Evaluating Mustard Seed Meal for Weed Suppression in Potato ( Solanum tuberosum )

Mustard seed meal (MSM) derived from Sinapis alba controls weeds for several weeks following application to soil, but also has potential to injure the planted crop. Producers of certified organic potatoes typically utilize a combination of cover crops, soil hilling, harrowing, and cultivation for weed control. Once the potato canopy nears row closure, most late emerging weeds are suppressed by the dense potato canopy. MSM may have value for early season weed suppression in potato, but has not been previously tested. Our objectives were to determine the weed control efficacy and potato tolerance to MSM. We evaluated response of potatoes and weeds to MSM applied at 1.1, 2.2, and 4.5 MT ha applied in a band on the potato hill just after a shallow harrowing and prior to potato emergence as part of an integrated weed management program in potatoes. MSM applied at 2.2 and 4.5 MT ha reduced early season grass and broadleaf weed density 73 to 99% and 54 to 98%, respectively, in potato and reduced late season broadleaf weed biomass 71 to 94% in all three years tested. Early season grass weed density in potato was reduced by MSM at 2.2 and 4.5 MT ha. Late season grass weed biomass was reduced by the highest rate of 4.5 MT ha. MSM at 4.5 MT ha caused minor injury to potato (3 to 15%) at 3 weeks after emergence, but did not reduce total tuber yields or percentage of US No.1 tubers. MSM could be a component of an integrated weed control program in potato.


Introduction
Consumer demand for certified organic produce grown without synthetic pesticides continues to increase and with the proliferation of organic agriculture, the need for new bioherbicides to control weeds has grown (Cai & Gu, 2016).Weed management in conventional potato production relies heavily on synthetic herbicides due to their outstanding efficacy, ease of use, and relatively low cost.More U.S. potato hectares (75%) are treated with the herbicide metribuzin than any other pesticide and 91% of potato hectares receive an herbicide (Guenthner, Wiese, Pavlista, Sieczka, & Wyman, 1999;USDA-NASS, 2014).Weed management in organically grown potatoes typically consists of a combination of an early cultivation that removes the top of the hill formed at planting (drag off), a cultivation to reform the hill (rehilling) that uproots and buries early emerging weed seedlings prior to potato emergence, additional shallow cultivations if needed, and reservoir tillage (dammer-diking) that creates shallow pockets in the furrows to improve water infiltration and prevent runoff (Boydston & Vaughn, 2002;Boydston, 2010).Once the potato canopy is near row closure, later emerging weeds are largely suppressed by the dense potato canopy.Excessive cultivation decreases soil organic matter, increases soil erosion, and has potential to injure crop roots.A biopesticide that could control weeds and decrease the need for cultivation would be useful.
Although MSM suppresses weed emergence and growth, it also can negatively affect crop seed germination and growth (Rice, Johnson-Maynard, Thill, & Morra, 2007;Earlywine, Smeda, Teuton, Sams, & Xiong, 2010;Boydston, Morra, Borek, Clayton, & Vaughn, 2011).MSM derived from Sinapis alba applied to the soil at 2.2 and 4.5 MT ha -1 reduced weed emergence by 76 to 91% for a period of 4 wks in onion (Allium cepa) (Boydston et al., 2011).However, onions were severely injured when MSM was applied at the 1-leaf stage or earlier.Beets (Beta vulgaris) and lettuce (Lactuca sativa) were injured when seeding into MSM amended soil (Rice et al., 2007).Wheat emergence was inhibited by seed meals from 15 glucosinolate-containing plant species when applied at 1% (w/w) to a sandy loam soil (Vaughn, Palmquist, Duval, & Berhow, 2006).Delaying planting of the crop following MSM application may reduce phytotoxicity of MSM to the crop, but would also limit the longevity of weed suppression in the crop.Transplanted broccoli and direct seeded spinach were tolerant of MSM incorporated into the soil at 1.2 and 4.5 MT ha -1 two weeks prior to planting (Shrestha, Rodriguez, Pasakdee, & Bañuelos, 2015).
Corn gluten meal is one of many products extracted from corn during the wet milling process and has been demonstrated to have herbicidal activity when applied preemergence to weeds (Christians, 1993;Yu & Morishita, 2014).Dried distillers grains with solubles (DDG) are a less expensive byproduct of ethanol production commonly fed to cattle and reduced emergence of common chickweed and annual bluegrass when applied to the surface of potting soil at relatively high rates of 8 to 16 MT ha -1 ) (Boydston, Vaughn, & Collins, 2008b).The relatively high rate of DDG required for herbicidal activity may make it less practical as a bioherbicide, but its nitrogen content (4.4% by wt.) makes it appealing as an organic fertilizer.
To be useful for weed control in potato production systems, rates and timing of applications of MSM must be determined in order to minimize the potential for phytotoxicity to the crop.Potato is a relatively deep-planted crop that is grown from a vegetative tuber which may allow it to escape injury typically observed when MSM is applied preemergence to crops grown from a true seed.Potatoes generally take 3 to 5 weeks to emerge depending on planting depth, soil temperatures, and seed condition.This allows for early cultivation and rehilling operations that control early weed flushes (Boydston & Vaughn, 2002;Boydston, 2010).A harrow or rod weeder operation that removes the top of the potato hill and leaves a flat topped hill (termed "drag off") is often practiced when potato shoots have sprouted, but still not emerged.Applying MSM following drag off may provide several weeks of weed control without damage to potato and possibly eliminate the need for additional cultivation, which can bring new weed seed near the soil surface where it can subsequently germinate (Grundy, Mead, & Burston, 1999).Once emerged, most potato cultivars grow rapidly and provide a dense canopy that is closed within 7 to 8 weeks, shading out later emerging weeds (Connell, Binning, & Schmitt, 1999).This research evaluated the response of potato and weeds to three rates of MSM applied to the soil surface after harrowing off the tops of potato hills prior to potato emergence.A single rate of DDG was included as a comparison treatment representing an organic amendment containing a similar amount of nitrogen, but with nominal herbicidal activity at the rate used.

Site Conditions and Management
All studies were conducted at the USDA-ARS Paterson research farm from 2006 to 2008 near Paterson, WA.The soil was a Quincy sand (mixed mesic Xeric Torripsamments) containing 0.4% organic matter and pH 7.0.Wheat preceeded potatoes each year and straw was removed.Fields were chisel plowed, disked, and harrowed in the spring to prepare ground for potato planting.Preplant fertilizer was applied based on soil tests and in-season nitrogen was applied through a center pivot irrigation system and based on leaf petiole sampling (Lang, Stevens, Thornton, Pan, & Victory, 1999). Potato, var. 'Umatilla Russet' was planted in 2006and var. 'Ranger Russet' in 2007and 2008 at a seed piece spacing of 22 cm in rows spaced 86 cm apart.Seed pieces were placed 17 cm deep with a 6-row potato planter.Three to four weeks after planting (Table 1), the top of potato hills that were formed during planting were harrowed with a shallow operated rod weeder to control early germinating weeds and to form a flat-topped hill which provided a level surface for the mustard seed meal application.Plots were irrigated using center pivot irrigation and standard fungicides and insecticides were applied as needed to control pests other than weeds according to standard potato pest management practices in the western United States (Strand, 2006).
DDG was obtained from a commercial ethanol plant near West Burlington, IA (Big River Resources, LLC).MSM from S. alba cultivar 'IdaGold' was obtained locally (McKay Seed Co., Moses Lake, WA).Mustard seed was cold pressed and extruded flakes were run through a hammer mill to obtain a more uniform and spreadable granular material.Glucosinolate content of MSM was determined in 2006 and 2008 by a modification of a high performance liquid chromatography (HPLC) method developed by Betz and Fox (1994) and described previously (Boydston et al., 2011).Sinalbin quantitation was determined from a standard curve prepared from purified sinalbin previously isolated from defatted white mustard seed meal using the method of Vaughn and Berhow (2005).Three subsamples were tested from each batch of MSM and the entire process repeated once.
MSM was applied at 1.1, 2.2, and 4.5 MT ha -1 and DDG at 2.2 MT ha -1 concentrated in a 31 cm wide band on the top of each potato hill shortly after harrowing (Table 1).Plots were sprinkler irrigated with 0.6 cm water within 4 h of application.A nontreated control treatment was included.In 2008, an additional treatment consisting of an application of rimsulfuron applied postemergence at 26 g ha -1 on May 13, 2008 and then hand weeded at weekly intervals to remove any weeds that escaped herbicide treatment.Treatments were arranged in a randomized complete block design replicated four times.Individual plots were 1.7 by 6.1 m and contained two potato rows.
All plots were reservoir tilled (dammer-diked) in early-to mid-May each year which controlled most early emerging weeds in the furrows and sides of the potato hill (Table 1).Potato injury was rated visually on a scale of 0 = no injury to 100 = death at 12, 20 to 27, and 31 to 40 days after emergence (DAE).Potato yield was determined by weighing tubers mechanically harvested from the entire plot.Tubers were graded by size and quality according to USDA grading standards (Anonymous, 1991).U.S. No. 1 potatoes were grouped into five size categories; 113 to 170 g, 171 to 227 g, 228 to 284 g, 285 to 340 g, and >340 g and the percentage by weight in each category determined.Although not considered U.S. No. 1, undersized tubers (< 113 g) that contained no blemishes or malformations were also included in size grading.Specific gravity was measured on ten, 230-to 280-g tubers in each treatment.

Data Analysis
Analysis of variance (ANOVA) was performed using the PROC GLIMMIX procedure in SAS (Statistical Analysis Systems®, 2014, version 9.4, SAS Institute Inc., SAS Campus Drive, Cary, NC 27513, USA).Blocks and years by blocks were treated as random factors.Data were pooled across years when no significant year by treatment interaction was detected.Data from each year were analyzed and presented separately when a significant year by treatment interaction was detected or when the elimination of treatments with no variance from the analysis (crop injury data) resulted in a different number of treatments among years.Mean separation was conducted using Tukey-Kramer for the calculated LSMEANS (P = 0.05).When crop injury data contained only one treatment mean greater than zero, t tests at 0.05 level were used to compare means to zero.

Weed Density and Biomass
In 2006, hairy nightshade accounted for 47% of total early season weed counts, followed by barnyardgrass and large crabgrass, redroot pigweed, and common lambsquarters.Hairy nightshade was the most prevalent broadleaf weed in 2007 accounting for 39% of the total weed early season weed counts, followed by barnyardgrass and large crabgrass, and lesser amounts of Russian thistle, common lambsquarters, and redroot pigweed.In 2008, no grass weeds were emerging on May 12 and hairy nightshade comprised 98% of the emerged weeds with sparse amounts of Russian thistle and common lambsquarters.
There were significant treatment effects and year by treatment effects on early season weed density so the data are presented by year.Early season grass weed density was reduced by the two highest rates of MSM and broadleaf weed densities were reduced by the highest rate of MSM compared to nontreated checks in 2006 (Table 2).DDG at 2.2 MT ha -1 and the 1.1 MT ha -1 rate of MSM did not reduce weed densities.In 2007, early season grass and broadleaf weed densities were reduced by the two highest rates of MSM compared to the nontreated checks.In 2007, the lowest rate of MSM and DDG at 2.2 MT ha -1 also reduced broadleaf weed density compared to the nontreated checks, but did not reduce grass weed density (Table 2).There were no early season grass weeds in 2008, but broadleaf weed density was significantly reduced by all MSM rates tested.DDG at 2.2 MT ha -1 did not reduce early season broadleaf weed density in 2008.Total (grass and broadleaf) early season weed density in nontreated checks was greatest in 2007 averaging 141 seedlings per m 2 .Note. a Lsmeans within a column and year followed by the same letter do not differ significantly at P = 0.05 according to Tukey-Kramer lsmeans test.
These results are in agreement with previous research where MSM at 2.2 to 4.5 MT ha -1 reduced early season weed density in field trials on onions, cucurbits, broccoli, and spinach (Boydston et al., 2011;Shrestha et al., 2015;Webber, White, Boydston, & Shrefler, 2017).The relatively poor weed suppression with DDG at 2.2 MT ha -1 is in agreement with greenhouse studies by Boydston et al. (2008b) in which rates of 400 g m -2 (equivalent to 4 MT ha -1 ) or lower applied to the soil surface did not suppress weeds.
There were both year and treatment effects on late season weed density and biomass, but not significant year by treatment interactions, so the combined data are presented (Table 3).Late season grass weed density was not impacted by MSM or DDG treatment, but grass weed biomass was reduced from 154 g m -2 in nontreated checks to 33 g m -2 by the 4.5 MT ha -1 rate of MSM.Both late season broadleaf weed density and biomass were reduced by the highest rate of MSM compared to the nontreated check.Broadleaf weed biomass was reduced from 644 g m -2 in nontreated checks to 40 g m -2 by MSM applied at 4.5 MT ha -1 (Table 3).The 1.1 and 2.2 MT ha -1 rates of MSM reduced late season broadleaf weed biomass compared to nontreated checks, but did not significantly reduce grass weed biomass.DDG did not significantly reduce late season grass or broadleaf weed counts or biomass.Handiseni et al. (2011), andWang, Gu, Niu, andBaumann (2015) reported greater suppression of broadleaf weed species than grass weed species with MSM from S. alba.In those studies, MSM derived from Brassica juncea was more effective on grass species than S. alba MSM.Late season weed biomass, consisting mainly of barnyardgrass, was not reduced in broccoli and spinach treated with up to 4.5 MT ha -1 MSM, indicating that the weed suppressive effects of MSM did not last season-long (Shrestha et al., 2015).
Field tests of MSM in several vegetable crops have resulted in lower early season weed density, but often failed to control later emerging weeds (Boydston et al., 2011;Shrestha et al., 2015;Webber et al., 2017).In the current study, we suspect the vigorous growing potato canopy likely prevented later emerging weeds from establishing or producing considerable biomass.When higher rates of MSM within the crop row was combined with reservoir tillage in the interrow, early season weeds were often suppressed and the potato canopy was able to provide good weed suppression for the remainder of the season.Note. a Lsmeans within a column and treatment or year followed by the same letter do not differ significantly at P = 0.05 according to Tukey-Kramer lsmeans test.Lsmeans not followed by a letter indicate there was no significant treatment or year effect.

Potato Response and Yield
Twelve days after potato emergence (DAE), slight injury was observed in all three years ranging from 4 to 6% for the 4.5 MT ha -1 rate of MSM (Table 4).In 2008, the 2.2 MT ha -1 rate of MSM also caused 2% injury at 12 DAE, but was not statistically different from nontreated controls.Injury consisted of chlorotic and stunted growth of the emerged potato shoots.The greatest potato injury of 15% was observed in 2008 at 22 DAE with the high rate of MSM.In 2008, MSM was applied only two days before potato emergence which may have increased the potential for MSM to injure potato shoots close to the surface.Potato injury was never extensive and declined after 3 to 4 weeks after emergence.
There was no significant year or year by treatment interaction on total potato tuber yield, so the combined data is presented (Table 4).Potatoes treated with MSM at 2.2 MT ha -1 yielded 70.5 MT ha -1 , significantly greater than nontreated checks and DDG treated potatoes, which averaged 60.2 and 59.0 MT ha -1 , respectively (Table 4).The lack of early season in-row weed control in the nontreated checks and DDG treatments likely contributed to the lower tuber yields.Potatoes treated with MSM at 1.1 and 4.5 MT ha -1 yielded 63.6 and 65.2 MT ha -1 , respectively.
In 2006, the percentage of tubers in the US No.1 category averaged 82% and was not impacted by soil amendments tested (Table 4).In 2007, MSM at 1.1 MT ha -1 averaged the greatest percentage of US No.1 tubers (93% by weight) and the nontreated checks and MSM at 4.5 MT ha -1 averaged the least (86 and 88% by weight, respectively) (Table 4).However, in 2008, MSM at 4.5 MT ha -1 averaged the greatest percentage of US No.1 tubers (94% by weight) and nontreated checks averaged the least (89% by weight) (Table 4).The individual positive effects of reduced early season weed competition and potential negative effects of potato phytotoxicity from MSM treatments on both total tuber yield and percentage of US No. 1 tubers cannot be completely sorted out in the current study.In 2008, a weed-free herbicide control treatment was included that yielded equal to the nontreated control indicating that uncontrolled in-row weeds in 2008 were not great enough to negatively impact potato tuber yield and percent US No. 1 tubers.
In 2006 and 2008 there were no significant differences in percentages of tubers in various size classes among treatments (data not shown).In 2006, the greatest percentage of tubers were in the 113 to 170 g (29% by weight) and 170 to 227 g (23% by weight) size categories and least percentage in the > 340 g size category (6% by weight) (data not shown).In 2008, the greatest percentage of tubers were in the > 340 g size category, averaging 33% by weight and the least percentage of tubers were in the < 113 g size category, averaging 5% by weight (data not shown).The later harvest date in 2008 compared to 2006 probably contributed to the larger percentage of tubers in the larger size categories.Note. a Lsmeans within a column and year followed by the same letter do not differ significantly at P = 0.05 according to Tukey-Kramer lsmeans test or according to t test at P = 0.05 (cases with only one mean greater than zero).Means not followed by a letter indicate there was no significant treatment effect.
In 2007, there was a significant treatment effect on the percentage of tubers in various size classes (Figure 1).MSM at 2.2 MT ha -1 resulted in a greater percentage of tubers in the 228 to 284 g and 285 to 340 g size classes and fewer undersized tubers (< 113 g) and small tubers 113 to 170 g size class compared to nontreated checks, which averaged a greater percentage of undersized and small tubers and lesser percentage in the 228 to 284 g and 285 to 340 g size categories.Similarly, MSM at 1.1 MT ha -1 also resulted in a greater number of tubers in the 228 to 284 g size class and fewer in the undersized class (< 113 g) than the nontreated checks.Weed competition in the nontreated checks likely contributed to the greater percentage of smaller-sized tubers (Boydston & Vaughn, 2002).Weed pressure (both early season weed density and late season weed biomass) tended to be greater in 2007 than in 2006 and 2008 and is likely the reason a significant response in tuber size was observed in 2007.
However, MSM at 4.5 MT ha -1 generally reduced early and late season weeds the most but resulted in tuber size profile similar to nontreated checks (Figure 1).Perhaps the early season phytotoxicity from the high rate of MSM in 2007 contributed to a slightly smaller tuber size profile compared to the 2.2 MT ha -1 rate of MSM.
Potato tuber specific gravity was significantly affected by year (P < 0.0001) and there was a significant year by treatment effect (P = 0.0197   MSM applied at 2.2 and 4.5 MT ha -1 just after a shallow harrowing and prior to potato emergence suppressed early season weed emergence and coupled with the shading of later emerging weeds by the crop canopy, prevented potato tuber yield loss from excessive weed competition.MSM could be a component of an integrated weed control program in potato if limitations of availability and spreading are overcome.

Table 1 .
Timing of potato planting, harrowing of hill tops, mustard seed meal application, reservoir tillage, and harvest operations in2006, 2007, and 2008 at Paterson, WA weight) was determined by clipping plants at the soil surface, drying at 40 °C for 48 h and weighing.

Table 2 .
Early season grass and broadleaf weed density in potatoes following preemergence treatment of mustard seed meal or dried distillers grains in2006, 2007, and 2008 near Paterson, WA

Table 3 .
Late season grass and broadleaf weed density and biomass in potatoes following preemergence treatment of mustard seed meal or dried distillers grains in2006, 2007, and 2008 near Paterson, WA

Table 4 .
Early season potato injury, total potato tuber yield, and percentage of U.S. No.1 tubers following preemergence treatment of mustard seed meal or dried distillers grains in2006, 2007, and 2008 near Paterson,  WA ). MSM or DDG . In 2008, ,t on tuber specific gravity in2006  and 2007.In 2008, , placement, rate, crop growth stage, and transplants versus direct seeded are crucial for greater adoption by producers.