. Cahoon. T. A. Boynton. A. Primary production and respiration in pelagic Mallin.. A. D. A. Factors contributing to hypoxia 123 . 1992.D. J. J. J. B. McIntire. metabolism in a transitional continental/marine area: The Dollar. Journal of Coastal Research 16: 40–47. A. Thomas & A. California. Solute-release mechanisms Scientific Diving Symposium. R. S. Methods and in the Georgia Bight. Ecology Progress Series 53: 169–178. Cahoon. ter for Marine Science Research. A. & J. Doyle. Canadian Journal of Fisheries Wheeler.. Community metabolism and surement of benthic respiration and production. Beretich Jr. Responses of tidal creek macrobenthic communities Davis. F. 1978. ton: Temporal. J. D. A. Sediment oxygen from bottom waters of Chesapeake Bay—roles of demand in the Lower Willamette River. Hopkinson.. S. B. University of North photosynthesis by the benthic community of a subtidal Carolina-Wilmington. G. T. 1993. M. K. Marine Johnson. P. de Jong. Baldwin (eds). Marine Kemp.. 1995. S. A.H. L. R.A. Marine 1991. K.. M.Hydrobiologia Boynton. salinity gradient. Maryland. M. Posey. Kemp. W. 1988. W. Standard Methods for the Examination of Water and Mallin. P. Peduzzi & N.. Albufera of Majorca (Balearic Islands. S. S. New York. Benthic microalgal Series 82: 187–197. L. Nahant. 1998.6. Sandee. M. Environmental quality of Wilmington and Aquatic Sciences 40(Suppl. Multiple hurricanes benthos to total ecosystem metabolism. Estuaries 23: gradients on the production dynamics of sediment. Garber.

H. V.S. K. USA. C. D. Thesis. 1982. Williams (eds). M. Roberts. R. 1992. to material flux across sediment/water interfaces in National Research Council. LeB. U. 405 pp.C. Duarte & J. L. Adams. Porcella & D.. nearshore environments of the lower Delmarva pen. 1972. In Del gress Series 65: 63–72. S. Washington. P. 310 pp. 1980.S. Technical Information Investigations Report 97-4103: 1–19. Jensen. marine environments. North Carolina. Heavy. Strickland. D. L. Center. Community metabolism of intertidal flats insula. B. Simmons. J. Hydrobiologia in rivers. L. A Practical Reay. Marine Biology 66: 95– 251–227. Limnology and Ocean. demand in three-phase aquatic microcosms. 1992–96. Geological Survey Water-Resources cosms in Ecological Research. Medine. versity of North Carolina at Wilmington. M. 63 pp.-P. G. 2005. A. Gattuso. 2002. Sediment oxygen metal and nutrient effects on sediment oxygen demand in the Tualatin River Basin. Gallagher & G.. D. J. 206–224. 1997. lakes and streams. 1995. F. D. W. National Academy Press. Van Es. C. during a late summer period. M. Marcher & M. & P. Handbook of Seawater Analysis. Sand-Jensen. Clean Coastal Waters. Importance of submarine tion in Aquatic Ecosystems. (USA) 52: 279–303. U. Hanover County. Oxford University Press. S. 2000. & T. Respira. A. A. M. Jr.. J. Marine Ecology Progress Series Understanding and Reducing the Effects of Pollution. Oregon. groundwater discharge (SGWD) and seawater cycling New York. Marine Ecology Progress Series 118: in the Ems-Dollard Estuary. 108. Micro. G.. Fisheries Research Sediment–water column oxygen and nutrient fluxes in Board of Canada.S. M. T. Marine Ecology Pro- Respiration in coastal benthic communities. Ottawa. Chemical constituents in the Peedee and Castle Hayne aquifers: Porters Neck Area. J.. Parsons. Doyle. Hansen. Rounds. Pelagic metabolism in eutrophic coastal waters Middelburg. Giorgio. New 123 . Department of Energy. 84: 173–184. Simmons Jr. 1990. Uni- ography 51: 690–701. & M.