This file was created by scanning the printed publication. Errors identified by the software have been corrected; however, some errors may remain. Riparian Vegetation Reduces Stream Bank and Row Crop Flood Damages1 Donald Roseboom and Kenneth Russe11 2 Abstract.--Large alluvial stream valleys in the prairies of western Illinois have been channelized to increase the size of row crop fields. Fish surveys at thirteen sites in the 61,760 acre watershed documented the elimination of large game fish in old channelized stream segments when upstream segments contained smallmouth bass up to 2.7 lbs. Channelized stream banks contributed much of the sediment, which eliminated instream habitat. INTRODUCTION and residential land management types were planimetered from ASCS aerial photographs for the entire watershed. Agricultural land uses were row crop (corn and soybean), pasture, and wooded pasture. Row crops were the largest single land use in the watershed (49 percent of the area). The Soil Conservation Service has desigLated 33 Illinois counties in the Mississippi and Illinois River basins of western Illinois as critical sediment producing areas (Crews, 1983). Western Illinois streams have highest instream sediment yields in the state (Bonini et al., 1983). Nonpoint pollution is generally regarded as the most common cause of stream degradation in Midwestern streams (Judy et al., 1984). The topography and soil types in the watershed were determined during the Wisconsin Glacier Age. Glacier deposits not only formed the flat prairie of western Illinois, but torrential runoff from melting glaciers formed very large stream valleys with extremely steep bluffs. Present day streams meandered down the wide floodplains formed by glacier runoff. Fehrenbacher et al. (1977) described the upland prairie and floodplain landforms of waterways in western Illinois. Soil types have developed as dark upland. prairie soil, lighter timber soil along the bluffs, and alluvium from soil deposition in the old glacier valleys. Row crop acreages are confined to the relatively flat upland prairie and the floodplain. Pastures and wooded pastures are found on bluffs bordering the floodplain because the steep slopes and infertile soils preclude other agricultural uses. Strip·mined land is also very steep and is almost entirely pasture. From 1980 to 1983, the Illinois Department of Energy and Natural Resources funded a study by the Illinois State Water Survey to determine the effects of land use on water quality in a 97.5 square mile watershed in western Illinois. By determining the extent of different land uses in 16 large subwatersheds during two years of stream sampling, the influences of agriculture, stripmining, urban activities, and stream modifications were assessed in watershed streams. GEOGRAPHIC AREA Court Creek flows east along the southern boundary of the watershed from Galesburg, IL to the Spoon River at Dahinda, IL. Three major tributaries (Middle Creek, North Creek, and Sugar Creek) drain 65 percent of the watershed from the north. .Ten land uses of agriculture, stripmining, METHODOLOGY Sixteen stream sampling stations were established on Court Creek and its tributaries. Subwatersheds are the watershed areas between sampling stations. The extent of each subwatershed area determined from U.S. Geological Survey topographical maps developed in 1978. Each quarter section of land use was placed in a subwatershed on the basis of topographic maps. The extent/ of land uses were measured for each quarter section of watershed from 5 sets of aerial photographs taken between 1940 and 1979. By this method the extent of agriculture, stripmining, residential housing, and stream length reduction was determined in each subwatershed. lPaper presented at Riparian Ecosystems and their Management. University of Arizona, Tucson, April 16-17, 1985. 2Donald Roseboom is Associate Professional Scientist, Illinois State Water Survey, Peoria, IL. Kenneth Russell is District Fishery Biologist, Illinois Department of Conservation, Galesburg, IL. 241 residential land uses to the major water quality problems at C2. Stream flow gages were established at nine of the 16 subwatershed sampling stations. Flows at ungaged sampling stations were calculated on the basis of watershed area. A continuous stage recorder was placed at the C2 sampling station. Flow-duration curves of staff gages were adjusted to the flow-duration curve of the C2 continuous stage recorder. Stream samples were obtained with a USGS depth-integrating DH-59 sampler. The M3 subwatershed is 71 percent row crop (3093 of 4320 acres) and had the highest annual sediment and nutrient yield of the three upland prairie subwatersheds. However, the agricultural subwatershed between the Ml and M3 sampling stations is only 40 percent row crop (834 of 2080 acres) but had much greater stream sediment and nutrient yields. The Ml subwatershed had annual sediment and nutrient yields similar to the net annual yields for the entire watershed at sampling station C2. Annual stream loadings were calculated by means of a flow-duration curve and a sediment or nutrient discharge curve. Stream loadings for individual storms were determined by sampling each event during the rise and fall of flood waters. Each stream sample represented a stream flow and a sampling time period. Sediment and nutrient loadings (lbs.) per sampling period (min.) were determined as the product of stream flow (cfs), sample concentration (mg/L), time period (min.) and a factor of 0.003747. The storm loadings were the sum of the sampling period loadings. During April of 1983 the amount of sediment and nutrients transported during a single storm (storm loadings) were determined. The Ml subwatershed contributed 74, 74, and 66 percent of the sediment, phosphorus, and ammonia loadings, even though it represents only 33 percent of the combined Ml and M3 subwatersheds. The peak stream flow at Ml was 320 cfs from 2.4 inches of rain while the peak flow at M3 was 76 cfs from 2.1 inches of rain. Fish survey sites were established on stream segments representing the upper and lower sectors of the principal streams in the watershed. Illinois Department of Conservation personnel performed the fish surveys and identified the collected fishes. Fish were collected Pith a rotenone-potassium permanganate technique with a blocking net at the lower end of the site. Each site contained a pool and riffle. Stream habitat evaluations for instream habitat, water depth, stream bed materials, and riparian vegetation were made by experienced Illinois Department of Conservation personnel. An analyses of the Ml and M3 subwatersheds revealed 58 percent of the Ml subwatershed and only 25 percent of the M3 subwatershed had steep slopes (>15 percent). Bluffs along the floodplain become even steeper in lower segments of Middle Creek. This steep topography increases the rate of storm water runoff as indicated by greater peak storm flows at Ml. A comparison of aerial photographs between 1940 and 1979 found a 7 percent stream length reduction in the Ml subwatershed. Channelization further increases the rate of storm water runoff. Even more importantly, 25 percent of Ml subwatershed lays in the floodplain below the steep bluffs. Since row crop fields are the dominant land use in the relatively flat floodplain, this cultivated land occurs downstream of areas in which storm water runoff reaches high velocity. The primary complaint of land owners in floodplain subwatersheds is damage to row crop fields from high velocity stream flows, not submergence of crops. Channelization efforts in the floodplain during the 1980's were attempts to divert high velocity stream flows from rwo crop fields. Either the stream is straightened (thus increasing stream velocity) or diverted into the bluff (eroding large earthern cliffs). RESULTS The C2 sampling station represented the combined effects of all land uses in the entire watershed (table 1). At the C2 station, sediment, total ammonia, and total phosphorus were the most pronounced water quality effects. The bulk of the row crops were in the flat upland prairie surrounding the headwaters of streams as in the M3 subwatershed. The sediment and nutrient yields of upland prairie subwatersheds are compared with yields from one floodplain subwatershed and the total watershed in Table 1. The three upland prairie subwatersheds of M3, S5, and Cl2 represent, respectively, the contribution of agriculture, stripmining, and Table 1. Landform type Land use Upland prairie Residential Annual stream yield on a per acre basis. Upland prairie Upland prairie Floodplain alluvium Mining Agriculture Agriculture Total watershed at C2 Sediment tons/ac/yr 0.7 1.0 1.7 9.2 7.5 Ammonia lbs/ac/yr 1.8 1.4 2.5 4.6 3.7 Phosphorus lbs/ac/yr 2.9 2.0 7.4 14.0 14.2 242 Downstream floodplain subwatersheds had even greater annual sediment yields of 52.2, 26.8, 23.4, 46.5, and 21.0 tons/acre/yr. On North Creek, the lower 25 percent of the watershed contributed 50 percent of the annual sediment yields. Stream nutrient yields for floodplain subwatersheds are closely correlated with sediment yields. As in Middle Creek, these floodplain subwatersheds are bordered by very steep bluffs, which have high rates of surface water runoff. Table 2. Type of yield Stream lengths in downstream floodplain subwatersheds have been reduced by an average of 15 percent since 1940. Stream reduction as great as 25 percent were determined from 1940 and 1979 aerial photographs. The broad floodplain of Court Creek had similar stream length reductions. Stream reconnaissance of floodplain subwatersheds found major bank erosion sites occurring at the sites of former stream meanders in the floodplain, especially sites bordering row crop fields. Percentage of storm yield eroded from seven bank erosion sites Storm date Sediment (tons) 12/2/82 12/24/82 4/1/83 13387 13206 33293 1391 639 2551 10.4 4.8 7.7 T. Ammonia (lbs) 12/2/82 12/24/82 4/1/83 4569 3524 8470 166.9 76.7 306.1 3.7 2.2 3.6 T. 12/2/82 Phosphorus12/24/82 (lbs) 4/1/83 37858 29311 53381 7511 3451 13775 19.8 11.8 25.8 1 Weights are unit weight erage total trations of Bank erosion measurements were begun during the second year of sampling when analysis of the 1981 annual sediment yields found large increases of sediment yield from the floodplain subwatersheds. In the eroding banks, Soil Conservation Service (SCS) personnel identified a sandy overburden of recent origin, a well developed soil layer developed over thousands of years, and a sandy glacial outwash dating from the glacial ages. A 48 inch steel pin was driven into each bank layer following the method described by J.M. Hooke (1979). Six foot wooden stakes were driv.en vertically into the stream bed and bank to mark the pin positions. The location of each set of three pins was marked by a wodden stake placed 15 feet from the edge of the bank. Bank erosion measurements were made with a steel surveyors tape after each storm. The weight of the eroded bank was the product of the length of the section, the height of the bank, the depth of erosion into the bank, and a unit weight of 90 lbs/cu. ft. SCS personnel from the Knox County soil survey estimated this unit weight during the inspection of the eroding banks. Amount Percent of eroded from storm yield Storm seven bank from eroded yield sites 1 banks calculated from bank soils with a of 90 lbs. per cubic foot with avammonia and total phosphorus concen60 mg/kg and 2700 mg/kg respectively. veys of Court Creek and its three tributaries, these seven sites are not estimated to contribute more than 10 percent of the total bank erosion occurring during major storms. In July of 1982, flood waters eroded as much as 1150 tons of soil from a single bank erosion site during one storm. These storms also caused the extensive uprooting of very large trees, which hung up on sand bars and diverted much of the stream flow into stream banks and across row crop fields. In an attempt to correlate the contribution of eroding stream bank~ to the sediment budget of large streams, the amount of bank erosion at seven sites was measured during the next three storms. The amount of bank soil eroded from all seven sites during the three storms is shown in column 4 of Table 2. Chemical analysis of bank soils for total ammonia and total phosphorus permitted the determination of the amounts of ammonia and phosphorus entering the stream with bank soil. Storm yields for the three storm events are also given in Table 2. Bank erosion from just seven bank erosion sites represents a significant proportion of the sediment, phosphorus, and ammonia leaving the entire watershed at C2. Given the extent of bank instability found during stream sur- 243 Fish surveys revealed instream habitat degradation in channelized floodplain stream segments, even though channelization occurred at least 20 years earlier. The habitat degradation and its effects on fish populations are illustrated by two fish survey sites on Court Creek - C3 and C6 (table 3). The C3 site was located below 72 square miles of watershed while the C6 site had a watershed area of 35 square miles. Both water depth and instream habitat at C3 were severely reduced by the deposition of fine sands. Stream banks had very little woody riparian vegetation and large bank erosion sites were present upstream. Approximately 20 percent of bank soils at major bank erosion sites were fine sand. Much of the sand was located in the lowest layer of bank soil. The C6 fish survey site was characterized by deep pools with instream cover of snags and exposed tree roots. High gradient riffles provided coarse gravel and large rock rubble. Large mature trees increased bank stability and provided shade. Stream fish populations reflected the habitat differences between C3 and C6. The standing crop of fish at C6 was 303 pounds per acre while the C3 site had only 100 pounds per acre even though both sites contained over 2600 fish. At C3 forage species represented 95 percent of the crop by number and 55 percent of the crop by weight. Bluntnose minnows and sand shiners were the most abundant species. Game fish comprised only 4 percent of the C3 crop by weight. At the C6 site, forage species comprised 88 percent of the crop by number but only 10 percent by weight. Game fish, mainly smallmouth bass, comprised 22 percent of the crop by weight. Eighty Table 3. Species and number of fishes found in unstable channelized stream site C3 and the stable site'C6. Species C3 Largemouth Bass Smallmouth Bass Bluegill Green Sunfish Channel Catfish Yellow Bullhead Stonecat River Carpsucker Quill back Whit'e Sucker Golden Redhorse Carp Stoneroller Hornyhead Chub Emerald Shiner Striped Shiner Bigmouth Shiner Red Shiner Sand Shiner Suckermouth Minnow Bluntnose Minnow Fathead Minnow Creek Chub Johnny Darter Orangethroated Darter Logperch No. of Fish No. of Species and scrub timber. Except for the constant filling of eroded banks by the local land owners, such field damage would be more appraent on aerial photographs. C6 22 8 29 19 11 12 7 Now the major reason for channelization is to divert floodwaters from eroding into floodplain fields. Erosion of floodplain fields also occurs as scour in areas away from the stream bank. Channelization is often practiced without legal authorization as the only method available for the local land owner to limit flood damages to fields. Aerial photography and bank erosion measurements suggest channelization is a short term solution, which results in long term damages to streams and surrounding row crop fields. 10 87 15 11 9 33 38 26 14 40 9 76 107 35 116 37 24 9 56 3 93 59 2 Individual land owners are attempting to stabilize stream banks, but are having limited success. Between areas of extensive bank erosion, certain channelized stream segments demonstrated remarkable stability. These stream segments were characterized by a well developed stream border of woody vegetation, which reduce stream flow velocity and prevent bank erosion. These green belts of trees act on high velocity water in the same manner that wind breaks of trees act on high velocity winds. They reduce erosion by decreasing the velocity of the flood waters at the surface of bank soils. They reduce the effects of upstream erosion since sedimentation occurs among the trees and forms a natural levee. Sand is deposited along the banks and not as sand bars in floodplain fields. 3 134 360 769 3 928 3 23 7 1963 31 6 3 1 2646 22 2607 23 LITERATURE CITED seven smallmouth bass were found with weights up to 2.7 lbs. Pool depth and instream habitat were the most important habitat differences found between the C6 and C3 survey sites. The major bank erosion sites upstream of the C3 site contributed tons of sand to the stream during flooding. The deposition of this sand is the major cause of stream degradation in the watershed. DISCUSSION Floodplains in Knox County have served as sediment traps for upstream erosion for thousands of years; indeed floodplain soils are characterized as alluvium. Forests in western Illinois were confined to the floodplain bluffs and large valleys formed by glacial melting. The major land use changes'in these floodplain areas has been theremoval of forests and the channelization of streams to maximize row crop fields and increase the rate of storm water runoff. The Court Creek study suggests this land management practice has been too extensive. A comparison of 1940 and 1979 aerial photographs of C2 subwatershed revealed the complete erosion of a small field at bank erosion site III. An average of 2000 tons of soil was eroded every year from 600 feet of bank. The soil deposited over thousands of years was replaced by sand, gravel, 244 Bonini, A.P., N.G. Bhowmik, R.L. Allgire, and D.K. Davie. 1983. Statewide instream sediment monitoring program for Illinois. SWS Contract Report 318A, 45 pp. IL. St. Water Sur. Champaign, IL. Crews, W. 1983. Erosion in the upper Mississippi River system: an analysis of the problem. 17 pp. Upper Mississippi River Basin Assoc. St. Paul, MN. Fehrenbacher, J.B., I.J. Jansen, B.W. Ray, J.D. Alexander, and T.S. Harris. 1977. Soil associations of Knox County, Illinois. Spec. Publ. 46. 27 pp. University of IL, College of Agriculture, Urbana, IL. Hooke, J.M. 1979. An analysis of the processes of river bank erosion. Jour. Hydro!. 42:39-62. Judy R.D., P.N. Seeley, T.M. Murray, S.C. Svirsky, M.R. Whitworth, and L.S. Ischinger. 1984. 1982 national fisheries survey val. I technical report: initial findings. 140 pp. FWS/ OBS-84/06. Fish and Wildlife Service. Washington, D.C. Roseboom, D.P., R.L. Evans, J.E. Erickson, and L.G. Brooks. 1982. An inventory of Court Creek watershed characteristics that may relate to water quality in the watershed. Doc. No. 83/ 23-A. 95 pp. IL. St. Water Survey. Peoria, IL.