To determine the quality effect of ecocertified wines, we study 74,148 wines from California that have vintages ranging from 1998 to 2009, from 3,842 wineries. California accounts for an estimated nine-tenths of U.S. wine production, making more than 276 million cases annually.To avoid relying too heavily on any one expert’s taste, we gather data from three influential publications by wine experts: WA, WE, and WS. WA is a bimonthly wine publication featuring the advice of wine critic Robert M. Parker, Jr. WE is a lifestyle publication that was founded in 1988 by Adam and Sybil Strum and covers wine, food, spirits, travel, and entertaining. WS is a lifestyle magazine that focuses on wine and wine culture. During our period of study, the main tasters for California wines for WA, WE, and WS were Robert Parker, Steve Heim off, and James Laube, respectively. Information on each publication rating system is provided in Table 1. All the publications claim blind review.Each wine review provides information regarding the wine’s winery, vintage, appellation, and varietal, and most also provide information on the price of the wine and the number of cases produced. Each review also contains a score, a short description of the wine, and the review date.A significant increase in sweet potato [Ipomoea batatas Lam.] production area in the southeastern United States has occurred in the past decade, increasing from 33,548 ha in 2007 to 51,800 ha in 2017 . Sweet potato has proven to be a valuable crop with a national farm gate value of $705.7 million in 2016, up from $298.4 million in 2006 . North Carolina is the largest sweet potato-producing state, accounting for 54% of U.S. production . North Carolina, California, Mississippi, and Louisiana account for 94% of sweet potato production in the United States . Unfortunately, due to its prostrate growth habit and relatively slow growth, sweet potato does not compete well with problematic weeds,hydroponic dutch buckets resulting in reduced yields . Palmer amaranth and large crabgrass [Digitaria sanguinalis Scop.] are among the top five most common weeds in North Carolina sweet potato, with A. palmeri being identified as the most troublesome weed .

Amaranthus palmeri has been reported to be taller, to have a faster growth rate and greater leaf area, and to produce more overall biomass when compared with other Amaranthus species . Season-long A. palmeri interference is seen in vegetable crops, with reduced yield of 94% in bell pepper , 67% in tomato , 36% to 81% in sweet potato , with the greater yield losses associated with higher A. palmeri densities. Limited herbicide options exist for use in sweet potato . Growers rely on PRE herbicides, which do not always provide efficacious weed control and require rainfall for activation. POST herbicide options for A. palmeri control in sweet potato are limited to between-row applications of carfentrazone or glyphosate . The lack of POST herbicides forces growers to use tillage for control of weeds until row closure, at which time growers have no additional control options for dicotyledonous weeds other than mowing weeds above the cropcanopy and hand weeding, which is a costly control measure . Digitaria sanguinalis is commonly found in fruit and vegetable crops but has not been highly ranked as a problematic weed due to efficacious POST herbicides such as clethodim, fluazifop, or sethoxydim . Although these graminicides can be effective, grasses escaping herbicide application or sprayed after substantial establishment may continue to compete with the crop and reduce yields. Furthermore, herbicide resistance management for D. sanguinalis should be considered, as resistance to acetyl-CoA carboxylase herbicides, including those registered for use in sweet potato has been reported . While its impact on sweet potato has not been reported, season-long, D. sanguinalis reduced yield in bell pepper by 46% , snap bean by 47% to 50% , and watermelon [Citrullus lanatus Matsum. & Nakai] by 82% . A better understanding of the interactions of A. palmeri and D. sanguinalis with sweet potato would allow for better decision making regarding their control. Thus, the objectives of this study were to determine the effect of five densities of A. palmeri and D. sanguinalis on sweet potato biomass and storage root yield and quality, the intraspecific response of A. palmeri and D. sanguinalis across five densities with and without sweet potato, and the effect of sweet potato on growth of A. palmeri and D. sanguinalis.

Field studies were conducted with ‘Covington’ sweet potato at the Horticultural Crops Research Station near Clinton, NC on a Norfolk loamy sand with humic matter 0.31% and pH 5.9 in 2016 and an Orangeburg loamy sand with humic matter 0.47% and pH 5.9 in 2017. Nonrooted ‘Covington’ sweet potato 20- to 30-cm-long cuttings were mechanically planted approximately 7.6-cm deep into ridged rows 1 m apart in the entire study at an in-row spacing of approximately 30 cm on June 9, 2016, and June 12, 2017. At 1 d after transplanting, sweet potato plants were removed by hand in the no-sweet potato treatments. On the same day, treatment rows assigned A. palmeri or D. sanguinalis were broadcast seeded on the soil surface and lightly raked to a depth of approximately 1.0 cm. After weed seeding, the entire study was irrigated with 1.3 cm of water using overhead irrigation to aid in weed seed establishment. No additional irrigation was applied, in either year, after the initial irrigation event. Treatments consisted of a single weed species at five weed densities grown with and without sweet potato arranged in a randomized complete block design with three replications . Amaranthus palmeri and D. sanguinalis were hand thinned to treatment densities of 0 , 1, 2, 4, and 8 and 0 , 1, 2, 4, and 16 plants m−1 of row, respectively, when A. palmeri was approximately 8 cm tall, and D. sanguinalis had two expanded leaves. At the time of weed thinning, sweet potato averaged one to two newly expanded leaves on each plant. Densities of A. palmeri and D. sanguinalis were based on those used in previous research . Plots consisted of two bedded rows, each 1-m wide by 5-m long, with the first row being a weed-free buffer row planted to sweet potato and the second row a treatment row. Treatment rows were maintained at specific weed treatment densities, and border rows were maintained free of weeds season-long by weekly removal by hand. Cultural practices for conventional sweet potato production in North Carolina were followed . Season-long rainfall and growing degree day data are presented in Table 1. Two days before sweet potato harvest, 5 sweet potato plants and 5 plants of each weed species were randomly harvested at the soil level from each plot to determine above ground biomass. Samples were placed in 2-ply paper yard waste bags measuring 40 by 30 by 89 cm and fresh biomass was recorded. Samples were then placed in a propane-heated,bato bucket forced-air drier for 96 h at 80 C. Once dry, samples were removed and weighed immediately to determine dry biomass. To determine fresh and dry sweet potato and weed biomass on a per plant basis, total sweet potato or weed biomass within a treatment and replication was divided by the number of plants harvested.

To determine dry biomass per meter of row, individual weed biomass was multiplied by sweet potato plant and/or weed number in 1 m of row, respectively. Sweet potato storage roots were harvested at 113 d after transplanting in 2016 and at 107 DAT in 2017. In both years storage roots were harvested with a tractor-mounted two-row chain digger and hand sorted into jumbo , no. 1 , and canner grades and weighed. Total marketable yield was calculated as the sum of jumbo and no. 1 grades. Data for crop biomass, individual weed biomass, weed biomass per meter of row, yield, and quality were subjected to ANOVA using PROC MIXED in SAS . Treatment, year, and treatment by year were considered fixed effects, while replication within year was treated as a random effect. Year was treated as a fixed effect to further evaluate components of the year by treatment interaction, such as year by weed density and year by crop presence or absence. If the treatment by year interaction was not significant, a contrast statement was used to test for a linear trend for dependent variables with increasing weed density, calculated separately for each weed species. All response variables, except canner yield, were square-root transformed to reduce both data skewness and variance heterogeneity before carrying out the mixed model ANOVA. Least squares means for these response variables are reported without the applied transformations and separated according to Tukey’s HSD . Results for estimated marketable yield loss per weed as weed density approaches zero for A. palmeri and D. sanguinalis were 119% and 61%, respectively. The higher estimated marketable yield loss as weed density approaches zero for A. palmeri relative to D. sanguinalis indicated higher competitive capacity of A. palmeri at low densities. These results for A. palmeri are consistent with another study in sweet potato but higher than in soybean [Glycine max Merr.] , peanut , and corn . Estimated yield loss as weed density approaches zero in the present study indicates that A. palmeri and D. sanguinalis, even at low densities, can greatly reduce sweet potato marketable yield. The initial yield loss as weed density approaches zero for D. sanguinalis was less than A. palmeri at lower densities. However, sweet potato yield loss from interference byD. sanguinalis was higher than yield loss reported in snap bean . For parameter A, the asymptote of the regression model estimating the maximum yield loss due to weed density was 87% for A. palmeri and 83% for D. sanguinalis. Meyers et al. estimated a maximum marketable yield loss of 90% at A. palmeri densities of 6.5 plants m−1 of sweet potato row. Findings from our study further support the findings of Meyers et al. , who also reported the highly competitive nature of A. palmeri with sweet potato. To reduce interference of A. palmeri and D. sanguinalis, which are commonly reported in sweet potato, growers should use a combination of efficacious PRE herbicides, as outlined by Meyers et al. , in combination with tillage, hand removal, and mowing . Although POST herbicides for A. palmeri are limited, POST herbicide options for selective grass control in sweet potato are available and should be used when D. sanguinalis is less than 10 cm to minimize yield loss. If D. sanguinalis resistance is suspected, then alternative methods should be analyzed for control. Growers should not dismiss the impact of either weed, as a single A. palmeri or D. sanguinalis per meter of row reduced marketable yield by 50% and 35%, respectively . Reduction in marketable yield loss was due to a decrease in weight of no. 1 and jumbo sweet potato grades. Amaranthus palmeri decreased the yield of no. 1 and jumbo grades at all densities greater than 1 plant m−1 row when compared with weed-free sweet potato yields . Similarly, D. sanguinalis at 1 plant m−1 row decreased the weight of sweet potato jumbo grade when compared with the weed-free check . Digitaria sanguinalis densities greater than 2 plants m−1 row decreased no. 1 grade sweet potato yield relative to the weed-free check, with 16 plants m−1 causing the greatest loss of no. 1 and jumbo grades.These findings further demonstrate the negative impact of A. palmeri and D. sanguinalis on sweet potato yields at low weed densities. Interspecific competition is also reflected in biomass reduction of one or both plant species competing with each other Interactions between year, crop versus no crop, and weed density were not significant ; therefore, means pooled over years were obtained for density and crop versus no crop combinations for each weed species. Biomass per meter of row of A. palmeri and D. sanguinalis increased with increasing weed density . The presence of sweet potato reduced overall biomass per meter of row for both weed species at densities of 1, 2, and 4 plants m−1 row. Furthermore, sweet potato reduced the rate of bio-accumulation for D. sanguinalis, as can be seen when comparing the slopes of biomass accumulation of both weeds . We believe that this was an effect of weed height, as A. palmeri quickly establishes and reduces the light reaching the sweet potato canopy, whereas D. sanguinalis does not exceed the sweet potato canopy height as quickly as A. palmeri and is therefore less competitive with sweet potato for light. The impact of A. palmeri on light interception with the sweet potato canopy has been documented by others . Individual weed biomass of A. palmeri and D. sanguinalis was similar across weed densities when grown with sweet potato .