'4*

1.,..11.4.01141'dir 7116 NN:1416" ANIN:11

* .10 —1/164144.."ZinMr" • 0 •• • ••• IP °It 1 IV '1,4 •_I v; 0 0-- • ItiV

• • •;1‘11

4411 014 I) a'OF a I

*1 I

qt,

III 41P le.144 ,

ab. vs I 1 At 44‘e,I Apr •dlebik'l

t#V4 1k ii„k 4 t ;it (OW 411..4111110.

444,

0

AuthorJOHN D. STEDNICK was hydrologist,USDA Forest Service, Tongass NationalForest, Sitka, Alaska. He is currentlyAssistant Professor of WatershedScience, Department of Earth Resources,Colorado State University, Fort Collins,Colorado 80523.

Abstract Introduction

Stednick, John D. Precipitation andstreamwater chemistry in an undis-turbed watershed in southeastAlaska. Res. Pap. PNW-291.Portland, OR: U.S. Department ofAgriculture, Forest Service, PacificNorthwest Forest and RangeExperiment Station; 1981. 8 p.

Water chemistry samples have beentaken from streamflow since 1976 andprecipitation since 1978 in Indian River,an undisturbed watershed on ChichagofIsland in Southeast Alaska. Volume-weighted concentrations of total nitrogen,ammonium nitrogen, nitrate nitrogen,total phosphorus, orthophosphate,sulfate sulfur, chloride, bicarbonate,silica, calcium, magnesium, sodium, andpotassium were used with precipitationand streamflow volumes to calculateannual input and output of elements. Totalnitrogen accumulated at 1.0 kg/ha peryear and ammonium-nitrogen at 3.3 kg/haper year; other monitored elementsshowed a net loss or export from 0.1 kg/haper year of total phosphorus to 256 kg/haper year of calcium. Precipitation andweathering of soil and bedrock materialaccount for these elemental losses instreamflow. The geochemistry of IndianRiver is compared to other studies donein mountainous forested watersheds.

Keywords: Water analysis, nutrientbudget, precipitation, streamflow,southeast Alaska.

Precipitation and streamflow measure-ments have been collected in Indian Riversince 1976 (fig. 1). Water chemistrysamples have been collected fromstreamflow waters since 1976 andprecipitation chemistry measured since1978. The objective is to characterize thehydrology and elemental balances in thestudy area. The valley forests are beinglogged for the first time, and logging andassociated impacts on water quantity andquality will be measured in later efforts.

Pretreatment results from a case-historystudy of precipitation and streamwaterchemistry are presented. From these, abasis for evaluating timber harvestimpacts in this watershed can bedeveloped.

Nutrient budgets have not been devel-oped for Southeast Alaska watershedsand we have little understanding of thebiogeochemical processes that controlstreamflow chemistry. This investigationof bulk inputs and outputs did not attemptto differentiate elemental exports frompedogenic or bedrock weathering.

To understand the biogeochemicalbehavior of terrestrial ecosystems, abudgetary approach, which assessesinputs to and outputs from small water-shed ecosystems, is often used.Investigations into the control of inputsand outputs of elements have been madethrough intensive measurements ofsingle watersheds. Paired watershedsare often used to measure land-management impacts. Departures inwater chemistry in the treatment water-shed are attributed to the managementactivity. This approach has producedhypotheses for the understanding ofspecific situations.

Water-balance measurements ofprecipitation and streamflow were usedin conjunction with volume-weightedconcentrations. Climate may affect waterquality by determining the absoluteamount of water input as a dilutinginfluence, as well as regulating losses toevapotranspiration as a concentratingeffect. Streamflow volume for the IndianRiver basin is 81 percent of the precipi-tation volume.

Cation outputs in streamwater from aforested watershed in the OregonCascade Range were related to theamounts in excess of the annualrequirements of the forest vegetation(Fredriksen 1972). The cation source waschemical weathering of the forest soil.Other studies, however, have indicatedbedrock weathering rather than soilweathering was primarily responsible forchemical composition of streamwater.Active chemical weathering was found inthe absence of biological and pedogenicprocesses (Reynolds and Johnson 1972).

The generation of mobile anions andconcomitant cation leaching will alsodefine the weathering rate and soluteremoval in some soils. Cations may beleached when the cation equivalence isgreater than the available capacity forcation exchange in the soil.

Outputs are considered to originate fromprecipitation and pedogenic and bedrockweathering. Definition of thesecomponents will occur as nutrientbudgets are better defined forecosystems in Southeast Alaska.

1

PACIFICOCEAN

..■■■■ WATERSHED BOUNDARYFOR GAGING STATIONS

Sitka

k(soeL CRE

WATERSHED BOUNDARY

Baranof Is

P PRECIPITATION RECORDER

G STREAM-LEVEL RECORDERChichagof Is.

■ INDIAN RIVERWATERSHED

Figure 1.—Key and vicinity maps for IndianRiver.

Methodology

Site DescriptionIndian River Valley is a broad U-shapedvalley near Tenakee Springs onChichagof Island in Southeast Alaska(fig. 1). The valley is approximately 50-percent forested with a mature coastalSitka spruce (Picea sitchenis (Bong.)Carr.) -western hemlock (Tsugaheterophylla (Raf .) Sarg.) forest. Thenonforested alpine areas are generallyabove 770 m with mountain peaks up to960 m. The watershed area is 72 km2.

GeologyFreshwater Bay is northeast of IndianRiver and is the northeast fold of asyncline in sedimentary and volcanicrocks that range in age from Silurian toMississippian (Loney et al. 1975). Theseare gently folded and are metamor-phosed in the vicinity of granitic rockmasses. The head of Indian River is asyncline in this material. The valleybottom is composed of unconsolidatedalluvium, colluvium, and glacialsediments. Subsurface water may flowthrough these sediments and not bemeasured as streamflow. Deeppercolation of groundwater is prevented,however, by a marine till that underlies thevalley bottom. The valley follows thenorthwest-southeast strike of the IndianRiver Fault, with a linear outcrop ofKennel Creek limestone on the east side.Upper elevation rocks are intrusiveigneous composed of hornblende,adamellite, biotite alaskite, and biotite-hornblende meladiorite. The westernvalley wall (of lower elevation) iscomposed of intensely folded, interlay-ered hornfels, schist, and amphibolite(Loney et al. 1975).

SoilAlpine areas are characteristically steepwith shallow soils and frequent bedrockoutcrops and talus slopes including alpineand subalpine meadows, brush slopes,and muskegs. The alpine soils are largelyorganic, poorly drained, and acidic. Soilson the brushy slopes are mineral andacidic but well drained. These soils havecharacteristically thick, rapidlydecomposing layers of surface litter over15-30 cm of dark and friable, gravelly siltloam. The alpine and brush soils mayoccupy slopes greater than 100 percent.

2

405

380

355

330

305

280

255

230

205

180

152

125

100

75

Hw

PRECIPITATION

Muskeg soils (Histosols) have a watertable at or close to the surface year-round and may support Alaska-cedar(Chamaecyparis nootkatensis (D.Don)Spach) and lodgepole pine (Pinuscontorta Dougl. ex Loud.) in areas wherethe water table is below the surface. Theforested mineral soils lower in the valleyare composed of alluvial and colluvialmaterials. These soils have thick surfaceorganic horizons and are typically welldrained.

VegetationThe valley forest is primarily westernhemlock and Sitka spruce with occasionalclusters of Alaska-cedar. A variety oflodgepole pine is found on poorly drainedsites. Dwarf mountain hemlock (Tsugamertensiana (Bong.) Carr.) is oftenassociated with muskegs and alpineareas. Alder (Alnus rubra Bong. and A.sinuata (Regel) Rydb.) line somestreams and dominate landslides or otherexposed mineral soil areas. Thelodgepole and alder species arenoncommercial.

The understory vegetation is character-istically blueberry ( Vaccinium ovalifoliumSm.), huckleberry ( V. parvifolium Sm.),rusty menziesia (Menziesia ferrugineaSm.), and devilsclub (Oplopanax horridus(Sm.) Miq.). Muskeg areas are dominatedby sphagnum mosses, sedges, rushes,and ericaceous shrubs. Alpine areas arelargely composed of heaths, grasses, anddeer cabbage (Fauria crista-galli(Menzies) Makino).

StreamflowIndian River annual streamflow rangesfrom 1800 to 2400 mm and averages2200 mm distributed in a bimodal pattern(fig. 2). Tenakee Springs (at sea level)has a long-term annual precipitationaverage of 1700 mm; short-termmeasurements in the valley indicateabout 2700 mm. Precipitation and

streamflow measurements in the valleyindicate orographic effects and storm-front funneling. Precipitation events mayproduce small cyclonic cells that result inlocalized precipitation. About 40 percentof the annual precipitation occurs asrainfall in September and October. Mostprecipitation occurs as snow in March orApril. Fall peakflows are reduced as snowaccumulates. Winter low flows occur untilMay or June when rain, snow melt, or bothoften create peakflows. Decreasingrainfall through summer results in agenerally receding hydrograph. Fogsand mists contribute an additional butundetermined quantity of water.

A continuous stream-level recorder wasinstalled at midvalley in 1976 at anelevation of 100 m and drainage area of34.2 km 2. The gaging cross section waslocated in a gravelly bottomed streamreach and was resurveyed for the stage-discharge calibration annually. Stream-flow measurements over the 4-yearperiod for weekly flows were accurate to± 10 percent. The results reported hereare from this midvalley gaging station.

Figure 2.—Average monthly Indian Riverstreamflow at gaging station (4 yr) andTenakee Springs precipitation (10-yraverage).

Water Sampling and AnalysisSurface grab-samples of Indian Riverwater were collected during instrumenta-tion servicing trips, about every 2 weeksduring the summer and every 3-4 weeksin winter, depending on access. Precipi-tation samples were taken as an aliquotof precipitation collected in weighingrecorders near the stream-level recorderand at the valley mouth. Spun fiberglasskept detritus and other debris out of thecollected precipitation water. Snowsamples were taken by collecting surfacesnow in wide-mouth sampling bottles.

3

Results

All samples were frozen, unfiltered aridunpreserved, and shipped to the USDAForest Service Forestry SciencesLaboratory in Corvallis, Oregon, foranalysis. Samples were analyzed for totalnitrogen, ammonium nitrogen, nitratenitrogen, total phosphorus, orthophos-phate, sulfate sulfur, chloride, bicar-bonate, silica, calcium, magnesium,sodium, and potassium.

A pH meter with an automatic titratormeasured alkalinity as bicarbonate(HCO3—C) and pH (American Public HealthAssociation 1975). A Varian 1200 AtomicAbsorption Emission Spectrophotometerwas used to measure dissolved cationsof calcium, magnesium, sodium, andpotassium. A Technicon II Autoanalyzerwas used to measure nitrogen andphosphorus forms. A copper-cadmiumreductor column reduced nitrate to nitritefor measurement (Wood et al. 1967).Total nitrogen and total phosphorus wereanalyzed after macro-Kjeldahl andperchioric acid digestions (AmericanPublic Health Association 1975) andmeasured by the Berthelot reaction andmolybdenum blue method, respectively(Technicon Industrial Systems 1971,Murphy and Riley 1962, Gales et al.1966). Sulfate-sulfur was measured bythe methylthymol blue technique (Galeset al. 1968). The molybdate-stannouschloride procedure was used to measuresilica (Golterman 1969). Chloride wasmeasured by using the ferricyanidemethod (O'Brien 1962).

Inputs and outputs were calculated on anannual basis, given that year's precipi-tation or streamflow and chemical data.The variance of the annual values wasused to calculate a standard error andsubsequently the 95-percent confidenceinterval. The net change was calculatedas the mathematical difference betweeninput and output. The confidence intervalwas calculated from a weighted averagevariance and standard error for thedifference (Steele and Torrie 1960).

Precipitation ChemistryIndian River precipitation is slightly acidic(pH = 6.5) and chemically dilute;however, the large precipitation influx(2300 to 2900 mm per year) results insubstantial deposition of elements.Precipitation originates over saltwaterand is reflected in the deposition ofchloride, calcium, magnesium, sodium,and potassium (Pritchett 1979, Patterson1976). The total nitrogen (ammonium,nitrate, nitrite, Kjeldahl, and organic) inputof 5.5 kg/ha per year must be assumed tooriginate from soil and ocean surfaces(Wollum and Davey 1975), becauseanthropogenic sources are not evident(table 1). No explanation is offered as tothe ammonium-nitrogen source. Theammonium-nitrogen concentrationaverage in precipitation was high andconsidered to be a valid measurement.Because samples were collected every2-4 weeks, interconversion amongnitrogen species may occur; however, thetotal nitrogen content would not change.

Amount of precipitation did not appear tobe related to element concentration.Separating element concentrations forseparate storms was impossible becauseof logistics and access to study sites. Astudy in the North Cascades of Wash-ington State indicated no consistentchemical differences in precipitationsamples collected at a range of eleva-tions and subsequently volumes (Dethier1979). Preliminary investigations did notindicate element concentration enrich-ment by dry deposition on surface snowlayers.

Element inputs to Indian River werecalculated by precipitation amounts andvolume-weighted concentrations andaveraged for the 1979 and 1980 wateryears. Streamflow outputs were deter-mined for water years 1977-80. Elementaltransfers in Indian River are comparableto geochemical studies of other water-sheds. (table 2).

Runoff ChemistryWater chemistry samples from IndianRiver have been collected since 1976.Volume (discharge)-weighted concen-trations and streamflow data were usedto calculate annual streamflow losses forwater years 1977-80. Water chemistrysamples often contained suspendedsediment; however, the transport ofelements associated with sediment wasnot measured. Sediment movementoccurred during the higher peakflows onthe rising limb of the hydrograph and wascomposed of 5- to 25-percent organicmaterials. The organic material could actas exchange sites. Therefore, elementoutputs in streamflow do not indicate totalwatershed output.

The alkalinity of runoff waters averaged10.47 mg/1, and the near neutral pH (7.2)was not apparently related to dischargeor season. Alkalinity increased fromprecipitation inputs because ofinteractions with the biological andpedological components. Changes in pHand alkalinity may be influenced by thepresence of organic acids as evidencedby solution color.

Dissolved cation outputs are about 400kg/ha per year with an anion output of 300kg/ha per year. Additional ion transportmay result from organic acids orassociation with suspended sediments.The output:input ratio of the cationscalcium, magnesium, sodium, andpotassium is 4.8 comparable to the 3.3of Jamieson Creek (Zeman 1975), and4.8 of the Olympics (Larson 1979), butless than the 12.0 of the H.J. AndrewsExperimental Forest in Oregon(Henderson et al. 1978).

Nutrient conservation occurred for totalnitrogen ( + 1.0 kg/ha per year) andammonium nitrogen ( + 3.3), but all otherelement transfers exhibited a net loss.Losses ranged from 0.1 kg/ha per year oftotal phosphorus to 256 kg/ha per year ofcalcium. The average net change(element gain or loss) was calculated asthe mathematical difference betweeninput and output (table 1).

4

Table 1-Summary of average precipitation input (1979-80), streamflow output(1977-80), and average net change all with 95-percent confidence intervals

Element Input Output Net change

Kilograms/hectare per yea Total N 5.5 ± 3.4 4.5 ± 0.8 + 1.0 ± 1.2NH 4 N 4.7 ± 2.8 1.4 ± 0.3 + 3.3 ± 0.5NO3 N 0.5 ± 0.3 2.2 ± 0.4 1.7 ± 0.5Total P 0.7 ± 0.4 0.8 ± 0.2 0.1 ± 0.3PO4 P 0.5 ± 0.8 0.4 ± 0.6 + 0.1 ± 0.1SO4S 9.6 ± 5.8 22.8 ± 4.2 13.2 ± 5.6C1 20.3 ± 12.2 90.1 ± 16.7 - 69.8 ± 21.8HCO3 23.4 ± 14.1 183.5 ± 33.7 - 160.1 ± 44.3Si 1.7 ± 0.9 34.1 ± 6.3 - 32.4 ± 8.2Ca 18.8 ± 11.3 275.0 ± 50.6 - 256.2 ± 66.2Mg 12.3 ± 7.4 24.0 ± 4.4 - 11.7 ± 5.9Na 38.4 ± 23.1 39.6 --.± 7.3 - 1.2 ± 9.6K 4.0 ± 2.4 10.3 ± 1.9 - 6.3 ± 2.5

Table 2-Summary of precipitation inputs and streamflow outputs for Indian River, Jamieson Creek (British Columbia),western Olympic Peninsula (Washington State), and Watershed West (New Hampshire)

Indian River(2-yr input)

(4-yr output)Jamieson Creek'

(1 yr)Olympic Peninsula2

(2-yr average)Watershed West3

(1 yr)Element Input Output Input Output Input Output Input Output

Kilograms/hectare per yearTotal N 5.5 4.5 3.8 2.2NH4N 4.7 1.4 0.5 0.4 2.8 0.4NO3 N 0.5 2.2 1.1 0.8 0.7 1.3 6.0 6.6Total P 0.7 0.8 1.1 1.0PO4 P 0.5 0.4 0.1 0.2 1.2 0.8SO4S 9.6 22.8 2.1 3.0 10.8 83.0 17.3 22.9Cl 20.3 90.1 23.1 38.1HCO3 23.4 183.5 7.6 37.2 90.5 1220.0 0 41.9Si 1.7 34.1 0.4 48.9 0 102.0Ca 18.8 275.0 7.3 41.7 10.5 320.0 3.1 26.4Mg 12.3 24.0 2.2 8.8 6.0 44.5 0.7 3.8Na 38.4 39.6 13.2 25.6 82.0 112.5 1.3 13.1K 4.0 10.3 0.9 2.6 3.8 12.0 2.4 7.0

'Zeman, 1975.2 Larson, 1979.3 Martin, 1979.

5

Table 3—Percent contribution of chemistry from tropospheric sources forstreamflow water from Indian River, Jamieson Creek (British Columbia), andWestern Olympic Peninsula (Washington State)

Constituent Indian River Jamieson Creek Olympics

Total N 122 173

NH4N 336 106

NO3N 23 138 54

Total P 88 110

PO4P 125 55 150

SO4S 42 74 13

C1 23 61

HCO3 13 20 7

Si 5 1 0

Ca 7 17 3

Mg 51 25 13

Na 97 51 73

K 39 34 32

Discussion

Chemical composition of streamflowwater can come either from troposphericfallout—including elements originallyderived from the ocean, terrestrialsources, or biological fixation. Thetropospheric contributions of eachelement can be expressed as a percent-age of streamflow output (table 3).Sulphate sulfur, chloride, bicarbonate,silica, calcium, magnesium, and potas-sium are largely derived from terrestrialsources.

The conservation of nitrogen in thesystem may be attributed to utilization ofammonium or nitrate nitrogen by forestflora, actual storage of the nitrogen in thesoil, or both. The input of nitrogen asammonia was greater than the nitratenitrogen input, a similar observation butof greater magnitude than that in theH. J. Andrews Experimental Forest(Henderson et al. 1978). The majority ofstreamwater nitrogen output fromundisturbed watersheds occurs in organicforms (Fredriksen 1972). Nitrogenbudgets are assessed through hydro-logic processes and do not reflecttransfers through gaseous forms,nitrogen fixation, or denitrification.Denitrification is probably not appre-ciable in the forest soils, because they areacid, well drained, and aerated. Denitri-fication and volatilization, however, maybe occurring in muskegs, where precipi-tation collectors were located, andaccount for the ammonium nitrogeninput.

The total loss of phosphorus of 0.1 kg/haper year results from weathering of theapatite series, which consists of ortho-phosphate (Hem 1970). Phosphorusconcentrations in naturally occurringsurface waters are usually small becauseof its utilization by aquatic vegetation(Hem 1970). Fixation of phosphate insoils containing appreciable amounts ofhydrated iron and aluminum oxides—typical of podzolized soils—restrictsphosphate movement (Zeman 1975). The

mechanisms of organic phosphorusretention by soils have not been fullyestablished. Although organic phos-phorus is reported to leach from soils, alarge proportion of the organic phos-phorus appears to be removed asparticulate matter rather than as dis-solved phosphorus (Fredriksen 1972).

Chloride is present in certain of theminerals of the igneous rocks, such aschlorapatite or as apatite in limestone(Mason and Berry 1968), but theseminerals have a limited distribution inIndian River Valley (Loney et al. 1975).Not all of the chloride output is believedto come from weathering of this parentmaterial. During dry periods, forestcanopies are effective interceptors ofchloride from sea salt in the dry state.Rains in late summer and fall wash outsalts accumulated on the foliage(Eriksson 1955).

The net loss of sulfur ( –13.2 kg/haper year) may be the result of weatheringof sulphur-bearing minerals within thewatershed. Pyrite occurrences are notuncommon (Loney et al. 1975).Decomposition of organic matter isanother possible source of vadose

sulphur compounds (Zeman 1975). Thismay be important in organic soil (muskeg)areas. The net outflow of sulfur isattributed to dissolution of sedimentaryrocks and from organic material contain-ing sulphur (Rainwater and Thatcher1960). A hot-water spring, containingsulfur, exists in the community of TenakeeSprings. The extent of this spring or otherhot springs in Indian River Valley is notknown.

The principal source of magnesium innatural waters is ferromagnesianminerals in igneous rocks andmagnesium carbonate in carbonate rocks(Hem 1970). The potassium content innatural waters is usually small. Potas-sium occurs in rocks in a form not easilybrought into solution, although severalgeochemical processes tend to removepotassium selectively and return it to thesolid phase (Hem 1970).

6

Conclusions

Sodium is very soluble and readilyleached from soil, mineral, or bedrock.Once in solution, sodium tends to remainthere. Calcium is dissolved frompractically all rocks, but is usually foundin greater quantities in waters leachinglimestone, dolomite, or gypsum deposits(Hem 1970).

The precipitation input of calcium (18.8kg/ha per year) as well as potassium,sodium, and magnesium largelyoriginates as an aerosol over saltwater(Pritchett 1979). The net calcium loss of256 kg/ha per year is from dissolution andweathering of a linear outcrop of KennelCreek limestone. Similarly, the losses ofmagnesium, sodium, and potassium at11.7, 1.2, and 6.3 kg/ha per year.respectively, may be dissolution andweathering of limestone, as well aspotassium and sodium feldspars and theavailable hornfels, schists, andamphiboles.

The elemental balances from Indian Riverare comparable to similar studies inforested watersheds. Indian River had anaverage annual precipitation of 2700 mm,Jamieson Creek in British Columbia had4540 mm for the 1970-71 water year(Zeman 1975), and the OlympicPeninsula averaged 4230 mm (Larson1979). No precipitation or streamflow datawere presented for Watershed West(Martin 1979).

Though precipitation was less than in theother studies, element inputs or outputsare comparable to them because ofhigher concentrations of elements. Thelarge input of total nitrogen to Indian Riveris not observed in the other cited studies;however, nitrogen conservation stillOMB's.

During the early stages of successionwhen production and biomass accumu-lation are rapid, nutrients essential toplant growth may be retained in thebiomass with a reduction in concentra-tion of streamwater. As the forest maturesand net production declines, uptake of

essential nutrients will be reduced andthese nutrients will be leached out of thesystem in streamwater (Vitousek 1977,Vitousek and Reiners 1975). Nitrogenand phosphorus accumulate while theother essential plant nutrients have a netloss in Indian River. The nutrient budgetsof nonessential ions—chloride, bicar-bonate, and sodium—should beunaffected by successional stage. Thebiogeochemistry of cations seems to bestrongly controlled by precipitation, soilwater movement, chemical weathering ofbedrock materials, and movement ofsuspended sediment, rather than forestsuccession.

Geochemical influences on streamwaterchemistry are evident in this forestedwatershed. Water quality is influenced bythe mineral composition and solubility ofunderlying rocks. Water percolatingthrough soil and substrata en route to thestream may be enriched with dissolvedsolids. A correlation between the mineralcomposition of water and that of thegeologic formation with which the waterhas been in contact seems reasonable.The major constituents of sodium,calcium, magnesium, bicarbonate,sulphate, chloride, and silica may bereadily influenced by bedrock weather-ing, depending on the bedrock withsecondary constituents of iron,potassium, carbonate, nitrate, fluoride,and boron influenced less (Davis andDeWiest 1966). Note that the weatheredelements coming from geologicformations can be overshadowed byother effects. Differences in climate orother influences on the weatheringprocess can produce very different typesof water from essentially similar rocksources (Hem 1970). A satisfactorysystem of classifying water based entirelyon composition of source rocks isunlikely.

Indian River receives about 140 kg/haper year of dissolved solids in 2700 mmof precipitation. Average annual losses instreamflow are 700 kg/ha per year ofdissolved solids and do not includeelemental losses associated withsediment transport or gaseous transfers.Total nitrogen accumulated in the water-shed at a rate of 1.0 kg/ha per year. Othermonitored elements had greater stream-flow losses than precipitation inputs.Precipitation (tropospheric) inputsaccounted for 5-336 percent of theelemental losses in streamflow.

Individual storm volumes and elementconcentrations were not investigatedbecause of limited access to the studyarea during the rainy fall months. Therelation of streamflow to elementconcentrations was not apparent;however, volume-weighted concentra-tions were used to portray elementalgains and losses more accurately.

The terrestrial sources of elements arefrom soil (pedogenic) and bedrockweathering. The presence of a linearoutcrop of Kennel Creek limestone andassociated minerals can account for thebulk of element addition to thestreamwater. Calcium as well as sodium,potassium, and magnesium losses werelargely from limestone. Potassium andsodium were also derived from availablehornfels, schists, and amphiboles.Phosphorus and sulfur were derivedlargely from apatite and pyrite,respectively. No effort was made toseparate elements weathered from thesoil and those from bedrock.

The elemental losses from Indian Riverare comparable to similar studies inmountainous and forested watersheds.The nitrogen budget distinguishes IndianRiver from other watersheds studied,however.

AcknowledgmentThe chemical analyses reported herewere done by Elly Holcombe, chemist,under cooperative agreement withRWU-PNW-1653, Pacific NorthwestForest and Range Experiment Station,Corvallis, Oregon.

7

Literature Cited

American Public Health Association. Standardmethods for the examination of water andwastewater. 14th ed. Washington, DC:American Public Health Association;1975. 1193 p.

Davis, S. N.; DeWiest, R. J. M. Hydrogeology.New York: John Wiley and Sons; 1966.463 p.

Dethier, D. P. Atmospheric contributions tostreamwater chemistry in the NorthCascade Range, Washington. WaterResour. Res. 15(4): 787-794; 1979.

Eriksson, E. Airborne salts and the chemicalcomposition of river waters. Tellus(7): 243-250; 1955.

Fredriksen, R. L. Nutrient budget of aDouglas-fir forest on an experimentalwatershed in western Oregon. In: Franklin,J. F.; Dempster, L. J.; Waring, R. H., eds.Research on coniferous forest eco-systems. Portland, OR: U.S. Departmentof Agriculture, Forest Service, PacificNorthwest Forest and Range ExperimentStation; 1972: 115-131.

Gales, M. E.; Julian, E.; Kroner, R. Methodsfor quantitative determination of totalphosphorus in water. J. Am. WaterworksAssoc. 58 (10): 1363; 1966.

Gales, M. E.; Kaylor, W. H.; Longbottom, J. E.Determination of sulphate by automaticcolorimetric analysis. Analyst 93:97; 1968.

Golterman, H. T. Methods for chemicalanalysis of fresh waters. 113P. Handbk. 8.Oxford: Blackwell Scientific Publications;1969. 166 p.

Hem, J. D. Study and interpretation of thechemical characteristics of natural water.Water Supply Paper 1473. Washington,DC: U.S. Department of the Interior;Geological Survey; 1970. 363 p.

Henderson, G. S.; Swank, W. T.; Waide, J. B.;Grier, C. C. Nutrient budgets of Appa-lachian and Cascade Region watersheds:a comparison. For. Sci. 24(3):385-397;1978.

Larson, A. G. Origin of the chemical compo-sition of undisturbed forested streams.Western Olympic Peninsula, WashingtonState. Seattle: University of Washington;1979. 216 p. Ph. D. dissertation.

Loney, R. A.; Brew, D. A.; Muffler, T. J. P.;Pomeroy, J. S. Reconnaissance geologyof Chichagof, Baranof and Kruzof Islands,Southeastern Alaska. Paper 792.Washington, DC: U.S. Department of theInterior, Geological Survey; 1975. 105 p.

Martin, D. W. Precipitation and streamwaterchemistry in an undisturbed forestedwatershed in New Hampshire. Ecology60(1):36-42;1979.

Mason, B.; Berry, L. G. Elements of mineral-ogy. San Francisco: W. H. Freeman andCo.; 1968. 500 p.

Murphy, J.; Riley, J. A modified single solu-tion for the determination of phosphate innatural waters. Anal. Chim. Acta(24):27-31;1962.

O'Brien, J. E. Automatic analysis of chloridesin sewage. Waste Eng. 33: 670-672; 1962.

Patterson, M. P. Is air chemistry monitoringworth its salt? In: Environmentalbiogeochemistry, Vol. 2. Ann Arbor, MI;Ann Arbor Science; 1976:77-78.

Pritchett, W. L. Properties and managementof forest soils. New York: John Wiley andSons; 1979. 500 p.

Rainwater, F. H.; Thatcher, L. L. Methods forcollection and analysis of water samples.Water Supply Paper 1454. Washington,DC: U.S. Department of the Interior,Geological Survey; 1960. 301 p.

Reynolds, R. C.; Johnson, N. M. Chemicalweathering in the temperate glacialenvironment of the northern CascadeMountains. Geochim. Cosmochim. Acta36: 537-554; 1972.

Steele, R. G. D.; Torrie, J. H. Principles andprocedures for statistics. New York:McGraw Hill, Inc.; 1960. 481 p.

Technicon Industrial Systems. Low-levelammonia in fresh and estuarine water.Industrial Method AAII, 108-71W.Tarrytown, NY: 1971.2 p.

Vitousek, P. M. The regulation of elementconcentrations in mountain streams in thenortheastern United States. Ecol. Monogr.47:65-87; 1977.

Vitousek, P. M.; Reiners, W. A. Ecosystemsuccession and nutrient retention: ahypothesis. Bioscience 25:376-381; 1975.

Wollum, A. G. II; Davey, C. B. Nitrogenaccumulation, transformation and trans-port in forest soils. In: Bernier, B.;Winget, C. H., eds. Forest soils and forestland management. Quebec City, PQ:Les Presses de L'Universite Laval; 1975:67-107.

Wood, E. D.; Armstrong, F. A. J.; Richards,F. A. Determination of nitrate in sea waterby cadmium-copper reduction to nitrite.J. Mar. Biol. Assoc. U. K.; 1967: 23-31.

Zeman, T. J. Hydrochemical balance of aBritish Columbia mountainous watershed.Catena 2:81-93; 1975.

8

Stednick, John D. Precipitation and streamwater chemistry in an undisturbedwatershed in southeast Alaska. Res. Pap. PNW-291. Portland, OR: U.S.Department of Agriculture, Forest Service, Pacific Northwest Forest andRange Experiment Station; 1981.8 p.

Water chemistry samples have been taken from streamflow since 1976 andprecipitation since 1978 in Indian River, an undisturbed watershed on ChichagofIsland in Southeast Alaska. Volume-weighted concentrations of total nitrogen,ammonium nitrogen, nitrate nitrogen, total phosphorus, orthophosphate, sulfatesulfur, chloride, bicarbonate, silica, calcium, magnesium, sodium, and potassiumwere used with precipitation and streamflow volumes to calculate annual input andoutput of elements. Total nitrogen accumulated at 1.0 kg/ha per year andammonium-nitrogen at 3.3 kg/ha per year; other monitored elements showed a netloss or export from 0.1 kg/ha per year of total phosphorus to 256 kg/ha per year ofcalcium. Precipitation and weathering of soil and bedrock material account for theseelemental losses in streamflow. The geochemistry of Indian River is compared toother studies done in mountainous forested watersheds.

Keywords: Water analysis, nutrient budget, precipitation, streamflow, southeastAlaska.

The Forest Service of the U.S. Department ofAgriculture is dedicated to the principle of multipleuse management of the Nation's forest resourcesfor sustained yields of wood, water, forage, wildlife,and recreation. Through forestry research,cooperation with the States and private forestowners, and management of the National Forestsand National Grasslands, it strives — as directed byCongress — to provide increasingly greater serviceto a growing Nation.

The U.S. Department of Agriculture is an EqualOpportunity Employer. Applicants for all Departmentprograms will be given equal consideration withoutregard to age, race, color, sex, religion, or nationalorigin.

Pacific Northwest Forest and RangeExperiment Station

809 NE Sixth AvenuePortland, Oregon 97232

GPO 995-359