Stream bank erosion adjacent to riparian forest buffers, row-crop fields, and continuously-grazed pastures along Bear Creek in central Iowa.
Stream bank erosion can supply [greater than or equal to] 50% of the sediment in streams (Lawler et al., 1999). The percentage supplied by stream bank erosion depends on the adjacent land-use, vegetation cover, topography, bank material, river morphology, weather cycles, and watershed area (Hagerty et al., 1981; Hooke, 1980). In Iowa, less than 10% of the land, including riparian areas, is covered by native vegetation (Burkhart et al., 1994). Most of the riparian areas have been converted from riparian forests and prairies to row-crop fields and/or continuously-grazed cool-season grass pastures where livestock have direct access to the stream bank and channel, thus accelerating stream bank erosion (Schumm et al., 1984). Channelization also has accelerated stream bank erosion (Schumm et al., 1984).
Lowrance et al. (2002) and Ritter (1988) state that research is needed to evaluate stream bank erosion and specifically the impact of riparian livestock grazing on stream bank erosion, channel morphology and quality of stream water and aquatic habitat. One of the stated functions of riparian forest buffers is stabilizing stream banks (USDA-NRCS, 1997). The first objective of this study was to compare stream bank erosion along; a) meandering channels bordered by riparian forest buffers, row-crop fields and continuously-grazed pastures and b) between row-crop fields along both meandering and channelized channels. The second objective was to estimate the potential reduction in stream bank erosion if non-buffered stream channels were restored to riparian forest buffers. Our hypothesis was that riparian forest buffers would have the least total stream bank crosion because of lower percentages of eroding lengths of stream banks and slower stream bank erosion rates resulting from the presence of undisturbed native perennial plant cover and roots.
Methods and Materials
Study site. The riparian zone along the continuous 11 km (6.8 mi) second order study reach of Bear Creek in Story and Hamilton counties in central Iowa contains riparian forest buffer, and continuously-grazed pastures bordering meandering channels and row-crop fields bordering both meandering and channelized channels (Figure 1). Bear Creek has been designated as a National Restoration Demonstration Watershed by the interagency team implementing the Clean Water Action Plan (1999) and as the Bear Creek Riparian Buffer National Research and Demonstration Area by the U.S. Department of Agriculture (1998) because of the extensive buffer research that has been conducted over the past decade (Bharati et al., 2002; Lee et al. 1999; 2000; 2003; Marquez et al., 1999; Simpkins et al., 2002; Tufeckioglu et al., 1999; 2001).
The Spillville and Coland soil series are the dominant soils in this study reach (Table 1). Both Coland (fine-loamy, mixed, mesic Cumulic Haplaquolls) and Spillville (fine-loamy, mixed, mesic Cumulic Hapludolls) are alluvial soils, on 0 to 2% slopes, that are moderately permeable, although Coland is deeper, finer in texture, and more poorly drained (DeWitt, 1984). The topography on the recently glaciated Des Moines Lobe in Iowa is flat with a poorly integrated natural drainage system that consists of recently incised stream channels, dredged ditches and field drainage tiles. Within the 11 km (6.8 mi) study reach there are no tributaries that enter Bear Creek except for several major classic gullies (see Figure 1). As a result, channel dimensions and discharge remain relatively constant throughout the study reach except for discharge from many small tiles [15-20 cm (6-8 in) in diameter] that contribute relatively little to the overall stream discharge compared to the less frequent large tiles (> 45 cm (18 in) in diameter) located outside of this reach (Webber, 2000). By focusing on a local reach, factors that influence stream bank erosion such as weather patterns, stream discharge, topography, soil, etc. are minimized. As a result land-use practices are the primary variable influencing the stream bank erosion that are described in detail in the following sections.
The 20 m (66 ft) wide, riparian forest buffer consists of three vegetation zones parallel to the stream course (Figure 2, Figure 3a) (Isenhart et al., 1998). The first zone [10 m (33 ft) wide], located nearest the stream, is composed of trees that stabilize the stream bank and provide long-term nutrient storage. Tree species include willows (Salix L. sp), hybrid poplar (Populus L. sp), silver maple (Acer saccharinum L.), and green ash (Fraxinus pennsylvanica Marsh.). Slower-growing species such as northern red oak (Quercus rubra L.), bur oak (Quercus macrocarpa Michx.), and black walnut (Juglans nigra L.) are planted further from the stream bank, where conditions permit. The second zone [3.6 m (12 ft) wide] includes shrubs that increase habitat diversity and reduce floodwater velocities. Shrub species include redosier dogwood (Cornus stolonifera Michx.), gray dogwood (Cornus racemosa Lam.), common chokecherry (Prunus virginiana L.), nannyberry (Viburnum lentago L.), ninebark [Physicarpus opulifolius (L.) Maxim.], nanking cherry (Prunus tomentosa Thunb.), silky dogwood (Cornus amomum Mill.) and highbush cranberry (Viburnum trilobum Marsh.). The third zone [6.4 m (21 ft) wide] consists of warm-season grasses and forbs that reduce sediment load and agricultural chemicals in overland flow. Pure switchgrass (Panicum virgatum L.) or combinations of big bluestem (Andropogon geraldii Vitman), Indian grass (Sorgastrum nutans L.) Canada wild rye (Elymus canadensis L.) and up to 15 different native forbs are planted.
Corn (Zea mays L.) and soybean (Glycine max (L.) Merr.) are grown in alternating years on the row-crop fields and most of them have a narrow strip [< 4m (13 ft)] of grasses and annual weeds along the stream bank (Figure 3b, 3c). The pastures consist of cool-season grasses [primarily Kentucky bluegrass (Poa prantensis L.)] and are confined to the riparian zone and backslope of the narrow stream valley (Figure 3d). The pastures in this study reach are continuously-grazed from about the beginning of May to the end of October, a practice typical of this region of Iowa. The pastures are not divided into paddocks allowing the livestock access to the whole pasture and stream channel during the entire grazing season.
Treatments. The 11 km (6.8 mi) stream reach used in this study was divided into three meandering and two channelized sub-reaches (Figure 1). The three meandering sub-reaches were divided into five treatment segments that included one bordered by a riparian forest buffer, two bordered by row-crop fields that are 6.8 km (4.2 mi) apart, and one cattle and adjacent horse pasture (Figure 1, Table 1). The riparian forest buffer included a 9 yr old and a 6 yr old buffer that were treated as one treatment segment because of the minimal difference in age, structure and development. The two channelized sub-reaches included three treatment segments bordered by row-crop fields (Figure 1, Table 1). Two are adjacent to each other while the other is 3.2 km (2.0 mi) downstream.
Rainfall data. Stream bank erosion data were correlated to total rainfall, mean rainfall, and number of rainfall events greater than 30 mm (1.2 in). The rainfall data were collected in a tipping bucket rain gauge that is part of a recording weather station located within the study reach.
Stream bank erosion pins. The erosion pin method (Wolman, 1959) was used to measure short-time-scale stream bank erosion rates with high resolution (Lawler, 1993) (Figure 4). Stream bank erosion rate (mm) was one of three different variables [the other two were total bank eroding length (%) and stream bank soil loss per unit length of stream bank (kg [m.sup.-1])] used in this study to compare stream bank erosion between the different treatments. Steel rod pins, 762 mm (30 in) long and 6.4 mm (0.25 in) in diameter, were inserted perpendicularly into the bank face. The length of 762 mm (30 in) was chosen because in the past we had witnessed erosion rates of up to 500 mm (19.6 in) per erosion event and Hooke (1979) recommends that at least one third of the pin needs to remain buried so it is not lost during an erosion event. Hooke, (1979) also states that pins up to 800 mm (31.5 in) in length do not interfere with stream bank erosion processes. The 6.4 mm (0.25 in) diameter was small enough to cause minimum disturbance to the stream bank but large enough not to bend under high discharge events (Lawler, 1993).
The horizontal and vertical distances between the pins in the plots were 1 m (3.3 ft) and 0.3 m (0.98 ft), respectively. Vertical distances were smaller because stream bank erosion is more variable vertically than horizontally (Lawler, 1993). Networks of erosion pins with similar vertical and horizontal distances have been found to not influence stream bank erosion processes (Hooke, 1979; Lawler, 1993).
Erosion pin networks were only placed on severe and very severe eroding stream banks because these should account for the majority of the sediment in the stream compared to the moderate and slight eroding bank sites that are more vegetated. As the percentage of vegetation on the stream bank increases stream bank erosion decreases. Beeson and Doyle (1995) found that stream bank erosion in non-vegetated stream bends can be 30 times greater than in vegetated stream bends. Severe eroding bank sites were defined as bare with slumps, vegetative overhang and/or exposed tree roots (USDA-NRCS, 1998). Very severe eroding bank sites were defined as bare with massive slumps or washouts, severe vegetative overhang and many exposed tree roots (Figure 4a) (USDA-NRCS, 1998). Specifically within each treatment segment we identified all the severe and very severe eroding sites and recorded them on a recent aerial photograph (1:24,000) so we could relocate the sites. Each site was assigned a number with number 1 the most downstream eroding site in the treatment segment and increasing upstream. A random numbers program was used to identify the location for the pin plots from the total number of severe and very severe in each treatment segment.
The number of pin plots within each treatment segment was based on the stream treatment segment bank length and the number of severe and very severe eroding bank sites in the treatment segment. Stream bank erosion is highly variable and in the treatment segments with more severe and very severe bank eroding sites more pin plots were established to better capture this variability. The meandering riparian forest buffer segment had seven plots (159 total pins), the two meandering row-crop segments had a total of eight (279 total pins), the two meandering pasture segments had a total of nine (343 total pins), and the three channelized row-crop segments also had a total of nine (291 total pins). Compared to other studies that have used pins, our study had a much larger number of erosion pins (1072) per treatment site, thus a larger number of observations, with frequent measurements (approximately every month except winter) (Lawler, 1993).
Pins were initially installed with approximately 50 mm (2.0 in) exposed. The lengths of the exposed portions of the pins were measured approximately every month from June 1998 to June 1999, except from December 1998 to February 1999 (Figure 4b). During these winter months most plots were not easily accessible or were covered with snow and ice. Stream bank erosion rate was estimated by subtracting the last measurement of the erosion pin from the previous measurement. If that difference was negative it was assumed that deposition and not erosion had occurred. When a pin was completely lost during an erosion event an erosion rate of 600 mm (23.6 in) was used (Hooke, 1979; Lawler, 1993). The mean stream bank erosion rate for the plot was estimated by averaging the bank erosion rate of all the pins in the plot.
Stream bank eroding length and height survey. The height at 2 m (6.6 ft) length intervals and total length of all severe and very severe stream bank eroding sites in each treatment segment were measured. The height of the eroding portion of the stream bank was estimated with a scaled height pole [accuracy of 1 cm (0.4 in)]. By dividing the treatment segment eroding stream bank length by its total stream bank length, the severe and very severe eroding stream bank length percentage was estimated (the second variable used to compare stream bank erosion between treatments) (USDA-NRCS, 1998). Eroding stream bank areas were calculated as the product of the average height [mean from measurements at 2 m (6.6 ft) length intervals] of an eroding length and its length, for each severe and very severe site within a treatment segment. The total stream bank eroding area for each treatment segment was determined as the sum of the eroding stream bank areas within a treatment segment. Mean bank eroding height and total bank eroding lengths and areas are given in Table 3.
Bulk density measurements. A 64 mm (2.5 in) diameter by 406 mm (16 in) long probe was used to collect soil bulk density samples (Blake and Hartge, 1986). The probe was inserted perpendicularly into the bank face at 0.3 m (0.98 ft) intervals (the same as the vertical spacing of the erosion pins) along a randomly selected vertical line on the face of the stream bank. Fifty percent of the erosion pin plots were randomly selected and sampled. When the eroding sites were entirely covered by pins, we collected the samples from the next adjacent eroding site with similar characteristics. The bulk density samples were weighed after drying for 1 d at 105[degrees] C (221[degrees] F).
Texture data. Soil samples for texture where collected at the same location as bulk density samples. The hydrometer method was used for particle size analysis (Gee and Bauder, 1986). In this method hydrometer readings were taken at 40 sec and at 2 hr. To estimate the textural percentages the following equations where used:
% clay + silt = (40 sec reading/sample weight) *100
% sand = 100 - (% clay + silt)
% clay = (2 hr reading/sample weight) *100
% silt = (% clay + silt) - (% clay)
Average stream bank soil textures for each treatment segment are reported in Table 1.
Total stream bank soil loss. To estimate the total stream bank soil loss (Mg) for each treatment segment the eroding stream bank area was multiplied by the mean stream bank erosion rate and the mean bulk density. The stream bank soil loss per unit length of stream bank (kg [m.sup.-1]) (the third variable used to compare stream bank erosion) was estimated by dividing the total stream bank soil loss for each treatment segment by the total stream bank length of the treatment segment.
Data analysis. The general linear model (GLM) in Statistical Analysis System (SAS) is an analysis of variance general purpose procedure (Statistical Analysis System Institute, Inc. 1996) that was used to compare the mean stream bank erosion rates and the mean bulk density of all treatments and treatment segments. The general linear model was used because the dataset was unbalanced. In addition to general linear model we used the SAS procedure stderr, that provides the standard error, and pdiff and contrast that can indicate if the difference between two treatments or treatment segments is statistically significant. The first stream bank erosion rate comparisons were made among the riparian forest buffer, the mean of the two row-crops and the mean of the two grazed pasture treatment segments along meandering sub-reaches (compared 3 means with each other). Stream bank erosion rates were also compared for all five treatment segments in the meandering sub-reaches, the riparian forest buffer, the two row-crops, and the horse and cow pastures (compared five means with each other). Another stream bank erosion rate comparison was made between the stream banks of meandering and channelized sub-reaches bordered with row-crop fields. All stream bank erosion rate comparisons where made for each erosion pin measuring period (Table 2) and for the entire year data was collected. Finally, bulk density comparisons were made among the riparian forest buffer, the mean of the two grazed pastures, and the two row-crop treatment segments along meanders, and the mean of the three channelized row-crop treatment segments.
Results and Discussion
Stream bank erosion is an episodic and seasonal phenomenon (Hagerty et al., 1981). Following these typical episodic stream bank erosion patterns the erosion pin plots in our study experienced 60 to 80% of their erosion during a few days in June 1998 and April 1999 (Table 2) showing the high temporal variability of stream bank erosion. Both periods were characterized by high total and daily rainfall, and repeated rainfall events greater than 30 mm in short periods of time (Table 2). In low order streams discharge is highly influenced by local precipitation (Junk et al., 1989). In June 1998, heavy rainfall led to over-bank flow, with fluid entrainment causing stream bank erosion (Lawler et al., 1999). Fluid entrainment again caused stream bank erosion in April 1999, but freeze and thaw may have been a significant erosional process as well. During the winter months, freeze and thaw cycles cause the soil to become friable (Hagerty et al., 1983; Lawler et al., 1999). Friable soil on the stream banks was observed right after the winter months. Additional evidence of the freeze and thaw effects was that the top erosion pins in half the erosion pin plots lost soil while middle and bottom pins accumulated soil during the winter months when there were no runoff events.
There were statistically significant differences over the entire period in mean stream bank erosion rates among the three treatments along the meandering sub-reaches. Mean stream bank erosion rates for the entire period (June 6th 1998- June 6th 1999) decreased in the following order: row-crops along meanders [387 mm (15.2 in)], continuously-grazed pastures [295 mm (11.6 in)], and riparian forest buffer [142 mm (5.6 in] (P = 0.06). Specifically, the meander riparian forest buffer treatment had a total erosion rate of 199 mm (7.8 in) less than the combined mean of the two row-crop and two continuously-grazed pasture treatments along meanders (P = 0.005). Similarly the riparian forest buffer had a total erosion rate of 79 mm (3.1 in) (P = 0.01) and 56 mm (2.2 in) (P = 0.04), less then the two row-crop and two continuously-grazed pasture treatments along meandering sub-reaches for the two major bank erosion events in June 1998 and April 1999, respectively. Finally, there were no differences in stream bank erosion rates between row-crop treatments along meandering or channelized sub-reaches.
When comparing stream bank erosion rates for spatial variability in the vertical profile there were no significant statistical differences. Specifically, we tested if there was a difference in erosion rates between the top, middle and bottom pins [pins were placed every 30 cm (0.98 ft) in the vertical profile]. In some cases, especially after the winter months, top pins lost soil while bottom pins gained soil (as mentioned above) in about half the pin plots. Stream bank erosion rates did not follow any specific trend in the horizontal and vertical profile indicating its high spatial variability. The mean stream bank erosion rate of pin plots in the same treatment segment was also highly variable indicating the spatial variability.
Total stream bank soil loss per unit stream bank length for each treatment segment was estimated by multiplying the total stream bank eroding area, by the mean stream bank erosion rate and mean bulk density and dividing this product by the treatment segment total bank length. The riparian forest buffer along the meandering reach had 324 kg [m.sup.-1] (218 lbs [ft.sup.-1]) and 224 kg [m.sup.-1] (151 lbs [ft.sup.-1]) less total soil loss than the meandering row-crop and continuously-grazed pasture treatments, respectively (Table 3). The total soil loss per unit length of stream bank for the riparian forest buffer was also lower than the per unit length stream bank soil loss for the channelized row-crop treatments (Table 3). Similarly, the riparian forest buffer treatment reach had a lower percentage of total eroding bank length, than the meandering row-crop and pasture treatment segments (Table 3). Finally, bulk densities followed a slightly different trend. In the meanders, the riparian forest buffer had the lowest mean bulk density whereas the row-crop fields and the continuously-grazed pasture were similar to each other (Table 3). Bulk density was greatest in the channelized row-crop segments (Table 2) (P = 0.08 for differences between bulk densities of the four treatments).
The riparian forest buffer, the row-crop fields, and the continuously-grazed pastures along the meandering reach had different mean stream bank erosion rates, percentage of eroding bank length, and total soil loss per unit length for the entire sampling period. The presence, density, and type of riparian and stream bank vegetation influences stream bank erosion (Beeson and Doyle, 1995). As the percentage of cultivated land in a riparian zone increases, streams deepen and widen (Hamlett et al., 1983). Riparian trees may be replaced with shallow-rooted row-crops that are present for only part of the year. Intensive continuous-grazing also reduces the number of roots on perennial pasture grasses that anchor the soil, while trampling along and on the stream bank increases stream bank instability (Belsky et al., 1999).
Trees, especially willows (Salix species) and silver maple (Acer saccharinum L.), protect the stream bank because of the high density and deep rooting habits that provide greater support to soil to resist fluid entrainment (Shields et al., 1995). Riparian forest buffers had the lowest stream bank erosion rate, percentage of eroding bank length, and soil loss per unit length (Table 3). Although the two ages of riparian forest buffer were treated as one, looking at the percentages of total eroding bank length showed an interesting trend. The riparian forest buffer that was 9 yr old had significantly less (6 %) total eroding bank length compared to the 6 yr old buffer. This difference suggests that as a riparian forest buffer becomes more established and trees mature, stream bank stabilization increases (Odgaard, 1987).
Surprisingly, mean stream bank erosion rates among the riparian forest buffer, the two row-crop fields and two pasture treatment segments along meanders, were also different for the entire period and for the June and April events (P = 0.02, 0.04, and 0.07, respectively), when each treatment segment was examined as a separate mean (comparing 5 means). We did not expect bank erosion rate differences among all of the five treatment segments, when each treatment segment was examined as a separate mean, because two treatment segments were bordered with row-crop fields, and two others were bordered with pastures. Mean stream bank erosion rates for the entire year decreased in the following order: meandering row-crop 1 [522 mm (20.6 in)], meandering cow pasture [408 mm (16.1 in)], meandering row-crop field 2 [253 mm (10.0 in)], meandering horse pasture [183 mm (7.2 in)] meandering riparian forest buffer [142 mm (5.6 in)] (Table 3). This was attributed to the influence of channelization upstream and differences in management for the same land-use practices.
The difference between meandering treatment segments row-crop 1 and 2 may be due to the channelization impacts. Specifically, above the meandering row-crop 1 treatment segment (which had the highest stream bank erosion rate) is a channelized treatment segment (Figure 1). Channelization increases stream velocity providing the water with more power to erode stream banks. In contrast, meandering row-crop 2 had meandering treatment segments upstream from it (Figure 1). Interestingly, between the two row-crop meandering treatment segments the percentages of total bank eroding lengths were similar 44 to 45% (Table 3).
The main difference between the cattle and horse pasture is probably related to differences in their management. The cattle pasture had three times more animal units (kg [ha.sup.-1] or lbs [ft.sup.-1]) than the horse pasture. Animal units (kg [ha.sup.-1] or lbs [ac.sup.-1]) were compared because they give a better estimate of the forage grazed than if the actual number of animals per area were compared (Western, 1997). In addition, cattle tend to spend a lot of time in and close to the stream compared to other livestock (Platt, 1981). The mean stream bank erosion rate was twice as much in the cattle pasture as in the horse pasture (Table 3). Similarly the soil losses per unit length and the percentage of stream eroding bank length also were greater for the cattle pasture than the horse pasture (Table 3). Intensive continuous-grazing increases stream bank erosion (Platt, 1981).
The total stream bank soil loss for the 11 km (6.8 mi) study reach based on our measurements (Table 3) from June 1998 to June 1999 was 4,114 Mg (4,535 tons). Of this total, only 417 Mg (460 tons) was derived from the 27% of stream bank length that was eroding in the 3.2 km (2 mi) long forest riparian buffer treatment segment. Assuming the stream banks of the other treatment segments [7.8 km (4.8 mi)] had been covered by 6 to 9 year old riparian forest buffer vegetation, only 27% of the stream bank length would be eroding and the mean stream bank erosion rate would be 142 mm (5.6 in) for the entire year. At these rates the estimated stream bank soil loss for this period would have been 1,051 Mg (1,159 tons), a decrease of 2,646 Mg (2,917 tons) or 72%.
According to Lawler et al., (1999) 50% or more of the stream sediment load comes from stream bank erosion. A 72% reduction in stream bank erosion would mean that the 50% load suggested by Lawler et al. (1999) would leave only 14% of the stream sediment load coming from bank erosion. Work in the Bear Creek watershed also has shown that riparian buffers of 20 m (66 ft) widths can reduce sediment delivery from overland flow by more than 90% (Lee et al., 2000; 2003). Assuming that overland flow is responsible for the other 50% of the stream sediment load, a 90% reduction from overland flow would leave only 5% of the sediment in streams coming from overland flow. Thus the total stream sediment load from both stream bank erosion and overland flow could potentially be reduced by 81% by riparian forest buffers, a significant reduction. With these kinds of reductions the National Buffer Initiative goal of buffering 3,218,00 km (2,000,000 mi) of streams could have a significant effect on sediment reduction in streams (Soil and Water Conservation Society, 2001). This projection does not take into account Lane's channel equilibrium model that suggests that an equal reduction in stream power would have to accompany a reduction in sediment load (Lane, 1955). If stream power were not reduced the channel would erode sediment from other reaches to maintain the equilibrium.
Summary and Conclusions
Stream bank erosion has high temporal and spatial variability. Although stream bank erosion exhibits high variability, land use has a significant impact on stream bank erosion rates, total bank eroding lengths and total soil losses. In this study intensive row-cropping and intensive continuous grazing sustained high levels of stream bank erosion. Riparian forest buffer vegetation with perennial roots reduced stream bank erosion, significantly. That reduction will probably increase as the buffers continue to mature. If the 6 to 9 year old riparian forest buffers were in place along all the row-crop fields and continuously-grazed pastures of the 7.8 km (4.8 mi) of the non-buffered study reach, stream bank erosion would have been reduced by 72%. While this paper reports results from a local watershed, we believe that the rankings of the treatments used in our study would remain the same in most landscape settings but that absolute stream bank erosion rates and soil losses might be different.
Stream bank erosion is a natural process (Henderson, 1986). The goal of sustainable land use management should sustain minimal levels of stream bank erosion. Riparian forest buffers accomplish this goal effectively. Riparian forest buffers also reduce sediment from overland flow from row-crop fields by up to 90% (Lee et al., 2000; 2003). By reducing the two major sources of sediment load in the streams, riparian forest buffers provide an alternative land-use for riparian zones that will decrease the major non-point source pollutant, sediment. At the same time, riparian forest buffers also are financially attractive to farmers because the Conservation Reserve Program (USDA-NRCS. 1997) subsidizes lost income for agricultural land planted in riparian forest buffers.
Table 1. Soil and channel characteristics of the eight treatment segments along Bear Creek in central Iowa. Texture and particle size percentages are only from the stream bank eroding sites. Treatment Soil series Texture Clay segments Eroding sites Eroding sites (%) Meandering Coland Clay loam 29 forest buffer Meandering Coland Sandy clay loam 23 row-crop 1 Meandering Spillville-Coland Clay loam 32 row-crop 2 complex Meandering Spillville-Coland Clay loam 33 horse pasture complex Meandering Spillville-Coland Clay loam 31 cow pasture complex Channelized Coland Sandy clay loam 29 row-crop 1 Channelized Coland Clay loam 33 row-crop 2 Channelized Coland Clay loam 35.5 row-crop 3 Treatment Silt Sand Average stream Total stream segments Eroding sites Eroding sites bank height bank length (%) (%) (m) (km) Meandering 28 43 1.8 6.37 forest buffer Meandering 20 57 2.0 1.41 row-crop 1 Meandering 29 39 2.3 2.75 row-crop 2 Meandering 31 36 2.3 2.21 horse pasture Meandering 32 37 2.1 2.14 cow pasture Channelized 26 45 2.0 1.17 row-crop 1 Channelized 25 42 1.8 3.75 row-crop 2 Channelized 34.5 30 2.3 2.21 row-crop 3 Table 2. Seasonal stream bank erosion rates (based on the means from all 33 erosion pin plots), mean and total rainfall amounts and number of events along Bear Creek in central Iowa. Only events greater than 30 mm resulted in high stream bank erosion rates. Period between erosion Stream bank Total Mean pin measurements erosion rate rainfall rainfall (mm)* (mm)# (mm*[d.sup.-1]) [section] 6 June - 21 June 1998 81 242 16.1 22 June - 7 July 1998 19 65 5.0 8 July - 21 July 1998 0 54 3.2 22 July - 8 Sept. 1998 2 125 1.9 9 Sept. - 13 Oct. 1998 1 72 2.0 14 Oct. - 28 Nov. 1998 0 81 1.8 29 Nov. 1998 - 27 Feb. 1999 10 33 0.4 28 Feb. - 4 April 1999 -4* 21 0.6 5 April - 18 April 1999 84 150 10.7 19 April - 9 May 1999 29 66 2.3 10 May - 6 June 1999 28 15 0.5 Period between erosion No. of rainfall No. of rainfall pin measurements events events 1-15 mm 16-30 mm 6 June - 21 June 1998 4 1 22 June - 7 July 1998 4 2 8 July - 21 July 1998 1 0 22 July - 8 Sept. 1998 14 1 9 Sept. - 13 Oct. 1998 8 1 14 Oct. - 28 Nov. 1998 7 2 29 Nov. 1998 - 27 Feb. 1999 8 0 28 Feb. - 4 April 1999 2 0 5 April - 18 April 1999 2 1 19 April - 9 May 1999 5 0 10 May - 6 June 1999 5 4 Period between erosion No. of rainfall pin measurements events [greater than or equal to] 31-mm 6 June - 21 June 1998 4 22 June - 7 July 1998 1 8 July - 21 July 1998 1 22 July - 8 Sept. 1998 1 9 Sept. - 13 Oct. 1998 0 14 Oct. - 28 Nov. 1998 0 29 Nov. 1998 - 27 Feb. 1999 0 28 Feb. - 4 April 1999 0 5 April - 18 April 1999 2 19 April - 9 May 1999 1 10 May - 6 June 1999 0 *Negative stream bank erosion rates denote deposition. #Includes all precipitation events. [section]Estimated by dividing the total precipitation by the number of days in that period. Table 3. Stream bank erosion parameters from eroding sites bordering different land-use practices along Bear Creek in central Iowa. Mean stream bank Treatment erosion rate* segments June '98# April '99[section] Year 1[dagger] (mm) (mm) (mm) Meandering 44 (23) 58 (20) 142 (50) forest buffer Meandering 172 (43) 183 (38) 522 (94) row-crop 1 Meandering 90 (23) 87 (22) 253 (54) row-crop 2 Meandering 131 (25) 135 (22) 387 (54) row-crop 1 & 2 Meandering 58 (23) 69 (20) 183 (50) horse pasture Meandering 170 (43) 116 (38) 408 (94) cow pasture Meandering 114 (24) 92 (22) 295 (54) cow & horse pastures Channelized 87 (22) 86 (23) 239 (54) row-crop 1, 2 & 3 Mean Total Total Treatment bulk stream eroding segments density bank length bank length (g*[mL.sup.-1]) (km) (km) (%)[double dagger] Meandering 1.24 (0.04) 6.37 1.73 27 forest buffer Meandering 1.34 (0.05) 1.41 0.63 45 row-crop 1 Meandering 1.33 (0.03) 2.75 1.21 44 row-crop 2 Meandering 1.33 (0.03) 4.16 1.84 44 row-crop 1 & 2 Meandering 1.34 (0.03) 2.21 0.71 32 horse pasture Meandering 1.33 (0.05) 2.14 1.10 51 cow pasture Meandering 1.33 (0.03) 4.35 1.81 42 cow & horse pastures Channelized 1.38 (0.03) 7.14 2.14 30 row-crop 1, 2 & 3 Mean Total Treatment eroding eroding segments bank height bank area** (m) ([m.sup.2]) Meandering 1.3 2369 forest buffer Meandering 1.4 1015 row-crop 1 Meandering 1.3 1642 row-crop 2 Meandering 1.4 2657 row-crop 1 & 2 Meandering 1.2 938 horse pasture Meandering 1.5 1638 cow pasture Meandering 1.4 2576 cow & horse pastures Channelized 1.3 2445 row-crop 1, 2 & 3 Soil Loss Treatment Total[double dagger][double dagger] Unit segments length length[parallel] (Mg) (kg*[m.sup.-1]) Meandering 417 65 forest buffer Meandering 910 650 row-crop 1 Meandering 725 259 row-crop 2 Meandering 1635 389 row-crop 1 & 2 Meandering 233 106 horse pasture Meandering 1009 480 cow pasture Meandering 1242 293 cow & horse pastures Channelized 820 115 row-crop 1, 2 & 3 *Values in parenthesis are standard errors estimated using SAS. #Measuring period was June 6th to June 26th 1998. [section]Measuring period was April 4th to April 18th 1999. [dagger]Measuring period was June 6th 1998 to June 6th 1999. [double dagger]Estimated by dividing total eroding bank length by total stream bank length and multiplied by 100. **Estimated by multiplying total eroding length and mean eroding height of each severe and very severe eroding bank site that was surveyed. NOTE that this was not estimated by multiplying total eroding length with average eroding bank height for each treatment segment. [double dagger][double dagger]Estimated by multiplying Year 1 mean stream bank erosion rate, mean bulk density, and total eroded bank area. [parallel]Estimated by dividing total soil loss by total stream bank length.
This research has been funded in part by the Iowa Department of Natural Resources under the Federal Nonpoint Source Management Program (Section 319 of the Clean Water Act), and the Leopold Center for Sustainable Agriculture, a State of Iowa Institution located at Iowa State University (ISU), Ames, Iowa. We would like to thank Dr. P. Dixon and Dr. P. Hinz for their help with the statistical analysis of the project. In addition, we would like to thank Jon Handrik and a number of ISU undergraduate and graduate students for helping collect the data.
Beeson, C.E. and P.F. Doyle. 1995. Comparison of bank erosion at vegetated and non-vegetated channel bends. Water Resource Bulletin 31:983-990.
Belsky. A.J., A. Martze, and S. Uselman. 1999. Survey of livestock influences on stream and riparian ecosystems in the western United States. Journal of Soil and Water Conservation 54(3):419-431.
Bharati, L., K-H Lee, T.M. Isenhart, and R.C. Schultz. 2002. Riparian zone soil-water infiltration under crops, pasture and established buffers. Agroforestry Systems 56:249-257.
Blake, G.R. and K.H. Hartge. 1986. Bulk Density. Pp. 363-375. In: Klute A. (ed.) Methods of soil analysis: Part 1. Physical and mineralogical methods. 2nd ed. Madison. Wisconsin. American Society of Agronomy and Soil Science Society of America. 364-367 pp.
Burkhart, M.R., S.L. Oberle, M.J. Hewitt, and J. Picklus. 1994. A framework for regional agroecosystems characterization using the national resources inventory. Journal of Environmental Quality 23:866-874.
DeWitt, T.A. 1984. Soil survey of Story County, Iowa. U.S. Department of Agriculture-Soil Conservation Service, Washington. D.C.
Federal Interagency Stream Restoration Working Group. 1998. Stream corridor restoration. Principles, processes and practices. The National Academy Press. Washington, D.C.
Gee, G.W. and J. W. Bauder. 1986. Particle-size analysis. Pp. 383-411. In: Klute A. (ed.) Methods of soil analysis: Part 1. Physical and mineralogical methods. 2nd ed. Madison, WI: American Society of Agronomy and Soil Science Society of America. 404-409 pp.
Hagerty, D.J., M.F. Spoor, and C.R. Ullrich, 1981. Bank failure and erosion on the Ohio River. Engineering Geology 17:141-158.
Hagerty. D.J., M. Sharifounnasab, and M.F. Spoor. 1983. Riverbank erosion: A case study. Bulletin of the Association of Engineering Geologists 20:411-437.
Hamlett, J.M., J.L. Baker, and H.P. Johnson. 1983. Channel morphology changes and sediment yield for a small agricultural watershed in Iowa. Transactions of the American Society of Agricultural Engineering 26:1390-1396.
Henderson, J.E. 1986. Environmental designs for streambank protection projects. Water Resource Bulletin 22:549-558.
Hooke, J.M. 1979. An analysis of the processes of river bank erosion. Journal of Hydrology 42:39-62.
Hooke, J.M. 1980. Magnitude and distribution of rates of river bank erosion. Earth Surfaces and Processes 5:143-157.
Iowa Department of Natural Resources (IDNR). 1997. Water quality in Iowa during 1994 and 1995. Iowa Department of Natural Resources. Des Moines, Iowa.
Isenhart. T.M., R.C. Schultz, and J.P. Colletti, 1998. Watershed restoration and agricultural practices in the Midwest: Bear Creek of Iowa. Chapter 19. Pp. 318-334. In: J.E. Williams, M.P. Dombeck, and C.A. Woods (eds.). Watershed restoration: Principles and practices. American Fisheries Society, Bethesda, Maryland, 322-324 pp.
Junk, W.J., P.B. Bayley, and R.E. Sparks. 1989. The flood-pulse concept in river-floodplain systems. Pp. 110-127. In: D.P. Dodge (ed.) Proceedings of the International Large River Symposium. Canadian Special Publications in Fisheries and Aquatic Sciences 106. Fisheries and Oceans, Communications Directorate. Ottawa, Canada, 113 pp.
Lane, E.W. 1955. The importance of fluvial morphology in hydraulic engineering. Proccedings of the American Society of Civil Engineers 81:1-17.
Lawler, D.M. 1993. The measurement of river bank erosion and lateral channel change: A review. Earth Surfaces and Processes and Landforms 18:777-821.
Lawler, D.M. J.R., Grove. J.S. Couperwaite, and G.J.L. Leeks. 1999. Downstream change in river bank erosion rates in the Swale-Ouse system, northern England. Hydrological Processes 13:977-992.
Lee, K-H., T.M. Isenhart, R.C. Schultz, and S.K. Mickelson. 1999. Nutrient and sediment removal by switchgrass and cool-season grass filter strips in Central Iowa. Agroforestry Systems 44:121-132.
Lee, K-H., T.M. Isenhart, R.C. Schultz, and S.K. Mickelson. 2000. Multi-species riparian buffer system in Central Iowa for controlling sediment and nutrient losses during simulated rain. Journal of Environmental Quality 29:1200-1205.
Lee, K-H., T.M. Isenhart, and R.C. Schultz. 2003. Sediment and nutrient removal in an established multi-species riparian buffer. Journal of Soil and Water Conservation 58(1):1-8.
Lowrance, R., S.M. Dabney, and R.C. Schultz. 2002. Improving water and soil quality with conservation buffers. Journal of Soil and Water Conservation 57(2):36A-43A.
Marquez. C.O., C.A. Cambardella. T.M. Isenhart, and R.C. Schultz. 1999. Assessing soil quality in a riparian buffer strip system by testing organic matter fractions. Agroforestry Systems 44:133-140.
Odgaard, A.J. 1987. Streambank erosion along two rivers in Iowa. Water Resource Research 23:1225-1236.
Platt, W.S. 1981. Effects of sheep grazing on a riparianstream environment. Note INT-307. U.S. Forest Service Research, Ogden, Utah.
Ritter, W.F. 1988. Reducing impacts of nonpoint pollution from agriculture: A review. Journal of Environmental Science and Health (Part A) 23:654-667.
Statistical Analysis System (SAS) Institute, Inc. 1996. SAS/STAT User's Guide. Release 6.12 ed. Cary. North Carolina.
Schumm, S.A., M.D. Harvey, and C.C. Watson. 1984. Incised channels. Morphology, dynamics and control. Water Resource Publication, Littleton, Colorado.
Shields, F.D., Jr., A.J. Bowie, and C.M. Cooper. 1995. Control of streambank erosion due to bed degradation with vegetation and structure. Water Research Bulletin 31:475-489.
Simpkins, W.W., T.R. Wineland, R.J. Andress, D.A. Johnston, T.M. Isenhart, and R.C. Schultz. 2002. Hydrogeological constraints on riparian buffers for reduction of diffuse pollution: Examples from the Bear Creek Watershed in Iowa. Journal of Water Science and Technology 45:61-68.
Soil and Water Conservation Society. 2001. Realizing the promise of conservation buffer technology. A report based on the National Conservation Buffer Workshop, June 12-13, 2001. Nebraska City, Nebraska. Soil and Water Conservation Society, Ankeny, Iowa.
Tufekcioglu, A., J.W. Raich, T.M. Isenhart, and R.C. Schultz. 1999. Root biomass, soil respiration, and root distribution in crop fields and riparian buffer zones. Agroforestry Systems 44:163-174.
Tufekcioglu, A., J.W. Raich, T.M. Isenhart, and R.C. Schultz. 2001. Soil respiration within riparian buffers and adjacent crop fields. Plant and Soil 229:117-124.
U.S. Department of Agriculture-Natural Resources Conservation Service (USDA-NRCS). 1997. Riparian forest buffer. Conservation Practice Standard, Code 392. U.S. Department of Agriculture and Natural Resources Conservation Service. Des Moines, Iowa.
U.S. Department of Agriculture-Natural Resources Conservation Service (USDA-NRCS). 1998. Erosion and sediment delivery. Field Office Technical Guide Notice No. IA-198. U.S. Department of Agriculture and Natural Resources Conservation Service, Des Moines, Iowa.
Webber, D.A. 2000. Comparing estimated surface flow-paths and sub-basins derived from digital elevation models of Bear Creek watershed in central Iowa. Thesis (M.S.). Iowa State University. Ames, Iowa.
Western, J.M. 1997. Cow-calf production from rotationally stocked alfalfa grass or brome grass pastures. Thesis (M.S.). Iowa State University, Ames, Iowa.
Wilkin, D.C. and S.J. Hebel. 1982. Erosion, redeposition, and delivery of sediment to Midwestern streams. Water Resource Research 18:1278-1282.
Wolman, M.G. 1959. Factors influencing erosion of a cohesive river bank. American Journal of Science 257:204-216.
George N. Zaimes is a graduate student, Richard C. Schultz is a professor, and Thomas M. Isenhart is an associate scientist in the Department of Natural Resources Ecology and Management at Iowa State University in Ames, Iowa.
|Printer friendly Cite/link Email Feedback|
|Author:||Zaimes, G.N.; Schultz, R.C.; Isenhart, T.M.|
|Publication:||Journal of Soil and Water Conservation|
|Date:||Jan 1, 2004|
|Previous Article:||A statewide assessment of the impacts of phosphorus-index implementation in Pennsylvania.|
|Next Article:||Storm runoff and soil erosion in south Florida as affected by water table fluctuations.|