Vernon Brock

University of Washington

School of Oceanography

7538 42^nd Avenue NE

Seattle, WA 98115

vbrock77@hotmail.com

https://members.tripod.com/vbrock77

7 June, 1999

The source and fate of terrestrial and oceanic Suspended Particulate Matter (SPM) in Hood Canal, Puget Sound, WA has been evaluated using stable carbon and nitrogen isotopes and elemental C/N ratios. Terrestrial and marine end member d ¹³C, d ¹⁵N and C/N ratios (respectively: terrestrial –26.88 ± .93, 4.02 ± .95, 19 ± 5.02; marine –21.26, 6.79 and 7.86) were distinct. The fractional contribution of terrestrial and marine derived organic matter was determined using a two end member mixing equation. The isotopic biological alterations of d ¹⁵N and C/N are large enough to ‘overprint’ the terrestrial and marine end member values. Production was assessed by assuming a two end member mixing model and calculating the fraction of SPM derived from production in the surface waters and terrestrial riverine input. Production accounts for >83%, 100% and >93% (using d ¹³C, d ¹⁵N and C/N, respectively) at all stations and depths with only a few exceptions. Vertical trends in d ¹³C, d ¹⁵N and C/N ratios show organic matter production and degradation with depth. d ¹³C is used, along with d ¹⁵N and C/N ratios, as an indicator of productivity and degradation and are indicators of an extremely productive estuary.

Tefry et al. (1994) argued that the knowledge of the sources, transport and fates of biogenic carbon onto the Louisiana Shelf is critical to understanding the development and persistance of bottom hypoxia, as well as the overall cycling of carbon. Results from a similar study conducted in Hood Canal, WA are presented here. The effects of observed levels of SPM on the development and persistance of bottom hypoxia, as well as the overall cycling of carbon, are discussed.

Hood Canal is a narrow, highly stratified estuary in the western edge of Puget Sound, WA (Fig. 1) which has sluggish circulation and produces characteristics similar to that of classic fjords (Paulson et al., 1993). The estuary is between 2 and 4 km wide throughout most of its length, and is separated from the main basin of Puget Sound by two deep sills (50 and 75 m). In the main basin of Hood Canal, depths reach a maximum of 175 m and tidal currents produce little net advective transport (~2 cm s^-1; Paulson et al., 1993). The main source of freshwater input is from the Skokomish River at the head of the estuary, several smaller rivers enter the western edge from the Great Bend to Dabob Bay (Fig. 1).

Much of the shoreline surrounding Hood Canal is rural in nature and with vacation homes served by septic tanks. The only National Pollutant Discharge Elimination System is located in Lynch Cove (Fig. 1) and serves a resort inn (Paulson et al., 1993). Levels of dissolved oxygen (DO) have declined from roughly 4 mg/l to presently ~2.5 mg/l since the construction of the Cushman #2 dam in 1930 (Fig. 1) (Albertson et al. unpublished literature). The distributions of benthic organisms (eg., Azinopsida serricata) reported in the Great Bend of Hood Canal are symptomatic of a community in stress with over half the organisms being species indicative of low DO conditions (Llanso, 1998).

The dam delays, by about 6 mo, approximately 40% of the freshwater flowing down the Skokomish River (Jay, 1996). Output form the dam is therefor very significant in monthly and seasonal averaging of Skokomish River output (Albertson et al., unpublished literature). Albertson et al., (unpublished literature) investigated the effects of stratification caused by the unnatural flow patterns of the Skokomish River caused by the Cushman #2 dam. By analyzing a three layer box model, DO levels at depth were found to decrease with an increase in stratification. Periodically, significant exchanges of relatively high oxygenated, oceanic seawater flush out much of the estuary. Such an event did occur in 1998 in Hood Canal, but DO levels quickly dropped back to prior levels (J. Newton, pers. comm.).

The fact that the DO levels quickly receded to low levels suggests that another process besides stratification is playing a significant role in controlling the DO levels in Hood Canal. Eutrophication (excessive nutrient loading causing high productivity at the surface resulting in increased respiration at depth) has been presented as one reason why observed DO levels have quickly declined in Hood Canal. Mackas and Harrison (1997) indicated that the most sensitive areas in Puget Sound to the effects of eutrophication are small tributaries and fjords that have slow flushing rates and adjoin urbanized shorelines. Most of these regions are in the south to west margins of Puget Sound. However, Jay and Simenstad (1996) found that the Skokomish River has not suffered major changes in nutrient dynamics and therefore eutrophication is not evident.

Thorton and McManus (1994) studied the use of carbon and nitrogen stable isotopes and C/N ratios of SPM as source indicators of organic provenance in estuarine systems. d ¹³C, d ¹⁵N and C/N ratios have distinct end member characteristics, while d ¹⁵N and C/N ratios increase due to biological fractionation and can be used to record the degree of diagenetic alteration during respiration (Thorton and McManus, 1994 and Owens, 1985). Only d ¹³C has been shown to be a conservative tracer of organic matter (Thorton and McManus, 1994; Simenstad and Wissmar, 1985).

SPM samples were taken at five stations in Hood Canal, Puget Sound, WA, on the R/V Thomas G. Thompson in April, 1999. SPM samples from the Skokomish River were collected ~1 mile up the river from where it empties into Hood Canal. Data from two stations (HC1 and HC 6; Fig. 1) are used as end members to which represent marine and terrestrial environments, respectively. Stations HC2, HC3, HC4 and HC 5 (Fig. 1) are used as representative values for their respective intermediate areas. Each station (with the exception of HC6) had a near-surface (5 m) sample taken and a near bottom (5 m above bottom) sample taken with samples taken at intermediate depths. Samples were taken in duplicate, with two samples being taken in triplicate to evaluate precision. Approximately 1-2 liters of water were filtered onto Whatman Grade QM-A Quartz filters with a vacuum pump.

SPM was analyzed for ¹³C/¹²C, ¹⁵N/¹⁴N, C/N, total carbon and total nitrogen using a Finnigan Delta Plus mass spectrometer coupled with a Carlos Erba HC2500 CN analyzer. The mass spectrometer reports isotope data in the form of a bulk ratio, which is converted to the standard d ¹³C, d ¹⁵N forms using equation (1).

All ¹³C/¹²C values were standardized relative to Pee Dee Beleminite (PDB) standards and ¹⁵N/¹⁴N values were standardized relative to atmospheric air. ¹³C/¹²C and ¹⁵N/¹⁴N values are expressed as d ¹³C and d ¹⁵N, respectively, and converted to this notation using the equation:

d ¹³C or d ¹⁵N = ((R_sample – R_standard)/R_standard)*1000 (1)

where R = ¹³C/¹²C or ¹⁵N/¹⁴N. Data quality control throughout the analysis was evaluated by running a reference standard after every 5 samples. This permitted corrections for machine drift to be made prior to calculation of isotope abundance.

The d ¹³C values range 7.06⁰/₀₀ from a high of –19.82⁰/₀₀ to a low of –26.88⁰/₀₀ (mean –22.53⁰/₀₀ ± 2.21⁰/₀₀). These values are larger than the range of 6.27⁰/₀₀ (4.02⁰/₀₀ to 10.29⁰/₀₀) observed for d ¹⁵N. C/N values range from 7.26 to 22.59 (mean 9.73 ± 4.59).

d ¹³C distribution within Hood Canal shows a seaward enrichment of ¹³C, resulting in less negative values of d ¹³C as SPM becomes enriched in marine derived organic matter (-26.88⁰/₀₀ ± .93⁰/₀₀ at HC6 to -22.76⁰/₀₀ ± .56⁰/₀₀ at HC1). Values of d ¹⁵N range from 4.02⁰/₀₀ at HC6 to 8.36⁰/₀₀ at HC1, showing higher d ¹⁵N values characteristic for marine SPM (Owens, 1985). C/N ratios show the greatest variability of values within the estuary with end member values of 19.11 ± 4.96 for the Skokomish River (HC6) to 9.56 ± .11 for the marine end member (HC1).

The fractional contribution of each end member (Table 1) can be estimated using the mixing equations (Schultz and Calder, 1976):

X = F_tX_t + F_mX_m (2)

Where X, X_t and X_m are the d ¹³C, d ¹⁵N and C/N of, respectively, the sample, the terrestrial end member and the marine end member. F_t and F_m are the fraction of, respectively, the terrestrial and marine end member in the sample. Rearranging equation (2) gives:

F_t = (X - X_m)/(X_t - X_m) (3)

Enabling us to calculate F_t and F_m as required. To calculate the fractions of terrestrial and marine contributions to the different stations and depths, average values were used for the d ¹³C, d ¹⁵N and the C/N end members. Terrestrial and marine end members were, respectively, the Skokomish River (HC6) and HC1 at 100 m depth.

Fractional values below 0.00 and above 1.00 for F_t represent marine and terrestrial end member fractions where negative d ¹³C, d ¹⁵N and C/N values of F_t are assumed to represent areas where marine carbon and nitrogen are thought to be entirely of marine origin. This is a safe assumption because HC1 is inside of the sills at the entrance to Hood Canal, so a purely marine end member is not known.

Table 1 lists the values obtained for F_t and F_m. The negative values that are common to the F_t of d ¹³C, d ¹⁵N and C/N show the importance of the marine input of carbon into Hood Canal. Conversely, negative values could be a sign of an unknown source of organic matter or significant alterations of organic matter occur in the water column. The high fractionation observed for nitrogen raises questions to the accuracy of the fractional contributions of SPM to the estuary as calculated using d ¹⁵N.

TRENDS WITH WATER DEPTH

Figure 5 shows the vertical profiles observed for d ¹³C, d ¹⁵N and C/N for stations HC1 through HC5. Figure 2 shows the relationship between d ¹³C, d ¹⁵N and C/N ratios for the surface samples. The same relationship is shown for all bottom samples (Fig. 3) and for all samples (Fig. 4). Regressions are strong (r² = .78 to .91) between C/N and d ¹³C in all figures. Regression analysis of d ¹³C and d ¹⁵N and d ¹⁵N and C/N (r² as low as .001 and .26, respectively, for the two relationships). This implies that as organic matter is created and sinks, there are significant alterations due to bacterial reworking. If there were no fractionation, regressions of all relationships would have r² close to 1.00, reflecting purely end member mixing. A detailed discussion of the vertical trends in d ¹³C, d ¹⁵N and C/N are in the following sections.

Table 2 shows the vertical profiles of C/N ratios and d ¹⁵N values for stations HC1 through HC5. Both C/N ratios and d ¹⁵N are a minimum at the surface (average value for all stations: C/N 6.86 ± .66; d ¹⁵N 7.89⁰/₀₀ ± 1.38⁰/₀₀) and maximum at depth (average values for all stations:C/N 8.97 ± .98; d ¹⁵N 8.54⁰/₀₀ ± 1.08⁰/₀₀). C/N ratios and d ¹⁵N values will increase with bacterial reworking because bacteria preferentially remineralize nitrogen over carbon (Andrews, 1998) and ¹⁴N over ¹⁵N (Owens, 1985). The idea of one end member having more of an influence over the other when looking at vertical profiles of C/N and d ¹⁵N is not valid because the trend for each of these parameters with depth is to become even less marine like. If these profiles were representative of profiles that were purely reflective of end member mixing, the trends would decrease with depth, opposite to observed trends.

The d ¹³C values at depth exceed the marine end member value taken at HC 1 at 100 m depth (-20.9). If there was no fractionation of d ¹³C, this would indicate that the bottom organic matter is more terrestrial like than the surface. A more plausible explanation is that at the time of sampling, inorganic carbon was not limiting in the surface water. The values at depth indicating a more terrestrial like signature represent organic matter that was produced during times of high productivity, where carbon was not fractionated, resulting in a terrestrial like signature in the d ¹³C values.

Using a two end member mixing equation, one is able to calculate the fraction of carbon at a given station and depth derived from production and terrestrial input. The fraction of carbon present due to production at the surface is calculated using a production value of -20.12⁰/₀₀, which is computed by taking the average of all the surface values. A terrestrial end member of -26.88⁰/₀₀ used to represent the Skokomish River (HC6) input of organic matter into Hood Canal. I rearrange equation (3) to calculate the fraction of carbon present due to production with X_m form equation (3) being replaced by X_t and X_t from equation (3) being replaced by a new variable, X_p, to yield the equation:

Where X_t and X_p represent the d ¹³C and C/N ratios of the terrestrial and phytoplankton end members, respectively. These values are reported in Table 2 and show the significant influence that production has on the isotopic composition of the bulk SPM. Organic matter produced at the surface accounts for 82% ± .17% and 93% ± .06% of the organic fractions, when averaged at every depth and station, and computed using d ¹³C and C/N ratios, respectively. The organic matter at the surface is most likely all phytoplankton, or so much so that any signal of it not being phytoplankton is not evident due to the highly Redfield like C/N ratios observed at the surface. Fractions calculated using the d ¹⁵N measurements are not included due to the high fractionation observed for d ¹⁵N (Fig. 2, 3, 4). Fractions that are larger than 1 for F_p are considered to be entirely of planktonic origin. The accuracy of this calculation decreases as the distance of station location increases away from Skokomish River end member.

River flow from the Skokomish River to the Cushman dam was blocked, by a landslide approximately two weeks before our samples were collected, resulting in no water diversion to the Cushman dam (J. Newton, pers. comm.). At 40 m depth at HC 4, terrestrial like characteristics of d ¹³C, d ¹⁵N and C/N (respectively -27.49 ⁰/₀₀, 5.41 ⁰/₀₀ and 22.59) were observed. This is in almost complete agreement with the Skokomish River values for d ¹³C, d ¹⁵N and C/N (respectively -26.88 ⁰/₀₀, 4.02 ⁰/₀₀ and 19) suggesting that the Skokomish River does have the potential to play a large role in the source of organic matter to the water column at the Great Bend.

Investigations into estuarine organic matter sources were conducted using stable carbon and nitrogen isotopes, coupled with bulk C/N ratios. The d ¹³C, d ¹⁵N and C/N of the end members, along with the sample values, were used in a two end member mixing equation to asses the importance of the terrestrial and oceanic influence on the organic matter composition of Hood Canal. Vertical trends in d ¹³C, d ¹⁵N and C/N were used to show degradation of the organic matter in the water column. Although estuarine mixing of respective organic matter pools (terrestrial and marine) were implied using equation (3), SPM d ¹³C, d ¹⁵N and C/N ratios are predominantly derived from biological processing. This is best shown by looking at the vertical trends in the water column, in addition to looking at the large fraction of calculated (equation 4) SPM from production (Table 2).

The Skokomish River also has the potential to play a large role in the source of organic matter southern Hood Canal (The Great Bend). This would be best concluded using a time series data set and analysis of sediment cores using stable carbon and nitrogen isotope data, coupled with C/N ratios, that can be used to evaluate the long term effects of river flow on Hood Canal.

I would like to thank Paul Quay, Dave Wilbur, Charles Simenstad, Jan Newton and Mark Woodworth for their discussions and review of this data. I would also like to thank Roy Carpenter for careful review of this manuscript. I would also like to thank Kathy Newell and the crew of the Thomas G. Thompson for their assistance in data collection.

Albertson, S.L., Newton, J. and Eadie, M. unpublished literature. Investigation of present versus historic circulation in Hood Canal with a three-layer box model.

Andrews, J.E., Greenaway, A.M. and Dennis, P.F. 1998. Conbined Carbon isotope and C/N ratios as indicators of source and fate of organic matter in a poorly flushed, tropical estuary: Hunts Bay, Kingston Harbour, Jamaica. Estuarine, Coastal and Shlef Science. 46, 743-756.

Jay, D.A. and Simenstad, C.A. 1996. Downstream effects of water withdrawal in a small, high gradient basin: erosion and deposition on the Skokomish River Delta. Estuaries 19, 501-517.

Llanso, R. 1998. Distribution of benthic communities in Puget Sound 1989-1993. Ecology, 98-328.

Mackas, D.L. and Harrison, P.J. 1997. Nitrogeonous nutrient sources and sinks in the Juan de Fuca Strait/Strait of Georgia/Puget Sound estuarine system: Assessing the potential for eutrophication. Estuarine, Coastal and Shelf Science 44, 1-21.

Owens, N.J.P. 1985. Variations in the natural abundance of ¹⁵N in estuarine suspended particulate matter: A specific indicator of biological processing. Estuarine, Coastal and Shelf Science 20, 505-510.

Paulson, A.J., Curl, H.C. and Feely, R.A. 1993. The biogeochemistry of nutrients and trace metals in Hood Canal, a Puget Sound fjord. Marine Chemistry 43, 157-173.

Rabalias, N.N., Wiseman, W.J. and Turner, R.E. 1994. Comparisons of continuous records of near-bottom dissolved oxygen from the hypoxic zone of Louisiana. Estauries 17, 850-861.

Schultz, D.J. and Calder, J.A. 1976. Organic carbon ¹³C/¹²C variations in estuarine sediments. Geochimica et Cosmochimica Acta 40, 381-385.

Simenstad, C.A. and Wissmar, R.C. 1985. d ¹³C evidence of the origins and fates of organic carbon in estuarine and nearshore food webs. Marine Ecology - Progress Series 22, 141-152.

Tefry, J.H., Simone, M., Nelson, T.A., Trocine, R.P. and Eadie, B.J. 1994. Transport of particulate organic carbon by the Mississippi River and its fate in the Gulf of Mexico. Estuaries 17, 839-849.

Thorton, S.F. and McManus, J. 1994. Application of organic carbon and nitrogen stable isotope and C/N ratios as source indicators of organic matter provenance in estuarine systems: evidence from the Tay Estuary, Scotland. Estuarine, Coastal and Shelf Science 38, 219-233.

Table 1: Averaged values of F_t and F_m as calculated using equation (3) for every depth and station. Values are expressed as decimals.

Table 2: Average values of F_p as calculated using equation (4) for every depth and station. Values are expressed as decimals.

Figure 1: Map of Hood Canal and surrounding Puget Sound. Station locations are indicated by solid black dots. HC1 (100 m) and HC6 (Skokomish River) were used as end members.

Figure 2: Relationships between d ¹³C, d ¹⁵N and C/N for samples taken from surface depths (5 m). The conservative properties of d ¹³C are observed in (a). Little changes in the values of C/N occur to the organic matter once it is in the estuary (a, c), indicating high abundance of organic matter in the estuary. Fractionation of nitrogen is indicated by (b).

Figure 3: Relationships between d ¹³C, d ¹⁵N and C/N for samples taken from samples at depth (5 m off bottom). The conservative properties of d ¹³C are observed in (a). Little changes in the C/N ratios in relation to d ¹³C since the organic matter has fallen from the surface. Little correlation was observed between d ¹⁵N and d ¹³C and C/N (b,c)

Figure 4: Relationships of d ¹³C, d ¹⁵N and C/N between all samples. (a) shows consistancy of d ¹³C as a tracer of organic matter by its good regression coefficient (.78) with C/N and (b) and (c) show the changes in d ¹⁵N due to fractionation.

Figure 5: Depth profiles of d ¹³C, d ¹⁵N and C/N for stations HC1 through HC5.