Ecological Archives E092-033-A1

Gordon W. Holtgrieve and Daniel E. Schindler. 2011. Marine-derived nutrients, bioturbation, and ecosystem metabolism: reconsidering the role of salmon in streams. Ecology 92:373–385.

Appendix A. Expanded materials and methods.

Pick Creek drains directly into Lake Nerka (surface area 201 km2), which is one of five important nursery lakes for sockeye salmon (Oncorhynchus nerka) within the Wood River system. Annual returns of sockeye salmon to the Wood River system have averaged 2.7 × 106 fish over the past 45 years and have sustained an important fishery since the late 1800s (Hilborn et al. 2003). The majority of salmon spawning occurs in the lower 2,500 m of the stream just below a spring-fed pond which contributes much of its total discharge (see Fig. 1 in print manuscript). The watershed area of Pick Creek is 20 km2. Riparian vegetation is primary willow (Salix spp.) with a diverse set of grasses and forbs. Riparian vegetation shaded only a small portion of the stream channel (Fig. A1).

photoA1
   FIG. A1. Photos of Pick Creek, southwestern Alaska, USA. The right panel shows post-spawning salmon carcasses. Photos: Gordon Holtgrieve.

Stream height (stage) was recorded at hour intervals during the open water season using a stilling well fitted with a water-level data logger (WT-HR-1000, Intech Instruments Ltd., Christchurch, New Zealand). Cross-section profiles of stream velocity were taken eight times during the season and used to calibrate stage to total discharge. At five times during the season measurements of stream width and cross-sectional depth were taken every 15 m longitudinally through the study reach to build relationships between stream height and average stream width and depth.

Stream water nutrients and periphyton

Individual spawning gravel sized rocks (mean width 37 ± 13 mm) were collected at four to six replicate locations in the wetted channel and cleaned to remove all periphyton. The removed material was quantitatively subsampled, collected on a GF/C filter, frozen, and analyzed for chlorophyll a content by methanol extraction and fluorescence at 440 nm (Marker et al. 1980). Rock dimensions (length × width × thickness) were used to express chlorophyll per unit rock area. The remainder of the periphyton sample was collected onto a second GF/C filter for isotopic analysis. Nitrogen isotopic ratios (15N to 14N) were measured on material scraped from the surface of the filter after it had been freeze-dried to a constant mass (24 h). Ten to 20 mg of scraped material was packed into tin boats, sealed, and analyzed by standard elemental analysis-isotope ratio mass spectrometry (EA-IRMS) methods.

Salmon nutrient mixing model

To estimate the proportion of N in periphyton from salmon (Ps) we used to two source isotope mass balance model of the form:

(A.1)

δ15NS is the isotopic signature of sockeye salmon equal to 11.2‰ (Johnson and Schindler 2009). δ15NWS is the background N isotopic signature from the watershed in the absence of salmon and was estimated from the average pre-salmon values among years (2.4‰), which is identical to the average δ15N of periphyton from a nearly stream without salmon (Cottonwood Creek). δ15NP is the measured nitrogen isotopic value of periphyton containing a mixture of salmon nitrogen and watershed nitrogen. Dissolved oxygen concentration and isotopes collections

Dissolved oxygen concentration and isotopes collections

The δ18O-O2, O2:Ar, and 13C-DIC were determined using methods similar to Barth et al. (2004). Twelve mL glass sample vials with butyl septa (Exetainers, Labco Limited) were washed with soap and water, ashed to 500 ºC, coated with a 40 µL saturated mercuric chloride as a preservative, and flushed with He to remove all atmospheric O2 two to six weeks prior to sampling. Field samples were collected in triplicate. Samples were obtained by fully immersing vials under water, opening each vial until completely filled and recapping while still underwater being careful not to contaminate with atmospheric air. Within 12 h vials were injected with an additional 30 µL saturated mercuric chloride through the septa to ensure all biologic activity ceased. A liberal amount of high vacuum silicon grease was placed over the septa and around the base of the cap to avoid air leaks. Samples were analyzed within two months at the University of Washington Oceanography Stable Isotope Lab. One day prior to analysis, pure helium was pumped into the vial until roughly half the water was displaced. Samples were acidified with 50 µL of 50% H3PO4 and allowed to come to equilibrium with the headspace for 12 to 18 h. Isotopic ratios of headspace gases (O2, Ar, and CO2) were determined by simultaneously measuring masses 32, 34, and 40 (18O–16O, 16O–16O, and Ar), and, 44 and 45 (12C–16O–16O and 13C–16O–16O) on a Finnigan Delta XL continuous-flow isotope ratio mass spectrometer (Thermo Electron Corporation). Mass 28 (N2) was also measured as an indicator of contamination by atmospheric air. In most cases, duplicate samples were analyzed for each diel time point and results averaged.

In situ dissolved oxygen concentration was determined based on the O2:Ar from isotope collections (masses 32 and 40). O2 and Ar have nearly identical solubility, and since Ar is not biologically cycled, it is assumed to be in temperature dependent equilibrium with the atmosphere. This allows determination of the dissolved Ar concentration as a function of water temperature (Weiss 1970) and oxygen concentration using the measured O2:Ar. Gas ratio measurements are a highly accurate measure of dissolved oxygen concentration (Emerson et al. 1999) and were used to calibrate optical probe based measurements.

All isotopic ratios are expressed in d notation using the following equation with units of per mil (‰):

(A.2)

where, R is the atomic ratio of the heavy (18O, 15N) to light (16O, 14N) isotopes.

Statistical analyses

The relationship between air-water gas-exchange of oxygen (k20, in units of m h-1), stream depth (D, in units of m), velocity (V, in units of m s-1), and salmon (S, in units of live fish m-2) was analyzed by nonlinear regression with the most parsimonious model given as set of competing models selected using AICc (Burnham and Anderson 2002). The full model equation was:

(A.3)

and is based on the reaeration model of O’Conner and Dobbins (1958) with an addition term for the effect of salmon density. Model parameters from the suite of models given in Table A1 were estimated through sum of squares minimization. The most parsimonious model included stream velocity and salmon density as predictor variables and explained 78% of the variation in k20.

TABLE A1. AIC table of competing reaeration models.

Model RSS AICc ΔAICc N k a α1 b α2 R2
k20 = aVa1 + bS 0.042 -21.63 0.00 10 3 0.00 2.48 0.70   0.78
k20 = aVa1Da2+ bS 0.081 -19.47 2.15 10 2 0.01 0.67* 0.55 -0.85* 0.58
k20 = aVa1Da2+ bS 0.042 -15.63 6.00 10 4 0.00 2.54 0.70 -0.05 0.78
k20 = aVa1Da2 0.208 -13.23 8.40 10 1 0.01 0.67   -0.85 0.02
k20 = aVa1Da2 0.188 -6.70 14.93 10 3 4.88 -0.62   0.92 0.02

*Parameter estimates from Owens et al. (1964).

Bioenergetics model of salmon metabolism

The Trudel et al. (2004) bioenergetics model used to calculate O2 consumption from salmon metabolism is:

(A.4)

where RT is O2 consumption for an individual fish (in mg O2 h-1) and is a function of water temperature (T, in °C), mass (W, in g), and swimming speed (U, in cm s-1). α is the standard metabolic rate of 1g fish at 0°C (0.060). β, φ, and ν are coefficients describing the metabolic costs of mass, temperature, and swimming, respectively (0.791, 0.086, and 0.0234). All coefficients are specific to sockeye salmon. The average swimming speed parameter was estimated to be 1.78 m s-1 by fitting the total amount of energy produced by an average salmon via metabolism, assuming an oxycalorific equivalent of 3.24 cal mg O2-1 (Trudel et al. 2004), to the previously measured change in sockeye salmon energy density from spawning until death (4987 kJ; Hendry and Berg 1999). The average mass of adult sockeye salmon is 2282 g and an individual fish resides in the stream for an average of 20 days (Hendry and Berg 1999). Individual rates of salmon metabolism were scaled up to the total number of salmon in the stream on a given day and converted to moles O2 consumed. Estimated values of R20 were linearly interpolated between sampling dates and scaled to daily ER as a function of average water temperature.

LITERATURE CITED

Barth, J. A. C., A. Tait, and M. Bolshaw. 2004. Automated analyses of O-18/O-16 ratios in dissolved oxygen from 12-mL water samples. Limnology and Oceanography: Methods 2:35–41.

Burnham, K. P., and D. R. Anderson. 2002. Model selection and multimodel inference: A practical information-theoretic approach. Second edition. Springer-Verlag, New York, New York, USA.

Emerson, S., C. Stump, D. Wilbur, and P. Quay. 1999. Accurate measurement of O2, N2, and Ar gases in water and the solubility of N2. Marine Chemistry 64:337–347.

Hendry, A. P., and O. K. Berg. 1999. Secondary sexual characters, energy use, senescence, and the cost of reproduction in sockeye salmon. Canadian Journal of Zoology-Revue Canadienne De Zoologie 77:1663–1675.

Hilborn, R., T. P. Quinn, D. E. Schindler, and D. E. Rogers. 2003. Biocomplexity and fisheries sustainability. Proceedings of the National Academy of Sciences of the United States of America 100:6564–6568.

Johnson, S. P., and D. E. Schindler. 2009. Trophic ecology of Pacific salmon (Oncorhynchus spp.) in the ocean: a synthesis of stable isotope research. Ecological Research 24:855–863.

Marker, A., C. Crowther, and R. Gunn. 1980. Methanol and acetone as solvents for estimation of chlorophyll-a and phaeopigments by spectrophotometry. Archiv für Hydrobiologie 14:52–69.

O'Connor, D. J., and W. E. Dobbins. 1958. Mechanism of reaeration in natural streams. Transactions of the American Society of Civil Engineers 123:641–684.

Owens, M., R. W. Edwards, and J. W. Gibbs. 1964. Some reaeration studies in streams. International Journal of Air and Water Pollution 8:469–486.

Trudel, M., D. R. Geist, and D. W. Welch. 2004. Modeling the oxygen consumption rates in Pacific salmon and steelhead: An assessment of current models and practices. Transactions of the American Fisheries Society 133:326–348.

Weiss, R. F. 1970. Solubility of nitrogen, oxygen and argon in water and seawater. Deep-Sea Research 17:721–735.


[Back to E092-033]