PAM APPLICATION METHOD AND ELECTROLYTE SOURCE EFFECTS ON PLOT SCALE RUNOFF AND EROSION

Transcription

1 PAM APPLICATION METHOD AND ELECTROLYTE SOURCE EFFECTS ON PLOT SCALE RUNOFF AND EROSION J. R. Peterson, D. C. Flanagan, J. K. Tishmack ABSTRACT. Previous research has indicated that polyacrylamide (PAM) soil amendments can be effective in reducing runoff and soil erosion by reducing soil sealing and stabilizing soil structure. Furthermore, the application of a multivalent electrolyte (Ca ++ ) in addition to PAM has been shown to further reduce runoff volume and sediment yield on some soils. A field study was conducted using simulated rainfall to test the effectiveness of the method of PAM application (dry or in solution) and the effectiveness of two sources of Ca ++ electrolytes. Treatments using an application of a liquid PAM solution that was allowed to dry on the soil surface were the most effective in reducing total runoff (62% to 76% reduction compared to control) and total sediment yield (93% to 98% reduction compared to control). Spraying of PAM in solution was significantly more effective in controlling runoff and erosion than was the dry granular application for the rainfall events simulated in this study. The slope of regressing sediment yield rate on runoff rate was used as a measure of erodibility. Sprayed PAM treatments dramatically reduced erodibility compared to the control. Dry PAM application also reduced soil erodibility compared to the control, but not as dramatically as the sprayed PAM. For intense rainstorms on initially dry soil, we recommend using a sprayed application of PAM for the best erosion control. More research is needed to determine whether this holds true for less intense storms under different field conditions. Keywords. Polyacrylamide, PAM, Soil amendments, Soil erosion, Runoff. Construction sites, roadside embankments, and landfill caps are examples of critical sites that are prone to serious erosion prior to establishment of vegetation. Daniel et al. (1979) reported that erosion rates at construction sites can approach 163 Mg ha 1 yr 1, which is about 18 times greater than maximum tolerable rates from agricultural lands. Common methods of reducing erosion at critical sites are mulch, sod, and man made blankets (e.g., jute, coconut fiber). Recently, the use of polyacrylamide (PAM) to control erosion at these sites has been investigated (Flanagan et al., 2002a, 2000b). Adoption of PAM for use in construction projects depends on demonstrated performance comparable to standard methods, cost effectiveness, and relative ease of application. The effectiveness of mulch at reducing runoff and erosion has been demonstrated by Mannering and Meyer (1963), Meyer et al. (1970), Lattanzi et al. (1974), and Krenitsky et al. (1998). Benefits of mulch addition were decreased area of bare soil exposed to rainfall detachment, Article was submitted for review in March 2002; approved for publication by the Soil & Water Division of ASAE in August The use of trade names does not imply endorsement by Purdue University or the USDA ARS. The authors are Joel R. Peterson, ASAE Member Engineer, Visiting Assistant Professor, and Dennis C. Flanagan, ASAE Member Engineer, Agricultural Engineer, USDA ARS National Soil Erosion Research Laboratory, Department of Agricultural and Biological Engineering, Purdue University, West Lafayette, Indiana, and Jody K. Tishmack, Ash Management Coordinator/Continuing Lecturer, Department of Agricultural and Biological Engineering, Purdue University, West Lafayette, Indiana. Corresponding author: Joel R. Peterson, NSERL, 1196 Soil Building, Purdue University, West Lafayette, IN 47907; phone: ; fax: ; e mail: petejoel@ecn.purdue.edu. restricted movement of soil by splash, reduced average flow velocity, and decreased amount of runoff. While mulch and man made materials can be effective, rilling can develop beneath the mulch or blanket (Meyer et al., 1972; Kramer and Meyer, 1969; Foster et al., 1982a; Thompson et al., 2001). The cost of PAM application can be small in comparison to traditional erosion control measures such as mulch. Chaudhari and Flanagan (1998) reported the cost of traditional mulch, adhering to Indiana Department of Transportation guidelines, was $1900 ha 1, compared to $160 ha 1 for PAM. Mulches, sod, and other erosion control materials limit erosion by reducing the effect of rainfall impact and by reducing the shear stress of flowing water imparted to the soil surface. Raindrop impact can lead to sealing of the soil surface. Seal formations are caused by two primary mechanisms: (1) physical disintegration of soil aggregates by raindrop impact, and (2) physicochemical dispersion and migration of clay particles in the 0.1 to 0.5 mm depth, where they can clog pores (Levy et al., 1992; Shainberg and Levy, 1994). Seal formations can also be prevented by introducing multivalent cations at the soil surface to reduce chemical dispersion of clay particles and/or by stabilizing aggregates at the soil surface by polymeric soil conditioners, such as PAM (Stern et al., 1991). Polyacrylamide is a water soluble, synthetic organic polymer high in molecular weight that primarily interacts with the clay fraction of soils (Seybold, 1994) and has been proven to be superior to other polymers in erosion control applications (Shainberg and Levy, 1994). Polyacrylamide stabilizes the soil by reducing repulsive forces among clay particles and acts as a bridge between soil particles in an aggregate by bonding with the particles (Ben Hur, 1994). Polyacrylamide can be synthesized in cationic, nonionic, or anionic forms, though anionic PAM has been found to be the Transactions of the ASAE Vol. 45(6): American Society of Agricultural Engineers ISSN

2 most effective for erosion control (Shainberg and Levy, 1994). The benefits of PAM are enhanced by the introduction of an electrolyte source (multivalent cations) that helps to create a cation bridge for the polymer to adsorb to the soil (Shainberg et al., 1990; Levy et al., 1992; Smith et al., 1990). Typically, the multivalent cations are introduced by application of a gypsiferous material. Phosphogypsum (PG), gypsum, and different types of coal combustion by products (CCBPs) are gypsiferous materials that have been used to enhance the electrolyte concentration of the soil solution. While several types of gypsiferous material have been used to enhance the electrolyte content of the soil solution, little research exists comparing the effectiveness of different gypsiferous materials in combination with PAM. Peterson et al. (2002) found that gypsum was a better source of electrolyte to use in combination with PAM than a class C ponded fly ash because of its greater solubility. Keren and Shainberg (1981) noted that the dissolution rate of gypsum is correlated with efficiency in maintaining high infiltration rates. Effectiveness of the different materials was related to particle size and solubility (ability to release electrolytes) of the gypsiferous material. Norton et al. (1993) concluded that gypsiferous materials with greater solubilities were more effective, and that the coarser the material, the less soluble it was due to surface area constraints. Flanagan et al. (1997a, 1997b) found that a fluidized bed combustion bottom ash (FBCBA) subject to simulated deionized rainfall significantly increased infiltration on interrill plots compared to untreated soil subject to simulated tap water or deionized rainfall. The FBCBA also significantly reduced sediment concentrations and sediment discharge rates on rill subplots on initially dry soil. Flanagan et al. (1997b) recommended using FBCBA in conjunction with PAM. Previous research indicates that PAM is more effective at controlling rill erosion than interrill erosion. The effect of PAM on interrill and rill runoff and soil loss was examined by Flanagan et al. (1997a, 1997b). Treatments included PAM applied at 20 kg ha 1 under both deionized and tap water simulated rainfall, a control, and an FBCBA under deionized rainfall. They found that interrill soil loss was not significantly different between treatments. For the rill subplots, the 20 kg ha 1 PAM treatments reduced detachment compared to the control. Flanagan et al. (2002 a, 2000b) conducted simulated and natural rainfall experiments on a steep slope (34% to 37%) to evaluate the effects of PAM on runoff, erosion, and seedling emergence of a grass mixture. Treatments were a control, PAM sprayed at 80 kg ha 1, and PAM plus gypsum applied at 80 kg ha 1 and 5 Mg ha 1, respectively. Both the PAM and PAM plus gypsum treatments significantly reduced runoff and sediment yield over the control. In the simulated rainfall experiment, the PAM plus gypsum treatment reduced runoff by 52% and sediment yield by 91%, compared to the control, over the sequence of rainfall events. Reductions in runoff from the natural rainfall study ranged from 28% to 32%, while sediment yield reductions ranged from 45% to 53% over the 3 month study period. For the simulated rainfall conditions, Flanagan et al. (2002a) noted a much more prominent rill network in the untreated control plots and in the PAM treated plots. Polyacrylamide is applied by dissolving it in water and then spraying it on the soil surface. Mitchell et al. (1996) noted that because of the high viscosity of PAM in solution, adequate dilution with water to spray the PAM solution may require excessive amounts of water, causing runoff during application or requiring several applications. For these reasons, Mitchell et al. (1996) suggested that dry application of PAM should be studied. Agassi and Ben Hur (1992) reported that the application of PAM in solution was problematic due to its low dissolution rate and high viscosity, causing them to suspend use of PAM in their experiment. In addition, Bjorneberg (1988) noted that the effectiveness of PAM, when sprayed as a solution, might be reduced through breaking or shearing action when placed in a recirculating pump. Dry application of PAM would alleviate the problems associated with of excessive amounts of water, high viscosity of the PAM solution, and shearing of the PAM molecules in a recirculating pump. Peterson et al. (2002) studied the effectiveness of granular application of PAM using small interrill erosion pans (32 cm wide by 45 cm long) in a laboratory setting. They found that sprayed PAM was slightly more beneficial in controlling runoff than granular application of PAM, but differences in total sediment yield were not statistically different. The study by Peterson et al. (2002) focused on interrill erosion. Research is needed to determine whether their conclusions apply to a plot scale situation where rill processes may become significant. Reduction of erosion and runoff through the use of PAM and gypsiferous material has been well documented. However, examination of PAM application method (granular versus solution) in combination with different sources of gypsiferous material in a rill dominated plot scale study has not been documented. Objectives of this study were (1) to determine the effect of application method (granular versus solution) of PAM on runoff and erosion and (2) to test the effectiveness of alternatives to gypsum as a means to increase electrolyte concentrations in runoff for a worst case scenario of a large storm event on a newly disturbed soil. Our main hypotheses were that sprayed application of PAM would lead to significantly greater reduction in runoff and erosion than granular PAM, and that no differences in runoff or sediment would result from application of two different gypsiferous materials. MATERIALS AND METHODS Twelve field plots measuring 3.0 m wide by 9.1 m long were constructed on a hillslope one mile south of the Purdue University campus in West Lafayette, Indiana. This location had previously been used as a sand and gravel quarry. The hillslope was first graded to an approximate 20% slope. The soil used in the experiment was a silty clay topsoil (18% sand, 40% silt, 42% clay, 3.4% OM, 325 mg kg 1 Mg, 3800 mg kg 1 Ca, 22.0 CEC meq 100 g 1 ) taken from the floodplain of the Wabash River by Purdue University Facilities, which uses the soil for various landscaping projects around campus. Plot locations were staked, and approximately 12 m 3 of soil was placed at each plot location with a front end loader and then raked and leveled by hand to evenly cover the plot to a depth of 25 cm to 35 cm. Plots were then bordered with corrugated, galvanized sheet metal to contain runoff. Average slope after raking and leveling was 16.6% µ0.3%. After plot construction, each plot was rototilled (7 to 10 cm depth) and 1860 TRANSACTIONS OF THE ASAE

3 subsequently covered with plastic sheeting to protect the plot from rainfall and to eliminate weed growth. Treatments in the study were (1) an untreated control, (2) PAM in liquid solution and Nutra Ash (PAMW+NA), (3) granular PAM and Nutra Ash (PAMD+NA), and (4) PAM in solution and SoilerLime (PAMW+SL). Polyacrylamide in solution will be referred to as PAMW, and dry PAM application will be referred to as PAMD. For all PAM treatments, the polymer was applied at a rate of 60 kg ha 1 in a 1300 mg L 1 solution with deionized water. The PAM used was anionic Percol 336 (now Magnafloc 336), a commercially available material (Ciba Specialty Chemicals Corp., Suffolk, Va.) having 32% charge density and a molecular mass of 20 Mg mol 1. SoilerLime (applied at 4.3 Mg ha 1 ) is a liming/fertilizer amendment made by blending composted alkaline coal ash from the Purdue University power plant and fermentation by product from the Eli Lilly Corporation. Nutra Ash (applied at 8.0 Mg ha 1 ) is marketed as a liming/fertilizer supplement. Nutra Ash is ponded class C fly ash that has been mined and crushed. Important properties of SoilerLime and Nutra Ash are listed in table 1. The application rates of SoilerLime and Nutra Ash were chosen so as to provide an equivalent amount of Ca as contained in the 5 Mg ha 1 of gypsum used by Chaudhari (1999). The day before each experimental run, each plot was rototilled and raked, soil moisture content samples were collected, and the plot was seeded with grass mixture. For control plots only, the moisture content samples were collected the day of the experiment. After collecting the moisture sample, soil amendments were applied as required. The gypsiferous material was applied before PAM application. The PAM was applied using a sprayer and a roller pump powered by a 2.2 kw motor. Plots were assigned treatments in a random fashion. Each plot underwent a sequence of three sub runs of simulated rainfall with de ionized water using a programmable rainfall simulator (Foster et al., 1982b). The first sub run (dry run) was targeted for a rainfall rate of 75 mm hr 1 for one hour followed by a period of one hour of no rainfall. This was followed by a wet run of 75 mm hr 1 for one hour followed by a half hour break. Then a very wet run was conducted in three stages with the following target rainfall sequence: 75 mm hr 1 for 15 min (very wet I), 28 mm hr 1 for 15 min (very wet II), and 100 mm hr 1 for 15 min (very wet III). The dry run corresponds to a 1 hr, 100 yr precipitation event for West Lafayette, Indiana. Cumulative rainfall amounts following the wet and very wet runs both correspond to events exceeding a 100 yr return period for their respective durations (Hershfield, 1961). The simulated rainfall sequence was chosen to test and compare the effects Table 1. Material characterization of SoilerLime and Nutra Ash. % Dry Basis SoilerLime Nutra Ash Calcium (Ca) Magnesium (Mg) Calcium carbonate equivalent (CCE) Passing U.S. sieve #8 (2.36 mm) Passing U.S. sieve #20 (850 µm) Passing U.S. sieve #60 (250 µm) of the treatments on initially dry soil, on moist field conditions, and on saturated field conditions. The time for concentrated runoff to begin was noted for each sub run, at which time sample collection began. After runoff initiation, samples were collected every three minutes. Discharge rate was measured by collecting runoff at the outlet trough in buckets for a known time. Replicate sediment concentration samples were collected in one liter bottles immediately after the discharge rate was measured. Total runoff volume was calculated as the integration of the discharge measurements over time. Sediment yield was calculated by multiplying the discharge rate by the average sediment concentration, averaging the replicated concentration samples, and integrating over time. Simulated rainfall depth was measured by placing rain gages at 0, 3, 6, and 9 m along each side and one gage at the mid point of the top of each plot. In addition, two small metal troughs (20 mm wide by 3.4 m long) were placed diagonally across the plot to collect rainfall. Data from the rain gages and troughs were averaged to produce a composite rainfall value. Runoff and sediment yield totals were used to determine treatment effects. Analysis of variance (ANOVA) was used to determine differences for runoff and sediment totals between treatments using least squares estimators. To account for differences in runoff caused by changes in precipitation intensity, the ANOVA procedure was carried out using the difference between rainfall and runoff. The Tukey multiple comparison procedure at P < 0.05 was used to compare treatment means. A sequential dilution dissolution analysis was performed by A&L Great Lakes Laboratories on both SL and NA in order to better understand each amendment s effect on electrolyte levels over the course of the experiment. Fifteen grams of material were placed in a 500 ml plastic bottle. Then, 300 ml of ultra pure water was added, and the bottle was mechanically shaken for 10 minutes. At the end of 10 minutes, an 11.0 cm Whatman #1 filter was placed in a buchner funnel, under suction, and the contents of the sample bottle were filtered. All solids were recovered from the filter paper, including the filter itself, and placed back into the mixing bottle. Another 300 ml of ultra pure water was added, and the procedure followed again. Thus, three samples, representing 10, 20, and 30 minutes, were obtained. The filtrate was analyzed for Ca, Mg, Na, K, S, B, Cu, Fe, Mn, and Zn. Only Ca, Mg, and Na are reported here. Samples were read using ICP water matrix standards. Concentrations of the selected constituents are indicated in table 2. Table 2. Results of sequential dilution dissolution study. Time Concentration (ppm) (min) Ca Mg Na SoilerLime < < <1 7 Nutra Ash < < <1 200 Vol. 45(6):

4 Treatment Dry Table 3. Total runoff and sediment yield for each treatment by sub run. [a] Total Runoff Total Sediment Yield mm % reduction of control Mg ha 1 % reduction of control Control 3.60 b 1.64 a PAMD+NA a a 39 PAMW+NA 0.09 b b 100 PAMW+SL 0.46 b b 100 Wet Control b a PAMD+NA a a 29 PAMW+NA c b 96 PAMW+SL 8.41 c b 99 Very wet I Control b 6.05 a PAMD+NA a b 33 PAMW+NA 2.53 c c 96 PAMW+SL 1.01 c c 99 Very wet II Control 3.72 b 0.85 a PAMD+NA 6.63 a ab 27 PAMW+NA 1.53 c b 93 PAMW+SL 1.00 c b 97 Very wet III Control b a PAMD+NA a b 45 PAMW+NA b c 85 PAMW+SL 6.94 c c 96 Total Control b a PAMD+NA a a 31 PAMW+NA c b 93 PAMW+SL c b 98 [a] When followed by the same letter, runoff and sediment values for a given sub run are not significantly different at α = 0.05 using the Tukey multiple comparison method. RESULTS AND DISCUSSION DISSOLUTION ANALYSIS Nutra Ash resulted in a lower concentration of calcium in solution than the SoilerLime (table 2). Both materials maintained fairly uniform Ca concentrations over the course of the dissolution analysis. In addition to the SoilerLime material being more soluble, it also had a finer particle size distribution, which means that it had a greater exposed surface area per unit mass than did the Nutra Ash. The SoilerLime may have been more soluble because it had a finer particle size distribution. A rapid decrease in the concentration of a given ion is often referred to as a first flush, in which easily dissolvable minerals are quickly depleted when water first contacts a material. Since Nutra Ash had been exposed to moisture for several years while in a landfill, it likely did not contain as many soluble minerals as it initially had. Grinding of the material may have exposed new surface area that had slightly higher solubility than the otherwise cemented matrix. Table 4. Final runoff and sediment yield rates. [a] Final Runoff Rate Final Sediment Yield Rate % reduction of control Mg ha 1 hr 1 % reduction of control mm hr 1 Dry Control 17.8 b a PAMD+NA 48.8 a ab 34 PAMW+NA 0.7 c b 100 PAMW+SL 3.3 bc b 100 Wet Control 49.9 a a PAMD+NA 70.0 a b 39 PAMW+NA 25.5 b c 93 PAMW+SL 16.0 b c 98 Very wet I Control 53.0 b a PAMD+NA 72.9 a b 34 PAMW+NA 21.1 c c 91 PAMW+SL 8.6 c c 99 Very wet II Control 16.4 a 4.03 a PAMD+NA 27.5 a a 30 PAMW+NA 2.8 b a 97 PAMW+SL 3.2 b a 97 Very wet III Control 81.0 ab a PAMD+NA a b 49 PAMW+NA 71.0 b bc 83 PAMW+SL 44.0 c c 94 [a] When followed by the same letter, runoff and sediment values for a given sub run are not significantly different at α = 0.05 using the Tukey multiple comparison method. RUNOFF Simulated rainfall amounts for the dry, wet, and very wet sub runs and the total were 79.0 µ1.6 mm, 79.2 µ1.9 mm, 56.7 µ1.4 mm, and µ4.6 mm, respectively (95% confidence limits). One of the keys to the effectiveness of PAM is its ability to preserve aggregate structure, thereby reducing surface sealing and promoting increased infiltration. This was evidenced in the increased time to runoff initiation exhibited by the PAMW treatments. The average time to runoff initiation was 64.4 min and 65.2 min for the PAMW+NA and PAMW+SL treatments, respectively. Average time to runoff initiation for the control and PAMD+NA treatments was significantly less ( = 0.05), 23.6 min and 20.6 min, respectively. Apparently, conditions were not suitable for the granular PAM to become activated during the first run on initially dry soil. The 1 hr, 100 yr precipitation event for West Lafayette, Indiana, is 70 mm. These results indicate that, on the average, liquid PAM treatments will produce negligible runoff for a 1 hr, 100 yr rainfall event under these field conditions. Table 3 lists runoff and sediment yield totals for each sub run as well as for all runs combined by treatment. Also shown is the percent reduction compared to the control. The PAMD+NA treatment resulted in the greatest amount of runoff for all runs and was significantly different than all other treatments. There were no significant differences between other treatments for the dry run because of variation among replicates. However, the mean values of total runoff 1862 TRANSACTIONS OF THE ASAE

5 Runoff Rate (mm/hr) a Control PAMW+NA PAMW+SL PAMD+NA Time (min) b Control PAMW+NA PAMW+SL PAMD+NA Runoff Rate (mm/hr) Time (min) c Control PAMW+NA PAMW+SL PAMD+NA Runoff Rate (mm/hr) Time (min) Figure 1. Runoff rate versus time for: (a) dry run, (b) wet run, and (c) very wet run. Lines shown are an average of three replicates. Vol. 45(6):

6 a b c Figure 2. Average sediment yield rate versus time for: (a) dry run, (b) wet run, and (c) very wet run. Lines shown are an average of three replicates. for the dry run suggest that both PAMW+NA and PAMW+SL were highly effective in reducing runoff. Mean runoff response for the PAMW+SL and PAMW+NA treatments was similar for the wet run, showing no significant difference. Both of these treatments were 1864 TRANSACTIONS OF THE ASAE

7 significantly different from the control (70% and 77% reduction in runoff for the PAMW+NA and PAMW+SL treatments, respectively). Over the entire experiment, both the PAMW+NA and PAMW+SL treatments resulted in significantly less runoff than the control or PAMD+NA. The effectiveness of both PAMW+SL and PAMW+NA diminished over time with respect to the control. However, the effectiveness of PAMW+NA decreased more than did PAMW+SL. The amounts of NA and SL applied were such that there was an equivalent amount of calcium. The SL was more effective because of its greater solubility, which was demonstrated in the dissolution analysis. Polyacrylamide treatments were also effective at reducing final runoff rates (table 4, fig. 1). The increase in runoff rate for the PAMD+NA treatment was dramatically greater than that of the other treatments (fig. 1a). This indicates that the infiltration rate was greatly reduced for the PAMD+NA treatment as compared to the control. It is suspected that the PAM granules migrated into pore spaces and became enlarged during wetting, which limited the infiltration rate, accounting for the significantly greater runoff. The PAMD+NA treatment also resulted in the greatest runoff rate (figs. 1b and 1c). The liquid PAM (PAMW) treatments resulted in consistently lower final runoff rates (table 4, fig. 1). The final runoff rate for PAMW+SL was significantly less than that of the control for all but the initial dry run, where large sample variation led to the insignificance of treatments. SEDIMENT YIELD General trends observed for runoff also held for sediment yield. Overall, the PAMW+NA and PAMW+SL treatments were the most effective, reducing sediment yield by 93% and 98%, respectively. Interestingly, the PAMD+NA treatment reduced total sediment yield by 31% (not statistically different from the control), even though average runoff was 64% greater than the control (table 3). For the runs on initially dry soil, PAMW+NA and PAMW+SL were extremely effective in reducing sediment yield (100% and 100% reduction, respectively). During the experiments with dry PAM, we noticed a considerable amount of PAM granules in the runoff water. A possible explanation is that individual PAM particles were adsorbed to small soil particles and were carried along with these soil particles in runoff, or that PAM granules did not have sufficient time to dissolve and were transported away by the runoff water. The effect of the PAMW treatments on final sediment yield rates was very pronounced. Final sediment yield rates for both PAMW treatments were significantly less than the control for all of the sub runs (fig. 2). Reductions in sediment yield rate for the most intense (100 mm hr 1 ) rainfall over the control were 83% for PAMW+NA and 94% for PAMW+SL (table 4). As shown in table 3, PAM was more effective at reducing sediment yield than runoff. If the PAM acted to promote infiltration only, one would expect the same reduction in sediment yield as the reduction in runoff. Because the reduction in sediment yield was greater than the reduction in runoff compared to the control, one can conclude that the erodibility of the soil was decreased. ERODIBILITY An alternative means of viewing the effectiveness of the treatments is to examine the relationship between sediment yield and runoff. Reductions in sediment yield due to the addition of amendments can be attributed to a decrease in soil erodibility or an increase in infiltration (i.e., reduction in shear by flowing water). The definition of erodibility can vary slightly depending on whether the rill or interrill process is being considered. Rill erodibility is a measure of the susceptibility of soil to become detached by concentrated flow and can be defined as the increase in soil detachment per unit increase in shear stress of clear water. Interrill erodibility is the rate at which sediment is delivered to rills as a function of rainfall intensity and runoff rate (Flanagan and Nearing, 1995). Sediment yield, not detachment, was measured in this study. However, sediment yield is proportional to soil detachment and shear stress under these experimental conditions. Thus, an estimate of soil erodibility can be made by regressing sediment yield rate on runoff rate. Figure 3 shows the best fit lines of sediment yield as a function of runoff for the wet run. Data from all three replicates were combined for each treatment. Both the control and PAMD+NA treatments were best represented by a power function, indicating that for these treatments, sediment yield did not increase linearly with respect to runoff. For the wet run, there was good agreement between sediment yield and runoff, as shown by the high R 2 values. This demonstrates that the relationship between sediment yield and runoff was consistent among replicates for each treatment. Good agreement was also found in the dry and very wet runs (data not shown). Huang and Bradford (1993) stated that under net detachment conditions, sediment yield rate (q s ) is a linear function of runoff rate (q w ). Under depositional conditions, the relationship can vary between linear and quadratic, depending on the importance of the ratio of deposition rate to erosion length scale. When deposition rate is great, q s should be a quadratic function of q w (Huang and Bradford, 1993). Therefore, from figure 3, both the control and PAMD+NA treatments were generally under net deposition conditions, indicating that the erosion process was transport limited. Both PAMW treatments exhibited a linear relationship, indicating a detachment limited condition. The slope of the best fit line was 138 kg ha 1 mm 1 for PAMW+NA and 60 kg ha 1 mm 1 for PAMW+SL. For low values of runoff, the lines for PAMW+NA and PAMW+SL were nearly identical; they diverged at an increasing rate for greater runoff amounts. The reduced erodibility exhibited by the PAMW+SL treatment supports the assertion that SL maintained greater electrolyte concentrations than NA, as suggested by the sequential dilution dissolution analysis (table 2). The greater electrolyte concentrations produced by SL may have allowed the PAM to bond better to the soil, may have acted independently to reduce clay dispersion, or, most likely, a combination of the two. For each unit increase in runoff, the resultant increase in sediment yield was greater for PAMD+NA than for either PAMW+NA or PAMW+SL (fig. 3). However, this change in sediment yield for PAMD+NA was less than that of the control. For the wet run, 62% more runoff but 29% less sediment yield was measured for the PAMD+NA treatment than for the control (table 3). This indicates that the dry PAM application hindered infiltration yet stabilized the soil. One possible explanation is that the PAM granules migrated into Vol. 45(6):

8 Sediment Yield (kg/ha) Control PAMD+NA PAMW+NA PAMW+SL Control PAMD+NA PAMW+SL PAMW+NA Control PAMD+NA PAMW+SL PAMW+NA y = x 1.44 R 2 = 0.95 y = x 1.45 R 2 = 0.94 y = x R 2 = 0.69 y = 59.70x 5.65 R 2 = Runoff (mm) Figure 3. Scatter plot of sediment yield versus runoff from the wet run. Each data point represents a 3 minute sampling period. pore spaces, either through raindrop impact or in the application process, and then enlarged during wetting, thus blocking water infiltration. Over the course of a run, soil may have been adsorbed to these saturated globules, or the saturated PAM may have acted as a mortar, to stabilize the soil to some degree. Another possibility is that the soil surface was sealed at the same time as the PAM became activated, impeding infiltration but, at the same time, strengthening the soil surface and thereby reducing detachment. Peterson et al. (2002) performed a laboratory study using the same soil as used in this study and applying 70 mm hr 1 rain for 2 hours. The small erosion pans (32 cm wide by 45 cm long) used in that study were used to study the effect of PAM and different gypsiferous materials on interrill erosion. Treated soils in that study reduced runoff by 12% and 35%, compared to the control, for a PAMW+NA treatment and a PAMW plus gypsum treatment, respectively. Reductions in sediment yield, compared to bare soil, were 27% and 74% for the same treatments, respectively. Reductions in runoff and sediment yield in the current study were much greater. This indicates that PAM treatments may be more effective in controlling rill erosion than interrill erosion. More research is needed to determine whether PAM treatments would be effective under the concentrated flow conditions in an ephemeral gully or other similar application. SUMMARY AND CONCLUSIONS Polyacrylamide applied in solution that was allowed to dry on the soil surface was most effective in reducing total runoff (62% to 76% reduction compared to control) and total sediment yield (93% to 98% reduction compared to control). Spraying of PAM in solution was significantly more effective in controlling runoff and erosion than was the dry granular application for the rainfall events simulated in this study. An analysis of erodibility showed that sprayed PAM treatments dramatically reduced erodibility compared to the control. Dry PAM application also reduced soil erodibility compared to the control, but not as dramatically as the sprayed PAM. Both liquid PAM application treatments resulted in significant reductions in sediment loss and runoff. Dry application of PAM had no benefit in reducing runoff and a marginal benefit in reducing sediment yield under these experimental conditions. Results indicated that SoilerLime (SL) at a lower application rate was as effective as Nutra Ash (NA) as an electrolyte source. This experiment was designed to represent a worst case scenario of a large rainfall on recently disturbed soil; other storm sequences could lead to different conclusions. For control of erosion from intense rainstorms, we recommend using application of liquid PAM solutions. More research is needed to determine if dry or wet PAM application is more effective for less intense storms under different field conditions. REFERENCES Agassi, M., and M. Ben Hur Stabilizing steep slopes with soil conditioners and plants. Soil Technology 5: Ben Hur, M Runoff, erosion, and polymer application in moving sprinkler irrigation. Soil Science 158(4): Bjorneberg, D. L Temperature, concentration, and pumping effects on PAM viscosity. Trans. ASAE 41(6): Chaudhari, K Polyacrylamide soil amendment effects on soil erosion from steep slopes. MS thesis. West Lafayette, Ind.: Purdue University. Chaudhari, K., and D. C. Flanagan Polyacrylamide effect on sediment yield, runoff, and seedling emergence on a steep slope. Presented at 1998 ASAE Annual International Meeting, Paper No St. Joseph, Mich.: ASAE. Daniel, T. C., P. E. McGuire, D. Stoffel, and B. Miller Sediment and nutrient yield from residential construction sites. J. Environmental Qual. 8(3): Flanagan, D. C., and M. A. Nearing USDA Water Erosion Prediction Project: Hillslope profile and watershed model 1866 TRANSACTIONS OF THE ASAE

9 documentation. NSERL Report No. 10. West Lafayette, Ind.: USDA ARS National Soil Erosion Research Laboratory. Flanagan, D. C., L. D. Norton, and I. Shainberg. 1997a. Effect of water chemistry and soil amendments on a silt loam soil Part 1: Infiltration and runoff. Trans. ASAE 40(6): b. Effect of water chemistry and soil amendments on a silt loam soil Part 2: Soil erosion. Trans. ASAE 40(6): Flanagan, D. C., K. Chaudhari, and L. D. Norton. 2002a. Polyacrylamide soil amendment effects on runoff and sediment yield on steep slopes: Part I. Simulated rainfall conditions. Trans. ASAE 45(5): b. Polyacrylamide soil amendment effects on runoff and sediment yield on steep slopes: Part II. Natural rainfall conditions. Trans. ASAE 45(5): Foster, G. R., C. B. Johnson, and W. C. Moldenhauer. 1982a. Critical slope lengths for unanchored cornstalk and wheat straw residue. Trans. ASAE 25(4): , 947. Foster, G. R., W. H. Neibling, and R. A. Natterman. 1982b. A programmable rainfall simulator. ASAE Paper No St. Joseph, Mich.: ASAE. Hershfield, D. M Rainfall frequency atlas of the United States for durations from 30 minutes to 24 hours and return periods from 1 to 100 years. Technical Paper No. 40. Washington, D.C.: National Weather Bureau. Huang, C., and J. M. Bradford Analyses of slope and runoff factors based on the WEPP erosion model. Soil Sci. Soc. Am. J. 57(5): Keren, R., and I. Shainberg Effect of dissolution rate on the efficiency of industrial and mined gypsum in improving infiltration of a sodic soil. Soil Sci. Soc. Am. J. 45(1): Kramer, L. A., and L. D. Meyer Small amounts of surface mulch reduce soil erosion and runoff velocity. Trans. ASAE 12(5): , 645. Krenitsky, E. C., M. J. Carroll, R. L. Hill, and J. M. Krouse Runoff and sediment losses from natural and man made erosion control materials. Crop Sci. 38(4): Lattanzi, A. R., L. D. Meyer, and M. F. Baumgardner Influences of mulch rate and slope steepness on interrill erosion. Soil Sci. Soc. Am. Proc. 38(6): Levy, G. J., J. Levin, M. Gal, M. Ben Hur, and I. Shainberg Polymers effects on infiltration and soil erosion during consecutive simulated sprinkler irrigations. Soil Sci. Soc. Am. J. 56(3): Mannering, J. V., and L. D. Meyer Effects of various rates of surface mulch on infiltration and erosion. Soil Sci. Soc. Am. Proc. 27(1): Meyer, L. D., W. H. Wischmeier, and G. R. Foster Mulch rates required for erosion control on steep slopes. Soil Sci. Soc. Am. Proc. 34(6): Meyer, L. D., C. B. Johnson, and G. R. Foster Stone and woodchip mulches for erosion control on construction sites. J. Soil Water Cons. 27(6): Mitchell, J. K., C. Ray, G. F. McIsaac, and J. G. O Brien Land treatment effects on soil erosion. In Proc. Managing Irrigation Induced Erosion and Infiltration with Polyacrylamide Conf., R. E. Sojka and R. D. Lentz, eds. Miscellaneous Pub. No Moscow, Idaho: University of Idaho. Norton, L. D., I. Shainberg, and K. W. King Utilization of gypsiferous amendments to reduce surface sealing in some humid soils of the eastern USA. In Soil Sealing and Crusting, J. W. A. Poesen and M. A. Nearing, eds. Catena Supplement 24. Cremlingen Destedt, Germany: Catena Verlag. Peterson, J. R., D. C. Flanagan, and J. K. Tishmack Polyacrylamide and gypsiferous material effects on runoff and erosion under simulated rainfall. Trans ASAE 45(4): Seybold, C. A Polyacrylamide review: Soil conditioning and environmental fate. Commun. Soil Sci. Plant Anal. 25(11&12): Shainberg, I., and G. J. Levy Organic polymers and soil sealing in cultivated soils. Soil Science 158(4): Shainberg, I., D. N. Warrington, and P. Rengasamy Water quality and PAM interactions reducing surface sealing. Soil Science 149(5): Smith, H. J. C, G. J. Levy, and I. Shainberg Water droplet energy and soil amendments: Effect on infiltration and erosion. Soil Sci. Soc. Am. J. 54(4): Stern, R., M. C. Laker, and A. J. van der Merwe Field studies on effect of soil conditioners and mulch on runoff from Kaolinitic and Illitic soils. Aust. J. Soil Res. 29(2): Thompson, A. M., B. N. Wilson, and H. V. Nguyen Effectiveness of erosion control blankets on reducing shear stress acting on soil particles. In Soil Erosion Research for the 21st Century: Proc. Int. Symp., J. C. Ascough II and D. C. Flanagan, eds. Honolulu, Hawaii. 3 5 January St. Joseph, Mich.: ASAE. Vol. 45(6):

10 1868 TRANSACTIONS OF THE ASAE