J. Great Lakes Res., 1979Internat. Assoc. Great Lakes Res. 5 (2):99-104

NON-POINT SOURCE POLLUTION FROM ABANDONEDAGRICULTURAL LAND IN THE GREAT LAKES BASIN

Thomas M. Burton and James E. Hook l

Institute of Water Researchand

Department of Crop and Soil SciencesMichigan State University

East Lansing, Michigan 48824

ABSTRACI'. The contribution ofabandoned farm fields to non-point source pollution in the Great LakesBasin was studied on a 7.73 ha old field watershed near East Lansing, Michigan, that had been abandoned18 years prior to the study. Exports ofnitrogen were low and approached values expected for undisturbedforests in the Great Lakes Basin. Exports of phosphorus were slightly elevated compared to undisturbedforests. Exports and concentra~ions of both Nand P were similar to values listed for cleared, unproductiveland for the Great Lakes region by the recent nationwide survey ofEPA. Annual exports of total P, P04 -P,N0 3 -N, N0 2 -N, NH4 -N, Organic-N, Cl, Na, Ca, and suspended solids are given. Most exports ofall con­stituents occurred during rainfall or snow-melt generated runoff events during the spring runoff period.Soil and soil-water nutrient concentrations were low, reflecting the low fertility of the study watershed.Nutrient concentrations in runoff reflect this low fertility and would be higher from recently abandonedfarm land. The contribution of abandoned fields to nutrient loading of the Great Lakes would decreasefrom levels typical of agricultural runoff at time of abandonment to levels typical of undisturbed forestswithin 15 to 20 years following abandonment.

INTRODUCTIONAbandoned, formerly cropped open land makes upa small but substantial percentage of land in theUnited States. In the contiguous 48 states there are4.7 million hectares of land in this category andanother 4.5 million hectares of cropland that istemporarily idle (Stewart and Woolhiser 1976).These 9.2~illion hectares comprise 5.2% of allcropland in the U.S. The percentage of cleared,formerly cropped farmland in the Great LakesBasin is unknown, since this category was not in­cluded in the land use categories of the recentsurvey conducted by the Pollution from Land UseActivities Reference Group (PLUARG) of theInternational Joint Commission OJC). Accordingto estimates of PLUARG, 18.6 million hectares, or32% of all lands in the Great Lakes Basin, are inagricultural usage with 33% of all agricultural landsoccurring in the Lake Michigan Basin [InternationalJoint Commission-Pollution from Land Use Activi­ties Reference Group (IJC-PLUARG) 1977a].

1Current Address: Department of Agronomy, Coastal Plains Experi­ment Station, Tifton, GA 31794.

99

There are estimates of open, formerly croppedfarmland for the U.S. portion of the Great LakesBasin (Drynan and Davis 1978, Appendix A). Forthe Lake Michigan Basin, 6.7% of all agriculturalland or 411,000 hectares were classified as open,formerly cropped farmland while another 132,000hectares (2.2%) were classified as temporarily idle.Thus, an estimate of the contribution of runofffrom such cleared, formerly cropped farmlands tonon-point source pollution would be helpful inany assessment of nutrient loadings to the GreatLakes.

Runoff from such formerly cropped lands shouldcontain higher nitrogen and phosphorus concentra­tions than runoff from forested lands, but lowerthan runoff from intensive agricultural or urbanwatersheds because of the residual effect of pastagricultural practices (Omernik 1977). Convertingthis formerly cropped land into agricultural pro­duction would increase nutrient loadings to theGreat Lakes while conversion to forest woulddecrease loadings. The recent nationwide survey ofstream nutrient levels in relation to land use didinclude such cleared, unproductive lands but the

100 BURTON and HOOK

relatively few watersheds in this survey allowedonly limited interpretation (Omernik 1977).Nutrient losses in surface runoff from native prairiein west central Minnesota have been intensivelystudied (Timmons and Holt 1977), but nutrientlosses from successional, old field watersheds in theGreat Lakes Basin have received little attention.Thus, a study of runoff losses from an abandonedfield was included as part of the Felton-HerronCreek Study. Felton-Herron Creek was one of thewatersheds selected for intensive study by the pilotwatershed group (Task C) of PLUARG. The objec­tive of this study was to quantify losses ofnitrogen,phophorus, and other nutrients from abandonedfarm lands in lower Michigan as a means of assess­ing the impact of abandoned farm fields on nutrientloadings to the Great Lakes.

MATERIALS AND METHODSAn old field, abandoned approximately 18 yearsago, was selected for study. The predominantvegetation on ths field was goldenrod (Solidagocanadensis and S. graminifolia) and quackgrass(Agropyron repens) , but a very diverse floraexisted. Most of this field was included in a 7.73ha subwatershed with well delineated topographicboundaries on the Michigan State Universitycampus, East Lansing, Michigan. An existingsubsurface drainage tile installed while the fieldwas in cultivation was still functional and provideda convenient place to sample runoff from this sub­watershed. The drainage tile emptied into an arti­ficial channel at the edge of the field. Dischargefrom this chalmel was measured with a V-notchweir and a Stevens Type F recorder. Water sampleswere taken during spring runoff and storm eventswith an ISCO sequential water sampler. Thesesamples were supplemented by low flow grabsamples.

Paired soil samples were taken at 24 sites inincrements to a depth of 150 cm along a "star"shaped group of criss-crossing transects in 1975.Soil-water samples were taken at weekly intervalsat 15 and 150 cm depths with porous cup tube­type vacuum soil-water samplers throughout the1976 and 1977 ice-free seasons.

Precipitation inputs were monitored withthree recording rain gauges at nearby localities.Precipitation for the study area averages 77.2cm/year. Precipitation was below normal duringboth years of the study and was 73.3 cm duringthe 1975-76 water year and 61.9 cm during the1976-77 water year.

All runoff samples were analyzed using standard

AutoAnalyzer techniques (U.S. EnvironmentalProtection Agency 1974). These techniquesincluded the automated ferric thiocyanate chloridemethod, automated colorimetric phenate ammonianitrogen method, automated diazotization nitritenitrogen method, automated cadmium reductionnitrate-nitrite nitrogen plethod, automated molyb­date reactive phosphorus method, persulfate diges­tion total phosphorus method, and automatedKjeldahl organic plus ammonia nitrogen method.Sodium and calcium analyses of runoff were donewith atomic absorption spectrophotometry andsuspended solids analyses followed StandardMethods (American Public Health Association1971). Nitrate and ammonium in soil-water sampleswere analyzed by ion-selective electrodes in 1976(Milham et al. 1970, Orion Research 1971) and bystandard_AutoAnalyzer techniques in 1977. Allother water-soil analyses were done by standardAutoAnalyzer techniques (U.S. EnvironmentalProtection Agency 1974). Soil samples wereanalyzed for nitrate extracted from wet sampleswith IN K2 S04 (approximately a 1:5 soil:solutionratio) (Bremner 1965); for available P extractedwith dilute acid-fluoride (1:8 soil:solution ratio)(Jackson 1958) and analyzed by the colorimetricmethod (Murphy and Riley 1962); for total P bythe method of Sommers and Nelson (1972); andfor total N by the method of Nelson and Sommers(1972).

Runoff loadings from the watershed were calcu­lated using the stratified, random sampling modelemploying a ratio estimator as suggested in theMarch 1977 revision of the IJC-PLUARG, QualityControl Handbook for Pilot Watershed Studies.

Soils on this site are very heterogeneous andinclude Miami, Conover, and Kalamazoo loams,Granby loamy sands, Barry and Corunna sandyloams, and Westland silty clay loams. In general,these loamy soils are developed on silt to loamglacial tills and are members of the mixed mesicfamily of Typic Hapludalfs. There apparently is afairly continuous clay lens underlying the lowercentral portion of this watershed, since relativelyimpermeable reduced clays were encountered atevery soil sampling site in the lower central partof the watershed. The existence of this clay lensresults in a perched shallow water table.

RESULTS AND DISCUSSIONWater budgets for this abandoned field water­shed were calculated using the technique of Thorn­thwaite and Mather (1967) and indicated thatlittle, if any, recharge of groundwater occurred

RUNOFF FROM ABANDONED AGRICULTURAL LAND 101

(Table 1). Almost all excess water appears to havebeen intercepted by the existing tile drainage sys­tem and exported from the watershed as runoff.Runoff from this site was excessive compared tonearby unirrigated forest and old field sites moni­tored as part of other studies. The excessive runoffprobably resulted from the clay lens underlying thelower central portion of this watershed and fromexistence of the tile drain. Since groundwaterrecharge on this site was minimal as a result of theclay lens and lower than average precipitation,runoff losses should be maximal for such a system.In many other Great Lakes watersheds, runofflosses would be reduced by losses to the ground­water pool.

Annual runoff loadings are presented in Table2. There was runoff in 1977 only during theFebruary to early June period. Exports during thespring runoff period (mid-February through May)

TABLE 1. Water budgets for the watershed (values in m 3 /ho).

represent 99.8% of total annual runoff. Thus, thespring runoff data in Table 2 can be used as anapproximate estimate of total export during thisdry year. Exports were much less than in the wetter1975-76 water year for all constituents. This vari­ability associated with variable rainfall suggests theneed for long-term data from this as well as otherwatershed studies in the Great Lakes Basin.

The annual flow weighted mean concentrationof 0.073 mg total P/L is low compared to the0.152 mg total P/L reported for intense agriculturalland for this region (Omernik 1977). It is inter­mediate between values reported for agriculturaland forest land uses as was predicted by Omernik(1977). Orthophosphorus concentrations (0.028mg P/L) also are intermediate between forest andagricultural land uses as was predicted. Somewhathigher concentrations are expected from water­sheds on soils with high clay content (IJC-PLUARG

Precipitation Evapotranspiration Runoff Recharge

ANNUAL 1975-76 7329.6 5548.7 2378.1 -597.2%of Input 100.0 75.7 32.45 -8.15

ANNUAL 1976-77 6188.8 5618.7 467.6 102.5%of Input 100.0 90.79 7.56 1.66

TABLE 2. Stream exports from the 7.73 ha abandoned farm field watershed.

1975-1976 1976-77

Total Exports Unit Area Percent Percent Exports in Unit(kgjyr ± one Exports Transported Transported Spring Runoff Area

Constituent std. dev.) (kgjhajyr) by Runoff* During 1977 Exports**·Events Spring (kg ± one std. (kgjha)

Runoff** dev.)

Molybdate Reactive P 0.83 ± 0.50 0.107 85.6 71.4 0.111 ± 0.013 0.014Total P 2.34 ± 0.93 0.303 84.1 76.6 0.363 ± 0.263 0.047Nitrate-N 1.35 ± 0.29 0.175 76.8 65.3 0.748 ± 0.103 0.097Ammonia-N 2.47 ± 2.32 0.320 82.8 91.2 0.183 ± 0.059 0.024Nitrite-N 0.39 ± 0.06 0.051 88.6 92.9 0.040 ± 0.Q11 0.005Total Inorganic N 4.21 ± 2.34 0.545 81.5 83.1 0.971 ± 0.119 0.126Organic N 13.76 ± 4.20 1.780 83.0 82.0 2.166 ± 0.423 0.280Chloride 488.35 ± 256.08 63.176 53.0 63.3 170.41 ± 26.44 22.045Sodium 440.26 ± 183.18 56.955 69.3 74.6 84.28 ± 16.84 10.903Calcium 107396 ± 98.46 138.934 65.3 76.5 142.80 ± 17.91 18.474Suspended Solids 307.43 ± 254.85 39.771 78.6 78.9 28.05 ± 16.72 3.629

Water (m3 ) 2378.1 307.6 80.0 467.6 60.5

*Runoff events were separated on the basis of distinct runoff peaks and include snowmelt and rain peak generated runoff during spring.**Spring runoff includes peak events plus baseflow and interflow conditions from snowmelt until constant runoff ceased.

***Spring only exports, but 99.8% of total runoff in 1976-77 occurred in this season.

102 BURTON and HOOK

1978). The concentrations are well above the con­centration of 0.010 mg P/L that is generallyaccepted as the level that can cause eutrophicationin lakes (Vollenweider 1968).

Annual unit area loads from this old field of0.1 1kg/ha/yr (Table 2) are within the 0.10 to 0.25 kg/ha/yr range predicted for losses from grasslandswith medium loam to clay soils and are slightlygreater than 0.05 to 0.10 kg/ha/yr losses predictedfor forested watersheds by the IJC-PLUARG studies(1978). These losses are essentially identical to themean losses from forested watersheds with sedi­mentary substrates and are higher than losses fromforests on igneous substrates reported by Dillonand Kirchner (1975) for their study of exports ofphosphorus from 34 watersheds in southernOntario. Thus, losses from this abandoned fieldwatershed are only slightly higher and are alreadyapproaching background levels expected fromundisturbed forested watersheds in the Great LakesBasin. Remedial measures on such abandoned farmfields are thus unnecessary and would result inlittle improvement in water quality in the GreatLakes. Some minor reductions in phosphorusloads to the Great Lakes could be achieved by con­version of these lands to forests. Such a conversionto production of useful biomass is desirable fromboth an economic and water quality standpoint.

Both total and inorganic nitrogen concentrationsfall in the range (0.47 N03 -N, 0.078 NH4 -N,0.017N02 -N) reported for forested land (Omernik1977). It is not surprising that nitrogen con­centrations in runoff from this abandoned farmfield are so low since nitrogen is normally one ofthe primary limiting factors to terrestrial plantproductivity. Inorganic N is readily immobilizedby plant uptake, storage in the organic layer ofsoils in uncultivated fields, or is rapidly lost byleaching to the groundwater (Harmsen and VanSchreven 1955). In addition, nitrogen applicationrates on agricultural lands were much lower in thelate 1950's when this field was abandoned thanthey are at present. Thus, any residual inorganic Nwould have long since been immobilized in theplant biomass, soil organic matter, or soil micro­bial community or would have been leached togroundwater. Total N exports from this formerlycropped field are 2.33 kg/ha/yr and are in therange of 0.5 to 6.3 kg/ha/yr reported for forestedand idle/perennial land uses by IJC-PLUARG(1978). The total N losses from this watershed arealmost identical to the 2.37 kg/ha/yr reportedfrom undisturbed forested watersheds in north­western Ontario (Nicholson 1977). Cropland

losses ranged from 4.3 to 31 kg N/ha/yr accordingto IJC, so losses from this old field are very lowcompared to cropland losses in the Great LakesBasin.

It is interesting to note the high percentage ofannual exports associated with spring runoff andwith runoff events (Table 2). Even during the springrunoff period, runoff events generated by snow­melt on warm days, rainfall, or a combination ofthe two dominated exports.

The runoff of nutrients from old fields dependson the nutrient status of the soils and soil-water.This nutrient status reflects the soil type, fertilizerand cropping practices prior to abandonment,number of years since abandonment, and the suc­cessional vegetation present at any particularpoint. Soils data identify the reservoir of nutrientsavailable for runoff, provide a means to relate run­off to nutrient content of that particular soil, andprovide the data necessary for design of manage­ment schemes to prevent release of nutrients.

Soil-water analyses indicate the very infertilenature of this abandoned farm field. For example,nitrate-N increased to a yearly high of 0.55 ± .52mg N/L at the 15 cm depth on March 24, 1977,after the soil began to warm, decreased rapidlyto less than 0.01 mg N/L by April 28, 1977, thenincreased slightly by mid-May with the weeklyaverage varying from 0.01 to 0.11 throughoutthe rest of the summer and fall. Nitrate-N concen­trations at the 150 cm depth were similar andvaried from a seasonal high of 0.30 ± .16 mg N/Lon April 7, 1977, down to a low of 0.06 ± .02 mgN/L in August. Weekly average concentrationsvaried between 0.06 and 0.13 mg N/L throughoutmost of the summer and fall. Spring peaks ofnitrate-N are the norm in fallow soils (Harmsen andVan Schreven 1955). Ammonia-N levels were alsovery low in soil-water, with weekly averages varyingbetween 0.02 and 0.25 mg N/L at the IS cm depthand between 0.05 and 0.27 mg N/L at the 150 cmdepth. Nitrite-N was always below limits of detec­tion (0.01 mg N/L). Weekly average organic-Nconcentrations varied from 0.34 to 0.93 mg N/Lat the 15 cm depth and from 0.08 to 0.82 mg N/Lat the ISO cm depth with no obvious seasonalcorrelation. Annual flow-weighted mean concentra­tions of inorganic nitrogen in runoff appear toreflect soil-water concentrations during the springwhen' concentrations are highest. Since more than80% of total runoff of nitrogen occurs during thespring (Table 2), these results are not surprising.

Total P concentration in soil-water was also verylow, with the weekly average varying from 0.02 to

RUNOFF FROM ABANDONED AGRICULTURAL LAND 103

0.36 mg P/L at the 15 cm depth and from 0.003to 0.32 mg P/L at the 150 cm depth. Most weeklyaverages were less than 0.080 mg P/L at bothdepths. There were no obvious seasonal trends insoil-water P. The annual flow-weighted mean con­centration of total P in runoff (0.073 mg/L) alsoappears to reflect soil-water concentrations duringspring.

Soil analyses also indicated the very low fertilityof this abandoned farm land (Table 3). Both avail­able (Bray extractable) phosphorus and nitratewere very low in these soils (Table 3). Both elementstended to decrease with depth, with highest con­centrations in the top 15 cm of soil where much ofthe root biomass and soil organic. matter werelocated. Total Kjeldahl-N also followed the trendof decreases with depth with Kjeldahl-N beingmore than an order of magnitude greater in thehighly organic surface soils (1095 J.1g/g dry soil inthe top 5 cmfthanafihe 150 cm depth (68 J.1g/gdry soil). Total P concentrations were also higherin the surface soils (342 ± 204 J.1g/g dry soil in thetop 5 cm), decreased rapidly to 254 ± 221 J.1g/gdry soil in the 31-45 cm increment, then leveledoff at a concentration of about 250 J.1g/g dry soiland remained at this level down to the 300 cmdepth sampled in 1975.

TABLE 3. Phosphorus and Nitrogen analyses olsoiis fromthe abandoned farm field watershed (values in pgjg dry soil± one standard deviation).

BrayDepth-cm Extractable P Total P Nitrate-N Kjeldahl-N

0-5 6.48 ±6.94 342 ± 204 3.84 ±4.24 1095 ± 3825-10 4.36 ±4.29 341 ±222 3.94 ±4.85 1008 ± 329

10-15 3.10 ± 2.31 324 ±212 3.65 ± 3.58 762 ± 32015-30 2.50 ± 2.16 283 ± 183 3.00 ± 2.87 599 ±29730-45 2.93 ±4.33 254 ± 221 2.02 ± 1.39 379 ± 24945-60 2.39 ±5.42 245 ±202 1.78 ± 1.14 247 ± 11760-75 2.60 ±4.97 252 ± 155 1.77 ± 1.04 222 ± 12175-90 2.34 ± 3.29 282 ± 171 1.81 ± 1.25 178 ± 11690-105 2.61 ± 3.43' 277 ± 162 1.77 ± 1.08 134 ± 91

105-120 2.19 ± 3.02 283 ±232 1.75 ± 1.14 102 ±75120-135 1.91 ± 2.92 236 ± 118 1.60 ± 1.00 81 ±61135-150 1.30 ± 2.15 231 ±89 1.61 ± 1.04 68 ±58

AVERAGE 2.89 ± 1.35 279 ± 39 2.38 ± .94 406 ± 369

These soil and soil-water data reflect the veryinfertile nature of this particular abandoned field.Certainly, recently abandoned fields would havelosses of nutrients in runoff approaching lossesfrom active agriculture. These data suggest thatfields that have been abandoned from 15-20 yearswill have nutrient losses approaching background

levels typical of undisturbed forests. Thus, toprecisely calculate nutrient losses from abandonedfarm fields to the Great Lakes would require dataon time of abandonment, soil nutrients status, etc.

CONCLUSIONS

Farm lands that have been abandoned for 15 to 20years are not major non-point sources of pollution.Phosphorus and nitrogen loadings from such water­sheds approach background levels for undisturbedforests. When calculating total loadings to theGreat Lakes from such abandoned fields, typicalnitrogen loadings for forests as generated by therecent PLUARG studies should be used whilephosphorus loadings slightly higher than thosetypical of forest loadings should be used. Nutrientlosses from recently abandoned fields shouldapproach losses typical of agriculture and dropover a 10 to 15 year period to losses typical offorests.

No remedial actions to reduce pollution fromthese abandoned farm lands appear to be practi­cable or needed. Disruption of still functionaldrainage tiles could possibly reduce suspendedsediment and total P losses from such lands andincrease groundwater recharge. Conversion .ofthese lands to forests would likely result in slightimprovements in water quality while producing aneconomically valuable product.

ACKNOWLEDGMENTS

Funds for this project were provided by the U.S.Environmental Protection Agency (Grant NumberR005143-Q 1) through the Pollution from Land UseActivities Reference Group of the InternationalJoint Commission established under the Canada­U.S. Great Lakes Water Quality Agreement of1972. Findings and conclusions are those of theauthors and do not necessarily reflect the views orrecommendations of the Reference Group, theInternational Joint Commission, or the Environ­mental Protection Agency. The authors wouldlike to thank C. S. Annett, W. Baker, P. Bent, J.Ervin, and D. O'Neill for field and technicalassistance.

REFERENCES

American Public Health Association. 1971. Standardmethods for the examination of water and wastewater.13th Edition. Washington D.C.

Bremner, J. M. 1965. Inorganic forms of nitrogen. In:Methods of Soil Analysis. C. A. Black, ed. AgronomyMonogr., 9:1179-1237. Amer. Soc. Agron., Madison,Wisconsin.

104 BURTON and HOOK

Dillon, P. J., and Kirchner, W. B. 1975. The effects ofgeology and land use on the export of phosphorusfrom watersheds. Water Research 9:135-148.

Drynan, W. R., and Davis, M. J. 1978. Application ofthe universal soil loss equation to the estimation ofnonpoint sources of pollutant loadings to the GreatLakes. Technical Report, International ReferenceGroup on Great Lakes Pollution from Land Use Activ­ities, International Joint Commission, Windsor, Ontario.

Harmsen, G. W., and Van Schreven, D. A. 1955. Mineraliza­tion of organic nitrogen in soil. In: Advances in Agron­omy, Vol. 7, 299-398. A. G. Norman, ed. AcademicPress, New York.

International Joint Commission-Pollution from Land UseActivities Reference Group. 1977a. Land use and landuse practices in the Great Lakes Basin. Joint SummaryReport-Task B, United States and Canada, Interna­tional Joint Commission, Windsor, Ontario.

_. 1977b. Quality control handbook for pilot watershedstudies. International Joint Commission, Windsor,Ontario.

_. 1978. Environmental management strategy for theGreat Lakes system. Final Report, International JointCommission, Windsor, Ontario.

Jackson, M. L. 1958. Soil chemistry analysis. Prentice-Hall,Inc., Englewood Cliffs, New Jersey.

Milham, P. J., Awad, A. S., Paull, R. E., and Bull, J. H.1970. Analyses of plants, soils and waters for nitrateby using an ion-selective electrode. Analyst 95 :751-753.

Murphy, J., and Riley, J. P. 1962. A modification singlesolution method for the determination of phosphatein natural waters. Anal. Chim. Acta. 27:31-36.

Nelson, D. W., and Sommers, L. E. 1972. A simple diges.tion procedure for estimating total N in soils and sedi­ments. J. Env. Qual. 1:423435.

Nicholson, J. A.1977 .Forested watershed studies. Summary

Technical Report, PLUARG Task C, Activity 2, Inter­national Joint Commission, Windsor, Ontario.

Omernik, J. M. 1977. Nonpoint source-stream nutrientlevel relationships: A nationwide survey. EcologicalResearch Series Report, EPA-600/3-77-105, U.S. Envi­ronmental Protection Agency.

Orion Research, Inc. 1971. Instruction manual. Am­monia electrode. Orion Research, Inc., Cambridge,Massachusetts.

Sommers, L. E., and Nelson, D. W. 1972. Determinationof total phosphorus in soils. A rapid perchloric aciddigestion procedure. Soil Sci. Soc. Amer. Proc. 36:902­904.

Stewart, B. A., and Woolhiser,.D. A. 1976. Introduction,pp. 1-5. In: Control of Water Pollution from Cropland.Vol. II-An overview. B. A. Stewart, D. A. Woolhiser,W. H. Wischmeier, J. H. Caro, and M. H. Frere. ReportNo. EPA-600/2-75-026b, U.S. Environmental ProtectionAgency and Report No. ARS-H-5 -2, Agricultural ResearchService, U.s. Department of Agriculture.

Thornthwaite, D. W., and Mather, J. R. 1967. Instructionsand tables for computing potential evapotranspirationand the water balance. Publ. In Climatology, Vol.10:185-311. Laboratory of Climatology, Drexel Insti­tute of Technology, Centerton, New Jersey.

Timmons, D. R., and Holt, R. F. 1977. Nutrient losses insurface runoff from a native prairie. J. Env. Qual.6(4):369-373.

U.S. Environmental Protection Agency. 1974. Methods forchemical analysis of water and wastes. EPA-625/6-74­003, U.S. Environmental Protection Agency.

Vollenweider, R. A. 1968. Scientific fundamentals of theeutrophication of lakes and flowing waters, with particu­lar reference to nitrogen and phosphorus as factors ineutrophication. Technical Report, Organisation forEconomic Co-operation and Development, Paris.