Predicting soil erosion in conservation tillage cotton production
systems using the revised universal soil loss equation (RUSLE)
E.Z. Nyakatawa
a, K.C. Reddy
a,*, J.L. Lemunyon
baDepartment of Plant and Soil Science, Alabama A&M University, PO Box 1208, Normal, AL 35762, USA bUSDA/NRCS, 510 W Felix St., Building 23, PO Box 6567, Fort Worth, TX 76115, USA
Received 6 March 2000; received in revised form 11 September 2000; accepted 27 September 2000
Abstract
Despite being one of the most pro®table crops for the southeastern USA, cotton (Gossypium hirsutumL.) is considered to create a greater soil erosion hazard than other annual crops such as corn (Zea maysL.) and soybeans (Glycine max(L.) Merr.). Reduced tillage systems and cover cropping can reduce soil erosion and leaching of nutrients into ground water. The objectives of this study, which was conducted in north Alabama from 1996 to 1998, were to assess the impact of no-till and mulch-till systems with a winter rye (Secale cerealeL.) cover crop and poultry litter on soil erosion estimates in cotton plots using the revised universal soil loss equation (RUSLE). Soil erosion estimates in conventional till plots with or without a winter rye cover crop and ammonium nitrate fertilizer were double the 11 t haÿ1yrÿ1tolerance level for the Decatur series
soils. However, using poultry litter as the N source (100 kg N haÿ1) gave soil erosion estimates about 50% below the tolerance level under conventional till. Doubling the N rate through poultry litter to 200 kg N haÿ1under no-till system gave the lowest soil erosion estimate level. No-till and mulch-till gave erosion estimates which were about 50% of the tolerance level with or without cover cropping or N fertilization. This study shows that no-till and mulch-till systems with cover cropping and poultry litter can reduce soil erosion in addition to increasing cotton growth and lint yields, and thus improve sustainability of cotton soils in the southeastern USA.#2001 Published by Elsevier Science B.V.
Keywords:Cotton; Cover crop; Mulch-till; No-till; Poultry litter; RUSLE; Soil erosion
1. Introduction
Soil erosion is a major environmental problem in the US and worldwide. Eroded soils carry nutrients, pesticides, and other harmful farm chemicals into rivers, streams, and ground water resources (Gallaher and Hawf, 1997). Soil erosion has been attributed to the loss of productivity of soils in the southern
Piedmont ranging from southern Virginia through central Alabama which were once some of the most productive cotton producing areas in the US. Accord-ing to Brown et al. (1985), cotton yields can decline by as much as 4% for each centimeter of top soil loss. Erosion has been suggested as being one of the major causes of static or declining cotton yields in some areas in the southeast USA. Unlike gully erosion which is easily recognized, sheet erosion often goes unnoticed by the producer while having a signi®cant impact on crop yields. In Alabama, soil erosion on crop lands averages about 25 t haÿ1yrÿ1, which can potentially decrease cotton yields by *Corresponding author. Tel.:1-256-858-4191;
fax:1-256-851-5429.
E-mail address: [email protected] (K.C. Reddy).
440±670 kg haÿ1, if no remedial actions are taken (Anon., 1991).
Any tillage and planting system that leaves at least 30% of the soil surface covered with crop residues can be called conservation tillage (CTIC, 1994; Gallaher and Hawf, 1997). Conservation tillage, crop rotations, and use of cover crops are acceptable mitigation techniques to soil erosion on highly erodible sites. The bene®ts of cover crops include improved soil structure and tilth, addition of organic matter, reduced soil erosion, and reduced leaching of nitrates into water resources (Kelley et al., 1992; Schertz and Kemper, 1994; Abdin et al., 1998; Sainju et al., 1998; Vyn et al., 1999). The 1985 Food Security Act and the 1990 US Federal Farm Bill restrict the production of cotton under conventional tillage on highly erodible sites for farmers participating in fed-eral commodity programs such as storage loans and subsidized prices on grain and cotton.
The revised universal soil loss equation (RUSLE) is an empirical soil erosion model revised from the universal soil erosion loss equation (USLE) (Wisch-meier and Smith, 1965, 1978; Renard et al., 1993). It computes the average annual soil erosion estimates caused by rainfall and its associated overland ¯ow. Published literature on soil erosion estimates in reduced tillage cotton production systems, particularly in combination with poultry litter and winter cover cropping is lacking. This paper describes soil loss by erosion estimated by RUSLE under conventional till, no-till, and mulch-till systems with winter rye cover cropping, and N from poultry litter from cotton plots on a Decatur silt loam soil in north Alabama. The experimental data will be useful for extension workers and farmers who want to come up with acceptable production practices in erosion prone areas.
2. Materials and methods
2.1. Site description and treatments
The study was conducted from 1996 to 1998 at the Alabama Agricultural Experiment Station, Belle Mina, AL (348410N, 868520W). The soil at the study site is a Decatur silt loam soil (clayey, kaolinitic thermic, Typic Paleudults). The study site has a slope of about 1.5%. The treatments consisted of three
tillage systems: conventional tillage, mulch-till and no-till; two cropping systems: cotton±winter fallow, i.e. cotton in summer and fallow in winter, and cotton± winter rye cropping, i.e. cotton in summer and rye (Secale cerealeL.) in winter; three nitrogen levels: 0, 100 and 200 kg N haÿ1; two nitrogen sources: ammo-nium nitrate and fresh poultry litter. Ammoammo-nium nitrate was used at one N rate (100 kg N haÿ1) only. An additional weed-free (bare) fallow treatment was included. The bare fallow plots were not tilled and cropped. They were kept weed-free by the use of Roundup (glyphosate) herbicide. The purpose of these control plots was to get an estimate of soil loss from plots without any vegetation canopy and surface resi-due protection. The experimental design was a ran-domized complete block design with four replications. Plot size was 8 m wide and 9 m long which resulted in eight rows of cotton, 1 m apart.
Conventional tillage included moldboard plowing to a depth of about 15 cm in November and disking in April. A ®eld cultivator was used to prepare a smooth seedbed after disking. Mulch-till included incorporat-ing surface crop residues to a depth of about 5±7 cm with a ®eld cultivator before planting, without per-forming any primary tillage. A ®eld cultivator was used for controlling weeds in the conventional tillage system, while spot applications of Roundup herbicide were used to control weeds in the no-till and mulch-till systems.
litter were applied to the plots 1 day before cotton planting. The experimental plots received a blanket application of 336 kg haÿ1of a 0±20±20 fertilizer to nullify the effects of P and K applied through poultry litter.
2.2. Winter rye cover crop establishment and cotton planting
A no-till planter was used to seed the winter rye cover crop, var. Oklon on 4 December 1996 in the ®rst year, which was killed with Roundup herbicide on 8 April 1997. In the second year, the winter rye cover crop was seeded on 24 November 1997 and was killed the same way on 28 February 1998. A herbicide mixture of Prowl (pendimethalin) at 2.3 l haÿ1, Cotoran (¯uometuron) at 3.5 l haÿ1, and Gramoxone extra (paraquat) at 1.7 l haÿ1was applied to all plots before planting cotton, var. Deltapine NuCotn 33B, on 8 May 1997 and 5 May 1998. All plots received 5.6 kg haÿ1 of Temik (aldicarb) prior to planting, for the control of thrips. No major weed problems were observed in the cotton plots. The major insect pests observed in cotton were aphids (Aphisspp.) and bollworms(Heliothisspp.). Aphids were controlled by spraying Bidrin (dicrotophos) at 0.4 kg haÿ1, whereas bollworms were controlled with Karate (cyperme-thrin) at 44 ml haÿ1.
2.3. Description of the RUSLE computer model
RUSLE incorporates four physical parameters asso-ciated with erosion by water: rainfall erosivity, soil erodibility, topography, and land-use management. The model calculates the average annual soil erosion on ®eld plots from the equationARKLSCP, where
Ais the predicted long-term average of annual sheet and rill soil loss from a de®ned slope (t haÿ1yrÿ1),R
the rainfall-runoff erosivity factor [(hundreds of ft-tons) in. acreÿ1hÿ1yrÿ1],Kthe soil erodibility factor as measured under unit plot conditions [tons acreÿ1 (hundreds of ft-tons acreÿ1in. hÿ1)ÿ1],LSthe erosion impact of the slope length (L) and steepness (S) on erosion in comparison to unit plot conditions (dimen-sionless),Cthe erosion impact of cover and manage-ment schemes on erosion in comparison to unit plot conditions (dimensionless), andPthe erosion impact of conservation support practices (e.g. contour tillage,
strip cropping, terraces, and drainage) on erosion in comparison to unit plot conditions (dimensionless).
Among the RUSLE factors,R,K, andLSare nature dependent or intrinsic to the landscape. They do not change when soil conservation practices are applied. Also, installing support practices that reduce the
P-value requires considerable monetary costs. There-fore, only C is the factor that can be managed with reasonable planning and minimal costs to in¯uence the soil losses in a given ®eld. The RUSLE model consists of three main databases: climate, crop, and soil databases. In this study, the calculations of soil erosion estimates were done by plot for each year, using RUSLE Ver. 1.06 computer program (RUSLE Users' Guide, 1993).
2.3.1. Climate database
The climate database contains information on monthly temperature and precipitation. The rainfall erosivity (R) factor is the product of the storm energy (E) and its maximum 30 min intensity (I30) for storms of at least 0.5 in. (15 mm). This product, called an
EI30-value, is directly proportional to the product of the storm amount andI30for storms of intensities of at least 0.5 in. hÿ1(Renard et al., 1993). TheR-value for a given site is the average ofEI30-values summed for each year for a 22-yr period. The values forEI30 for 10-yr periods have been computed and entered into the city database utility ®le for each location and have been used to draw maps which show R-values for different locations.
2.3.2. Crop database
rainfall in®ltration rate. Each of these parameters is assigned a subfactor value, and these are multiplied together to yield a soil-loss ratio (SLR), thus SLRPLU CC SC SR SM. Each of these SLR values is then weighted by the fraction of rainfall and runoff erosivity (EI) associated with the corresponding time period of the cover and management, and these weighted values are combined into an overallC-factor value (Renard et al., 1993). The ®eld operations database utility ®le contains information on how ®eld operations such as tillage, planting, fertilizer applica-tion, and harvesting operations affect soil, vegetaapplica-tion, and residue.
2.3.3. Soil database
The soil database contains soil survey and soil characterization data. It holds information on the soil erodibility (K) factor, the slope length (L) and steep-ness (S), which comprise theLS-factors. TheK-factor represents the susceptibility of a soil to erode and the amount and rate of runoff, as measured under the standard unit plot condition. The unit plot condition is a continuously cultivated fallow plot, 72.6 ft (22.1 m) long with a slope of 9%. The soil erodibility nomo-graph (Renard et al., 1993) calculatesK-values based on soil texture, structure, permeability, and percent organic matter. The slope length (L) factor computes the effect of slope length on erosion, while the slope steepness (S) factor computes the effect of slope steepness on erosion. The values for L and S under the unit plot conditions equal 1, hence values forLand
S for any given location are relative to the unit plot conditions.
2.4. Cover management (C) factor data collection
Most of the information required for predicting soil erosion using RUSLE, such as rainfall and soil data can be obtained from published records. The excep-tion is the C-factor which varies with season and production system. Conservation tillage such as no-till and mulch-no-till affect theC-factor by reducing soil degradation, reducing runoff, and increasing in®ltra-tion and soil organic matter, which in turn reduce soil erosion. Crop data collected for the RUSLEC-factor calculation in this study included winter rye biomass, surface residue cover (SRC) after cotton planting, cotton CC, effective fall height (FH) from the cotton
canopy, and cotton surface root mass. SRC was mea-sured in all plots using the camline transect method (Renard et al., 1993; Reddy et al., 1994) immediately after cotton planting. CC was determined by measur-ing the width of the crop canopy of each row from the four central rows on each plot using a ruler and expressing the ®gure as a percentage of the row width. Effective FH is the distance a raindrop falls after striking the crop canopy. This was calculated from the equation:
FH1
2 THÿBH BH (1)
where TH and BH are the top and bottom heights of the cotton canopy, respectively.
Root biomass was determined by sampling plants with their roots intact from 0.5 m2 quadrants from each plot. Roots were extracted out of the soil by removing soil from both sides of the row and lifting the intact plants from the base with a garden fork. The roots were cut from the shoots, washed in water to remove soil and placed in separate bags. The shoot and root samples were oven dried to constant weight at 658C for 72 h before weighing. The cotton crop growth data were taken in each plot at 15-day intervals.
2.5. Data analysis
Statistical analysis of the data were performed using the GLM procedures of the Statistical Analysis Sys-tem. Contrast analysis procedures were used to com-pare the main effect treatment means for tillage systems, cropping systems, and N treatments. The analyses were performed separately for each year. There were signi®cant tillagecropping system and tillagenitrogen source interactions forC-factor and soil erosion estimates.
3. Results and discussion
3.1. Surface residue cover
cropping systems which left at least 30% of the soil surface covered with residues qualify to be described as conservation tillage. As SRC approaches 100%, soil erosion approaches zero (Moldenhauer and Langdale, 1995). With 50% residue cover, soil erosion reduction is about 83% and with 10% residue cover, soil erosion is about 30%. Based on the above ®gures, soil erosion in plots with no-till or mulch-till under cotton±rye sequential cropping system is expected to be very low.
3.2. Cotton growth parameters used as RUSLE input data
Cotton FHs, CC, and root biomass as in¯uenced by tillage systems and N levels at 2-week intervals during 1997 and 1998 growing seasons are presented in Figs. 2 and 3. Drier weather and higher temperature in the summer of 1998 (Fig. 1) resulted in a faster rate of cotton growth development which resulted in 2-week earlier maturity in 1998 compared to 1997. Plants under no-till and mulch-till systems averaged 10 cm taller than those under conventional till system at each sampling period in both years. As a result, FHs for cotton plants under no-till and mulch-till were generally higher than those under conventional till (Fig. 2). Similar trends were observed with CC, which was highest between 12 and 14 weeks after planting. A fast growing canopy is a desirable attribute for cotton. Since cotton is planted in wide rows, most of the inter-row spacing is exposed to the direct impact of rain drops which can result in high soil erosion rates. Therefore, treatments which enable cotton to rapidly
develop a good CC will most likely result in reduced soil erosion. Plant height was signi®cantly correlated to CC, r0:88 and 0.83 in 1997 and 1998, respec-tively. Maximum CC for cotton plants under no-till and mulch-till systems was 7 and 11% higher (P<0:05) than that for plants under conventional till in 1997 and 1998, respectively (Fig. 2).
Cotton root mass in the top 10 cm of the soil at 10± 14 weeks after planting for plants under no-till was 20 and 38% higher than that for plants under mulch-till and conventional till systems, respectively, in 1997 (Fig. 2). However, in 1998, cotton root mass in the top 10 cm of the soil at 10±14 weeks after planting for plants under no-till was 15% lower than that for plants under conventional till or mulch-till systems (Fig. 2). This can be largely attributed to the drought in the months of August and September 1998. Roots and crop residues control soil erosion directly by physi-cally holding soil particles together, providing a car-bon source for microbial soil processes that promote water in®ltration, and providing a mechanical barrier to water and eroded sediment movement. In addition, roots and crop residues exude binding agents and serve as a food source of micro-organisms which increase soil aggregation and thereby reducing its susceptibility to erosion.
In both years, cotton growth parameters for plants which received N in the form of ammonium nitrate or poultry litter were consistently higher than those for plants which did not receive N (Fig. 3). Growth parameters for cotton plants which received 100 kg N haÿ1in the form of ammonium nitrate were
Table 1
Surface residue cover in cotton plots immediately after seeding as affected by tillage systems and cotton±winter fallow (CF) and cotton±winter rye (CR) sequential cropping systems, Belle Mina, AL, 1997 and 1998
Tillage systems (%)a
Conventional till Mulch-till No-till
1997 1998 1997 1998 1997 1998
Cropping systems
CF 1.3 a 1.2 a 6.1 a 5.7 a 16.8 a 12.6 a
CR 20.4 b 19.8 b 64.8 b 51.3 b 99.8 b 100.0 b
S.E.b 9.5 9.3 25.3 21.9 41.5 43.7
CVc(%) 4.8 4.3 4.8 4.3 4.8 4.3
aMeans for tillage systems within a column followed by different letters are signi®cantly different at the 1% level. bStandard error of the difference between means.
higher than those for plants which received 100 kg N haÿ1in the form of poultry litter. This can be attributed to the slower availability of N from poultry litter. However, doubling the N rate to 200 kg N haÿ1 in the form of poultry litter resulted in signi®cantly higher cotton growth parameters than all the other N levels (Fig. 3). Improved SM conserva-tion in no-till system was largely responsible for better plant growth and higher yield parameters in cotton (Nyakatawa et al., 1998; Nyakatawa and Reddy, 2000).
3.3. Cotton lint yields
In addition to reducing soil erosion, conservation tillage practices such as no-till and mulch-till may give higher cotton lint yields compared to conventional
tillage. Cotton lint yield under no-till (1360 kg haÿ1) was 24 and 18% greater than that under conventional till (1100 kg haÿ1) and mulch-till (1150 kg haÿ1) systems, respectively, in 1997 (Table 2). In 1998, cotton lint yield under no-till system (1440 kg haÿ1) was 7% greater than that under conventional till (1350 kg haÿ1). There were no signi®cant yield increases in mulch-till compared to conventional till. Poultry litter at 100 kg N haÿ1gave similar cotton lint yield to ammonium nitrate, whereas at 200 kg N haÿ1, lint yields were 23 and 38% greater than those at 100 kg N haÿ1in the form of ammonium nitrate and poultry litter, respectively, in 1997; and 12 and 25% greater than those at 100 kg N haÿ1 in the form of ammonium nitrate and poultry litter, respectively, in 1998 (Table 2).
SM measurements in the top 7 cm of the soil showed signi®cantly higher volumetric SM content in no-till plots compared to conventional till plots with and without winter rye cover cropping and in plots which received N in the form of poultry litter vs. ammonium nitrate (Nyakatawa et al., 1998). The
improved SM conservation which resulted in early seedling emergence, higher seedling vigor, and better plant growth in no-till system, cotton±winter rye cropping, and 200 kg N haÿ1in form of poultry litter were attributed to the greater lint yield of cotton.
3.4. Cover management (C-factor) and soil erosion estimates
HigherC-factor values indicate higher soil erosion loss since theC-factor is a ratio of soil loss in a cover management sequence to soil loss from the unit plot.
C-factors for cotton±winter rye cover cropping under
conventional till were signi®cantly reduced by 15% (0.487 vs. 0.423) and 28% (0.525 vs. 0.410) compared to cotton±winter fallow cropping in 1997 and 1998, respectively (Table 3). C-factors under conventional till were about four times greater than those under no-till, under cotton±winter fallow cropping in 1997 and 1998, respectively (Table 3). However, under
cotton±winter rye cropping,C-factors under conven-tional till were up to seven times greater than those under no-till both years. The main factor which caused differences in theC-factor is SRC since most of the other parameters such as rainfall and soil factors are
important role of slowing surface runoff and protect-ing soil from the direct impact of raindrops whose energy breaks the soil particles apart which can then be carried away by moving water. The reduced
C-factors explain the 15% lower (15.7 vs. 18.0 t haÿ1yrÿ1) and the 25% lower (15.8 vs. 19.7 t haÿ1yrÿ1) soil erosion estimates in conven-tional till in cotton±winter rye cover cropping com-pared to cotton±winter fallow cropping in 1997 and 1998, respectively (Fig. 4).
reduced soil erosion (McDowell and McGregor, 1984; Stein et al., 1986; Blevins et al., 1990; Seta et al., 1993). Without cover cropping, soil erosion estimates in no-till plots were about one-third of that in conven-tional till in both years. Similar results were found by Stevens et al. (1992) who reported that no-till cotton cropping without cover cropping may reduce soil erosion by 70% compared to conventional till system. Surface application of poultry litter at 100 kg N haÿ1 gave signi®cantly lowerC-factors compared to control
plots and 100 kg N haÿ1 in the form of ammonium nitrate under conventional till in both years (Table 3). Similar results were obtained under no-till. The lower
C-factor values with use of poultry litter indicate that soil erosion rates were lower than those in plots which received an equal amount of N but in the form of ammonium nitrate or those which did not receive any N. The better erosion control with poultry litter can be attributed to the fact that when surface applied, the litter material acts as a residue cover which protects
soil from the direct impact of raindrops, slows runoff water movement, and also increases in®ltration. The RUSLE model takes into account of this fact in the calculation of theC-factor.
At each level of N, soil erosion estimates in no-till plots was below 5 t haÿ1yrÿ1 compared to over 15 t haÿ1yrÿ1 under conventional till at 0 or 100 kg N haÿ1ammonium nitrate levels in both years (Fig. 4). Plots which received 100 kg N haÿ1in the form of poultry litter had 10, 3 and 3 t haÿ1yrÿ1less soil erosion rates under conventional till, no-till and mulch-till systems, respectively, compared to plots which received the same amount of N in the form of ammonium nitrate in 1997. Similar ®gures for 1998 were 9, 2, and 3 t haÿ1yrÿ1. Doubling the N rate through poultry litter to 200 kg N haÿ1under no-till system gave the lowest soil erosion estimate levels of less than 2 t haÿ1yrÿ1in both years (Fig. 4). Applica-tion of N in the form of ammonium nitrate or poultry litter reduced soil erosion rates due to higher cotton CC, root biomass, and more cotton residues compared to control plots (Fig. 3).
The tolerance levels of soil erosion loss for these Decatur series soils is 11 t haÿ1yrÿ1 (Reddy et al., 1994). Results from this study clearly show that growing cotton under conventional till with or without a winter rye cover crop using ammonium nitrate fertilizer at 100 kg N haÿ1 on the Decatur silt loam soils results in soil erosion rates up to 10 t haÿ1yrÿ1in excess of the tolerance level (Fig. 4). However, using poultry litter as the N source under conventional till gave soil erosion estimates 4±5 t haÿ1yrÿ1below the tolerance level (Fig. 4). Soil erosion estimates under no-till and mulch-till systems were about 50% of the tolerance level. The reduction in soil erosion rates will reduce the amount of nutrients and pesticide residues which end up in surface and ground water resources.
4. Conclusions
Soil erosion estimates under conventional till and continuous cotton cropping system with the use of ammonium nitrate fertilizer were similar to that in weed-free fallow plots, which are about twice the tolerance levels for northern Alabama. Adoption of no-till or mulch-till systems with winter rye cover cropping and poultry litter in the current cotton
pro-duction systems can signi®cantly reduce soil erosion to tolerable levels and thus improve sustainability of cotton soils in the southeastern USA where erosion and safe disposal of poultry litter that is being pro-duced in increasing quantities from the bourgeoning poultry industry are major problems.
References
Abdin, O., Coulman, B.E., Cloutier, D., Faris, M.A., Zhou, X., Smith, D.L., 1998. Yield and yield components of corn interseeded with cover crops. Agron. J. 90, 63±68.
Anon., 1991. Agriculture and Natural Resources Timely Informa-tion. Alabama Cooperative Extension Service, Auburn Uni-versity, Auburn, AL, pp. 1±14.
Blevins, R.L., Frye, W.W., Baldwin, P.L., Robertson, S.D., 1990. Tillage effects on sediment and soluble nutrient losses from a Maury silt loam. J. Environ. Qual. 19, 683±686.
Bremner, J.M., Mulvaney, C.S., 1982. Nitrogen-total. In: Page, A.L., Miller, R.H., Keeney, D.R. (Eds.), Methods of Soil Analysis. Part 2. Chemical and Microbiological Properties, 2nd Edition. American Society of Agronomy, Madison, WI, pp. 595±625.
Brown, S.M., Whitwell, T., Touchton, J.T., Burmester, C.H., 1985. Conservation tillage for cotton production. Soil Sci. Soc. Am. J. 49, 1256±1260.
Castellanos, J.Z., Pratt, P.F., 1981. Mineralization of manure nitrogen Ð correlation with laboratory indexes. Soil Sci. Soc. Am. J. 45, 354±357.
CTIC, 1994. National Crop Residue Management Survey Ð Executive Summary. Conservation Technology Information Center, West Lafayette, IN.
Gallaher, R.N., Hawf, L., 1997. Role of conservation tillage in production of a wholesome food supply. In: Gallaher, R.N. (Ed.), Proceedings of the 20th Annual Southern Conservation Tillage for Sustainable Agriculture, University of Florida, Gainesville, FL, USA, June 24±26, pp. 23±27.
Keeling, K.A., Hero, D., Rylant, K.E., 1995. Effectiveness of composted manure for supplying nutrients. In: Proceedings of the Conference on Fertilizer, Aglime and Pest Management. Department of Agronomy, University of Wisconsin, Madison, WI, January 17±18, pp. 77±81.
Kelley, K.R., Mortvedt, J.J., Soileau, J.M., Simmons, K.E., 1992. Effect of winter cover crops on nitrate leaching. Agronomy Abstracts. American Society of Agronomy, Madison, WI, p. 282.
McDowell, L.L., McGregor, K.C., 1984. Plant nutrient losses in runoff from conservation tillage corn. Soil Till. Res. 4, 79±91. Moldenhauer, W.C., Langdale, G.W., 1995. Crop residue manage-ment to reduce soil erosion and improve soil quality. USDA/ ARS Report No. 39, January 1995, pp. 1±47.
Nyakatawa, E.Z., Reddy, K.C., 2000. Tillage, cover cropping, and poultry litter effects on cotton: 1. Germination and seedling growth. Agron. J. 92, 992±999.
Reddy, K.C., Weesies, G.A., Lemunyon, J.L., 1994. Predicting soil erosion in different cotton production systems with the revised universal soil loss equation (RUSLE). In: Bushan, L.S., Abrol, I.P., Rao, M.S. (Eds.), Proceedings of the Eighth International Soil Conservation Organization Conference on Soil and Water Conservation: Challenges and Opportunities, Vol. 1, Indian Association of soil and water conservationsists, Dehradun, India, December 4±8, pp. 1±11.
Renard, K.G., Foster, G.R., Weesies, G.A., Yoder, D.C., 1993. Predicting soil erosion by water. A guide to conservation planning with the universal soil loss equation (RUSLE). Agricultural Handbook No. 703. US Dept. of Agriculture, p. 47.
RUSLE Users' Guide, 1993. Revised Universal Soil Loss Equation. Soil and Water Conservation Society, Ankeny, IA.
Sainju, U.M., Singh, B.P., Whitehead, W.F., 1998. Cover crop root distribution and its effects on soil nitrogen cycling. Agron. J. 90, 511±518.
Schertz, D.L., Kemper, W.D., 1994. Report on Field Review of No-till Cotton. USDA/ARS/NRCS/Auburn University/
Alabama A&M University, Huntsville, AL, September 22±23.
Seta, A.K., Blevins, R.L., Frye, W.W., Bal®eld, B.J., 1993. Reducing soil erosion and agricultural chemical losses with conservation tillage. J. Environ. Qual. 22, 661±665.
Stein, O.R., Neibling, W.H., Logan, T.J., Moldenhauer, W.C., 1986. Runoff and soil loss as in¯uenced by tillage residue cover. Soil Sci. Am. J. 50, 1527±1531.
Stevens, W.E., Johnson, J.R., Varco, J.J., Parkman, J., 1992. Tillage and winter cover management effects on fruiting and yield of cotton. J. Prod. Agric. 5, 570±575.
Vyn, T.J., Janovick, K.J., Miller, M.H., Beauchamp, E.G., 1999. Soil nitrate accumulation and corn response to preceding small grain fertilization and cover crops. Agron. J. 91, 17±24. Wischmeier, W.A., Smith, D.A., 1965. Predicting rainfall-erosion
losses from cropland east of the Rocky Mountains: guide for selection practices for soil and water conservation. Agricultural Handbook No. 282, US Dept. of Agric. Sci and Edu. Admin., Agric. Res., 47 pp.
