Evaluation of Bromoxynil, Pyraflufen

Evaluation of Bromoxynil, Pyraflufen-ethyl/2,4-D, and Tiafenacil for the Control of Glyphosate-resistant Canada fleabane (Conyza canadensis) in Soybean (Glycine max) and Metribuzin for the Control of Waterhemp (Amaranthus tuberculatus) with Two Mechanisms of Resistance to Photosystem II-inhibiting Herbicides

David Bernard Westerveld

A Thesis presented to The University of Guelph

In partial fulfilment of requirements for the degree of Master of Science in Plant Agriculture

Guelph, Ontario, Canada

EVALUATION OF BROMOXYNIL, PYRAFLUFEN-ETHYL/2,4-D, AND TIAFENACIL FOR THE CONTROL OF GLYPHOSATE-RESISTANT CANADA FLEABANE (Conyza canadensis) IN SOYBEAN (Glycine max) AND METRIBUZIN FOR THE CONTROL OF WATERHEMP (Amaranthus tuberculatus) WITH TWO MECHANISMS OF RESISTANCE TO PHOTOSYSTEM II-INHIBITING HERBICIDES

David Bernard Westerveld Advisor: University of Guelph, 2021 Dr. P. H. Sikkema

Fifteen field experiments were conducted in 2019 and 2020 to evaluate bromoxynil, pyraflufen-ethyl/2,4-D, and tiafenacil, alone and tankmixed with metribuzin, for the control of glyphosate-resistant (GR) Canada fleabane applied preplant in soybean. The biologically- effective-dose (BED) of bromoxynil +/- metribuzin, pyraflufen-ethyl/2,4-D +/- metribuzin, and tiafenacil +/- metribuzin was determined. Bromoxynil, pyraflufen-ethyl/2,4-D, and tiafenacil, tankmixed with metribuzin, provided similar control to the current industry standards at 8 weeks after application (WAA). Ten field and eight greenhouse experiments were conducted to evaluate metribuzin, applied preemergence (PRE) and postemergence (POST), for the control of photosystem II (PS II)-inhibitor resistant waterhemp conferred by target-site resistance (TSR) and non-target-site resistance (NTSR) mechanisms. In both the field and greenhouse studies

NTSR PS II-inhibitor resistant waterhemp was controlled with metribuzin applied PRE and

POST, in contrast TSR PS II-inhibitor resistant waterhemp was not controlled with metribuzin.

Acknowledgements

Over the past two years I have have had the privilege of working and learning from many great people. I would like to acknowledge my advisor Dr. Peter Sikkema for his guidance, enthusiasm, and unwavering support. Thank you for your encouragement and feedback on both my writing and research. I would also like to acknowledge Dr. Darren Robinson and Dr. David Hooker for their assistance during my studies as committee members. Your input and thought- provoking questions, both in the field and during the writing process, have been tremendously helpful. Thank you both for your valued input on my thesis and for your overall contributions to my education. I would like to awknowledge the many people who assisted me in completing this research. To Chris Kramer, thank you for all your technical assistance with the field and greenhouse experiments. Your expertise and hardwork is appreciated. To Christy Shropshire, thank you for your technical assistance with writing and for all your help with statistical analysis. To Lynette Brown, thank you for encouragement and inclusion of all graduate students. To Dr. Michelle Edwards, thank you for your assistance with statistics, your wiliness to teach and share is greatly appreciated. To Dr. Nader Soltani, thank you for all your work preparing manuscripts for publication and your encouragement of graduate students. To Dr. Francois Tardif, I would like to thank you for your assistance with the growth room experiments in Guelph and for your dedication to the agriculture students in the undergraduate and graduate programs. To fellow graduate students: Jessica Quinn, Nicole Langdon, Christian Willemse, Meghan Dilliot, and John Fluttert, thank you for your assistance and friendship over the past two years. To summer students Nicole and Alyssa, thank you for all your hard work during the field season. To Nigel Buffone, thank you for all the discussions and advice during my studies. Thank you to the farmers who allowed use of their land for these research trials. Thank you to the Grain Farmers of Ontario, the Ontario Agri-Food Innovation Alliance, BASF, Bayer CropScience, Corteva, FMC, Nufarm, Syngenta, and Valent for funding this research. To my family and friends, I would like to thank you for all your support and encouragement. To Kate, thank you for supporting and helping me countless times over the past two years. To Landon, thank you for being a great friend during university and graduate school. To Oma, thank you for your support and for always being willing to have me over for a visit. I would also like to thank my siblings, Micah and Eryn, John, and Michael for their support and interest in my research. Finally, I would like to acknowledge and thank my Mom and Dad, Andrew and Linda, for their motivation and support during my studies.

iii

Table of Contents

ABSTRACT...... ii Acknowledgements ...... iii Table of Contents ...... iv List of Tables ...... viii Chapter 1: Literature Review ...... 1 Section 1.0 – Waterhemp ...... 1 1.1 Botanical classification ...... 1 1.2 Morphology and Identification ...... 2 1.3 Distribution and Habitat ...... 3 1.4 Germination and Emergence ...... 3 1.5 Vegetative Growth and Competition ...... 5 1.6 Reproduction...... 6 1.7 Herbicide Resistance ...... 7 Section 2.0 - Canada Fleabane ...... 8 2.1 Introduction and Botanical information ...... 8 2.2 Morphology and Identification ...... 8 2.3 Distribution and Habitat ...... 9 2.4 Germination and Emergence ...... 10 2.5 Reproduction and Dispersal ...... 12 2.6 Herbicide Resistance ...... 12 2.7 Impact ...... 13 2.8 Control ...... 14 Section 3.0 - Weed Control in Soybean ...... 17 3.1 Soybean Production in Canada and Ontario ...... 17 3.2 Soybean Yield loss due to Weed Interference ...... 18 3.3 Other Benefits of Weed Control in Soybean ...... 19 3.4 Critical Weed-Free Period in Soybean ...... 20 3.5 Integrated Weed Management ...... 21 3.6 Cultural Weed Control ...... 21 3.7 Mechanical Weed Control ...... 23 3.8 Biological Weed Control ...... 24 iv

3.9 Chemical Weed Control ...... 25 3.10 Herbicide-Resistant Crops ...... 26 Section 4.0 - Herbicide-Resistant Weeds ...... 27 4.1 Introduction into Herbicide Resistance ...... 27 4.2 Mechanisms of Resistance ...... 28 4.3 Target-Site Resistance ...... 29 4.4 Non-Target-Site Resistance ...... 29 4.5 Glyphosate Resistance ...... 31 4.6 Photosystem II-Inhibitor Resistance ...... 33 4.7 Glyphosate-Resistant Canada Fleabane ...... 35 4.8 Metribuzin-Resistant Waterhemp ...... 36 4.9 Cost and Management of Herbicide Resistance Weeds ...... 37 Chapter 5.0 - Glyphosate, Bromoxynil, Pyraflufen-ethyl/2,4-D, Metribuzin, and Tiafenacil ...... 39 5.1 Glyphosate ...... 39 5.2 Bromoxynil ...... 40 5.3 Pyraflufen-ethyl ...... 43 5.4 Metribuzin...... 45 5.5 Tiafenacil ...... 48 Section 6.0 – Hypotheses and Objectives ...... 50 6.1 Hypotheses ...... 50 6.2 Objectives ...... 50 Chapter 2: Biologically-Effective-Dose of Bromoxynil, Applied Alone and Mixed with Metribuzin, for the Control of Glyphosate-Resistant Canada fleabane [Conyza canadensis (L.) Cronq.] in Soybean ...... 53 2.1 Abstract ...... 53 2.2 Introduction...... 54 2.3 Materials and Methods...... 57 2.4 Results and Discussion ...... 60 2.4.1 Soybean Injury ...... 60 2.4.2 Biologically-Effective-Dose of Bromoxynil and Bromoxynil + Metribuzin for the Control of GR Canada fleabane ...... 60 2.4.3 Bromoxynil, Metribuzin, Bromoxynil + Metribuzin Compared to Industry Standards ...... 63 2.5 Conclusion ...... 65

Chapter 3: Biologically-Effective-Dose of Pyraflufen-ethyl/2,4-D, Applied Preplant Alone or Tankmixed with Metribuzin for Glyphosate-Resistant Canada Fleabane [Conyza canadensis (L.) Cronq.] Control in Soybean ...... 71 3.1 Abstract ...... 71 3.2 Introduction...... 72 3.3 Materials and Methods...... 75 3.4. Results and Discussion ...... 80 3.4.1 Soybean Injury ...... 80 3.4.2 Biologically-Effective-Dose of Pyraflufen-ethyl/2,4-D and Pyraflufen-ethyl/2,4-D + Metribuzin for the control of GR Canada Fleabane ...... 80 3.4.3 Pyraflufen-ethyl/2,4-D, Metribuzin, Pyraflufen-ethyl/2,4-D + Metribuzin compared to Industry Standards...... 83 3.4.4 Interaction – Pyraflufen-ethyl/2,4-D, Metribuzin compared to Pyraflufen-ethyl/2,4-D + Metribuzin...... 84 3.5.1 Conclusion ...... 85 Chapter 4: Biologically-Effective-Dose of Tiafenacil Applied Preplant Alone or Tankmixed with Metribuzin for Glyphosate-Resistant Canada Fleabane [Conyza canadensis (L.) Cronq.] Control in Soybean ...... 91 4.1 Abstract ...... 91 4.2 Introduction...... 92 4.3 Materials and Methods...... 94 4.4. Results and Discussion ...... 99 4.4.1 Soybean Injury ...... 99 4.4.2 Biologically-Effective-Dose of Tiafenacil and Tiafenacil + Metribuzin for the control of GR Canada fleabane ...... 99 4.4.3 Tiafenacil, Metribuzin, Tiafenacil + Metribuzin compared to Industry Standards ...... 102 4.4.4 Interaction – Tiafenacil, Metribuzin compared to Tiafenacil + Metribuzin ...... 103 4.5.1 Conclusion ...... 103 Chapter 5: Biologically-Effective-Dose of Metribuzin, Applied Preemergence and Postemergence, for the Control of Waterhemp (Amaranthus tuberculatus) with Different Mechanisms of Resistance to Photosystem II-inhibiting herbicides ...... 110 5.1 Abstract ...... 110 5.2 Introduction...... 111 5.3 Materials and Methods...... 114 5.4. Results and Discussion ...... 120

5.4.1 Biologically-Effective-Dose of Metribuzin applied Preemergence for the control of Metribuzin- Sensitive and Resistant Waterhemp ...... 120 5.4.2 Biologically-Effective-Dose of Metribuzin applied Postemergence for the control of Metribuzin-Sensitive and Resistant Waterhemp ...... 121 5.4.3 Metribuzin-Sensitive and Resistant Waterhemp control with Metribuzin Compared to Industry Standards...... 122 5.4.4.1 Biologically-Effective-Dose of Metribuzin applied Preemergence for control of Metribuzin- Sensitive and Resistant Waterhemp ...... 124 5.4.4.2 Biologically-Effective-Dose of Metribuzin applied Postemergence for the control of Metribuzin-Sensitive and Resistant Waterhemp ...... 125 5.4.4.3 Metribuzin-Sensitive and Resistant Waterhemp control compared to Industry Standards ... 125 5.5.1 Conclusion ...... 126 Chapter 6: General Discussion ...... 138 6.1 Contributions ...... 138 6.2 Limitations...... 140 6.3 Future Research ...... 141 Chapter 7: Literature Cited ...... 145 Chapter 8: Appendix ...... 170 8.1 Determining Non-linear Regression Models ...... 170 8.2 SAS Code for Chapter 2 ...... 171 8.2 SAS Code for Chapter 3 ...... 174 8.3 SAS Code for Chapter 4 ...... 180 8.4 SAS Code for Chapter 5 ...... 186

vii

List of Tables

Table 2.1 Year, location, application information, crop information and soil characteristics for five field experiments conducted on the biologically-effective-dose of bromoxynil, applied alone and mixed with metribuzin, for the control of glyphosate-resistant Canada fleabane [Conyza canadensis (L.) Cronq.] in soybean in 2019 and 2020 in Ontario, Canada...... 66

Table 2.2 Regression parameters and predicted effective dose of bromoxynil and bromoxynil plus metribuzin for 50, 80 and 95% glyphosate-resistant Canada fleabane [Conyza canadensis (L.) Cronq.] control at 2, 4 and 8 weeks after application (WAA), for 50, 80 and 95% reduction in biomass and plant density at 8 WAA and to achieve 50, 80 and 95% of the yield from five field experiments completed in 2019 and 2020 in Ontario, Canada...... 67

Table 2.3 Glyphosate-resistant Canada fleabane [Conyza canadensis (L.) Cronq.] control at 2, 4 and 8 weeks after application (WAA) and reduction in plant density and dry biomass at 8 WAA from five field experiments conducted in Ontario, Canada in 2019 and 2020...... 68

Table 2.4 Glyphosate-resistant Canada fleabane [Conyza canadensis (L.) Cronq.] control at 2, 4 and 8 weeks after application (WAA) and plant density and dry biomass at 8 WAA with metribuzin, bromoxynil and bromoxynil + metribuzin applied preplant from five field experiments conducted in 2019 and 2020 in Ontario, Canada...... 69

Table 3.2 Parameter estimates and calculated effective doses of pyraflufen-ethyl/2,4-D and pyraflufen-ethyl/2,4-D plus metribuzin for 50, 80, and 95% glyphosate-resistant Canada fleabane [Conyza canadensis (L.) Cronq.] control at 2, 4, and 8 weeks after application (WAA), for 50, 80, and 95% reduction in biomass and density at 8 WAA and to achieve 50, 80, and 95% of yield from five experiments completed in 2019 and 2020 in Ontario, Canada...... 87

Table 3.3 Glyphosate-resistant Canada fleabane [Conyza canadensis (L.) Cronq.] control at 2, 4, and 8 weeks after application (WAA), density and dry biomass at 8 WAA, and soybean yield from five field experiments conducted in Ontario, Canada in 2019 and 2020...... 88

viii

Table 4.2 Parameter estimates and calculated effective dose of tiafenacil applied preplant for glyphosate-resistant Canada fleabane [Conyza canadensis (L.) Cronq.] control, reduction in biomass and plant density, and yield from five trials completed in 2019 and 2020 in Ontario, Canada...... 105

Table 4.3 Parameter estimates and calculated effective dose of tiafenacil + metribuzin applied preplant for glyphosate-resistant Canada fleabane [Conyza canadensis (L.) Cronq.] control, reduction in biomass and plant density, and yield from five trials completed in 2019 and 2020 in Ontario, Canada...... 106

Table 4.4 Glyphosate-resistant Canada fleabane [Conyza canadensis (L.) Cronq.] control at 2, 4, and 8 weeks after application (WAA), density and dry biomass at 8 WAA, and soybean yield from five field trials conducted in Ontario, Canada in 2019 and 2020...... 107

Table 5.5 Metribuzin-sensitive and resistant waterhemp (Amaranthus tuberculatus) control, and density and dry biomass reduction at 8 WAA for herbicides applied preemergence from ten field experiments conducted in Ontario, Canada in 2019 and 2020...... 132

Table 5.6 Metribuzin-sensitive and resistant waterhemp (Amaranthus tuberculatus) control, and density and dry biomass reduction at 8 WAA for herbicides applied postemergence from ten field experiments conducted in Ontario, Canada in 2019 and 2020...... 133

Table 5.7 Parameter estimates and calculated effective doses of metribuzin applied preemergence on control and fresh weight reduction of metribuzin-sensitive and resistant waterhemp (Amaranthus tuberculatus) biotypes from four greenhouse experiments completed in 2020 in Ontario, Canada...... 134

Table 5.8 Parameter estimates and calculated effective doses of metribuzin applied postemergence on control and fresh weight reduction of metribuzin-sensitive and resistant waterhemp (Amaranthus tuberculatus) biotypes from four greenhouse experiments completed in 2020 in Ontario, Canada...... 135

Table 5.9 Metribuzin-sensitive and resistant waterhemp (Amaranthus tuberculatus) control 1, 3 and 5 weeks after application (WAA) and fresh weight 5 WAA, with herbicide applied preemergence from four greenhouse experiments conducted in Ontario, Canada in 2020...... 136

Table 5.10 Metribuzin-sensitive and resistant waterhemp (Amaranthus tuberculatus) control at 1, 3 and 5 weeks after application (WAA) and fresh weight 5 WAA with herbicide applied postemergence from four greenhouse experiments conducted in Ontario, Canada in 2020...... 137

Chapter 1: Literature Review

Section 1.0 – Waterhemp

1.1 Botanical classification

Waterhemp, Amaranthus tuberculatus (Moq.) J.D. Sauer, is a small seeded, summer annual broadleaf species that is part of the genus Amaranthus along with 70 other species worldwide

(Costea et al. 2005). Other members of the Amaranthaceae family include Palmer amaranth

(Amaranthus palmeri S. Watson), green pigweed (Amaranthus powellii S. Watson), redroot pigweed (Amaranthus retroflexus L.), and smooth pigweed (Amaranthus hybridus L.). Two subspecies of waterhemp, A. tuberculatus var. tuberculatus and A. tuberculatus var. rudis, are present in Ontario (Costea and Tardif 2003; Costea et al. 2005).

The botanical classification of waterhemp has changed over time. Uline and Bray (1895) proposed two varieties under one polymorphous species. By the mid-20th century a species separation was proposed with A. tuberculatus being the plant found east of the Mississippi River with indehiscent utricles and no pistillate tepals, and A. rudis common to areas west of the

Mississippi River with the dehiscent‐fruited taxon with one pistillate tepal (Sauer 1955; Sauer

1957; Sauer 1972). Pratt and Clark’s (2001) taxonomic study re-iterated the idea of a single polymorphous species with a degree of high variability over different geographic areas. Most recently, Costea and Tardif (2003) supported the concept of waterhemp as a single species, but supported recognition at a varietal level. Although there is still confusion within the scientific community, both tall (A. tuberculatus var. tuberculatus) and common (A. tuberculatus var. rudis) waterhemp are currently referred to by the common name ‘waterhemp’ according to the Weed

Science Society of America (WSSA) in 2019. 1

1.2 Morphology and Identification

Waterhemp is similar in size and appearance to the other members of the Amaranth family. The hypocotyl is approximately 2.5 cm long and glabrous (Costea et al. 2005). Stems of a mature waterhemp plant are erect with limited to no hair. The hairless stems are either round or ridged and green or pinkish red in colour. Leaves are hairless with a somewhat shiny surface

(Horak et al. 1994). Leaves are lanceolate, ovate, or oblong in shape with smooth margins

(Costea et al. 2005). Petioles are long and narrow, in most cases shorter than the leaves.

Waterhemp can grow greater than 3 m in height under favourable environmental conditions

Waterhemp is a dioecious species; the male and female flowers that are found on separate plants. Each plant has either staminate (male) or pistillate flowers (female) (Costea et al. 2005).

Bracts of male flowers are 1.5 to 2 mm long consisting of 5 sepals, 5 stamens, and no petals; it is surrounded by 1 to 3 narrow bracts with pointed tips. Female flowers are 1.5 to 2.5 mm long, have 1 or 2 tepals, usually 1 sepal, an ovary, and no petals. Waterhemp can flower from late summer to early fall; flowering lasts about 1 to 2 months. Waterhemp seeds are about 1.5 mm long, dark reddish-brown in colour, and elliptic to obovate in shape.

In the early stages of waterhemp development differentiation from other Amaranthus species is difficult. Waterhemp is distinguishable from other Amaranthus species since its cotyledons that are more egg shaped. The leaves can be a useful indicator of species. Waterhemp leaves are long and narrow (lanceolate) and appear to have a waxy surface (Costea et al. 2005).

Stem and leaf surfaces are smooth for the entirety of the plant’s life (Horak et al. 1994). The presence of male and female flowers on separate plants is another characteristic of waterhemp, whereas pigweeds such as redroot, green and smooth pigweed are monecious. Within A.

2 tuberculatus, var. rudis can be distinguished from var. tuberculatus by the manner in which the seed capsule fractures when threshed with A. tuberculatus var. rudis breaking apart into two cup- like sections, while var. tumberculatus breaks irregularly.

1.3 Distribution and Habitat

The distribution of waterhemp has expanded in recent decades. Waterhemp is native to the Great Plains region of the United States. Amaranthus tuberculatus is native to North America

(Sauer 1955). Prior to the 21st century, only Amaranthus tuberculatus var. tuberculatus was identified in Canada near freshwater biomes, and was not reported as a problem in agricultural fields (Costea et al. 2005). Within the United States, A. tuberculatus var. rudis occurs in 35 states, but was mostly absent east of the Mississippi River Basin (Sauer 1972). Amaranthus tuberculatus var. rudis was first found in fields in Essex, Huron, and Lambton counties in southwestern Ontario in 2002 and 2003 (Costea et al. 2005).

Waterhemp has the ability to thrive under a range of habitats and growing conditions.

Waterhemp is a C4 plant that exhibits rapid growth (Elmore and Paul 1983). Waterhemp is a thermophyte, thriving at elevated temperatures, and is also considered a mesophyte to hygrophyte, preferentially growing in areas with medium to high amounts of water (Costea et al.

2005). Although waterhemp prefers nutrient rich soils, it is adapted to a broad range of soil types and textures with a pH range from 4.5 to 8.

1.4 Germination and Emergence

Waterhemp requires dormancy for successful germination. Dormancy allows a portion of the seeds to remain viable in the soil for several years resulting in extended weed seed bank life

(Burnside et al. 1996). Germination was observed by Leon et al. (2004) after 12 weeks of 3 exposure to temperatures of 4°C. Waterhemp seeds that were stratified had 4 times greater germination than non-chilled seeds (Leon and Owen 2003). Successful germination can also be influenced by light quality; a higher red to far-red light ratio results in greater germination.

Overall, waterhemp has relatively broad environmental requirements for germination allowing it to emerge throughout the growing season in Ontario.

Waterhemp germination is affected by temperature. The minimum temperature for germination is 10°C (Steckel et al. 2004; Leon et al. 2004) to 12°C (Steckel et al. 2007) with an optimal temperature range of 25 to 35°C, after which germination begins to decrease (Guo and

Al-Khatib 2003). Leon et al. (2004) found there was no specific optimal germination temperature but rather germination was decreased as the amplitude of the temperature alternations increased.

In Ontario, waterhemp predominantly emerges from May to August but can emerge as late as

October (Vyn et al. 2007; Schryver et al. 2017b).

Annual emergence of common waterhemp seed has been reported to rarely exceed 7% of the initial seedbank, with cumulative emergence averaging approximately 15% during a 3- to 4- year period (Hartzler et al. 1999; Buhler and Hartzler 2001). Four years after burial, only 12% of the initial seedbank was recovered from the soil (Buhler and Hartzler 2001). Burnside et al.

(1996) reported that seeds can remain dormant for up to 17 years; conversely, Steckel et al.

(2007) found that the waterhemp seedbank declined to less than 1% of the original amount after

4 years.

Waterhemp is well adapted to reduced-tilled systems. Leon and Owen (2006) reported that waterhemp germination in no-till soils was four times greater compared to soil that had been chisel or moldboard plowed. Steckel et al. (2007) found greater initial germination of waterhemp

4 in no-till systems and greater seed persistence in tilled soils. Refsell and Hartzler (2009) concluded that the concentration of waterhemp seed near the soil surface resulted in overall greater emergence in reduced-till systems.

1.5 Vegetative Growth and Competition

Waterhemp’s rapid growth, and its’ ability to survive and reproduce in diverse environments has made it a problematic weed in a wide geographical area. Waterhemp can grow as much as

0.16 cm per growing degree day, contributing to its competitiveness for resources, resulting in up to 2.5 cm of growth per day (Horak and Loughin 2000). Final height of waterhemp is not impacted when they are grown in up to 68% shade, regardless of emergence date (Steckel et al.

2003). However, Steckel and Sprague (2004a) found that decreasing soybean row width from 76 to 19 cm decreased waterhemp biomass, density, seed production, and soybean yield loss. The density of waterhemp plants that emerged in wide-row soybean after the V2 growth stage was twice that of narrow-row soybean.

Waterhemp is a concern for growers within Ontario and the United States. If uncontrolled, waterhemp can reduce corn yields up to 74% (Steckel and Sprague 2004b). In soybean, Bensch et al. (2003) reported a 56% yield loss due to season-long waterhemp interference while Hager et al. (2002) reported a 43% yield decrease when interference occurred up to ten weeks after soybean unifoliate leaf expansion. Steckel and Sprague (2004a) reported that soybean yield was reduced due to waterhemp that emerged at V4-V5 soybean for both row widths; however, these reductions were less than 10%. Overall, waterhemp control is critical to minimize crop yield loss to weed interference.

1.6 Reproduction

Waterhemp is a dioecious plant, making cross-pollination necessary for reproduction.

Cross-pollination has shown to greatly increase the genetic diversity of a population, often resulting in a wide range of physical and biological characteristics (Steckel. 2007; Sarangi et al.

2017). Waterhemp pollen can stay viable for up to 5 days after release and can fertilize a female plant as far as 800 meters from the pollen source, but most commonly fertilizes plants within 25 meters of the male plant (Liu et al. 2012). It is possible for two different Amaranthus species, to produce fertile hybrids from cross-pollination. Trucco et al. (2005) found that successful hybridization between waterhemp and smooth pigweed could occur when waterhemp was the maternal parent.

The number of waterhemp seeds produced per plant varies greatly. Sellers et al. (2003) reported an average of 288 950 seeds plant-1 in Missouri. Steckel et al. (2003) reported mean seed production in excess of 1 million seeds plant-1 in Illinois, and Hartzler et al. (2004) reported

4.8 million seeds from one single plant in Iowa. Hartzler et al. (2004) also reported female waterhemp plants emerging with soybean produced an average of 300 000 seeds plant-1, while a single waterhemp plant that emerged 50 days after planting soybean produced 3000 seeds.

Waterhemp is also able to produce viable seed in low light environments, allowing it to reach full maturity even when emergence is delayed. While grown under 68% shade Steckel et al.

(2003) observed waterhemp to produce up to 400 000 seeds per plant. Seed longevity in the soil, and a tremendous seed bank of dormant and non-dormant weed seed, assures that farm managers will have weed issues year-after-year.

1.7 Herbicide Resistance

Herbicide resistance can be defined as “the inherited ability of a plant to survive and reproduce following exposure to a dose of herbicide normally lethal to the wild type. In a plant, resistance may be naturally occurring or induced by such techniques as genetic engineering or selection of variants produced by tissue culture or mutagenesis” (Vencill et al. 2012). The first reported case of herbicide resistance in waterhemp was triazine resistance in Fillmore County,

Nebraska in 1990 (Anderson et al. 1996). The first confirmation of glyphosate-resistant waterhemp occurred in Platte County, Missouri, in 2005 (Legleiter and Bradley 2008). This waterhemp biotype had a glyphosate resistance factor of 19 and was later confirmed to also be resistant to acetolactate synthase (ALS)- and protoporphyrinogen (PPO)-inhibiting herbicides.

Waterhemp biotypes resistant to multiple modes of action now exist in several combinations in different geographical locations. The first occurrence was in 1996 in a field with waterhemp resistant to both triazine and ALS-Inhibiting herbicides (Foes et al. 1998). Currently, waterhemp is resistant to seven modes of action including 4-hydroxyphenylpyruvate dioxygenase (HPPD) inhibitors, photosystem II (PS II)-inhibitors, acetolactate synthase (ALS)-inhibitors, protoporphyrinogen oxidase (PPO)-inhibitors, and 5-enolpyruvylskikimate-3-phosphate synthase

(EPSPS)-inhibitors, very long chain fatty acid (VLCFA)-elongases inhibitors, and the synthetic auxin herbicides such as 2,4-D (Heap 2020). Furthermore, some waterhemp populations are resistant to several modes of actions (MOA) being classified as multiple-resistant (MR) resulting in plants with resistance to herbicides from three, four, five, and six MOA groups (Bell et al.

2013; Evans 2016; Shergill et al. 2018). Since waterhemp is a dioecious, obligiate-outcrossing, wind pollinated plant, there is wide genetic variability resulting in high potential for resistance genes to be transferred among populations (Tranel et al. 2011). 7

Currently within Ontario, there are waterhemp biotypes that are resistant to four herbicide modes of action including ALS-, PS II-, EPSPS-, and PPO-inhibitors, representing WSAA

Herbicide Groups 2, 5, 9, and 14, respectively (Heap 2020). Resistance was first recorded to imazethapyr and atrazine in 2002 (Costea et al. 2005; Vyn et al. 2007), glyphosate and imazethapyr in 2014 (Schryver et al. 2017a), and lactofen in 2017 (Benoit et al. 2019). In 2017,

4-way multiple resistance was identified in Ontario with a waterhemp population waterhemp resistant to imazethapyr, atrazine, glyphosate, and lactofen.

Section 2.0 - Canada Fleabane

2.1 Introduction and Botanical information

Canada fleabane, Conyza canadensis (L.) Cronq., is a winter or summer annual that is a member of the Asteraceae, or Compositae family (Weaver 2001). Other common names include horseweed, coltstail, marestail, and butterweed. Canada fleabane is a cosmopolitan weed, meaning it is present on all continents except Antarctica. There are three other fleabane species found in Canada (Frankton and Mulligan 1987). Erigeron philadelphicus L. is a perennial herbaceous plant with alternate, simple leaves that clasp the stem, and pale pink or white flowers.

Erigeron annuus (L.) Pers. and Erigeron strigosus Muhl ex. Willd are annual or biennial plants with flowers that are white to purple in colour (Weaver 2001). Canada fleabane is readily distinguished from these species by its leaves, which are more numerous and narrower. The flowers of Canada fleabane are smaller and the inflorescence has a bushy appearance.

2.2 Morphology and Identification

Canada fleabane seedlings that emerge in the fall start as a basal rosette with sparsely hairy, dark green leaves that lack evident veins (Weaver 2001). Spring emerging Canada fleabane does not form a basal rosette; the stem elongates shortly after emergence. The stem is erect growing up to 180 cm in height. As the stem elongates the basal rosette diminishes (Frankton and

Mulligan 1987). Stem leaves are alternate, hairy, lanceolate to linear, crowded on the stem, with leaf margins being entire or toothed.

The composite flowers of Canada fleabane are small relative to other species. The inflorescence is an elongated panicle near the apex of the principal stem, made up of flower heads ranging in size from 1 to 5 mm (Weaver 2001; Bryson and DeFelice 2009). Seeds are yellowish to transparent, 1 to 2 mm in length, narrow towards the base, with an attached pappus that is 2 to 5 mm long (Frankton and Mulligan 1987).

2.3 Distribution and Habitat

Canada fleabane is a widespread problematic weed, most commonly found in temperate northern regions around the world. Canada fleabane is thought to be native to North America and it is thought that humans have spread it to many regions where it is currently found (Cronquist

1943; Frankton and Mulligan 1987). Its introduction into Western Europe from North America is believed to have occurred during the 17th century; it has since spread so extensively that it is now considered naturalized in Europe (Thebaud and Abbott 1995). Canada fleabane can be found in areas of agricultural production, meadows, gardens, woodland trails, roadsides, and waste places throughout the Americas, Europe, Australia, and Asia (Frankton and Mulligan 1987; Thebaud and Abbott 1995; Weaver 2001). Within Canada, it is found in agricultural fields, rangeland, and forests south of latitude N55 in all provinces except Newfoundland (Weaver 2001).

Canada fleabane’s adaptability makes it suitable for a wide array of habitats. It is commonly found in grasslands and in moist undisturbed sites including riparian buffers, ditches, and wetland areas (Tilley 2012). Canada fleabane thrives in a range of conditions and soil types, however Nandula et al. (2006) noted that it emerged faster and in larger quantities on coarse soils than compared to fine-textured soils. It grows the best on slightly acidic to neutral soils in areas receiving 410 to 1400 mm in annual precipitation (Weaver at al 2001; Tilley 2012). The plants produce large quantities of windborne seed, making it an effective colonizer (Regehr and Bazzaz

1979). The adaptability of this weed allowed it to become problematic across North America.

2.4 Germination and Emergence

Canada fleabane can emerge any time during the year when conditions are favourable; however, germination occurs predominantly in the spring and/or fall (Weaver 2001; Main et al.

2006). Buhler and Owen (1997) observed fall emergence of Canada fleabane in Minnesota and

Iowa at densities of 78 to 151 plants m-2; spring emergence also occurred but accounted for only

5 to 32% of the total population. Tozzi and Van Acker (2014) reported fall emergence to be greater compared to spring emergence in southern Ontario. Peak fall emergence occurred from

August 27 to September 9, with a density of 448 plants m-2. Comparatively, peak emergence in the spring occurred from May 14 to May 27 with 42 plants m-2 in the same study. Plants that germinate in the fall behave as winter annuals, overwintering as rosettes and flowering and producing seed the following summer (Bhowmik and Bekech 1993). Fall-emerged seedlings have lower survivability than spring-emerged seedings but have greater fecundity than spring emerged plants when growing in a mixed population (Regehr and Bazzaz 1979). Spring emerging Canada fleabane does not form a basal rosette, the stem elongates shortly after

10 emergence; they flower and produce seed within the same growing season (Davis and Johnson

2008).

Canada fleabane does not have dormancy requirements, which allows for germination and establishment after seed release from the mother plant (Buhler and Owen 1997; Nandula et al.

2006). Canada fleabane seeds can germinate immediately after seed-shed when there is a day/night temperature fluctuation of 22/16°C (Buhler and Owen 1997). Nandula et al. (2006) found a temperature fluctuation of 24/20°C resulted in the highest germination of Canada fleabane seeds; seeds did not germinate when subjected to temperatures of 12/6°C and germination decreased as temperatures increased to 36/30°C. Canada fleabane can germinate in complete darkness but overall germination increases with a 13-hour photoperiod. Canada fleabane germination is also influenced by soil pH. Nandula et al. (2006) found that a pH between 6 and 10 resulted in seed germination of 24 to 36%, and germination dropped to 19% when the pH was below 5.

Germination of Canada fleabane is greatest when seed is on the soil surface and no seedlings emerge when burial depth exceeds 0.5 cm (Nandula et al. 2006). The small seed size and low energy reserves contribute to Canada fleabane’s inability to emerge from greater depths. In a seedbank study, 80% of Canada fleabane seeds from the top 2 cm of soil in a no tillage system did germinate, but seed from depths greater than 6 cm failed to germinate (Bhowmik and Bekech

1993). Canada fleabane seed viability is short. In a laboratory experiment, seed remained viable for up to 3 years (Comes et al. 1978). However, viable seed has been discovered in the soil in long-term pasture and abandoned fields, despite Canada fleabane not being part of the existing flora (Tsuyuzaki and Kanda 1996; Leck and Leck 1998). Regehr and Bazzaz (1979) estimated annual seedbank recruitment to be around 7%.

2.5 Reproduction and Dispersal

Canada fleabane has high fecundity, which contributes to its dispersal potential. Canada fleabane’s only means of reproduction is by seed (Weaver 2001). Flowering occurs in July in

Canada, and plants set seed in August through September. Canada fleabane is primarily self- pollinated, however, Smisek (1995) found an average of 4% outcrossing in an Essex County

Canada fleabane population. Canada fleabane seeds can travel great distances due to their small size and attached parachute-like pappus. Regehr and Bazzaz (1979) recorded seeds being transported 122 m downwind in a corn field, and Dauer et al. (2007) showed that seed travels hundreds of meters with prevailing winds. Canada fleabane seed has been found in the planetary boundary layer, illustrating the potential for long distance seed dispersal of up to 550 km

(Shields et al. 2006). A single Canada fleabane plant can produce more than one million seeds per plant (Davis et al. 2009b) with seed production correlated with plant height. Weaver (2001) found that a 1.5-m tall plant can produce up to 230 000 seeds per plant. Other studies have found similar results with Canada fleabane producing between 31 000 and 72 000 seeds per plant when growing in a soybean field and up to 200 000 seeds per plant without a crop at a density of 10 plants m-2 (Bhowmik and Bekech 1993; Davis and Johnson 2008)

2.6 Herbicide Resistance

Canada fleabane has evolved resistance to several herbicide modes of action. Herbicide- resistant Canada fleabane is currently found in 18 countries worldwide (Heap 2020). Glyphosate- resistant (GR) Canada fleabane was first confirmed in Delaware, USA in 2000 (VanGessel

2001), and since then it has been detected in 13 countries including Brazil, Canada, China, Czech

Republic, Greece, Hungary, Italy, Japan, Poland, Portugal, South Korea, Spain, and the USA

(Heap 2020). In the USA, Canada fleabane has evolved resistance to the ALS-, PS II-, and

EPSPS-inhibitors and photosystem 1 diverters representing WSSA Groups 2, 5, 9 and 22, respectively.

In 2010, GR Canada fleabane was first confirmed at eight sites in Essex County, Ontario

(Byker et al. 2013a). In Canada, Canada fleabane has evolved resistance to cloransulam-methyl glyphosate, and paraquat representing WSSA groups 2, 9 and 22, respectively (Heap 2020).

Multiple-herbicide-resistant Canada fleabane is present in Canada with some biotypes resistant to both cloransulam-methyl and glyphosate.

The glyphosate resistance factor in Canada fleabane is biotype specific. Koger et al.

(2004) found a Mississippi biotype to have 8- to 12-fold resistance, VanGessel (2001) reported

8- to 13-fold resistance in Delaware, Mueller et al. (2003) found Canada fleabane to be resistant to 4 times the normal glyphosate application rate which is similar to Dinelli et al. (2006) who reported a glyphosate resistance factor of 4 to 5. Plant growth and seed production remain constant between sensitive and resistant biotypes, indicating that there is no fitness penalty associated with glyphosate resistance (Davis et al. 2009b).

2.7 Impact

Canada fleabane interference can substantially reduce the yield of agronomic crops. At densities of 100 to 200 plants m-2 Canada fleabane reduced soybean yields by up to 90% (Bruce and Kells 1990). Eubank et al. (2008) found a 62 to 97% soybean yield loss when comparing the untreated control to the most efficacious herbicide treatment. Budd et al. (2016a) and Hedges et al. (2018a), from studies completed in Ontario, reported that GR Canada fleabane interference reduced soybean yield by 73 and 67%, respectively. Byker et al. (2013b) found that GR Canada

13 fleabane interference reduced soybean yield by 83 to 93% when no weed management practices were implemented. In grain corn, Ford et al. (2014) and Metzger et al. (2019) reported that

Canada fleabane interference caused yield losses of up to 69 and 47%, respectively. In winter wheat there was no yield loss due to GR Canada fleabane interference (Quinn et al. 2020). The impact of Canada fleabane interference on crop yield can be substantial, consequentilly control measures must be implemented to maximize yield and profit.

2.8 Control

Canada fleabane is more common in reduced- and no-till systems; as these crop management systems increased in Ontario so did the prevalence of Canada fleabane. In a 3-year study, Brown and Whitwell (1998) observed that tillage prior to seeding a fall crop or spring disking to prepare the seedbed reduced Canada fleabane populations to minimal levels. In

Illinois, Kapusta (1979) compared weed control systems for no-, minimum- and conventional-till soybean; Canada fleabane was present in no-till plots, whereas there was no Canada fleabane in the minimum- and conventional-till systems. Barnes et al. (2004) surveyed 389 fields in Indiana and reported that Canada fleabane was present in 8 and 61% of conventional- and no-till tillage fields, respectively. Overall no- or minimal-tillage practices provide a more conducive environment for Canada fleabane germination, emergence and growth (Brown and Whitwell.

1988; Nandula et al. 2006). Utilization of tillage in the future may be one component of a successful integrated Canada fleabane management program

Crop rotation can have an impact on the germination, emergence, and survivability of

Canada fleabane. The first GR Canada fleabane population was documented in a continuous soybean rotation field (VanGessel 2001). Survey data from Indiana suggested that Canada fleabane is primarily an issue in a continuous soybean rotation (Davis et al. 2008). Davis et al. 14

(2009a) found that crop rotation influenced Canada fleabane density with a soybean-corn rotation consistently having lower density than a continuous soybean rotation. There are discrepancies in the peer-reviewed literature as some studies reported that crop rotation did not impact Canada fleabane density (Davis et al. 2007).

Cover crops can reduce GR Canada fleabane population. Winter annual cover crop mixtures and monocultures seeded in the late summer/early fall have reduced GR Canada fleabane in the subsequent crop grown the following growing season (Cholette et al. 2018;

Pittman et al. 2019). Cereal rye (Secale cereale L.) and cereal rye mixtures have been documented to suppress GR Canada fleabane (Pittman et al. 2019; Wallace et al. 2019).

Przepiorkowski and Gorski (1994), based on a greenhouse study, found that germination of

Canada fleabane was reduced 50% by aqueous extracts of rye shoot tissue and by soil containing rye roots. Essman et al. (2020) found that planting cereal rye at 50 kg ha-1 reduced GR Canada fleabane density, but should not be the sole means of control prior to soybean planting. Cover crops are one component of a diversified, integrated GR Canada fleabane management program.

Herbicides are the primary strategy for the control of GR Canada fleabane, but several herbicides do not provide satisfactory control. Initial research found that a single application of imazethapyr, imazaquin, linuron, metribuzin and paraquat provide variable GR Canada fleabane control (Bruce and Kells 1990; Moseley and Hagood 1990). Similarly, glufosinate, applied preplant, provided controlled GR Canada fleabane 33 and 88% in studies conducted by Owen et al. (2009) and Eubank et al. (2008), respectively. The increasing number of cases of herbicide resistance demands that growers diversify their GR Canada fleabane control programs.

Although some herbicides provide inconsistent or poor GR Canada fleabane control, there are some efficacious herbicide options. Co-application of burndown plus residual 15 herbicides, applied in the spring, were found to be most effective for GR Canada fleabane control since a percentage of the populations emerged in the spring (Davis and Johnson 2008).

Although fall-applied herbicides can provide excellent control of the fall-emerging GR Canada fleabane cohort, spring-applied herbicides are still required to control the spring-emerging cohort

(Owen et al. 2009; Davis et al. 2010). Davis et al. (2010) found that fall-applied chlorimuron + tribenuron provided greater than 90% control of spring emerged Canada fleabane. There are few effective herbicide options for the control of multiple-herbicide-resistant (Group 2 and 9) Canada fleabane in soybean. Prior to the evolution of GR Canada fleabane, glyphosate provided excellent control of sensitive Canada fleabane (Brown and Whitwell 1988; Bruce and Kells

1990; VanGessel 2001). Kruger et al. (2010) reported that synthetic auxin herbicides, 2,4-D and dicamba, controlled Canada fleabane that was greater than 30 cm tall 81 and 97%, respectively;

2,4-D (amine and ester) and dicamba (diglycolamine and dimethylamine salt) provided ≥90%

GR Canada fleabane control that was ≤30 cm in height. Eight weeks after application, Byker et al. (2013b) found that dicamba (600 g ae ha-1), applied preplant, controlled GR Canada fleabane

90 to 100%, Everitt and Keeling (2007) reported similar results, dicamba at 140 and 280 g ae ha-

1 controlled GR Canada fleabane 93 and 98%, respectively. These herbicides could play a role in

GR Canada fleabane control with the recent release of 2,4-D- and dicamba-resistant soybean traits. Saflufenacil and chlorimuron + tribenuron, applied in late spring, reduced Canada fleabane density ≥ 97 and 92%, respectively (Davis et al. 2010). Although an effective herbicide, salflufenacil (25 g ai ha-1) provides variable control of GR Canada fleabane with 37% (Ikley

2010) to 100% control (Byker et al. 2013b). The addition of metribuzin (400 g ai ha-1) to salflufenacil (25 g ai ha-1) has improved the level and consistency of GR Canada fleabane control

16 to 96 to 99% (Budd et al. 2016) and 86 to 92% control (Soltani et al. 2016). There are effective herbicide options for the control of GR Canada fleabane.

Section 3.0 - Weed Control in Soybean

3.1 Soybean Production in Canada and Ontario

Soybean (Glycine max (L.) Merr) is a nitrogen fixing crop that is a member of the

Leguminosae family utilized globally for oil and protein production. The soybean plant is an annual, erect, bushy plant that can grow up to 1.5 m in height (CFIA 1996). The primary leaves are unifoliate, positioned opposite from each other and are ovate in shape (Fehr 1980). The secondary leaves are trifoliate, alternate and ovate in shape. The stems and branches are commonly covered with trichomes, but there are glabrous types. Soybean plants have a long taproot from which a lateral root system arises. The pods vary in length from 2 to 7 cm and are straight or slightly curved (CFIA 1996). The seed is oval and tan to yellow which varies among cultivars (Fehr 1980).

Soybean has a long history as part of agricultural production in Canada. Breeding efforts have been recorded as far back as the early 1920’s at the Ontario Agricultural College and the

Harrow and Ottawa Research Stations (Bernard et al. 1998). Soybean production in Canada has been estimated to have increased from 63 000 hectares in 1951 to 1.2 million hectares in 2006, and 2.3 million hectares in 2019 (Dorff 2007; Cober and Voldeng 2012; Statistics Canada

2019a). In 2019, it is estimated that Canadian farmers produced 6.2 million tonnes of soybean

(Statistics Canada 2019b).

Soybean is a major field crop produced worldwide. Hartman et al. (2011) reported that soybean is grown on an estimated 6% of the world’s arable land and since the 1970s its 17 percentage of area of production has increased the most of the major field crops. Currently, the major producers of soybean in the world include the USA, Brazil, Argentina, China, and India with more than 92% of the world’s soybean production (Rodríguez-Navarro et al. 2011; Pagano and Miransari 2016). Canada ranks seventh in global soybean production.

Soybean has many food and industrial uses. Soybean composition is approximately 40% protein and 20% oil and is desired as it has less saturated fat in comparison to palm oil

(Goldsmith 2008). Soybean seed is utilized as a protein source for livestock nutrition and as an edible vegetable oil for human consumption. It is estimated that 2% of soybean production is consumed by humans directly as food with the remainder being pressed into soybean meal and oil for animal feed or industrial uses. In Canada, food products produced from soybean include infant formula, soya sauce, tofu, and simulated meat and milk products (Hartman et al. 2011).

Margarines, cooking oils, salad oils, and shortenings are made from purified soybean oil (CFIA

1996). Industrial uses of soybean include yeasts, soaps, plastics, and disinfectants.

3.2 Soybean Yield loss due to Weed Interference

Weed interference can reduce soybean yield substantially. Oerke and Dehne (2004) reported a 37% soybean yield loss due to weed competition. Soltani et al. (2017b) in a meta- analysis found that weed inference caused an average 52% soybean yield loss when no control measures were implemented from trials conducted in 2007-2013 in North America. Soybean yield loss due to weed interference is attributed to direct competition for water, nutrients and light and indirect competition through changes in the red to far-red ratio which negatively affects soybean emergence, growth, and seed production (Ghersa et al. 2000; Chandler et al. 2001; Van

Roekel et al. 2015).

Weed species vary in their impact on soybean yield loss. Prickly lettuce (Lactuca serriola

L.) interference reduced soybean yield up to 80% (Weaver et al. 2006), similar to common cocklebur (Xanthium strumarium L.) (Barrentine and Oliver. 1977). Giant foxtail (Setaria faberi

Herrm.) interference reduced soybean yield 28% when allowed to compete with soybean for the entire growing season (Knake and Slife 1962). Moolani et al. (1964) found that smooth pigweed interference reduced soybean yield 55%. Bruce and Kells (1990) reported up to 90% yield loss in soybean due to Canada fleabane interference. Shurtleff and Coble (1985) reported that redroot pigweed, common lambsquarters (Chenopodium album L.), and common ragweed (Ambrosia artemisiifolia L.) interference reduced soybean yield 22, 15, and 12%, respectively. Palmer amaranth competition reduced soybean yield 68% (Klingaman and Oliver 1994), while waterhemp interference reduced soybean seed yield an average of 43% from a 3-year study

(Hager et al. 2002). Haizel and Harper (1973) concluded that weed species mixtures reduce soybean yield less than the sum of their independent impacts.

3.3 Other Benefits of Weed Control in Soybean

There are several other benefits of weed control in soybean in addition to minimizing yield loss. Wilson et al. (1995) found that weed seed return to the soil seedbank was directly correlated with weed biomass. Chandler et al. (2001) reported that as the weed-free period in soybean increased the weed seed return to the soil seed bank decreased. In addition, poor weed control resulted in increased foreign matter in the soybean seed sample at harvest and the presence of some weeds can cause staining in soybean (Burnside et al. 1969; Ellis et al. 1998).

Nave and Wax (1971) found that harvesting a large amount of weed biomass reduced harvest efficiency and increased separating losses. Additionally, the presence of weeds can act an alternate host for disease and insect pests, increasing potential damage to the crop (Zandstra and 19

Motooka 1978; Capinera 2005). Eliminating weeds from a soybean field has numerous benefits beyond yield losses.

3.4 Critical Weed-Free Period in Soybean

The critical weed-free period (CWFP) is defined as the interval in the life cycle of a crop when it must be kept weed-free to avoid significant yield loss (Van Acker et al. 1993). The

CWFP is influenced by various factors including relative time of crop and weed emergence, weed density, environment, row spacing, and weed species present (Knezevic et al. 2002;

Swanton et al. 2008). The identification of the CWFP helps weed management practitioners implement weed management strategies that minimize soybean yield losses due to weed interference (Nieto et al. 1968; Van Acker et al. 1993; Halford et al. 2001).

Many studies have been completed that have ascertained the CWFP in soybean. In

Argentina, Eyherabide and Cendoya (2002) reported a wide range in soybean yield loss when weeds were present from emergence up to V7 and R1. Mulugeta and Boerboom (2000) found the

CWFP to range from V2 to R1, with changes in different tillage systems and soybean row spacing. In a conventional tillage system in Ontario, Van Acker et al. (1993) reported that to prevent soybean yield losses of greater than 5%, the CWFP is from V2 to V3; to prevent soybean yield loss greater than 2.5% the CWFP was reported to be from VE to V4. Halford et al. (2001) found the CWFP to be from the V1 to R1 stage and determined that it was longer in a no-till compared to a conventional tillage system. Green-Tracewicz et al. (2012) reported that soybean should be weed-free from V1 to V3 due to a change in the R: FR ratio which resulted in a reduction in yield. Maintaining soybean weed-free during the CWFP minimizes yield loss due to weed interference.

3.5 Integrated Weed Management

Integrated weed management (IWM) is the utilization of multiple weed management tactics to minimize crop yield loss due to weed interference. IWM employs cultural, mechanical, biological, and chemical weed control techniques (Baldwin and Santelmann 1980). Each individual weed control tactic is not expected to provide adequate weed control, but when utilized together acceptable weed control can be achieved (Swanton and Weise 1991). IWM incorporates cultural weed control techniques including crop rotation, high seeding rates, narrow row widths, strategic fertilizer application timing and placement, and competitive cultivars.

Mechanical weed control techniques include primary and secondary tillage and interrow cultivation; biological weed control techniques utilizes crop production tactics that increase weed seed predation by insect, fungi and bacteria, and chemical weed control to minimize the effect of weed interference on crop yield (Swanton and Murphy 1996). By reducing the weed population below threshold levels, the weed seed bank can be reduced and future yield lose due to weed interference will be reduced.

3.6 Cultural Weed Control

Cultural weed control is manipulating a cropping system in a way that will minimize weed interference. Some forms of cultural weed control include crop rotation, high seeding rates, narrow row-widths, strategic row orientation, optimal fertilizer placement, timing and rate, competitive cultivars, and the use of cover crops.

Cover crops are most commonly grown between cash-generating crops to reduce weed seedling emergence, weed growth, and weed seed return to the soil (Swanton and Weise 1991).

Rapid cover crop growth and dense canopy development after summer annual crops will limit

21 weed seedling emergence and growth. Annual cover crops may be killed by the harsh winter environment while perennial cover crops may persist until the following spring and will then need to be terminated by mechanical or chemical methods (Swanton and Weise 1991). The resulting dead cover crop biomass can suppress weeds through the CWFP of the following cash- generating crop. Weed suppression by cover crops has been attributed to competition by the living cover crop, physical obstruction of weed emergence, and allelopathy (Moore et al. 1994;

Hartwing and Ammon 2002). Teasdale (1996) found that as cover crop residue biomass increased, early season weed suppression also increased. This study also found that cover crops are most effective on small-seeded annual weed species with a light requirement for germination.

Smith et al. (2019) found that multiple-species cover crop mixtures were no more effective in suppressing weeds than a single-species cover crop. Teasdale (1996) also raised the inherent risk of cover crops stating, “Any living mulch competitive enough to suppress weeds will also be competitive enough to interfere with crop growth and yield.”

Soybean row width, planting density, and row orientation can influence weed germination, emergence, and growth. Traditionally, soybean has been planted in rows spaced 76 cm apart, but in recent years there has been a trend to reduce soybean row width to 50, 38, or 19 cm. One of the benefits of reduced row-width is more rapid canopy closure which ensures maximum solar radiation interception to maximize soybean yield and also reduce weed growth through reduced seed germination and growth (Taylor 1980; Board et al. 1990). Harder et al.

(2007) found reduced weed seedling emergence when soybean was seeded in 19 cm rows compared to 76 cm rows following an application of glyphosate. Kells et al. (2004) and Young et al. (2001) reported similar results where there was reduced weed growth following the application of glyphosate when soybean was seeded in rows spaced 19 or 38 cm apart compared

22 to 76 cm rows. The impact of soybean row width on weed suppression is species specific. Howe and Oliver (1987) found that soybean planted in rows spaced 20 cm apart reduced pitted morningglory (Ipomoea lacunosa L.) seed production and leaf area index compared to soybean seeded in rows spaced 1 m apart. Légère and Schreiber (1989) found a 20% increase in control of redroot pigweed as soybean row spacing decreased from 76 to 25 cm. Bradley (2006) suggests that weeds with extended emergence patterns, such as waterhemp, are better controlled by narrow soybean row spacing when compared to early emerging weed species. Soybean row spacing has a species-specific impact on weed germination, emergence, and growth.

3.7 Mechanical Weed Control

Mechanical weed control is the utilization of physical forces to kill or inhibit the growth of weeds. This includes cultivating, harrowing, hoeing, and flaming prior to and after crop emergence (Leblanc and Cloutier 1996; Van der Weide et al. 2008). Tillage often buries weed seeds deep into the soil preventing weed seedling emergence (Yenish et al. 1992; Mohler 1993).

Plowing or other forms of aggressive tillage can place the weed seed deep in the soil profile, allowing for the seed to remain dormant until another tillage event that moves the weed seed closer to the soil surface where acceptable conditions occur for weed seed germination and emergence (Yenish et al. 1992). The first flushes of germinating annual weeds can be controlled utilizing shallow in-crop field cultivation (Swanton and Weise 1991). Many forms of mechanical weed control can be utilized to keep the crop weed-free through the CWFP (Van Acker et al.

1993). Weed species with large seeds and a long dormancy period are well-adapted to conventional tillage systems. Large-seeded annual broadleaves such as velvetleaf (Abutilion theophrasti Medik.) ideally germinate from a depth of 2.5 cm below the soil surface (Mester and

Buhler 1990); seeds are frequently placed in this zone with primary tillage. Tillage can be 23 effective for the control of biennial and perennial weeds including wild carrot (Daucus carota

L.), dandelion (Taraxacum officinale F. H. Wigg.), Canada thistle [Cirsium arvense (L.) Scop.], and quackgrass [Elymus repens (L.) Gould] (Derksen et al. 1993; Murphy et al. 2006).

Tillage can negatively effect weed management and the environment. Tillage can bring buried seeds to the soil surface where the conditions are ideal for weed seed germination and seedling emergence (Buhler 1997; Jha and Norsworthy 2009). It has also been found that weed seed mortality increased when seeds remained on the surface of the soil compared to deep burial by tillage (Blackshaw et al. 2008). Crop residues may delay and inhibit weed seed germination and initial growth by adding an additional barrier to weed seedling establishment. Furthermore, reduced-tillage systems prevent soil erosion and reduce pesticides and fertilizer run-off while conserving soil moisture (Lafond et al. 1990; Derksen et al. 1993; Fawcett et al. 1994; Swanton and Murphy 1996). The benefits of mechanical weed control and tillage must be balanced with the negative effects on the environment.

3.8 Biological Weed Control

Biological weed control is the intentional use of living organisms such as insects, bacteria and fungi to suppress or control a weed population. Eradication of a problem weed may not be the ultimate goal with biological weed control as the weed may be needed as a food source for continued success of the natural enemy. Rather, the goal of biological weed control may be to suppress a weed species to a level where its economic impact is negligible (Swanton and Weise

1991). Invasive weeds are often targeted for biological control. Mile-a-minute (Persicaria perfoliata (L.) H. Gross) is an aggressive annual vine native to Asia that is now normalized along the eastern coast of the United States. In 2004, Rhinoncomimus latipes Korotyaev, a host-specific weevil from China, was released and has reduced the effects of this deleterious weed by feeding 24 on it (Hough-Goldstein et al. 2008; Hough-Goldstein et al. 2012). Similarly, Canada’s west coast rangeland has been plagued with diffuse knapweed (Centaurea difusa Lamarck), a short-lived perennial plant native to Eurasia that is a poor forage and displaces native grasses (Harris and

Cranston. 1979; Lejeune and Seastedt 2001). A weevil, Larinus minutus, was released between

1996 and 1999 which feeds on rosettes and bolting stems resulting in lower diffuse knapweed densities and a decline in seed production at sites across British Columbia (Myers et al. 2009).

Plant pathogens can also be used as a form weed control (Bajwa et al. 2015). DiTommaso et al.

(1996) found that the fungal pathogen Colletotrichum coccodes reduced velvetleaf height in soybean compared to uninoculated plants. C. coccodes application resulted in short velvetleaf plants that were unable to grow above the soybean canopy and compete for available light.

Overall, the utilization of biological weed control in field crops has been limited as there are inefficiencies and more economical ways to control a wider range of weed species.

3.9 Chemical Weed Control

Chemical weed control is the most common method of weed control in modern agriculture. The use of chemical weed control began with the introduction of the synthetic auxin herbicides, specifically 2,4-D in 1946 (Green and Owen 2011) and later atrazine in 1958.

Herbicides stop plant growth by inhibiting plant functions necessary to plant life, with the most common target sites being enzymes (Hall et al. 1999). The introduction of glyphosate-resistant

(GR) crops in 1996 increased the adoption rate of chemical weed control (Green and Owen

2011). Soybean producers rapidly adopted the GR technology since it is effective, low cost, relatively easy to apply, has minimal labour requirements and a wide margin of crop safety. With the increased use of glyphosate, and other chemical herbicides, herbicide resistance has become a serious issue for many grain and oilseed producers worldwide. 25

Herbicides can be used in many ways to effective eliminate weeds. In soybean, herbicides are applied preplant (PP), preplant incorporated (PPI), preemergence (PRE) and postemergence (POST). PP herbicide applications are made prior to seeding soybean for the control of emerged weeds and full-season residual weed control (OMAFRA 2019). PPI herbicide applications are made prior to planting the crop and are incorporated into the soil utilizing tillage that should be no deeper than 10 cm. PRE herbicide applications are made after seeding soybean and prior to crop and weed emergence. Soil-applied herbicides require rainfall within 7 to 14 days of application to dissolve the herbicide into the soil water solution so that if can be taken up by the developing weed seedling (Stewart et al. 2010). POST herbicide applications are made after the crop and weeds have emerged (OMAFRA 2019).

3.10 Herbicide-Resistant Crops

The release of herbicide-resistant (HR) crops has changed weed management in soybean.

In 1996, glyphosate-resistant (GR) soybean was commercialized in the USA resulting in a dramatic increase in the use of glyphosate (Young 2006; Givens et al. 2009). GR crops were rapidly adopted by producers across many production regions around the globe because of the wide margin of crop safely, excellent weed control efficacy, flexible application timing, no residues in the soil affecting future crops in the rotation, low cost, simplicity, and relatively low mammalian and environmental toxicity (Gulden et al. 2009; Green and Owen 2011; Krausz et al.

2001,). In Ontario there are four HR soybean technologies available including GR, glufosinate- resistant, 2,4-D-resistant and dicamba-resistant soybean (OMAFRA 2019). Furthermore, some

HR soybean cultivars have transgenes that confer resistant to more than one herbicide; this is known as “stacking” of herbicide resistance traits. As the adoption of HR crops has increased, and specifically the widespread adoption of GR soybean, there has been rapid evolution of GR 26 weeds which is a serious issue for crop production. Herbicide resistance develops as a result of the high selection pressures imposed by herbicides on a weed population from the overuse of chemistries with similar modes of action (Shaner 1995; Green and Owen 2011).

Section 4.0 - Herbicide-Resistant Weeds

4.1 Introduction into Herbicide Resistance

Herbicide resistance is the ability of a plant to survive and reproduce following the application of a dose of herbicide that is normally lethal to the susceptible plant (Vencill et al.

2012). There is confusion, and the incorrect use, of the terms “herbicide tolerance” and

“herbicide resistance”. The Weed Science Society of America (WSSA) defines herbicide tolerance as “the inherent ability of a species to survive and reproduce after herbicide treatment”

(Vencill et al. 2012). Tolerance occurs naturally with no selection pressure exerted and no genetic changes within the plant. Herbicide resistance is most often a result of selection pressure placed on a weed population through the repeated application of a herbicide(s) with the same mode of action (Jasieniuk et al. 1996; Vencill et al. 2012). As a result of this selection pressure, biotypes with natural resistance to that mode of action will survive, produce seed, and eventually become the dominant biotype within a local population.

Herbicide resistance may also be induced by techniques such as genetic engineering, plant breeding, or mutagenesis.

Herbicide resistance is not a new concept. The first reported cases of herbicide resistance were in 1957; 2,4-D-resistant wild carrot was found in Ontario, Canada (Switzer 1957) and

2,4-D-resistant spreading dayflower (Commelina diffusa L.) was identified on the island of

Hawaii, USA (Hilton 1957).

Worldwide there are a total of 512 unique cases of herbicide-resistant (HR) weeds (Heap

2020). Unique cases are defined as different species by site of action. In total, herbicide resistance has been found in 262 different species in 70 different countries. Resistance has been found in 23 of the 26 known herbicide modes of action, for a total of 167 different herbicides.

ALS-inhibitors have the most number of HR cases with 165, PS II-inhibitors are second with 74, and acetyl CoA carboxylase inhibitors are third with 49 cases. Evolution of HR weeds within agronomic cropping systems is a major threat to crop productivity and farm economical sustainability across the world.

4.2 Mechanisms of Resistance

Plants evolve resistance to herbicides through several mechanisms and genetic alterations within the plant. Herbicide resistance is due to metabolic and physiological changes within a plant that most often interfere with the herbicide reaching and binding to its target site in lethal quantities (Holt et al. 1993). These changes are referred to as mechanisms of herbicide resistance

(Vencill et al. 2012). Generally, mechanisms of herbicide resistance are divided into target-site and non-target-site mechanisms. It is thought that high doses of a herbicide select for target-site resistance, while low doses result in non-target-site resistance, however this theory has been challenged recently (Gardner et al. 1998; Yuan et al. 2007). Cross resistance occurs when one mechanism of resistance confers resistance to multiple herbicides in a HR weed biotype (Beckie and Tardif 2012). Multiple resistance occurs when a HR weed biotype has two or more mechanisms of resistance that confer resistance to two of more herbicides with different sites of action (Holt et al. 1993). In rare cases negative cross resistance occurs were weeds that may be resistant to one herbicide and as a result may also be more susceptible to other herbicides

(Gressel and Segel 1990). 28

4.3 Target-Site Resistance

Herbicides control weeds by targeting specific enzymes or proteins that ultimately inhibit plant growth and life (Vencill et al. 2012). Target-site resistance occurs when a herbicide reaches the target site at a normally a lethal dose, but a gene mutation results in amino acid changes which prevents the herbicide from binding as efficiently to the target enzyme rendering the herbicide molecule to be ineffective in inhibiting that specific enzyme (Yuan et al. 2007; Duke and Powles 2008). Another form of target-site resistance is gene overexpression where there is an overexpression of the target enzyme and the herbicide is unable to inhibit it at its usual lethal dose (Powles and Yu. 2010). Sometimes there is a fitness penalty with target-site resistance. A fitness penalty in a resistant biotype is associated with reduced growth and fecundity and over time, in the absence of selection pressure, would result in the population reverting to the susceptible biotype (Vila-Aiub et al. 2009). Target-site resistance in most cases are monogenic as they are controlled by a singular gene. Target-site is the primary mechanism of resistance reported in ACC-, ALS-, PS II, and PPO-inhibitors (Powles and Preston 2006; Powles and Yu

2010; Heap 2020).

4.4 Non-Target-Site Resistance

Non-target-site resistance occurs when the herbicide is prevented from reaching the target site at a lethal dose, allowing the plant to survive. Mechanisms include decreased herbicide absorption, decreased herbicide translocation, increased herbicide sequestration and enhanced herbicide metabolism (Powles and Yu 2010). Increased activity of enzymes such as glutathione

S-transferases (GST) and cytochrome P450 monooxygenases (CYP450) are involved in enhanced herbicide metabolism (Powles and Yu 2010; Jugulam and Shyam 2019). GSTs and

CYP450’s are not exclusively in HR weeds but they are also utilized in herbicide safeners, allowing for the crop to metabolize the herbicide more rapidly to reduce the potential for herbicide injury in crops (Yuan et al. 2007). GST has a significant role in herbicide detoxification in many crops, such as atrazine metabolism in corn (Guddewar et al. 1979).

Non-target-site herbicide resistance includes distinctly different mechanisms. Reduced herbicide absorption occurs when the herbicide is not taken up in sufficient quantities to reach the target site in a lethal amount. This mechanism has been noted in glyphosate-resistant (GR) annual ryegrass (Lolium multiflorum Lam.) and GR Palmer amaranth (Nandula et al. 2008,

Palma-Bautista et al. 2019). Palma-Bautista et al. (2019) found that a glyphosate susceptible biotype of Palmer amaranth absorbed 10% more glyphosate than the resistant biotype (Gaines et al. 2011; Ward et al. 2013). Reduced herbicide translocation results in reduced herbicide movement within the plant, resulting in reduced amounts reaching the target site. Resistance due to reduced translocation appears to be a common mechanism of resistance in rigid ryegrass

(Lolium rigidum Gaud.) and Canada fleabane (Ge et al. 2010; Powles and Yu 2010); this mechanism of resistance may cause a fitness penalty (Preston and Wakelin 2008). Herbicide sequestration occurs when a herbicide is inactivated through binding or is removed from the active region of the cell and redirected to the cell vacuole preventing it from impacting the target site (Vencill et al. 2012). Sequestration of the herbicide can happen in various locations depending on the herbicide and species, but it is most commonly sequestered in the vacuole (Ge et al. 2011; Sammons and Gaines 2014). Enhanced metabolism is a mechanism a plant uses to detoxify a foreign compound such as herbicide. The plant quickly degrades the herbicide to a non-toxic metabolite preventing the active herbicide from reaching the site of action at lethal dose. Enhanced metabolism can confer multiple and/or cross resistance across various site of

30 actions while reduced translocation and target-site substitution usually confer resistance to herbicides with the same site of action (Beckie and Tardif 2012).

4.5 Glyphosate Resistance

The increased cultivation of GR crops and concomitant increased use of glyphosate has resulted in the evolution of GR weeds (Young 2006). Glyphosate use has increased almost

15-fold since introduction of GR soybean in 1996 (Benbrook 2016). The increased use of glyphosate can be attributed, in part, due to broad-spectrum weed control, wide margin of crop safety, no carryover injury to subsequent crops in the rotation, low cost, and low environmental impact (Powles 2008; Duke and Powles 2008). It was suggested that glyphosate resistance in weeds was unlikely since there were no documented cases over the first 20 years of its use, little to no metabolism of glyphosate in plants, and site-directed plant EPSPS mutants provided little resistance (Bradshaw et al. 1997). The widespread adoption of GR crops contributed to reduced herbicide diversification and an increase in the evolution of GR weeds (Dill 2005; Powles 2008;

Beckie 2011).

Glyphosate-resistant weeds are now a global issue. Globally, there are 48 weeds species that have evolved resistance to glyphosate (Heap 2020). Interestingly, there are 24 monocotyledonous and 24 dicotyledonous weed species that are resistant to glyphosate. The first case of glyphosate resistance reported was in rigid ryegrass in an Australian apple orchard in

1996 (Powles et al. 1998). In Canada, there are currently six weed species that have evolved resistance to glyphosate, four of which occur in Ontario (Heap 2020). The first confirmed GR weed in Ontario was giant ragweed (Ambrosia trifida L.) in 2008 (Vink et al. 2012), followed by

Canada fleabane in 2010 (Byker at al. 2013), common ragweed in 2011 (Van Wely et al. 2015) and waterhemp in 2014 (Schryver et al. 2017a). 31

There are many mechanisms that confer resistance to glyphosate. The target-site mutation of 5-enolpyruvylshikimate-3-phosphate synthase (EPSPS) was first identified in a goosegrass

(Eleusine indica L.) with resistance being caused by a proline to serine substitution at position

106 (Baerson et al. 2002); additional substitutions at position 106 have been identified that confer resistance to glyphosate including alanine, leucine, and threonine (Powles and Preston

2006; Page et al. 2018). While single amino acid substitution at position 106 generally endows low-levels of resistance, the single amino acid substitution of glycine to alanine at position 101 and threonine to isoleucine at position 102 confer greater levels of resistance to glyphosate

(Eichholtz et al. 2001; Alibhai et al. 2010; Sammons and Gaines 2014). Double mutations such as threonine to isoleucine substitution at position 102 and the proline to serine substitutions at position 106, also known as TIPS, has been described in transgenic crops as providing high-level glyphosate resistance in corn (Lebrun et al. 2003). Yu et al. (2015) found the first double amino acid substitution in a weed species when it was found a biotype of goosegrass had a threonine to isoleucine substitution at position 102 and a substitution of proline to serine at position 106. The double amino acid substitution of glycine to alanine at position 101 plus glycine to asparagine at position 144 and glycine to alanine at position 101 plus alanine to threonine at position 192 have been shown to confer glyphosate resistance in transgenic crops such as corn and canola (Brassica napus L.) respectively (Howe et al. 2002; Kahrizi et al. 2007). Gene amplification of the EPSPS gene has been shown to confer high level of glyphosate resistance in Palmer amaranth. Gaines et al. (2010) documented up to 100 fold EPSPS gene amplification. Reduced glyphosate translocation was first identified in rigid ryegrass (Lorraine-Colwill et al. 2003; Powles and Yu

2010). Some species have both an EPSPS Pro-106 target-site mutation and reduced translocation

(Yu et al. 2007; Preston et al. 2009). Enhanced metabolism and vacuolar sequestration also

32 confer resistance to glyphosate in some weeds species (Gonzalez-Torralva et al. 2010).

Glyphosate resistance has evolved rapidly worldwide through many different mechanisms of resistance.

4.6 Photosystem II-Inhibitor Resistance

Photosystem II (PS II)-inhibitor resistance has been researched extensively. PS II- inhibitors disrupt the photosynthetic process in susceptible plants by binding to specific sites within the PS II complex in the plant chloroplasts (Hess 2000). The PS II-inhibitors are further divided by the WSSA mode-of-action classification system into three separate Groups (5, 6, and

7) because, although all PS II-inhibitors bind to the D1 protein, the binding occurs at different sites and in different ways. WSSA Group 5 herbicides form hyrdogen bounds with serine at position 264 of the QB-binding niche while WSSA Group 6 herbicides bind with histidine 215 with in the QB-binding niche on the D1 protein (Trebst 1987; Tietjen et al. 1991). The following herbicides are classified based on their attachment site with the C1 group including herbicides belonging to pyridazinones, triazines, triazinones, and uracils; C2 group including the amides and ureas, and C3 group including benzothiadiazinones, nitriles, and phenylpyridazines (Shaner

2014).

Constant and consistent use of the PS II inhibiting herbicides, in particular the triazine herbicides, has led to the evolution of resistance. The first weed species that was confirmed to be resistant to PS II-inhibitors was common groundsel (Senecio vulgaris L.) in 1968 in Washington

State (Ryan 1970; Heap 2020). Currently there are 74 weed species worldwide that are resistant to the PS II inhibiting herbicides. In Canada, there are 13 species resistant to PS II inhibiting herbicides with waterhemp being the most recent in 2017 (Heap 2020). With few exceptions PS

II-inhibitor resistance is conferred by a target-site mutation in the psbA gene which encodes for 33 the D1 protein. A single amino acid substitution of serine to glycine at position 264 in the D1 protein is most common amongst triazine-resistant weeds (Murphy and Tranel 2019). As a result of this substitution, hydrogen bonds are removed from the target site and herbicide binding is not achieved (Hess 2000). These target-site mutations are encoded in the plastid DNA, making resistance maternally inherited. The amino acid mutation which endows resistance to triazine herbicides also reduces substrate affinity, thereby reducing the electron transfer rate in PS II reducing photosynthetic efﬁciency and resulting in a reduction of the overall plant growth rate

(McCloskety et al. 1990; Devine and Shukla 2000; Vila-Aiub 2019). Cases of other target-site substitution that confer resistance to PS II inhibiting herbicides do exist. A population of common purslane (Portulaca oleracea L.) was discovered to be resistant to atrazine and linuron due to a single amino acid substitution of serine to threonine at position 264 (Masabni and

Zandstra 1999). In annual bluegrass (Poa annua L.) a population was found to have a valine to isoleucine substitution at position 219 that conferred resistance to both metribuzin and diuron

(Mengistu et al. 2000). A leucine to valine substitution at position 218 that was reported in common lambsquarters in Germany that resulted in resistance to the triazinone herbicides, but not to the triazine herbicides (Thiel and Varrelmann 2014).

Enhanced metabolism, a non-target-site mechanism of resistance, confers resistance to the PS II inhibiting herbicides in some weed species. Waterhemp (Ma et al. 2013), velvetleaf

(Anderson and Gronwald 1991; Burnet et al. 1993) and annual bluegrass (Preston 2004) have biotypes that are resistant to the PS II inhibitors via metabolic detoxification. Enhanced metabolism is due to GST and/or CYP450 mediated detoxiﬁcation mechanisms.

4.7 Glyphosate-Resistant Canada Fleabane

There are a number of mechanisms that confer resistance to glyphosate in Canada fleabane. Initially non-target-site resistance mechanisms were discovered and thought to be the primary and most widespread mechanisms of glyphosate resistance in Canada fleabane. Reduced glyphosate translocation in Canada fleabane was thought to be the most common mechanism worldwide (Feng et al. 2004; Dinelli et al. 2006; Shaner 2009). Ge et al. (2010) found selective sequestration of glyphosate into the vacuole and subsequent reduced translocation of glyphosate to the growing points conferred resistance in Canada fleabane. This mechanism has been found to be temperature dependent, with decreased sequestration and increased glyphosate efficacy observed under cooler temperatures (Ge et al. 2011). This could allow for growers to make efficacious applications in colder seasons. González-Torralva et al. (2012) found both reduced glyphosate translocation and increased herbicide metabolism are possible in some populations of

Canada fleabane.

Target-site resistant mechanisms also confer resistance to glyphosate in Canada fleabane.

Powles and Preston (2006) found that an amino acid substitution of the EPSPS enzyme in

Canada fleabane at position 106 from proline to serine and alanine or threonine that confer low level resistance to glyphosate. In Ontario, target-site resistance was found with an amino acid substitution of position 106 from proline to serine which appears to be the most common mechanism of resistance for GR Canada fleabane in Ontario (Page et al. 2018). Interestingly, in the Ontario biotypes the amino acid substitutions confers a high level of resistance to glyphosate.

Overexpression of the EPSPS enzyme has also been reported to allow Canada fleabane to continue to grow after the application of glyphosate (Mueller et al. 2003; Dinelli et al. 2006).

4.8 Metribuzin-Resistant Waterhemp

There is limited information in the scientific literature on metribuzin-resistant waterhemp, although triazine resistance is well documented. In Ontario, waterhemp control with metribuzin has been variable. Vyn et al. (2007) reported that metribuzin (1120 g ai ha-1), applied PRE, controlled waterhemp 24 and 99% near Petrolia and Cottam, ON, respectively. The results from the site near Cottam are similar to those of Hausman et al. (2013) in Illinois where metribuzin

(420 g ai ha-1) controlled waterhemp 92%.

PS II resistance in waterhemp is conferred by target- and non-target-site mechanisms of resistance. Triazine-resistant waterhemp was first confirmed in 1990 in Nebraska (Anderson et al. 1996). In survey of 59 Illinois waterhemp populations, Patzoldt et al. (2002) reported that PS

II resistance was present in 14 populations. Target-site resistance, a serine to glycine amino acid substitution at position 264 of the D1 protein, was the first mechanism of resistance identified in waterhemp (Foes et al. 1998). Ma et al. (2013) reported enhanced metabolism, due to increased

GST activity, conferred resistance to atrazine in a waterhemp population in Illinois. Evans et al.

(2017) found that a single GST enzyme, AtuGSTF2, correlated strongly with PS II resistance and is believed to be responsible for detoxifying atrazine in some resistant waterhemp biotypes. In

Illinois, rapid metabolism of atrazine in PS II-resistant waterhemp resulted in a several hundred- fold resistance level compared to atrazine sensitive plants (Evans 2016). Vennapusa et al. (2018) also reported populations of waterhemp in Nebraska that rapidly metabolized atrazine, mediated by GST conjugation, conferring resistance to atrazine but not to metribuzin. PS II-inhibitor resistance can be conferred by both target-site and non-target-site mechanisms of resistance.

4.9 Cost and Management of Herbicide Resistance Weeds

Herbicide-resistant weeds come at a high economic cost to the grower. The loss of revenue due to decreased yield has been noted as a largest cost for producers (Norsworthy et al. 2012).

Other costs that producers incur due to HR weeds include increased cost of alternative herbicides, increase cost of cultural, mechanical, or biological control tactics, reduced land value, and reduced price of the commodity at the point of sale due weed seed contamination. Mueller et al. (2005) found in the USA, the presence of GR Canada fleabane increased the cost of soybean production by $28 per hectare. Similarly, control of GR Palmer amaranth has been estimated to cost $40 per hectare for corn, $52 per hectare for soybean, and $74 per hectare for cotton due to additional weed control measures (Carpenter and Gianessi 2010). The increased cost for the control of GR waterhemp in soybean is estimated to be up to $89 per hectare (Mueller et al.

2005). Carpenter and Gianessi (2010) state that moving from a glyphosate only herbicide program to a diversified herbicide program for resistance management could cost growers an additional $105 per hectare in soybean and $88 per hectare in corn, although resistance management programs have been modeled to pay-off in the long term (Livingston et al. 2016).

The presence of HR weeds reduces producer profitability.

Even with the looming herbicide resistance crisis many growers are reluctant to implement herbicide resistance mitigation management practices. Using multiple effective modes of effective action is one of the least utilized practices by growers for reducing herbicide resistance on their farms (Shaner 1995; Duke and Powles 2008, Frisvold et al. 2009). Furthermore, maximizing short term economic return is thought to be one of the primary reasons why producers do not implement HR management tactics (Beckie 2006). In a nationwide survey,

Schroeder et al. (2018) found that amongst industry stakeholder 95% of all survey participants 37 agreed or strongly agreed that HR is a problem in agriculture. Furthermore, these same stakeholders wanted new herbicides with new mechanisms of action, less government involvement, increased utilization of integrated weed management and stated that the current agricultural economy was not conducive to managing HR weeds. The hope for herbicides with a novel mode of action may not come to fruition since no new herbicide site of action has been introduced commercially in more than 20 years (Dukes and Powles 2008; Duke 2012).

Many researchers have given thought on how to mitigate herbicide resistance going forward.

Norsworthy et al. (2012) suggested several production practices that will reduce the evolution of herbicide resistance including; scouting and understanding the lifecycle of hard to control weeds, utilization of multiple methods of effective weed control including cultural, mechanical, biological, and chemical methods, reducing weed seed return to the seed back, stopping the transportation of weed seed from field-to-field, and utilizing herbicides at the label rate. These strategies will reduce the selection pressure and slow the evolution of HR weeds. Shaw et al.

(2018) proposed several actionable items in response to Schroeder et al. (2018). These recommendations included reducing the weed seedbank through diversified weed control programs, communicating that discovery of new herbicide sites of actions are rare, promoting the application of full-labeled rates at the correct weed and crop growth stage, increasing research funding to address knowledge gaps, and demonstrating the benefits of a diversified weed- management system for the mitigation of HR weeds. Overall, a shift to long term cropping systems approach, including less dependence on herbicides, and an overall diversification of cropping systems may need to be adopted in fields were HR weeds reside (Beckie 2006).

Utilizing a balanced integrated weed management approach is among the most promising strategies for combatting herbicide resistance.

Chapter 5.0 - Glyphosate, Bromoxynil, Pyraflufen-ethyl/2,4-D, Metribuzin, and Tiafenacil

5.1 Glyphosate

Glyphosate is a systemic, non-selective, herbicide used globally for weed management in crop production. Glyphosate was first synthesized in 1950 by a Swiss chemist Dr. Henri Martin, who worked for a small pharmaceutical company called Cilag; no pharmaceutical uses were identified (Franz et al. 1997). Glyphosate was sold to a number of companies including

Monsanto. Dr. John Franz identified the potential herbicidal properties of this compound in

1970. In 1974, the formulated herbicide Roundup® was sold commercially by Monsanto (Duke and Powles 2008). Through the 1970s and 1980’s, glyphosate had a wide range of uses including weed management in orchards and forestry, and as a preplant burn down, a pre-harvest desiccant or post-harvest burndown in agronomic and horticultural crops (Benbrook 2016). The use of glyphosate increased dramatically following the introduction of GR (Roundup Ready®) crops in

1996.

Glyphosate inhibits 5-enolpyruvyl shikimate-3-phosphate synthase (EPSPS), an enzyme in the shikimate biosynthetic pathway (Franz et al. 1997). The shikimate pathway, also known as the shikimic acid pathway, is a seven-step metabolic pathway that results in the synthesis of chorismate, a precursor for the production of aromatic amino acids (Shaner 2014). EPSPS catalyzes the reaction between shikimate-3-phosphate (S3P) and phosphoenolpyruvate (PEP) which produces 5-enolpyruvyl shikimate-3-phosphate. The inhibition of the aforementioned step prevents the formation of aromatic amino acids: tryptophan, tyrosine, and phenylalanine. It is thought that this is the explanation for plant death following the application of glyphosate (Duke and Powles 2008), but the exact cause of death is not fully known.

Glyphosate is an efficacious herbicide that must be applied postemergence to the weeds since it is rapidly adsorbed to soil colloids and biologically unavailable (Shaner 2014).

Symptoms of glyphosate injury on susceptible plants include chlorosis and necrosis of the leaves between 4 and 20 days depending on the weed species susceptibility, growth stage of the weed at application, and weather conditions; symptomology is most pronounced in the growing point of the plant (Shaner 2014). Glyphosate is absorbed by the foliage of the plant with transport across the plasmalemma being slower than most nonpolar herbicides (Franz et al. 1997). Glyphosate is primarily symplastically translocated via the phloem; it accumulates in metabolic sinks of sucrose such as the roots, newly formed leaves, and meristematic tissue (Duke and Powles 2008;

Shaner 2014). Glyphosate is slowly metabolized in plants to amino methylphosphonic acid

(AMPA); the breakdown in plants is catalyzed, in part, by the enzyme glyphosate oxidoreductase

(Reddy et al. 2004; Duke and Powles 2008; Shaner 2014). Glyphosate has essentially no soil activity; this zwitterionic compound is rapidly and tightly adsorbed to organic matter constituents and metal ions in the soil (Franz et al. 1997). The adsorption of glyphosate to soil colloids results in minimal leaching of the herbicide.

5.2 Bromoxynil

Bromoxynil is a selective, contact herbicide applied POST in monocot and some dicot crops for annual broadleaf weed control. The synthesis of bromoxynil was first reported in 1896 in Germany by Auwers and Reis (1896) via a four-step synthesis process. Bromoxynil was first registered in the United States in 1965 for broadleaf weed control in wheat and barley (USDA

1988). In Canada it is currently sold under the trade names Bromotril®, Brotex®, Koril®, and

Pardner®. Bromoyxnil is currently labeled for use in canola, alfalfa (Medicago sativa L.), barley

(Hordeum vulgare L.), canary seed (Phalaris canariensis L.), carrot (Daucus carota L.), corn,

40 fall rye (Secale cereale L.), fallow, flax (Linum usitatissimum L.), oats (Avena sativa L.), and wheat (Triticum aestivum L.)(Anonymous 2019a). For some of those crops, bromoxynil can only be applied preplant since in-crop post emergence applications result in crop injury.

Bromoxynil is classified as a PS II-inhibitor at site B. Bromoxynil is a Group 6 herbicide that belongs to the benzonitrile family; other herbicide families in the Group 6 herbicides include the benzothiadiazoles and phenyl-pyridazines. Photosynthesis inhibitors act in the chloroplast thylakoid membrane with the herbicide binding to the QB-binding niche on the D1 protein of the

PS II complex (Shaner 2014). Although the Group 5 herbicides (triazines) and the Group 6 herbicides are both PS II-inhibitors, the benzonitriles and benzothiadiazole binds to histidine at position 215, a different amino acid, on the D1 protein (Trebst 1987). This herbicide molecule displaces a plastoquinone molecule, resulting in an inhibition of electron flow in the electron transport chain from QA to QB as the herbicide does not accept two electrons. The result is a loss of CO2 fixation and a halt in NADPH production.

Bromoxynil controls of a wide spectrum of annual broadleaf weeds. Carpenter et al.

(1964) listed weeds in the Crucifereae, Compositaceae, Polygonum and Chenopodium family as being susceptible, with the Crucifereae being the most susceptible and the Chenopodium family being least. In the same study, grasses were the most tolerant and some legumes, such as alfalfa, also showed greater tolerance to higher doses than were needed for broadleaf weed control.

Bromoxynil (336 g ai ha-1) controls wild mustard (Sinapis arvensis L.), kochia (Kochia scoparia

L.), common ragweed, redroot pigweed, velvetleaf, and lambsquarters (Anonoymous 2019a). In corn, bromoxynil + atrazine (280 +1500 g ai ha-1) controlled GR Canada fleabane ≥93% at 8

WAA (Metzger et al. 2019; Mahoney et al. 2016b).

Bromoxynil is a contact herbicide with no residual activity that only controls susceptible weeds that have emerged at the time of application. Limited root uptake, adsorption on soil particles, and rapid degradation in the soil are three reasons for no residual control (Zimmerman

1969). Bromoxynil is readily absorbed by plant foliage and has very little basipetal translocation within the plant (Shaner 2014). Translocated effects are generally sublethal and are observed as chlorosis extending beyond treated areas; translocation is not necessary for phytotoxic effects

(Carpenter et al. 1964; Shaner 2014). Herbicide symptomology begins with chlorosis 1 to 2 days after application followed by complete necrosis 3-6 days after application (Shaner 2014).

Crop selectivity with bromoxynil is relatively complex. Initial research by Carpenter et al. (1964) showed spray retention, distribution of the herbicide deposit, growth stage and environment as possible bases for crop selectivity. It was later discovered that enhanced metabolism occurs in wheat; hydrolysis of the nitrile group results in the production of an amide which is then converted to a carboxyl group followed by decarboxylation (Buckland et al. 1973;

Shaner 2014). In contrast, several bacteria species, such as Klesbsiella pneumonia subsp.

Ozaenae, hydrolyze the cyano group of bromoxynil and use it as a nitrogen source (McBride et al. 1986). The gene responsible for this hydroxylation, bromoxynil-nitrilase (bxn) has been used to develop bromoxynil-resistant crops through rapid hydroxylation (Stalker et al. 1998b).

Bromoxynil-resistant cotton (Gossypium hirsutum L.) in the United States and canola in Canada were commercialized in 1995 and 1997, respectively; both of these transgenic crops are no longer commercially available. Low adoption of bromoxynil-resistant crops is due to the limited spectrum of weeds controlled with bromoxynil (Duke 2005).

There are only three bromoxynil-resistant weeds globally; two of which are in Ontario,

Canada (Heap 2020). The first bromoxynil-resistant weed discovered was common groundsel in

1995 in Oregon, USA in a peppermint (Mentha x piperita L.) field (Mallory-Smith 1998). The mechanism of resistance was identified by Park and Mallory-Smith (2006) as a point mutation of asparagine to threonine at position 266. In 2004 and 2005, bromoxynil-resistant redroot pigweed and smooth pigweed were confirmed in corn fields in Ontario, respectively (Heap 2020).

5.3 Pyraflufen-ethyl

Pyraflufen-ethyl is a low-use rate, contact herbicide developed by Nihon Nohyau Co.

Limited. Pyraflufen-ethyl is most commonly used as a preplant burndown in various crops since its registration in Canada in 2014. Pyraflufen-ethyl controls several broadleaf weeds including cleavers (Galium aparine L.), common chickweed (Stellaria media (L.) Vill.), and prickly sida

(Sida spinosa L.)(Mabuchi et al. 2002; Scroggs et al. 2006). Westra (2016) reported that pyraflufen-ethyl controlled GR kochia in greenhouse and field trials only 12 and 7%, respectively. It is currently labeled for the control of emerged broadleaf weeds prior to the emergence of many crops including wheat, barley, oat, rye, triticale, canola, lentil (Lens culinaris

L.), corn and soybean. In the USA it has also shown potential as a harvest aid in cotton and potato (Solanum tuberosum L.) (Ivany 2005; Griffin et al. 2010).

Pyraflufen-ethyl is not currently sold as a stand-alone herbicide in Canada but is registered under the tradename Pyro® (Anonymous 2018). In Western Canada, pyraflufen-ethyl

(13.5 g/L) and MCPA Ester (420 g a.e/L) are sold under the tradename GoldWing® as a preplant and preemergence herbicide in various crops. In Western and Eastern Canada, pyraflufen-ethyl

(6.1 g/L) and 2,4-D ester (473 g a.e/L) is sold in a premix as BlackHawk®. BlackHawk® is currently registered prior to emergence of barley, field corn, oats, rye, soybean, and both spring and winter wheat (Anonymous 2019b). BlackHawk® (527 g ai ha-1) controls broadleaf weeds such as lambsquarters and cleavers and suppression of Canada fleabane. 43

Pyraflufen-ethyl is a protoporphyrinogen IX oxidase (PPO) inhibitor. It is a member of the phenyl pyrazole family (Shaner 2014). The mode of action has been reported to be the same as the diphenyl ethers such as acifluorfen and fomesafen (Derrick et al. 1988). PPO-inhibitors work by inhibiting PPO, the enzyme that catalyzes the oxidation of protoporphyrinogen IX

(PPGIX) to protoporphyrin IX (PPIX). As a result, there is reduced synthesis of both chlorophyll and heme. Heme is important as it acts as a regulator of the porphyrin pathway, with inhibition leading PPGIX accumulation in the chloroplast (Duke et al. 1991). PPGIX leaks into the cytoplasm where it is converted to PPIX. In the presence of light, singlet oxygen forms which results in lipid peroxidation. The loss of chlorophyll, carotenoids, plastiquinone and tocopherols results in leaky membranes and ultimately plant death.

Pyraflufen-ethyl is not herbicidally active and is classified as a pro-herbicide, requiring plant processes to activate its herbicidal properties. Once it has entered a plant, the ethyl group is lost via hydrolysis and as a result the herbicidal metabolite pyraflufen is formed (Murata et al.

2002). Like other PPO inhibiting herbicides, pyraflufen-ethyl is considered photodynamic with phytotoxic effects only occurring in the presence of light (Duke et al. 1991). Symptomology of pyraflufen-ethyl includes speckling, bronzing and necrosis of the leaves and stems. Pyraflufen- ethyl is a contact herbicide with no significant uptake by the roots or emerging shoots of the plants and limited translocation (Pest Management Regulatory Agency, 2014; Shaner 2014).

Currently there are only two weeds species resistant to pyraflufen-ethyl worldwide. In

2005, common ragweed was reported be resistant to pyraflufen-ethyl in Delaware, USA along with many other PPO inhibiting herbicides, as a result of a R98L PPX2 mutation (Rousonelos et al. 2012; Heap 2020). In 2016, Palmer Amaranth was found to also be resistant in Arkansas with the same R98L PPX2 mutation (Giacomini et al. 2017; Heap 2020).

5.4 Metribuzin

Metribuzin is an asymmetrical triazine, PS II-inhibiting herbicide. Introduced in 1969 by

Du Pont Company, sales of metribuzin began in 1973 under the trade name Sencor® (Eue 1972;

Shaner 2014). Metribuzin is considered an asymmetrical triazine, as the molecular structure of a heterocyclic ring contains three nitrogen and three carbon atoms which are arranged asymmetrically around the ring with an oxygen attached at position one. Atrazine, a close relative, has three nitrogen atoms and three carbons alternating in the heterocyclic ring, therefore it is a symmetrical triazine. While the molecular structure of these molecules may be similar, the structural differences result in differences in crop selectivity and weed control efficacy.

Metribuzin is primarily used for controlling weeds in soybean, potato and tomato; atrazine is primarily used in corn (Friesen and Wall. 1984; Green et al. 1988). Metribuzin is available in both dry and liquid formulations under the trade names Sencor®, Squadron®, and Tricor®. In eastern Canada, Sencor® is registered for use in soybean, potato, field corn, tomato, asparagus

(Asparagus officinalis L.), carrot and fruit trees (Anonymous 2017). Metribuzin is most commonly applied preplant, preplant incorporated or preemergence, but POST applications are utilized in potato and tomato (Gawronski 1986; Anonymous 2017).

Metribuzin controls a wide spectrum of weed species. Metribuzin provides primarily annual broadleaf weed control of cocklebur, jimsonweed (Datura stramonium L.), lady’s-thumb

(Persicaria maculosa Gray), lambsquarters, mustard species, pigweed species, common ragweed, and velvetleaf and provides suppression or control of annual grasses including barnyardgrass (Echinochloa crus-galli (L.) P. Beauv.), crabgrass species, fall panicum (Panicum dichotomiflorum Michx.) and foxtail species (Anonymous 2017, Shaner 2014). Soltani et al.

(2017a) found metribuzin (659 g ha-1) controlled glyphosate-resistant Canada fleabane 80% at 4

45 weeks after application (WAA) while Byker et al. (2013b) reported that metribuzin (1120 g ha-1) plus glyphosate (900 g ae ha-1) controlled GR Canada fleabane 99%. Eubank et al. (2008) also reported that metribuzin (420 g ha−1) plus glyphosate (860 g ae ha−1) controlled GR Canada fleabane 53 to 60% at 4 WAA in the southern USA. Most commonly in Ontario metribuzin is applied as a tank mix partner prior to emergence of soybean.

Metribuzin is a photosynthesis inhibitor at PS II, site A. This group includes the triazines, triazinones, uracils, phenyl-carbamates, and pyridazinones (Shaner 2014). Metribuzin acts in the chloroplast thylakoid membrane where it binds to the QB-binding niche on the D1 protein of the

PS II complex displacing plastoquinone, effectively blocking electron transport (Shaner 2014).

This stops CO2 fixation and production of ATP and NADPH which are needed for plant growth.

There is a buildup of electrons at QA which results in the development of highly reactive oxygen species (ROS) mainly triplet chlorophyll and singlet oxygen (Hess 2000). The ultimate cause of plant death is due to lipid peroxidation of organelle and cell membranes, causing cells to leak and disintegrate.

Metribuzin can be applied as a soil or foliar-applied herbicide. It is apoplastically translocated in the xylem (Shaner 2014). The phytotoxic effects of metribuzin are influenced by temperature, humidity, light and plant transpiration rate (Devlin et al. 1987; Shaner 2014).

Metribuzin symptomology appears as chlorosis and necrosis of the leaves, with injury symptoms being most pronounced on the older leaves (Shaner 2014). The destruction of chlorophyll via the photo-oxidation results in leaf chlorosis, while leaf necrosis is caused by destruction of organelle and cell membranes (Hess 2000; Shaner 2014). The metabolism of metribuzin in tolerant species results in three major metabolites: deaminated-metribuzin (DA), diketo-metribuzin (DK), and

46 deaminated diketo-metribuzin (DADK) which are less harmful to the plants (Fedtke 1991;

Shaner 2014)

Metribuzin activity is influenced by soil characteristics which impacts weed control efficacy and crop safety. Metribuzin can leach on low organic matter, sandy soils but is less available for plant uptake on high organic matter clay soils (Savage 1976, Sharom and

Stephenson 1976). At greater soil depths there is reduced breakdown of metribuzin due to lower temperature and an associated decrease in microbial activity (Banks and Robinson 1982). Soil pH has effect on metribuzin plant uptake and mobility in the soil (Ladlie et al. 1976). In low pH soils, there is an increased amount of the protonated form of the alkaline metribuzin, resulting in an in increase in the amount of metribuzin that can be retained on the cation exchange capacity and reduced availability for the plant uptake which results in reduced crop injury and weed control efficacy. On higher pH soils, metribuzin is adsorbed less on soil colloids, which results in greater mobility and increased plant uptake (Shaner 2014). Metribuzin is primarily degraded through soil microbial activity. Metribuzin degradation is biphasic with a period of rapid degradation immediately after application followed by a slower degradation.

Metribuzin is a potential environmental hazard since it is detrimental to algae and other aquatic plant species, but rapid photolysis reduces the environmental impact of metribuzin in most aquatic environments (Shaner 2014). Like other triazine and triazinone herbicides, metribuzin is prone to runoff into surface waters due to its high water solubility (1100 mg L-1)

(Shaner 2014). Contamination of surface waters by metribuzin results from surface runoff to water courses and leaching into groundwater. Losses have been found to primarily occur through movement through soil runoff as opposed to translocation with eroded soil sediment (Glotfelty et

47 al. 1984). Metribuzin has been found freshwater in Ontario at a concentration of 0.8 to 18 μg L-1, far below the health reference level of 91µg L-1 (Frank et al. 1990; EPA 2003).

5.5 Tiafenacil

Tiafenacil is a new PPO inhibiting, pyrimidinedione, contact, non-selective herbicide, being developed by the company FarmHannong of South Korea (Duke et al. 1991; Park et al.

2018). Similar to other PPO-inhibiting herbicides, tiafenacil prevents chlorophyll and heme biosynthesis. The ultimate cause of plant death is due to increased levels of singlet oxygen and subsequent lipid peroxidation of the plant cells (Shaner 2014). Tiafenacil is being developed as a preplant burndown herbicide that has activity on both dicots and monocots. Other members of the pyrimidinedione family include butafenacil and saflufenacil, which were commercialized in

2001 and 2009, respectively. Both herbicides are primarily applied prior to crop emergence for control of emerged weeds in numerous crops; only saflufenacil has been commercialized in

Canada (Grossman et al. 2010)

Minimal research has been conducted on weed control efficacy of tiafenacil. Park et al.

(2018) found that the half-maximal inhibitory tiafenacil concentration (IC50), required for 50% inhibition in vitro, was 20 to 29 nM for waterhemp species, soybean, arabidopsis (Arabidopsis thaliana L.) and rapeseed (Brassica napus L.), while butafenacil and saflufenacil had a range of

26 to 52 nM. Acifluorfen was found to be 72- to 134-fold less potent than tiafenacil with an IC50 value of 2.0 to 2.9 μM. The pyrimidinediones, such as tiafenacil, may have increased binding affinity and herbicidal activity compared to other PPO inhibitors due to its heterocyclic structure

(Park et al. 2018; Narita et al. 1996). Tiafenacil has greater herbicidal effect on specific monocots when compared to saflufenacil (Park et al. 2018). Barnyardgrass was controlled with

30 μM of tiafenacil while over 50 μM of saflufenacil was required. In two-week-old corn plants, 48 treatment with 4 μM tiafenacil resulted in death; however, 8 μM of saflufenacil was not lethal to corn. Park et al. (2018) concluded that tiafenacil doses of less than 85 g ai ha-1 are sufficient to control both dicot and monocot weeds, which is approximately 5 to 40 μM concentrations of tiafenacil. Haring and Hanson (2020) from the University of California, Davis, found that tiafenacil (25 g ai ha-1) controlled Canada fleabane 71 and 29% at 7 and 31 days after treatment

(DAT), respectively. In the same study, tiafenacil (25 g ai ha-1) provided 84% control of velvetleaf, 50% control of prostrate pigweed (Amaranthus blitoides L.), 46% control of common purslane, and 17% control of common lambsquarters at 31 DAT. It should be noted that in the aforementioned study weeds were 30-60 cm in height at the time of application; efficacy may have reduced due to the advance growth stage of the weeds.

Section 6.0 – Hypotheses and Objectives

6.1 Hypotheses

1. Bromoxynil (280 g ai ha-1) + metribuzin (400 g ai ha-1) applied preplant (PP) will

control glyphosate-resistant (GR) Canada fleabane ≥80%.

2. Pyraflufen-ethyl/2,4-D (527 g ai ha-1) + metribuzin (400 g ai ha-1) applied PP will

control GR Canada fleabane ≥80%.

3. Tiafenacil (25 g ai ha-1) + metribuzin (400 g ai ha-1) applied PP will control GR

Canada fleabane ≥80%.

4. The addition of metribuzin (400 g ai ha-1) to bromoxynil, pyraflufen-ethyl/2,4-D, and

tiafenacil applied preplant will improve GR Canada fleabane control.

5. Metribuzin (560 g ai ha-1) applied preemergence (PRE) and postemergence (POST)

will control of PS II-inhibitor resistant waterhemp biotypes, with enhanced metabolism

mechanism of resistance, ≥80% control, while the same rate will not control PS II-

inhibitor resistant waterhemp biotypes with altered target-site mechanism of resistance

(serine264glycine within the D1 protein).

6.2 Objectives

1. To determine the biologically-effective-dose (BED) of bromoxynil and bromoxynil +

metribuzin applied PP for the control of GR Canada fleabane.

2. To determine the BED of pyraflufen-ethyl/2,4-D and pyraflufen-ethyl/2,4-D +

metribuzin applied PP for the control of GR Canada fleabane.

3. To determine the BED of tiafenacil and tiafenacil + metribuzin applied PP for the control of GR Canada fleabane in soybean.

4. To determine the BED of metribuzin, applied PRE and POST, for the control of waterhemp with two mechanism of resistance to the PS II-inhibiting herbicides (altered target-site and enhanced metabolism).

This manuscript was submitted to Weed Technology and accepted on January 26, 2021

Chapter 2: Biologically-Effective-Dose of Bromoxynil, Applied Alone and Mixed with Metribuzin, for the Control of Glyphosate-Resistant Canada fleabane [Conyza canadensis (L.) Cronq.] in Soybean

2.1 Abstract

Glyphosate-resistant (GR) Canada fleabane was first confirmed in Ontario in 2010. GR

Canada fleabane interference can reduce soybean yield up to 97%. Bromoxynil is a photosystem

II-inhibiting herbicide that is primarily used for annual broadleaf weed control in monocot crops.

The objective of this study was to determine the biologically-effective-dose (BED) of bromoxynil applied alone and when mixed with metribuzin applied preplant (PP) for control of

GR Canada fleabane in soybean in Ontario. Five field experiments were conducted over a two- year period (2019-2020) to determine the predicted dose of bromoxynil +/- metribuzin that controls GR Canada fleabane 50, 80 and 95%. No soybean injury was observed. The predicted doses of bromoxynil for 50 and 80% GR Canada fleabane control were 98 and 277 g ai ha-1, respectively; at 8 weeks after application (WAA). When mixed with metribuzin (400 g ai ha-1), the predicted doses of bromoxynil for 50, 80 and 95% GR Canada fleabane control were 10, 25, and 54 g ai ha-1, respectively. Bromoxynil (280 g ai ha-1) plus metribuzin (400 g ai ha-1) controlled GR Canada fleabane 97% which was similar to the industry standards of saflufenacil

+ metribuzin (99%) and glyphosate/dicamba + saflufenacil (100%) at 8 WAA. This study concludes that bromoxynil + metribuzin applied PP provides excellent control of GR Canada fleabane in soybean.

2.2 Introduction

Canada fleabane, Conyza canadensis (L.) Cronq., is a winter or summer annual that is member of the Asteraceae family (Weaver 2001). It has high fecundity, with seed production being proportional to plant height. Canada fleabane can grow up to 1.5 m in height and produce in excess of 230,000 small, wind-blown seeds per plant (Weaver 2001). Dispersal of seed by wind is enhanced by the attachment of a small pappus; seed can be dispersed over 500 km from the mother plant (Shields et al. 2006; Weaver 2001). Canada fleabane’s prolonged emergence pattern and small seed size make it well adapted to a no-till cropping system (Main et al. 2006).

Glyphosate-resistant (GR) Canada fleabane was first confirmed in Delaware, USA in

2001 (VanGessel 2001). In Canada, it was first confirmed in Essex County in 2010 (Byker et al.

2013a); it has now been confirmed in 30 counties across southern Ontario (Budd et al. 2017).

Worldwide, Canada fleabane is resistant to five modes of action (Heap 2020). In the USA,

Canada fleabane has evolved resistance to the acetolactate synthase (ALS)-inhibitors, photosystem II (PS II)-inhibitors, 5-enolpyruvyl shikimate-3-phosphate synthase (EPSPS)- inhibitors, and photosystem 1 diverters, representing WSSA Groups 2, 5, 9 and 22, respectively.

Canada fleabane interference can reduce the yield of agronomic crops. At densities of

100 to 200 plants m-2, Canada fleabane reduced soybean yield 90% (Bruce and Kells 1990). GR

Canada fleabane interference has been found to reduce soybean yield from 73 to 97% (Budd et al. 2016a; Byker et al. 2013b; Eubank et al. 2008; Hedges et al. 2018a).

The utilization of burndown plus residual herbicides applied in the spring are effective for controlling both fall and spring emerged GR Canada fleabane (Davis and Johnson 2008;

VanGessel 2001). Davis et al. (2010) reported an increase in spring-emerged Canada fleabane

54 when a non-residual herbicide was applied in the fall. GR Canada fleabane should be controlled prior to emergence to minimize yield loss in most soybean production systems (Bruce and Kells

1990; Byker et 2013c). Herbicides applied postemergence (POST) provide poor control of GR

Canada fleabane in identity-preserved (IP), non-genetically modified and glyphosate-resistant soybean. If required, dicamba and 2,4-D choline, applied POST can provide acceptable control in dicamba-resistant (XtendTM) and 2,4-D-resistant (EnlistTM) soybean, respectively.

Bromoxynil, a member of the benzonitrile family, is a PS II-inhibitor (WSSA Group 6).

Bromoxynil acts in the chloroplast thylakoid membrane where the herbicide binds with histidine

215 within the QB-binding niche on the D1 protein of the PS II complex which stops electron transport resulting in the production of free radicals and subsequent lipid peroxidation (Trebst

1987; Shaner 2014). In Canada and the USA, bromoxynil is primarily used for annual broadleaf weed control in wheat (Triticum aestivum L.), oat (Avena sativa L.), barley (Hordeum vulgare

L.) and field corn (Zea mays L.) (Anonymous 2019a). Bromoxynil is classified as a contact herbicide with limited translocation in most species (Shaner 2014). Bromoxynil controls a wide range of annual broadleaf weeds at doses of 210 to 560 g ai ha-1 and is often used in combination with other active ingredients, in order to provide effective control of GR Canada fleabane

(Anonymous 2019a; Shaner 2014). In such instances, a reduced rate of bromoxynil can be used.

For example, bromoxynil + pyrasulfotole (174 + 31 g ai ha-1) controlled GR Canada fleabane

97% in winter wheat at 8 weeks after application (WAA) (Mahoney et al. 2016a). Metzger et al.

(2019) reported that bromoxynil + atrazine (280 + 1500 g ai ha-1) controlled multiple-herbicide- resistant (MHR) Canada fleabane 94% in corn at 8 WAA. The aforementioned studies suggest that there is potential for bromoxynil applied preplant (PP) in soybean for GR Canada fleabane control.

Metribuzin is an asymmetrical triazine (WSSA Group 5), PS II-inhibiting herbicide that is commonly applied as a mix partner to improve annual broadleaf weed control in soybean

(Budd et al. 2016a; Eubanks et al. 2008; Shaner 2014). Similar to bromoxynil, metribuzin also works within the QB-binding niche on the D1 protein of the PS II complex but binds with serine at position 264 (Tietjen et al. 1991). Metribuzin has provided variable control of GR Canada fleabane when applied alone. Eubanks et al. (2008) found that glyphosate + metribuzin (860 +

420 g ai ha-1) controlled 15 to 30 cm tall GR Canada fleabane 53 and 63% in 2005 and 2006, respectively in Mississippi, USA. Tardif and Smith (2003) reported that metribuzin at 1120 g ai ha-1 controlled 1 to 10 cm tall GR Canada fleabane 73%, while Byker et al. (2013d) reported the same rate provided 97 to 99% control of 8 to 11 cm tall GR Canada fleabane. When mixed with other burndown herbicides including saflufenacil, 2,4-D ester, pyraflufen-ethyl/2,4-D, or S- metolachlor/dicamba, metribuzin (400 g ai ha-1) improved the control of GR Canada fleabane

(Budd et al. 2016a; Soltani et al. 2020a).

Biologically-effective-dose (BED) is defined as the dose that provides a specific response level (eg- ED50, 80, 95) of control, reduction in density, or reduction in biomass (Dieleman et al.

1996; Knezevic et al. 2007). Knezevic et al. (1998) suggested determining the BED can reduce the use of excess amounts of herbicides, reducing the impact on the environment and maximizing profits for growers. There is little published research on the BED of bromoxynil applied alone and mixed with metribuzin for the control of GR Canada fleabane. The control of

GR Canada fleabane does not appear on the bromoxynil label in Canada and it is not registered as a PP application prior to soybean planting (Anonymous 2019a). The objective of this study was to determine the BED of bromoxynil applied alone and mixed with metribuzin for control of GR Canada fleabane applied PP to soybean in Ontario.

2.3 Materials and Methods

Experimental Methods

Field experiments were completed over a two-year period (2019, 2020) in five different commercial farm fields in southwestern Ontario with previously confirmed GR Canada fleabane populations. In this study, treatments within each experiememt were arranged in a randomized complete block design (RCBD) with four replications, each replicate included a weedy (non- treated) control. Glyphosate (450 g ha-1) was applied POST to the entire experimental area to remove the confounding effect of susceptible Canada fleabane biotypes and all other weed species. Glyphosate/dicamba-resistant soybean (DKB 12-16) were no-tilled in early-May to early-June at a rate of approximately 416 000 seeds ha-1. Plots measured 2.25 m wide (3 rows spaced 75 cm apart) by 8 m long with a 2 m alley between replicates.

To determine the BED of bromoxynil and bromoxynil plus metribuzin, a rate titration was utilized with bromoxynil doses of 35, 70, 140, 280, 560, 1120 g ai ha-1 and a titration of bromoxynil plus metribuzin of 35 + 400, 70 + 400, 140 + 400, 280 + 400, 560 + 400, and 1120 +

400 g ai ha-1. Saflufenacil + metribuzin (25 + 400 g ai ha-1) and glyphosate/dicamba + saflufenacil (1800 + 25 g ai ha-1) were included to represent the current industry standards for control of GR Canada fleabane applied PP in glyphosate and glyphosate/dicamba-resistant soybean, respectively (Budd et al. 2016a; Hedges et al. 2018a).

Herbicides were applied using a CO2 pressurized backpack sprayer equipped with four

ULD 11002 nozzles (Pentair, New Brighton, MN, USA) calibrated to deliver 200 L ha-1 at a pressure of 260 kPa. Herbicide treatments were applied PP when GR Canada fleabane reached an average of 10 cm in height. Soybean was planted 1 to 10 days after herbicide application.

Trial year, location, herbicide application date, Canada fleabane size and density at application, soybean planting, emergence and harvest dates, and soil characteristics are presented in Table 1.

Visible GR Canada fleabane control was assessed at 2, 4, and 8 WAA on a scale of 0 (no control) to 100 (complete control), relative to the non-treated control. Crop injury was assessed at 1, 2 and 4 weeks after emergence (WAE) on a scale of 0 to 100; 0 represented no visible soybean injury and 100 represented complete plant death. Weed density and aboveground biomass were determined at 8 WAA and expressed as a percent of the non-treated control within each replicate. Canada fleabane density within each plot was determined from two randomly placed 0.25 m2 quadrats, GR Canada fleabane plants were counted, cut at the soil surface, placed in a paper bag and dried in a kiln to a constant moisture and weighed to determine aboveground biomass. The centre two rows of each plot were combined at maturity of the soybean crop, with moisture content and weight of soybean recorded. Soybean seed yields were adjusted to 13.5% seed moisture content prior to statistical analysis.

Statistical Analysis

Nonlinear Regression

Visible weed control at 2, 4 and 8 WAA and soybean yield [expressed as percent of the yield of the industry standard (glyphosate/dicamba + saflufenacil) within each replicate] were regressed against bromoxynil and bromoxynil + metribuzin by specifying an exponential to maximum model (Equation 1) within PROC NLIN in SAS v 9.4 (SAS Institute, Cary, NC).

Weed density and biomass, expressed as a percent of the non-treated control within each replicate, were regressed using the inverse exponential model (Equation 2). The parameters generated from each regression analysis were used to calculate the BED (ED50, ED80, ED95) of

58 bromoxynil and bromoxynil + metribuzin to achieve 50, 80, and 95% control, reduction in density, biomass and soybean yield. Where a predicted dose could not be calculated by the model, it was expressed as non-estimable (Non-est).

Equation 1: Exponential to maximum

y = a - b (e -c dose)

Where a is the upper asymptote, b is the magnitude, and c is the slope

Equation 2: Inverse exponential

y = a + b (e –c dose)

Where a is the lower asymptote, b is the reduction in y from intercept to asymptote, and

c is the slope

Least-Square means comparisons

The data across all five site-years were pooled for analysis to determine the BED and to compare with the industry standards. PROC GLIMMIX was used in SAS v. 9.4 (SAS Institute,

Cary, NC) to perform a mixed-model variance analysis. Variance was partitioned into the fixed effect of herbicide treatment and the random effects of environment, replication within environment, and the treatment-by-environment interaction. The Shapiro-Wilk test of normality was used to check for normality among the data along with visual assessments of studentized residuals and scatterplots. Residuals of visible GR Canada fleabane control data at 2, 4 and 8

WAA and yield data followed a normal distribution. The non-treated control data were not

59 included in the analysis. Biomass and density data were assigned a lognormal distribution; means were back transformed using the omega method within PROC GLIMMIX (M. Edwards, OAC

Statistician, University of Guelph, personal communication). Least-square means were separated using the Tukey-Kramer multiple-range test (α = 0.05).

2.4 Results and Discussion

2.4.1 Soybean Injury

No soybean injury was observed at 1, 2 and 4 weeks after emergence (data not presented).

2.4.2 Biologically-Effective-Dose of Bromoxynil and Bromoxynil + Metribuzin for the Control of GR Canada fleabane

Bromoxynil Applied Alone

The predicted doses of bromoxynil to control GR Canada fleabane 50, 80 and 95% were

53, 148 g ai ha-1 and non-estimable, respectively, at 2 WAA (Table 2). The predicted dose of bromoxynil to control GR Canada fleabane 50, 80 and 95% increased to 70, 197 g ai ha-1 and non-estimable, respectively, at 4 WAA, and 98 and 277 g ai ha-1 and non-estimable, respectively, at 8 WAA. The predicted doses of bromoxynil to reduce GR Canada fleabane density 50, 80 and

95% were 126, 287 and 509 g ai ha-1, respectively, at 8 WAA. The predicted doses of bromoxynil to reduce GR Canada fleabane biomass were 259, 565 and 1027 g ai ha-1, respectively, at 8 WAA. The predicted doses of bromoxynil that would result in 50, 80 and 95% of the yield of the industry standard were non-estimable, 177 and 497 g ai ha-1, respectively.

Similar rates of bromoxynil have been reported to be effective on other weed species.

Corbett et al. (2004) found that bromoxynil at 420 and 560 g ai ha-1 controlled velvetleaf

(Abutilon theophrasti Medik.) 93 and 98%, respectively. Jordan et al. (1993) reported that bromoxynil at 560 to 840 g ai ha-1 controlled velvetleaf. Bromoxynil (420 g ai ha-1) controlled

Palmer amaranth (Amaranthus palmeri S. Wats.) and redroot pigweed (Amaranthus retroflexus

L.) 55 and 53%; increasing the rate to 560 g ai ha-1 improved control to 68% and 76% control, respectively (Corbett et al. 2004). Culpepper and York (1997) found that bromoxynil at similar rates to the aforementioned study controlled Palmer amaranth 74-96 % dependent upon location and year. Corbett et al. (2004) found that bromoxynil at 420 and 560 g ai ha-1 controlled common ragweed 90 and 96%, respectively; Ganie and Jhala (2017) reported that bromoxynil (420 g ai ha-1) controlled common ragweed 99% 3 WAA. Bromoxynil at 420 and 560 g ai ha-1 controlled common lambsquarters (Chenopodium album L.) 94 and 97%, respectively (Corbett et al. 2004).

Culpepper and York (1997) reported that bromoxynil (400 g ai ha-1) controlled common lambsquarters 99% in bromoxynil-resistant cotton. Results from this study found that the rate of bromoxynil required for the control of GR Canada fleabane is similar to many other annual broadleaf weed species.

The predicted bromoxynil doses for a reduction in GR Canada fleabane density and biomass were higher than the predicted dose for GR Canada fleabane control at 8 WAA (Table

2). The predicted dose of bromoxynil for 50% control and decrease in density and biomass was

98, 126 and 259 g ai ha-1, respectively; the same trend was observed for 80% control and decrease in density and biomass. Ganie and Jhala (2017) reported that bromoxynil (420 g ai ha-1) applied POST in corn reduced GR common ragweed biomass 80%. The higher predicted dose of bromoxynil to achieve a 50, 80, and 95% reduction in biomass compared to density is because

61 plants that survived bromoxynil application were larger at application resulting in increased biomass at 8 WAA; relatively few plants survived but those that did rapidly accumulated biomass. Similar results were reported by Budd et al. (2016a) who found that a higher dose of saflufenacil and metribuzin was required to achieve 90 and 95% reduction in biomass when compared to density.

Bromoxynil at 177 and 497 g ai ha-1 resulted in 80 and 95% of the soybean yield of the industry standard (glyphosate/dicamba + saflufenacil), respectively (Table 2). ED50 could not be calculated because GR Canada fleabane interference did not reduce soybean yield 50% in this study.

Bromoxynil + Metribuzin

The predicted doses of bromoxynil + metribuzin to control GR Canada fleabane 50, 80 and 95% was 9 to 10, 22 to 25, and 47 to 54 g ai ha-1 respectively at 2, 4 and 8 WAA (Table 2).

The predicted dose of bromoxynil + metribuzin to reduce GR Canada fleabane density 50, 80 and 95% was 8, 19, and 36 g ai ha-1 and biomass was 14, 33, and 61 g ai ha-1, respectively at 8

WAA. The predicted dose of bromoxynil that would result in 80 and 95% soybean yield of the industry standard was 17 and 38 g ai ha-1, respectively. Limited research exists on the efficacy of bromoxynil + metribuzin. The addition of metribuzin to other herbicides applied preplant in soybean has resulted in improved control of GR Canada fleabane. The addition of metribuzin

(420 g ai ha-1) to paraquat (840 g ai ha-1) increased GR Canada fleabane control from 55 to 94% and 55 to 79% 4 WAA in 2005 and 2006, respectively (Eubank et al. 2008). The addition of metribuzin (400 g ai ha-1) to saflufenacil (25 g ai h ha-1) improved GR Canada fleabane control from 96 to 99% (Budd et al. 2016a) and 93 to 98% control 8 WAA (Soltani et al. 2020a).

Similar to bromoxynil alone, the predicted doses of bromoxynil + metribuzin for a reduction in GR Canada fleabane biomass were slightly higher than the predicted doses for GR

Canada fleabane control and density reduction at 8 WAA (Table 2). The predicted doses of bromoxynil + metribuzin for 50% control and decrease in density and biomass were 10, 8 and 14 g ai ha-1, respectively; the same trend was observed for the predicted doses at 80% and 95%. The authors suggest that the higher predicted dose for reducing biomass was because the few GR

Canada fleabane that survived were larger in size.

Bromoxynil + metribuzin at 17 and 38 g ai ha-1 resulted in 80 and 95% of the soybean yield of the industry standard (glyphosate/dicamba + saflufenacil), respectively. ED50 could not be calculated because GR Canada fleabane interference did not reduce soybean yield by 50% in this study.

2.4.3 Bromoxynil, Metribuzin, Bromoxynil + Metribuzin Compared to Industry Standards

Bromoxynil (280 g ai ha-1), metribuzin (400 g ai ha-1) and bromoxynil + metribuzin (280

+ 400 g ai ha-1) were compared to the industry standards of saflufenacil + metribuzin (25 + 400 g ai ha-1) and glyphosate/dicamba + saflufenacil (1800 + 25 g ai ha-1). The industry standard of saflufenacil + metribuzin was selected based on research by Budd et al. (2016a) for GR Canada fleabane control in GR and identity-preserved (IP), non-genetically engineered soybean; glyphosate/dicamba + saflufenacil was selected based on research by Hedges et al. (2018a) for

GR Canada fleabane control in glyphosate/dicamba-resistant soybean. Metribuzin or bromoxynil, applied alone, controlled GR Canada fleabane similarly at 76 to 80 and 77 to 85%, respectively. The mixes of bromoxynil + metribuzin, saflufenacil + metribuzin and glyphosate/dicamba + saflufenacil controlled GR Canada fleabane 97 to 100% at 2, 4 and 8

WAA; there was no difference in control among the three mixes evaluated (Table 3). Hamill and 63

Zhang (1995) reported that bromoxynil (280 g ai ha-1) plus low rates of metribuzin (100 g ai ha-1) provided effective control of annual broadleaf weeds. Soltani et al. (2020a) found glyphosate/dicamba + saflufenacil and saflufenacil + metribuzin controlled GR Canada fleabane

97 and 98%, respectively. GR Canada fleabane control with bromoxynil declined over time from

85 to 77% at 2 and 8 WAA, respectively; the decrease in control was likely due to the incomplete death of fall-emerged Canada fleabane. There is little research on the use of bromoxynil applied alone for the control of GR Canada fleabane. Since bromoxynil is a contact herbicide with no residual activity, the authors attribute the decline in control over time due to recovery of large weeds present at the time of application. Eubanks et al. (2008) reported that metribuzin (420 g ai ha-1) controlled GR Canada fleabane 53 to 63% at 4 WAA, which was considerably lower than the results from this study. Byker et al. (2013c) found that when metribuzin was applied at a much higher rate of 1120 g ai ha-1, GR Canada fleabane control was

97 to 99 % at 8 WAA, which was similar to Soltani et al. (2016), who reported 91% GR Canada fleabane control with 1120 g ai ha-1 of metribuzin. The improved control in those two studies can be attributed to the much higher rate of metribuzin. GR Canada fleabane control with metribuzin remained relatively constant in this study at 2, 4 and 8 WAA, which was similar to the findings of Byker et al. (2013d). The mixes of bromoxynil + metribuzin, saflufenacil + metribuzin and glyphosate/dicamba + saflufenacil provided similar GR Canada fleabane control (97 to 100%) at

2, 4 and 8 WAA.

Bromoxynil and metribuzin reduced GR Canada fleabane density 84 to 87% and biomass

64 to 69%; there was no difference between the two herbicides (Table 3). Bromoxynil + metribuzin, saflufenacil + metribuzin, and glyphosate/dicamba + saflufenacil reduced GR

Canada fleabane density 99 to 100% and biomass 92 to 100%; there were no differences among the three mixes evaluated.

GR Canada fleabane interference reduced soybean yield up to 48% in this study (Table

3). Reduced GR Canada fleabane interference with the herbicide treatments evaluated resulted in higher soybean yield. Although there were differences in GR Canada fleabane control, density, and biomass, there were no statistical differences detected in soybean yield across herbicide treatments. Soltani et al. (2020a) reported similar results with few differences in soybean yield due to GR Canada fleabane interference among a number of herbicide treatments evaluated.

2.5 Conclusion

In conclusion, bromoxynil, metribuzin, bromoxynil + metribuzin, saflufenacil + metribuzin and glyphosate/dicamba + saflufenacil applied PP caused no visible soybean injury in this study. Bromoxynil applied alone provided marginal control of GR Canada fleabane; the predicted doses for 50, 80 and 95% control were 98, 277 g ai ha-1 and non-estimable, respectively, at 8 WAA. The addition of metribuzin to bromoxynil improved GR Canada fleabane control; the predicted doses for 50, 80 and 95% control were reduced to 10, 25 and 54 ai ha-1 at 8 WAA. Bromoxynil + metribuzin (280 + 400 g ai ha-1) applied PP provided similar GR

Canada fleabane control to the industry standards. Saflufenacil + metribuzin and glyphosate/dicamba + saflufenacil controlled GR Canada fleabane 97 to 100% at 2, 4 and 8

WAA. This study indicates that bromoxynil + metribuzin applied preplant provides comparable

GR Canada fleabane control to the current industry standards.

cm plants m-2 %

Bothwell June 7 7 172 June 8 June 14 October 10 Sand 2.6 6.7

2019 Moraviantown June 18 7.5 368 June 19 June 25 October 8 Sand 2.2 6.1

Thamesville June 11 9 80 June 12 June 20 October 7 Loamy 1.8 5.6 sand

Ridgetown May 26 6.5 599 June 5 June 10 October 1 Sandy 1.9 7.1 2020 loam Duart June 1 8 318 June 5 June 10 October 6 Sandy 4.8 7.3 loam aAbbreviation: OM, organic matter. bSize measured as height of bolting plants. Mean of eight measurements per experiment at time of application. cMean density based on two stand counts per replication within each experiment.

Table 2.2 Regression parameters and predicted effective dose of bromoxynil and bromoxynil plus metribuzin for 50, 80 and 95% glyphosate- resistant Canada fleabane [Conyza canadensis (L.) Cronq.] control at 2, 4 and 8 weeks after application (WAA), for 50, 80 and 95% reduction in biomass and plant density at 8 WAA and to achieve 50, 80 and 95% of the yield from five field experiments completed in 2019 and 2020 in Ontario, Canada.

Parameter estimates Predicted bromoxynil dose Variable a b c ED50 ED80 ED95

------±SE ------g ai ha-1 ------Control 2 WAAa 89.45 (1.71) 86.99 (3.15) 0.015 (0) 53 148 Non-est.c

Control 4 WAAa 89.89 (2.10) 86.23 (3.54) 0.011 (0) 70 197 Non-est.

Control 8 WAAa 89.34 (2.60) 85.94 (3.87) 0.008 (0) 98 277 Non-est. Densityb 0.00 (0) 106.30 (7.51) 0.006 (0) 126 287 509

Dry Biomassb 0.00 (0) 108.80 (6.76) 0.003 (0) 259 565 1027

Yieldad 98.80 (8.88) 45.62 (10.36) 0.005 (0) Non-est. 177 497

Predicted bromoxynil + metribuzin dose

Variable a b c ED50 ED80 ED95

------±SE ------g ai ha-1 ------

Control 2 WAAa 96.87 (0.78) 96.85 (1.87) 0.079 (0) 9 22 50

Control 4 WAAa 97.42 (0.78) 97.40 (1.87) 0.079 (0) 9 22 47

Control 8 WAAa 97.28 (0.91) 97.26 (2.15) 0.069 (0) 10 25 54

Densityb 0 (0) 99.99 (1.75) 0.084 (0) 8 19 36

Dry Biomassb 0 (0) 99.70 (3.44) 0.049 (0) 14 33 61

Yieldad 105.1 (4.28) 51.50 (9.62) 0.043 (0.02) Non-est. 17 38 aRegression parameters: y = a - b (e -c dose); Where a is the upper asymptote, b is the magnitude, and c is the slope. bRegression parameters: y = a + b (e –c dose); Where a is the lower asymptote, b is the reduction in y from intercept to asymptote, and c is the slope. cNon-est., predicted dose for parameter could not be computed by the model. dExpressed as percent of yield in the industry standard [glyphosate/dicamba + saflufenacil (1800 + 25 g ai ha-1)] among replications. 67

Non-treated controlc - - - - 0 c 0 c 1.6 b

Metribuzin 400 78 b 80 b 76 b 84 b 64b 2.5 a

Bromoxynil 280 85 b 83 b 77 b 87 b 69 b 2.6 a

Bromoxynil + metribuzin 280 + 400 97 a 98 a 97 a 99 ab 92 ab 2.9 a

Saflufenacil + metribuzin b 25 + 400 99 a 99 a 99 a 100 a 98 a 3.1 a

Glyphosate/dicamba + 1800 + 25 99 a 100 a 100 a 100 a 100 a 3.1 a saflufenacil b aMeans followed by the same letter within a column do not significantly differ from one another according to Tukey-Kramer’s multiple range test α=0.05. bIncluded Merge® (0.5% vol/vol). cNon-treated control plots received only glyphosate (900 g ae ha-1). dDensity and biomass of GR Canada fleabane expressed as a percent reduction of the non-treated control within replications.

------% ------% % Non-treated Controlc - - - - 100 a 100 ab Metribuzin 400 78 cd 80 cd 76 cd 16 cd 36 cd Bromoxynil 35 43 a 39 a 31 a 100 a 104 a Bromoxynil 70 61 b 54 ab 42 ab 60 b 99 ab Bromoxynil 140 72 bc 66 bc 60 bc 44 bc 63 bc Bromoxynil 280 85 cde 83 cde 77 cde 13 cd 31 cd Bromoxynil 560 88 cde 87 de 84 cde 11 cd 2 cd Bromoxynil 1120 98 e 97 de 96 de 3 d 5 d Bromoxynil + metribuzin 35 + 400 91 de 92 de 89 de 5 d 14 d Bromoxynil + metribuzin 70 + 400 92 de 93 de 94 de 4 d 12 d Bromoxynil + metribuzin 140 + 400 96 e 96 de 95 de 4 d 10 d Bromoxynil + metribuzin 280 + 400 97 e 98 de 97 de 1 d 8 d Bromoxynil + metribuzin 560 + 400 99 e 100 e 99 de 0 d 3 d Bromoxynil + metribuzin 1120 +400 99 e 100 e 100 e 0 d 2 d a Density and biomass expressed as a percent of the non-treated control within replications. b Means followed by a different letter within a column differ significantly according to a Tukey-Kramer multiple range test at α =0.05. c Non-treated control plots received only glyphosate (900 g ae ha-1)

This manuscript was submitted to Weed Technology and is currently under review

3.1 Abstract

Glyphosate resistance in weed species has presented immense challenges for farmers in

Ontario. Glyphosate-resistant (GR) Canada fleabane was first confirmed in Ontario in 2010 and is the most challenging GR species for soybean producers to control. In Ontario, yield reduction due to GR Canada fleabane interference has been reported to be up to 93%. Utilization of burndown plus residual herbicides has proven successful for the control of GR Canada fleabane in soybean. Pyraflufen-ethyl/2,4-D is a premixed herbicide formulation sold under the tradename

BlackHawk®. Five field experiments were conducted over a two-year period (2019, 2020) in commercial fields in southwestern Ontario to establish the biologically-effective-dose of pyraflufen-ethyl/2,4-D, applied alone, or tankmixed with metribuzin, for the control of GR

Canada fleabane applied preplant to soybean. Soybean visible injury was <15%. The calculated doses of pyraflufen-ethyl/2,4-D required for 50, 80, and 95% GR Canada fleabane control were

390, 1148, and >2108 g ai ha-1, respectively at 8 weeks after application (WAA). The addition of metribuzin (400 g ai ha-1) to pyraflufen-ethyl/2,4-D reduced the dose of pyraflufen-ethyl/2,4-D

-1 for 50, 80, and 95% GR Canada fleabane control to 19, 46, 201 g ai ha , respectively at 8 WAA.

Pyraflufen-ethyl/2,4-D + metribuzin (527 + 400 g ai ha-1) controlled GR Canada fleabane 97%, similar to the current industry standards of saflufenacil + metribuzin (25+ 400 g ai ha-1) and glyphosate/dicamba + saflufenacil (1800 + 25 g ai ha-1) which provided 99% control at 8 WAA.

The results of this study conclude that pyraflufen-ethyl/2,4-D + metribuzin (527 + 400 g ai ha-1) applied preplant in soybean provided similar GR Canada fleabane control to the current industry standards. 71

3.2 Introduction

Canada fleabane is a winter or summer annual weed that is native to North America

(Weaver 2001). Canada fleabane does not have dormancy requirements which allows for germination and establishment shortly after seed release from the mother plant (Buhler and

Owen 1997). Fall germinated seeds develop a winter annual growth pattern; forming rosettes that overwinter followed by stem elongation in the spring. Seeds that germinate in the spring bypass the rosette stage; stem elongation commences shortly after emergence (Weaver 2001). Canada fleabane’s extended and variable emergence pattern make control difficult, as weed size at application is variable (Buhler and Owen 1997; Kruger et al. 2010a). Canada fleabane has the ability to produce large amounts of seed; a 1.5 m tall plant has been reported to produce up to

230,000 per plant (Weaver 2001). The seeds germinate close to the soil surface with a maximum depth of seedling recruitment of 0.5 cm, resulting in suitable conditions for germination in no-till cropping systems (Nandula et al. 2006). Canada fleabane seed is commonly dispersed via the wind; with the majority of seed falling within 100 m of the source but it has been found in the planetary boundary layer and can move up to 500 km (Dauer 2007; Shields et al. 2006)

Glyphosate resistance is a global issue in many cropping systems. Glyphosate [N-

(phosphonomethyl) glycine] is broad-spectrum, systemic herbicide that inhibits 5- enolpyruvylshikimate 3-phosphate synthase (EPSPS) resulting in a blocked shikimic acid pathway (Duke and Powles 2008; Shaner 2014). Farmers’ adoption of glyphosate-resistant (GR)

(Roundup Ready®) crops has grown rapidly since the release of GR canola and soybean in 1996

(Owen 2008). The utilization of GR crop technology has allowed for an increase in the number of glyphosate applications per growing season, which increased the evolution of GR weed

72 biotypes (Vencill et al. 2012). Currently there are 51 different GR weed species documented worldwide (Heap 2020).

GR Canada fleabane was first reported in 2001 in Delaware, USA (VanGessel 2001). In

Ontario, GR Canada fleabane was first identified in 2010 in Essex County, and as of 2015, it has been found across southern Ontario from the Michigan to the Quebec border (Budd et al. 2017;

Byker et al. 2013a). There are target-site and non-target-site mechanisms of resistance that confer resistance to glyphosate in Canada fleabane. Non-target-site mechanisms of resistance include vacuolar sequestration and enhanced metabolism (Ge et al. 2010; González-Torralva et al. 2012). Recently, Page et al. (2018) identified a target-site mutation of a proline to serine at position 106 of the EPSPS enzyme in the majority of GR Canada fleabane accessions in Ontario.

GR Canada fleabane interference can reduce soybean yield dramatically. Both Bruce and

Kells (1990) and Byker et al. (2013b) report soybean yield reductions of >90% with Canada fleabane densities of 100 to 200 plants m-2. Postemergence (POST) herbicides have been reported to not sufficiently control of GR Canada fleabane in identity-preserved (IP), non- genetically modified and GR soybean (Byker et al. 2013c). Therefore, control of GR Canada fleabane should be accomplished using preplant herbicides that provide control of emerged

Canada fleabane and full-season residual control to minimize soybean yield loss (Bruce and

Kells 1990). The POST use of dicamba in glyphosate/dicamba-resistant soybean and 2,4-D choline in glyphosate/2,4-D-resistant soybean have activity on GR Canada fleabane. Burndown plus residual herbicides applied prior to the planting or emergence of soybean is recommended for the control of fall and spring emerged GR Canada (Budd et al. 2016a; Davis and Johnson

2008).

Pyraflufen-ethyl/2,4-D is a pre-formulated herbicide sold under the trade name

BlackHawk® that consists of pyraflufen-ethyl (6.1 g L-1) and 2,4-D ester (473 g a.e L-1).

Pyraflufen-ethyl/2,4-D is currently registered for the control of emerged broadleaf weeds and can be applied prior to emergence of barley (Hordeum vulgare L.), field corn (Zea mays L.), oat

(Avena sativa L.), rye (Secale cereale L.), soybean (Glycine max L.), and spring and winter wheat (Triticum aestivum L.) (Anonymous 2019b). Pyraflufen-ethyl is classified as a protoporphyrinogen IX oxidase (PPO)-inhibitor that belongs to the phenylpyrazole chemical family with plants often displaying necrosis of foliage prior to plant death (Shaner 2014).

Pyraflufen-ethyl provides excellent control of broadleaf weeds such as cleavers (Galium aparine

L.) and common chickweed [Stellaria media (L.) Vill.] (Mabuchi et al. 2002). There is a limited amount of published research that exists on GR Canada fleabane control with pyraflufen- ethyl/2,4-D (532 g ai ha-1); one study reported variable and generally unsatisfactory control of

60% at 8 weeks after application (WAA) (Soltani et al. 2020a). 2,4-D, a member of the phenoxy carboxylic acids, is a commonly used synthetic auxin, or growth regulator, herbicide for broadleaf weed control. Eubank et al. (2008) reported that 2,4-D + glyphosate (840 + 860 g ae ha-1) provided 99 and 95% control of GR Canada fleabane at 4 WAA in 2005 and 2006, respectively. Kruger et al. (2010a) reported that 2,4-D controlled 30 cm tall Canada fleabane 90 to 97%. In Ontario, 2,4-D (500 g ae ha-1) applied preplant to soybean has provided inconsistent control of GR Canada fleabane from 53% to 92% at 8 WAA (Byker et al. 2013d; Soltani et al.

2020b). Glyphosate/2,4-D choline (1720 g ae ha-1) applied preplant in corn provided 79% control of GR Canada fleabane 8 WAA, while a sequential application of glyphosate/2,4-D choline

(1720 g ae ha-1) provided 98 to 100% control at 8 WAA (Ford et al. 2014).

Metribuzin is an annual broadleaf herbicide that is classified as a photosystem II (PS II)- inhibitor. Metribuzin is commonly used in combination with other herbicides at a preplant timing in no-till soybean (Budd et al. 2016a; Soltani et al. 2016). The co-application of metribuzin with saflufenacil, glufosinate, pyraflufen-ethyl/2,4-D, s-metolachlor/dicamba and 2,4-D provides good control of GR Canada fleabane (Budd et al. 2016a; Eubank et al. 2008; Soltani et al.

2020a). Soltani et al. (2020a) reported that 2,4-D ester (500 g ae ha-1) or pyraflufen-ethyl/2,4-D

(532 ai ha-1) tankmixed with metribuzin (400 g ai ha-1) provided 91 and 95% control respectively of GR Canada fleabane at 8 WAA.

Little published information exists on the biological activity of pyraflufen-ethyl/2,4-D, applied alone, or tankmixed with metribuzin, for the control of GR Canada fleabane. The objective of this study was to determine the biologically-effective-dose (BED) of pyraflufen- ethyl/2,4-D applied alone, and tankmixed with metribuzin, for control of GR Canada fleabane applied preplant (PP) to soybean in Ontario.

3.3 Materials and Methods

Experimental Methods

Five experiments, set up as a randomized complete block design (RCBD), were conducted in commercial fields in southwestern Ontario over two years (2019, 2020) at sites with confirmed GR Canada fleabane populations (Table 3.1). Plots measured 2.25 m in width (3 soybean rows spaced 0.75 m apart) and 8 m in length. To determine the BED, pyraflufen- ethyl/2,4-D was applied at doses of 66, 132, 264, 527, 1054, and 2108 g ai ha-1 and to determine the dose of pyraflufen-ethyl/2,4-D plus metribuzin, pyraflufen-ethyl/2,4-D was applied at the

75 aforementioned doses with metribuzin at 400 g ai ha-1. The current industry standards of saflufenacil + metribuzin (25 + 400 g ai ha-1) was derived from research by Budd et al. (2016a) for the control of GR Canada fleabane in identity-preserved (IP), non-genetically engineered and

GR soybean; glyphosate/dicamba + saflufenacil (1800 + 25 g ai ha-1) was included based on research by Hedges et al. (2018a) for the control of GR Canada fleabane in glyphosate/dicamba- resistant soybean. Glyphosate (900 g ae ha-1) was applied with all treatments to eliminate all non-

GR Canada fleabane plants. A backpack CO2 pressurized sprayer was used for all herbicide treatment applications. The sprayer was equipped with a 1.5 m hand-held boom, producing a 2 m spray width, and was set to 260 kPa to deliver a water volume of 200 L ha-1. The spray boom was equipped with four ULD 11002 nozzles (Pentair, New Brighton, MN, USA), spaced 50 cm apart which produced a 2 m spray width. Treatments were applied once the diameter or height of the GR Canada fleabane had reached 10 cm. All herbicide treatments were applied PP with soybean planted 1-10 days after herbicide applications. Glyphosate/dicamba-resistant soybean

(DKB 12-16) were planted using a no-till planter in early-May to early-June at approximately

416 000 seeds ha-1. Trial year, location, herbicide application date, Canada fleabane size and density at application, soybean planting, emergence and harvest dates, and soil characteristics are presented in Table 3.1. The previous crop at all sites was soybean. An application of glyphosate

(450 ae ha-1) was made to the entire experimental area after soybean emergence using a tractor mounted sprayer to remove susceptible GR Canada fleabane biotypes and non-target weed species.

Crop injury was assessed visually at 1, 2 and 4 weeks after emergence (WAE) on a scale of 0 to 100, with 0 representing no soybean injury and 100 total plant death. GR Canada fleabane control was assessed visually at 2, 4 and 8 WAA on a scale of 0 (no control) to 100 (complete

76 control); control ratings are a visual estimate of the GR Canada fleabane biomass reduction relative to the non-treated control within each replicate. Canada fleabane density and aboveground biomass were collected at 8 WAA and expressed as plants m-2 and g m-2, respectively. Canada fleabane density within each plot was determined from two randomly placed 0.25 m2 quadrats, GR Canada fleabane plants were counted, cut at the soil surface, collected and dried to a uniform moisture and weighed to determine dry biomass. Soybean yield was collected upon harvest maturity by combining the centre two soybean rows of each plot.

Soybean moisture content was collected and adjusted to 13.5% moisture prior to statistical analysis. Soybean weight was recorded and reported as tonnes ha−1.

Statistical Analysis

Non-linear Regression

Data from the rate titration of pyraflufen-ethyl/2,4-D and pyraflufen-ethyl/2,4-D + metribuzin were analyzed using nonlinear regression (PROC NLIN) in SAS v. 9.4 (Statistical

Analysis Systems, Cary, NC). GR Canada fleabane control at 2, 4 and 8 WAA and soybean yield

[expressed as percent of the yield of the industry standard (dicamba + saflufenacil) within each replicate] were regressed against pyraflufen-ethyl/2,4-D and pyraflufen-ethyl/2,4-D + metribuzin by specifying an exponential to maximum model (Equation 1) within PROC NLIN in SAS 9.4

(SAS Institute, Cary, NC). Weed density and biomass were expressed as a percent of the non- treated control within each replicate and regressed using an inverse exponential model (Equation

2). The parameters generated from each regression analysis were used to calculate the BED

(ED50, ED80, ED95) of pyraflufen-ethyl/2,4-D and pyraflufen-ethyl/2,4-D + metribuzin to achieve

50, 80, and 95% control, reduction in density, biomass and soybean yield. Where a dose was calculated to be greater than the highest dose evaluated in this study (>2108 g ai ha-1) it was

77 expressed as non-estimable (Non-est.) as it would be unacceptable to extrapolate outside of the range tested.

Equation 1: Exponential to maximum

Y = a - b (e - c Dose)

Where:

a = upper asymptote

b = magnitude

c = slope

Equation 2: Inverse exponential

Y = a + b (e – c Dose)

Where: a = lower asymptote b = reduction in y from intercept to asymptote

c = slope

Least-Square means comparisons

The data across all five site-years were pooled for analysis to compare pyraflufen- ethyl/2,4-D and pyraflufen-ethyl/2,4-D + metribuzin to the current industry standards in Ontario,

Canada. PROC GLIMMIX was used in SAS v. 9.4 (Statistical Analysis Systems, Cary, NC) to perform a mixed-model variance analysis. Variance was divided into the fixed effect (herbicide treatment) and the random effects (environment, replication within environment, and the treatment-by-environment interaction). Least-square means were separated using a Tukey-

Kramer multiple-range test (α = 0.05). Data normality was checked using the Shapiro-Wilk test

78 of normality along with visual assessments of studentized residuals and scatterplots. The residuals from visible GR Canada fleabane control data at 2, 4 and 8 WAA and yield followed a normal distribution. The biomass and density data were analyzed using a lognormal distribution and back transformed for presentation using the omega method within the GLIMMIX procedure

(M. Edwards, OAC Statistician, University of Guelph, personal communication).

To determine the interactions of pyraflufen-ethyl/2,4-D and metribuzin, expected values for control, density and biomass were calculated using the Colby’s Equations (Colby 1967).

PROC TTEST was utilized in SAS 9.4 (SAS Institute, Cary, NC) to compare the observed and expected values. If the values did not significantly differ, the interaction was considered additive.

The interaction was considered antagonistic or synergistic if there was a significant difference

(P<0.05 or P<0.01) between the expected and observed values.

Equation 3: Colby’s Equation – Visible Weed Control

E= X+Y– (푋푌/100)

Where:

E = expected control with pyraflufen-ethyl/2,4-D + metribuzin

X = observed control with pyraflufen-ethyl/2,4-D

Y = observed control with metribuzin

Equation 4: Colby’s Equation- Density and Biomass

E = (푋푌/푍)

Where:

E = expected density or biomass with pyraflufen-ethyl/2,4-D + metribuzin

X = observed density or biomass pyraflufen-ethyl/2,4-D

Y = observed density or biomass of metribuzin

Z = density or biomass of the non-treated control

3.4. Results and Discussion

3.4.1 Soybean Injury

No treatment caused greater than 15% visible injury. Pyraflufen-ethyl/2,4-D (2108 g ai ha-1) resulted in the highest soybean injury of 13 and 9% injury at 1 and 4 WAE, respectively.

The addition of metribuzin did not increase soybean injury. The registered rate of pyraflufen- ethyl/2,4-D (527 g ai ha-1) caused 5 and 1% soybean injury at 1 and 4 WAE, respectively.

Similarly, pyraflufen-ethyl/2,4-D + metribuzin at 527 + 400 g ai ha-1 caused 3 and 1% injury at 1 and 4 WAE, respectively. Soybean injury consisted of stunting, twisting of the stems and leaf strapping consistent with 2,4-D injury. Vink et al. (2012) reported that 2,4-D ester (500 g ae ha-1) caused up 7% soybean injury at 4 WAA. Similar soybean injury has been observed with 2,4-D ester (560 g ae ha-1) in Tennessee, USA (Thompson et al. 2007).

3.4.2 Biologically-Effective-Dose of Pyraflufen-ethyl/2,4-D and Pyraflufen-ethyl/2,4-D + Metribuzin for the control of GR Canada Fleabane

Pyraflufen-ethyl/2,4-D Applied Alone

The calculated doses of pyraflufen-ethyl/2,4-D to control GR Canada fleabane 50, 80, and 95% were 185, 568 and >2108 g ai ha-1, respectively at 2 WAA (Table 3.2). At 4 and 8

WAA there was an increase in the calculated doses required for 50, 80, and 95% control of 244, 80

709, and >2108 g ai ha-1 at 4 WAA and 390, 1148, and >2108 g ai ha-1 at 8 WAA, respectively.

The calculated doses for a 50, 80, and 95% reduction in density and biomass were 876, 1793, and > 2108 g ai ha-1 and 934, 1851, and >2108 g ai ha-1, respectively at 8 WAA. The calculated doses of <66, 516, and 988 g ai ha-1 were required for 50, 80 and 95% soybean yield relative to the industry standard.

Pyraflufen-ethyl/2,4-D is a relatively new preformulated herbicide mixture, however, similar rates of pyraflufen-ethyl and 2,4-D, when applied alone, have shown similar results for the control of GR Canada fleabane and other weed species. Based on the concentrations of pyraflufen-ethyl (6.1 g/L) and 2,4-D ester (473 g a.e/L) in BlackHawk®, the calculated dose for

80% GR Canada fleabane control of 1148 g ai ha-1 consists of approximately 15 g ai ha-1 of pyraflufen-ethyl and 1133 g ai ha-1 of 2,4-D ester. Pyraflufen-ethyl (8 g ai ha-1) controlled GR

Canada fleabane and other Conyza species in Europe (Tahmasebi et al. 2018). Pyraflufen-ethyl

(3 to 6 g ai ha-1) controlled cleavers, common chickweed and Persian speedwell (Veronica persica Poiret.) (Mabuchi et al. 2002; Mirua et al. 2003). Scroggs et al. (2006) reported that pyraflufen-ethyl + glyphosate (11 + 630 g ai ha-1) controlled pitted morninglory (Ipomoea lacunosa L.) 90 and 88% at 7 and 28 days after application (DAA); hemp sesbania [Sesbania herbacea (Mill.) McVaugh.] control was 57% at the same evaluation timings. Previous research on the efficacy of 2,4-D ester for the control of GR Canada fleabane is far more extensive.

Kruger et al. (2010a) calculated that the dose of 2,4-D ester for 90% control of GR Canada fleabane ranged from 316 to 684 g ae ha-1. In the same study, 560 g ae ha-1 of 2,4-D ester controlled 7 to 15 cm tall GR Canada fleabane 97% at 4 WAA; the authors concluded that “a

2,4-D ester use rate of 500 g ae ha-1 or more should be considered for control of Canada fleabane”. Byker et al. (2013d) reported that 2,4-D ester + glyphosate (500 + 900 g ai ha-1)

81 applied PP controlled GR Canada fleabane 78 to 92% 4 WAA dependent upon location. Soltani et al. (2020a) and Soltani et al. (2020b) reported that 2,4-D ester (500 g ai ha-1) provided lower levels of GR Canada fleabane control of 58 and 53 % at 8 WAA, while Soltani et al. (2016) found the same rate controlled GR Canada fleabane 68 and 61% control at 4 and 8 WAA respectively. The results of this study found that GR Canada fleabane control with pyraflufen- ethyl/2,4-D are consistent with previous studies where pyraflufen-ethyl and 2,4-D ester were applied individually.

The calculated doses of pyraflufen-ethyl/2,4-D for 50% GR Canada fleabane control and

50% decrease in density and biomass were 390, 876 and 934 g ai ha-1, respectively at 8 WAA

(Table 3.2). There was regrowth of GR Canada fleabane following the application of pyraflufen- ethyl/2,4-D which may have contributed the higher calculated dose for biomass reduction.

Regrowth of injured plants after 2,4-D application was noted by Zimmer et al. (2019); surviving plants in some scenarios were able to produce seed (Kruger et al. 2010b). The calculated doses of pyraflufen-ethyl/2,4-D that resulted in 80 and 95% of the soybean yield of the industry standard

(glyphosate/dicamba + saflufenacil) were 516 and 988 g ai ha-1, respectively.

Pyraflufen-ethyl/2,4-D + Metribuzin

The addition of metribuzin (400 g ai ha-1) to pyraflufen-ethyl/2,4-D improved GR Canada fleabane control compared to pyraflufen-ethyl/2,4-D applied alone. The calculated doses of pyraflufen-ethyl/2,4-D + metribuzin to control GR Canada fleabane 50, 80, and 95% were 17,

42, and 145 g ai ha-1, respectively at 2 WAA (Table 3.2). At 8 WAA, the calculated doses of pyraflufen-ethyl/2,4-D + metribuzin increased to 19, 46, and 201 g ai ha-1 for 50, 80, and 95%, respectively. The calculated doses of pyraflufen-ethyl/2,4-D for 50, 80, and 95% reduction in

82 density and biomass were 15, 34, and 64 g ai ha-1 and 26, 62, and 115 g ai ha-1, respectively at 8

WAA. The calculated doses of pyraflufen-ethyl/2,4-D for 80 and 95% of the soybean yield of the industry standard (glyphosate/dicamba + saflufenacil) were 1 and 2 g ai ha-1, respectively.

3.4.3 Pyraflufen-ethyl/2,4-D, Metribuzin, Pyraflufen-ethyl/2,4-D + Metribuzin compared to Industry Standards

Pyraflufen-ethyl/2,4-D (527 g ai ha-1 ), metribuzin (400 g ai ha-1 ), and pyraflufen- ethyl/2,4-D + metribuzin (527 + 400 g ai ha-1 ) were compared to the industry standards of saflufenacil + metribuzin (25 + 400 g ai ha-1 ) and glyphosate/dicamba + saflufenacil (1800 + 25 g ai ha-1 ). Pyraflufen-ethyl/2,4-D and metribuzin controlled GR Canada fleabane 59 to 69% and

75 to 78%, respectively at 2, 4, and 8 WAA respectively (Table 3.3). Metribuzin alone provided

75% control of GR Canada Fleabane at 8 WAA which was greater than pyraflufen-ethyl/2,4-D which provided 59% control. Soltani et al. (2020a) found that pyraflufen-ethyl/2,4-D (532 g ai ha-1) controlled GR Canada fleabane 63 and 60% at 4 and 8 WAA, respectively. Pyraflufen- ethyl/2,4-D reduced GR Canada fleabane density 39%; there was no effect on biomass.

Metribuzin, applied alone, reduced fleabane density by 90% but had no effect on biomass.

Pyraflufen-ethyl/2,4-D + metribuzin (527 + 400 g ai ha-1) and the industry standards saflufenacil

+ metribuzin and glyphosate/dicamba + saflufenacil controlled GR Canada fleabane similarly at

97 to 100% at 2, 4, and 8 WAA. Soltani et al. (2020a) reported that pyraflufen-ethyl/2,4-D + metribuzin (532 + 400 g ai ha-1) provided 95% control of GR Canada fleabane, similar to saflufenacil + metribuzin (25 + 400 g ai ha-1) and glyphosate/dicamba + saflufenacil (1800 + 25 g ai ha-1) with 99 and 98%, respectively at 8 WAA. Pyraflufen-ethyl/2,4-D + metribuzin, saflufenacil + metribuzin, and glyphosate/dicamba + saflufenacil reduced GR Canada fleabane density and biomass similarly at ≥98 and ≥97%, respectively at 8 WAA. Soybean yield was

83 reduced up to 60% in this study due to GR Canada fleabane interference (Table 3.3). Although there were differences in GR Canada fleabane control, density, and biomass among the herbicide treatments evaluated, there were no statistical differences in soybean yield. Similar soybean yield reductions due to GR Canada fleabane interference have been reported by Budd et al. (2016a) and Soltani et al. (2020a).

3.4.4 Interaction – Pyraflufen-ethyl/2,4-D, Metribuzin compared to Pyraflufen-ethyl/2,4-D + Metribuzin

Additive and synergistic interactions were identified with the co-application of pyraflufen-ethyl/2,4-D and metribuzin for GR Canada fleabane control and reductions in density and biomass (Table 3.4). There was a synergistic improvement in GR Canada fleabane control when pyraflufen-ethyl/2,4-D at 66, 264, 527, and 1054 g ai ha-1 was co-applied with metribuzin

(400 g ai ha-1 ) at 2 WAA, when pyraflufen-ethyl/2,4-D was applied at 264, 527, and 1054 g ai ha-1 with metribuzin (400 g ai ha-1 ) at 4 WAA, and when pyraflufen-ethyl/2,4-D was applied at

66, 264, 527, and 1054 g ai ha-1 with metribuzin (400 g ai ha-1 ) at 8 WAA; there were similar synergistic decreases in Canada fleabane density and biomass reduction with the co-application of pyraflufen-ethyl/2,4-D and metribuzin. All other interactions were additive. Han et al. (2020) reported that sequential applications of 2,4-D amine followed by metribuzin resulted in a synergistic interaction with increased metribuzin translocation to the new leaves in wild oats

(Avena sterilis L.). In corn, the co-application of 2,4-D amine with atrazine, a PS II-inhibitor, resulted in a synergistic improvement in the control of prostrate knotweed (Polygonum aviculare

L.) (Ludwig 1973).

3.5.1 Conclusion

This study concludes the calculated doses of pyraflufen-ethyl/2,4-D for 50, 80, and 95%

GR Canada fleabane control were 390, 1148, and >2108, respectively; the addition of metribuzin

(400 g ai ha-1) to pyraflufen-ethyl/2,4-D reduced the calculated doses for 50, 80, and 95% GR

Canada fleabane control to 19, 46, and 201 g ai ha-1, respectively at 8 WAA. Pyraflufen- ethyl/2,4-D (527 g ai ha-1) and metribuzin (400 g ai ha-1) provided 59 and 75% control of GR

Canada fleabane, respectively. The co-application of pyraflufen-ethyl/2,4-D + metribuzin (527 +

400 g ai ha-1) controlled GR Canada fleabane 97% similar to the industry standards of saflufenacil + metribuzin (25 + 400 g ai ha-1) and glyphosate/dicamba + saflufenacil (1800 + 25 g ai ha-1) at 8 WAA. Based on Colby’s Equation there was an additive or synergistic improvement in GR Canada fleabane control when pyraflufen-ethyl/2,4-D was co-applied with metribuzin. Overall, this study concludes that GR Canada fleabane control with pyraflufen- ethyl/2,4-D + metribuzin, applied PP, is comparable to the industry standards of saflufenacil + metribuzin and glyphosate/dicamba + saflufenacil.

Table 3.1 Year, location, application information, crop information, and soil characteristics for five field experiments conducted on the biologically-effective-dose of pyraflufen-ethyl/2,4-D, applied alone and tankmixed with metribuzin, for the control of glyphosate- resistant Canada fleabane [Conyza canadensis (L.) Cronq.] in soybean in 2019 and 2020 in Ontario, Canada. Canada Canada Soybean Soybean Soybean Year Site Application fleabane fleabane planting emergence harvest Texture SOMa Soil date sizeb densityc date date date pH (cm) (plants m-2) (%) 2019 Bothwell June 7 7.5 127 June 8 June 14 October 10 Sand 2.6 6.7

Moraviantown June 18 7.5 321 June 19 June 25 October 8 Sand 2.2 6.1

Thamesville June 11 7 88 June 12 June 20 October 7 Loamy 1.8 5.6 sand 2020 Duart June 1 8.5 352 June 5 June 10 October 6 Sandy 4.8 7.3 loam Ridgetown May 26 6.5 470 June 5 June 10 October 1 Sandy 1.9 7.1 loam aAbbreviation: SOM, soil organic matter of the upper 15 cm of the soil profile bSize measured as height or diameter of plant on day of application. Mean of two measurements per replication. cMean density based on two density counts in non-treated control within each replication.

Parameter estimates (SE ) Calculated pyraflufen-ethyl/2,4-D dose Variable a b c ED50 ED80 ED95

(g ai ha-1)

Control 2 WAAa 88.29 (2.24) 80.33 (3.04) 0.004 (0) 185 568 Non-est.c

Control 4 WAAa 89.88 (2.44) 82.52 (3.03) 0.003 (0) 244 709 Non-est.

Control 8 WAAa 88.43 (2.90) 83.80 (3.17) 0.002 (0) 390 1148 Non-est.

Densityb 0 (0) 120.1 (10.10) 0.001 (0) 876 1793 Non-est.

Dry Biomassb 0 (0) 127.3 (9.09) 0.001 (0) 934 1851 Non-est.

Yieldad 116.9 (16.23) 66.88 (15.26) 0.001 (0) Non-est. 516 988

Calculated pyraflufen-ethyl/2,4-D + metribuzin dose Variable a b c ED50 ED80 ED95

(g ai ha-1)

Control 2 WAAa 95.16 (0.73) 95.14 (1.74) 0.044 (0.005) 17 42 145

Control 4 WAAa 95.67 (0.75) 95.63 (1.78) 0.040 (0.004) 18 45 124

Control 8 WAAa 95.03 (0.88) 94.91 (2.09) 0.040 (0.005) 19 46 201 Densityb 0 (0) 99.99 (1.58) 0.047 (0.005) 15 34 64

Dry Biomassb 0 (0) 99.39 (3.18) 0.026 (0.003) 26 62 115

Yieldad 118.6 (4.66) 66.52 (12.24) 0.457 (0.042) Non-est. 1 2 aRegression parameters: Y = a - b (e –c Dose); a = upper asymptote, b = magnitude, and c = slope bRegression parameters: Y = a + b (e –c Dose); a = lower asymptote, b = reduction in y from intercept to asymptote, and c = slope cGreater than the highest dose evaluated in this study (>2108 g ai ha-1) or could not be calculated by the model. dExpressed as percent of yield in the industry standard [Glyphosate/dicamba + saflufenacil (1800 + 25 g ai ha-1)] among replications

Parameter

Treatment Rate Visual Control Density Dry Soybean Biomass Yield 2 WAA 4 WAA 8 WAA g ai ha-1 ---%--- plants m-2 g m-2 tonnes ha-1

Non-treated controlc - 0 c 0 c 0 d 315 c 217 b 1.0 b

Metribuzin 400 77 b 78 b 75 b 31 b 111 b 1.7 a

Pyraflufen-ethyl/2,4-D 527 69 b 67 b 59 c 191 c 128 b 1.7 a

Pyraflufen-ethyl/2,4-D + metribuzin 527 + 400 98 a 98 a 97 a 4 a 6 a 2.4 a

Saflufenacil + metribuzin b 25 + 400 99 a 100 a 99 a 1 a 1 a 2.3 a

Glyphosate/dicamba + saflufenacil b 1800 + 25 99 a 99 a 99 a 1 a 1 a 2.2 a aMeans followed by the same letter within a column are not statistically different at α=0.05 according to Tukey-Kramer’s multiple range test. bIncluded Merge (0.5% vol/vol). cNon-treated control plots received glyphosate (900 g ae ha-1) alone

Table 3.4 Glyphosate-resistant Canada fleabane [Conyza canadensis (L.) Cronq.] control at 2, 4, and 8 weeks after application (WAA), plant density and dry biomass at 8 WAA with metribuzin, pyraflufen-ethyl/2,4-D, and pyraflufen-ethyl/2,4-D + metribuzin applied preplant from five field experiments conducted in 2019 and 2020 in Ontario, Canada.c Treatment Rate Visual Control Density Biomass g ai ha-1 2 WAA 4 WAA 8 WAA Plants m-2 g m-2 --%-- Non-treated control d - 0 g 0 g 0 g 315 g 217 g Metribuzin 400 77 cd 78 cd 75 bc 31 de 111 def Pyraflufen-ethyl/2,4-D 66 32 f 28 f 20 e 365 g 297 g Pyraflufen-ethyl/2,4-D 132 44 ef 39 ef 30 de 269 g 206 g Pyraflufen-ethyl/2,4-D 264 53 e 49 e 38 d 252 g 205 g Pyraflufen-ethyl/2,4-D 527 69 d 67 d 59 c 191 fg 128 efg Pyraflufen-ethyl/2,4-D 1054 82 bcd 82 bcd 77 bc 46 ef 37 de Pyraflufen-ethyl/2,4-D 2108 95 abc 94 ab 89 ab 10 cd 10 abcd Pyraflufen-ethyl/2,4-D + 66 + 400 91 abc (85a)* 90 abc (84a) 88 ab (80a)* 9 bcd (24b)* 25 bcd (110b)* metribuzin Pyraflufen-ethyl/2,4-D + 132 + 400 86 abc (87) 86 abc (87) 84 ab (83) 36 de (30) 56 cde (98) metribuzin Pyraflufen-ethyl/2,4-D + 264 + 400 94 ab (90)* 94 ab (89)** 92 ab (85)** 10 bcd (26)* 17 abcd (90)** metribuzin Pyraflufen-ethyl/2,4-D + 527 + 400 98 a (93)** 98 a (93) ** 97 a (90)** 4.abc (16) ** 56 abc (45)** metribuzin Pyraflufen-ethyl/2,4-D + 1054 + 400 98 a (96)* 98 a (96)* 97 a (94)* 2 ab (4)* 1 ab (12)* metribuzin Pyraflufen-ethyl/2,4-D + 2108 +400 99 a (99) 100 a (99) 98 a (98) 0.4 a (1) 1 a (3) metribuzin * or ** denote a significant difference of P<0.05 or P<0.01, respectively, between observed and expected values based on a t-test, indicating synergistic interactions of pyraflufen-ethyl/2,4-D + metribuzin. a Expected values for herbicide combinations are shown in parentheses following observed values based on Colby’s Equation (Equation 3). b Expected values for herbicide combinations are shown in parentheses following observed values based on Colby’s Equation (Equation 4). c Means followed by the same letter within a column are not statistically different at α=0.05 according to Tukey-Kramer’s multiple range test. d Non-treated control plots received glyphosate (900 g ae ha-1 ) at the time of application.

This manuscript was submitted to Weed Technology and is currently under review

Chapter 4: Biologically-Effective-Dose of Tiafenacil Applied Preplant Alone or Tankmixed with Metribuzin for Glyphosate-Resistant Canada Fleabane [Conyza canadensis (L.) Cronq.] Control in Soybean

4.1 Abstract

Tiafenacil is a recently developed protoporphyrinogen IX oxidase (PPO)-inhibiting herbicide that is a member of the pyrimidinedione family and is proposed for use as a preplant

(PP) burndown in soybean. Glyphosate-resistant (GR) Canada fleabane is a winter or summer annual weed often found in no-till systems that can dramatically reduce soybean yield; control in soybean has been variable. Five field trials were conducted in 2019 and 2020 in commercial soybean fields with GR Canada fleabane to determine the biologically-effective-dose (BED) of tiafenacil and tiafenacil + metribuzin, and to compare their efficacy to the currently accepted industry standard herbicide treatments in identity-preserved (IP), GR, and dicamba-resistant

(DR) soybean systems. No soybean injury was observed in this study. The calculated doses of tiafenacil for 50, 80, and 95% control of GR Canada fleabane control were 21, 147, and >200 g ai ha-1, respectively, at 8 weeks after application (WAA). At 8 WAA, the calculated doses of tiafenacil for 50, 80, and 95% GR Canada fleabane control were 0.6, 3.3, and >200 g ai ha-1 respectively, when metribuzin (400 g ai ha-1) was included. Tiafenacil + metribuzin at 25 + 400 and 50 + 400 g ai ha-1 provided 88 and 93% control of GR Canada fleabane, respectively which was similar control to the industry standards of saflufenacil + metribuzin (25 + 400 g ai ha-1) and glyphosate/dicamba + saflufenacil (1800 + 25 g ai ha-1) that provided 98 to 100% control at 8

WAA. This study presents the potential utility of tiafenacil + metribuzin as a GR Canada fleabane management strategy in soybean.

4.2 Introduction

Canada fleabane is a winter or summer annual weed that is native to North America and is a problematic weed across Ontario (Budd et al. 2017; Weaver 2001). Canada fleabane plants are highly prolific producing up to 500,000 seeds (Davis et al. 2009a). Canada fleabane seed dispersion is assisted by its small seed size (1 to 2 mm long) and the attachment of a pappus that aids with wind dispersal (Weaver 2001). While most seeds land within 100 m of the mother plant, seeds have been found in the planetary boundary layer, with potential to travel hundreds of kilometers (Shields et al. 2007; Steckel et al. 2010). Canada fleabane seed emerges from a maximum burial depth of 0.5 cm, making it adapted to reduced-, strip- and no-till crop production systems (Nandula et al. 2006: Weaver 2001).

Glyphosate-resistant (GR) Canada fleabane was first reported in Delaware, USA in 2001 after the recurrent use of glyphosate in GR soybean (VanGessel 2001). In Ontario, the first incidence of GR Canada fleabane was reported in Essex County from seed collected in 2010

(Byker et al.2013a). The control of GR Canada fleabane is difficult as there no efficacious post- emergent weed control options in identity-preserved (IP, non-GMO) and GR soybean (Byker et al. 2013b; VanGessel 2001). The preplant (PP) application of saflufenacil + metribuzin (25 +

400 g ai ha-1) is currently the most efficacious and consistent option for GR Canada fleabane control in IP and GR soybean (Budd et al. 2016a). Glyphosate/dicamba + saflufenacil (1800 + 25 g ai ha-1) is considered the best PP herbicide option in dicamba-resistant (DR) soybean (Hedges et al. 2018a).

Weed control in soybean in crucial for producers to maximize yield and net returns.

Soltani et al. (2017b) found that in North America if no weed management tactics were implemented, a 52% reduction in soybean yield would occur on account of weed interference. In

92 soybean, Van Acker et al. (1993) described the critical weed free period (CWFP) to be from emergence (VE) to the fourth node (V4), with yield loss <2.5% when there was no weed interference throughout this period. In Ontario, GR Canada fleabane interference has been reported to reduce soybean yield 53 to 73% (Budd et al. 2016a; Hedges et al. 2018a; Soltani et al. 2020a). The utilization of burndown plus residual herbicides applied prior to soybean emergence provide the most consistent control of both fall and spring emerged GR Canada fleabane in IP and GR soybean (Davis and Johnson 2008). Therefore, PP burndown applications are key to controlling GR Canada fleabane through the CWFP to minimize soybean yield loss.

Metribuzin is a photosystem II-inhibiting herbicide commonly used as a tankmix partner prior to planting soybean (Budd et al. 2016b; Eubank et al. 2008; Kapusta 1979; Soltani et al.

2020a). The inclusion of metribuzin (400 and 420 g ai ha-1) when tankmixed with saflufenacil, glufosinate, 2,4-D ester, pyraflufen-ethyl/2,4-D, or s-metolachlor/dicamba improved the control of GR Canada fleabane (Budd et al. 2016a; Eubanks et al. 2008; Soltani et al. 2020a).

Tiafenacil is a new protoporphyrinogen IX oxidase (PPO)-inhibiting, non-selective, contact herbicide that is a member of the pyrimidinedione family developed by FarmHannong

Co., Ltd., Korea (Park et al. 2018, Anonymous 2020a). PPO-inhibitors stop the production of protoporphyrin IX (PPIX) from protoporphyrinogen IX (PPGIX), thereby preventing chlorophyll and heme biosynthesis, resulting in increased levels of singlet oxygen leading to lipid peroxidation, cell membrane destruction and ultimately plant death (Shaner 2014). Tiafenacil has recently been proposed for registration in the USA and Canada (Environmental Protection

Agency 2020; Health Canada 2018). Tiafenacil can be applied preplant or preemergence in corn

(Zea mays L.), cotton (Gossypium hirsutum L.), wheat (Triticum aestivum L.), and soybean for the control of emerged annual grass and broadleaf weeds (Anonymous 2020a). The proposed

93 rates of tiafenacil for all application timings and crops is between 25 and 75 g ai ha-1.

Preliminary research by Park et al. (2018) reported that tiafenacil has activity on weed species such as waterhemp (Amaranthus tuberculatus L.) and arabidopsis (Arabidopsis thaliana), as well as crop species including soybean and rapeseed (Brassica napus L.). A half-maximal inhibitory concentration (IC50) was reported to be similar to other pyrimidinediones such as saflufenacil and butafenacil ranging from of 22 to 28 nM. Park et al. (2018) concluded that tiafenacil doses at

< 85 g ai ha-1 are sufficient for the control of selected monocot and dicot weeds. Haring and

Hanson (2020) reported that tiafenacil at 25, 50 and 75 g ai ha-1 controlled GR Canada fleabane

71 to 73%, at 7 days after treatment (DAT) with control being reduced to 20 to 29% at 31 DAT.

Minimal amounts of published information exists on the utilization of tiafenacil, applied alone or tankmixed with metribuzin, for the control of GR Canada fleabane. The objective of this study was to determine the biologically-effective-dose (BED) of tiafenacil and tiafenacil + metribuzin for control of GR Canada fleabane applied PP in soybean in Ontario.

4.3 Materials and Methods

Experimental Methods

Five field trials were conducted in 2019 and 2020 in commercial soybean fields located in southern Ontario with confirmed GR Canada fleabane. Trial application information, and site- specific soil characteristics are listed in Table 4.1. Two trials were conducted at one site in 2020 and were separated in space and time with herbicide applications occurring on different days.

Each trial was set up as a randomized complete block design with four replications per trial. The trials consisted of 18 treatment including a non-treated control, metribuzin (400 g ai ha-1), tiafenacil (3.125, 6.25, 12.5, 25, 50, 100, and 200 g ai ha-1), tiafenacil + metribuzin (3.125 + 400, 94

6.25 + 400, 12.5 + 400, 25 + 400, 50 + 400, 100 + 400, and 200 + 400 g ai ha-1), saflufenacil + metribuzin (25 + 400 g ai ha-1), and glyphosate/dicamba + saflufenacil (1800 + 25 g ai ha-1) applied PP. Saflufenacil + metribuzin (25 + 400 g ai ha-1) and glyphosate/dicamba + saflufenacil

(1800 + 25 g ai ha-1) were included as the current industry standards for the control of GR

Canada fleabane in IP and GR soybean and DR soybean, respectively (Budd et al. 2016a;

Hedges et al. 2018a). Methylated seed oil (MSO) was included in all tiafenacil and tiafenacil + metribuzin treatments in 2020 at a rate of 1% vol/vol. Glyphosate (Roundup WeatherMAX®) was included in all treatments at a rate of 900 g ae ha-1 with the exception of glyphosate/dicamba

+ saflufenacil (1800 + 25 g ai ha-1) as it is preformulated with 1200 g ae ha-1 of glyphosate. All treatments were applied PP when the average Canada fleabane reached 10 cm in height or diameter. A CO2 pressurized backpack sprayer was utilized for all applications and was calibrated to deliver 200 L ha-1 of water at a pressure of 260 kPa. The spray boom was equipped with four ULD 11002 nozzles (Pentair, New Brighton, MN, USA), spaced 50 cm apart, producing a spray width of 2.0 m.

Dicamba-resistant (DR) soybean (DKB 12-16) was no-tilled at a seeding rate of approximately 416 000 seeds ha-1 with planting dates listed in Table 4.1. Plots measured 2.25 m wide (3 soybean rows spaced 75 cm apart) by 8 m long. Glyphosate (450 g ae ha-1) was applied

POST to all trials to eliminate all other weed species, including glyphosate-susceptible Canada fleabane.

Weed control assessments were completed at 2, 4, and 8 weeks after application (WAA).

Treatments were assessed visually on a scale of 0 to 100%, with 0 being no control and 100 being complete control. Crop injury assessments were completed visually at 1, 2, and 4 weeks

95 after soybean emergence (WAE). Crop injury was assessed on a scale of 0 to 100%, with 0 being no soybean injury and 100 being complete soybean death. At 8 WAA, Canada fleabane plants were counted in two 0.25 m2 quadrats representative of each plot. The Canada fleabane plants were cut at the soil surface, bagged, and dried in a kiln to a constant moisture, and weighed.

Soybean yield was collected from each plot by combining the centre two rows at harvest maturity. Soybean moisture was adjusted to 13.5% prior to statistical analysis.

Statistical Analysis

Although MSO was included in 2020 and not in 2019, the data across all five site-years were pooled for analysis as no statistically significant treatment by year interactions were found.

For GR Canada fleabane control at 2, 4, and 8 WAA the p-values for treatment by year interaction were 0.128. 0.178, and 0.263, respectively; for GR Canada fleabane density and biomass, and soybean yield the p-values were 0.610, 0.189, and 0.456, respectively.

Non-linear Regression

Percent GR Canada fleabane control at 2, 4, and 8 WAA was regressed against tiafenacil or tiafenacil + metribuzin using the NLIN procedure in SAS v. 9.4 (SAS Institute, Cary NC) using the rectangular hyperbola equation (Equation 1) (Cousens 1985). Yield was expressed as a percent of the industry standard [saflufenacil + glyphosate/dicamba (25 + 1800 g ai ha-1)] within each replicate and were regressed against tiafenacil or tiafenacil + metribuzin. (Equation 1). The effective dose (ED) for 50, 80, and 95% Canada fleabane control for each assessment timing was calculated using the parameter estimates generated from the regression analysis. An inverse exponential equation (Equation 2) was used to regress density and dry biomass, relative to the non-treated control within each replicate, against tiafenacil or tiafenacil + metribuzin. Similar to

96 control, the ED of tiafenacil and tiafenacil + metribuzin was calculated for 50, 80, and 95% reduction in density and dry biomass. Where the ED could not be calculated by the regression model or was outside the set of doses utilized, “non-est” (non-estimable) was presented in the tables.

Equation 1: Rectangular Hyperbola

Y = (i*dose) / [1+(i*dose/a)]

Where: a is the upper asymptote; i is the initial slope

Equation 2: Inverse exponential

Y = a + b (e – c dose)

Where: a is the lower asymptote; b is the reduction in y from intercept to asymptote; c is the slope

Least-Square means comparisons

PROC GLIMMIX was used in SAS v. 9.4 (SAS Institute, Cary, NC) to perform a generalized linear mixed-model analysis of variance comparing the non-treated control, tiafenacil, metribuzin, tiafenacil + metribuzin, saflufenacil + metribuzin, and glyphosate/dicamba

+ saflufenacil. Similarly, all rates of tiafenacil and tiafenacil + metribuzin were compared.

Variance was partitioned into the random effects of environment, replication within environment, and the treatment by environment interaction. Herbicide treatment was considered a fixed effect. Control and soybean yield data were fit to normal distribution with the identity link function. GR Canada fleabane density and biomass data were fit to a lognormal distribution with an identity link; the data was back transformed using the omega method for presentation

97 purposes (M. Edwards, Ontario Agricultural College Statistician, University of Guelph, personal communication). Least-square means of each parameter were separated using the Tukey-Kramer test (p=0.05).

Interaction – Tiafenacil and Metribuzin

Colby’s Equations were utilized to determine whether the interaction between tiafenacil and metribuzin was antagonistic, additive, or synergistic (Colby 1967). Expected values were calculated for each dose of herbicide combination by replication for Canada fleabane control at

2, 4, and 8 WAA (Equation 3) and for density and biomass (Equation 4). The expected and observed values were compared using PROC TTEST in SAS v. 9.4 (SAS Institute, Cary, NC) at

P<0.05 and P<0.01. The co-application of tiafenacil with metribuzin at specific rates was considered to be: i) additive if the expected values were non-significant and numerically similar to the observed, ii) synergistic if the observed was significantly greater than the expected, and iii) antagonistic if the observed was significantly lower than the expected.

Equation 3: Colby’s Equation – GR Canada Fleabane Control

E = X+Y – (푋푌/100)

Where: E is the expected control with tiafenacil + metribuzin; X is the observed control with tiafenacil; and Y is the observed control with metribuzin.

Equation 4: Colby’s Equation- GR Canada Fleabane Density and Dry Biomass

E = (푋푌/푍)

Where: E is the expected density or biomass with tiafenacil + metribuzin; X is the observed density or biomass with tiafenacil; Y is the observed density or biomass with metribuzin; and Z is the density or biomass of the non-treated control.

4.4. Results and Discussion

4.4.1 Soybean Injury

No soybean injury was observed at 1, 2, and 4 WAE (data not presented). Budd et al.

(2016b) found that saflufenacil, another PPO-inhibiting herbicide and member of the pyrimidinedione family, caused less than 10% soybean injury when applied PP at 200 g ai ha-1.

Miller et al. (2012) tested 12 GR soybean cultivars and reported the majority of soybean cultivars were tolerant to saflufenacil PRE, although sensitivity varied with 52 and 59 g ai ha-1 of saflufenacil causing 10% soybean injury in two cultivars. Injury was also influenced by environmental conditions shortly after application.

4.4.2 Biologically-Effective-Dose of Tiafenacil and Tiafenacil + Metribuzin for the control of GR Canada fleabane

Tiafenacil Applied Alone

Based on the regression analysis, the calculated dose of tiafenacil for 50% GR Canada fleabane control was 8 g ai ha-1, with no dose ≤200 g ai ha-1providing 80% or 95% control at 2

WAA (Table 4.2). At 4 and 8 WAA, 50% GR Canada fleabane control could be achieved with

11 and 21 g ai ha-1 of tiafenacil, respectively. The calculated doses of tiafenacil for 80% GR

Canada fleabane were 103 and 147 g ai ha-1 at 4 and 8 WAA, respectively. At 4 and 8 WAA, the

99 tiafenacil dose for 95% GR Canada fleabane control were >200 g ai ha-1 and could not be calculated based on the doses evaluated in this study.

Density and biomass required higher doses of tiafenacil compared to GR Canada fleabane control ratings at 8 WAA. The calculated tiafenacil doses for 50% control and reduction in density and biomass were 21, 47, and 90 g ai ha-1, respectively. For 80% reduction in GR Canada fleabane density and biomass the calculated doses of tiafenacil were 169 and >200 g ai ha-1, respectively; similarly, the calculated dose of tiafenacil for a 95% reduction in density and biomass was >200 g ai ha-1. The calculated doses of tiafenacil that would result in 50, 80 and

95% of the soybean yield of the industry standard [dicamba + saflufenacil (1800 + 25 g ai ha-1)] were 11, 139, and >200 g ai ha-1, respectively.

With limited research on tiafenacil, comparisons to saflufenacil are appropriate since they are both pyrimidinedione herbicides with similar use patterns (Park et al. 2018). Research on the

BED of saflufenacil found that lower doses are required for GR Canada fleabane control compared to tiafenacil. Budd et al. (2016b) found that saflufenacil at 25 and 36 g ai ha-1 controlled GR Canada fleabane 90 and 95%, respectively at 8 WAA, while glyphosate (900 g ai ha-1) + saflufenacil at doses of 25, 34, and 47 g ai ha-1 controlled GR Canada fleabane 90, 95 and

98%, respectively. Similar to tiafenacil, increased saflufenacil doses were required for biomass reduction when compared to density reduction. The calculated saflufenacil dose for a 90% GR

Canada fleabane biomass and density reduction was 26 and 16 g ai ha-1, respectively. Other research has found the addition of adjuvants decrease the required dose of saflufenacil for

Canada fleabane control. Knezevic et al. (2009) found that the required doses of saflufenacil compared to saflufenacil + MSO for 90% Canada fleabane control was 217 and 78 g ha-1, respectively at 4 WAA. Mellendorf et al. (2013) found that as Canada fleabane height increased

100 at the time of application there was decreased control with saflufenacil (50 g ai ha-1). Mellendorf et al. (2015) also found that spray-solution pH, adjuvant, inclusion of glyphosate, and light intensity all effected the efficacy of saflufenacil on Canada fleabane. Further research should be conducted to identify if GR Canada fleabane density, height, and application variables impact tiafenacil efficacy.

Tiafenacil + Metribuzin

The addition of metribuzin (400 g ai ha-1) to tiafenacil decreased the calculated doses of tiafenacil for all GR Canada fleabane parameters assessed (Table 4.3). The calculated doses of tiafenacil when co-applied with metribuzin for 50% GR Canada fleabane control were 0.6, 0.5 and 0.6 g ai ha-1, while the calculated doses for 80% GR Canada fleabane control were was 3.3,

2.6, and 3.3 g ai ha-1 at 2, 4, and 8 WAA, respectively. The calculated doses for 95% GR Canada fleabane control were >200 g ai ha-1 at all assesement timings.

The calculated doses of tiafenacil when co-applied with metribuzin for a 50 and 80% decrease in GR Canada fleabane density were 0.8 and 2.0 g ai ha-1, respectively similar to the doses required for 50 and 80% control. The calculated dose for a 95% decrease in GR Canada fleabane density was >200 g ai ha-1. The calculated dose for a 50% decrease in GR Canada fleabane biomass was 2.6 g ai ha-1; > 200 g ai ha-1 of was required for an 80 and 95% reduction in biomass. For yield, the calculated doses of 0.6 and 3.1 g ai ha-1 of tiafenacil + metribuzin resulted in 50 and 80% of the yield of the industry standard [dicamba + saflufenacil (1800 + 25 g ai ha-1]; > 200 g ai ha-1 was required for 95% soybean yield relative to the industry standard.

101

4.4.3 Tiafenacil, Metribuzin, Tiafenacil + Metribuzin compared to Industry Standards

Least square means were used to compared metribuzin (400 g ai ha-1), tiafenacil (25 and

50 g ai ha-1), tiafenacil + metribuzin (25 + 400 and 50 + 400 g ai ha-1), saflufenacil + metribuzin

(25 + 400 g ai ha-1), glyphosate/dicamba + saflufenacil (1800 + 25 g ai ha-1) and the non-treated control.

Tiafenacil (25 and 50 g ai ha-1) provided 65 to 66% control of GR Canada fleabane at 2 and 4 WAA, respectively. Metribuzin (400 g ai ha-1) controlled GR Canada fleabane 71 to 73% which was similar to tiafenacil. Tiafenacil + metribuzin (25 + 400 g ai ha-1 and 50 + 400 g ai ha-

1) controlled GR Canada fleabane 87 to 93% which was greater than tiafenacil applied alone. The co-application of tiafenacil + metribuzin (25 + 400 g ai ha-1) controlled GR Canada fleabane 87 and 90% control at 2 and 4 WAA, respectively, which was similar to tiafenacil + metribuzin (50

+ 400 g ai ha-1), saflufenacil + metribuzin, and glyphosate/dicamba + saflufenacil.

At 8 WAA, tiafenacil at 25 and 50 g ai ha-1 provided 55 and 61% control, respectively of

GR Canada fleabane which was similar to metribuzin. Tiafenacil + metribuzin at 25 + 400 and

50 + 400 g ai ha-1 provided 88 and 93%, respectively which was greater than tiafenacil but similar to metribuzin. Tiafenacil + metribuzin (25 + 400 and 50 + 400 g ai ha-1), saflufenacil + metribuzin, and glyphosate/dicamba + saflufenacil provided at 88 to 100% control of GR Canada fleabane.

Tiafenacil (25 and 50 g ai ha-1) did not reduce GR Canada fleabane density (Table 4.4).

Metribuzin, tiafenacil + metribuzin, saflufenacil + metribuzin, and glyphosate/dicamba + metribuzin reduced GR Canada fleabane density 81, 93 to 94, 98, and 100%, respectively at 8

WAA. Tiafenacil and metribuzin applied alone did not reduce GR Canada fleabane biomass.

102

Tiafenacil + metribuzin, saflufenacil + metribuzin, and glyphosate/dicamba + metribuzin reduced GR Canada fleabane biomass 63 to 73, 95, and 100%, respectively at 8 WAA.

GR Canada fleabane interference reduced soybean yield up to 67% in this study. No difference in soybean yield was found among the herbicide treatments in this study.

4.4.4 Interaction – Tiafenacil, Metribuzin compared to Tiafenacil + Metribuzin

All interactions between tiafenacil and metribuzin were considered additive as there was no significant difference (P<0.05 and P<0.01) between the observed and expected values (Table

4.5).

4.5.1 Conclusion

No crop injury was observed in this study. This study concludes the calculated doses of tiafenacil for 50 and 80% control of GR Canada fleabane control were 21 and 147 g ai ha-1 respectively, and the calculated doses of tiafenacil when co-applied with metribuzin (400 g ai ha-

1) were 0.6 and 3.3 g ai ha-1 respectively, at 8 WAA. Tiafenacil at 25 and 50 g ai ha-1 and metribuzin (400 g ai ha-1) controlled GR Canada fleabane 55, 61, and 74%, respectively at 8

WAA. Tiafenacil + metribuzin at 25 + 400 and 50 + 400 g ai ha-1 controlled GR Canada fleabane

88 and 93%, respectively, which was similar to the industry standards of saflufenacil + metribuzin (25+ 400 g ai ha-1) providing 98% control and glyphosate/dicamba + saflufenacil

(1800 + 25 g ai ha-1) providing 100% control at 8 WAA. The co-application of tiafenacil with metribuzin resulted in an additive increase in GR Canada fleabane control, and an additive decrease in density and biomass. This study presents the efficacy of tiafenacil + metribuzin for

GR Canada fleabane control relative to the current industry standards.

103

Table 4.1 Year, location, application information, soil characteristics, and crop information for five field trials conducted on the biologically-effective-dose of tiafenacil, applied alone and tankmixed with metribuzin, for the control of glyphosate-resistant Canada fleabane [Conyza canadensis (L.) Cronq.] in soybean in Ontario, Canada, in 2019 and 2020. Canada Canada Soil Soybean Soybean Soybean Year Location Application fleabane fleabane Texture organic pH planting emergence harvest date sizeb densityc mattera date date date (cm) (plants m-2) (%) 2019 Bothwell June 7 7.5 276 Sand 2.6 6.7 June 8 June 14 October 10 Thamesville June 11 9.0 108 Loamy 1.8 5.6 June 12 June 20 October 7 sand 2020 Duart June 1 7.5 163 Sandy 4.8 7.3 June 5 June 10 October 6 loam Duart June 4 10 188 Sandy 4.8 7.3 June 5 June 10 October 6 loam Ridgetown May 26 7.0 558 Sandy 1.9 7.1 June 5 June 10 October 1 loam aBased on soils test of the upper 15 cm of the soil profile. bSize measured on application date with a mean of eight measurements per trial. cMean density based on eight stand counts among the non-treated controls of each trial.

104

g ai ha-1

Variable a i ED50 ED80 ED95

Control 2 WAAa 81.25 (2.43) 15.93 (1.98) 8 Non-est c Non-est

Control 4 WAAa 86.24 (2.85) 10.75 (1.25) 11 103 Non-est

Control 8 WAAa 89.13 (4.00) 5.30 (0.65) 21 147 Non-est

Yieldad 84.55 (6.23) 10.70 (2.84) 11 139 Non-est

a b c

Densityb 18.07 (12.09) 94.28 (12.93) 0.023 (0.00) 47 169 Non-est

Biomassb 23.98 (18.16) 92.06 (17.24) 0.014 (0.00) 90 Non-est Non-est aRegression parameters: Y = (i*dose)/[1+(i*dose/a)] ; Where a is the upper asymptote, i is the initial slope. bRegression parameters: Y = a + b (e –c dose); Where a is the lower asymptote, b is the reduction in y from intercept to asymptote, and c is the slope. cNon-est., effective dose could not be estimated by the model or was beyond the set of doses used in this study. dExpressed as percent of yield in the industry standard [glyphosate/dicamba + saflufenacil (1800 + 25 g ai ha-1)] among replications.

105

g ai ha-1

Variable a i ED50 ED80 ED95

Control 2 WAAa 91.30 (0.97) 196.2 (38.17) 0.6 3.3 Non-estc

Control 4 WAAa 93.08 (0.92) 218.2 (42.62) 0.5 2.6 Non-est

Control 8 WAAa 92.63 (1.12) 178.1 (35.94) 0.6 3.3 Non-est

Yieldad 92.53 (4.58) 187.6 (159.7) 0.6 3.1 Non-est

a b c

Densityb 6.27 (0.83) 93.72 (2.16) 0.94 (0.14) 0.8 2.0 Non-est

Biomassb 25.33 (3.06) 74.20 (7.49) 0.42 (0.12) 2.6 Non-est Non-est aRegression parameters: Y = (i*dose)/[1+(i*dose/a)] ; Where a is the upper asymptote, i is the initial slope. bRegression parameters: Y = a + b (e –c dose); Where a is the lower asymptote, b is the reduction in y from intercept to asymptote, and c is the slope. cNon-est., effective dose could not be estimated by the model or was beyond the set of doses used in this study. dExpressed as percent of yield in the industry standard [glyphosate/dicamba + saflufenacil (1800 + 25 g ai ha-1)] among replications.

106

Parameter

Treatment Rate Visual Control Density Dry biomass Grain yieldd 2 WAA 4 WAA 8 WAA g ai ha-1 ---%--- plants m-2 g m-2 tonnes ha-1

Non-treated control - 0 d 0 d 0 d 289 e 240 e 1.0 b

Metribuzin 400 71 bc 73 bc 74 bc 54 cd 139 cde 2.3 a

Tiafenacilc 25 65 c 66 c 55 c 175 de 140 de 2.1 a

Tiafenacilc 50 66 c 66 c 61 c 177 de 141 de 2.2 a

Tiafenacil + metribuzinc 25 + 400 87 ab 90 ab 88 ab 17 bc 88 bcd 2.8 a

Tiafenacil + metribuzinc 50 + 400 91 a 93 a 93 ab 21 bc 66 bc 2.9 a

Saflufenacil + metribuzin b 25 + 400 98 a 98 a 98 a 5 ab 13 ab 3.0 a

Glyphosate/dicamba + saflufenacil b 1800 + 25 100 a 99 a 100 a 1 a 1 a 3.0 a aMeans followed by the same letter within a column are not significantly different from one another according to Tukey-Kramer’s multiple range test at P <0.05. bIncluded Merge (0.5% vol/vol). cIncluded MSO (1% vol/vol) in 2020. dSoybean moisture adjusted to 13.5%.

107

Table 4.5 Glyphosate-resistant Canada fleabane [Conyza canadensis (L.) Cronq.] control at 2, 4, and 8 weeks after application (WAA), plant density and dry biomass at 8 WAA with metribuzin, tiafenacil, and tiafenacil + metribuzin applied preplant from five field trials conducted in Ontario, Canada in 2019 and 2020.c Treatment Rate Visual control Density c Biomass c 2 WAA 4 WAA 8 WAA Plants m-2 g m-2 g ai ha-1 --%-- Non-treated Control - 0 f 0 e 0 e 289 e 240 cd

Metribuzin 400 71 bc 73 bc 74 bc 54 bcd 139 bcd

Tiafenacild 3.125 39 e 32 d 25 d 286 e 278 d

Tiafenacild 6.25 46 e 40 d 28 d 361 e 236 cd

Tiafenacild 12.5 51 ed 46 d 32 d 247 e 239 cd

Tiafenacild 25 65 cd 66 c 55 c 175 de 140 bcd

Tiafenacild 50 66 cd 66 c 61 c 177 de 141 bcd

Tiafenacild 100 79 abc 81 abc 74 bc 118 cd 123 bcd

Tiafenacild 200 88 ab 90 ab 88 ab 33 abc 69 abc

Tiafenacil + metribuzind 3.125 + 400 81 abc (81a) 84 abc (80a) 82 ab (80a) 47 abc (91b) 147 bcd (191b)

Tiafenacil + metribuzind 6.25 + 400 83 ab (85) 85 abc (85) 83 ab (81) 46 abc (176) 96 abc (173)

Tiafenacil + metribuzind 12.5 + 400 88 ab (87) 88 ab (87) 87 ab (83) 42 abc (52) 146 bcd (163)

Tiafenacil + metribuzind 25 + 400 87 ab (88) 90 ab (89) 88 ab (87) 17 ab (41) 88 abc (132)

Tiafenacil + metribuzind 50 + 400 91 a (89) 93 a (89) 93 ab (89) 21 ab (32) 66 ab (127)

Tiafenacil + metribuzind 100 + 400 90 a (93) 93 a (94) 93 ab (93) 19 ab (28) 20 a (117)

Tiafenacil + metribuzind 200 + 400 95 a (97) 96 a (97) 96 a (97) 9 a (22) 24 a (57)

* denotes a significant difference (P<0.05) and ** denotes a significant difference (P<0.01) between observed and expected values based on a t-test, indicating synergistic interactions of tiafenacil + metribuzin. a Expected values for herbicide combinations are shown in parentheses following observed values based on Colby’s Equation (Equation 3). b Expected values for herbicide combinations are shown in parentheses following observed values based on Colby’s Equation (Equation 4). c Means followed by the same letter within a column are not significantly different from one another according to Tukey-Kramer’s multiple range test at P <0.05. dIncluded MSO (1% vol/vol) in 2020.

108

This manuscript was submitted to Weed Science and is currently under review

109

Chapter 5: Biologically-Effective-Dose of Metribuzin, Applied Preemergence and Postemergence, for the Control of Waterhemp (Amaranthus tuberculatus) with Different Mechanisms of Resistance to Photosystem II-inhibiting herbicides

5.1 Abstract

Waterhemp is a highly competitive, difficult to control, broadleaf weed found in cropping systems across North America. Target-site resistance (TSR) and non-target-site resistance

(NTSR) mechanisms confer photosystem II (PS II)-inhibitor resistance in Ontario waterhemp populations. Metribuzin-resistant (MR) waterhemp is due to TSR, a serine to glycine substitution at position 264. Conversely, in other populations of PS II-resistant waterhemp, plants are resistant to atrazine but metribuzin-sensitive (MS). PS II resistance in MS waterhemp is due to enhanced metabolism, a form of NTSR. The objective of this study was to determine the biologically-effective-dose of metribuzin applied preemergence (PRE) and postemergence

(POST) for the control of MS and MR PS II-resistant waterhemp. Ten field experiments were conducted in 2019 and 2020 to determine the effective doses of metribuzin for 50, 80, and 95% control of MS and MR waterhemp. Metribuzin applied PRE at the calculated doses of 133, 350, and 1070 g ai ha-1 controlled MS waterhemp 50, 80, and 95%, respectively, whereas the calculated doses of 7868 and 17533 g ai ha-1 controlled MR waterhemp 50 and 80%, respectively at 12 WAA. Metribuzin applied POST at the calculated doses of 245 and 1480 g ai ha-1 controlled MS waterhemp 50 and 80%, respectively; the calculated dose for 50% MR waterhemp control was greater than the highest dose (17920 g ai ha-1) included in this study. Metribuzin at

560 and 1120 g ai ha-1 and pyroxasulfone/flumioxazin applied PRE controlled MS waterhemp

88, 95, and 98%, respectively at 12 WAA. The aforementioned treatments controlled MR waterhemp 0, 4, and 93%, respectively at 12 WAA. Metribuzin at 560 and 1120 g ai ha-1 and

110 fomesafen applied POST controlled MS waterhemp 65, 70, and 78%, and MR waterhemp 0, 1, and 49%, respectively at 12 WAA. Results from greenhouse studies supported field observations. This study illustrates the effect of PS II resistance mechanisms on PS II-resistant waterhemp control. NTSR PS II-resistant waterhemp (enhanced metabolism) is controlled with metribuzin applied PRE and POST; in contrast, TSR PS II-resistant waterhemp (serine264glycine altered target site) is not controlled with metribuzin. This study illustrates the value of molecular understanding of resistance mechanisms to weed management practitioners.

5.2 Introduction

Waterhemp (Amaranthus tuberculatus) is a highly prolific, dioecious, broadleaf weed that is of great economic importance in many cropping systems across North America.

Waterhemp is a member of the Amaranthaceae family and is among the most competitive weed species (Costea et al. 2005). Waterhemp interference can reduce soybean and corn yield up to 73 and 74%, respectively (Steckel and Sprague 2004b; Vyn et al. 2007). Furthermore, waterhemp has high fecundity; Steckel et al. (2003) reported waterhemp seed production in excess of 1 million plant-1 in Illinois, while Hartzler et al. (2004) reported 4.8 million seeds from a single plant in Iowa. As an obligate outcrossing species, waterhemp has high genetic diversity resulting in increased potential for selection of herbicide-resistant (HR) biotypes (Liu et al. 2012; Steckel

2007).

Globally, waterhemp has been confirmed to be resistant to herbicides targeting seven distinct sites of action including resistance to the acetolactate synthase (ALS)-inhibitors [Weed

Science Society of America Site of Action (WSSA SAO) Group 2], synthetic auxins (WSSA

SOA Group 4), photosystem II (PS II)-inhibitors (WSSA SOA Group 5), 5- enolpyruvylshikimate 3-phosphate synthase (EPSPS)-inhibitors (WSSA SOA Group 9),

111 protoporphyrinogen oxidase (PPO)-inhibitors (WSSA SOA Group 14), very long chain fatty acid

(VLCFA) elongases-inhibitors (WSSA SOA Group 15), and 4-hydroxyphenylpyruvate dioxygenase (HPPD)-inhibitors (WSSA SOA Group 27) (Heap 2021). Even more troublesome are reports of six-way multiple-herbicide-resistant (MHR) waterhemp in Missouri and Illinois

(Shergill et al. 2018; Strom et al. 2019). In Ontario, HR waterhemp was first reported in 2002 in

Lambton County with a population resistant to both ALS and PS II-inhibitors (Costea et al. 2005;

Vyn et al. 2007). In 2014, Schryver et al. (2017a) reported a population to be resistant to Group

2, 5, and 9 herbicides. Most recently in Ontario, waterhemp populations have been confirmed resistant to Group 2, 5, 9, and 14 herbicides (Benoit et al. 2019).

PS II-inhibitor resistance is not a new phenomenon. Ryan (1970) reported the first PS II- inhibitor resistant weed when both atrazine and simazine failed to control a population of common groundsel (Senecio vulgaris L.). PS II-resistant waterhemp was first reported in 1990 in

Nebraska (Anderson et al. 1996). The mechanisms of herbicide resistance are divided into target- site resistance (TSR) and non-target-site resistance (NTSR). Initial reports found PS II-inhibitor resistance in waterhemp was conferred by TSR, specifically, by a mutation in the psbA gene, which encodes the D1 protein (Foes et al. 1998). The most common substitution that confers PS

II-inhibitor resistance in waterhemp is the serine264glycine amino acid substitution (Heap 2021).

Patzoldt et al. (2003) identified NTSR in waterhemp in Illinois that conferred resistance to atrazine (symmetrical triazine) but not to metribuzin (asymmetrical triazinone). Ma et al. (2013) further determined this NTSR was due to increased glutathione-S-transferase (GST) activity, which conferred resistance to atrazine in a waterhemp population in Illinois. Evans et al. (2017) found that a single GST enzyme, AtuGSTF2, correlated strongly with PS II resistance in waterhemp. In other weeds species, NTSR have been found to confer PS II-inhibitor resistance

112 by CYP450 monooxygenase, which was recently reported to confer metribuzin resistance in rigid ryegrass (Lolium rigidium Gaud.) in Australia (Ma et al. 2020). Waterhemp with PS II resistance due to NTSR is resistant to atrazine but sensitive to metribuzin (Patzodlt et al. 2003;

Vennapusa et al. 2018). The chlorine at the one position of the atrazine molecule provides a binding site for the GST enzyme, resulting in conjugation of atrazine to glutathione; the GST enzyme does not bind to the oxygen at the one position of the metribuzin molecule (O’Brien et al. 2018; Tranel, unpublished data). In contrast, waterhemp resistant to atrazine due to the serine264glycine amino acid substitution was found to also be resistant to metribuzin (Heap 2021;

Tranel, unpublished data).

Metribuzin is a PS II-inhibitor asymmetrical triazinone, commonly used preemergence

(PRE) for weed control prior to soybean emergence (Anonymous 2020b). Metribuzin is also utilized postemergence (POST) in processing tomatoes. In Ontario, waterhemp control with metribuzin has been variable. Vyn et al. (2007) reported metribuzin (1120 g ai ha-1) applied PRE controlled waterhemp 24 and 99% at Petrolia and Cottam, Ontario, respectively at 10 weeks after emergence (WAE). Schryver et al. (2017b) reported metribuzin (1120 g ai ha-1) controlled waterhemp 88% at 8 weeks after application (WAA) at sites located on Walpole Island and near

Cottam, Ontario.

There is little published research on the biological activity of metribuzin on PS II- inhibitor resistant waterhemp conferred by NSTR and TSR mechanisms. Within Ontario, waterhemp populations possess either TSR or NTSR PS II-inhibitor mechanisms of resistance

(or maybe both mechanisms), with atrazine providing poor control, while metribuzin has provided variable control dependent upon the mechanism of resistance (Vyn et al. 2007;

Schryver et al. 2017b). The objective of this study was to determine the biologically-effective-

113 dose (BED) of metribuzin applied PRE and postemergence (POST) for control of MR and MS

PS II-inhibitor-resistant waterhemp in Ontario.

5.3 Materials and Methods

Field Studies Two separate field studies were conducted over two years. Each study had three experiments in 2019 and two in 2020 for a total of 5 site-years per study. The locations near

Cottam and on Walpole Island, ON were previously confirmed to have NTSR PS II-inhibitor resistant waterhemp conferred by enhanced metabolism, while waterhemp at the sites near

Petrolia, ON were confirmed to have PS II-inhibitor TSR waterhemp due to a serine264glycine substitution (Tranel, unpublished data; Laforest, unpublished data). For the remainder of this manuscript TSR and NTSR will be referred to as metribuzin-resistant (MR) and metribuzin- sensitive (MS), respectively.

All experiments were conducted in non-cropped areas because metribuzin POST would kill soybean. The MS waterhemp sites were prepared with one pass of a tandem disc followed by a second pass with a field cultivator in 2019 and 2020. The MR waterhemp sites were prepared with a tractor-mounted roto-tiller set to a shallow depth (<10 cm) in 2019, and two passes of a field cultivator in 2020. Due to the differences in waterhemp resistance profile among sites, different cover sprays were used to control all other weed species (Table 1). At the MS sites, glyphosate (Roundup WeatherMAX®, Bayer CropScience, Calgary, AB) (450 g ae ha-1) was applied POST to control all other species. At the MR sites, imazethapyr (Pursuit®, BASF

Canada, Mississauga, ON) (75 g ai ha-1) applied PRE followed by quizalofop-p-ethyl (ASSURE®

II, AMVAC Canada, Vancouver, BC) (72 g ha-1) + Sure-Mix™ (AMVAC Canada, Vancouver,

114

BC)(0.5% v/v) applied POST were used to control annual broadleaf and grass weeds, respectively. Experimental sites were hand-weeded and hand-hoed as necessary to control all weeds other than waterhemp.

Experiments were set up in a randomized complete block design with four replications.

Plots measured 2.25 m wide by 8 m long with a 2 m alley between replicates. Each replicate included a non-treated control. Metribuzin (Sencor® 480 F, Bayer CropScience Canada, Calgary,

AB) was applied at 70, 140, 280, 560, 1120, 2240, and 4480 g ai ha-1 PRE and POST at the MS waterhemp sites. At MR waterhemp sites, metribuzin was applied at 280, 560, 1120, 2240, 4480,

8960, and 17920 g ai ha-1 PRE and POST. Pyroxasulfone/flumioxazin (Fierce®, Valent Canada,

Guelph, ON) (240 g ai ha-1) PRE was included as it is currently considered the industry standard for waterhemp control in soybean (Schryver et al. 2017b; Hedges et al. 2018b). Fomesafen

(Reflex®, Syngenta Canada, Guelph, ON) (240 g ai ha-1) + Turbocharge® (Syngenta Canada,

Guelph, ON) (0.5%) was included in all experiments as the industry standard POST (Vyn et al.

2007). Herbicides were applied using a CO2 pressurized backpack sprayer calibrated to deliver a spray volume of 200 L ha-1 at 260 kPa. The spray boom was equipped with four ULD 11002 nozzles (Pentair, New Brighton, MN, USA) spaced 50 cm apart, producing a spray width of 2.0 m. PRE herbicides were applied after the last tillage pass. Rainfall was recorded at each site

(Table 5.2). POST herbicides were applied when waterhemp reached 10 cm in height.

Waterhemp control was assessed visually as the waterhemp biomass reduction relative to the non-treated control at 2, 4, 8, and 12 WAA, on a 0 to 100 scale with 0 representing no control and 100 complete control. Waterhemp density and biomass were collected at 8 WAA. Density was determined by randomly placing two 0.25 m2 quadrats in each plot and counting the waterhemp plants inside the quadrat. The waterhemp plants were then cut at the soil surface,

115 placed in paper bags, and dried in a kiln to a constant moisture, and weighed to determine dry biomass.

Greenhouse Studies

Greenhouse experiments were conducted in 2020 in Ridgetown and Guelph, ON.

Waterhemp seeds were collected from four populations; two MS locations (collected near

Cottam and on Walpole, ON) and two MR locations (collected near Petrolia, ON). MS and MR experiments were repeated twice in time, with metribuzin being applied PRE and POST, for a total of eight greenhouse experiments. The waterhemp seeds were stratified via refrigeration at a constant temperature of 3.5°C for seven weeks prior to germination.

For the metribuzin POST experiments, waterhemp seeds were germinated in trays of soilless mixture (Agro-Mix® G5, Fafard®, Agawam, MA), transplanted into individual pots (10 cm in diameter) at the first true leaf stage containing the same soilless mixture. When the average waterhemp height was 10 cm, metribuzin was applied at 70, 140, 280, 560, 1120, 2240, and 4480 g ai ha-1 for the MS populations and 280, 560, 1120, 2240, 4480, 8960, and 17920 g ai ha-1 for the MR populations. Fomesafen (240 g ai ha-1) + Turbocharge® (0.5%) was included in both the MR and MS experiments to represent the industry standard. Applications were made in a spray chamber calibrated to deliver 200 L ha-1 using a flat fan nozzle at 280 kPa while moving

2.15 km h-1.

For metribuzin PRE experiments, trays (14.5 cm long, 10 cm wide, and 5 cm deep) were filled with approximately 3.5 cm of top soil (loamy sand, 8.4% OM, 7.3 pH). Seventy waterhemp seeds were spread evenly on soil surface of each tray and then covered with 0.5 cm of topsoil; the trays were watered and then allowed to equilibrate overnight. Metribuzin was

116 applied the following day at the rates mentioned in the previous paragraph.

Pyroxasulfone/flumioxazin (240 g ai ha-1) PRE was included to represent the industry standard for PRE herbicides. Herbicides were applied in a spray chamber; in Ridgetown, the spray chamber was calibrated to deliver 200 L ha-1 using flat fan nozzles at 280 kPa while moving 2.15 km h-1 and in Guelph, the spray chamber was calibrated to deliver 210 L ha-1 at 276 kPa while moving 4 km h-1. Following herbicide application, the trays were then returned to the greenhouse

(Ridgetown) or growth room (Guelph) and watered to ensure herbicide activation. Greenhouse experiments were set up in a completely randomized design (CRD) with two non-treated controls per replicate.

Waterhemp control was assessed visually as the waterhemp biomass reduction relative to the non-treated control at 1, 3, and 5 WAA, on a 0 to 100 scale, with 0 representing no control and 100 complete control. Fresh weight was determined at 5 WAA by cutting the plant material at the soil surface and weighing. Fresh weight is presented as fresh weight per tray.

Statistical Analysis

Both MR and MS field studies and greenhouse studies were analyzed separately.

Non-linear Regression

PROC NLIN in SAS v. 9.4 (SAS Institute, Cary NC) was used to determine parameter estimates, which were then utilized to calculate the effective doses (ED) for 50, 80, and 95% waterhemp control for each assessment timing, PRE and POST, for both the MS and MR waterhemp populations. Waterhemp control at 2, 4, 8, and 12 WAA at the PRE application timing were regressed against metribuzin doses using an ascending four parameter log-logistic

117 model (Equation 1). PRE waterhemp density and biomass were regressed against metribuzin doses using a descending four parameter log-logistic model (Equation 2). The ED of metribuzin was calculated for 50, 80, and 95% reduction in density and biomass relative to the non-treated control. For the POST study, waterhemp control data for the MS experiments were fit to an ascending four parameter log-logistic model (Equation 1), while for the MR experiments the data were fit to an exponential to maximum model (Equation 3); waterhemp density and biomass were fit to an inverse exponential model (Equation 4). Where the ED could not be calculated by the regression model or was outside the range of doses utilized, non-estimable (non-est) is presented in the Tables, as extrapolation outside the doses evaluated in these studies would be unsuitable.

Equation 1: Four Parameter Log-logistic equations (ascending)

(-b*(log(dose)-log(I ))) y = C + (D-C)/(1+e 50 )

Where: C is the lower asymptote of percent control; D is the upper asymptote of percent

control; b is the slope about I50; I50 is the dose eliciting a response equidistant between C

and D

Equation 2: Four Parameter Log-logistic equations (descending)

(b*(log(dose)-log(I ))) y = C + (D-C)/(1+e 50 )

Where: C is the lower asymptote of percent relative density or dry biomass reduction; D

is the upper asymptote of percent relative density/biomass reduction, b is slope about I50;

I50 the dose eliciting a response equidistant between C and D

118

Equation 3: Exponential to maximum

y = a - b (e -c dose)

Where a is the upper asymptote, b is the magnitude, and c is the slope

Equation 4: Inverse exponential

y = a + b (e –c dose)

Where a is the lower asymptote, b is the reduction in y from intercept to asymptote, and

c is the slope

Least-Square Means Comparisons

PROC GLIMMIX was used in SAS v. 9.4 (SAS Institute, Cary, NC) to perform a generalized linear mixed-model analysis of variance comparing the non-treated control, metribuzin (560 and 1120 g ai ha-1) and the industry standards of pyroxasulfone/flumioxazin,

PRE and fomesafen, POST. The random effects were environment, replication within environment, and the treatment by environment interaction. Herbicide treatment was considered the fixed effect. For greenhouse experiments, the random effects were environment and treatment by environment interaction. PROC UNIVARIATE was used to test for normality using the Shapiro-Wilk statistic. Waterhemp control data were modelled using a normal distribution with the identity link function. Waterhemp density and biomass were modelled using a lognormal distribution with an identity link, with data being back transformed using the omega method (M. Edwards, Ontario Agricultural College Statistician, University of Guelph, personal 119 communication). Treatment means were separated using Tukey-Kramer test (p<0.05) and letters were assigned for presentation.

5.4. Results and Discussion

5.4.1 Biologically-Effective-Dose of Metribuzin applied Preemergence for the control of Metribuzin-Sensitive and Resistant Waterhemp

The calculated metribuzin doses applied PRE for 50, 80, and 95% MS waterhemp control were 65, 175, and 538 g ai ha-1 at 2 WAA (Table 5.3). At 12 WAA, the calculated metribuzin doses applied PRE for 50, 80, and 95% MS waterhemp control increased to 133, 350, and 1070 g ai ha-1, respectively. The higher calculated doses at 12 WAA is attributed to metribuzin degradation in the soil and prolonged waterhemp emergence. The calculated metribuzin doses applied PRE for a 50, 80, and 95% decrease in MS waterhemp density were 211, 622, and 2026 g ai ha-1 and biomass were 158, 315, and 681 g ai ha-1, respectively. The higher calculated doses for a density compared to biomass reduction is attributed to the presence of greater number of small waterhemp plants of low biomass at 8 WAA.

The calculated metribuzin doses applied PRE for 50, 80, and 95% MR waterhemp control were 2461, 9385, and >17920 g ai ha-1 at 2 WAA; and increased to 7868, 17533, and >17920 at

12 WAA. The higher calculated doses, as the time interval between application and assessment increased is attributed to metribuzin degradation in the soil and prolonged waterhemp emergence. The calculated dose for a 50% reduction in MR waterhemp density was 4338 g ai ha-

1 at 8 WAA, the calculated dose for a 80 and 95% reduction in MR waterhemp density and 50,

80 and, 95% reduction in MR waterhemp biomass were greater than the highest doses used in this study. 120

The metribuzin resistant factor for the MR waterhemp biotype for a 50% control at 2, 4, 8 and 12 WAA and 50% reduction in density was 38, 47, 75, 59, and 20, respectively (Table 5.3).

The resistance factor for a 50% reduction in MR waterhemp biomass was not calculable from the doses include in this study.

Field applied metribuzin dose depends on crop tolerance and soil characteristics. In

Ontario, the registered rate of metribuzin PRE in soybean ranges from 408 to 1080 g ai ha-1

(Anonymous 2020b). For the MS waterhemp populations, the calculated dose for 95% control 12

WAA was 1070 g ai ha-1, which is similar to the maximum dose on the label. Sweat et al. (1998) reported that metribuzin (420 g ai ha-1) applied PRE controlled waterhemp 98 to 100% at 4

WAA. Hausman et al. (2013) found that metribuzin (420 g ai ha-1) controlled waterhemp 92 and

73% in 2010 and 2011, respectively at 60 days after treatment. Evans et al. (2019) documented that metribuzin applied PRE at 280 and 560 g ai ha-1 controlled PS II-resistant waterhemp 79 and

95%, respectively at 4 WAA; PS II-inhibitor resistance was due to NTSR enhanced metabolism.

PS II-resistant waterhemp control with metribuzin at 280 and 560 g ai ha-1 decreased to 67 and

88%, respectively at 6 WAA. Vennapusa et al. (2018) found similar results, metribuzin (560 g ai ha-1) controlled NTSR PS II-inhibitor (GST mediated) resistant waterhemp. In contrast, Vyn et al. (2007) found that metribuzin (1120 g ai ha -1) controlled TSR PS II-resistant waterhemp 62 and 24% at 4 and 10 WAE, respectively. In the same study, metribuzin (1120 g ai ha-1) did not reduce TSR PS II-resistant waterhemp density and biomass.

5.4.2 Biologically-Effective-Dose of Metribuzin applied Postemergence for the control of Metribuzin-Sensitive and Resistant Waterhemp

Higher calculated metribuzin doses were required when applied POST compared to PRE for waterhemp control and density and biomass reduction; metribuzin is less efficacious applied

121

POST for MR and MS waterhemp. The calculated metribuzin doses applied POST for 50, 80, and 95% MS waterhemp control were 87, 512, and 3749 g ai ha-1 at 2 WAA (Table 5.4). The calculated metribuzin doses applied POST for MS waterhemp control increased as the time interval between application and assessment increased. At 12 WAA, the calculated metribuzin doses applied POST increased to 245, 1480, and >4480 g ai ha-1 for 50, 80, and 95% control, respectively, of MS waterhemp. The calculated metribuzin dose for a 50% reduction in MS waterhemp density was 1970 g ai ha-1 and calculated metribuzin doses for a 50 and 80% reduction in MS waterhemp biomass were 1248 and 2275 g ai ha-1, respectively.

The calculated metribuzin doses applied POST for 50, 80, and 95% MR waterhemp control and 50, 80 and 95% reduction in density and biomass were non-estimable from the doses evaluated in this study.

There is limited published research on metribuzin applied POST. Patzoldt et al. (2003) reported that metribuzin applied POST at 32 and 44 g ai ha-1 reduced growth by 50% of uniformly sensitive and segregating for PS II-inhibitor resistant waterhemp populations, respectively in a greenhouse study.

5.4.3 Metribuzin-Sensitive and Resistant Waterhemp control with Metribuzin Compared to Industry Standards.

Metribuzin preemergence compared to flumioxazin/pyroxasulfone

Metribuzin (560 and 1120 g ai ha-1) was compared to the current industry standard pyroxasulfone/flumioxazin (240 g ai ha-1) applied PRE for both MS and MR waterhemp (Table

5.5). Metribuzin and pyroxasulfone/flumioxazin controlled MS waterhemp 97-100%, 96-100%,

94-99%, and 88-98% at 2, 4, 8 and 12 WAA; there was no difference in control among the three

122 treatments evaluated (Table 5.5). All treatments reduced MS waterhemp density and biomass similarly at 89-100%.

Metribuzin (560 and 1120 g ai ha-1) PRE controlled MR waterhemp 19-31%, 18-26%, 3-

8%, and 0-4% at 2, 4, 8 and 12 WAA, respectively; there was no difference in control among the two rates of metribuzin (Table 5.5). In addition, metribuzin PRE did not reduce MR waterhemp density and biomass. In contrast, pyroxasulfone/flumioxazin controlled MR waterhemp 93-99% at 2, 4, 8, and 12 WAA and reduced MR waterhemp density and biomass 97%. Schryver et al.

(2017b) and Hedges et al. (2018b) reported that pyroxasulfone/flumioxazin (240 g ai ha-1) controlled waterhemp 97% control at 12 WAA.

Metribuzin postemergence compared to fomesafen

Similar to the BED of metribuzin for both MS and MR waterhemp, control was reduced when metribuzin was applied POST compared to PRE. Metribuzin (560 and 1120 g ai ha-1) was compared to the current industry standard fomesafen (240 g ai ha-1) applied POST for bothMS and MR waterhemp (Table 5.6). Metribuzin and fomesafen controlled MS waterhemp 80-88%,

76-87%, 74-82% and 65-78% at 2, 4, 8 and 12 WAA, there was no difference in control among the three treatments evaluated. All treatments reduced MS waterhemp density and biomass similarly at 18-81% and 41-50%, respectively.

Metribuzin (560 and 1120 g ai ha-1) POST controlled MR waterhemp 17%, 5-6%, 3-4%, and 0-1% at 2, 4, 8, and 12 WAA, respectively; there was no difference in control among the two rates of metribuzin (Table 5.6). In addition, metribuzin POST did not reduce MR waterhemp density and biomass. Fomesafen controlled MR waterhemp 75, 55, 57, and 49% at 2, 4, 8, and 12

WAA, respectively and reduced MR biomass 47% but there was no decrease in density.

123

The lower control of MR compared to MS waterhemp with fomesafen may be due to greater late season emergence of waterhemp at the MR sites (Table 5.1). Prior to wide-spread

PPO-inhibitor-resistant waterhemp in Illinois, Patzoldt et al. (2002) found that fomesafen (330 g ai ha-1) controlled all 59 waterhemp populations tested in a greenhouse survey. Fomesafen (395 g ai ha-1) controlled 10-15 cm tall waterhemp 89% 2 WAA, in IL (Hausmen et al. 2016). Sweat et al. (1998) reported variable waterhemp control with fomesafen (280 g ai ha-1) POST; it provided

74-79% and 86-87 % control at Richmond and Ashland Bottoms KS. Vyn et al. (2007) reported that fomesafen (240 g ai ha-1) controlled waterhemp at Petrolia, ON in 2004 100% at 4 and 10

WAA and reduced waterhemp density and biomass similar to the weed-free control; in contrast, in 2005, fomesafen controlled waterhemp 66 and 60 % at 4 and 10 WAA, respectively. The reduced waterhemp control with fomesafen at Petrolia, ON in 2005 was attributed to prolonged waterhemp emergence. At Cottam, ON, the same rate of fomesafen controlled waterhemp 98% at

4 and 10 WAA. Field studies conducted by Hager et al. (2003) documented that fomesafen at 87,

175, and 350 g ai ha-1 controlled 5 cm tall waterhemp 63, 74, and 85% and controlled 10 cm tall waterhemp 53, 61, and 77%, respectively, at 3 WAA. Fomesafen POST provides variable waterhemp control depending on density, weed size, emergence pattern and weather conditions.

5.4.4 Greenhouse Experiments 5.4.4.1 Biologically-Effective-Dose of Metribuzin applied Preemergence for control of Metribuzin-Sensitive and Resistant Waterhemp

Greenhouse experiments were conducted in 2020; lower metribuzin doses were required for MS and MR waterhemp control when compared to the field studies. The calculated metribuzin doses applied PRE for 50, 80, and 95% MS waterhemp control were 6, 88, and 1868 g ai ha-1 at 1 WAA, respectively (Table 5.7). The higher calculated dose for 95% control was due

124 to the short time interval between application and assessment; waterhemp had germinated but was not yet controlled. At 5 WAA, the calculated metribuzin doses for 50, 80, and 95% MS waterhemp control were 9, 21, and 50 g ai ha-1, respectively. The calculated metribuzin doses applied PRE for 50, 80, and 95% MS waterhemp fresh weight reduction were 21, 68, and 100 g ai ha-1, respectively. The calculated metribuzin dose applied PRE for 50, 80, and 95% MR waterhemp control at 1, 3, and 5 WAA and MR waterhemp fresh weight reduction was ≥ 2396 g ai ha-1. The metribuzin resistant factor for the MR waterhemp biotype for a 50% control at 1, 3, and 5 WAA and 50% reduction in fresh weight was 917, 218, 266, and 163, respectively.

5.4.4.2 Biologically-Effective-Dose of Metribuzin applied Postemergence for the control of Metribuzin-Sensitive and Resistant Waterhemp

The metribuzin dose applied for 50, 80 and 95% MS and MR waterhemp control was not calculable at 1 WAA from the doses evaluated in this study (Table 5.8). The calculated metribuzin POST doses for 50, 80, and 95% MS waterhemp control were 284, 711, and 1983 g ai ha-1, respectively at 5 WAA. The calculated metribuzin POST doses for a 50, 80, and 95% MS waterhemp fresh-weight reduction were 178, 595, and 2310 g ai ha-1, respectively. The calculated metribuzin POST doses for 50, 80, and 95% MR waterhemp control were 5238, 8684, and 15285 g ai ha-1, respectively at 5 WAA. The calculated metribuzin POST doses for a 50, 80, and 95% MR waterhemp fresh-weight reduction were 5575, 8389, and 14233 g ai ha-1, respectively. The metribuzin resistant factors for the MR waterhemp biotype for a 50% control at

3 and 5 WAA and 50% reduction in fresh weight was 18, 18 and 31, respectively.

5.4.4.3 Metribuzin-Sensitive and Resistant Waterhemp control compared to Industry Standards

Metribuzin preemergence compared to flumioxazin/pyroxasulfone

125

Metribuzin (560 and 1120 g ai ha-1) and pyroxasulfone/flumioxazin PRE controlled MS waterhemp 93-100% at 1, 3, and 5 WAA, there was no difference in MS waterhemp control among the three herbicide treatments evaluated (Table 5.9). Metribuzin (560 and 1120 g ai ha-1) and pyroxasulfone/flumioxazin all reduced fresh weight by 100%.

Metribuzin (560 and 1120 g ai ha-1) PRE controlled MR waterhemp 1-12% at 1, 3, and 5

WAA, and did not reduce MR waterhemp fresh weight; there was no difference in MR waterhemp control between the two rates of metribuzin evaluated. In contrast, pyroxasulfone/flumioxazin PRE controlled MR waterhemp 100% and reduced fresh weight

100%.

Metribuzin postemergence compared to fomesafen

Metribuzin (560 and 1120 g ai ha-1) and fomesafen POST controlled MS waterhemp 20-

69, 85-93, and 74-97% at 1, 3, and 5 WAA, respectively and reduced MS waterhemp fresh weight 70-96%; there was no difference among the three herbicide treatments evaluated (Table

5.10). Metribuzin (560 and 1120 g ai ha-1) controlled MR waterhemp 6-8%, 3-7%, and 1-4% at

1, 3, and 5 WAA, respectively and did not reduced MR waterhemp fresh weight. In contrast, fomesafen controlled MR waterhemp 73-85% at 1, 3, and 5 WAA and did not reduce fresh weight.

5.5.1 Conclusion

Metribuzin applied PRE provided better control of MS and MR waterhemp than metribuzin applied POST. Based on the field studies, the calculated metribuzin doses applied PRE for 50, 80, and 95% MS waterhemp control were 133, 350, and 1070 g ai ha-1, respectively; the calculated

126 metribuzin doses for 50 and 80% MR waterhemp control were 7868 and 17 533 g ai ha-1, respectively at 12 WAA. The calculated metribuzin doses applied POST for 50 and 80% MS waterhemp control were 245 and 1480 g ai ha-1, respectively, at 12 WAA. The metribuzin doses applied POST in this study did not achieve 50% control of MR waterhemp. Metribuzin (560 g ai ha-1 and 1120 g ai ha-1), and pyroxasulfone/flumioxazin PRE controlled MS waterhemp 88, 95, and 98%, respectively; the same treatments controlled MR waterhemp 0, 4, and 93%, respectively, at 12 WAA. Metribuzin (560 g ai ha-1 and 1120 g ai ha-1), and fomesafen POST controlled MS waterhemp 65, 70, and 78%, respectively, while the same treatments controlled MR waterhemp 0,

1, and 49% control, respectively, at 12 WAA. Results from greenhouse studies followed similar trends. The metribuzin PRE resistance factor for the MR Waterhemp biotype for a 50% control at

2, 4, 8 and 12 WAA and 50% reduction in density was 20-75. The metribuzin POST resistant factor for the MR waterhemp biotype could not be calculated from the rates evaluated in this study.

There was a substantial effect of PS II resistance mechanisms on two PS II-resistant waterhemp biotypes. The waterhemp with NTSR PS II-resistant (enhanced metabolism) was controlled with metribuzin applied PRE and POST while the TSR PS II-resistant (serine264glycine altered target site) waterhemp was not controlled with metribuzin. This study illustrates the need for determining the herbicide resistance mechanism on a field-by-field basis since it influences weed management on commercial farms in Ontario.

127

Table 5.1 Year, location, resistance profile, and application information for ten field experiments conducted on the biologically- effective-dose of metribuzin, applied preemergence and postemergence, for the control of waterhemp (Amaranthus tuberculatus) with different mechanisms of resistance to photosystem II-inhibiting herbicides in 2019 and 2020 in Ontario, Canada. Percent Resistantc % Waterhemp Waterhemp 2 5 5 9 14 Application Application heighta densityb Year Location (ATR) (MET) timing Date (cm) (plants m-2) 2019 Petrolia 1 99 98 - 18 - PRE June 28 - - POST July 23 9.5 62 Petrolia 2 - - - - - PRE June 28 - - POST July 23 10 86 Petrolia 3 - - - - - PRE June 28 - - POST Jul 23 9 95 Cottam 96 34 - 100 - PRE June 21 - - POST July 10 9 220 Walpole 22 6 - 79 - PRE June 27 - - Island 1 POST July 16 10 43 Walpole - - - - - PRE June 27 - - Island 2 POST July 16 9 52 2020 Petrolia 1 95 92 98 23 29 PRE May 22 - - POST June 24 10 280 Petrolia 2 96 94 98 26 31 PRE May 25 - - POST June 26 9.5 231 Cottam 68 54 29 64 43 PRE May 23 - - POST June 30 8.5 517 Walpole 54 30 18 96 17 PRE June 9 - - Island POST July 7 9.0 53 aAverage height of plant on day of application. Mean of two measurements per replication. bMean density based on the average of two density counts in non-treated control within each replication. cNumbers representing WSSA Group 2 (ALS-inhibitors), Group 5 (Photosystem II-inhibitors); ATR- Atrazine; Met- Metribuzin, Group 9 (ESPS-inhibitors), and Group 14 (PPO-inhibitors)

128

Table 5.2 Year, location, soil characteristics, and rainfall information for ten field experiments conducted on the biologically- effective-dose of metribuzin, applied preemergence and postemergence, for the control of waterhemp (Amaranthus tuberculatus) with different mechanisms of resistance to photosystem II-inhibiting herbicides in 2019 and 2020 in Ontario, Canada.a Year Location Soil Characteristics Rainfall (mm) 2019 SOM Texture (%) pH 0-7 DAAb 8-14 DAA 0-84 DAA Petrolia 1 Sandy clay loam 2.7 7.2 9.8 23.3 170.1 Petrolia 2 Sandy clay loam 1.9 7.6 9.8 23.3 170.1 Petrolia 3 Clay 4.1 6.6 9.8 23.3 170.1 Cottam Sandy loam 2.6 6.0 0.3 29.4 175.6 Walpole Island 1 Loamy sand 2.2 7.6 9.3 28.0 198.6 Walpole Island 2 Loamy sand 2.5 7.6 9.5 31.6 212.0 2020 Petrolia 1 Clay loam 3.7 7.5 16.8 15.6 261.5 Petrolia 2 Clay loam 2.3 7.7 6.6 12.7 314.2 Cottam Sandy loam 2.6 5.9 17.6 5.6 154.0 Walpole Island Sandy loam 2.7 8.0 22.0 26.2 233.2 aAbbreviation: SOM, soil organic matter of the upper 15 cm of the soil profile; DAA, days after application bDays after preemergence application.

129

Table 5.3 Parameter estimates and calculated effective doses of metribuzin applied preemergence (PRE) for control, and dry biomass and density reduction of metribuzin-sensitive and resistant waterhemp (Amaranthus tuberculatus) biotypes from ten field experiments completed in 2019 and 2020 in Ontario, Canada.a Metribuzin-sensitive Parameter estimates ( SE ) Calculated metribuzin dose Variable C D b I50 ED50 ED80 ED95 (g ai ha-1) Control 2 WAA 0.08 (2.46) 100 (0) 1.39 (0.14) 64.73 (5.11) 65 175 538 Control 4 WAA 0 (0) 100 (0) 1.46 (0.15) 63.23 (4.41) 63 163 475

Control 8 WAA 0.02 (2.58) 100 (0) 1.45 (0.13) 83.32 (5.66) 83 217 635 Control 12 WAA 0 (0) 99.68 (2.56) 1.44 (0.17) 132.3 (10.25) 133 350 1070 Densityd 0 (0) 93.93 (22.83) 1.33 (0.78) 232.7 (160.60) 211 622 2026

Dry Biomassd 0 (0) 99.43 (10.81) 2.02 (0.69) 159.1 (34.50) 158 315 681 Metribuzin-resistant Calculated metribuzin dose Variable C D b I50 ED50 ED80 ED95

(g ai ha-1) Control 2 WAA 2.44 (3.65) 100 (0) 1.05 (0.11) 2581.5 (329.6) 2461 9385 Non-est.c Control 4 WAA 3.64 (3.27) 100 (0) 1.13 (0.12) 3156.7 (359.9) 2953 10331 Non-est. Control 8 WAA 1.16 (2.28) 100 (0) 1.61 (0.19) 6277.9 (470.7) 6187 14717 Non-est. Control 12 WAA 0 (0) 100 (0) 1.73 (0.17) 7867.6 (466.7) 7868 17533 Non-est. Densityd 23.30 (16.73) 100 (0) 2.44 (1.79) 3354.4 (1283.7) 4338 Non-est. Non-est. Dry Biomassd 48.44 (18.55) 100 (0) 4.24 (6.58) 8898.6 (2512.7) Non-est. Non-est. Non-est. aAbbreveiation: WAA, weeks after application b (-b*(log(dose)-log(I )))) (b*(log(dose)-log(I )))) Regression parameters; y = C + (D-C)/(1+e 50 (ascending); y = C + (D-C)/(1+e 50 (descending); C = lower asymptote of percent control or percent relative density/dry biomass, D = upper asymptote of percent control or percent relative density/biomass, b = slope about I50, I50 = the dose eliciting a response equidistant between C and D, EDn, effective dose to elicit response level n. cNon-est., predicted dose for any parameter was outside of the range of doses used in this study or could not be computed by the model. dPercent of density and dry biomass relative to non-treated control plots within replications

130

Table 5.4 Parameter estimates and calculated effective doses of metribuzin applied postemergence (POST) for control, and dry biomass and density reduction of metribuzin-sensitive and resistant waterhemp (Amaranthus tuberculatus) biotypes from ten field experiments completed in 2019 and 2020 in Ontario, Canada.a Metribuzin-sensitive Parameter estimates ( SE ) Calculated metribuzin dose Variable C D b I50 ED50 ED80 ED95 (g ai ha-1) Control 2 WAAb 0.19 (6.14) 100 (0) 0.78 (0.13) 87.48 (22.25) 87 512 3749 Control 4 WAAb 0 (0) 100 (0) 0.77 (0.12) 114.5 (21.94) 115 693 Non-est.e

Control 8 WAAb 0 (0) 99.82 (13.4) 0.77 (0.27) 175.4 (76.28) 176 1074 Non-est. Control 12 WAAb 0 (0) 100 (0) 0.77 (0.13) 244.5 (46.39) 245 1480 Non-est. A B C Densitydf 0 (0) 90.30 (14.20) 0.0003 (0) 1970 Non-est. Non-est. Dry Biomassdf 0 (0) 105.7 (10.06) 0.0006 (0) 1248 2275 Non-est. Metribuzin-resistant Calculated metribuzin dose Variable A B C ED50 ED80 ED95 (g ai ha-1) Control 2 WAAc 49.47 (5.04) 41.61 (4.79) 0.0002 (0) Non-est. Non-est. Non-est. Control 4 WAAc 29.96 (6.31) 28.38 (5.91) 0.0001 (0) Non-est. Non-est. Non-est. Control 8 WAAc 28.86 (7.22) 28.74 (6.78) 0.0001 (0) Non-est. Non-est. Non-est. Control 12 WAAc 67.42 (27.61) 69.17 (26.98) 0.00005 (0) Non-est. Non-est. Non-est. Densitydf 0 (0) 81.22 (4.83) 0.00001 (0) Non-est. Non-est. Non-est. Dry Biomassdf 0 (0) 98.49 (3.85) 0.00001 (0) Non-est. Non-est. Non-est. aAbbreveiation: WAA, weeks after application b (-b*(log(dose)-log(I )))) Regression parameters; y = C + (D-C)/(1+e 50 (ascending); C = lower asymptote of percent control, D = upper asymptote of percent control b = slope about I50, I50 = the dose eliciting a response equidistant between C and D, EDn, effective dose to elicit response level n. cRegression parameters: y = a - b (e -c dose); Where a is the upper asymptote, b is the magnitude, and c is the slope dRegression parameters: y = a + b (e –c dose); Where a is the lower asymptote, b is the reduction in y from intercept to asymptote, and c is the slope eNon-est., predicted dose for any parameter was outside of the range of doses used in this study or could not be computed by the model. fPercent of density and dry biomass relative to non-treated control plots within replications

131

g ai ha-1 ---%--- plants m-2 g m-2

Metribuzin Sensitive Non-treated control - 0 b 0 b 0 b 0 b 223 b 229 b

Metribuzin 560 97 a 96 a 94 a 88 a 24 a 11 a

Metribuzin 1120 100 a 100 a 99 a 95 a 5 a 1 a

Pyroxasulfone/flumioxazin 240 99 a 99 a 99 a 98 a 1 a 1 a

Metribuzin resistant Non-treated control - 0 c 0 c 0 c 0 b 110 b 283 b

Metribuzin 560 19 b 18 b 3 bc 0 b 133 b 257 b

Metribuzin 1120 31 b 26 b 8 b 4 b 93 b 281 b

Pyroxasulfone/flumioxazin 240 99 a 99 a 96 a 93 a 4 a 10 a aAbbreviations: WAA, weeks after application bMeans followed by the same letter within a column, under the same heading, are not statistically different at α=0.05 according to Tukey-Kramer’s multiple range test.

132

g ai ha-1 ---%--- plants m-2 g m-2

Metribuzin sensitive Non-treated control - 0 b 0 b 0 b 0 b 114 b 326 a

Metribuzin 560 80 a 76 a 74 a 65 a 93 ab 192 a

Metribuzin 1120 86 a 80 a 75 a 70 a 71 ab 163 a

Fomesafenc 240 88 a 87 a 82 a 78 a 22 a 162 a

Metribuzin resistant Non-treated control - 0 c 0 b 0 b 0 b 106 a 512 b

Metribuzin 560 17 b 5 b 3 b 0 b 86 a 459 b

Metribuzin 1120 17 b 6 b 4 b 1 b 72 a 400 b

Fomesafenc 240 75 a 66 a 57 a 49 a 86 a 272 a aAbbreviations: WAA, weeks after application bMeans followed by the same letter within a column, under the same heading, are not statistically different at α=0.05 according to Tukey-Kramer’s multiple range test. cIncluded Turbocharge® at 0.5% vol/vol.

133

(g ai ha-1)

Control 1 WAAb 0.02 (2.32) 100 (0) 0.51 (0.09) 5.81 (3.47) 6 88 1868

Control 3 WAAb 0 (0.64) 100 (0) 1.41 (0.27) 11.27 (4.34) 11 30 91

Control 5 WAAb 0 (0.4) 100 (0) 1.75 (0.59) 9.30 (6.59) 9 21 50

Fresh Weightbe 0 (0) 100 (1.05) 1.90 (0.52) 21.31 (7.39) 21 68 100

Metribuzin resistant Calculated metribuzin dose

Variable C D b I50 ED50 ED80 ED95

(g ai ha-1)

Control 1 WAAb 2.08 (2.64) 100 (0) 2.63 (0.27) 5993.7 (274.8) 5504 9381 16993

Control 3 WAAb 0 (0) 95.05 (3.55) 2.43 (0.36) 2295.3 (166.4) 2396 4564 Non-est.

Control 5 WAAb 0 (0) 96.18 (3.40) 2.51 (0.36) 2322.0 (157.0) 2397 4389 13340

Fresh Weightbe 1.54 (7.91) 100 (0) 2.74 (0.8) 3388.2 (546.1) 3427 5786 11351 aAbbreveiation: WAA, weeks after application b (-b*(log(dose)-log(I )))) (b*(log(dose)-log(I )))) Regression parameters; y = C + (D-C)/(1+e 50 (ascending); y = C + (D-C)/(1+e 50 (descending); C = lower asymptote of percent control or percent relative fresh weight, D = upper asymptote of percent control or percent relative fresh weight, b = slope about I50, I50 = the dose eliciting a response equidistant between C and D, EDn, effective dose to elicit response level n. dNon-est., predicted dose for any parameter was outside of the range of doses used in this study or could not be computed by the model. ePercent of fresh weight relative to non-treated control tray within replication. 134

(g ai ha-1)

Control 1 WAAb 0 (0) 24.43 (9.98) 0.75 (0.63) 268.7 (368.1) Non-est.c Non-est. Non-est.

Control 3 WAAb 0.35 (4.96) 91.31 (3.17) 2.67 (0.59) 181.0 (18.42) 194 367 Non-est.

Control 5 WAAb 0.73 (7.15) 100 (0) 1.52 (0.32) 287.2 (52.73) 284 711 1983

Fresh Weightbd 0 (0) 99.29 (9.06) 1.15 (0.27) 179.7 (47.89) 178 595 2310

Metribuzin resistant Calculated metribuzin dose

Variable C D b I50 ED50 ED80 ED95

(g ai ha-1)

Control 1 WAAb 2.30 (3.99) 42.72 (6.04) 1.84 (0.97) 2327.0 (735.20) Non-est. Non-est. Non-est.

Control 3 WAAb 2.47 (2.04) 82.04 (3.04) 4.03 (0.70) 3246.9 (213.5) 3581 8007 Non-est.

Control 5 WAAb 1.16 (1.82) 100 (0) 2.76 (0.32) 5283.1 (254.7) 5238 8684 15285

Fresh Weightbd 1.62 (8.49) 99.45 (3.37) 3.53 (1.16) 5541.0 (614.7) 5575 8389 14233 aAbbreveiation: WAA, weeks after application b (-b*(log(dose)-log(I )))) (b*(log(dose)-log(I )))) Regression parameters; y = C + (D-C)/(1+e 50 (ascending); y = C + (D-C)/(1+e 50 (descending); C = lower asymptote of percent control or percent relative fresh weight, D = upper asymptote of percent control or percent relative fresh weight, b = slope about I50, I50 = the dose eliciting a response equidistant between C and D, EDn, effective dose to elicit response level n. cNon-est., predicted dose for any parameter was outside of the range of doses used in this study or could not be computed by the model. dPercent of fresh weight relative to non-treated control trays within replication. 135

Treatment Rate Control Fresh weight 1 WAA 3 WAA 5 WAA g ai ha-1 ---%--- g

Metribuzin sensitive Non-treated control - 0 b 0 b 0 b 20 b

Metribuzin 560 93 a 100 a 100 a 0 a

Metribuzin 1120 94 a 100 a 100 a 0 a

Pyroxasulfone/flumioxazin 240 100 a 100 a 100 a 0 a

Metribuzin resistant Non-treated control - 0 b 0 b 0 b 19 b

Metribuzin 560 1 b 2 b 1 b 19 b

Metribuzin 1120 4 b 12 b 10 b 18 b

Pyroxasulfone/flumioxazin 240 100 a 100 a 100 a 0 a aAbbreviations: WAA, weeks after application bMeans followed by the same letter within a column, under the same heading, are not statistically different at α=0.05 according to Tukey-Kramer’s multiple range test.

136

Treatment Rate Control Fresh weight 1 WAA 3 WAA 5 WAA g ai ha-1 ---%--- G

Metribuzin sensitive Non-treated control - 0 b 0 b 0 c 2.7 b

Metribuzin 560 20 ab 85 a 74 b 0.4 b

Metribuzin 1120 23 ab 93 a 97 a 0.1 b

Fomesafenc 240 69 a 85 a 82 a 0.8 b

Metribuzin resistant Non-treated control - 0 b 0 b 0 b 2.6 a

Metribuzin 560 6 ab 3 b 1 b 2.4 a

Metribuzin 1120 8 ab 7 b 4 b 2.3 a

Fomesafenc 240 73 a 85 a 80 a 0.7 a aAbbreviations: WAA, weeks after application bMeans followed by the same letter within a column, under the same heading, are not statistically different at α=0.05 according to Tukey-Kramer’s multiple range test. cIncluded Turbocharge® at 0.5% vol/vol

137

Chapter 6: General Discussion

6.1 Contributions

This research has advanced the science of GR Canada fleabane control in soybean with preplant (PP) applications of bromoxynil, pyraflufen-ethyl/2,4-D, and tiafenacil, applied alone and when tankmixed with metribuzin. Furthermore, this research advanced our understanding of photosystem II (PS II)-inhibitor resistance in waterhemp conferred by non-target-site resistance

(NTSR) and target-site resistance (TSR) mechanisms and its implications on management of these weed biotypes on commercial farms in Ontario.

The predicted biologically-effective-doses (BED) of bromoxynil for 50 and 80% GR

Canada fleabane control were 98 and 277 g ai ha-1 at 8 weeks after application (WAA). When tankmixed with metribuzin (400 g ai ha-1), the predicted doses of bromoxynil for 50, 80, and

95% GR Canada fleabane control were 10, 25, and 54 g ai ha-1, respectively, at 8 WAA.

Bromoxynil (280 g ai ha-1) plus metribuzin (400 g ai ha-1) controlled GR Canada fleabane 97% which was similar to the industry standards of saflufenacil + metribuzin (25+ 400 g ai ha-1) and

-1 glyphosate/dicamba + saflufenacil (1800 + 25 g ai ha ) providing 99 and 100% control, respectively at 8 WAA.

For calculated BED of pyraflufen-ethyl/2,4-D for 50, 80, and 95% GR Canada fleabane control were 390, 1148, and >2108 g ai ha-1, respectively at 8 WAA. The addition of metribuzin

(400 g ai ha-1) to pyraflufen-ethyl/2,4-D reduced the doses of pyraflufen-ethyl/2,4-D for 50, 80,

-1 and 95% GR Canada fleabane control to 19, 46, and 201 g ai ha , respectively at 8 WAA.

Pyraflufen-ethyl/2,4-D + metribuzin (527 + 400 g ai ha-1) controlled GR Canada fleabane 97%,

138 similar to the current industry standards of saflufenacil + metribuzin and glyphosate/dicamba + saflufenacil, which provided 99% control at 8 WAA.

For calculated BED of tiafenacil for 50, 80, and 95% GR Canada fleabane control were

21, 147, and >200 g ai ha-1, respectively at 8 WAA. The calculated doses for tiafenacil + metribuzin (400 g ai ha-1) for 50, 80, and 95% GR Canada fleabane control were 1, 3, and >200 g ai ha-1, respectively at 8 WAA. Tiafenacil + metribuzin at 25 + 400 and 50 + 400 g ai ha-1 controlled GR Canada fleabane 88 and 93%, respectively, which was similar control to the industry standards of saflufenacil + metribuzin and glyphosate/dicamba + saflufenacil that provided 98 to 100% control, at 8 WAA.

Metribuzin applied PRE at the calculated doses of 133, 350, and 1070 g ai ha-1 controlled metribuzin-senestive (MS) waterhemp 50, 80, and 95%; the calculated doses of 7868 and 17533 g ai ha-1 controlled metribuzin-resistant (MR) waterhemp 50 and 80%, respectively at 12 WAA.

Metribuzin applied POST at the calculated doses of 245 and 1480 g ai ha-1 controlled MS waterhemp 50 and 80%, respectively; the calculated doses for 50, 80, and 95% MR waterhemp control was greater than the highest dose (17920 g ai ha-1) included in this study. Metribuzin at

560 and 1120 g ai ha-1 and pyroxasulfone/flumioxazin (240 g ai ha-1) applied PRE controlled MS waterhemp while only pyroxasulfone/flumioxazin controlled MR waterhemp. Similarly, metribuzin at 560 and 1120 g ai ha-1 and fomesafen applied POST controlled MS waterhemp but only fomesafen controlled MR waterhemp. Greenhouse studies were conducted with similar responses. NTSR PS II-resistant waterhemp (enhanced metabolism) is controlled with metribuzin applied PRE and POST, in contrast TSR PS II-resistant waterhemp (serine264glycine altered target site) is not controlled with metribuzin.

139

6.2 Limitations

After completing my MSc research, I would like to offer the following suggestions to improve future studies. Due to the variable emergence pattern of Canada fleabane, average plant height at application was difficult to assess and varied from plot-to-plot and replicate-to-replicate within a location. The variable height at time of application may have had an effect on herbicide efficacy. The prolonged Canada fleabane emergence pattern and concomitant range in plant size at the time of herbicide application, may have disproportionately impacted the contact herbicides used, such as bromoxynil and tiafenacil. Some Canada fleabane seedlings were only at the cotyledon stage at application with the larger fleabane plants creating a canopy over these emerging seedlings resulting in reduced herbicide interception and absorption. This canopy may have inhibited the contact herbicides from reaching the smaller weeds in sufficient quantity for control. In future research, fall applications of glufosinate could be made to ensure more uniform

Canada fleabane height at the time of herbicide application in the spring. Greenhouse studies could also be conducted to determine the impact of weed height on control.

The study on the BED of tiafenacil for control of GR Canada fleabane included methylated seed oil (MSO) in 2020, but not in 2019. No difference in control was observed between 2019 and 2020, however, for consistency MSO should be included in future studies.

When this research was initiated, there was only one area in the province with known MR

(TSR) waterhemp. All MR waterhemp sites were with in one kilometer of each other resulting in all sites having similar rainfall and soil characteristics. Furthermore, overall weed density at

Cottam and Walpole Island varied, resulting in differing levels of control. Ideally, sites would be selected with similar weed densities. The MS and MR waterhemp populations were confirmed with PS II-inhibitor resistance endowed by enhanced metabolism and an amino acid substitution,

140 respectively. The MR sites had a higher prevalence of PS II-inhibitor resistance compared to the

MS sites. Therefore, a larger portion of waterhemp plants at the MS locations that had no PS II- inhibitor resistance, while this is the nature of field research variability, comparing populations with similar percentages of PS II-inhibitor resistance would be ideal. All waterhemp trials were conducted in non-cropped area therefore there was no crop competition. This may have decreased late-season waterhemp control ratings and increased waterhemp density and biomass compared to a cropped system.

In all four studies conducted, “non-estimable” was utilized when the calculated herbicide dose was greater than the highest dose in the research protocol. Utilizing 95% control, reduction in biomass and density relative to the non-treated control, and percent of the industry standard yield is a high standard, particularly with high weed density and large weeds at the time of application. In future studies, a 90% response level could be utilized to increase the number of calculable doses.

6.3 Future Research

The completion of my MSc research has raised several research topics regarding GR

Canada fleabane and MHR waterhemp management that would benefit Ontario soybean producers.

While effective herbicide solutions were identified in these studies, herbicides alone are not the long-term solution for herbicide-resistant weeds. Integrated weed management (IWM) programs should be developed to ensure long-term sustainable weed management on Ontario farms. For GR Canada fleabane, utilization of tillage and cover crops has provided variable

141 control. Future research should focus on the development of IWM programs for GR Canada fleabane control including tillage, cover crops, and crop rotations that would be beneficial to the long term success of GR Canada fleabane control in Ontario. Establishment of a long-term GR

Canada fleabane IWM project similar to what is currently being conducted in waterhemp would be valuable as a systems approach would be more sustainable than relying on herbicides alone.

This IWM Canada fleabane study could include, but is not limited to, crop rotation, tillage, cover crops, row spacing and multiple effective herbicide modes of action. Furthermore, cover crops that are economical for grain farmers such as red clover (Trifolium pratense L.), and for livestock farmers, such as cereal rye (Secale cereale L.) and triticale (X Triticosecale), should be investigated further for both Canada fleabane control and overall economic benefits.

For the BED of bromoxynil, pyraflufen ethyl/2,4-D and tiafenacil, metribuzin (400 g ai ha-1) was consistently used as a tank mix partner. Lower doses of metribuzin should be investigated to determine efficacy on Canada fleabane when tankmixed with the aforementioned herbicides. Lower rates of metribuzin could result in reduced environmental loading of pesticides and cost savings for farmers. Although bromoxynil and metribuzin share the same mode of action, but different binding sites, a need for other tankmix partners may be necessitated in the case of PS II-inhibitor resistance in Canada fleabane. Research should be conducted on other tankmix partners for preplant (PP) applications such as saflufenacil and glufosinate in IP and RR soybean, dicamba in dicamba-resistant soybean, and 2,4-D in Enlist™ soybean system.

Glyphosate was included in all GR Canada fleabane treatments. With increased societal pressure to reduce the use of glyphosate, future research should be conducted without the use of glyphosate to reflect potential future scenarios for Ontario farmers and determine the impact this may have on weed control.

142

Bromoxynil is a contact herbicide with no residual control which allows bromoxynil to be at a PP application timing to control GR Canada fleabane. Crop tolerance should be examined in additional crops such as dry common beans and azuki bean, as there are limited options to control of GR Canada fleabane in many low acreage crops. Research has been conducted and found synergistic interactions between PS II-inhibiting herbicides and 4-hydroxyphenylpyruvate dioxygenase (HPPD)-inhibiting herbicides. With the pending registration of GT27™ soybean, the interaction between bromoxynil and isoxaflutole should be investigated to determine efficacy on GR Canada fleabane, and other troublesome weeds within Ontario, as well as determine the joint interactions of the tankmix combination. Good spray coverage is essential to provide the best weed control with contact herbicides. If bromoxynil utilization as a PP burndown in soybean is increased, studies should be conducted to determine the impact of water volume and nozzle selection of bromoxynil efficacy on GR Canada fleabane.

Pyraflufen-ethyl/2,4-D alone provided poor control of GR Canada fleabane at the label rate (527 g ai ha-1). Future research should be conducted to determine if pyraflufen-ethyl and 2,4-

D applied alone at the label rates provide effective control. Both herbicides must be effective to be considered multiple effective modes of action. Tankmix interactions could analyzed to determine the joint interactions of the herbicide combination. Glyphosate/dicamba + saflufenacil has proven to be effective in all three GR Canada fleabane studies conducted. If pyraflufen-ethyl is effective in controlling GR Canada fleabane, pyraflufen-ethyl + dicamba may be an effective tankmix mix option for the control of GR Canada fleabane.

Tiafenacil is a relatively new herbicide with limited previous research on the majority of crops and weed species. Further research that should be conducted to classify the interaction between tiafenacil and glyphosate, determine the most efficacious adjuvant for tiafenacil, time-

143 of-day studies, and the impact of weed height at application timing. Further research should also be conducted on the utilization of tiafenacil in other crops such as edible beans, corn, and wheat.

Saflufenacil is commonly utilized as a pre-harvest desiccant and a PP burndown, research should be conducted to determine if tiafenacil may also have a similar, or potentially better, fit at a pre- harvest application timing. Greenhouse studies should be conducted to determine if tiafenacil could be effective on other weed species such as multiple-herbicide-resistant waterhemp.

Furthermore, effective tankmix partners should be identified for tiafenacil. Glufosinate has been reported to be synergistic when mixed with other PPO inhibiting herbicides, utilizing tiafenacil in a similar manner could be effective at controlling problematic weeds in Ontario.

MHR waterhemp is expanding in Ontario and its control in corn and soybean is an increasing challenge. To date the numbers of fields with waterhemp in Ontario remains relatively low but is growing every year. Effective control options in corn and soybean have been identified in previous research. Future research should be conducted to determine how waterhemp seed is spreading into, and within, Ontario. Research should be conducted to determine the potential adaptability and feasibility of programs and policies that would limit the introduction and spread of waterhemp in Ontario. Identifying economic incentive programs, best management practices, and predicted trends would help limit the spread of waterhemp in

Ontario. Since waterhemp’s long emergence pattern allows it to germinate, grow, and produce seed far into the growing season beyond the residual activity of many herbicides, harvest weed seed control (HWSC) is a relatively new innovation that should be evaluated. HSWC may have potential to reduce the overall weed seedbank and be one component of long-term diversified integrated waterhemp management program for Ontario farmers.

144

Chapter 7: Literature Cited

Alibhai, M. F., Cajacob, C., Feng, P. C., Heck, G. R., Qi, Y., Flasinski, S., & Stallings, W. C. (2010). U.S. Patent No. 7,723,575. Washington, DC: U.S. Patent and Trademark Office.

Anderson, M. P., & Gronwald, J. W. (1991). Atrazine resistance in a velvetleaf (Abutilon theophrasti) biotype due to enhanced glutathione S-transferase activity. Plant Physiol. 96:104-109.

Anderson, D. D., Roeth, F. W., & Martin, A. R. (1996). Occurrence and control of triazine- resistant common waterhemp (Amaranthus rudis) in field corn (Zea mays). Weed Technol. 10:570-575.

Anonymous. (2017). Sencor 480F Herbicide. Bayer CropScience. Inc. Calgary, Alberta.

Anonymous. (2018). Pyro Herbicide. Nufarm Agriculture Inc. Calgary, Alberta.

Anonymous. (2019a). Pardner 280 EC Herbicide Label. Bayer CropScience Inc. Calgary, Alberta.

Anonymous. (2019b). Blackhawk Herbicide. Nufarm Agriculture Inc. Calgary, Alberta.

Anonymous. (2020a). Tiafenacil 339SC Herbicide label. Concord, Ohio, United States: ISK Biosciences Corporation. EPA Registration No. 71512

Anonymous. (2020b). Sencor 480 F Herbicide. Bayer CropScience. Inc. Calgary, Alberta.

Auwers, K., & Reis, J. (1896). Ueber einige neue Derivate des p‐Oxybenzaldehyds, des p‐ Cyanphenols und der p‐Oxybenzoësäure. Berichte der deutschen chemischen Gesellschaft. 29:2355-2360.

Baerson, S. R., Rodriguez, D. J., Tran, M., Feng, Y., Biest, N. A., & Dill, G. M. (2002). Glyphosate-resistant goosegrass. Identification of a mutation in the target enzyme 5- enolpyruvylshikimate-3-phosphate synthase. Plant Physiol. 29:1265-1275.

Bajwa, A. A., Mahajan, G., & Chauhan, B. S. (2015). Nonconventional weed management strategies for modern agriculture. Weed Sci. 63:723-747.

Baldwin, F. L., & Santelmann, P. W. (1980). Weed Sci in integrated pest management. BioScience. 30:675-678.

Banks, P. A., & Robinson, E. L. (1982). The influence of straw mulch on the soil reception and persistence of metribuzin. Weed Sci. 30:164-168.

Barnes, J., Johnson, B., Gibson, K., & Weller, S. (2004). Crop Rotation and Tillage System Influence Late‐Season Incidence of Giant Ragweed and Horseweed in Indiana Soybean. Online. Crop Management. doi:10.1094/CM‐2004‐0923‐02‐BR.

145

Barrentine, W. L., & Oliver, L. R. (1977). Competition, threshold levels, and control of common cocklebur in soybeans. Miss. Agric. Exp. Stn. and Arkansas Agric. Exp. Stn. Bull. No. 83. P. 27.

Beckie, H. J. (2006). Herbicide-resistant weeds: management tactics and practices. Weed Technol. 20:793-814.

Beckie, H. J. (2011). Herbicide‐resistant weed management: focus on glyphosate. Pest Manag. Sci. 67:1037-1048.

Beckie, H. J., & Tardif, F. J. (2012). Herbicide cross resistance in weeds. J. Crop Prot. 35:15-28.

Bell, M. S., Hager, A. G., & Tranel, P. J. (2013). Multiple resistance to herbicides from four site- of-action groups in waterhemp (Amaranthus tuberculatus). Weed Sci. 61:460-468.

Benbrook, C. M. (2016). Trends in glyphosate herbicide use in the United States and globally. Environ. Sci. Eur. 28:3–19.

Benoit, L., Hedges, B., Schryver, M. G., Soltani, N., Hooker, D. C., Robinson, D. E., & Sikkema, P. H. (2019). The first record of protoporphyrinogen oxidase and four-way herbicide resistance in eastern Canada. Can. J. Plant Sci. 100:327-331.

Bensch, C., Horak, M., & Peterson, D. (2003). Interference of redroot pigweed (Amaranthus retroflexus), Palmer amaranth (A. palmeri), and common waterhemp (A. rudis) in soybean.Weed Sci. 51: 37-43.

Bernard, R.L., G.A. Juvik, E.E. Hartwig, & C.J. Edwards, Jr. (1988). Origins and pedigrees of public soybean varieties in the United States and Canada. USDA Technical Bulletin No. 1746. U.S. Gov. Print. Office, Washington, DC.

Bhowmik, P. C. & M. M. Bekech. (1993). Horseweed (Conyza canadensis) seed production, emergence, and distribution in no-tillage and conventional-tillage corn (Zea mays). Agron. Trends Agric. Sci. 1:67–71.

Blackshaw, R. E., Harker, K. N., O'Donovan, J. T., Beckie, H. J., & Smith, E. G. (2008). Ongoing development of integrated weed management systems on the Canadian prairies. Weed Sci. 56:146-150.

Board, J. E., B. G. Harville, A. M. Saxton. (1990). Narrow-row seed-yield enhancement in determinate soybean. Agron. J. 82:64–68.

Bradley, K. W. (2006). A review of the effects of row spacing on weed management in corn and soybean. Crop Manag. 5:1-10.

Bradshaw, L. D., Padgette, S. R., Kimball, S. L., & Wells, B. H. (1997). Perspectives on glyphosate resistance. Weed Technol. 11:189-198.

146

Brown, S. M., & Whitwell, T. (1988). Influence of tillage on horseweed, Conyza canadensis. Weed Technol. 2:269-270.

Bruce, J. A., & Kells, J. J. (1990). Horseweed (Conyza canadensis) control in no-tillage soybeans (Glycine max) with preplant and preemergence herbicides. Weed Technol. 4:642-647.

Bryson, C. T., & DeFelice, M. S. (2010) Weeds of the Midwestern United States and Central Canada. University of Georgia Press. P.83. Athens, Georgia, United States.

Buckland, J. L., Collins, R. E., & Pullin, E. M. (1973). Metabolism of bromoxynil octanoate in growing wheat. Pestic. Sci. 4:149-162.

Budd, C. M., Soltani, N., Robinson, D. E., Hooker, D. C., Miller, R. T., & Sikkema, P. H. (2016a). Control of glyphosate resistant Canada fleabane with saflufenacil plus tankmix partners in soybean. Can. J. Plant Sci. 96:989-994.

Budd, C. M., Soltani, N., Robinson, D. E., Hooker, D. C., Miller, R. T., & Sikkema, P. H. (2016b). Glyphosate-resistant horseweed (Conyza canadensis) dose response to saflufenacil, saflufenacil plus glyphosate, and metribuzin plus saflufenacil plus glyphosate in soybean. Weed Sci. 64:727-734.

Budd, C. M., Soltani, N., Robinson, D. E., Hooker, D. C., Miller, R. T., & Sikkema, P. H. (2017). Distribution of glyphosate and cloransulam-methyl resistant Canada fleabane [Conyza canadensis (L.) Cronq.] in Ontario. Can. J. Plant Sci. 98:492-497.

Buhler, D. D. (1997). Effects of tillage and light environment on emergence of 13 annual weeds. Weed Technol. 11:496-501.

Buhler, D. D. & Owen, M. D. K. (1997). Emergence and survival of horseweed (Conyza canadensis). Weed Sci. 45:98–101.

Buhler, D. D., Kohler, K. A., & Thompson, R. L. (2001). Weed seed bank dynamics during a five- year crop rotation. Weed Technol. 15:170-176.

Buhler, D. D. & Hartzler, R. G. (2001). Emergence and persistence of seed of velvetleaf, common waterhemp, woolly cupgrass, and giant foxtail. Weed Sci. 49:230-235.

Burnet, M. W., Loveys, B. R., Holtum, J. A., & Powles, S. B. (1993). Increased detoxification is a mechanism of simazine resistance in Lolium rigidum. Pestic. Biochem. Physiol. 46:207-218.

Burnside, O. C., Wicks, G. A., Warnes, D. D., Somerhalder, B. R., & Weeks, S. A. (1969). Effect of weeds on harvesting efficiency in corn, sorghum, and soybeans. Weed Sci. 17:438-441.

Burnside, O. C., Wilson, R. G., Weisberg, S., & Hubbard, K. G. (1996). Seed longevity of 41 weed species buried 17 years in eastern and western Nebraska. Weed Sci. 44:74-86.

147

Byker, H. P., Soltani, N., Robinson, D. E., Tardif, F. J., Lawton, M. B., & Sikkema, P. H. (2013a). Occurrence of glyphosate and cloransulam resistant Canada fleabane (Conyza canadensis L. Cronq.) in Ontario. Can. J. Plant. Sci. 93:851-855.

Byker, H. P., Soltani, N., Robinson, D. E., Tardif, F. J., Lawton, M. B., & Sikkema, P. H. (2013b). Control of glyphosate-resistant horseweed (Conyza canadensis) with dicamba applied preplant and postemergence in dicamba-resistant soybean. Weed Technol. 27:492–496.

Byker, H. P., Soltani, N., Robinson, D. E., Tardif, F. J., Lawton, M. B., & Sikkema, P. H. (2013c). Glyphosate-resistant Canada fleabane (Conyza canadensis (L). Cronq.): Dose response to glyphosate and control with postemergence herbicides in soybean in Ontario. Can. J. Plant Sci. 93:1187-1193.

Byker, H.P., Soltani, N., Robinson, D.E., Tardif, F.J., Lawton, M.B., & Sikkema, P.H. (2013d). Control of glyphosate-resistant Canada fleabane [Conyza canadensis (L.) Cronq.] with preplant herbicide tankmixes in soybean [Glycine max.(L). Merr.]. Can. J. Plant Sci. 93:659- 667.

Canadian Food Inspection Agency (CFIA) (1996). BIO1996-10L.The Biology of Glycine mac (L.) Merr. https://inspection.gc.ca/plants/plants-with-novel-traits/applicants/directive-94- 08/biology-documents/glycine-max-l-merr-/eng/1330975306785/1330975382668. Accessed: November 26, 2019.

Capinera, J. L. (2005). Relationships between insect pests and weeds: an evolutionary perspective. Weed Sci. 53:892-901.

Carpenter, J. E., & Gianessi, L. P. (2010) Economic impacts of glyphosate-resistant weeds. Nandula VK, ed. Glyphosate Resistance in Crops and Weeds: History, Development, and Management. P. 297–312. Hoboken, NJ, United States.

Carpenter, K., Cottrell, H. J., DeSilva, W. H., Heywood, B. J., Leeds, W. G., Rivett, K. F., & Soundy, M. L. (1964). Chemical and biological properties of two new herbicides-ioxynil and bromoxynil. Weed Res. 4:175-195.

Chandler, K, Shrestha, A, & Swanton, C. J. (2001). Weed seed return as influenced by the critical weed-free period and row spacing of no-till glyphosate-resistant soybean. Can. J. Plant. Sci. 81:877-880.

Cholette, T. B., Soltani, N., Hooker, D. C., Robinson, D. E., & Sikkema, P. H. (2018). Suppression of glyphosate-resistant Canada fleabane (Conyza canadensis) in corn with cover crops seeded after wheat harvest the previous year. Weed Technol. 32:244-250.

Cober, E. R., & Voldeng, H. D. (2012). A retrospective look at short-season soybean cultivar development in Ontario. Can. J. Plant. Sci. 92:1239-1243.

Colby, S. R. (1967). Calculating synergistic and antagonistic responses of herbicide combinations. Weeds. 15:20-22.

148

Comes, R. D., Bruns, V. F., & Kelley, A. D. (1987). Longevity of certain weed and crop seeds in fresh water. Weed Sci. 26:336-344.

Corbett, J. L., Askew, S. D., Thomas, W. E., & Wilcut, J. W. (2004). Weed efficacy evaluations for bromoxynil, glufosinate, glyphosate, pyrithiobac, and sulfosate. Weed Technol. 18:443- 453.

Costea, M. & Tardif, F. J. (2003). Conspectus and notes on the genus Amaranthus in Canada. Rhodora. 105:260-281.

Costea, M., Weaver, S. E., & Tardif, F. J. (2005). The biology of invasive alien plants in Canada. 3. Amaranthus tuberculatus (Moq.) Sauer var. rudis (Sauer) Costea & Tardif. Can. J. Plant. Sci. 85:507-522.

Cousens, R. (1985). A simple model relating yield loss to weed density. Ann Appl Biol. 107:239– 252.

Cronquist, A. (1943). The separation of Erigeron from Conyza. Bull. Torrey Bot. Club. 70:629- 632.

Culpepper A. S., & York, A. C. (1997) Weed management in no-tillage bromoxynil-tolerant cotton (Gossypium hirsutum). Weed Technol 11:335-345.

Dauer, J. T., Mortensen, D. A., & VanGessel, M. J. (2007). Temporal and spatial dynamics of long-distance Conyza canadensis seed dispersal. J. Appl. Ecol. 44:105–114.

Davis, V. M., Gibson, K. D., & Johnson, W. G. (2008). A field survey to determine distribution and frequency of glyphosate-resistant horseweed (Conyza canadensis) in Indiana. Weed Technol. 22:331-338.

Davis V. M., & Johnson, W. G. (2008). Glyphosate-resistant horseweed (Conyza canadensis) emergence, survival, and fecundity, in no-till soybeans. Weed Sci. 56:231-236.

Davis, V. M., Gibson, K. D., Bauman, T. T., Weller, S. C., & Johnson, W. G. (2009a). Influence of weed management practices and crop rotation on glyphosate-resistant horseweed (Conyza canadensis) population dynamics and crop yield-years III and IV. Weed Sci. 57:417-426.

Davis, V. M., Kruger, G. R., Stachler, J. M., Loux, M. M. & Johnson, W. G., (2009b). Growth and seed production of horseweed (Conyza canadensis) populations resistant to glyphosate, ALS inhibiting, and multiple (glyphosate + ALS-inhibiting) herbicides. Weed Sci. 57:494-504.

Davis, V. M., Kruger, G. R., Young, B. G., & Johnson, W. G. (2010). Fall and spring preplant herbicide applications influence spring emergence of glyphosate-resistant horseweed (Conyza canadensis). Weed Technol. 24:11-19.

Derksen, D. A., Thomas, A. G., Lafond, G. P., Loeppky, H. A., & Swanton, C. J. (1994). Impact of agronomic practices on weed communities: fallow within tillage systems. Weed Sci. 42:184-194.

149

Derrick, P. M., Cobb, A. H., & Pallett, K. E. (1988). Ultrastructural effects of the diphenyl ether herbicide acifluorfen and the experimental herbicide M&B 39279. Pestic. Biochem. Physiol. 32:153-163.

Devine, M. D., & Shukla, A. (2000). Altered target sites as a mechanism of herbicide resistance. Crop Prot. 19:881-889.

Devlin, D. L., Gealy, D. R., & Morrow, L. A. (1987). Retention, absorption, and loss of foliage- applied metribuzin. Weed Sci. 775-779.

Dieleman, A., Hamill A. S., Weise S. F., & Swanton, C. J. (1995). Empirical models of pigweed (Amaranthus spp.) interference in soybean (Glycine max). Weed Sci. 3:612-618.

Dieleman A., Hamill A. S., Fox G. C., Swanton, C. J. (1996). Decision rules for postemergence control of pigweed (Amaranthus spp.) in soybean (Glycine max). Weed Sci 44:126-132.

Dill, G. M. (2005). Glyphosate‐resistant crops: history, status and future. Pest Manag. Sci. 61:219- 224.

Dinelli, G., Marotti, I., Bonetti, A., Minelli, M., Catizone, P., & Barnes, J. (2006). Physiological and molecular insight on the mechanisms of resistance to glyphosate in Conyza canadensis (L.) Cronq. Biotypes. Pestic. Biochem. Physiol. 86:30-41.

Ditommaso, A., Watson, A. K., & Hallett, S. G. (1996). Infection by the fungal pathogen Colletotrichum coccodes affects velvetleaf (Abutilon theophrasti)-soybean competition in the field. Weed Sci. 44:924-933.

Dorff, E. (2007). The soybean, agriculture’s jack-of-all-trades, is gaining ground across Canada. Statistics Canada. Available at: http://www.statcan.gc.ca/pub/ 96-325- x/2007000/article/10369-eng.htm. Accessed November 18 2019.

Duke, S. O., Lydon, J., Becerril, J. M., Sherman, T. D., Lehnen, L. P., & Matsumoto, H. (1991). Protoporphyrinogen oxidase-inhibiting herbicides. Weed Sci. 39:465-473.

Duke, S. O. (2005) Taking stock of herbicide-resistant crops ten years after introduction. Pest Manage Sci. 61:211–218.

Duke, S. O. (2012). Why have no new herbicide modes of action appeared in recent years? Pest Manag. Sci. 68:505-512.

Duke, S. O., & Powles, S. B. (2008). Glyphosate: a once‐in‐a‐century herbicide. Pest Manag. Sci. 64:319-325.

Eichholtz, D. A., Gasser, C. S., & Kishore, G. M. (2001). U.S. Patent No. 6,225,114. Washington, DC: U.S. Patent and Trademark Office.

Ellis, J. M., Shaw, D. R., & Barrentine, W. L. (1998). Soybean (Glycine max) seed quality and harvesting efficiency as affected by low weed densities. Weed Technol. 12:166-173.

150

Elmore, D., & Paul, R. N. (1983). Composite List of C4 Weeds. Weed Sci. 31:686–692.

Environmental Protection Agency (EPA). (2003). Health Effects Support Document for Metribuzin. Washington (DC): Environmental Protection Agency; (EPA-733-R-04–001). Acessed March 21, 2020.

Environmental Protection Agency (EPA). (2020). Office of Pesticide Programs. EPA Proposes Registration of New Herbicide to Aid in Resistance Management. https://www.epa.gov/pesticides/epa-proposes-registration-new-herbicide-aid-resistance- management. Accessed January 5, 2021.

Essman, A. I., Loux, M. M., Lindsey, A. J., Dobbels, A. F., & Regnier, E. E. (2020). The effects of integrating a cereal rye cover crop with herbicides on glyphosate-resistant horseweed (Conyza canadensis) in no-till soybean. Weed Sci. 68: 527-533.

Eubank, T. W., Poston, D. H., Nandula, V. K., Koger, C. H., Shaw, D. R., & Reynolds, D. B. (2008). Glyphosate-resistant horseweed (Conyza canadensis) control using glyphosate-, paraquat-, and glufosinate-based herbicide programs. Weed Technol. 22:16-21.

Eue, L. (1972). Sencor, a herbicide of the triazinone group. Pflanzenschutz Nachr. 25:175.

Evans Jr, A. F., O'Brien, S. R., Ma, R., Hager, A. G., Riggins, C. W., Lambert, K. N., & Riechers, D. E. (2017). Biochemical characterization of metabolism‐based atrazine resistance in Amaranthus tuberculatus and identification of an expressed GST associated with resistance. Plant Biotechnol. J. 15:1238-1249.

Evans, C. M. (2016). Characterization of a novel five-way-resistant population of waterhemp (Amaranthus tuberculatus). University of Illinois. P. 30-53. Masters Thesis.

Evans, C. M., Strom, S. A., Riechers, D. E., Davis, A. S., Tranel, P. J., & Hager, A. G. (2019). Characterization of a waterhemp (Amaranthus tuberculatus) population from Illinois resistant to herbicides from five site-of-action groups. Weed Technol. 33:400-410.

Everitt, J. D., & Keeling, J. W. (2007). Weed control and cotton (Gossypium hirsutum) response to preplant applications of dicamba, 2,4-D, and diflufenzopyr plus dicamba. Weed Technol. 21:506-510.

Eyherabide, J. J., & Cendoya, M. G. (2002). Critical periods of weed control in soybean for full field and in-furrow interference. Weed Sci. 50:162-166.

Fawcett, R. S., B. R. Christensen, & D. P. Tierney. (1994). The impact of conservation tillage on pesticide runoff into surface water: a review and analysis. J. Soil Water Cons. 49:126-13.

Fedtke, C. (1991). Deamination of metribuzin in tolerant and susceptible soybean (Glycine max) cultivars. Pestic. Sci. 175-183.

Fehr, W. R. (1989). Oil Crops of the World: Their Breeding and Utilization, McGraw-Hill. P. 283-300. New York, United States.

151

Feng, P. C., Tran, M., Chiu, T., Sammons, R. D., Heck, G. R., & Jacob, C. A. (2004). Investigations into glyphosate-resistant horseweed (Conyza canadensis): retention, uptake, translocation, and metabolism. Weed Sci. 52:498-505.

Foes, M. J., Tranel, P. J., Wax, L. M., & Stoller, E. W. (1998). A biotype of common waterhemp (Amaranthus rudis) resistant to triazine and ALS herbicides. Weed Sci. 46:514-520.

Ford, L., Soltani, N., Robinson, D. E., Nurse, R. E., McFadden, A., & Sikkema, P. H. (2014). Canada fleabane (Conyza canadensis) control with preplant applied residual herbicides followed by 2,4-D choline/glyphosate DMA applied postemergence in corn. Can. J. Plant. Sci. 9: 1231-1237.

Frank, R., Clegg, B. S., Sherman, C., & Chapman, N. D. (1990). Triazine and chloroacetamide herbicides in Sydenham River water and municipal drinking water, Dresden, Ontario, Canada, 1981–1987. Arch. Environ. Contain. Toxicol. 19: 319-324.

Frankton, C., & Mulligan, G. A. (1987). Weeds of Canada (No. 948). NC Press Limited. p. 156- 157. Ottawa, ON, Canada.

Franz, J. E., Mao, M. K., & Sikorski, J. A. (1997). Glyphosate: a unique global herbicide. American Chemical Society. Washington, DC, USA.

Friesen, G. H. & Wall, D. A. (1984). Response of potato (Solanum tuberosum) cultivars to metribuzin. Weed Sci. 32:442–444.

Frisvold, G. B., T. M. Hurley, & P. D. Mitchell. (2009). Adoption of best management practices to control weed resistance by cotton, corn, and soybean growers. In The 2009 Annual Meeting, July 26–28, 2009. Milwaukee, WI: Agricultural and Applied Economics Association. Agbioforum.12: 370–381.

Gaines, T. A., Shaner, D. L., Ward, S. M., Leach, J. E., Preston, C., & Westra, P. (2011). Mechanism of resistance of evolved glyphosate-resistant Palmer amaranth (Amaranthus palmeri). J. Agric. Food Chem. 59:5886-5889.

Gaines, T. A., Zhang, W., Wang, D., Bukun, B., Chisholm, S. T., Shaner, D. L. & Grey, T. L. (2010). Gene amplification confers glyphosate resistance in Amaranthus palmeri. Proceedings of the National Academy of Sciences.107:1029-1034.

Ganie Z. A., & Jhala, A. J. (2017) Glyphosate-resistant common ragweed (Ambrosia artemisiifolia) in Nebraska: confirmation and response to postemergence corn and soybean herbicides. Weed Technol. 31:225-237.

Gardner, S. N., Gressel, J., & Mangel, M. (1998). A revolving dose strategy to delay the evolution of both quantitative vs major monogene resistances to pesticides and drugs. Int. J. Pest Manag. 44:161-180.

152

Gawronski, S. W., Haderlie, L. C., Callihan, R. H., & Gawronska, H. (1986). Mechanism of metribuzin tolerance: herbicide metabolism as a basis for tolerance in potatoes. Weed Res. 26:307-314.

Ge, X., d'Avignon, D. A., Ackerman, J. J., & Sammons, R. D. (2010). Rapid vacuolar sequestration: the horseweed glyphosate resistance mechanism. Pest Manag. Sci. 66:345- 348.

Ge, X., d'Avignon, D. A., Ackerman, J. J., Duncan, B., Spaur, M. B., & Sammons, R. D. (2011). Glyphosate‐resistant horseweed made sensitive to glyphosate: low‐temperature suppression of glyphosate vacuolar sequestration revealed by 31P NMR. Pest Manag. Sci. 67:1215-1221.

Ghersa, C. M., Benech-Arnold, R. L., Satorre, E. H., & Martınez-Ghersa, M. A. (2000). Advances in weed management strategies. Field Crops Res. 67:95-104.

Giacomini, D. A., Umphres, A. M., Nie, H., Mueller, T. C., Steckel, L. E., Young, B. G, & Tranel, P. J. (2017). Two new PPX2 mutations associated with resistance to PPO‐inhibiting herbicides in Amaranthus palmeri. Pest Manag. Sci.73:1559-1563.

Givens, W. A., Shaw, D. R., Johnson, W. G., Weller, S. C., Young, B. G., Wilson, R. G., & Jordan, D. (2009). A grower survey of herbicide use patterns in glyphosate-resistant cropping systems. Weed Technol. 23:156-161.

Glotfelty, D. E., Taylor, A. W., Isensee, A. R., Jersey, J., & Glenn, S. (1984). Atrazine and simazine movement to Wye River estuary. Journal of Environ. Qual.13:115-121.

Goldsmith, P. (2008). Economics of soybean production, marketing, and utilization. In L. A. Johnson, P. J. White, & R. Galloway (Eds.), Soybeans: Chemistry, production, processing, and utilization. P. 117-150. Urbana, Illinois, United States.

González-Torralva, F., Rojano-Delgado, A. M., de Castro, M. D. L., Mülleder, N., & De Prado, R. (2012). Two non-target mechanisms are involved in glyphosate-resistant horseweed (Conyza canadensis L. Cronq.) biotypes. J. Plant Physiol. 169:1673-1679.

Green, J. M., Obrigawitch, T. T., Long, J. D., & Hutchison, J. M. (1988). Metribuzin and chlorimuron mixtures for preemergence broadleaf weed control in soybeans, Glycine max. Weed Technol. 2:355-363.

Green, J. M., & Owen, M. D. (2011). Herbicide-resistant crops: utilities and limitations for herbicide-resistant weed management. J. Agric. Food Chem. 59:5819-5829.

Green-Tracewicz, E., Page, E. R., & Swanton, C. J. (2012). Light quality and the critical period for weed control in soybean. Weed Sci. 60:86-91.

Gressel, J., & Segel, L. A. (1990). Modelling the effectiveness of herbicide rotations and mixtures as strategies to delay or preclude resistance. Weed Technol. 4:186-198.

153

Griffin J. L., Boudreaux J. M., & Miller, D. K. (2010) Herbicides as harvest aids. Weed Sci. 58:355–358.

Grossmann, K., Niggeweg, R., Christiansen, N., Looser, R., & Ehrhardt, T. (2010). The herbicide saflufenacil (Kixor™) is a new inhibitor of protoporphyrinogen IX oxidase activity. Weed Sci. 58:1-9.

Guddewar, M. B., & Dauterman, W. C. (1979). Purification and properties of a glutathione-S- transferase from corn which conjugates S-triazine herbicides. Phytochemistry. 18:735-740.

Gulden, R. H., Sikkema, P. H., Hamill, A. S., Tardif, F., & Swanton, C. J. (2009). Conventional vs. glyphosate-resistant cropping systems in Ontario: weed control, diversity, and yield. Weed Sci. 57:665-672.

Guo, P., & Al-Khatib, K. (2003) Temperature effects on germination and growth of redroot pigweed (Amaranthus retroflexus), Palmer amaranth (A. palmeri), and common waterhemp (A. rudis). Weed Sci. 51:869–875.

Hager, A. G., Wax, L. M., Stoller, E. W., & Bollero, G. A. (2002). Common waterhemp (Amaranthus rudis) interference in soybean. Weed Sci. 50:607-610.

Hager, A. G., Wax, L. M., Bollero, G. A., & Stoller, E. W. (2003). Influence of Diphenylether Herbicide application rate and timing on common waterhemp (Amaranthus rudis) control in soybean (Glycine max). Weed Technol.17:14-20.

Haizel, K. A., & Harper J. L. (1973). The effects of density and the timing of removal on interference between barley (Hordeum vulgare L.), white mustard (Sinapis alba L.), and wild oats (Avena fatua L.). J. Appl. Ecol. 10:23-31.

Halford, C., Hamill, A. S., Zhang, J., & Doucet, C. (2001). Critical period of weed control in no- till soybean (Glycine max) and corn (Zea mays). Weed Technol. 15:737-744.

Hall, L., Beckie, H., & Wolf, T. M. (1999). Agriculture, Food and Rural Development. How herbicides work. How herbicides work: Biology to application. P. 1-12. Edmonton, AB, Canada

Hamill, A. S., & Zhang, J. (1995) Herbicide reduction in metribuzin-based weed control programs in corn. Can. J. Plant Sci. 75:927-933.

Harder, D. B., Sprague, C. L., & Renner, K. A. (2007). Effect of soybean row width and population on weeds, crop yield, and economic return. Weed Technol. 21:744-752.

Haring, B., & Hanson, B. (2020). Glufosinate and Tiafenacil Burndown Trial. University of California, Davis. https://ucanr.edu/repository/fileaccess.cfm?article=178644&p=NFLFGK. Accessed: January 4, 2020.

Harris, P., & Cranston, R. (1979). An economic evaluation of control methods for diffuse and spotted knapweed in western Canada. Can. J. Plant. Sci. 59:375-382.

154

Hartman, G. L., West, E. D., & Herman, T. K. (2011). Crops that feed the World 2. Soybean- worldwide production, use, and constraints caused by pathogens and pests. Food Secur. 3:5- 17.

Hartwig, N. L., & Ammon, H. U. (2002). Cover crops and living mulches. Weed Sci. 50:688-699.

Hartzler, R. G., D. D. Buhler, & D. E. Stoltenberg. (1999). Emergence characteristics of four annual weed species. Weed Sci. 47:578-584.

Hartzler, R. G., Battles, B. A., & Nordby, D. (2004). Effect of common waterhemp (Amaranthus rudis) emergence date on growth and fecundity in soybean. Weed Sci. 52:242-245.

Hausman, N. E., Tranel, P. J., Riechers, D. E., Maxwell, D. J., Gonzini, L. C., & Hager, A. G. (2013). Responses of an HPPD inhibitor-resistant waterhemp (Amaranthus tuberculatus) population to soil-residual herbicides. Weed Technol. 27:704-711.

Hausman, N. E., Tranel, P. J., Riechers, D. E., & Hager, A. G. (2016). Responses of a waterhemp (Amaranthus tuberculatus) population resistant to HPPD-inhibiting herbicides to foliar- applied herbicides. Weed Technol. 30: 106-115.

Health Canada. (2018). Tiafenacil Herbicide. Ottawa, ON, Canada: Pest Management Regulatory Agency Consumer Product Safety. Application number: 2018-1276, 2018-1277, and 2018- 1301. https://pr-rp.hc-sc.gc.ca/pi-ip/rba-epa-eng.php?p_actv=TIAF%45NACIL. Acessed: January 5, 2021.

Heap, I. (2020). The International Herbicide-Resistant Weed Database. Online. Available: www.weedscience.org.

Heap, I. (2021). The International Herbicide-Resistant Weed Database. Online. Available: www.weedscience.org.

Hedges, B. K., Soltani, N., Robinson, D. E., Hooker, D. C., & Sikkema, P. H. (2018a). Control of glyphosate-resistant Canada fleabane in Ontario with multiple effective modes-of-action in glyphosate/dicamba-resistant soybean. Can. J. Plant. Sci. 99:78-83.

Hedges, B. K., Soltani, N., Hooker, D. C., Robinson, D. E., & Sikkema, P. H. (2018b). Control of glyphosate-resistant waterhemp with preemergence herbicides in glyphosate-and dicamba- resistant soybean. Can. J. Plant Sci. 99:34-39.

Hess, F. D. (2000). Light-dependent herbicides: an overview. Weed Sci. 48:160-170.

Hilton, H. W. (1957). Herbicide tolerant strains of weeds. In Hawaiian Sugar Planters Association Annual Report. P. 69-72. Honolulu, HI: Hawaiian Sugar Planters Association.

Holt, J. S., Powles, S. B., & Holtum, J. A. (1993). Mechanisms and agronomic aspects of herbicide resistance. Annu. Rev. Plant Biol.44: 203-229.

155

Horak, M. J., Peterson, D. E., Chessman, D. J., & Wax, L. M. (1994). Pigweed identification: A pictorial guide to the common pigweeds of the Great Plains. P. 8.

Horak, M. J. & Loughin, T. M. (2000). Growth analysis of four Amaranthus species. Weed Sci. 48:347-355.

Hough-Goldstein, J., Schiff, M., Lake, E., & Butterworth, B. (2008). Impact of the biological control agent Rhinoncomimus latipes (Coleoptera: Curculionidae) on mile-a-minute weed, Persicaria perfoliata, in field cages. Biol. Control. 46:417-423.

Hough-Goldstein, J., Lake, E., & Reardon, R. (2012). Status of an ongoing biological control program for the invasive vine, Persicaria perfoliata in eastern North America. Biol. Control. 57:181-189.

Howe, O. W., & Oliver, L. R. (1987). Influence of soybean (Glycine max) row spacing on pitted morningglory (Ipomoea lacunosa) interference. Weed Sci. 35:185-193.

Howe, A. R., Gasser, C. S., Brown, S. M., Padgette, S. R., Hart, J., Parker, G. B., & Armstrong, C. L. (2002). Glyphosate as a selective agent for the production of fertile transgenic maize (Zea mays L.) plants. Mol. Breed. 10:153-164.

Ikley, J.T. (2012). The utility of saflufenacil on glyphosate-resistant horseweed and its effect on select soybean varieties. University of Maryland, College Park, Maryland. P. 78. Master’s Thesis.

Ivany, J. A. (2005). Response of three potato (Solanum tuberosum) cultivars to pyraflufen-ethyl used as a desiccant in Canada. Crop Prot. 24:836-841.

Jasieniuk, M., Brûlé-Babel, A. L., & Morrison, I. N. (1996). The evolution and genetics of herbicide resistance in weeds. Weed Sci. 44:176-193.

Jha, P., & Norsworthy, J. K. (2009). Soybean canopy and tillage effects on emergence of Palmer amaranth (Amaranthus palmeri) from a natural seed bank. Weed Sci. 57:644-651.

Jordan D. L., Smith M. C., McClelland M. R., & Frans, R. E. (1993). Weed control with bromoxynil applied alone and with graminicides. Weed Technol. 7:835-839.

Jugulam, M., & Shyam, C. (2019). Non-Target-Site Resistance to Herbicides: Recent Developments. Plants. 8:417.

Kahrizi, D., Salmanian, A. H., Afshari, A., Moieni, A., & Mousavi, A. (2007). Simultaneous substitution of Gly96 to Ala and Ala183 to Thr in 5-enolpyruvylshikimate-3-phosphate synthase gene of E. coli (k12) and transformation of rapeseed (Brassica napus L.) in order to make tolerance to glyphosate. Plant Cell Reports. 26:95-104.

Kapusta, G. (1979). Seedbed tillage and herbicide influence on soybean (Glycine max) weed control and yield. Weed Sci. 27:520-526.

156

Kells, J. J. Dalley, C. D., & Renner, K. A. (2004). Effect of glyphosate application timing and row spacing on corn (Zea mays) and soybean (Glycine max) yields. Weed Technol. 18:165-176.

Klingaman, T. E., & Oliver, L. R. (1994). Palmer amaranth (Amaranthus palmeri) interference in soybeans (Glycine max). Weed Sci. 42:523-527.

Knake, E. L., & Slife, F. W. (1962). Competition of Setaria faberii with corn and soybeans. Weeds. 10:26-29.

Knezevic S. Z., Sikkema P. H., Tardif F., Hamill A. S., Chandler K., & Swanton, C. J. (1998). Biologically effective dose and selectivity of RPA 201772 (isoxaflutole) for preemergence weed control in corn. Weed Technol. 12:670–676.

Knezevic, S. Z., Evans, S. P., Blankenship, E. E., Van Acker, R. C., & Lindquist, J. L. (2002). Critical period for weed control: the concept and data analysis. Weed Sci. 50:773-786.

Knezevic, S. Z., Streibig, J. C., Ritz, C. (2007). Utilizing R software package for dose-response studies: the concept and data analysis. Weed Technol. 21:840-848.

Knezevic, S. Z., Datta, A., Scott, J., & Charvat, L. D. (2009). Interactions between saflufenacil and glyphosate on selected broadleaf weeds. Crop Manage. 8:1-15.

Koger, C. H., D. H. Poston, R. M. Hayes, & R. F. Montgomery. (2004). Glyphosate-resistant horseweed (Conyza canadensis) in Mississippi. Weed Technol. 18:820–825.

Krausz, R. F., Young, B. G., Kapusta, G., & Matthews, J. L. (2001). Influence of weed competition and herbicides on glyphosate-resistant soybean (Glycine max). Weed Technol. 15:530-534.

Kruger, G. R., Davis, V. M., Weller, S. C., & Johnson, W. G. (2010a). Control of horseweed (Conyza canadensis) with growth regulator herbicides. Weed Technol. 24:425-429.

Kruger, G. R., Davis, V. M., Weller, S. C., & Johnson, W. G. (2010b). Growth and seed production of horseweed (Conyza canadensis) populations after exposure to postemergence 2,4-D. Weed Sci. 58:413-419.

Ladlie, J. S., Meggitt, W. F., & Penner, D. (1976). Effect of soil pH on microbial degradation, adsorption, and mobility of metribuzin. Weed Sci. 24:477-481.

Lafond, G., Brandt, S., McAndrew, D., Stobbe, E., & Tessier, S. (1990). Tillage systems and crop production in G. P. Lafond and D. B. Fowler, eds. Crop Management for Conservation. P. 155-201. Univ. Saskatchewan, Saskatoon, SK, Canada.

Leblanc, M. & Cloutier, D. C. (1996). Effect of false seedbed technique on annual weed emergence in maize. Proceedings 1996 Xth International Symposium on the Biology of Weeds, Dyon, France. P. 29–34. Association Nationale pour la Protection des Plantes.

157

Lebrun, M., Sailland, A., Freyssinet, G., & DeGryse, E. (2003). U.S. Patent No. 6,566,587. Washington, DC: U.S. Patent and Trademark Office.

Leck, M. A., & Leck, C. F. (1998). A ten-year seed bank study of old field succession in central New Jersey. J. Torrey Bot. Soc. 125:11-32.

Légère, A., & Schreiber, M. M. (1989). Competition and canopy architecture as affected by soybean (Glycine max) row width and density of redroot pigweed (Amaranthus retroflexus). Weed Sci. 37:84-92.

Legleiter, T. R., & Bradley, K. W. (2008). Glyphosate and multiple herbicide resistance in common waterhemp (Amaranthus rudis) populations from Missouri. Weed Sci. 56:582-587.

Legleiter, T. R., Bradley, K. W., & Massey, R. E. (2009). Glyphosate-resistant waterhemp (Amaranthus rudis) control and economic returns with herbicide programs in soybean. Weed Technol. 23:54-61.

Lejeune, K. D., & Seastedt, T. R. (2001). Centaurea species: the forb that won the west. Conserv. Biol. 15:1568-1574.

Leon, R. G., & Owen, M. D. (2003). Regulation of weed seed dormancy through light and temperature interactions. Weed Sci. 51:752-758.

Leon, R. G., Knapp, A. D., & Owen, M. D. (2004). Effect of temperature on the germination of common waterhemp (Amaranthus tuberculatus), giant foxtail (Setaria faberi), and velvetleaf (Abutilon theophrasti). Weed Sci. 52:67-73.

Leon, R. G., & Owen, M. D. (2006). Tillage systems and seed dormancy effects on common waterhemp (Amaranthus tuberculatus) seedling emergence. Weed Sci. 54:1037-1044.

Liu, J., Davis, A. S., & Tranel, P. J. (2012). Pollen biology and dispersal dynamics in waterhemp (Amaranthus tuberculatus). Weed Sci. 60:416-422.

Livingston, M., Fernandez-Cornejo, J., & Frisvold, G. B. (2016). Economic returns to herbicide resistance management in the short and long run: the role of neighbor effects. Weed Sci. 64:595-608.

Lorraine-Colwill, D. F., Powles, S. B., Hawkes, T. R., Hollinshead, P., Warner, S. A. J., & Preston, C. (2002). Investigations into the mechanism of glyphosate resistance in Lolium rigidum. Pestic Biochem Physiol. 74:62-72.

Ludwig, J. W. (1973). The use of a low dose of atrazine alone and in mixtures with other herbicides in the maize crop. Weed Res. 13:12-18.

Ma, R., Kaundun, S. S., Tranel, P. J., Riggins, C. W., McGinness, D. L., Hager, A. G., & Riechers, D. E. (2013). Distinct detoxification mechanisms confer resistance to mesotrione and atrazine in a population of waterhemp. Plant Physiol. 163:363-377.

158

Ma, H., Lu, H., Han, H., Yu, Q., & Powles, S. (2020). Metribuzin resistance via enhanced metabolism in a multiple herbicide resistant Lolium rigidum population. Pest Manag. Sci 76:3785-3791.

Mabuchi, T., Miura, Y., & Ohtsuka, T. (2002). Herbicidal activity and characteristics of pyraflufen-ethyl for controlling broad-leaved weeds in cereals. J. Pest. Sci. 27:39-46.

Mahoney, K. J., McNaughton, K. E., & Sikkema, P. H. (2016a). Control of glyphosate-resistant horseweed in winter wheat with pyrasulfotole premixed with bromoxynil. Weed Technol. 30: 291-296.

Mahoney, K. J., Shropshire, C., & Sikkema, P. H. (2016b). Post-emergence herbicides for control of glyphosate-resistant Canada fleabane in corn. Can. J. Plant. Sci. 97:206-209.

Main, C. L., Steckel, L. E., Hayes, R. M., & Mueller, T. C. (2006). Biotic and abiotic factors influence horseweed emergence. Weed Sci. 54:1101-1105.

Mallory-Smith, C. (1998). Bromoxynil-resistant common groundsel (Senecio vulgaris). Weed Technol. 12:322-324.

Masabni, J. G., & Zandstra, B. H. (1999). A serine-to-threonine mutation in linuron-resistant Portulaca oleracea. Weed Sci. 47:393-400.

McBride E, K. E., Kenny, J. W., & Stalker, D. M. (1986). Metabolism of the herbicide bromoxynil by Klebsiella pneumoniae subsp. ozaenae. Appl. Environ. Microbiol. 52:325- 330.

McCloskey, W.B., & Holt, J. S. (1990). Triazine resistance in Senecio vulgaris parental and nearly isonuclear backcrossed biotypes is correlated with reduced productivity. Plant Physiol. 92:954–962.

Mellendorf, T. G., Young, J. M., Matthews, J. L., & Young, B. G. (2013). Influence of plant height and glyphosate on saflufenacil efficacy on glyphosate-resistant horseweed (Conyza canadensis). Weed Technol. 27:463-467.

Mellendorf, T. G., Young, J. M., Matthews, J. L., & Young, B. G. (2015). Influence of application variables on the foliar efficacy of saflufenacil on horseweed (Conyza canadensis). Weed Sci. 63:578-586.

Mengistu, L. W., Mueller‐Warrant, G. W., Liston, A., & Barker, R. E. (2000). psbA mutation (valine219 to isoleucine) in Poa annua resistant to metribuzin and diuron. Pest Manag. Sci. 56:209-217.

Mester, T. C., & Buhler, D. D. (1990). Effect of planting depth on velvetleaf (Abutilon theophrasti) seedling development and response to cyanazine. Weed Sci. 38:34-38.

Metzger, B. A., Soltani, N., Raeder, A. J., Hooker, D. C., Robinson, D. E., & Sikkema, P. H. (2019). Multiple herbicide-resistant horseweed (Conyza canadensis) dose response to

159

tolpyralate and tolpyralate plus atrazine and comparison to industry standard herbicides in corn. Weed Technol. 33:366-373.

Miller, R.T., Soltani, N., Robinson, D.E., Kraus, T.E., & Sikkema, P.H. (2012). Soybean (Glycine max) cultivar tolerance to saflufenacil. Can. J. Plant Sci. 92:1319-1328.

Miura, Y., Mabuchi, T., Higashimura, M., & Amanuma, T. (2003). Development of a new herbicide, pyraflufen-ethyl. J. Pest. Sci. 28:219–220.

Mohler, C. L. (1993). A model of the effects of tillage on emergence of weed seedlings. Ecol. Appl. 3:53-73.

Moolani, M. K., Knake, E. L., & Slife, F. W. (1964). Competition of smooth pigweed with corn and soybeans. Weeds. 12:126-128.

Moore, M. J., Gillespie, T. J., & Swanton, C. J. (1994). Effect of cover crop mulches on weed emergence, weed biomass, and soybean (Glycine max) development. Weed Technol. 8:512- 518.

Moseley, C. M., & Hagood, E. S. (1990). Horseweed (Conyza canadensis) control in full-season no-till soybeans (Glycine max). Weed Technol. 4:814-818.

Mueller, T. C., Massey, J. H., Hayes, R. M., Main, C. L., & Stewart, C. N. (2003). Shikimate accumulates in both glyphosate-sensitive and glyphosate-resistant horseweed (Conyza canadensis L. Cronq.). J. Agric. Food Chem. 51: 680-684.

Mueller, T. C., Mitchell, P. D., Young, B. G., & Culpepper, A. S. (2005). Proactive versus reactive management of glyphosate-resistant or-tolerant weeds. Weed Technol. 19:924-933.

Mulugeta, D., & Boerboom, C. M. (2000). Critical time of weed removal in glyphosate-resistant Glycine max. Weed Sci. 48:35-42.

Murata, S., Yamashita, A., Kimura, Y., Motoba, K., Mabuchi, T., & Miura, Y. (2002). Mechanisms of selective action of a protoporphyrinogen IX oxidase-inhibiting herbicide pyraflufen-ethyl between wheat (Triticum aestivum) and cleavers (Galium aparine). Pest Manag. Sci. 27:47-52.

Murphy, B. P., & Tranel, P. J. (2019). Target-site mutations conferring herbicide resistance. Plants. 8:382.

Murphy, S. D., Clements, D. R., Belaoussoff, S., Kevan, P. G., & Swanton, C. J. (2006). Promotion of weed species diversity and reduction of weed seedbanks with conservation tillage and crop rotation. Weed Sci. 54:69-77.

Myers, J. H., Jackson, C., Quinn, H., White, S. R., & Cory, J. S. (2009). Successful biological control of diffuse knapweed, Centaurea diffusa, in British Columbia, Canada. Biol. Control. 50:66-72.

160

Nandula, V. K., Eubank, T. W., Poston, D. H., Koger, C. H., & Reddy, K. N. (2006). Factors affecting germination of horseweed (Conyza canadensis). Weed Sci. 54:898-902.

Nandula, V. K., Reddy, K. N., Poston, D. H., Rimando, A. M., & Duke, S. O. (2008). Glyphosate tolerance mechanism in Italian ryegrass (Lolium multiflorum) from Mississippi. Weed Sci. 56:344-349.

Nave, W. R., & Wax, L. M. (1971). Effect of weeds on soybean yield and harvesting efficiency. Weed Sci. 19:533-535.

Nieto, H. J., Brondo, M. A., & Gonzalez, J. T. (1968). Critical Periods of the Crop Growth Cycle for Competition from Weeds, Pest Articles & News Summaries. Section C. Weed Control. 14:159-166.

Norsworthy, J. K., Ward, S. M., Shaw, D. R., Llewellyn, R. S., Nichols, R. L., Webster, T. M., & Witt, W. W. (2012). Reducing the risks of herbicide resistance: best management practices and recommendations. Weed Sci. 60:31-62.

O'Brien, S. R., Davis, A. S., & Riechers, D. E. (2018). Quantifying resistance to isoxaflutole and mesotrione and investigating their interactions with metribuzin POST in waterhemp (Amaranthus tuberculatus). Weed Sci. 66:586-594.

Oerke, E. C., & Dehne, H. W. (2004). Safeguarding production—losses in major crops and the role of crop protection. Crop Prot. 23:275-285.

OMAFRA (2019). Publication 75: Guide to Weed Control 2016-2017, Toronto, Ontario. Queens Printer.

Owen, M. D. (2008). Weed species shifts in glyphosate‐resistant crops. Pest Manag Sci. 64:377- 387.

Owen, L. N., L. E. Steckel, C. H. Koger, C. L. Main, & T. C. Mueller. (2009). Evaluation of spring and fall burndown application timings on control of glyphosate-resistant horseweed (Conyza canadensis) in no-till cotton. Weed Technol. 23:335-33

Pagano, M. C., & Miransari, M. (2016). The importance of soybean production worldwide. In Abiotic and Biotic Stresses in Soybean Production. Academic Press. P.1-26. Cambridge, Massachusetts.

Page, E. R., Grainger, C. M., Laforest, M., Nurse, R. E., Rajcan, I., Bae, J., & Tardif, F. J. (2018). Target and non-target site mechanisms confer resistance to glyphosate in Canadian accessions of Conyza canadensis. Weed Sci. 66:234-245.

Palma-Bautista, C., Torra, J., Garcia, M. J., Bracamonte, E., Rojano-Delgado, A. M., Alcántara-de la Cruz, R., & De Prado, R. (2019). Reduced absorption and impaired translocation endows glyphosate resistance in Amaranthus palmeri harvested in glyphosate-resistant soybean from Argentina. J. Agric. Food Chem. 67:1052-1060.

161

Park, J., Ahn, Y. O., Nam, J. W., Hong, M. K., Song, N., Kim, T., & Sung, S. K. (2018). Biochemical and physiological mode of action of tiafenacil, a new protoporphyrinogen IX oxidase-inhibiting herbicide. Pestic Biochem. Physiol. 152:38-44.

Park, K. W., & Mallory‐Smith, C. A. (2006). psbA mutation (Asn 266 to Thr) in Senecio vulgaris L. confers resistance to several PS II‐inhibiting herbicides. Pest Manag. Sci. 62:880-885.

Patzoldt, W. L., Tranel, P. J., & Hager, A. G. (2002). Variable herbicide responses among Illinois waterhemp (Amaranthus rudis and A. tuberculatus) populations. Crop Prot. 21: 707-712.

Patzoldt, W. L., Dixon, B. S., & Tranel, P. J. (2003). Triazine resistance in Amaranthus tuberculatus (Moq) Sauer that is not site‐of‐action mediated. Pest Manag. Sci. 59: 1134- 1142.

Pest Management Regulatory Agency (PMRA). (2014). Evaluation Report: Pyraflufen-ethyl. Accessed at https://www.canada.ca/en/health-canada/services/consumer-product- safety/reports-publications/pesticides-pest-management/decisions-updates/evaluation- report/2014/pyraflufen-ethyl-erc2014-03.html. Accessed on May 26, 2020.

Pittman, K. B., Barney, J. N., & Flessner, M. L. (2019). Horseweed (Conyza canadensis) suppression from cover crop mixtures and fall-applied residual herbicides. Weed Technol. 33: 303-311.

Powles, S. B., Lorraine-Colwill, D. F., Dellow, J. J., & Preston, C. (1998). Evolved resistance to glyphosate in rigid ryegrass (Lolium rigidum) in Australia. Weed Sci. 46:604-607.

Powles, S. B., & Preston, C. (2006). Evolved glyphosate resistance in plants: biochemical and genetic basis of resistance. Weed Technol. 20:282-289.

Powles, S. B. (2008). Evolution in action: glyphosate-resistant weeds threaten world crops. Outlooks Pest Manag. 19:256-259.

Powles, S. B., & Yu, Q. (2010). Evolution in action: plants resistant to herbicides. Annual Rev. Plant Biol. 61:317-347.

Pratt, D. B., & Clark, L. G. (2001). Amaranthus rudis and A. tuberculatus, one species or two? J. Torrey Bot. Soc. 128:282-296.

Preston, C. (2004). Herbicide resistance in weeds endowed by enhanced detoxification: complications for management. Weed Sci. 52:448-453.

Preston, C., & Wakelin, A. M. (2008). Resistance to glyphosate from altered herbicide translocation patterns. Pest Manag. Sci. 64:372-376.

Preston, C., Wakelin, A. M., Dolman, F. C., Bostamam, Y., & Boutsalis, P. (2009). A decade of glyphosate-resistant Lolium around the world: mechanisms, genes, fitness, and agronomic management. Weed Sci. 57:435-441.

162

Przepiorkowski, T., & Gorski, S. F. (1994). Influence of rye (Secale cereale) plant residues on germination and growth of three triazine-resistant and susceptible weeds. Weed Technol. 8:744-747.

Quinn, J., Soltani, N., Ashigh, J., Hooker, D. C., Robinson, D. E., & Sikkema, P. H. (2020). Halauxifen-methyl controls glyphosate-resistant horseweed (Conyza canadensis) but not giant ragweed (Ambrosia trifida) in winter wheat. Weed Technol. 34: 607-612.

Reddy, K. N., Rimando, A. M., & Duke, S. O. (2004). Aminomethylphosphonic acid, a metabolite of glyphosate, causes injury in glyphosate-treated, glyphosate-resistant soybean. J. Agric. Food Chem. 52:5139-5143.

Refsell, D. E., & Hartzler, R. G. (2009). Effect of tillage on common waterhemp (Amaranthus rudis) emergence and vertical distribution of seed in the soil. Weed Technol. 23:129-133.

Regehr D.L., & Bazzaz, F.A. (1979). The population dynamics of Erigeron canadensis, a successional winter annual. J Ecol. 67:923–933

Rodríguez-Navarro, D. N., Oliver, I. M., Contreras, M. A., & Ruiz-Sainz, J. E. (2011). Soybean interactions with soil microbes, agronomical and molecular aspects. Agron. Sustain. Dev. 31:173-190.

Rousonelos, S. L., Lee, R. M., Moreira, M. S., VanGessel, M. J., & Tranel, P. J. (2012). Characterization of a common ragweed (Ambrosia artemisiifolia) population resistant to ALS-and PPO-inhibiting herbicides. Weed Sci. 60: 335-344.

Ryan, G. F. (1970). Resistance of common groundsel to simazine and atrazine. Weed Sci. 18:614– 616.

Sammons, R. D., & Gaines, T. A. (2014). Glyphosate resistance: state of knowledge. Pest Manag. Sci. 70:1367-1377.

Sarangi, D., & Jhala, A. J. (2017). Biologically effective rates of a new premix (atrazine, bicyclopyrone, mesotrione, and S-metolachlor) for preemergence or postemergence control of common waterhemp [Amaranthus tuberculatus (Moq.) Sauer var. rudis] in corn. Can. J. Plant Sci. 97:1075-1089.

Sarangi, D., Tyre A. J., Patterson, E. L., Gaines, T. A., Irmak, S., Knezevic, S. Z., Lindquist, J. L., & Jhala, A. J. (2017). Pollen-mediated gene flow from glyphosate-resistant common waterhemp (Amaranthus rudis Sauer): consequences for the dispersal of resistance genes. Sci Rep. 7:1-16.

Sauer, J. (1955). Revision of the dioecious amaranths. Madroño. 13: 5-46.

Sauer, J. (1957). Recent migration and evolution of the dioecious amaranths. Evolution. 11-31.

Sauer, J. D. (1972). The dioecious amaranths: a new species name and major range extensions. Madroño. 21:426-434.

163

Savage, K. E. (1976). Adsorption and mobility of metribuzin in soil. Weed Sci. 24:525-528.

Schroeder, J., Barrett, M., Shaw, D. R., Asmus, A. B., Coble, H., Ervin, D., & Culpepper, A. S. (2018). Managing wicked herbicide-resistance: Lessons from the field. Weed Technol. 32: 475-488.

Schryver, M. G., Soltani, N., Hooker, D. C., Robinson, D. E., Tranel, P. J., & Sikkema, P. H. (2017a). Glyphosate-resistant waterhemp (Amaranthus tuberculatus var. rudis) in Ontario, Canada. Can. J. Plant Sci. 97: 1057-1067.

Schryver, M. G., Soltani, N., Hooker, D. C., Robinson, D. E., Tranel, P. J., & Sikkema, P. H. (2017b). Control of glyphosate-resistant common waterhemp (Amaranthus tuberculatus var. rudis) in soybean in Ontario. Weed Technol. 31:811-821.

Schryver, M. G., Soltani, N., Hooker, D. C., Robinson, D. E., Tranel, P. J. & Sikkema, P. H. (2017c). The Distribution and Control of Glyphosate-Resistant Waterhemp (Amaranthus tuberculatus var. rudis) in Soybean (Glycine max) in Ontario. University of Guelph, Guelph, ON. P.59-83. Masters Thesis.

Scroggs, D. M., Miller, D. K., Vidrine, P. R., & Downer, R. G. (2006). Evaluation of weed control and crop tolerance with co-application of glyphosate and pyraflufen-ethyl in glyphosate- resistant soybean (Glycine max). Weed Technol. 20:1035-1039.

Sellers, B. A., Smeda, R. J., Johnson, W. G. Kendig, J. A., Ellersieck, & M. R. (2003). Comparative Growth of Six Amaranthus Species in Missouri. Weed Sci. 51:329–333.

Shaner, D. L. (1995). Herbicide resistance: where are we? How did we get here? Where are we going? Weed Technol. 9:850-856.

Shaner, D. L. (2009). Role of translocation as a mechanism of resistance to glyphosate. Weed Sci. 57:118–123.

Shaner, D. L. (2014). Herbicide Handbook of the Weed Science Society of America. Champaign, IL: Weed Science Society of America.

Sharom, M. S., & Stephenson, G. R. (1976). Behavior and fate of metribuzin in eight Ontario soils. Weed Sci. 24:153-160.

Shaw, D. R., Barrett, M., Schroeder, J., Asmus, A. B., Ervin, D., Jussaume, R. A., & Coble, H. (2018). Critical Next Steps in Combating Herbicide Resistance: Our View. Weed Sci. 66: 559-561.

Shergill, L. S., Barlow, B. R., Bish, M. D., & Bradley, K. W. (2018). Investigations of 2,4-D and multiple herbicide resistance in a Missouri waterhemp (Amaranthus tuberculatus) population. Weed Sci. 66:386-394.

Shields, E. J., Dauer, J. T., VanGessel, M. J., & Neumann, G. (2006). Horseweed (Conyza canadensis) seed collected in the planetary boundary layer. Weed Sci. 54:1063-1067.

164

Shurtleff, J. L., & Coble, H. D. (1985). Interference of certain broadleaf weed species in soybeans (Glycine max). Weed Sci. 33:654-657.

Sit, V., & Poulin-Costello, M. (1994). Catalogue of curves for curve fitting. Forest Science Research Branch, Ministry of Forests. Victoria, Bristish Columbia, Canada.

Smisek, A. (1995). Resistance to paraquat in Erigeron canadensis L. University of Western Ontario, London, ON. P.102. Master’s Thesis.

Smith, R. G., Warren, N. D., & Cordeau, S. (2020). Are cover crop mixtures better at suppressing weeds than cover crop monocultures?. Weed Sci. 68:186-194.

Soltani, N., Brown, L. R., & Sikkema, P. H. (2016). Control of glyphosate-resistant Canada fleabane in soybean with preplant herbicides. Can. J. Plant. Sci. 97:408-410.

Soltani, N., Jhala A. J., Shropshire, C., & Sikkema, P. H. (2017a). Biologically effective rate of metribuzin for glyphosate-resistant Canada fleabane control in soybean. Can. J. Plant Sci. 97:771-774.

Soltani, N., Dille, J. A., Burke, I. C., Everman, W. J., VanGessel, M. J., Davis, V. M., & Sikkema, P. H. (2017b). Perspectives on potential soybean yield losses from weeds in North America. Weed Technol. 31:148-154.

Soltani, N., Shropshire, C., & Sikkema, P.H. (2020a). Control of glyphosate-resistant marestail in identity-preserved or glyphosate-resistant and glyphosate/dicamba-resistant soybean with preplant herbicides. Am. J. Plant Sci. 11: 851-860.

Soltani, N., Shropshire, C., & Sikkema, P.H. (2020b). Glyphosate-resistant Canada fleabane control with three-way herbicide tankmixes in soybean. Am. J. Plant Sci. 11:1478-1486.

Stalker, D. M., Malyj, L. D., & McBride, K. E. (1988a). Purification and properties of a nitrilase specific for the herbicide bromoxynil and corresponding nucleotide sequence analysis of the bxn gene. J. Biol. Chem. 263:6310-6314.

Stalker, D. M., McBride, K. E., & Malyj, L. D. (1988b). Herbicide resistance in transgenic plants expressing a bacterial detoxification gene. Science. 242:419-423.

Statistics Canada. (2019a). Production of principal field crops, July 2019. https://www150.statcan.gc.ca/n1/daily-quotidien/180831/dq180831b-eng.htm. Accessed November 26, 2019.

Statistics Canada. (2019b). Production of principal field crops, June 2019. https://www150.statcan.gc.ca/n1/daily-quotidien/190626/dq190626b-eng.htm. Accessed November 26, 2019.

Steckel, L. E., Sprague, C. L., Hager, A. G., Simmons, F. W., & Bollero, G. A. (2003). Effects of shading on common waterhemp (Amaranthus rudis) growth and development. Weed Sci. 51:898-903.

165

Steckel, L. E., & Sprague, C. L. (2004a). Late-season common waterhemp (Amaranthus rudis) interference in narrow-and wide-row soybean. Weed Technol. 18: 947-952.

Steckel, L. E., & Sprague, C. L. (2004b). Common waterhemp (Amaranthus rudis) interference in corn. Weed Sci. 52: 359-364.

Steckel, L. E., C. L. Sprague, E. W. Stoller, G. Bollero, & L. M. Wax. (2004). Temperature effects on germination of nine Amaranthus species. Weed Sci. 52:217-221.

Steckel, L. E. (2007). The dioecious Amaranthus spp.: here to stay. Weed Technol. 21:567-570.

Steckel, L. E., Sprague, C. L., Stoller, E. W., Wax, L. M., & Simmons, F. W. (2007). Tillage, cropping system, and soil depth effects on common waterhemp (Amaranthus rudis) seedbank persistence. Weed Sci. 55: 235-239.

Steckel, L. E., Main C. L., & Mueller, T. C. (2010). Glyphosate-resistant horseweed in the United States. in Nandula VK, ed. Glyphosate Resistance in Crops and Weeds: History, Development, and Management. P. 185–193. Hoboken, NJ, United States.

Stewart, C. L., Nurse, R. E., Hamill, A. S., & Sikkema, P. H. (2010). Environment and soil conditions influence pre-and postemergence herbicide efficacy in soybean. Weed Technol. 24: 234-243.

Strom, S. A., Gonzini, L. C., Mitsdarfer, C., Davis, A. S., Riechers, D. E., & Hager, A. G. (2019). Characterization of multiple herbicide–resistant waterhemp (Amaranthus tuberculatus) populations from Illinois to VLCFA-inhibiting herbicides. Weed Sci. 67:369-379.

Swanton, C. J., & Weise, S. F. (1991). Integrated weed management: the rationale and approach. Weed Technol. 5:657-663.

Swanton, C. J., & Murphy, S. D. (1996). Weed Sci beyond the weeds: the role of integrated weed management (IWM) in agroecosystem health. Weed Sci. 44:437-445.

Swanton, C. J., Mahoney, K. J., Chandler, K., & Gulden, R. H. (2008). Integrated weed management: knowledge-based weed management systems. Weed Sci. 56: 168-172.

Sweat, J. K., Horak, M. J., Peterson, D. E., Lloyd, R. W., & Boyer, J. E. (1998). Herbicide efficacy on four Amaranthus species in soybean (Glycine max). Weed Technol. 12:315-321.

Switzer, C. M. (1957). The existence of 2,4-D resistant strains of wild carrot. Proceedings of the North Eastern Weed Control Conference. 11:315-318.

Tahmasebi, B. K., Alebrahim, M. T., Roldán-Gómez, R. A., da Silveira, H. M., de Carvalho, L. B., Alcántara-de la Cruz, R., & De Prado, R. (2018). Effectiveness of alternative herbicides on three Conyza species from Europe with and without glyphosate resistance. Crop Prot. 112:350-355.

166

Tardif, F., & Smith, P. (2003) Alternative herbicides for the control of Canada fleabane in soybeans. Crop Advances: Field Crop Reports (Interim Report). Online. Available: http://www.ontariosoilcrop.org/wp-content/uploads/2015/07/V1Soy7.pdf

Taylor, H. M. (1980). Soybean Growth and Yield as Affected by Row Spacing and by Seasonal Water Supply. Agron. J. 72:543-547.

Teasdale, J. R. (1996). Contribution of cover crops to weed management in sustainable agricultural systems. J. Prod. Agric. 9:475-479.

Thebaud, C. & Abbott, R. J. (1995). Characterization of invasive Conyza species (Asteraceae) in Europe: Quantitative trait and isozyme analysis. Am. J. Bot. 82: 360–368.

Thiel, H., & Varrelmann, M. (2014). Identification of a new PSII target site psbA mutation leading to D1 amino acid leu218val exchange in the Chenopodium album D1 protein and comparison to cross‐resistance profiles of known modifications at positions 251 and 264. Pest Manag. Sci. 70:278-285.

Thompson, M. A., Steckel, L. E., Ellis, A. T., & Mueller, T. C. (2007). Soybean tolerance to early preplant applications of 2,4-D ester, 2,4-D amine, and dicamba. Weed Technol. 21:882-885.

Tietjen, K. G., Kluth, J. F., Andree, R., Haug, M., Lindig, M., Müller, K. H., & Trebst, A. (1991). The herbicide binding niche of photosystem II‐a model. Pestic Sci 31:65-72

Tilley, D. (2012) Plant Guide for Canadian Horseweed (Conyza canadensis). USDA Natural Resources Conservation Service, Aberdeen, ID Plant Materials Center.

Tozzi, E., & Van Acker, R. C. (2014). Effects of seedling emergence timing on the population dynamics of horseweed (Conyza canadensis var. canadensis). Weed Sci. 62: 451-456.

Tranel, P. J., C. W. Riggins, M. S. Bell, & A. G. Hager. (2011). Herbicide resistances in Amaranthus tuberculatus: a call for new options. J. Agric. Food Chem. 59:5808–5812

Trebst, A. (1987). The three-dimensional structure of the herbicide binding niche on the reaction center polypeptides of photosystem II. Z. Naturforsch. Sect. C Biosci. 42:742-750.

Trucco, F., Jeschke, M. R., Rayburn, A. L., & Tranel, P. J. (2005). Amaranthus hybridus can be pollinated frequently by A. tuberculatus under field conditions. Heredity. 94:64-70.

Tsuyuzaki, S., & Kanda, F. (1996). Revegetation patterns and seedbank structure on abandoned pastures in northern Japan. Am. J. Bot. 83:1422-1428.

Uline, E. B., & Bray, W. L. (1895). Synopsis of North American Amaranthaceae II. Bot. Gaz. 20: 155-161

Van Acker, R. C., Swanton, C. J., & Weise, S. F. (1993). The critical period of weed control in soybean [Glycine max (L.) Merr.]. Weed Sci. 41:194-200.

167

Van der Weide, R. Y., Bleeker, P. O., Achten, V. T. J. M., Lotz, L. A. P., Fogelberg, F., & Melander, B. (2008). Innovation in mechanical weed control in crop rows. Weed Res. 48:215-224.

Van Roekel, R. J., Purcell, L. C., & Salmerón, M. (2015). Physiological and management factors contributing to soybean potential yield. Field Crops Res. 182:86-97.

Van Wely, A. C., Soltani, N., Robinson, D. E., Hooker, D. C., Lawton, M. B., & Sikkema, P. H. (2015). Glyphosate and acetolactate synthase inhibitor resistant common ragweed (Ambrosia artemisiifolia L.) in southwestern Ontario. Can. J. Plant. Sci. 95:335-338.

VanGessel, M. J. (2001). Glyphosate-resistant horseweed from Delaware. Weed Sci. 49:703-705.

Vencill, W. K., Nichols, R. L., Webster, T. M., Soteres, J. K., Mallory-Smith, C., Burgos, N. R., & McClelland, M. R. (2012). Herbicide resistance: toward an understanding of resistance development and the impact of herbicide-resistant crops. Weed Sci. 60:2-30.

Vennapusa, A. R., Faleco, F., Vieira, B., Samuelson, S., Kruger, G. R., Werle, R., & Jugulam, M. (2018). Prevalence and mechanism of atrazine resistance in waterhemp (Amaranthus tuberculatus) from Nebraska. Weed Sci. 66:595-602.

Vila‐Aiub, M. M., Neve, P., & Powles, S. B. (2009). Fitness costs associated with evolved herbicide resistance alleles in plants. New Phytol. 184:751-767.

Vila-Aiub, M. M. (2019). Fitness of Herbicide-Resistant Weeds: Current Knowledge and Implications for Management. Plants. 8:469.

Vink, J. P., Soltani, N., Robinson, D. E., Tardif, F. J., Lawton, M. B., & Sikkema, P. H. (2012). Occurrence and distribution of glyphosate-resistant giant ragweed (Ambrosia trifida L.) in southwestern Ontario. Can. J. Plant. Sci. 92:533-539.

Vyn, J. D., Swanton, C. J., Weaver, S. E., & Sikkema, P. H. (2007). Control of herbicide-resistant common waterhemp (Amaranthus tuberculatus var. rudis) with pre-and post-emergence herbicides in soybean. Can. J. Plant. Sci. 87: 175-182.

Wallace, J. M., Curran, W. S., & Mortensen, D. A. (2019). Cover crop effects on horseweed (Erigeron canadensis) density and size inequality at the time of herbicide exposure. Weed Sci. 67:327-338.

Ward, S. M., Webster, T. M., & Steckel, L. E. (2013). Palmer amaranth (Amaranthus palmeri): a review. Weed Technol. 27:12-27.

Weaver, S. E. (2001). The Biology of Canadian weeds. 115. Conyza canadensis. Can. J. Plant. Sci. 81: 867-875.

Weaver, S., Cluney, K., Downs, M., & Page, E. (2006). Prickly lettuce (Lactuca serriola) interference and seed production in soybeans and winter wheat. Weed Sci. 54:496-503.

168

Westra, E. P. (2016). Glyphosate-resistant kochia (Kochia scoparia) management in the central Great Plains and western Canada. Colorado State University, Fort Collins, Colorado. P. 82- 83. Doctoral dissertation.

Wilson, B. J., Wright, K. J., Brain, P., Clements, M., & Stephens, E. (1995). Predicting the competitive effects of weed and crop density on weed biomass, weed seed production and crop yield in wheat. Weed Res. 35:265-278.

Yenish, J. P., Doll, J. D., & Buhler, D. D. (1992). Effects of tillage on vertical distribution and viability of weed seed in soil. Weed Sci. 40:429-433.

Young, B. G., Young, J. M., Gonzini, L. C., Hart, S. E., Wax, L. M., & Kapusta, G. (2001). Weed management in narrow-and wide-row glyphosate-resistant soybean (Glycine max). Weed Technol. 15:112-121.

Young, B. G. (2006). Changes in herbicide use patterns and production practices resulting from glyphosate-resistant crops. Weed Technol. 20:301-307.

Yu, Q., Cairns, A., & Powles, S. (2007). Glyphosate, paraquat and ACCase multiple herbicide resistance evolved in a Lolium rigidum biotype. Planta. 225:499-513

Yu, Q., Jalaludin, A., Han, H., Chen, M., Sammons, R. D., & Powles, S. B. (2015). Evolution of a double amino acid substitution in the 5-enolpyruvylshikimate-3-phosphate synthase in Eleusine indica conferring high-level glyphosate resistance. Plant Physiol. 167:1440-1447.

Yuan, J. S., Tranel, P. J., & Stewart Jr, C. N. (2007). Non-target-site herbicide resistance: a family business. Trends Plant Sci. 12:6-13.

Zandstra, B. H., & Motooka, P. S. (1978). Beneficial effects of weeds in pest management—a review. Pans. 24:333-338.

Zimmer, M., Young, B. G., & Johnson, W. G. (2018). Weed control with halauxifen-methyl applied alone and in mixtures with 2,4-D, dicamba, and glyphosate. Weed Technol. 32: 597- 602.

Zimmerman, M. J. (1969). The influence of light, temperature and humidity on the herbicidal activity of bromoxynil (3, 5-dibromo-4-hydroxybenzo-nitrile). Oregon State University, Corvallis, Oregon. P. 1-13; 16-30. Masters Thesis.

169

Chapter 8: Appendix

8.1 Determining Non-linear Regression Models

To determine fit of non-linear regressions the following steps were used:

1. Visual inspections of the data to gain a general idea of the curve (log-scale on x-axis).

2. Find the SS error in the SAS Summary Statistics (ANOVA) table for each regression. To

determine mean-square error (MSE) divide SS error by the total number of observations

used (n). To determine the root mean square error (RMSE) take the square root of the

MSE for any given variable's regression. MSE and RMSE units are the same scale as

your variable (i.e. % for % control). The lower the RMSE the better the fit. In field

research an RMSE under 20 is adqueate, and under 10 is very good.

3. The relative efficiency (RE) of each mode1 can be determined by the calculation 1-(SS

error divided by SS model). A value of 1 means the model perfectly predicts the observed

data. Realistically, an RE of at least 0.5 - closer to 1 is ideal. An RE over 0 means that the

model predicts the observed data better than the mean of the observed data.

Utlilze the equation that best represents/fits the data as determined by the steps above.

Sarangi and Jhala (2017) and Sit and Poulin-Costello (1994) provide a good overview of

determing fit of a nonlinear regression.

Credit and appreciation to Christy Shropshire for assistance and development of these

guidelines.

170

8.2 SAS Code for Chapter 2

BED Bromoxynil and Bromoxynil + Metribuzin – Control title 'Bromoxynil and Bromoxynil + Metribuzin Nlin BED'; data first; input plot trt rep dose env cont14 cont28 cont56; datalines; ; **/exponential to maximum ; title 'Exponential to Maximum Model'; proc sort data=first; by dose env rep; proc plot; plot cont56*dose; run;

Proc univariate data=first normal plot; var cont56; run;

*Equation B* FOR 14, 28 and 56 DAA Control *a=upper asymptote, b=magnitude constant, c=slope (dose constant)*; proc nlin method=marquardt; parameters a=100 b=100 c=0.01; model cont56=a-b*(exp(-c*dose)); output out=b predicted=bp run; proc corr; var cont56 bp; run; proc sort data=b; by dose; run; proc sgplot data=b; scatter x=dose y=cont56; series x=dose y=bp; run;

BED Bromoxynil and Bromoxynil + Metribuzin – Biomass, Density, and Yield

title 'Bromoxynil and Bromoxynil + Metribuzin NLIN DEN DM Yield'; data first; input plot trt rep dose env Den DM Yield; datalines; 171

; proc sort data=first; by dose env rep; proc plot; plot DM*dose; run;

**For DM and DEN**; **Where a=lower asymptote, b=reduction in y from intercept to asymptote and c=slope (from intercept to a); *proc nlin method=marquardt; *parms a=0 b=40 c=0.002; *bounds a<=0; *bounds a>=0; *model DM=a+b*exp(-c*dose); *output out=b predicted=bp; *run;

*For Yield ; *a=upper asymptote, b=magnitude constant, c=slope (dose constant)*; proc nlin method=marquardt; parameters a=100 b=100 c=0.01; model Yield=a-b*(exp(-c*dose)); output out=b predicted=bp run; proc corr; var DM bp; run; proc sort data=b; by dose; run; proc sgplot data=b; scatter x=dose y=DM; series x=dose y=bp; run;

GLIMMIX - Industry Standard Comparison and Bromoxynil and Bromoxynil + Metribuzin

Comparison Control title 'Bromoxynil Glimmix'; data first; input plot trt rep dose env cont14 cont28 cont56; datalines;

;

172

Run; proc sort data=first; by env trt rep; run; proc means data=first; class trt; var cont56; run; proc glimmix data=first /*nobound method=laplace*/; class env trt rep; model cont56=trt;* /dist= link= ; random env rep(env) env*trt; covtest 'envvar=0' 0 . . ; covtest 'repvar=0' . 0 . ; covtest 'trt*env=0' . . 0 ; lsmeans trt /pdiff adjust=tukey lines ilink; ods output diffs=ppp lsmeans=mmm; output out=second predicted=pred residual=resid residual(noblup)=mresid student=studentresid student(noblup)=smresid; run; proc sort; by studentresid; proc print; run; */Linearity of fixed effects - both a scatter & boxplot; proc sgplot data=second; scatter y=studentresid x=trt; refline 0; proc sgplot data=second; vbox studentresid / group=trt datalabel; run; */Homogeneity of effects; proc sgscatter; plot studentresid*(pred env trt rep); run; */Q-Q plot of normal distribution; */Conditional: Normal distribution, normality plot and Shapiro-Wilk test; proc univariate normal plot data=second; var studentresid; histogram studentresid/normal kernel; run;

GLIMMIX - Industry Standard Comparison and Bromoxynil and Bromoxynil + Metribuzin

Comparison – Density and Biomass title 'Bromoxynil Glimmix DEN DM Yield'; data first; input plot trt rep dose env Den DM Yield; if Den=0 then Den=0.0000000001; if DM=0 then Den=0.0000000001; datalines; 173

; Run; proc sort data=first; by env trt rep; run; proc means data=first; class trt; var yield; run; proc glimmix data=first nobound; /*method=laplace*/; if yield = 0 then yield = 0.0001; class env trt rep; model yield=trt/ dist=normal; *link=log; random env rep(env) env*trt; covtest 'envvar=0' 0 . . ; covtest 'repvar=0' . 0 . ; covtest 'trt*env=0' . . 0 ; lsmeans trt /pdiff adjust=tukey lines ilink; ods output lsmeans=mmm ; output out=second predicted=pred residual=resid residual(noblup)=mresid student=studentresid student(noblup)=smresid; *ods html exclude lsmeans diffs; run; proc sort; by studentresid; proc print; run; proc sort; by studentresid; proc print; run;

8.2 SAS Code for Chapter 3

BED Pyraflufen-ethyl/2,4-D and Pyraflufen-ethyl/2,4-D + Metribuzin - Control

174 title 'Pyraflufen-ethyl/2,4-D and Pyraflufen-ethyl/2,4-D Metribuzin Nlin BED'; data first; input plot trt rep dose env cont14 cont28 cont56; datalines;

; title 'Exponential to Maximum Model'; proc sort data=first; by dose env rep; proc plot; plot cont56*dose; run;

Proc univariate data=first normal plot; var cont56; run;

*Equation A* FOR 14, 28 and 56 DAA Control *a=upper asymptote, b=magnitude constant, c=slope (dose constant)*; proc nlin method=marquardt; parameters a=100 b=100 c=0.01; model cont56=a-b*(exp(-c*dose)); output out=b predicted=bp run;

** Equation B** ****a=intercept , b=magnitude , c=slope**;

*proc nlin method=marquardt; *by timing; *parameters a=0 b=100 c=0.01; *bounds a>=0; *bounds a<=0; *model cont14=a+b*(1-exp(-c*dose)); *output out=b predicted=bp; *run;

**Equation C**Linear; *a=upper asymptote, b=dose constant, c=magnitude constant); *proc nlin method=marquardt; *by timing; *parameters a=100 b=0.001 c=100; *model cont14=a-c*(b*dose); *output out=b predicted=bp; *run;

175

proc corr; var cont56 bp; run; proc sort data=b; by dose; run; proc sgplot data=b; scatter x=dose y=cont56; series x=dose y=bp; run;

BED Pyraflufen-ethyl/2,4-D and Pyraflufen-ethyl/2,4-D + Metribuzin – Biomass, Density and

Yield title 'Pyraflufen-ethyl/2,4-D and Pyraflufen-ethyl/2,4-D + Metribuzin NLIN DEN DM Yield'; data first; input plot trt rep dose env Den DM Yield;

datalines; ; proc sort data=first; by dose env rep; proc plot; plot Den*dose; run; **For DM and DEN**; proc nlin method=marquardt; parms a=0 b=40 c=0.02; bounds a<=0; bounds a>=0; model Den=a+b*exp(-c*dose); output out=b predicted=bp; run; *For Yield ;

*For Yield ; *proc nlin method=marquardt; *by timing; *parameters a=100 b=100 c=0.01; *model Yield=a-b*(exp(-c*dose)); *output out=b predicted=bp; *run; proc corr; var Den bp; run; proc sort data=b; by dose;

176 run; proc sgplot data=b; scatter x=dose y=Den; series x=dose y=bp; run;

GLIMMIX – Industry Standard Comparison and Pyraflufen-ethyl/2,4-D + Metribuzin

Comparison Control title 'Pyraflufen-ethyl/2,4-D Pyraflufen-ethyl/2,4-D + Metribuzin Glimmix'; data first; input plot trt rep dose env cont14 cont28 cont56;

datalines; ; Run; proc sort data=first; by env trt rep; run; proc means data=first; class trt; var cont14; run; proc glimmix data=first /*nobound method=laplace*/; class env trt rep; model cont14=trt /dist=normal link=identity ; random env rep(env) env*trt; covtest 'envvar=0' 0 . . ; covtest 'repvar=0' . 0 . ; covtest 'trt*env=0' . . 0 ; lsmeans trt /pdiff adjust=tukey lines ilink; ods output diffs=ppp lsmeans=mmm; output out=second predicted=pred residual=resid residual(noblup)=mresid student=studentresid student(noblup)=smresid; run; proc sort; by studentresid; proc print; run;

177

*/Conditional: Normal distribution, normality plot and Shapiro-Wilk test; proc univariate normal plot data=second; var studentresid; histogram studentresid/normal kernel; run;

GLIMMIX – Industry Standard Comparison and Pyraflufen-ethyl/2,4-D + Metribuzin

Comparison - Biomass, Density and Yield title 'BlackHawk-Glimmix DEN DM Yield'; data first; input plot trt rep dose env Den DM Yield; datalines;

/***Log Transformation***/; data first; set first; TDen=Den+1; logavar=log(TDen); *TDM=DM+1; *logavar=log(TDM); run;

*proc sort data=first; *by env trt rep; *run; proc means data=first; class trt; var TDen; run; proc glimmix data=first /*nobound; /method=laplace*/; class env trt rep; model TDen=trt /dist=lognormal link=identity; random env rep(env) env*trt; covtest 'envvar=0' 0 . . ; covtest 'repvar=0' . 0 . ; covtest 'trt*env=0' . . 0 ; lsmeans trt /pdiff adjust=tukey lines ilink; output out=second predicted=pred residual=resid residual(noblup)=mresid student=studentresid student(noblup)=smresid; ods output diffs=ppp lsmeans=mmm; run; proc sort; by studentresid; proc print; run; proc sort; by studentresid; proc print; run;

178

/***to calculate standard deviation for lognormal***/; proc print data=mmm; run; proc means data=first noprint; class trt; var logavar; output out=tim mean=mns std=dev; run; proc sort data=tim (firstobs=2); by trt; run; proc print data=tim; run; data combined; merge tim mmm; by trt; run; proc print data=combined; run;

*/Linearity of fixed effects - both a scatter & boxplot; proc sgplot data=second; scatter y=studentresid x=trt; refline 0; proc sgplot data=second; vbox studentresid / group=trt datalabel; run; */Homogeneity of effects; proc sgscatter; plot studentresid*(pred env trt rep); run; */Q-Q plot of normal distribution; */Conditional: Normal distribution, normality plot and Shapiro-Wilk test; proc univariate normal plot data=second; var studentresid; histogram studentresid/normal kernel; run; proc means data=first; class trt; var yield; run; /***back-transformation from lognormal_mean and SE***/ data btdata; set combined; fred=dev*dev/2.5; **if high variability, increase to 2.5-3.5; omega=exp(stderr*stderr); btlsmean=exp(estimate+fred)-1; btvar=exp(2*estimate)*omega*(omega-1); btse_mean=sqrt(btvar); run; title'Backtransformations'; proc print data=btdata; var trt btlsmean btse_mean; run; 179

T-Test for Observed Versus Expected (Colby’s)

**T Test for comparison of Observed vs Expected (Colby’s) title 'T Test for Pyraflufen-ethyl/2,4-D and Metribuzin Colby's'; data first; input plot trt rep dose env cont14 cont28 cont56 den dm exp14 exp28 exp56 eden edm; datalines;

Run; proc sort data=first; by trt env rep; run; proc means data=first; class trt; var dm edm; run; proc ttest data=first; by trt; *paired cont14*exp14; *paired cont28*exp28; *paired cont56*exp56; *paired den*eden; paired dm*edm;

Run;

8.3 SAS Code for Chapter 4

BED Tiafenacil and Tiafenacil + Metribuzin – Control title 'Tiafenacil + Metribuzin Nlin BED'; data first; input plot trt rep dose env cont14 cont28 cont56;

datalines;

;

*title 'Rectangular Hyperbola Model'; proc sort data=first; by dose env rep; proc plot; plot cont28*dose; run; Proc univariate data=first normal plot; var cont28; run; **Equation D**; 180

/***Rectangular hyperbola (Cousen) Ascending - a=upper asymptote, i=initial slope***/ title 'Rectangular Hypoerbola'; proc nlin data=first; parameters a=8.2 i=2; *bounds a<=100; model cont28=(i*dose)/(1+((i*dose)/a)); output out=second predicted=bp; run; proc corr; var cont28 bp; run; proc sort data=second; by dose; run; proc sgplot data=second; scatter x=dose y=cont28; series x=dose y=bp; run;

BED Tiafenacil and Tiafenacil + Metribuzin – Biomass, Density and Yield

title 'Tiafenacil NLIN DEN DM Yield'; data first; input plot trt rep dose env Den DM yield; datalines; run; proc sort data=first; by dose env rep; proc plot; plot yield*dose; run; **For DM and DEN**; *proc nlin ; *parms a=0 *b=40 *c=0.02; *bounds a<=0; *bounds a>=0; *model Den=a+b*exp(-c*dose); *output out=b predicted=bp; *run;

*For Yield ; proc nlin data=first; parameters 181 a=8.2 i=2; *bounds a<=100; model yield=(i*dose)/(1+((i*dose)/a)); output out=b predicted=bp; run; proc corr; var yield bp; run; proc sort data=b; by dose; run; proc sgplot data=b; scatter x=dose y=yield; series x=dose y=bp; run;

GLIMMIX – Industry Standard Comparison and Tiafenacil + Metribuzin Comparison Control title 'Tiafenacil and Tiafenacil + metribuzin Glimmix'; data first; input plot trt rep dose env cont14 cont28 cont56;

*parameter=cont28; *parameter=cont56; *parameter=cont84; datalines;

; Run; proc sort data=first; by env trt rep; run; proc means data=first; class trt; var cont56; run; proc glimmix data=first /*nobound method=laplace*/; class env trt rep; model cont56=trt /*dist=normal link=identity*/ ; random env rep(env) env*trt; covtest 'envvar=0' 0 . . ; covtest 'repvar=0' . 0 . ; covtest 'trt*env=0' . . 0 ; lsmeans trt /pdiff adjust=tukey lines ilink; ods output diffs=ppp lsmeans=mmm; output out=second predicted=pred residual=resid residual(noblup)=mresid student=studentresid student(noblup)=smresid; *ods html exclude lsmeans diffs;

182 run; proc sort; by studentresid; proc print; run;

GLIMMIX – Industry Standard Comparison and Tiafenacil + Metribuzin Comparison Biomass,

Density and Yield title 'Tiafenacil and Tiafenacil + Metribuzin DEN DM Yield'; data first; input plot trt rep dose env Den DM Yield; datalines;

/***Log Transformation***/; data first; set first; *TDen=Den+1; *logavar=log(TDen); TDM=DM+1; logavar=log(TDM); run; proc means data=first; class trt; var TDM; run; proc glimmix data=first /*nobound; /method=laplace*/; class env trt rep; model TDM=trt /dist=lognormal link=identity; random env rep(env) env*trt; covtest 'envvar=0' 0 . . ; covtest 'repvar=0' . 0 . ; covtest 'trt*env=0' . . 0 ; lsmeans trt /pdiff adjust=tukey lines ilink; output out=second predicted=pred residual=resid residual(noblup)=mresid student=studentresid student(noblup)=smresid; ods output diffs=ppp lsmeans=mmm; run;

183 proc sort; by studentresid; proc print; run; proc sort; by studentresid; proc print; run;

*/Linearity of fixed effects - both a scatter & boxplot; proc sgplot data=second; scatter y=studentresid x=trt; refline 0; proc sgplot data=second; vbox studentresid / group=trt datalabel; run; */Homogeneity of effects; proc sgscatter; plot studentresid*(pred env trt rep); run; */Q-Q plot of normal distribution; */Conditional: Normal distribution, normality plot and Shapiro-Wilk test; proc univariate normal plot data=second; var studentresid; histogram studentresid/normal kernel; run; proc means data=first; class trt; var yield; run; /***back-transformation from lognormal_mean and SE***/ data btdata; set combined; fred=dev*dev/3.5; **if high variability, increase to 2.5-3.5; omega=exp(stderr*stderr); 184 btlsmean=exp(estimate+fred)-1; btvar=exp(2*estimate)*omega*(omega-1); btse_mean=sqrt(btvar); run; title'Backtransformations'; proc print data=btdata; var trt btlsmean btse_mean; run;

T-Test for Observed Versus Expected (Colby’s)

**T Test for comparison of Observed vs Expected (Colby’s) title 'Tiafenacil'; data first; input plot trt rep dose env cont14 cont28 cont56 den dm exp14 exp28 exp56 eden edm;

datalines;

Run;

Site by Year Determination title 'Tiafenacil Test for Site'; data first; input plot trt rep year site cont14 cont28 cont56 Den DM yield; datalines;

; /***ANOVA to determine which locations could be combined***;*/ proc glimmix data=first /*nobound method=laplace*/; 185 class year site trt rep; model yield=trt year trt*year /dist=normal link=identity ; random site rep(site) site*trt; covtest 'sitevar=0' 0 . . ; covtest 'repvar=0' . 0 . ; covtest 'trt*site=0' . . 0 ; lsmeans trt /pdiff adjust=tukey lines ilink; ods output diffs=ppp lsmeans=mmm; output out=second predicted=pred residual=resid residual(noblup)=mresid student=studentresid student(noblup)=smresid; run;

8.4 SAS Code for Chapter 5

BED Metribuzin on MS/MR Waterhemp Control input plot trt rep dose env cont14 cont28 cont56 cont84; datalines;

; proc sort data=first; by dose env rep; proc plot; plot cont14*dose; run;

Proc univariate data=first normal plot; var cont14; run;

*Equation A* *a=upper asymptote, b=magnitude constant, c=slope (dose constant)*;

*proc nlin method=marquardt; *by timing; *parameters a=100 b=100 c=0.01; *model cont56=a-b*(exp(-c*dose)); *output out=b predicted=bp *run;

** Equation B** ****a=intercept , b=magnitude , c=slope**; *proc nlin method=marquardt; *by timing; *parameters a=0 b=100 c=0.01; *bounds a>=0; *bounds a<=0; *model cont14=a+b*(1-exp(-c*dose));

186

*output out=b predicted=bp; *run;

**Equation C** *a=upper asymptote, b=dose constant, c=magnitude constant); *proc nlin method=marquardt; *by timing; *parameters a=100 b=0.001 c=100; *model cont14=a-c*(b*dose); *output out=b predicted=bp; *run;

**Equation D**; /***Rectangular hyperbola (Cousen) Ascending - a=upper asymptote, i=initial slope***/ *title 'Rectangular Hypoerbola';

*proc nlin data=first; *parameters a=20 i=2; *bounds a<=100; *model cont56=(i*dose)/(1+((i*dose)/a)); *output out=second predicted=bp; *run;

*Equation E* **/Log-logistic Regression (ascending); **Where d=upper asymptote, c=lower asymptote i50=ED50, b=slope; title 'Log-logistic Model'; proc nlin data=first; parameters d=100 c=0 i50=250 b=2 ; bounds c>=0; bounds d<=100; if dose=0 then model cont14=c; else model cont14=c+(d-c)/(1+EXP(-b*(log(dose)-log(i50)))); output out=b predicted=bp; run; proc corr; var cont14 bp; run; proc sort data=b; by dose; run; proc sgplot data=b; scatter x=dose y=cont14; series x=dose y=bp; 187 run; BED Metribuzin on MS/MR Waterhemp Biomass title 'Waterhemp BED-RES PRE-Density & Biomass'; data first; input plot trt rep dose env Den DM; datalines;

; proc sort data=first; by dose env rep; proc plot; plot Den*dose; run;

Proc univariate data=first normal plot; var Den; run;

*Equation F* **/Log-logistic Regression (descending); **Where d=lower asymptote, c=upper asymptote i50=ED50, b=slope; title 'Log-logistic Model'; proc nlin data=first; parameters d=100 c=0 i50=5000 b=2 ; bounds c>=0; bounds d<=100; if dose=0 then model Den=d; else model Den=c+(d-c)/(1+EXP(b*(log(dose)-log(i50)))); output out=b predicted=bp; run;

*Equation G* **/Inverse Exponential; **Where a=lower asymptote, b=reduction in y from intercept to asymptote and c=slope (from intercept to a); *title 'Inverse Exponential'; *proc nlin ; *parms a=0 b=40 c=0.02; *bounds a<=0; *bounds a>=0; *model DM=a+b*exp(-c*dose); *output out=b predicted=bp; *run; *;

**Equation H**Linear; *a=upper asymptote, b=dose constant, c=magnitude constant);

188

*proc nlin method=marquardt; *by timing; *parameters a=100 b=0.001 c=100; *model Den=a-c*(-b*dose); *output out=b predicted=bp; *run; proc corr; var Den bp; run; proc sort data=b; by dose; run; proc sgplot data=b; scatter x=dose y=Den; series x=dose y=bp; run;

GLIMMIX – Industry Standard Comparison and Tiafenacil + Metribuzin Comparison Control title 'MR Waterhemp Glimmix Control '; data first; input plot trt rep dose env cont14 cont28 cont56 cont84;

**Use following adjustment for arcsine square root trans; *title2 'Control 56 DAA (arcsine square root transformation)'; *y1=cont28; *if y1=100 then y1=100-0.01; *if y1=0 then y1=0+0.01; *y2=y1/100; *y=arsin(sqrt(y2));

datalines;

; Run; proc glimmix data=first /*nobound method=laplace*/; class env trt rep; model y=trt /dist=normal link=id ; random env rep(env) env*trt; covtest 'envvar=0' 0 . . ; covtest 'repvar=0' . 0 . ; covtest 'trt*env=0' . . 0 ; lsmeans trt /pdiff adjust=tukey lines ilink; ods output diffs=ppp lsmeans=mmm; output out=second predicted=pred residual=resid residual(noblup)=mresid student=studentresid student(noblup)=smresid; title "Control 14 DAA"; *ods html exclude lsmeans diffs; 189 run;

**Back transformation of arcsine-square root transformed data; *data btarcdata; *set second; *btlsmean=sin(estimate)*sin(estimate)*100-0.01; *run;

*proc print data=btarcdata; *title3 'arcsine square root backtransfrom'; *var trt btlsmean; *run; proc means data=first; class trt; var cont28; run;

GLIMMIX – Industry Standard Comparison and Tiafenacil + Metribuzin Comparison

Biomass, Density and Yield title 'Waterhemp BED-Density & Biomass'; data first; input plot trt rep dose env Den DM; datalines;

; proc sort data=first; by dose env rep; proc plot; plot DM*dose; run; 190

Proc univariate data=first normal plot; var DM; run;

*Equation F* **/Log-logistic Regression (descending); **Where d=lower asymptote, c=upper asymptote i50=ED50, b=slope; *title 'Log-logistic Model';

*proc nlin data=first; *parameters d=100 c=0 i50=250 b=5 ; *bounds c>=0; *bounds d<=100; *if dose=0 then model DM=d; *else model DM=c+(d-c)/(1+EXP(b*(log(dose)-log(i50)))); *output out=b predicted=bp; *run;

*Equation G* **/Inverse Exponential; **Where a=lower asymptote, b=reduction in y from intercept to asymptote and c=slope (from intercept to a); title 'Inverse Exponential'; proc nlin ; parms a=0 b=40 c=0.02; bounds a<=0; bounds a>=0; model DM=a+b*exp(-c*dose); output out=b predicted=bp; run;

*Equation H* **/Linear; *proc nlin method=marquardt; *parameters a=1 b=0.1; *model DM=(a*trt)+b; *output out=b predicted=bp; *run; ; proc corr; var DM bp; run; proc sort data=b; by dose; run; proc sgplot data=b; scatter x=dose y=DM; series x=dose y=bp; run;

191