Dry beans such as adzuki bean (Vigna angularis L.), black bean (Phaseolus vulgaris L.), kidney bean (Phaseo- lus vulgaris L.) and white bean (Phaseolus vulgaris L.) are important cash crops grown in southwestern Ontario. Short physical stature of these crops makes them not very competitive against weeds. Weed interference has been shown to reduce seed yield of dry bean as much as 70% [1] . Common annual weeds in dry bean production in Ontario include redroot pigweed (Amaranthus retroflexus L.), common lambsquarters (Chenopodium album L.), common ragweed (Ambrosia artemisiifolia L.), velvetleaf (Abutilon theophrasti Medicus), ladysthumb (Po- lygonum persicaria L.), eastern black nightshade (Solanum ptycanthum Dun.), and annual grasses such as Seta- ria, Digitaria, Echinochloa, and Panicum species [2] .

Saflufenacil is a new herbicide developed by BASF for burndown and residual broadleaf weed control in corn, soybean, and other crops applied prior to crop emergence. Saflufenacil is a pyrimidinedione that inhibits the protoporphyrinogen-IX-oxidase (PPO) enzyme which is critical for chlorophyll synthesis. Saflufenacil is ab- sorbed by both roots and foliage of plants. Saflufenacil is mainly translocated in the xylem and has limited mo- bility in the phloem. Plants susceptible to saflufenacil show injury symptoms within a few hours and die in 1 to 3 days after treatment (Anonymous 2008). Saflufenacil is applied at relatively low doses and has low environ- mental, toxicological and eco-toxicological impact with minimal residual carryover as the herbicide does not persist in the soil [3] . Saflufenacil effectively controls several broadleaf weeds in Ontario including redroot pigweed, common ragweed, common lambsquarters, velvetleaf, common cocklebur (Xanthium strumarium L.), giant ragweed (Ambrosia trifida L.) and common waterhemp (Amaranthus rudis Sauer) including triazine-, ace- tolactate synthase- and glyphosate-resistant biotypes [3] [4] . Saflufenacil is not very efficacious on annual grasses such as Setaria, Digitaria, Echinochloa, and Panicum species.

Dimethenamid-p is a chloroacetanilide herbicide that controls annual grasses such as green foxtail, yellow foxtail [Setaria glauca auct., non (L.) P. Beauv.], giant foxtail (Setaria faberi R.A.W. Herrm.), barnyardgrass [Echi- nochloa crus-galli (L.) Beauv.], witchgrass (Panicum capillare L.), fall panicum (Panicum dichotomiflorum Michx.), smooth crabgrass [Digitaria ischaemum (Schreb.) ex Muhl.], and large crabgrass [Digitaria sanguinalis (L.) Scop.] and some small-seeded broadleaf weeds such as redroot pigweed and eastern black nightshade in- cluding triazine- and acetolactate synthase-resistant biotypes [5] -[7] .

Saflufenacil/dimethenamid-p is a new pre-packaged herbicide mixture available for use in soybean at a max- imum dose of 245 g ai ha⁻¹ (25:220 g ai ha⁻¹ of saflufenacil:dimethenamid-p), in Ontario [2] . Saflufenacil used at a rate of 100 or 200 g ai ha⁻¹ in Ontario has been shown to cause significant injury in some market classes of dry bean [8] . Dimethenamid-p and other chloroacetamide herbicides have been shown to have an adequate mar- gin of crop safety in various market classes of dry beans [5] [9] -[11] .

There is little information on the tolerance of black, white, adzuki and kidney beans to saflufenacil/dime- thenamid-p herbicide mixture under Ontario environmental conditions. Saflufenacil at the proposed reduced rate of 25 g ai ha⁻¹ in mixture with dimethenamid (220 g ai ha⁻¹) may be an effective option for broad-spectrum con- trol of annual grass and broadleaf weeds in dry beans.

The objective of this research was to evaluate the sensitivity of black, white, adzuki and kidney bean to saflu- fenacil/dimethenamid-p applied preemergence at the 1X and 2X manufacturer’s recommended rates for soybean under Ontario environmental conditions.

Field trials were conducted at the Huron Research Station at Exeter, ON, Canada and at Ridgetown Campus of the University of Guelph at Ridgetown, ON, Canada in 2012 and 2013. The soil at Exeter was a Brookston clay loam (Orthic Humic Gleysol, mixed, mesic, and poorly drained) with 41% sand, 35% silt, 24% clay, 3.2% or- ganic matter and pH of 7.9 in 2012 and 29% sand, 44% silt, 27% clay, 3.6% organic matter and pH of 7.7 in 2013. The soil at Ridgetown was a Watford (Grey to Brown Brunisolic, mixed, mesic, sandy, and imperfectly drained)-Brady (Gleyed Brunisolic Grey to Brown Luvisol, mixed, mesic, sandy, and imperfectly drained) sandy loam with 56% sand, 27% silt, 17% clay, 5.4% organic matter and pH of 6.7 in 2012 and 52% sand, 28% silt, 20% clay, 5.9% organic matter and pH of 6.4 in 2013. Seedbed preparation at all sites consisted of fall moldboard plowing followed by two passes with a field cultivator with rolling basket harrows in the spring.

The study was established as a split-plot in a randomized complete block design with four replications. Main plots consisted of untreated check, dimethenamid-p (220 g ai ha⁻¹), dimethenamid-p (440 g ai ha⁻¹), saflufenaci (25 g ai ha⁻¹), saflufenacil (50 g ai ha⁻¹), dimethenamid-p/saflufenacil (220 + 25 g ai ha⁻¹), and dimethenamid- p/saflufenacil (440 + 50 g ai ha⁻¹) and sub-plots were market classes of dry bean which consisted of one row each of black bean (“Black Velvet”), white bean (“T9905”), adzuki bean (“Erimo”) and kidney bean (“Red Hawk”) spaced 0.75 m apart. Plots were 3 m wide at both locations and 8 m long at Ridgetown and 10 m long at Exeter. Black, white, and adzuki bean were seeded at 250,000 seeds ha⁻¹ and kidney bean was seeded at 200,000 seeds ha⁻¹ using different seed disks at a depth of 5 cm in late May to early June.

Herbicide treatments were applied PRE to the soil surface one or two days after seeding. Herbicide treatments were applied with a CO₂-pressurized backpack sprayer calibrated to deliver 200 L∙ha⁻¹ of spray solution at a pressure of 200/241 kPa using low drift nozzles (ULD120-02, Spraying Systems Co., P.O. Box 7900. Wheaton, IL 60188). The boom was 2.5 m wide with six nozzles spaced 0.5 m apart. All plots including the non-treated control were kept weed-free by inter-row cultivation and hand weeding as required.

Estimates of crop injury were evaluated 1, 2, and 4 weeks after emergence (WAE) using a scale of 0% to 100%, where a rating of 0 was no plant injury and a rating of 100 was plant death. At 3 WAE, a 1 m section of row for each plot was hand harvested at the ground level, oven dried and the shoot dry weight was recorded. Plant stand was also counted at 3 WAE. At 6 WAE, ten plants from each plot were randomly selected and the height from the soil surface to the highest growing point was measured with a measuring stick and recorded as plant height. Yield was measured at maturity for each crop by combining the remaining 9 m from each plot at Exeter and 7 m from each plot at Ridgetown. Seed moisture content was adjusted to 18% for black, white and kidney bean and 13% for adzuki bean.

Data were analyzed as a 2-way factorial using PROC MIXED in SAS 9.2. Fixed effects included dry bean market class and herbicide treatment, as well as their interaction; random effects included year-location combi- nations (environment), interactions between environment and the fixed effects, and replicate nested within envi- ronment. Significance of fixed effects were tested using F-tests and random effects were tested using a Z-test of the variance estimate. The UNIVARIATE procedure was used to test data for normality and homogeneity of variance. To satisfy the assumptions of the variance analyses, injury 1, 2 and 4 WAE were arcsine square root transformed, percent plant stand, percent dry weight and percent yield were square root transformed and seed moisture content at harvest was log-transformed. For all injury ratings, the untreated check (assigned a value of zero) was excluded from the analysis. However, all values were compared independently to zero to evaluate treatment differences with the untreated check. Plant stand, dry weight, height and yield were converted to a percent of the untreated check for analysis. Treatment comparisons were made using Fisher’s Protected LSD at a level of P < 0.05 and any data compared on the transformed scale were converted back to the original scale for presentation of results.

Statistical analysis indicated that for main effects of market class was significant for seed moisture content (Ta- ble 1). Also, herbicide treatment was significant for injury 1, 2 and 4 WAE, plant stand, dry weight, height, seed moisture content and yield (Table 1). Market class by treatment interaction was also significant for plant stand, dry weight, height and yield (Table 1).

Based on these results, saflufenacil and dimethenamid-p/saflufenacil applied PRE at the proposed rate of 25 g ai ha⁻¹ can be safely used for weed management in adzuki bean. There is also potential for use of saflufenacil and dimethenamid-p/saflufenacil applied PRE at the 25 g ai ha⁻¹ in kidney bean. However, saflufenacil and dime- thenamid-p/saflufenacil applied PRE at the proposed rates cause substantial injury and a reduction in plant stand, height, shoot dry weight and seed yield in black and white bean. Results showed that black and white bean are the most sensitive to saflufenacil and dimethenamid-p/saflufenacil followed by kidney bean and then adzuki bean.

Dimethenamid-p/saflufenacil with its novel mode of action would be a desirable compliment to the current weed management programs for PRE control of troublesome broadleaf weeds in adzuki bean. More research is needed to evaluate other market classes of dry bean to saflufenacil and dimethenamid-p/saflufenacil.

The authors would like to acknowledge Todd Cowan for his expertise and technical assistance in these studies. Funding for this project was provided in part by BASF and the GF2 program of the Agricultural Adaptation Council.