Evaluation of Cover Crops and Conservation Tillage for Conventional and Organic Sweetpotato (Ipomoea batatas) Production in North Carolina

Final Report for GS00-006

Project Type: Graduate Student
Funds awarded in 2000: $9,927.00
Projected End Date: 12/31/2004
Region: Southern
State: North Carolina
Graduate Student:
Major Professor:
Dr. Nancy Creamer
North Carolina State University
Expand All

Project Information


Sweetpotato production is an important part of North Carolina’s economy, but less than three percent of sweetpotato acreage is certified organic. Despite the demand for organic produce, few production recommendations exist for organic fresh market vegetables from land grant universities. This study was conducted to explore management options for conventional growers interested in producing organically managed sweetpotatoes. Local conventional growers were consulted prior to the initiation of this project to identify specific research needs and concerns related to the adoption of organic systems. Specific management protocols in organic systems and the conventionally managed system are representative of organic sweetpotato growers in our area and best management practices, respectively. A three-year field study on the effect of management of a cover crop mixture of hairy vetch and ‘Wrens Abruzzi’ rye in organically managed sweetpotato was compared to a conventional control using a systems approach. The organic systems included 1) no cover crop, 2) cover crop incorporated prior to transplanting, and 3) reduced tillage. All three years, experiments were conducted at the Center for Environmental Farming Systems (CEFS) in Goldsboro, NC. We examined weed suppression, nutrient management, wireworm larvae insect density, insect feeding damage to sweetpotato roots and costs and returns of each system


North Carolina has produced the majority of the nation’s sweetpotatoes since 1971 (USDA ERS, 2003). In 2004, North Carolina supplied 46% of the nations’ market and generated 79 million dollars for the state (NCDA & CS, 2005). Although there were over 16,000 ha in production that year (NCDA &CS, 2005), only 405 ha were managed according to federal organic standards (K. Hardison, personal communication).
Growers implementing best management practices for sweetpotato in North Carolina typically employ multiple tillage events including subsoiling, discing, field leveling, hilling, and cultivation. Due to heavy pest infestations typical of the warm and humid seasonal climate of southeastern NC, pest control is largely preventative rather than curative and includes soil fumigation prior to planting and multiple applications of preplant and postemergent herbicides and insecticides. According to a 1996 survey of North Carolina sweetpotato growers, 66% of respondents reported they used insecticides, 70% use herbicides, and 97% used cultivation to manage weeds (Toth et al., 1997). Soil-dwelling insect larvae of wireworm (Melanotus communis Gyllenhal and Conoderus spp.) pose the greatest risk to growers. Despite repeated applications, conventional insecticides post emergence dicot weed control are only marginally effective. Additionally, some of the insecticidal materials are currently being reviewed under the Food Quality Protection Act because of their potential for environmental harm.
In conventionally managed systems, nutrients are supplied with high analysis fertilizers. Although the actual amount supplied depends on preseason soil test results, producers typically apply 62 kg NO3-N, 112-168 kg K, 112 kg P, and 0.56 kg B ha-1 for control of blister. Common rotational crops include tobacco, cotton, soybean, and mixed vegetables. Although the soil in our area has low organic matter, few producers incorporate organic matter via cover crops or compost prior to planting. The extensive tillage and high number of vehicle passes required to manage sweetpotato as well as its associated rotational crops can lead to a loss of organic matter, decline in soil structure, and erosion.

Organic soil amendments such as compost and cover crops are an integral component of organic management systems. Hairy vetch and rye are species that establish easily, overwinter successfully, produce sufficient biomass, and are easily killed in the spring by mechanical methods (Creamer et al., 1997). Integration of this cover crop mixture in cropping systems has been shown to reduce nitrate leaching and carbon losses (Drinkwater et al., 1998; Rannells and Wagger, 1997), improve nutrient use efficiency (Staver and Brinsfield, 1998), increase the nitrogen fixation rate of legume species (Rannells and Wagger, 1997), increase water retention (Teasdale and Mohler, 1993) and decrease soil bulk density (Jackson et al., 2004). Cover crop residues that remain on the surface can inhibit weed seed germination due to a reduction of light penetration to the soil surface (Teasdale and Mohler, 1993) as well as physical interference (Creamer et al., 1996; Hutchinson and McGiffen, 2000). Chemical interference of weed germination has been demonstrated for a number of cover crop species including rye (Creamer et al., 1996; Reburg-Horton et al., 2005), members of the Cruciferae family (Blum et al., 1997) and crimson clover (Creamer et al., 1996).

There are few published reports on the effects of cover crops on sweetpotato. Growers implementing conventional and organic sweetpotato systems will benefit from research on the effects of cover crops on weeds, insects and nutrient management.

Project Objectives:
  • Identify the organically managed system with the best weed and insect suppression.

    To confirm yield was not limited by nutrient availability among systems.

    To investigate the impact of organically managed systems on crop productivity including foliar biomass, sweetpotato quality and yield.

    Evaluate the economics of conventional verses organically managed sweetpotato production in terms of cost effectiveness and product return.

    To participate in outreach and education events for growers and extension agents.


Materials and methods:

The experimental design was a randomized complete block with six replications. Three organically managed systems including 1) no cover crop, 2) cover crop incorporated prior to sweetpotato planting, and 3) cover crop reduced tillage were compared to a conventionally managed control. We made management decisions throughout the season that would optimize the production each system. Because site conditions varied among years in soil type, weed populations, field history, and climate, some variation of methods was necessary.

Sweetpotato management

All three organically managed treatments received compost composed of materials approved for use in organic production by the Organic Materials Review Institute (OMRI) at a rate of 20 kg ha-1 in May of 2001 and November 2002 and 2004. Following analysis, it was estimated that 90 kg N ha-1 would potentially be available to the sweetpotato crop during the growing season. Additionally, 0.56 kg ha-1 boron was applied to all treatments prior to planting.

Treatment management.
Organic, no cover crop. In May, overwintering weeds were incorporated with a tandem disc harrow, followed by one additional discing approximately two weeks later. Hills were formed with a ripper-bedder several days prior to sweetpotato planting. Slips were planted with a two row finger-type transplanter. Following sweetpotato planting, seasonal weed control was by rolling cultivator until prohibited by sweetpotato vine growth, at which time weeds were removed by mowing or by hand.

Organic, cover crop incorporated. In November of all three years, a cover crop mixture of 45 kg ha-1 hairy vetch inoculated with Rhizobium spp. and 67 kg ha-1 rye ‘Wrens Abruzzi’ was seeded with a grain drill on a disked flat soil surface. Cover crops were flail mowed when hairy vetch was in mid bloom (Creamer et al., 1995) on 23 May 2001, 3 May 2002, and 3 June 2004. Cover crops were incorporated in 2001 with a moldboard plow followed by a tandem disc harrow, a tandem disc harrow in 2002 and an articulating spading machine (Celli S.p.A., Forli, Italy) in 2004. Based on analysis and assuming 50% of N was available for plant uptake during the season (Baldwin and Creamer, 2005), N contributions from cover incorporated crop were estimated to be approximately 93, 71, and 64 kg N ha-1 in 2001, 2002 and 2004. Plots were disked twice in 2001 and 2004 and once 2002 prior to sweetpotato planting to ensure adequate distribution and decomposition of cover crop. Slips were planted with the same transplanter as in the organic, no cover treatment.

Organic, reduced tillage. In November of all three years, hairy vetch and rye cover crop mixture was seeded at the same rates as above on preformed hills. In 2000, cover crops were hand seeded with a broadcast seeder spreader (PlantMates Inc., Gallatin, TX). In 2000, the germination rate of vetch was poor on the crest of the hills compared to the furrows presumably due to the inability of the round seeded vetch to remain in place. Therefore, in 2001, hand seeding was followed by seed incorporation with a seeder cultipacker with notched rollers (SS-10, Brillion Iron Works, Brillion, Wis.). In 2003, the cultipacker was used to seed and incorporate the cover crop. In May 2002 and June 2004, cover crop was killed by rolling using the cultipacker. Plant available nitrogen from cover crop residues that remain on the surface is typically less than that of incorporated residue. Therefore, based on analysis, we predicted a maximum of 40% of reported nitrogen would potentially be available for plant uptake (75, 57 and 51 kg N ha-1 for 2001, 2002 and 2004, respectively) (Baldwin and Creamer, 2005). Sweetpotato slips were transplanted with the same transplanter as other treatments in 2001. In 2002 and 2004, slips were transplanted with subsurface tiller-transplanter (SST-T) (B&B No-Till, Laurel Fork, Va.) (Morse et. al., 1993). This transplanter cut through surface residue with a double disk coulter, opened a furrow, and placed water below the sweetpotato slip, placed the slip in the row, and closed the furrow with weighted press wheels.

Conventional. Recommended cultural and pest management practices for sweetpotato ‘Beauregard’ were followed throughout the season (Schultheis et al, 1999). Plots were fumigated (1,3-dichloropropene) primarily for nematode control two weeks prior to planting. EPTC (S-ethyl dipropylthiocarbamate) was applied for control of annual weeds mixed with chlorpyrifos [O,O-diethyl-O-(3,5,6-trichloro-2-pyridinyl) phosphorothioate] for control of soil dwelling insects [wireworm (Melanotus and Conoderus spp.), flea beetle (Systena spp.) and sweetpotato flea beetle (Chaetocnema confinis Crotch)] one week prior to planting in 2001 and two weeks prior to planting in 2002 and 2004. In addition, napropramide was applied to control dicot weeds one day after plant for weed control. Ammonium nitrate (NH4NO3) was banded and incorporated 28 days after plant each year to provide 56 kg ha-1. Slips were planted with the same transplanter as in the organic, no cover treatment.

Data collection

To minimize variation of mechanical practices as well as the effects of pesticide drift, we collected data in the center 8 rows and 18.3 m of each plot. Due to the extensive destructive sampling for this trial, a stratified random sampling pattern was used to ensure sample removal was from a previously undisturbed area.

Weed density and composition. Prior to weed removal events and biomass sampling at harvest, two 0.5-m2 frames were randomly placed perpendicular to the row in each plot, and weeds were identified and counted.

Plant biomass. Above ground biomass was removed from one randomly placed 0.5-m2 frame per plot one day prior to cover crop kill all three years. Biomass was separated into rye, vetch, and weed samples. The samples were analyzed for total N using a micro-Kjeldahl procedure (Keeney and Nelson, 1982). To determine if sweetpotato foliar growth was different among treatments, and to assess the proportion of weed biomass to sweetpotato foliar biomass, sweetpotato vines and weeds were removed at harvest from two 0.5-m2 frames per plot by cutting at the soil surface. Dry weights recorded for sweetpotato, monocot, and dicot weed species separately. Cover crop residue remaining in organic reduced tillage plots was collected at harvest. Soil was removed from the residue by sifting. All biomass samples were oven dried at 65 °C and weighed.

Insect density. Systems were baited for soil dwelling wireworm larvae species using a misxture of corn and wheat seeds in a 1:1 ratio. Plots were baited in two locations each sampling date. Samples were collected six times each season. Insects were removed from the soil approximately 1 week after baiting, and specimens were preserved in 70% ethanol solution until identified.

Leaf tissue sampling. At 30 and 60 days after planting (DAP) each year, 24 most recently mature sweetpotato leaves with petioles attached were randomly collected from each plot. Tissue was submitted for nutrient analysis to NCDA plant analysis laboratory.

Yield. Sweetpotato roots were collected from a 6.1 m long previously undisturbed section in the center of each of two adjacent rows approximately 100 days after planting. Roots were sorted by size according to USDA market grade standards and weighed by grade in the field.

Statistical analysis. All data were subjected to analysis of variance using PROC GLM (SAS V.8.2, 2001) to test the main effect of management system. Management and year were treated as fixed effects and blocks and appropriate error terms as random effects. Mean comparisons among management systems were generated using Fisher’s protected LSD at P = 0.05. When significant year by treatment interactions existed, data were analyzed and are presented by year. Weed density, dicot and total weed biomass and sweetpotato vine biomass were square root transformed to satisfy assumptions of normality and homogeneity of variances prior to analysis. For these data, statistical conclusions were derived from transformed data, and back transformed means derived from transformed data are reported here.

Research results and discussion:

Weed density in organically managed treatments varied with cover crop and tillage. In general, weed densities were similar between Org-NC and Org-CI treatments and intermediate between Org-RT and Conv. In 2001 and 2002, grass and sedge weeds were poorly controlled in Org-RT compared to other treatments. Although control was achieved by hand removal, the high frequency of goose grass in 2001 was especially difficult to manage in Org-RT. In 2001, the cover crop was flail chopped and it was difficult to remove the weeds by hand without removing the residue. Rolling the cover crop at kill (2002 and 2004) proved to be a much more effective method of preserving biomass on the bed surface. Rolling the cover crop rather than flail chopping was also associated with an increase in cover crop residue at sweetpotato harvest. Although some weeds were observed in the conventional system, cultivation and herbicides proved effective throughout the season, and weed density in this treatment was lower or similar to density in organically managed treatments. In 2004, there were no differences in weed densities or biomass among all four management systems. An increase in the number of weed control events, and implementing cultural controls earlier in the season contributed to optimal management in the final year of the trial.

In 2001, sweetpotato vine biomass was similar among systems until 60 days after planting (DAP). Previous research on sweetpotato ‘Beauregard’ indicated no yield advantage by continuing weed control efforts after about 6 weeks (Seem et al., 2003). In this trial, the final weed control event took place between 40 and 50 days after planting each year. Therefore, at 60 days after planting, some weeds were present in each system each year. At 60 DAP, foliar biomass was greatest in the Org-CI, intermediate and similar between Org-NC and Conv, and lowest in Org-RT. In 2002, the conventionally managed treatment established faster and produced more foliar biomass at 30 DAP than organic treatments. However, by 45 DAP and continuing through 60 DAP, organically managed treatments without cover crop and with cover incorporated were similar to Conv. Growth in the Org-RT treatment was lower than other treatments all year. These differences in sweetpotato foliar biomass reflected differences in weed density each year of the trial.

During the course of this trial there were at least six baiting events per year. In all, over 200 wireworm species were captured and identified. Most of the species present were Melanotus communis, although Conoderus vestpertinus was also important. Analysis is ongoing to determine if species density was different among treatments, and if density is a predictor to root feeding damage.

For all systems, nutrients including N, P, and K were within sufficiency ranges established for sweetpotato each year. These results demonstrate organically managed treatments with or without an incorporated cover crop prior to planting can perform as well as conventionally managed sweetpotato.

Yield of No.1, canners, and total roots in organically managed systems without cover crop and with cover crop incorporated prior to planting was equal to the conventional system each year. The reduced tillage treatment produced the same yield as remaining treatments in 2001 and 2004. Yield in this treatment was reduced in 2002 due to a significant increase in grass and sedge weed density and biomass, and consequent decline in sweetpotato foliage biomass.

Although economic analysis is still ongoing, preliminary data indicate costs among Org-NC, Org-CI and Conv will be similar. Costs associated with high analysis fertilizers and pesticides in the conventional system were matched by costs of compost in organic systems. Growers producing their own compost onsite may have reduced management costs. The outcome of the economic analysis will depend on the predicted income based on crop quality.

Literature cited

Baldwin, K. R. and N. G. Creamer. 2005. Cover crops for organic farming. North Carolina Department of Agriculture and Consumer Services. Submitted.

Blum, U., L. D. King, T. M. Gerig, M. E. Lehman, A. D. Worsham. 1997. Effects of clover and small grain cover crops and tillage techniques on seedling emergence of some dicotyledonous weed species. Amer. J. Alternative Agric. 4:146-161.

Creamer, N. G., B. Plassman, M. A. Bennett, R. K. Wood, B. R. Stinner, and J. Cardina. 1995. A method for mechanically killing cover crops to optimize weed suppression. J. Alt. Agr. 10:157-163.

Creamer, N. G., M. A. Bennett, B. R. Stinner, J. Cardina and E. E. Regnier. 1996. Mechanisms of weed suppression in cover crop-based production systems. HortScience 31(3):410-413.

Creamer, N. G., M. A. Bennett, B. R. Stinner. 1997. Evaluation of cover crop mixtures for use in vegetable production systems. HortScience. 32(5):866-870.

Drinkwater, L. E., D. K. Letourneau, F. Workneh, A. H. VanBruggen and C. Shennan. 1995. Fundamental differences between conventional and organic tomato agroecosystems in California. Ecol. Applic. 5(4):1098-1112.

Hardison, K. 2004. Personal communication. NCDA & CS. Horticulture Marketing Specialist. 2004 certified organic growers.

Hutchinson, C. M., and M. E. McGiffen, Jr. 2000. Cowpea cover crop mulch for weed control in desert pepper production. HortScience. 35(2):196-108.

Jackson, L. E., I. Ramirez, R. Yokota, S. A. Fennimore, S. T. Koike, D. M. Henderson, W. E. Chaney, F. J. Calderon and K. Klonsky. 2004. On-farm assessment of organic matter and tillage management on vegetable yield, soil, weeds, pests, and economics in California. Agric. Ecosystems Environ. 103(3):443-463.

Morse, R.D., D. H. Vaughan, and L. W. Belcher. 1993. Evaluation of conservation tillage systems for transplanting crops-Potential role the subsurface tillage transplanter. p. 145-151. In: P. K. Bollich (ed.). The evolution of conservation tillage systems. Proc. Southern Conservation Tillage Conference for Sustainable Agriculture. Monroe, Louisiana, 15-17 June.

[NCDA & CS] North Carolina Department of Agriculture and Consumer Services. 2004. North Carolina Agricultural Statistics 2004. North Carolina Department of Agriculture and Consumer Services and the U.S. Department of Agriculture.

Reburg-Horton, S. C., J. D. Burton, D. A. Danehower, G. Ma, D. W. Monks, J. P. Murphy, N. N. Ranells, J. D. Williamson, and N. G. Creamer. 2005. Changes in time in the allelochemical content of ten cultivars of rye (Secale cereale L.). J. Chem. Ecol. 31(1):179-193.

Ranells, N. N. and M. G. Wagger. 1997. Grass-legume bicultures as winter annual cover crops. Agron. J. 89:659-665.

Seem, J. E., N. G. Creamer, and D. M. Monks. 2003. Critical weed-free period for ‘Beauregard’ sweetpotato (Ipomoea batatas). Weed Tech. 17:686-695.

Staver, K. W. and R. B. Brinsfield. 1998. Using cereal grain winter cover crops to reduce groundwater nitrate contamination in the mid-Atlantic coastal plains. J. Soil Water Cons. 53:230-240.

Teasdale, J. R. and C. L. Mohler. 1993. Light transmittance, soil temperature, and soil moisture under residue of hairy vetch and rye. Agron. J. 85:673–680.

Toth, S. J., T. Melton, D. W. Monks, J. R. Schultheis, and K. A. Sorensen. 1997. Sweetpotato pesticide use survey in North Carolina. Data report for the Southern Region Pesticide Impact Assessment Program. Raleigh, NC: North Carolina State University. 91 p.

[USDA, ERS] U.S. Department of Agriculture and the Economic Research Service. 2003. Sweet Potato Statistics. Report No. 03001. 20 Jan. 2004. .

Villani, M. G. and F. Gould. 1986. Use of radiographs for movement analysis of the corn wireworm, Melanotus communis (Coleoptera, Elateridae). Environ. Entomol. 15(3): 462-464.

Participation Summary

Educational & Outreach Activities

Participation Summary

Education/outreach description:

Management of organic sweetpotato. Southeastern Vegetable and Fruit Expo. Greensboro, NC. Dec. 2001.

Conservation tillage in organic sweetpotato. National Sweetpotato Collaborators Meeting. Orlando, FL. Jan. 2003.

Sweetpotatoes – Undercover! National Sweetpotato Collaborators Meeting. Mobile, AL. Jan. 2003.

Management of cover crops and conservation tillage in organic
sweetpotato. Invited speaker. North Carolina Agricultural and Tobacco Foundation Meeting of Deans and Directors. Raleigh, NC. Feb 2003.

Evaluation of a cover crop mixture of hairy vetch (Vicia villosa Roth) and rye (Secale cereale L.) in organic sweetpotato (Ipomoea batatas [L.] Lam.) production in North Carolina. Abstr. Proceedings Southern Region American Society for Horticultural Science Conference. Mobile, AL. Feb 1-5, 2003.

Influence of reduced tillage on insects. Carolina Farm Stewardship Association’s Sustainable Agriculture Conference. November 2004. Asheville, NC.

Papers and Reports

Evaluation of organically managed sweetpotato systems in North Carolina. Ph.D. Thesis. Department of Horticultural Science. North Carolina State University. Expected completion date May 2005.

Project Outcomes

Project outcomes:

It will take several more years for the results of this trial to be incorporated into a production system that can be readily adopted by growers in our area. The organically managed systems performed as well and sometimes better than the conventionally managed system in terms crop productivity. However specific information on the effects of variety selection, plant spacing and planting date on weed suppression and yield are needed for organically managed systems.


Areas needing additional study

Due to the long growing season of sweetpotato (100 days), insects have ample time to feed on roots and can cause extensive damage even in systems receiving synthetic insecticides. Preliminary analysis of the 3 year field study indicated an increase in wireworm density in organically managed cover crop incorporated treatments compared to organically managed reduced tillage on several occasions. Cover crops influence soil bulk density either by reducing bulk density through the addition of green manures or by increasing bulk density when residues are allowed to remain on the surface as in conservation tillage systems. An understanding of the rate of movement under different bulk densities will also be beneficial to scouting efforts. Currently, baiting the soil with wheat and corn in the fall is the only way to determine if wireworms may be present in a field. This method is time consuming and not always informative. Because the distribution of larvae in the field is highly aggregated, baiting results are not always accurate. These results will help guide growers, extension agents and crop consultants make informed decisions about the distribution and frequency of baits, as well as provide some basis for understanding movement of larvae in cover cropped systems.


A SARE Enhancement Grant #EG032-002 (end date March 31, 2005) to conduct a laboratory study to observe wireworm movement in the soil under different soil physical conditions including bulk density and soil moisture. Little research has been completed to improve our understanding of the basic biology and ecology of wireworm. Previous research has indicated larvae movement is inhibited in soils with high bulk densities (Villani and Gould, 1985). Using methods first described by Gould and Villani, radiographs of M. communis were taken with a soft x-ray system (Faxitron 43805 N, Hewlett-Packard, Palo Alto, CA) in five environmentally different arenas that varied in soil bulk density and soil moisture. This technique makes it possible to view subterranean movement in an nondestructive manner.

Five plexiglass chambers (43 x 36 x 3 cm) were filled with sterilized field soil. Soil was ground with a 16 mm wire mesh screen, mixed, and autoclaved prior to filming. A bait composed of 1:1 mixture of untreated corn and wheat was soaked overnight when applicable to the treatment, and excess moisture removed by blotting. M. communis were removed from the field using baiting methods described above. One randomly selected larvae of similar length and weight were placed singularly in each of five arenas. We hypothesized larvae detect a food source through the delivery of sugars and other carbohydrates through the soil solution. Therefore, arenas were chosen to provide a range of responses to feeding stimuli through a substrate with a known bulk density and soil moisture. The five arenas include: 1) completely dry soil, no bait, 2) completely dry soil with dry bait, 3) completely dry soil with dry bait, water added immediately prior to addition of larvae 4) moist soil and dry bait and 5) moist soil and moist bait. Controls include arena 1, that tests the rate of movement with no feeding stimulus, and 2, that tests the rate of movement with food but no vector (soil moisture) to signal the presence of food. These five treatments were considered 1 replicate. Each treatment was repeated 5 times. M. communis were placed in one end of the box, while a bait of 1:1 corn and wheat is placed in the other end. The box was large enough to allow free movement of wireworms towards the bait. After wireworms were placed in the arenas, time was varied to allow for movement of wireworms towards the bait prior to radiography. Lead foil chips (1 mm diameter) were affixed to the abdominal segment (9th dorsal plate) by a thin layer of clear nail polish. The chip is the densest thing in the soil matrix and appeared as a white dot. Radiographs were taken with an x-ray system (Faxitron 43805N) using milliamperage (mA) of 2.8 and a kilovoltage of around 80. Exposure time was automatically controlled for the best image, and film was manually processed onsite. At this time, data are still being analyzed. A published report on the results of this trial and the insect portion of the field trial will be available to growers and extension agents following completion of data analysis.

In the field study, additional information is needed before organic systems can be adopted on a large scale in our area. Although weed control in the organic systems was successful, the additional hand weeding required in the reduced tillage system proved time consuming. Since weed emergence is frequently inhibited beneath dense crop canopies, reducing the time to achieve maximum foliar biomass through cultural practices may be beneficial, especially in the reduced tillage system. Because sweetpotato ‘Beauregard’ achieves maximum foliar biomass slower than other commercially available cultivars, future research could explore sweetpotato varieties and in row plant spacing best suited for organic production.

Any opinions, findings, conclusions, or recommendations expressed in this publication are those of the author(s) and do not necessarily reflect the view of the U.S. Department of Agriculture or SARE.