Evaluation of Tolpyralate for Weed Management in Field Corn (Zea Mays L.)

Evaluation of Tolpyralate for Weed Management in Field Corn (Zea mays L.)

By:

Brendan A. Metzger

A Thesis Presented to The University of Guelph

In partial fulfilment of requirements For the degree of

Master of Science

Plant Agriculture

Guelph, Ontario, Canada © Brendan A. Metzger, January, 2019

ABSTRACT

EVALUATION OF TOLPYRALATE FOR WEED MANAGEMENT IN FIELD CORN (Zea mays L.)

Brendan A. Metzger Advisors: University of Guelph, 2019 Dr. P. H. Sikkema Dr. D. E. Robinson

Twenty-seven field experiments were conducted from 2015-2018 at nine locations to evaluate tolpyralate, a new pyrazolone 4-hydroxyphenyl-pyruvate dioxygenase inhibitor, for weed management in corn. The biologically-effective dose of tolpyralate and tolpyralate + atrazine was established in eight common annual weed species, and separately in multiple herbicide-resistant (MR) Canada fleabane and waterhemp biotypes. Co-application of tolpyralate with atrazine improved weed control; tolpyralate + atrazine at label rates provided similar or improved control of eight common weed species, MR Canada fleabane and MR waterhemp compared to industry standards. Four weed species were controlled equally regardless of POST application timing and two species were controlled with tolpyralate + atrazine applied pre- emergence; annual grass control declined with delayed applications. Corn was at greatest risk of injury with tolpyralate + atrazine applied at V1 or V3, and at 2X rates. Injury was influenced by environmental variables, but did not translate to grain yield loss.

Acknowledgements

As my time as a student comes to a close, I would like to thank all those who have played a role in making the past two years not only successful but truly enjoyable. To my advisors Dr.’s Peter Sikkema and Darren Robinson, I can’t express how much I appreciate your guidance, support, advice and encouragement. Despite extremely full schedules and extensive field programs, you consistently bend over backwards to make your graduate students and their projects a priority, and go out of your way to make them feel welcome and valued as members of the team. You’ve taught me to be a better researcher, writer and presenter, but I’ve also learned what it means to be an outstanding leader. Thank you also to Dr.’s Dave Hooker and Alan Raeder for serving on my advisory committee, and a special thanks to Alan for making numerous trips north for meetings and tours. From writing to research tours, I greatly appreciate the time, effort and perspective both of you contributed to my project; the quality of this research is improved thanks to your involvement. A sincere thank you to Dr. Nader Soltani, for your tireless efforts in getting manuscripts fit for publication, your dedication to the program, and for always being the graduate students’ number one fan.

Field research is truly a group effort, and I owe a huge thank you to everyone who helped with this project at ground level. To my fellow graduate students: Andrea Smith, Lauren Benoit, Brittany Hedges, Elizabeth Buck, Jessica Quinn and Nicole Langdon, thank you doesn’t begin to describe my appreciation for all of the help, comradery, moral support and comic relief throughout the past two years. Thank you all for welcoming a male presence into the crew; I couldn’t have asked for a better group of people to share a small, windowless office with. To Todd Cowan, Christy Shropshire, Chris Kramer and Lynette Brown, thank you for the time each of you dedicate to making sure graduate students are successful, while still managing busy field programs of your own. I greatly appreciate your help in planning, training and trial execution, and also the independence that you grant graduate students to truly manage their research. A special thanks to Christy for your patience in working through statistics with me and all of the graduate students. To all of the Sikkema summer students at Ridgetown and at Huron, a big thanks to each of you for your help with some of the most gruelling aspects of this project. Whether it was staking, hoeing or colossal weed harvests, I appreciate the sweat equity that each of you have put into this research project and many others. Thank you also to Dr. Michelle Edwards for all of your assistance with statistical analysis. You go above and beyond for all of the graduate students; we and the entire OAC are lucky to have you.

Thank you to ISK Biosciences Corporation, Grain Farmers of Ontario, and the Growing Forward 2 program for their financial contributions that made this project possible, as well as landowners who generously allowed the use of their land.

Finally, I would like to thank my friends and family members for their encouragement and support throughout my studies. I’m privileged to have such a solid group of supportive people in my life.

iii

Table of Contents

ABSTRACT ...... ii Acknowledgements ...... iii Table of Contents ...... iv List of Tables ...... ix List of Figures ...... xiv Chapter 1: Literature Review ...... 1 Section I - Weed Control in Corn ...... 1 1.1 Corn Yield Loss Due to Weed Interference ...... 1 1.2 Effects of Weed Competition in Corn ...... 2 1.3 Development of the Critical Weed-Free Period ...... 3 1.4 Factors Influencing the Critical Weed-Free Period in Corn ...... 4 1.5 Evolution of Weed Management Strategies ...... 6 1.6 Herbicide-Resistant Crops ...... 6 1.7 Herbicide Management Strategies and Two-Pass Weed Control ...... 7 1.8 No-till Farming and Implications for Herbicide Stewardship ...... 9 Section II – HPPD Inhibiting Herbicides and Tolpyralate...... 10 2.1 HPPD Mode of Action ...... 10 2.2 HPPD-Inhibitors: Discovery and Designations ...... 11 2.3 HPPD-Inhibitors: Uptake and Translocation ...... 12 2.4 HPPD Inhibitors: Selectivity ...... 13 2.5 Photosystem II Inhibitors: Interactions and Synergy ...... 14 2.6 Tolpyralate: Discovery and Development ...... 15 2.7 Tolpyralate: Application Timing, Activity and Control Spectrum ...... 16 2.8 Tolpyralate: Comparisons to Industry Standards ...... 17 2.9 Tolpyralate: Crop Tolerance and Re-Cropping Restrictions...... 19 Section III - Effect of Weed Size on Herbicide Efficacy ...... 20 3.1 Effects of Delaying Weed Control ...... 20 3.2 Contact Herbicides ...... 21 3.3 Systemic Herbicides: ALS-Inhibitors, ACCase-Inhibitors and Synthetic Auxins ...... 23 3.4 Systemic Herbicides: Glyphosate ...... 24 3.5 Factors Affecting Herbicide Efficacy ...... 25

3.6 Overcoming the Influence of Weed Size on Herbicide Efficacy ...... 26 Section IV - Glyphosate-Resistant Weeds and Canada Fleabane (Conyza canadensis (L.) Cronq.) ...... 29 4.1 Glyphosate: Discovery and Development ...... 29 4.2 Glyphosate: Mode of Action ...... 29 4.3 Glyphosate: Application, Uptake and Selectivity ...... 30 4.4 Glyphosate-Resistant Crops: Development ...... 30 4.5 Glyphosate-Resistant Crops: Adoption ...... 31 4.6 Glyphosate-Resistant Weeds ...... 33 4.7 Mechanisms of Glyphosate-Resistance: Target Site/Non-Target Site Resistance ...... 34 4.8 Canada fleabane: Biology ...... 36 4.9 Glyphosate-Resistant Canada fleabane: Resistance Mechanisms and Distribution ...... 37 4.10 Herbicide-Resistant Canada fleabane: Agronomic Implications and Management Strategies in Corn...... 38 Section V - Hypotheses and Objectives ...... 40 5.1 Hypotheses ...... 40 5.2 Objectives ...... 40 Chapter 2: Tolpyralate Efficacy: Part I. Biologically-effective dose of tolpyralate for control of annual grass and broadleaf weeds in corn ...... 42 2.1 Abstract ...... 42 2.2 Introduction ...... 43 2.3 Materials and Methods ...... 45 Experimental Methods ...... 45 Statistical Analysis ...... 47 2.4 Results and Discussion ...... 49 Weed Control ...... 49 Yield and Crop Injury ...... 54 2.5 Conclusions ...... 54 Chapter 3: Tolpyralate Efficacy: Part II. Comparison of three group 27 herbicides applied POST for annual grass and broadleaf weed control in corn ...... 63 3.1 Abstract ...... 63 3.2 Introduction ...... 64 3.3 Materials and Methods ...... 66 Experimental Methods ...... 66 Statistical Analysis ...... 68

3.4 Results and Discussion ...... 69 Common Lambsquarters ...... 69 Velvetleaf ...... 70 Pigweed Species ...... 70 Common Ragweed ...... 71 Ladysthumb ...... 73 Wild Mustard ...... 74 Green Foxtail ...... 74 Barnyardgrass ...... 75 3.5 Conclusions ...... 76 Chapter 4: Influence of weed size and herbicide rate on the efficacy of tolpyralate plus atrazine for control of annual grass and broadleaf weeds ...... 85 4.1 Abstract ...... 85 4.2 Introduction ...... 86 4.3 Materials and Methods ...... 88 Experimental Methods ...... 88 Statistical Analysis ...... 90 4.4 Results and Discussion ...... 91 Common Ragweed ...... 91 Common Lambsquarters ...... 93 Green Pigweed ...... 96 Velvetleaf ...... 97 Ladysthumb ...... 98 Flower-of-an-hour ...... 100 Barnyardgrass ...... 103 Green Foxtail ...... 105 Crop Injury ...... 109 Grain Yield ...... 109 4.5 Conclusions ...... 111 Chapter 5: Multiple herbicide-resistant Canada fleabane (Conyza canadensis (L.) Cronq.) dose response to tolpyralate and tolpyralate plus atrazine, and comparison to industry standard herbicides ...... 132 5.1 Abstract ...... 132 5.2 Introduction ...... 133

5.3 Materials and Methods ...... 135 Experimental Methods ...... 135 Statistical Analysis ...... 137 5.4 Results and Discussion ...... 140 Crop Injury ...... 140 Dose-Response ...... 140 Tolpyralate and Herbicide Standards ...... 143 Corn Grain Yield ...... 145 5.5 Conclusions ...... 146 Chapter 6: Multiple herbicide-resistant waterhemp (Amaranthus tuberculatus (Moq.) J. D. Sauer) dose response to tolpyralate and tolpyralate plus atrazine, and comparison to industry standard herbicides ...... 152 6.1 Abstract ...... 152 6.2 Introduction ...... 153 6.3 Materials and Methods ...... 156 Experimental Methods ...... 156 Statistical Analysis ...... 158 6.4 Results and Discussion ...... 160 Crop Injury ...... 160 Dose-Response ...... 161 Tolpyralate, Tolpyralate plus Atrazine and Herbicide Standards ...... 165 Corn Grain Yield ...... 167 6.5 Conclusions ...... 167 Chapter 7: Effect of hybrid, application timing and herbicide rate on corn tolerance to tolpyralate plus atrazine ...... 174 7.1 Abstract ...... 174 7.2 Introduction ...... 175 7.3 Materials and Methods ...... 177 Experimental Methods ...... 177 Statistical Analysis ...... 179 7.4 Results and Discussion ...... 181 Factorial Analysis of Fixed Effects ...... 181 Multiple Stepwise Regression of Predictor Variables ...... 185 7.5 Conclusions ...... 187

vii

Chapter 8: Discussion ...... 198 8.1 Contributions...... 198 8.2 Limitations ...... 201 8.3 Opportunities for Future Research ...... 202 Chapter 9: Literature Cited ...... 205 Chapter 10: Appendix ...... 226 10.1 – SAS Code for Chapter 2 Non-linear Regression Analysis ...... 226 10.2 – SAS Code for Chapter 3 GLIMMIX Analysis ...... 228 10.3 – SAS Code for Chapter 4 Two-Factor GLIMMIX Analysis ...... 230 10.4 – SAS Code for Chapter 5 and 6 GLIMMIX Analysis and Non-linear Regression ...... 232 10.5 – SAS Code for Chapter 7 Split-Split Block GLIMMIX and Mixed-Model Multiple Stepwise Regression ...... 235

viii

List of Tables

Table 2.1. Soil characteristics, planting, spraying and harvest dates for trials near Ridgetown and Exeter, Ontario, Canada in 2015, 2016 and 2017……………………………………………………………..56

Table 2.2. Non-linear regression parameters (± se) and predicted tolpyralate or tolpyralate + atrazine dose required for 50, 80 and 90% control of velvetleaf (ABUTH), pigweed species (AMASS), common ragweed (AMBEL), common lambsquarters (CHEAL), barnyardgrass (ECHCG), ladysthumb (POLPE), green foxtail (SETVI) and wild mustard (SINAR) at 2 wk after application in field studies conducted in Ontario, Canada in 2015, 2016 and 2017…………………………………57

Table 2.3. Non-linear regression parameters (± se) and predicted tolpyralate and tolpyralate + atrazine dose required for 50, 80 and 90% control of velvetleaf (ABUTH), pigweed species (AMASS), common ragweed (AMBEL), common lambsquarters (CHEAL), barnyardgrass (ECHCG), ladysthumb (POLPE), green foxtail (SETVI) and wild mustard (SINAR) at 4 wk after application in field studies conducted in Ontario, Canada in 2015, 2016 and 2017…………………………………58

Table 2.4. Non-linear regression parameters (± se) and predicted tolpyralate and tolpyralate + atrazine dose required for 50, 80 and 90% control of velvetleaf (ABUTH), pigweed species (AMASS), common ragweed (AMBEL), common lambsquarters (CHEAL), barnyardgrass (ECHCG), ladysthumb (POLPE), green foxtail (SETVI) and wild mustard (SINAR) at 8 wk after application in field studies conducted in Ontario, Canada in 2015, 2016 and 2017…………………………………59

Table 2.5. Non-linear regression parameters (± se) and predicted tolpyralate and tolpyralate + atrazine dose required for 50, 80 and 90% reduction in velvetleaf (ABUTH), pigweed species (AMASS), common ragweed (AMBEL), common lambsquarters (CHEAL), barnyardgrass (ECHCG), ladysthumb (POLPE), green foxtail (SETVI) and wild mustard (SINAR) density relative to non- treated check plots within blocks at 8 wk after application in field studies conducted in Ontario, Canada in 2015, 2016 and 2017………………………………………………………………………60

Table 2.6. Non-linear regression parameters (± se) and predicted tolpyralate and tolpyralate + atrazine dose required for 50, 80 and 90% control of velvetleaf (ABUTH), pigweed species (AMASS), common ragweed (AMBEL), common lambsquarters (CHEAL), barnyardgrass (ECHCG), ladysthumb (POLPE), green foxtail (SETVI) and wild mustard (SINAR) dry biomass relative to non- treated check plots within blocks at 8 wk after application in field studies conducted in Ontario, Canada in 2015, 2016 and 2017………………………………………………………………………61

Table 2.7. Non-linear regression parameters (± se) and predicted tolpyralate, or tolpyralate + atrazine dose required to obtain 50, 80 and 90% grain yield of weed-free control plots within blocks in field studies conducted in Ontario, Canada in 2015, 2016 and 2017………………………………………62

Table 3.1 - CHEAL. Control 1, 2, 4 and 8 weeks after application (WAA), density and dry biomass of common lambsquarters (Chenopodium album (L.)) following treatment with tolpyralate, tolpyralate + atrazine, topramezone + atrazine and mesotrione + atrazine in field studies conducted in Ontario, Canada in 2015, 2016 and 2017………………………………………………………………………77

Table 3.2 - ABUTH. Control 1, 2, 4 and 8 weeks after application (WAA), density and dry biomass of velvetleaf (Abutilon theophrasti Medik.) following treatment with tolpyralate, tolpyralate + atrazine, topramezone + atrazine and mesotrione + atrazine in field studies conducted in Ontario, Canada in 2015, 2016 and 2017………………………………………………………………………………….78

Table 3.3 - AMASS. Control 1, 2, 4 and 8 weeks after application (WAA), density and dry biomass of green/redroot pigweed (Amaranthus powelli (S.) Watson/Amaranthus retroflexus (L.)) following treatment with tolpyralate, tolpyralate + atrazine, topramezone + atrazine and mesotrione + atrazine in field studies conducted in Ontario, Canada in 2015, 2016 and 2017…………………………………79

Table 3.4 - AMBEL. Control 1, 2, 4 and 8 weeks after application (WAA), density and dry biomass of common ragweed (Ambrosia artemisiifolia (L.)) following treatment with tolpyralate, tolpyralate + atrazine, topramezone + atrazine and mesotrione + atrazine in field studies conducted in Ontario, Canada in 2015, 2016 and 2017………………………………………………………………………80

Table 3.5 - POLPE. Control 1, 2, 4 and 8 weeks after application (WAA), density and dry biomass of ladysthumb (Persicaria maculosa Gray) following treatment with tolpyralate, tolpyralate + atrazine, topramezone + atrazine and mesotrione + atrazine in field studies conducted in Ontario, Canada in 2015, 2016 and 2017………………………………………………………………………………….81

Table 3.6 - SINAR. Control 1, 2, 4 and 8 weeks after application (WAA), density and dry biomass of wild mustard (Sinapis arvensis (L.)) following treatment with tolpyralate, tolpyralate + atrazine, topramezone + atrazine and mesotrione + atrazine in field studies conducted in Ontario, Canada in 2015, 2016 and 2017………………………………………………………………………………….82

Table 3.7 - SETVI. Control 1, 2, 4 and 8 weeks after application (WAA), density and dry biomass of green foxtail (Setaria viridis L. P. Beauv.) following treatment with tolpyralate, tolpyralate + atrazine, topramezone + atrazine and mesotrione + atrazine in field studies conducted in Ontario, Canada in 2015, 2016 and 2017………………………………………………………………………………….83

Table 3.8 - ECHCG. Control 1, 2, 4 and 8 weeks after application (WAA), density and dry biomass of barnyardgrass (Echinochloa crus-galli L. P. Beauv.) following treatment with tolpyralate, tolpyralate + atrazine, topramezone + atrazine and mesotrione + atrazine in field studies conducted in Ontario, Canada in 2015, 2016 and 2017………………………………………………………………………84

Table 4.2. Effect of rate and application timing on common ragweed (Ambrosia artemisiifolia) control 2, 4 and 8 WAE/WAA, and density and dry biomass reduction with tolpyralate + atrazine in field experiments conducted in Ontario, Canada in 2017/2018…………………………………………...115

Table 4.3. Interaction of herbicide rate and application timing on control of common ragweed (Ambrosia artemisiifolia) 2 and 4 WAE/WAA, and density and dry biomass reduction with three rates of

tolpyralate + atrazine applied PRE, ePOST, mPOST or lPOST in field experiments conducted in Ontario, Canada in 2017/2018……………………………………………………………………….116

Table 4.4. Effect of rate and application timing on common lambsquarters (Chenopodium album) control 2, 4 and 8 WAE/WAA, and density and dry biomass reduction with tolpyralate + atrazine in field experiments conducted in Ontario, Canada in 2017/2018…………………………………………...117

Table 4.5. Interaction of herbicide rate and application timing on control of common lambsquarters (Chenopodium album) 2 and 4 WAE/WAA, and density and dry biomass reduction with three rates of tolpyralate + atrazine applied PRE, ePOST, mPOST or lPOST in field experiments conducted in Ontario, Canada in 2017/2018……………………………………………………………………….118

Table 4.6. Effect of rate and application timing on green pigweed (Amaranthus powelli) control 2, 4 and 8 WAE/WAA, and density and dry biomass reduction with tolpyralate + atrazine in field experiments conducted in Ontario, Canada in 2017/2018………………………………………………………...119

Table 4.7. Effect of rate and application timing on velvetleaf (Abutilon theophrasti) control 2, 4 and 8 WAE/WAA, and density and dry biomass reduction with tolpyralate + atrazine in field experiments conducted in Ontario, Canada in 2017/2018………………………………………………………...120

Table 4.8. Interaction of herbicide rate and application timing on control of velvetleaf (Abutilon theophrasti) 2 and 4 WAE/WAA, and density and dry biomass reduction with three rates of tolpyralate + atrazine applied PRE, ePOST, mPOST or lPOST in field experiments conducted in Ontario, Canada in 2017/2018……………………………………………………………………….121

Table 4.9. Effect of rate and application timing on ladysthumb (Persicaria maculosa) control 2, 4 and 8 WAE/WAA, and density and dry biomass reduction with tolpyralate + atrazine in field experiments conducted in Ontario, Canada in 2017/2018………………………………………………………...122

Table 4.10. Interaction of herbicide rate and application timing on control of ladysthumb (Persicaria maculosa) 8 WAE/WAA with three rates of tolpyralate + atrazine applied PRE, ePOST, mPOST or lPOST in field experiments conducted in Ontario, Canada in 2017/2018…………………………..123

Table 4.11. Effect of rate and application timing on flower-of-an-hour (Hibiscus trionum) control 2, 4 and 8 WAE/WAA and density and dry biomass reduction with tolpyralate + atrazine in field experiments conducted in Ontario, Canada in 2017/2018………………………………………………………...124

Table 4.12. Interaction of herbicide rate and application timing on control of flower-of-an-hour (Hibiscus trionum) 4 and 8 WAE/WAA, and density reduction with three rates of tolpyralate + atrazine applied PRE, ePOST, mPOST or lPOST in field experiments conducted in Ontario, Canada in 2017/2018...... …………………………………………………………………….125

Table 4.13. Effect of rate and application timing on barnyardgrass (Echinochloa crus-galli) control 2, 4 and 8 WAE/WAA and density and dry biomass reduction with tolpyralate + atrazine in field experiments conducted in Ontario, Canada in 2017/2018………………………………………….126

Table 4.14. Interaction of herbicide rate and application timing on control of barnyardgrass (Echinochloa crus-galli) 2 WAE/WAA and dry biomass reduction with three rates of tolpyralate + atrazine applied

PRE, ePOST, mPOST or lPOST in field experiments conducted in Ontario, Canada in 2017/2018……………………………………………………………………………………………127

Table 4.15. Effect of rate and application timing on green foxtail (Setaria viridis) control 2, 4 and 8 WAE/WAA and density and dry biomass reduction with tolpyralate + atrazine in field experiments conducted in Ontario, Canada in 2017/2018………………………………………………………...128

Table 4.16. Interaction of herbicide rate and application timing on control of green foxtail (Setaria viridis) 2 and 4 WAE/WAA, and density and dry biomass reduction with three rates of tolpyralate + atrazine applied PRE, ePOST, mPOST or lPOST in field experiments conducted in Ontario, Canada in 2017/2018…………………………………………………………………………………………129

Table 4.18. Interaction of herbicide rate and application timing on corn injury 1, 2 and 4 WAA, and grain yield with three rates of tolpyralate + atrazine applied PRE, ePOST, mPOST or lPOST in field experiments conducted in Ontario, Canada in 2017/2018…………………………………………...131

Table 5.2. Regression parameters and predicted effective dose of tolpyralate and tolpyralate plus atrazine for 50, 80 and 95% visible control of multiple-resistant Canada fleabane (Conyza canadensis L. Cronq.) 2, 4 and 8 WAA, and to achieve 50, 80 and 95% of the yield obtained in weed-free control plots based on the log-logistic dose response equation (Equation 1)………………………………..148

Table 5.3. Multiple-resistant Canada fleabane (Conyza canadensis L. Cronq.) visible control 2, 4 and 8 WAA and reduction in plant density and dry biomass 8 WAA provided by commercial rates of tolpyralate, tolpyralate + atrazine, dicamba/atrazine and bromoxynil + atrazine applied postemergence in field studies conducted in Ontario, Canada during 2017/2018……………………….151

Table 6.2. Regression parameters and the predicted dose of tolpyralate required to obtain 50, 80 and 95% visible control of multiple herbicide-resistant common waterhemp (Amaranthus tuberculatus (Moq.) J.D. Sauer) 2, 4, 8 and 12 WAA, a 50, 80 and 95% reduction in density and dry biomass, and 50, 80 and 95% of the yield in WFC plots, from field experiments conducted in Ontario, Canada in 2018, based on the log-logistic dose response equation (Equation 1)……………………………………...171

xii

biomass, and 50, 80 and 95% of the yield in WFC plots, from field experiments conducted in Ontario, Canada in 2018, based on the log-logistic dose response equation (Equation 1)……………………172

Table 6.4. Percent visible control of multiple herbicide-resistant waterhemp (Amaranthus tuberculatus (Moq.) J.D. Sauer) 2, 4, 8, and 12 WAA, percent reduction in density and dry biomass, and corn grain yield provided by tolpyralate, tolpyralate + atrazine and two industry standard herbicides from trials conducted in Ontario, Canada in 2018………………………………………………………………173

Table 7.1. Planting dates, harvest dates, soil characteristics and application information for experiments conducted near Ridgetown and Exeter, Ontario in 2017 and 2018………………………………….189

Table 7.3. Interaction of tolpyralate + atrazine application timing and corn hybrid on visible injury 1 WAA in field experiments conducted in Ontario, Canada in 2017/2018……………………………191

Table 7.7. Summary of mixed-model multiple stepwise regression analysis of significant (P<0.05) predictor variables for visible crop injury 2 WAA following application of tolpyralate + atrazine at two rates and three timings to four corn hybrids, from experiments conducted near Ridgetown and Exeter, Ontario, Canada in 2017 and 2018. …………………………………………………………195

Table 7.8. Summary of mixed-model multiple stepwise regression analysis of significant (P<0.05) predictor variables for visible crop injury 4 WAA following application of tolpyralate + atrazine at two rates and three timings to four corn hybrids, from experiments conducted near Ridgetown and Exeter, Ontario, Canada in 2017 and 2018. …………………………………………………………196

Table 7.9. Summary of mixed-model multiple stepwise regression analysis of significant (P<0.05) predictor variables for visible crop injury 8 WAA following application of tolpyralate + atrazine at two rates and three timings to four corn hybrids, from experiments conducted near Ridgetown and Exeter, Ontario, Canada in 2017 and 2018.. ………………………………………………………...197

xiii

List of Figures

Figure 5.1. Inverse exponential function (Equation 2) and predicted effective dose of tolpyralate for a 50, 80 and 95% reduction in multiple-resistant Canada fleabane (Conyza canadensis L. Cronq.) dry biomass, based on six field experiments conducted in Ontario, Canada in 2017/2018…………..…149

Figure 5.2. Inverse exponential function (Equation 2 ) and predicted effective dose of tolpyralate + atrazine for a 50, 80 and 95% reduction in multiple-resistant Canada fleabane (Conyza canadensis L. Cronq.) dry biomass, based on six field experiments conducted in Ontario, Canada in 2017/2018...149

Figure 5.3. Inverse exponential function (Equation 2) and predicted effective dose of tolpyralate for a 50, 80 and 95% reduction in multiple-resistant Canada fleabane (Conyza canadensis L. Cronq.) density, based on six field experiments conducted in Ontario, Canada in 2017/2018……………...………..150

Figure 5.4. Inverse exponential function (Equation 2) and predicted effective dose of tolpyralate plus atrazine for a 50, 80 and 95% reduction in multiple-resistant Canada fleabane (Conyza canadensis L. Cronq.) density, based on six field experiments conducted in Ontario, Canada in 2017/2018…...... 150

xiv

Chapter 1: Literature Review

Section I - Weed Control in Corn

1.1 Corn Yield Loss Due to Weed Interference Effective weed control in field corn (Zea mays L.) is critical to minimize yield loss due to weed interference. Aside from unforeseen environmental variables, weed competition represents the most important factor impacting corn yield (Rajcan and Swanton 2001). A recent meta- analysis of corn-producing US states and Canadian provinces reported an average yield loss of

51% in corn crops where no weed control measures were implemented (Soltani et al. 2016).

Quackgrass (Elytrigia repens (L.) Desv. ex B.D. Jacks) has been determined to reduce corn yield by up to 37% (Young et al. 1984). In studies conducted in Nebraska, shattercane (Sorghum bicolor L.) interference reduced corn yields by over 75% (Hayden and Burnside 1987).

Increasing densities of grass weed species, including shattercane, have been linearly correlated with decreasing corn yield (Beckett et al. 1988). Dicotyledonous weed species also impact corn yield substantially. Cocklebur (Xanthium strumarium L.), common lambsquarters (Chenopodium album L.), Palmer amaranth (Amaranthus palmeri S.), giant ragweed (Ambrosia trifida L.) and

Canada fleabane (Conyza canadensis (L.) Cronq.) at densities as low as 5-14 plants m-1 row reduced corn grain yield by 27, 12, 91, >90, and 68% respectively (Beckett et al. 1988; Massinga et al. 2003; Harrison et al. 2001; Ford et al. 2014). In Ontario, redroot pigweed (Amaranthus retroflexus L.) infestations at even lower densities (2 plants m-1 of row) reduced corn yield by over 20% (Knezevic et al. 1994). Regardless of species and density, effective weed management programs must be implemented to minimize corn yield loss due to weeds.

1.2 Effects of Weed Competition in Corn Yield loss in corn due to weed interference is caused by a concatenation of environmental, physical and physiological processes, which extend beyond basic resource competition. Fundamentally, if resources such as light, water or nutrients are limiting, competition for these resources with weeds will result in measureable crop yield loss (Ford et al.

1994; Reid et al. 2014). Weed species that emerge early and have a rapid and aggressive seedling growth habit relative to that of corn may compete directly for available sunlight, reducing corn grain yield simply through shading (Beckett et al. 1988). Conversely, if corn exhibits rapid early- season growth, it may be able to effectively shade out many competing weed seedlings (Ford et al. 1994). Beckett et al. (1988) observed that moderate, early-season weed pressure in corn reduced corn ear weight, whole ear yield and kernel weight, even when all other plant resources were non-limiting. Cerrudo et al. (2012) reported that early-season weed competition in corn delayed silking by 4.5 days, subsequently leading to a reduction in total plant dry matter and kernel number per plant at harvest, even though there was excess water and nutrients available.

These results indicate that weed competition ultimately alters physiological processes of resource partitioning within corn plants. Common cocklebur competition has been correlated with a linear decline in corn leaf area index (LAI) in irrigated studies (Hussain et al. 2011). For each 1 kg ha-1 increase in common cocklebur biomass, there was a decline of 1.28-1.35 kg ha-1 in aboveground corn biomass, leading researchers to conclude that weed competition results in changes in carbon partitioning (Hussain et al. 2011). In response to these findings, it has been postulated that the change in red/far-red light ratio due to the presence of weeds may be the signalling mechanism which triggers the “shade avoidance response” leading to decreased LAI, taller/thinner plants and reduced overall plant biomass (Rajcan and Swanton 2001; Rajcan et al. 2004; Ballaré et al.

1990; Page et al. 2009). Ultimately, while the final outcome of weed-crop competition in corn is

2 yield loss, the underlying factors contributing to the physiological alterations which cause this yield loss are a complex interaction of multiple biotic and abiotic processes.

1.3 Development of the Critical Weed-Free Period The impact of weed competition on crop yield is not constant throughout all stages of crop development, leading to the development of the “critical weed-free period” (CWFP). In

Ontario, understanding the CWFP was initially thought to be a potential way to reduce prophylactic use of herbicides in tomatoes (Weaver and Tan 1983), soybean (Van Acker et al.

1993) and corn (Hall et al. 1992). Additionally, it was suggested that post-emergence (POST) herbicide applications could be applied at an optimal timing and dosage with knowledge of the

CWFP, thereby increasing economic return on in-season herbicide applications (Knezevic et al.

2002; Swanton and Weise 1991).

Knezevic et al. (1994) observed that redroot pigweed interference in field corn prior to

V5 caused yield loss, while pigweed that emerged after V5 did not reduce corn yield. Secondary to direct resource competition, studies designed to mimic weed competition at various times throughout the season found that reduction in corn growth rate due to shading was influenced by timing (pre-anthesis, silking and post-silking) (Liu and Tollenaar 2009). Accordingly, the CWFP represents this specific time period during which weed interference must be minimized in order to prevent yield loss, and is generally thought to consist of two main components: a) length of time weeds can remain in the crop before yield loss occurs, and b) length of time that weeds must be controlled to prevent yield loss (Weaver and Tan 1983). More broadly, the CWFP can be defined as the period of time during the life of the crop when weed interference will cause crop yield loss (Zimdahl 2004); however, determination of the CWFP is influenced by relative time of weed and crop emergence, weed species composition, weed density, weed leaf area

3 relative to that of the crop, and weather conditions (Kropff and Spitters 1991; Gower et al. 2002).

To define “yield loss” an acceptable threshold must be discerned, typically ranging from 5-20%

(Swanton et al. 1999). Due to differences in crop morphology and physiology however, it is accepted that the CWFP is crop-specific (Knezevic et al. 2002). An Ontario study observed that the CWFP in corn was from V1-V11, with weed interference occurring during this timeframe reducing leaf number per plant, LAI, and yield by >5% (Hall et al. 1992). Similarly, Tursun et al.

(2016) observed that the CWFP in field corn ranged from V1 to V12. Gantoli et al. (2013) reported the CWFP for corn was from V4-VT. Conversely, Cox et al. (2006) reported corn yield losses of 0.7 kernel rows per ear when POST herbicide application was delayed from the V3-V4 stage, to the V4-V6 leaf stage, suggesting there may be an inconsistent rate of yield loss within the originally postulated V1-V12 timeframe. It has also been determined that weeds present at corn emergence may induce linear yield loss until V3, compared to corn emerging into a weed- free environment (Page et al. 2009). Generally, controlling weeds from the time of corn emergence until 3-5 weeks after crop emergence minimizes corn yield loss due to weed interference (Hall et al. 1992).

1.4 Factors Influencing the Critical Weed-Free Period in Corn The onset and duration of the CWFP in corn is dependent on several factors related to the crop, weeds, management practices and environmental variables. Weed species composition and density has a significant impact on the CWFP. In field studies with low to medium weed density, corn yield loss was similar when weeds were controlled at V2 or V8 (Myers et al. 2005). In contrast, in studies with high weed density, weeds had to be controlled by V4 in order to avoid yield loss (Myers et al. 2005). With most weed species, the beginning of the CWFP is relatively constant, but weed species composition does impact the length of the CWFP, particularly when

POST, non-residual herbicides are used (Cox et al. 2006). In terms of management, studies

4 conducted in Nebraska showed that the addition of 120 kg ha-1 nitrogen (N) fertilizer both delayed onset and decreased the length of the CWFP in irrigated corn, even when the effects of

N fertilizer on yield were not detectable (Evans et al. 2003). The findings of Evans et al. (2003) suggest that the addition of N may increase the competitiveness or stress tolerance of corn, reducing its sensitivity to weed interference and subsequent yield loss. It has also been postulated that increasing corn planting density and decreasing corn row width could delay or shorten the

CWFP (Murphy et al. 1996; Teasdale 1995). Teasdale (1995) reported that corn grown on 38 cm row spacing reduced light transmittance through the crop canopy 7 days earlier compared to corn grown at standard 76 cm spacing, and suggested that the end of the CWFP may occur earlier in narrow-row corn due to complete canopy closure. In contrast, Norsworthy and Oliveira (2004) observed that row spacing did not impact the CWFP. Studies conducted with sweet corn determined that the CWFP was also influenced by seeding date; corn seeded in early May had an earlier onset and longer duration of the CWFP, and there was greater yield loss compared to corn seeded in mid-June (Williams 2006). Tillage practices also alter the CWFP of corn. Halford et al.

(2001) reported a more stable, predictable CWFP, as well as a general shift to earlier onset and shorter duration of the CWFP under no-till compared to conventional-tillage. Corn type and hybrid has been observed to affect CWFP due to differing plant architecture and plant growth habits (Tursun et al. 2016; Travlos et al. 2011). In a study conducted in 2013-2014, it was determined that the CWFP varied depending on the type of corn; in field corn the CWFP was

V1-V12, in popcorn the CWFP was VE-V10, and in sweet corn the CWFP was V2-V10 (Tursun et al. 2016). In general, studies agree that early-season weed control in corn is critical, but there is considerable range in the onset and duration of the defined CWFP in corn based on a suite of management factors and external environmental variables beyond a manager’s control.

1.5 Evolution of Weed Management Strategies There has been an evolution of weed control practices in corn over the past century, with herbicides now the primary method of weed management in North America. Prior to the 1950’s, weed control in corn was accomplished primarily through mechanical tillage and inter-row cultivation (Timmons 1970). However, following the discovery and commercialization of selective herbicides such as the phenoxy carboxylic acids and triazines, which controlled select weed species without injury to the crop, interest grew in chemical weed control for corn production (Timmons 1970). A 9-year study in Wisconsin in the 1950’s-1960’s determined that application of simazine resulted in equal or greater weed control and higher corn yield than two in-season mechanical cultivations (Bucholtz and Doersch 1968). Consequently, corn producers adopted the use of herbicides that provided more consistent weed control than mechanical techniques, and there was a concurrent increase in corn yield (Olson and Eidman 1992). As availability of herbicides increased, adoption of herbicides for weed control was rapid; by the mid 1980’s over 96% of US corn acres were treated with herbicides (Olson and Eidman 1992).

Though the specific herbicides used changed over the past 6 decades, the percent of acres treated with an herbicide remain relatively constant; over 95% of corn grown in the US is treated with at least one herbicide on an annual basis (Gianessi and Reigner 2007). Due to improved weed control, lower cost and increased corn yield, the use of herbicides has quickly become the most widely-used method of weed control in field corn.

1.6 Herbicide-Resistant Crops The development of herbicide-resistant crops (HRC) led to drastic changes in weed management programs used in corn production. Glyphosate-resistant (GR) corn was introduced in 1998, and was rapidly adopted by North American corn producers (Dalley et al. 2006).

Glyphosate is a non-selective, systemic herbicide that inhibits the enzyme 5-

6 enolpyruvylshikimate 3-phosphate synthase (EPSPS) in susceptible plants (Cobb and Reade

2010). Prior to the introduction of HRC, corn herbicide programs relied heavily on selective, soil-applied herbicides such as atrazine, acetochlor, and metolachlor applied at relatively high rates (Myers et al. 2005; Green 2012). However, with the introduction of HRC, which exhibited resistance to herbicides such as glyphosate, reliance on soil-applied herbicides was drastically reduced (Devine 2005; Gower et al. 2002). In 2014, over 62% of US corn acreage received at least one glyphosate application (National Agricultural Statistics Service (NASS) 2015). Rapid adoption of GR corn was due to characteristics of glyphosate that made it highly desirable. The

GR system was seen as a method that was easy-to-use, economical, and provided effective, consistent weed control across a range of weed species and environmental conditions (Green and

Owen 2011; Gulden et al. 2009). This general shift to POST herbicides such as glyphosate in corn production further increased the value associated with knowing the CWFP in corn, given that herbicide applications could be timed accordingly (Fickett et al. 2013). However, reliance on a single herbicide mode of action (MOA) increases the potential for the evolution of herbicide- resistant (HR) weed biotypes (Gulden et al. 2009), an issue which has become clear in Ontario in recent years, particularly with regard to glyphosate (Beckie et al. 2014). For this reason, the continued use of multiple herbicide MOA in sequence or in combination is recommended

(Gower et al. 2002; Beckie et al. 2014; Nurse et al. 2006). Overall, the introduction of HRC and realization of the benefits associated with them have drastically changed weed control programs in corn.

1.7 Herbicide Management Strategies and Two-Pass Weed Control The use of multiple herbicide modes of action which overlap the CWFP and provide season-long weed control in corn is considered to be a sound weed control strategy. The traditional herbicides used before the introduction of HR corn largely included soil-applied

7 residual compounds such as triazines or chloroacetamides, which were applied prior to seeding or crop emergence in order to prevent early weed emergence (Green 2012; Grichar and Minton,

2006). However, the use of a pre-emergence-only herbicide program which fails to control late germinating weed seeds has been implicated in contributing to the buildup of weed seeds in the soil, potentially causing economic loss in subsequent crops in the rotation, even if yield is maintained in the current season (Gower et al. 2002). Additionally, weed control with soil- applied residual herbicides is variable depending on precipitation after application (Stewart et al.

2012). In contrast, POST herbicides such as glyphosate and glufosinate do not provide any residual weed control, meaning any weeds which emerge after application are not controlled

(Tharp and Kells 2002; Ferrell and Witt, 2002). One proposed solution is to apply POST herbicides such as glyphosate or glufosinate in combination with residual herbicides. Tharp and

Kells (2002) observed that glufosinate or glyphosate applied as a tank mixture with a residual herbicide to HR corn POST, provided improved and extended control of velvetleaf (Abutilon theophrasti Medik.), common ragweed (Ambrosia artemisiifolia L.) and redroot pigweed compared to sequential POST applications of either glyphosate or glufosinate alone. Two-pass weed control programs of an effective soil-applied residual herbicide followed by (fb) glyphosate

POST, or tank-mix combinations of glyphosate with a residual applied POST provide more consistent weed control. Grichar and Minton (2006) observed that a soil-applied herbicide fb glyphosate POST provided improved control of broadleaf signalgrass (Urochloa platyphylla

(Munro ex C. Wright) R.D. Webster) and prostrate pigweed (Amaranthus albus L.) in GR corn, compared to either herbicide applied alone, particularly in years with above-average soil moisture. Similarly, Culpepper and York (1999), Hellwig et al. (2003) and Lindsey et al. (2012), observed that two-pass weed management strategies of a soil-applied residual herbicide fb a

8 broad-spectrum POST herbicide provided the greatest control of grass and broadleaf weeds, and highest economic return. Overall, the use of multiple herbicide MOA in combination or sequentially as part of a two-pass strategy, results in improved and extended weed control in corn.

1.8 No-till Farming and Implications for Herbicide Stewardship The increased employment of no- or reduced-tillage crop production systems will increase the reliance on herbicides for weed management in corn. In the US, over 62,500,000 acres are under no-till management (Huggins and Reganold 2008). In Ontario, adoption of no- or reduced-till corn production has been slower, with just over 10% of total corn acreage in the province being managed without tillage (Soil Conservation Council of Canada 2017). The implementation of a no- or reduced-till production system provides a range of ecological benefits, such as increased nutrient cycling, reduced potential for soil erosion, improved soil structure, increased water infiltration and decreased fuel and labour costs (Uri 2000; Lal et al.

2007). Without tillage as a component of a weed management program however, there is an increased reliance on herbicides for weed control (Snyder et al. 2016). The evolution of herbicide technology has also been cited as one of the main drivers of the success of no-till crop production systems (Derpsch 2008). In particular, herbicides which control a wide range of monocotyledonous and dicotyledonous weed species are needed in order to facilitate successful no-till systems (Soane et al. 2012). Weed species shifts may occur with the transition from conventional- to no-till production systems, further increasing the importance of efficacious, broad-spectrum herbicides for effective weed management (Soane et al. 2012). Certainly, the continued adoption of no- or reduced-tillage systems which eliminate or reduce mechanical weed control will place increased reliance on effective, broad-spectrum herbicides as the primary method of weed control in field corn and other crops.

Section II – HPPD Inhibiting Herbicides and Tolpyralate

2.1 HPPD Mode of Action Hydroxyphenyl-pyruvate dioxygenase-inhibiting herbicides kill susceptible plants through competitive inhibition of the 4-hydroxyphenyl-pyruvate dioxygenase enzyme (HPPD).

In plants, HPPD catalyzes the oxidative decarboxylation and hydroxylation of 4-hydroxyphenyl- pyruvate, an intermediate in the tyrosine degradation pathway which ultimately yields homogentisic acid (HGA) (Hawkes 2012; Kakidani and Hirai 2003). Homogentisic acid is an intermediate involved in the biosynthesis of plastoquinone (PQ) and α-tocopherols within plants

(Matsumoto et al. 2002; Tsegaye et al. 2002). Homogentisic acid undergoes decarboxylation and prenylation to yield 2-methyl-6-phytyl-1,4-benzoquinone, which is further methylated to produce

PQ or, alternatively, 2,3-dimethyl-6-phytyl-1,4-benzoquinone, the final intermediate in the biosynthesis of tocopherols (Hawkes 2012). Plastoquinone is required as an electron acceptor in the photosynthetic electron transport chain and is also involved in the biosynthesis of carotenoid pigments in subsequent processes (Schulz et al. 1993). Plastoquinone and tocopherols both act as antioxidants in plants, serving to dissipate singlet oxygen and other reactive oxygen species

(ROS) produced under excessive sunlight, and are also involved in stress signaling processes

(Ahrens et al. 2013; Hyun et al. 2010; Liu and Lu 2016). Without protection from ROS, depletion of carotenoid pigments results in the generation and proliferation of singlet oxygen, which destroys proteins, lipids and the photosynthetic complex, releasing triplet chlorophyll

(Hawkes 2012). Triplet chlorophyll induces further photo-damage to leaf pigments, leading to the striking whitening/bleaching symptoms characteristic of HPPD-inhibiting herbicides (Ahrens et al. 2013; Hawkes 2012; Siddall et al. 2002). In general terms, inhibition of the HPPD enzyme in plants leads to white bleaching of plant tissue and subsequent cell death, due to depletion of

PQ and α-tocopherols, photo-degradation of carotenoid pigments and ultimately, the destruction of chlorophyll by singlet oxygen.

2.2 HPPD-Inhibitors: Discovery and Designations The HPPD-inhibiting class of herbicides used for weed management in corn can be divided into 3 main chemical families which include the triketones, isoxazoles (diketone derivatives) and the pyrazolones (benzoylpyrazoles). In 1977, scientists at Stauffer Chemical observed bleaching herbicidal properties exhibited by leptospermone, a compound exuded from the roots of Callistemon citrinus (Curtis) Skeels, commonly known as the bottle-brush tree

(Mitchell et al. 2001). Using leptospermone as a model, chemists optimized the chemical structure and subsequently developed the first HPPD-inhibiting herbicides, classified as triketones (Edmunds and Morris 2012). These first HPPD-inhibitors included sulcotrione and mesotrione, released commercially in 1993 and 2001, respectively, for use in corn (Edmunds and

Morris 2012; Mitchell et al. 2001). Concurrently, the isoxaflutole molecule was developed as another HPPD-inhibiting herbicide for use in corn, with a chemical structure which differed from the triketones (Ahrens et al. 2013). Isoxaflutole is considered a pro-herbicide and is degraded to its herbicidally-active diketonitrile (DKN) derivative in water, soil and vegetation via opening of the isoxazole ring within its structure (Beltran et al. 2001; Pallet et al. 1998). Most recently, pyrazolone (or benzoylpyrazole) herbicides such as topramezone have been introduced, representing a third chemical family that exhibits the HPPD-inhibiting mode of action (MOA)

(Edmunds and Morris 2012). Several herbicide molecules exhibit HPPD-inhibition in plants, however their differing structural conformations have led to their separate designations as triketones, isoxazoles (diketones) or pyrazolones.

2.3 HPPD-Inhibitors: Uptake and Translocation Although the fundamental MOA is consistent, there is variation among HPPD-inhibitors in terms of herbicide uptake, translocation and metabolism due to differences in chemical structure. Despite conformational differences which designate chemical families, certain HPPD- inhibitors share common characteristics. The DKN of isoxaflutole and sulcotrione share a common central functional group which mimics the ketoacid substrate involved in HPPD- mediated reactions, resulting in virtually identical competitive inhibition of HPPD by each herbicide (Viviani et al. 1998). Additionally, nearly all current HPPD-inhibiting herbicides bind to the iron atom on the ‘open’ form of HPPD, contributing to a similar structure-activity relationship among all three classes of HPPD-inhibitors (Ahrens et al. 2013; Edmunds and

Morris 2012). However, structural differences relate more closely to the physico-chemical behavior of the specific herbicide molecule, contributing to variation among HPPD-inhibitors in terms of uptake, translocation to the target-site, and metabolism (Ahrens et al. 2013; Edmunds and Morris 2012). In the case of triketones, specific substituents on the aromatic ring influence the acidity of the entire molecule, thereby affecting herbicide translocation within plants

(Edmunds and Morris 2012; Lee et al. 1998). Varying these substituents may also affect herbicide affinity for the HPPD target site (Lee et al. 1998). Additionally, certain substituents on

HPPD-inhibitors may improve herbicidal activity in grasses by actively impeding metabolic processes within the plant (Edmunds and Morris 2012), indicating that herbicide selectivity could be altered based on structural variations as well. All three families of HPPD-inhibiting herbicides exhibit the same MOA, but structural variations and substitutions contribute to differences in herbicide uptake, translocation and metabolism.

2.4 HPPD Inhibitors: Selectivity Differential metabolism is a key factor in the selectivity and crop tolerance of HPPD- inhibiting herbicides. Increased metabolic degradation represents a major factor in corn tolerance to members of all three classes of HPPD-inhibiting herbicides, including isoxaflutole, mesotrione and topramezone (Grossmann and Erhardt 2007; Mitchell et al. 2001; Pallet et al. 1998). In the case of mesotrione, the metabolic processes in corn are so rapid that almost none of the active form is translocated beyond the initial site of entry (Mitchell et al. 2001). Initial absorption of mesotrione by corn foliage, shoots and roots also occurs at a significantly slower rate relative to that of susceptible weed species (Mitchell et al. 2001). Furthermore, decreased sensitivity of the target HPPD enzyme in tolerant crops plays a role in the selectivity of mesotrione. Hawkes

(2012) reported a minimum 100-fold difference in the potency of mesotrione between broadleaf- derived HPPD and grass-derived HPPD, correlating to the drastically higher activity of mesotrione in broadleaf weeds compared to grasses. Selectivity of topramezone is due in part to differential metabolism; a reduced sensitivity of the enzymatic target to the herbicide in tolerant species such as corn has also been determined to play a role (Ahrens et al. 2013; Grossmann and

Ehrhardt 2007). Unlike mesotrione however, foliar absorption of topramezone has been determined to occur at nearly identical rates in both corn and susceptible weed species

(Grossmann and Ehrhardt 2007). Isoxaflutole is taken up readily by roots and shoots of both corn and susceptible weed species; however, the initial metabolic conversion to its active DKN occurs much more rapidly in susceptible species, leading to relatively good tolerance in corn to isoxaflutole applied PRE (Pallet et al. 2001). Despite this inherent tolerance, crop injury was observed under certain conditions when isoxaflutole was applied at rates sufficient for acceptable weed control (Ahrens et al. 2013).

This led to the development of cyprosulfamide, an herbicide safener for use in conjunction with isoxaflutole (Ahrens et al. 2013). Herbicide safeners are chemical compounds that selectively reduce phytotoxicity of herbicides to the crop without reducing weed control, through physiological or molecular mechanisms such as increasing expression of genes encoding herbicide-metabolizing enzymes (Davies and Caseley 1999; Hatzios and Burgos 2004; Riechers et al. 2010). The addition of cyprosulfamide to isoxaflutole allowed for higher rates to be used and extended the application window to include POST timing (Ahrens et al. 2013). Although multiple mechanisms of tolerance to HPPD-inhibitors exist in corn, differential metabolism plays a central role in tolerance to all three families of the HPPD-inhibiting herbicides.

2.5 Photosystem II Inhibitors: Interactions and Synergy Due to their complementary modes of action, a synergistic effect has been observed when some HPPD-inhibiting herbicides are applied in combination with photosystem-II (PSII) - inhibiting herbicides. PSII-inhibitors, including triazines and substituted ureas, compete with PQ for the QB-binding site on the D1 protein within the photosynthetic electron transport chain (Hess

2000). When the herbicide occupies the binding site on the D1 protein, PQ is displaced and electron flow through photosystem II is interrupted, leading to increased generation of damaging triplet chlorophyll and ROS, in turn causing lipid peroxidation and cell death (Hess 2000).

Although carotenoid pigments serve to dissipate ROS and free radicals, the levels generated within treated plant tissues following application of PSII-inhibitors far exceeds the quenching ability of carotenoids (Hess 2000). Abendroth et al. (2006), Kim et al. (1999) and Kohrt and

Sprague (2017a) among others suggest that a synergistic response is induced when PSII- inhibitors (particularly atrazine) are applied in combination with HPPD-inhibitors, due to their interrelated modes of action. HPPD-inhibitors limit PQ, carotenoid and tocopherol biosynthesis within the plant as PSII-inhibitors compete with PQ for the binding site on the D1 protein,

14 ultimately making atrazine a more efficient PSII-inhibitor and exacerbating lipid peroxidation by subsequently produced free radicals due to depletion of carotenoids and α-tocopherols (Armel et al. 2005). The addition of atrazine to mesotrione applied POST has demonstrated synergistic responses in red morninglory (Ipomoea coccinea L.), common cocklebur (Xanthium strumarium

L.), and even atrazine-resistant (AR) Palmer amaranth (Amaranthus palmeri (S.) Wats), and resulted in improved control of redroot pigweed and velvetleaf compared to either herbicide applied alone (Abendroth et al. 2006; Armel et al. 2007; Creech et al. 2004; Kohrt and Sprague

2017a). Similarly, the addition of atrazine to tembotrione resulted in a synergistic response in AR

Palmer amaranth (Kohrt and Sprague 2017a). Atrazine added to PRE applications of isoxaflutole has also increased control of entireleaf morninglory (Ipomoea hederacea var. integruiscula) and Palmer amaranth (Stephenson and Bond 2012). Thus, the co-application of

HPPD-inhibitors with PSII-inhibitors is well-documented as a practice that can increase overall herbicidal activity, and broaden the weed control spectrum due to interconnected herbicide modes of action.

2.6 Tolpyralate: Discovery and Development Tolpyralate is a new pyrazolone herbicide molecule which exhibits the HPPD-inhibiting

MOA and has been developed for weed management in corn. Tolpyralate (SL-573) was discovered by Ishihara Sangyo Kaisha Ltd. in 2008, and classified as a WSSA Group 27 pyrazolone herbicide, designated by the presence of a pyrazole ring within the molecular structure (Kikugawa et al. 2015). Tolpyralate is considered a pro-herbicide, with the active species being the first metabolite of the tolpyralate molecule (Jeanmart et al. 2016). Commercial development of tolpyralate as an herbicide began in 2009 on a limited basis (Tonks et al. 2015).

In 2011, full development of the molecule for use as a selective, post-emergence herbicide for use in corn began in the USA, with a combination of university trials and in-house experiments

15 overseen by ISK Biosciences Corporation, a subsidiary of Ishihara Sangyo Kaisha Ltd. (Tonks et al. 2015). Registration for tolpyralate was subsequently granted in the USA and Canada in 2017

(ISK Biosciences Corporation 2018; Health Canada 2018). Experiments conducted in the USA have targeted rate response across several weed species, crop tolerance in field, seed, popcorn and sweet corn, carryover/re-cropping limitations for following crops, and performance of tank- mixtures with existing herbicides (Tonks et al. 2015). Tolpyralate is formulated as a 400 g L-1 suspension concentrate, with a proposed use rate of between 30 to 40 g ai ha-1 (Anonymous

2017a). Tolpyralate represents the second pyrazolone herbicide developed for weed management in corn, after topramezone.

2.7 Tolpyralate: Application Timing, Activity and Control Spectrum Tolpyralate provides effective control of several annual grass and broadleaf weeds, particularly when applied POST in a tank mixture with atrazine. The optimal herbicide application timing for tolpyralate is reported to be POST, potentially due to the relatively low water solubility of the molecule (26.5 mg L-1) (Kikugawa et al. 2015; Tonks et al. 2015). This contrasts with certain other HPPD-inhibitors such as mesotrione, which has water solubility of

~15,000 mg L-1 and exhibits good herbicidal activity in susceptible weed species following applications made to the soil (PRE) (Gillespie et al. 2011; United States Environmental

Protection Agency 2001). The addition of methylated seed oil (MSO) and 28% urea ammonium nitrate (UAN) or ammonium sulphate (AMS) as adjuvants to POST applications of tolpyralate are reported to provide best results (Tonks et al. 2015). In 2014 field trials at Michigan State

University, tolpyralate applied POST at 30 and 40 g ha-1 with MSO and AMS provided 95% control of annual grasses, (compared to 90% and 0% with topramezone and mesotrione plus atrazine respectively), and 97 to 99% control of common lambsquarters, green pigweed, common ragweed and velvetleaf (Sprague and Powell 2014). The addition of atrazine (560 g ha-

1) to tolpyralate (40 g ha-1) increased control of annual grasses marginally from 97% to 99%

(Sprague and Powell 2014); however, the benefit of adding atrazine for annual grass control with tolpyralate is not consistent across individual species. Tonks et al. (2015) reported increases of up to 18% in the control of fall panicum (Panicum dichotimiflorum L.) when atrazine was included, compared to tolpyralate applied alone (Tonks et al. 2015). In all species highlighted by

Tonks et al. (2015) except large crabgrass (Digitaria sanguinalis (L.) Scop.), the addition of atrazine at 560 to 1120 g ha-1 to tolpyralate at 30 and 40 g ha-1 increased the level of weed control to some extent. Tonks et al. (2015) reported that tolpyralate at 40 g ha-1 controlled velvetleaf 87%, Amaranthus spp. 90%, common ragweed 90%, giant ragweed 91%, common lambsquarters 88%, kochia (Bassia scoparia (L.) A. J. Scott) 82% and cocklebur 88% across multiple trials. The addition of atrazine (560-1120 g ha-1) to tolpyralate (40 g ha-1) increased control to between 93 and 97% for all the above broadleaf species evaluated (Tonks et al. 2015).

Control of ivyleaf morningglory (Ipomoea hederacea L.) with tolpyralate (40 g ha-1) applied

POST increased from 29% to 88% control when atrazine was included (Tonks et al. 2015).

These results support the hypothesis that weed control is improved when HPPD-inhibitors are applied with PSII-inhibitors, although responses are not always synergistic. Overall, tolpyralate applied POST is reported to provide broad spectrum control of many annual broadleaf and grass weeds, particularly when atrazine is included in the application.

2.8 Tolpyralate: Comparisons to Industry Standards Initial field experiments comparing the efficacy of tolpyralate to currently available

HPPD-inhibitors have revealed some key differences. Kikugawa et al. (2015) compared tolpyralate to topramezone, and two triketones (mesotrione and tembotrione) for the control of three glyphosate resistant (GR) weed species: tall waterhemp (Amaranthus tuberculatus (Moq.)

Sauer), giant ragweed and Canada fleabane. Tolpyralate (30 g ha-1) controlled GR tall

17 waterhemp, giant ragweed and Canada fleabane 99, 96 and 98% respectively (Kikugawa et al.

2015). Similarly, tolpyralate plus atrazine controlled common waterhemp 99% in University of

Illinois field experiments; although these treatments also included a PRE herbicide of S- metolachlor, S-metolachlor plus atrazine, dimethenamid-P plus saflufenacil, acetochlor or acetochlor plus atrazine (University of Illinois 2014). Tolpyralate and tolpyralate plus atrazine tank mixtures provided better control (98%) of 8 and 15 cm-tall AR Palmer amaranth compared to mesotrione plus atrazine, tembotrione plus atrazine and topramezone plus atrazine (Kohrt and

Sprague 2017a). Control of GR giant ragweed and Canada fleabane with tembotrione was similar to that observed with tolpyralate, however tembotrione provided only 80% control of GR tall waterhemp (Kikugawa et al. 2015). Kikugawa et al. (2015) reported that tolpyralate and tembotrione provided similar control of GR giant ragweed and Canada fleabane; however, tolpyralate was more efficacious than tembotrione for the control of GR tall waterhemp.

Furthermore, Kikugawa et al. (2015) reported that mesotrione provided similar levels of control for GR giant ragweed as tembotrione and tolpyralate, but control of GR tall waterhemp and

Canada fleabane with mesotrione was only 65% and 85%, respectively. Due to their similar chemical structure (pyrazolone), tolpyralate and topramezone could be expected to exhibit comparable herbicidal properties and weed control spectrum. Control of several annual broadleaf weed species reported by Kikugawa et al. (2015), Sprague and Powell (2014) and Tonks et al.

(2015) was similar for both tolpyralate and topramezone, particularly when each herbicide was applied with atrazine. However, tolpyralate (30 g ha-1) provided 3-14% greater control of annual grasses including green foxtail (Setaria viridis (L.) P. Beauv.) giant foxtail (Setaria faberi

Herm.) fall panicum, woolly cupgrass (Eriochloa villosa (Thunb.) Kunth), and large crabgrass than topramezone (18.4 g ha-1) (Tonks et al. 2015). Additionally, topramezone controlled GR tall

18 waterhemp, GR giant ragweed, and GR Canada fleabane only 55, 88 and 71%, compared to tolpyralate which provided 99, 96 and 98% control, respectively (Kikugawa et al. 2015). In contrast, Purdue University researchers observed that tolpyralate plus atrazine controlled giant ragweed 72%, compared to 85% control with topramezone plus atrazine, however S-metolachlor was applied PRE before each treatment in this case (Johnson 2014). Despite some inconsistencies across environments, initial results suggest that even with their common designation as pyrazolone herbicides, the herbicidal activity of topramezone and tolpyralate differs in certain weed species. Taken together, initial field research concludes that there are some notable differences with respect to the efficacy of tolpyralate compared to current industry standards.

2.9 Tolpyralate: Crop Tolerance and Re-Cropping Restrictions There is a wide margin of crop safety in field, seed and sweet corn to tolpyralate applied

POST, and many crops can be seeded the following year without deleterious effects from tolpyralate residues. Tolpyralate plus atrazine applied POST with MSO and AMS, following application of a PRE herbicide did not cause any phytotoxicity in field corn (University of

Illinois 2014). Tolpyralate (30-40 g ha-1) plus atrazine applied POST at V3 and V5 corn, tank- mixed with glyphosate or glufosinate caused less than 5% crop injury (Johnson 2014; Owen et al. 2014). Tolpyralate (30 or 40 g ha-1) alone or with atrazine, or atrazine plus either glyphosate or glufosinate applied POST to V4 corn did not cause any crop injury at 7, 14 and 28 DAT

(Sprague and Powell 2014). Tonks et al. (2015) reported that in trials from 23 seed companies covering more than 120 corn hybrids, including sweet corn and popcorn, over 90% of trials reported less than 1% crop injury when tolpyralate was applied with and without atrazine; less than 0.5% of all trials resulted in greater than 10% crop injury (n=214). ISK Biosciences proposes that the application timing for tolpyralate in corn be from V1 to V6 stage (Anonymous

2017a). Additionally, re-cropping restrictions for tolpyralate are relatively short (3 months for cereal crops and grasses; 9 months for soybean, legumes, Brassicaceae crops, cucurbits, cotton and tomato; 18 months for sugarbeets) (Tonks et al. 2015). In a study by Kikugawa et al. (2015), oilseed rape was sown at various times after application of tolpyralate, tembotrione, mesotrione and topramezone (Kikugawa et al. 2015). No injury was observed when oilseed rape was sown

34 days after application (DAA) of tolpyralate, tembotrione and mesotrione, but topramezone still injured oilseed rape seeded 56 DAA (Kikugawa et al. 2015). While of less importance in a continuous corn production system, re-cropping restrictions relative to the persistence of herbicides in soil have significant implications for producers employing a wider crop rotation, particularly those including horticultural crops. Initial results suggest that field corn, sweet corn and popcorn exhibit good tolerance to tolpyralate, and re-cropping intervals for several common field and horticultural crops do not appear to restrict the usability of tolpyralate.

Section III - Effect of Weed Size on Herbicide Efficacy

3.1 Effects of Delaying Weed Control In general, the efficacy of post-emergence (POST) herbicides declines as weed size increases at the time of application; however, there are several secondary factors that can influence control of large weeds. Ultimately, in order to be effective, an herbicide must be intercepted by the target plant, and reach its site of action in sufficient concentration to be lethal

(Rojano-Delgado et al. 2014). Contact herbicides exhibit limited translocation within treated plants, often killing only the treated foliage, meaning thorough spray coverage is imperative

(Rojano-Delgado et al. 2014; Stoller et al. 1975). This can lead to decreased control of larger weeds simply due to insufficient coverage (Steckel et al. 1997; Stoller et al. 1975). However, the efficacy of some systemic herbicides is also affected by weed size (Johnson and Norsworthy

2014). The underlying reasons for this decrease in control with increasing size have been attributed to species-specific morphological traits such as leaf orientation (Lee and Oliver 1982) or to molecular/chemical properties of the herbicide itself, such as water solubility (Steckel et al.

1997). In some cases, increasing herbicide rate has been shown to minimize or negate the decreased control of larger weeds; however, these rates may not be within herbicide label recommendations (Blackshaw 1989; King and Oliver 1992; Lee and Oliver 1982; Steckel et al.

1997).

There is an exception to the notion of decreased weed control with increased weed size; in some cases, herbicide efficacy is increased by delaying application. Delayed application of

POST herbicides with limited residual activity improves control of weeds with prolonged emergence patterns (Carey and Kells 1995; Johnson and Norsworthy 2014; Tharp and Kells

1999). Improved weed control with delayed herbicide application in these cases can be attributed to a greater number of weeds which are emerged at the time of application, and therefore are exposed to a POST, non-residual herbicide, rather than a function of weed size at time of application.

Aside from a noteworthy exception, the delayed application of POST herbicides generally results in reduced control of annual grass and broadleaf weeds, a longer period of weed interference, and a concomitant decrease in crop yield (Carey and Kells 1995). The underlying causes of this trend however, depend on a number of biological, environmental and chemical factors.

3.2 Contact Herbicides The efficacy of contact (non-systemic) herbicides decreases with increasing weed size at the time of application. Herbicides such as bentazon and glufosinate exhibit little to no

21 translocation once absorbed by plant foliage (Rojano-Delgado et al. 2014; Steckel et al. 1997).

The herbicidal effects of bentazon on susceptible species are primarily limited to the foliage that intercepts the herbicide (Stoller et al. 1975). Treated leaves turn necrotic soon after treatment, but due to limited translocation, new leaves can emerge normally if meristems of larger weeds are not killed (Stoller et al. 1975). In greenhouse studies, bentazon controlled biennial wormwood (Artemisia biennis Wild) >80% when applied to 6-20 cm-tall seedlings but provided less than 44% control of plants larger than 28 cm (Kegode and Fronning 2005). A similar decline in glufosinate efficacy on larger weeds has been widely reported in the literature. Steckel et al.

(1997) observed that glufosinate consistently controlled green foxtail and Pennsylvania smartweed (Polygonum pensylvanicum L.) when weeds were 5-10 cm tall at application but reported erratic control of 15 cm-tall weeds, which was attributed to insufficient coverage of the larger weeds at the time of application. Similarly, Blackshaw (1989) determined that glufosinate

(1000 g ai ha-1) provided greater control of green foxtail, Russian thistle (Salsola tragus L.), wheat (Triticum aestivum (L.) cv. ‘Katepwa’) and barley (Hordeum vulgare (L.) cv. ‘Klages’) when applied to small compared to larger plants. Coetzer et al. (2002) reported that glufosinate provided inconsistent control of larger common waterhemp (Amaranthus rudis Sauer), redroot pigweed and Palmer amaranth, which was attributed to weed height at time of application.

Corbett et al. (2004) observed that control of ladysthumb (Persicaria maculosa Gray) and

Pennsylvania smartweed with glufosinate (291 g ai ha-1) declined from 94 and 98%, to 73 and

75%, when application was delayed from 2-5 cm to 8-10 cm-tall plants, respectively. This was in contrast, however, to common ragweed, velvetleaf and smooth pigweed (Amaranthus hybridus

L.), all of which were controlled >90% at both application timings with the same rate of glufosinate (Corbett et al. 2004). These results may reflect the differential sensitivity of various

22 weed species to glufosinate. Although species-specific differences exist with respect to the impact of weed size at the time of application, there is sufficient evidence in the literature that the efficacy of contact herbicides generally declines as weed size increases.

3.3 Systemic Herbicides: ALS-Inhibitors, ACCase-Inhibitors and Synthetic Auxins Weed size at the time of application and the efficacy of systemic herbicides are also correlated. For successful weed control, systemic herbicides must be intercepted and retained by treated plants, and be internally translocated beyond the point of contact to their specific site of action where they accumulate and elicit activity (Franz et al. 1997). This requirement may affect the potential for plant-herbicide interactions influenced by weed size or stage, and such interactions have been documented with several systemic herbicides and weed species.

Nicosulfuron, an acetolactate synthase (ALS)-inhibiting herbicide, applied to 15, or 30 to 45 cm- tall johnsongrass (Sorghum halepense (L.) Pers.), provided 90 and <74% control, respectively

(Johnson and Norsworthy 2014). When this application was delayed further to 60 cm-tall johnsongrass, control declined to 26% (Johnson and Norsworthy 2014). Contrary to these results,

Obrigawitch et al. (1990) observed better control and less regrowth of johnsongrass with applications of nicosulfuron made to 30 or 60 cm plants compared to 10 cm plants. The efficacy of clethodim, an acetyl coenzyme-A carboxylase (ACCase)-inhibitor, for the control of johnsongrass was reduced by 22% when the application was delayed from boot stage to full- panicle stage (Johnson and Norsworthy 2014). Similarly, control of dogfennel (Eupatorium capillifolium (Lam.) Small) with synthetic auxin herbicides 2,4-D amine, dicamba, and triclopyr

+ fluroxypyr has been shown to decline as weed height at time of application increases (Sellers et al. 2009).

3.4 Systemic Herbicides: Glyphosate Several studies have examined the effect of weed size on the efficacy of glyphosate.

Glyphosate is rapidly absorbed by foliage, symplastically-translocated, and exhibits high activity within treated plants (Franz et al. 1997). Considering such properties, some studies have reported little interaction between glyphosate efficacy and weed size (Corbett et al. 2004; Johnson and

Norsworthy 2014). Other studies however, have reported a decrease in control of larger plants with glyphosate, compared to plants treated in early growth stages. Kegode and Fronning (2005) found that control of biennial wormwood with glyphosate (420 g ae ha-1) declined from 100 to

25% when plants were sprayed at 15-20 compared to 39-47 cm timing, respectively. Similarly, glyphosate provided variable control of prickly sida (Sida spinosa L.) and velvetleaf which had

4-6 or 7-10 leaves, compared to applications at the 1-3 leaf stage (Jordan et al. 1997). In general, control of redroot pigweed, common ragweed, common lambsquarters, barnyardgrass and green foxtail with glyphosate declined when applications were made to 30 compared to 10 or 20 cm- tall weeds (Soltani et al. 2016). Krausz et al. (1996) reported similar trends in some species, while no weed-size effect was seen in other species. Glyphosate at 560 g ai ha-1 (isopropylamine salt formulation (480 g ai L-1)) applied to 10 and 20 cm-tall ivyleaf morningglory (Ipomoea hederacea Jacq.) provided 91% and 33% control, respectively (Krausz et al. 1996). Conversely, control of giant foxtail, fall panicum and redroot pigweed was 100%, regardless of application timing (Krausz et al. 1996). These results were in agreeance with those of Tharp and Kells

(1999), who reported no difference in control of giant foxtail with glyphosate applied ePOST compared to lPOST. Gower et al. (2002) observed that control of giant foxtail was improved to

≥90% when glyphosate application was delayed from 10 cm timing to 23 cm timing, while control of common lambsquarters and velvetleaf declined, and did not exceed 71% at these timings. While these results were attributed to re-infestation of giant foxtail after application

(Gower et al. 2002), such findings would suggest that giant foxtail is more sensitive to glyphosate at later application timings compared to other weed species. Similarly, Hartzler et al.

(2006) reported that there is interspecific sensitivity to glyphosate. Overall, several studies have documented interactions between weed size and the efficacy of systemic herbicides.

3.5 Factors Affecting Herbicide Efficacy Various biotic and abiotic factors contribute to the interactions between weed size/stage at the time of application and herbicide efficacy. Aside from spray deposition and coverage playing a major role, other factors during and shortly after application may be equally as important. One significant factor in uptake of any substance applied to plants is leaf cuticle thickness or robustness (Hess and Falk 1990; Hull 1975). Several environmental and morphological elements may contribute to cuticle composition and thickness, and thus affect uptake of herbicides (Hull 1975). In general, young, actively growing plants have thinner, less developed cuticles than older plants (Hull 1975; King and Oliver 1992; Steckel et al. 1997).

These thinner, more permeable cuticles typically allow for greater penetration of water-soluble herbicides, contributing to greater herbicide efficacy (Steckel et al. 1997). Additionally, the presence, density and hydrophilicity/hydrophobicity of trichomes on leaf surfaces greatly impact herbicide coverage, retention and absorption (Hess and Falk 1990; Hull 1975). Factors impacting absorption and penetration of 2,4-D have been extensively studied and reported in the literature

(Sargent and Blackman 1972). Sargent (1965) reported a decline in leaf permeability to 2,4-D correlated to leaf age, however it was acknowledged that environmental factors also have a significant influence on cuticle permeability. Sargent and Blackman (1972) determined that 2,4-

D penetration into leaves was related to light intensity, with increased light intensity corresponding to greater herbicide absorption and translocation. Subsequent research has confirmed that time of day and ambient air temperature at application play a role in weed control

25 with several herbicides (Stopps et al. 2013). It has also been reported that, generally, younger plants may be more sensitive to herbicides simply due to higher metabolic activity and translocation, assuming temperature and light are in excess (Agbakoba and Goodin 1969;

Coetzer et al. 2002).

Leaf orientation relative to time of day or, more importantly, stage of growth may play a role in herbicide interception, retention, absorption and subsequent herbicide activity

(Norsworthy et al. 1999; Stoller et al. 1975). In field studies, Stoller et al. (1975) documented improved control of older yellow nutsedge (Cyperus esculentus L.) with bentazon compared to younger plants, which was attributed to the fact that older yellow nutsedge had larger leaf blades, which leaned horizontal to the spray, whereas young plants had short, erect leaves oriented nearly vertical to the spray (Stoller et al. 1975). Additionally, Steckel et al. (1997) proposed that decreased control of 5 cm cocklebur relative to larger plants with glufosinate could be due to insufficient leaf area for herbicide interception. Such results can be viewed as the exception however, with the majority of published literature reporting the opposite relationship; better spray coverage is obtained with applications made to small, younger plants (Coetzer et al. 2002;

Krausz et al. 1996; Steckel et al. 1997). Numerous variables related to weed species, herbicide active ingredient and environment collectively influence the relationship between herbicide efficacy and weed size at application.

3.6 Overcoming the Influence of Weed Size on Herbicide Efficacy Several application modifications can be made to mitigate the general decline in control of large weeds with herbicides. In many cases, the decline in weed control as weed size at the time of application was increased could be minimized or overcome by increasing herbicide rate

(Edmund and York 1987; Johnson and Norsworthy 2014; Lee and Oliver 1982; Steckel et al.

1997). Control of johnsongrass, giant foxtail, redroot pigweed and velvetleaf with glufosinate applied at later growth stages was improved substantially when higher rates of the herbicide were used (Johnson and Norsworthy 2014; Steckel et al. 1997; Tharp and Kells 1999). Similar trends were reported with bentazon for the control of cocklebur 3 days after emergence (DAE) compared to 6-12 DAE (King and Oliver 1992). Lee and Oliver (1982) observed that control of

2-leaf cocklebur with acifluorfen was improved 17% when the rate was increased from 100 to

300 g ai ha-1; when applied at 4-leaf stage 1100 g ai ha-1 was required for comparable control

(Lee and Oliver 1982). Similarly, control of larger weeds with systemic herbicides including 2,4-

D, dicamba and glyphosate has been improved simply by increasing the rate of the herbicide

(Sellers et al. 2009; Krausz et al. 1996; Soltani et al. 2016; Tharp and Kells 1999). The control of larger weeds with contact and systemic herbicides is also improved with increased spray volume.

Krausz et al. (1996) reported a 55% increase in control of 20 cm-tall ivyleaf morningglory with

560 g ai ha-1 of the isopropylamine salt of glyphosate (480 g ai L-1) in 187, compared to 93 L ha-1 water volume. Lee and Oliver (1982) reported an increase in control of pitted morningglory with acifluorfen as spray volume was increased; however, control of hemp sesbania (Sesbania herbacea (Mill.) McVaugh) was optimized with the lowest tested water volume. The authors attributed these divergent results to high water volume causing improved spray coverage and retention by pitted morningglory, but increased herbicide runoff in hemp sesbania due to its waxy cuticle (Lee and Oliver 1982). In certain cases, the addition of various adjuvants has been shown to improve control of larger weeds, or those with thicker cuticles; it is known that surface- active adjuvants can increase herbicide retention and absorption into the treated leaf (Baur and

Aponte 2014). Hull et al. (1975) found that the addition of surfactant allowed mature leaves of mesquite which had higher amounts of epicuticular wax compared to young leaves, to be

27 adequately wetted. Hess and Falk (1990) reported that adjuvants can reduce the negative influence of unfavourable leaf topography and epicuticular wax, however acknowledged that such additives cannot reliably provide a uniform covering of herbicide across the leaf surface.

Baur and Aponte (2014) concluded that due to the complexity of interactions between specific active ingredients in adjuvants and herbicides, several factors must be considered when determining absorption potential. Such results further compound the challenge of obtaining optimal control of a range of weed species with a single POST herbicide application.

Although control of large weeds may be improved with a variety of modifications to application procedures, or with the formulation of specific herbicides, experimental results document the benefit of early-season weed control, highlighting the advantage of timely application of POST herbicides. Soltani et al. (2016) determined that corn yield was maximized with applications of glyphosate made to 10 cm weeds, with yield declining as applications were delayed to 20 and 30 cm timing. Gower et al. (2003) reported similar findings, with corn yield averaging 79% of the weed-free control where glyphosate application was delayed until weeds reached 30 cm. There are many studies in the literature that confirm the benefits of applying

POST contact and systemic herbicides to young, actively growing weeds less than 10-15 cm in height (Ateh and Harvey 1999; Hull 1970; Jordan et al. 1997; King and Oliver 1992; Lee and

Oliver 1982; Soltani et al. 2016). Where late emerging weeds are of concern, use of soil-applied residual herbicides, sequential POST applications, or integration of mechanical weed control have been proposed as viable options to maintain full-season weed control, while still facilitating optimal early application timing of the POST herbicide(s) (Coetzer et al. 2002; Steckel et al.

1997; Tharp and Kells 1999). Overall, despite efforts to mitigate the widely-observed decline in control of larger weeds with contact and systemic herbicides, as well as some noteworthy

28 exceptions to this relationship (Obrigawitch et al. 1990; Stoller et al. 1975), the benefits of early

POST applications are clear, both in optimizing herbicide efficacy as well as minimizing crop yield losses due to early-season weed interference.

Section IV - Glyphosate-Resistant Weeds and Canada Fleabane (Conyza canadensis (L.)

Cronq.)

4.1 Glyphosate: Discovery and Development Glyphosate, N-(phosphonomethylglycine) was developed over the course of several decades, and has been called a once-in-100-years herbicide. Glyphosate was first synthesized in

1950, although its potential as an herbicide was not realized until 1970 (Bader 1988; Franz et al.

1997). Subsequently, glyphosate was commercialized in 1974 by Monsanto Company (Dill et al.

2010). Glyphosate is a white, crystalline, acidic amino acid, with relatively low water solubility

(Franz et al. 1997). As such, glyphosate was formulated as water-soluble monoanionic salts, such as isopropylamine (IPA), trimethylsulfonium (TMS) or potassium (K) often with a surfactant, and originally used for preplant, post-emergence directed, and post-harvest weed control (Dill et al. 2010; Shaner 2014). Although glyphosate was discovered in the mid-20th century, it took several decades for its potential as an herbicide to be fully realized.

4.2 Glyphosate: Mode of Action Glyphosate is classified as a unique aromatic amino acid synthesis inhibitor. The mode of action of glyphosate is competitive inhibition of 5-enolpyruvylshikimate-3-phosphate synthase

(EPSPS). This enzyme, found in the shikimate biosynthetic pathway, is present only in plants and microorganisms, and is required for the biosynthesis of aromatic amino acids (Herrmann and

Weaver 1999). EPSPS catalyzes the synthesis of 5-enolpyruvylshikimate-3-phosphate from phosphoenolpyruvate (PEP) and shikimate 3-phosphate (S3P), a vital step in the shikimate

29 biosynthetic pathway which ultimately produces chorismate (Herrmann and Weaver 1999).

Chorismate is a necessary intermediate in the synthesis of the essential aromatic amino acids tryptophan, tyrosine and phenylalanine (Herrmann and Weaver 1999). This mechanism of

EPSPS inhibition is unique to glyphosate (Franz et al. 1997).

4.3 Glyphosate: Application, Uptake and Selectivity Glyphosate is a non-selective, POST, systemic herbicide which exhibits no residual herbicidal activity in the soil. In water, glyphosate salts readily dissociate to the glyphosate acid anion, and the salt cation (Franz et al. 1997). Negatively-charged glyphosate anions are rapidly adsorbed and bound tightly to organic matter constituents and metal ions in the soil (Franz et al.

1997). Glyphosate adsorption to soil takes place within one hour of application, and subsequent degradation by soil microorganisms is rapid; therefore, glyphosate provides no residual herbicidal activity in soil (Sprankle et al. 1975). Applied POST, glyphosate is rapidly absorbed by plant foliage (Dill et al. 2010). Glyphosate readily penetrates cell membranes and is extensively translocated both acropetally and basipetally to above and below-ground meristematic tissues via phloem sieve tubes (Franz et al. 1997). Given this rapid and extensive translocation pattern, glyphosate is highly efficacious, and controls over 300 annual, biennial and perennial grass and broadleaf weed species (Franz et al. 1997). This broad-spectrum activity, combined with low toxicity to non-target organisms and a rapid degradation pattern in soil, led to the widespread popularity of glyphosate as a non-selective, POST herbicide (Franz et al. 1997).

4.4 Glyphosate-Resistant Crops: Development Due to its lack of selectivity, the usefulness of glyphosate in agronomic crops was limited until the introduction of glyphosate-resistant (GR) crops, which rapidly expanded its utility (Dill et al 2010). Prior to 1996, glyphosate was commonly applied prior to planting in reduced and no- tillage systems, in a growing crop using spot/wick applicators, or in non-crop areas such as

30 ditches and right-of-ways (Feng et al. 2010). In the 1980-1990’s, due to widespread cases of weed biotypes resistant to existing herbicides including Group 2 acetolactate synthase (ALS)- inhibitors and Group 5 (PSII-inhibitors), there was interest in the development of GR crops (Holt and Lebaron 1990; Feng et al. 2010). Initial attempts to create GR crops focused on either isolation of cell lines which contained DNA encoding overexpression of the EPSPS gene, or employing glyphosate detoxification mechanisms in plants to confer resistance; both of these methods proved to be commercially unsuccessful (Dill 2005; Shah et al. 1986). Instead, the development of GR crops stemmed from the chance discovery of a glyphosate-insensitive

EPSPS enzyme in a soil Agrobacterium species surviving in wastewater from a glyphosate manufacturing facility (Barry et al. 1992; Dill 2005). The insensitive EPSPS gene from this

Agrobacterium tumafaciens strain CP4 (referred to as CP4 EPSPS) was subsequently used to develop several transgenic GR crops, including GR soybean, canola (Brassica spp.), corn, cotton

(Gossypium spp.), sugar beet (Beta vulgaris var. altissima) and alfalfa (Medicago sativa L.)

(Feng et al. 2010). The first commercially available GR crops were soybean (GTS line 40-3-2) and canola released in 1996, with corn and cotton following shortly thereafter (Beckie et al.

2006; Feng et al. 2010). The release of GR crops dramatically expanded the use pattern for glyphosate in the mid-1990’s.

4.5 Glyphosate-Resistant Crops: Adoption A confluence of several factors led to widespread adoption of GR crops, and dramatically increased the use of glyphosate. Of all GR crops, the most rapid and comprehensive adoption occurred with soybean (Benbrook 2016; Green and Castle 2010). In 1996, the year of introduction, just over 20% of total soybean hectares in the USA were planted to GR cultivars; by 2006 this share had surpassed 95% of total soybean hectares (Benbrook 2016). Although adoption was slower in Eastern Canada, over 60% of total soybean area was seeded to GR

31 cultivars by 2005, growing to 78% by 2017 (Beckie et al. 2006; Michael Reidy, Stratus Ag

Research, personal communication). There has also been widespread acceptance of GR corn and cotton. In 2014, corn, soybean and cotton collectively accounted for 80% of the total agricultural use of glyphosate in the USA (Benbrook 2016). Thirty-eight percent of US corn hectares planted in 2014 received at least one glyphosate application (National Agricultural Statistics Service,

2015). In contrast to soybean, 97% of the total corn area in Eastern Canada was planted to GR hybrids in 2017 (Michael Reidy, Stratus Ag Research, personal communication).

The extensive adoption of GR crops was driven by several interconnected factors. GR crops allowed farmers to apply glyphosate to field crops in place of herbicides which were often more expensive and controlled a narrower spectrum of weeds (Green and Castle 2010). In GR corn, glyphosate was seen as an effective replacement for atrazine, providing equal or better control of several weed species, particularly where triazine-resistant biotypes were prevalent

(Ferrell and Witt 2002). Similarly, in soybean, glyphosate can provide equal or better broad- spectrum weed control compared to commonly used herbicides such as imazethapyr, bentazon and metribuzin (Stewart et al. 2010). Furthermore, tank-mixing glyphosate with existing herbicides in soybean has been shown to provide no benefit in terms of weed control relative to glyphosate alone (Walsh et al. 2014). Glyphosate eliminated the need for prescription herbicide combinations dependent on the crop type or weed spectrum and provided growers with a simplified weed management system (Duke and Powles 2009; Green 2012). Given the low acute toxicity of glyphosate, GR crop systems were recognized to be more environmentally friendly than the conventional weed management systems they replaced (Mamy et al. 2010; Nelson and

Bullock 2003). The ability to control weeds POST with glyphosate facilitated a reduction in the requirement for tillage as a means of weed control, encouraging the adoption of no-till crop

32 production (Barrows et al. 2014; Green 2012). This movement away from tillage reduced the energy use and negative environmental impacts, such as soil erosion, associated with conventional-tillage crop production (Duke and Powles 2009). Of paramount concern to growers, GR crops have increased farm profitability, due to higher yields as a result of improved weed control, as well as reduced fuel and labour costs associated with a reduction in tillage

(Barrows et al. 2014; Dill 2005; Duke and Powles 2009; Green 2012). Over a short time, several agronomic, environmental and economic factors drove GR crops to become the most rapidly- adopted technology in the history of agriculture, concomitantly increasing the global use of glyphosate (Green 2012; Johnson et al. 2009).

4.6 Glyphosate-Resistant Weeds Due in part to the overreliance on glyphosate for weed control, GR weed biotypes have proliferated, and become a serious management challenge. Herbicide resistance (HR) can generally be defined as the inherited ability of a plant biotype to survive and reproduce following exposure to a previously lethal dose of herbicide (Weed Science Society of America 1998).

Herbicide resistance is not a new occurrence; however, the spread of HR with respect to glyphosate is of great concern given its widespread use (Yuan et al. 2006). Although HR biotypes are naturally-occurring, their proliferation is dependent on selection pressure induced by the repeated use of a specific herbicide (Johnson et al. 2009; Reddy and Norsworthy 2010).

Following the initial introduction of GR crops in 1996, one or two applications of glyphosate alone proved to be an economical and effective weed management strategy in GR crops

(Benbrook 2016). In many cases where GR corn and soybean were grown in rotation, growers applied glyphosate alone each season, often repeatedly, and over wide geographic areas (Duke and Powles 2009; Green 2012). This shift to a weed management system which relied nearly exclusively on glyphosate for weed control constituted an intense selection pressure for GR weed

33 biotypes (Duke and Powles 2009). Glyphosate-resistance was first confirmed in 1998, in an

Australian population of rigid ryegrass (Lolium rigidum Gaud.) (Powles et al. 1998). Just three years later, GR Canada fleabane was confirmed in Delaware, marking the first documented GR weed biotype within a GR crop (VanGessel 2001). Since then, 43 species globally have been confirmed resistant to glyphosate; 6 of these species (common ragweed, giant ragweed, Canada fleabane, kochia, birdsrape mustard (Brassica rapa B. campestris) and waterhemp (Amaranthus tuberculatus (Moq.) J.D. Sauer) occur in Canada (Heap 2018). Among the most troublesome GR species in the USA corn-soybean producing regions, are dioecious Amaranthus spp. (including waterhemp and Palmer amaranth), Canada fleabane, and Ambrosia spp., ─many of which have evolved resistance to multiple herbicide modes of action (Beckie 2011; Heap 2018). Glyphosate- resistant weed populations unquestionably require growers to modify their weed management strategies, increasing the cost for weed control and in some cases, necessitating drastic steps such as hand weeding (Green and Owen 2011). Attributable to an overreliance on glyphosate, GR weed biotypes have become an arduous management challenge for crop producers.

4.7 Mechanisms of Glyphosate-Resistance: Target Site/Non-Target Site Resistance Mechanisms which confer resistance to glyphosate can be broadly categorized into target site resistance (TSR) and non-target site resistance (non-TSR). Target site resistance prevents glyphosate from eliciting its normal activity through specific modifications at the EPSPS target site (Beckie 2011). Various alterations in the EPSPS amino acid sequence confer glyphosate resistance (Bianco de Carvalho and Alves 2012). A Pro-Ser mutation at position 106 in EPSPS of GR Malaysian goosegrass (Eleusine indica (L.) Gaertn.) was first documented by Baerson et al. (2002); this mutation was subsequently discovered in GR populations of Italian ryegrass

(Lolium perenne L. spp. multiflorum) and tall waterhemp (Bell et al. 2013; Jasieniuk et al. 2008;

Nandula et al. 2013). Other amino acid substitutions at position 106, including Pro-Thr, have

34 also been documented to confer glyphosate resistance (Wakelin and Preston 2006). In 2010, a second TSR mechanism was discovered in GR Palmer amaranth, in which resistance resulted from EPSPS gene amplification and subsequent over-expression of the EPSPS target site (Gaines et al. 2010). EPSPS gene amplification has since been discovered in GR populations of Italian ryegrass, tall waterhemp and kochia (Jugulam et al. 2014; Salas et al. 2015; Lorentz et al. 2014;

Wiersma et al. 2015). In other cases of GR weeds, non-TSR mechanisms exist which prevent glyphosate from reaching the EPSPS target site at a lethal dose (Beckie 2011). Non-target site resistance to glyphosate includes mechanisms which reduce herbicide uptake or translocation, or lead to sequestration of the herbicide away from the target site (Huffman et al. 2016). In samples collected from Arkansas, Riar et al. (2011) reported a 28% reduction in glyphosate translocation from the treated leaves of a GR johnsongrass biotype compared to the susceptible biotype. Vila-

Aiub et al. (2011) also observed a significant reduction in glyphosate absorption and translocation in a GR johnsongrass population from Argentina. A similar exclusion mechanism, reduced cellular uptake of glyphosate, has since been reported in GR Palmer amaranth and tall waterhemp (Sammons and Gaines 2014). Conversely, in Canada fleabane, one of the most prevalent GR species, resistance is typically conferred by selective sequestration of glyphosate in the vacuole of GR biotypes (Ge et al. 2010). Although several TSR and non-TSR mechanisms exist, any single mechanism typically confers relatively low resistance to glyphosate, and thus multiple mechanisms of resistance often occur within a single GR biotype (Sammons and Gaines

2014). Both TSR and non-TSR mechanisms have been reported to act in tandem, conferring high level glyphosate resistance in populations of common waterhemp and Digitaria spp. (Bianco de

Carvalho and Alves 2012; Nandula et al. 2013; Chatham et al. 2015). Therefore, although

35 distinctions can be made between TSR and non-TSR mechanisms, overlap of both mechanisms commonly occurs within individual GR biotypes.

4.8 Canada fleabane: Biology Canada fleabane is a member of the Asteraceae family, and exhibits an extended emergence pattern, high fecundity and wind-dispersed seeds. Canada fleabane is a winter or summer annual plant native to North America and is found in every Canadian province except

Newfoundland (Weaver 2001). Typically, Canada fleabane germinates in late summer and early autumn, forming overwintering rosettes which bolt and produce numerous, elongated flowering branches the following summer; however, a portion also germinates in spring and completes its life cycle in a single season (Buhler and Owen 1997; Steckel et al. 2010; Weaver 2001). The ability to form an overwintering rosette makes Canada fleabane highly competitive relative to spring-germinated weed species, because of the potential for significant vegetative growth in early spring (Buhler and Owen 1997). Canada fleabane is predominately self-pollinated, and seed production in southern Canada typically peaks in August and September (Huang et al. 2015;

Weaver 2001). Canada fleabane has few specific climactic requirements; its emergence patterns are highly variable and poorly correlated to photoperiod, soil or air temperature, although germination only occurs from depths of less than 0.5 cm (Main et al. 2006; Nandula et al. 2006;

Weaver 2001). Regehr and Bazzaz (1979), however, observed greater fall germination when soil moisture was high. Canada fleabane is a highly-prolific seed producer, with seed numbers being proportional to plant height (Regehr and Bazzaz 1979). A 1.5 m -tall plant can produce over 230

000 seeds which are widely wind-disseminated, facilitated by small pappi attached to the seed

(Weaver 2001). Canada fleabane seeds have been collected from heights of 140 m above the ground, in the Planetary Boundary Layer, where they can be dispersed to distances greater than

500 km from the mother plant (Shields et al. 2006). This highly effective dispersal mechanism,

36 coupled with its inordinate fecundity and extensive emergence pattern, has made Canada fleabane a highly successful ruderal Asteraceae weed species.

4.9 Glyphosate-Resistant Canada fleabane: Resistance Mechanisms and Distribution Glyphosate-resistant biotypes of Canada fleabane have expanded across a wide geographic area of the USA and Ontario. Due to its ability to readily germinate at or near the soil surface, Canada fleabane is highly successful in no-till production systems (Davis et al. 2009;

Main et al. 2004; Nandula et al. 2006). The adoption of GR crops facilitated the widespread use of no-till, which in turn increased the use of glyphosate and created immense selection pressure for GR biotypes (Barrows et al. 2014; Nandula et al. 2006). Glyphosate-resistant Canada fleabane was first reported in 2001 in Delaware, and subsequently in Tennessee in 2003 (Mueller et al. 2003; VanGessel 2001). The primary mechanism of glyphosate resistance in most GR

Canada fleabane is vacuolar sequestration (Ge et al. 2010). This mechanism is temperature- dependent, with decreased sequestration and increased glyphosate efficacy observed under cooler temperatures (Ge et al. 2010). Populations of GR fleabane have also been discovered which exhibit enhanced metabolism of glyphosate in addition to vacuolar sequestration

(Gonzalez-Torralva et al. 2012), and in Ontario biotypes, Page et al. (2018) recently confirmed the presence of target site resistance mechanisms (Pro-Ser mutation at position 106) in addition to non-target site mechanisms. Glyphosate-resistant Canada fleabane has spread rapidly to become one of the most pervasive GR weed species globally (Beckie 2011). Weather patterns and wind can distribute Canada fleabane seeds and pollen across a vast spatial area, contributing to the rapid spread of resistant biotypes (Barrows et al. 2014; Huang et al. 2015; Wang et al.

2017). Glyphosate-resistant Canada fleabane is present on five continents and is found throughout much of the USA (Ge et al. 2011). Glyphosate-resistant Canada fleabane was first discovered in Ontario in 2010, in Essex County (Byker et al. 2013a). By 2015, GR Canada

37 fleabane had spread to 30 counties throughout Ontario, spanning from the Michigan to Quebec borders (Budd et al. 2018). Furthermore, many of these populations exhibit multiple herbicide resistance (MR) to glyphosate and acetolactate synthase (ALS)-inhibitors (Budd et al. 2018).

Due to several compounding factors, the range of GR/MR Canada fleabane has expanded rapidly across several US states and Ontario.

4.10 Herbicide-Resistant Canada fleabane: Agronomic Implications and Management Strategies in Corn Interference from GR Canada fleabane in GR crops can cause significant yield loss, however some management strategies have proved to be effective. At densities of 100-200 plants m-2, Canada fleabane can reduce soybean yields by up to 90% (Bruce and Kells 1990). In

Ontario, interference from GR Canada fleabane left uncontrolled in corn reduced yield by 69%

(Ford et al. 2014). Despite the efficiencies and simplicity afforded by a glyphosate-alone weed management system, effective control of GR Canada fleabane is critical to avoid such yield losses. One of the most common management strategies adopted by growers is the use of herbicides with alternative MOA to control GR Canada fleabane (Scott and VanGessel 2007). In corn, several soil-applied herbicides, including S-metolachlor + flumetsulam + clopyralid, dicamba/atrazine, mesotrione + atrazine, saflufenacil/dimethenamid-P and dicamba provide

>90% control of GR Canada fleabane (Brown et al. 2016; Ford et al. 2014). Postemergence herbicide options that provide effective GR Canada fleabane control (>90%) in corn are currently limited to dicamba/atrazine and bromoxynil + atrazine (Mahoney et al. 2017).

However, the introduction of new HR traits in corn such as those imparting resistance to 2,4-D will increase the number of viable POST herbicide options available to growers (Ford et al.

2014). Furthermore, the prevalence of Canada fleabane biotypes with resistance to multiple herbicide modes of action in addition to glyphosate illustrates the importance of integrating non-

38 chemical management strategies (Byker et al. 2013a). Despite the environmental advantages to no-till crop production, tillage is an effective management strategy for GR Canada fleabane

(Nandula et al. 2006; Steckel et al. 2010). Conversely, the presence of grain corn residue left to overwinter can reduce Canada fleabane emergence up to 79% compared to bare ground, due to the shading and interception of wind-blown seeds by the residue (Main et al. 2006). Such results suggest the benefit of tillage for GR Canada fleabane control may be crop-dependent, and specific to certain environments or timings. In some instances, crop rotation has reduced Canada fleabane emergence and depleted the seedbank within the soil; however, other studies have shown no effect of rotation compared to a monoculture (Davis et al. 2009; Davis et al. 2007).

Overall, these inconsistencies underscore the importance of integrating multiple weed control practices, both chemical and non-chemical, in order to enhance the durability of existing weed management systems, prevent further spread of GR/MR biotypes and avoid crop yield loss caused by GR/MR Canada fleabane interference.

Section V - Hypotheses and Objectives

5.1 Hypotheses

1. Tolpyralate (30 g ai ha-1) + atrazine (1000 g ai ha-1) will provide ≥80% control of the

common annual grass and broadleaf weeds in corn in Ontario.

2. The addition of atrazine to tolpyralate will improve the control of common annual grass

and broadleaf weeds.

3. Weed control with tolpyralate + atrazine will decline as time of application is delayed and

weed size increases.

4. Tolpyralate (40 g ai ha-1) + atrazine (1000 g ai ha-1) will cause ≤10% crop injury in corn.

5.2 Objectives

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

atrazine applied POST, for control of common annual grass and broadleaf weeds in corn

2. To determine the relative efficacy of tolpyralate + atrazine, topramezone + atrazine and

mesotrione + atrazine applied POST for the control of common annual grass and

broadleaf weeds in corn

3. To establish the BED of tolpyralate and tolpyralate + atrazine for control of GR/MR

Canada fleabane in corn.

4. To establish the BED of tolpyralate and tolpyralate + atrazine for control of MR

waterhemp in corn.

5. To determine the effect of application timing and weed size on the efficacy of tolpyralate

+ atrazine.

6. To determine the effect of herbicide rate, application timing and hybrid selection on the

tolerance of corn to tolpyralate + atrazine.

Chapter 2: Tolpyralate Efficacy: Part I. Biologically-effective dose of tolpyralate for control of annual grass and broadleaf weeds in corn

2.1 Abstract

Tolpyralate is a new 4-hydroxyphenyl-pyruvate dioxygenase (HPPD)-inhibiting herbicide for postemergence (POST) weed management in corn; however, there is limited information regarding its efficacy. Six field studies were conducted in Ontario, Canada over three years (2015 to 2017), to determine the biologically-effective dose of tolpyralate for the control of eight annual weed species. Tolpyralate was applied POST at six doses from 3.75 to

120 g a.i. ha-1 and tank-mixed at a 1:33.3 ratio with atrazine at six doses from 125 to 4000 g ai ha-1. Regression analysis was performed to determine the effective dose (ED) of tolpyralate, and tolpyralate + atrazine, required to achieve 50, 80 or 90% control of eight weed species, 1, 2, 4, and 8 weeks after application (WAA). The ED of tolpyralate for 90% control (ED90) of velvetleaf (Abutilon theophrasti Medik.) [ABUTH], common lambsquarters (Chenopodium album L.) [CHEAL], common ragweed (Ambrosia artemisiifolia L.) [AMBEL], redroot/green pigweed (Amaranthus powelli S. Watson/Amaranthus retroflexus L.) [AMASS] and green foxtail

(Setaria viridis (L.) P. Beauv.) [SETVI] at 8 WAA was ≤15.5 g ha-1; however, tolpyralate alone did not provide 90% control of wild mustard (Sinapis arvensis L.) [SINAR], barnyardgrass

(Echinochloa crus-galli (L.) P. Beauv.) [ECHCG] or ladysthumb (Persicaria maculosa Gray)

[POLPE] 8 WAA at any dose evaluated in this study. In contrast, the ED90 for all species in this study with tolpyralate + atrazine was ≤13.1 + 436 g ha-1, indicating that tolpyralate plus atrazine can be highly efficacious at low field doses.

2.2 Introduction

Competition from weeds represents one of the principal factors affecting corn grain yield.

Herbicides are regarded as an effective and economical form of weed management, and are applied to over 95% of corn hectares in North America (Gianessi and Reigner 2007).

Development of the critical weed-free period (CWFP) in corn has determined that corn yield loss due to weed interference is most probable during early growth stages, prior to V8 (Hall et al.

1992). Introduction of selective herbicides and glyphosate-resistant (GR) corn hybrids facilitated timely control of weeds during the CWFP with POST herbicide applications; however, diversity of chemical weed management programs has generally declined (Duke and Powles 2009).

Evolving weed management challenges, including those associated with managing GR weed biotypes, have spurred renewed interest in the development of new herbicide active ingredients to broaden the number of available herbicides.

Herbicides which inhibit the 4-hydroxyphenyl-pyruvate dioxygenase (HPPD) enzyme in susceptible plants impede the biosynthesis of plastoquinone (PQ) and α-tocopherols, thereby inhibiting biosynthesis of carotenoid pigments (Hawkes 2012; Matsumoto et al. 2002; Shulz et al. 1993). Carotenoids act both as accessory light harvesting pigments and quenchers of high- energy triplet chlorophyll (Hawkes 2012). Carotenoid depletion by way of HPPD inhibition leaves chlorophyll susceptible to oxidative degradation by reactive oxygen species (ROS), resulting in white bleaching of plant tissues, protein and lipid destruction, and subsequent plant death (Ahrens et al. 2013; Hawkes 2012). The HPPD-inhibitors include triketones, isoxazoles and pyrazolones, and are currently used for weed management in corn, rice and cereals (Hawkes

2012).

Photosystem-II (PSII)-inhibiting herbicides, including atrazine, are commonly tank- mixed with HPPD-inhibitors because of their complementary mechanisms of action (Hess 2000).

The HPPD-inhibitors are presumed to increase efficiency of atrazine binding on the D1 protein of PSII via depletion of PQ, while concurrently intensifying cell membrane destruction by subsequently produced ROS, due to their inhibition of antioxidant biosynthesis (Armel et al.

2005; Kim et al. 1999). The addition of atrazine to mesotrione or tembotrione has been documented to induce herbicide synergy in some instances (Abendroth et al. 2006; Armel et al.

2007; Kohrt and Sprague 2017a); however, additive effects are more widely reported with topramezone plus atrazine, which suggests that the benefit of atrazine addition is specific to the

HPPD-inhibitor and weed species (Kohrt and Sprague 2017a).

Tolpyralate is a new pyrazolone-type HPPD-inhibiting herbicide which has recently been registered in the United States and Canada for use in corn (Health Canada 2018). Tolpyralate has relatively low water solubility (26.5 mg L-1), low potential for volatilization, and has not been found to pose significant risk to humans or the environment (Health Canada 2017). Applied

POST at 30 to 40 g ha-1 alone or in combination with atrazine at 560 to 1000 g ha-1, tolpyralate has been reported to control a range of annual grass and broadleaf weed species and exhibit selectivity in all types of corn (Kikugawa et al. 2015). Currently, there is limited information in the published literature on the use of tolpyralate in North America and globally. Therefore, the objective of this research was to determine the efficacy of tolpyralate in corn for the control of several weed species across environments. The results of this research are presented in two papers.

The purpose of this paper was to develop weed species-specific dose-response curves, with tolpyralate alone or tank-mixed with atrazine, to ascertain a biologically-effective dose

(BED) of tolpyralate and tolpyralate plus atrazine for several weed species. The subsequent companion paper: 1) examines tolpyralate efficacy applied alone or in combination with atrazine to determine the benefit of atrazine addition, and 2) compares the efficacy and selectivity of tolpyralate with existing HPPD-inhibiting herbicides.

2.3 Materials and Methods

Experimental Methods Six field experiments were conducted over a 3 year period (2015 to 2017) near

Ridgetown and Exeter, Ontario, Canada on field research sites managed under corn-soybean

(Glycine max (L.) Merr.)-winter wheat (Triticum aestivum L.) rotations. Seedbed preparation consisted of fall moldboard plowing, followed by two passes with a field cultivator with rolling basket harrows in the spring. Sites were fertilized in accordance with soil test results and crop requirements each year prior to planting. No herbicides aside from treatments described herein were applied to the trial sites during the years of study.

Each field experiment was organized as a randomized complete block with four replications. Plots were 3 m wide (4 rows of corn spaced 76 cm apart) and 8 or 10 m long at

Ridgetown and Exeter, respectively. Glyphosate-resistant corn was seeded to a depth of 4 to 5 cm, at 78,000 to 82,000 seeds ha-1. Hybrids were selected for each site based on geographic suitability and were DKC42-42RIB and DKC53-56RIB (Monsanto Co., St. Louis, MO) at Exeter and Ridgetown, respectively. Information pertaining to soil characteristics, planting/harvest dates and spray application dates are presented in further detail in Table 2.1.

Herbicide treatments were applied using a CO2-pressurized backpack sprayer, calibrated to deliver 187 L ha-1 at 240 kPa through four ULD 12002 nozzles (Pentair, New Brighton, MN)

45 spaced 50 cm apart. Applications were made POST when native weed populations in the non- treated control plots reached an average of 10 cm in height. Crop stage at time of application ranged from V4 to V6. Weed-free control plots were maintained free of weeds for the entirety of the trial period with S-metolachlor (1600 g ai ha-1) plus atrazine (1280 g ai ha-1) plus mesotrione

(140 g ai ha-1) (Lumax® EZ, Syngenta Canada Inc., Guelph, ON) applied PRE, followed by glyphosate (900 g ae ha-1) applied POST and subsequent hand weeding as needed.

Treatments consisted of tolpyralate at 3.75, 7.5, 15, 30, 60 and 120 g ha-1 and a tank mixture of tolpyralate plus atrazine at a 1:33.3 ratio, at doses of 3.75 + 125, 7.5 + 250, 15 + 500,

30 + 1000, 60 + 2000 and 120 + 4000 ha-1. Adjuvants were included in accordance with herbicide manufacturer recommendations. All tolpyralate applications included methylated seed oil (MSO Concentrate®, Loveland Products Inc., Loveland, CO) at 0.50% v/v and 28% N urea ammonium nitrate (UAN) (2.50% v/v).

Crop injury was evaluated at 1, 2 and 4 wk after application (WAA) on a scale of 0 to

100 with 0 representing no injury and 100 representing complete plant death. Visible weed control was assessed 1, 2, 4 and 8 WAA, where control of each species was evaluated relative to the non-treated control plot and assigned a value from 0 indicating no control, to 100, indicating complete control. Following the final weed control assessment at 8 WAA, density and dry weight of each weed species were determined by counting the number of weeds within two randomly-placed 0.5 m2 quadrats per plot. The weeds were cut at the soil surface, separated by species into paper bags, dried at 60 C to constant moisture, and the dry weight recorded.

At maturity, the centre two rows of each plot were harvested with a small plot combine.

Moisture content and grain weight was recorded, and grain yields were calculated and adjusted to 15.5% moisture for analysis.

Statistical Analysis Percent control of each weed species at 1, 2, 4 and 8 WAA was regressed against the dose of tolpyralate alone, and the combined dose of tolpyralate + atrazine using NLIN procedures in SAS v. 9.4 (SAS Institute, Cary, NC) with one of two exponential to a maximum equations. Where tolpyralate was applied alone, Equation 1 was fit to the data. Where tolpyralate

+ atrazine were applied, Equation 2 was used due to a better fit, as determined by Pseudo-R2 values and standard errors associated with parameter estimates of each model. Yield data was expressed as a percentage of the yield of weed-free control plots within each replication, and regressed against tolpyralate dose (using Equation 1), and tolpyralate + atrazine dose (using

Equation 2). Weed density (plants m-2) and dry biomass (g dry matter m-2) were regressed against tolpyralate and tolpyralate + atrazine dose using an inverse exponential equation

(Equation 3). Predicted values generated from regression analyses were used to compute the effective dose (ED) of tolpyralate and tolpyralate + atrazine required to provide 50, 80 and 90% control of each weed species at each assessment timing, and a 50, 80 or 90% reduction in weed density/dry weight. Where the predicted value could not be computed or was beyond the dosage range used in this study, it is expressed as a dash (─) in tables. The following equations were used for non-linear regression analysis.

Equation 1 – Exponential to a Maximum

y = a – b (e-c*dose)

Where: y = response parameter a = upper asymptote b = magnitude c = slope

Equation 2 – Exponential to a Maximum Alternate

y = a – c (bdose)

Where: y = response parameter a = upper asymptote b = slope c = magnitude

Equation 3 – Inverse Exponential y = a + b(-c*dose)

Where:

y = response parameter

a = lower asymptote

b = reduction in y from intercept to asymptote

c = slope

2.4 Results and Discussion

Weed Control

The eight weed species analyzed in this study were naturally-occurring at each trial site, and reflect typical native weed populations encountered in corn production systems in southwestern Ontario, Canada. Four of these species were ranked among the top ten most troublesome weed species by Ontario farmers in a 2016 opinion poll conducted by Bilyea

(2016). Broadleaf weed species included common lambsquarters (average density 14 plants m-2), velvetleaf (average density 5 plants m-2), common ragweed (average density 50 plants m-2), ladysthumb (average density 7 plants m-2), wild mustard (average density 20 plants m-2), and pigweed species [AMASS] (average density 14 plants m-2). Pigweed species grouping was used because sites comprised a heterogeneous population of green pigweed and redroot pigweed, which have similar morphology and exhibit the potential to hybridize with one another (Weaver

2009). Grass weed species included green foxtail (average density 17 plants m-2) and barnyardgrass (average density 38 plants m-2).

There was interspecific variation in sensitivity to tolpyralate at each assessment timing as indicated by predicted ED values; however, control generally improved with increasing herbicide dose. Based on regression analysis, tolpyralate alone at the tested doses did not provide ≥80% control of any species in this study 1 WAA (data not presented). Weed injury symptoms 1 WAA consisted of bleaching, stunting and slight leaf necrosis. At 1 WAA, 50% control of common lambsquarters, velvetleaf, pigweed species, common ragweed, green foxtail, barnyardgrass and ladysthumb was recorded with tolpyralate rates of 2.8 to 6.4 g ha-1; however, wild mustard was less sensitive and required 25.5 g ha-1 to achieve equivalent control. Tolpyralate efficacy 1 WAA

49 has not been previously reported; however, these results are consistent with experiments using other HPPD-inhibitors. Woodyard et al. (2009) reported 52 to 68% control of common lambsquarters and 53 to 75% control of waterhemp (Amaranthus tuberculatus (Moq.) J. D.

Sauer) with mesotrione (105 g ha-1) 10 days after application (DAA).

At 1 WAA, the addition of atrazine to tolpyralate improved control of all species (data not presented). These results are similar to Woodyard et al. (2009), who found that the addition of atrazine to mesotrione increased control of common lambsquarters and common waterhemp

10 DAA relative to mesotrione applied alone. Similarly, Abendroth et al. (2006) observed greater leaf necrosis in Palmer amaranth (Amaranthus palmeri S. Wats) and velvetleaf with mesotrione

+ atrazine compared to mesotrione alone. At 1 WAA, all eight weed species were controlled

80% with tolpyralate + atrazine at doses 10.4 + 345.9 g ha-1, respectively or less; velvetleaf, pigweed species, common lambsquarters, barnyardgrass and common ragweed showed greater sensitivity to tolpyralate + atrazine compared to ladysthumb, green foxtail and wild mustard. At

1 WAA, the ED90 of tolpyralate + atrazine, for common lambsquarters, velvetleaf, pigweed species and common ragweed was 10.4 + 347.8 g ha-1 or less.

At 2 WAA, common lambsquarters, velvetleaf, pigweed species, common ragweed, green foxtail, barnyardgrass and ladysthumb were more sensitive to tolpyralate alone compared with wild mustard (Equation 1; Table 2.2). Regression analysis indicated that common lambsquarters, velvetleaf and pigweed species were controlled 80% with tolpyralate alone at 4.4,

4.5 and 4.5 g ha-1, respectively, while common ragweed required 10.6 g ha-1 for equivalent control. At 2 WAA, regression analysis could not estimate the tolpyralate dose required for 80% control of four species (green foxtail, barnyardgrass, ladysthumb and wild mustard) and no dose provided ≥90% control. At 2 WAA, when atrazine was added to tolpyralate, 90% control of all

50 species was achieved with doses of 3.6 + 121 to 12.4 + 412.9 g ha-1 of tolpyralate + atrazine.

Ladysthumb, green foxtail, wild mustard and barnyardgrass required comparatively higher doses of tolpyralate + atrazine to achieve 90% control compared to other species; common lambsquarters and velvetleaf had lowest ED values to achieve the same level of control.

Weed control with tolpyralate applied alone improved considerably from 2 to 4 WAA. At

4 WAA, 90% control of common lambsquarters, velvetleaf, pigweed species and common ragweed was achieved with 3.4, 3.8, 6.9 and 8.4 g ha-1 tolpyralate, respectively (Table 2.3).

Previous research by Kohrt and Sprague (2017a) and Tonks et al. (2015) investigated tolpyralate efficacy 3 to 4 WAA at 30 to 40 g ha-1. Kohrt and Sprague (2017a) reported 96% control of atrazine-resistant Palmer amaranth with tolpyralate (40 g ha-1) 3 WAA. In contrast to the results from this study, Tonks et al. (2015) reported that on average, tolpyralate (30 g ha-1) controlled velvetleaf, common ragweed, Amaranthus spp. and common lambsquarters <90% 30 d after application. At 4 WAA, green foxtail was also controlled 90% in this study; however, a comparatively higher dose of tolpyralate (29.6 g ha-1) was required for equivalent control of this species compared with common lambsquarters, velvetleaf, pigweed species and common ragweed. Consistent with these results, Tonks et al. (2015) reported 91% control of green foxtail

30 d after application with tolpyralate at 30 g ha-1. Tolpyralate alone provided <90% control of barnyardgrass, and <80% control of ladysthumb and wild mustard 4 WAA. However, tolpyralate

+ atrazine at doses of 13.1 + 436.7 g ha-1 or less provided 90% control of all weed species evaluated at the same timing. At 4 WAA, differences in the tolpyralate and tolpyralate + atrazine

-1 ED90 for common lambsquarters, velvetleaf and pigweed species were less than 0.4 g ha ,

-1 however the ED90 for green foxtail was reduced from 29.6 to 9.6 g ha when atrazine was included.

At 8 WAA, control of wild mustard and ladysthumb with tolpyralate alone at the doses evaluated in this study was less than 80%. Wild mustard and ladysthumb were not adequately controlled with tolpyralate alone and they recovered from injury and resumed growth. Previous research has not investigated tolpyralate efficacy on either of these species, however Pannacci and Covarelli (2009) found that mesotrione applied alone did not provide over 90% control of ladysthumb based on regression analysis. Control of green foxtail improved from 4 WAA to 8

WAA. At 8 WAA, tolpyralate controlled common lambsquarters, velvetleaf, pigweed species, common ragweed and green foxtail 90% at predicted doses of 15.5 g ha-1 or less (Table 2.4).

Tolpyralate + atrazine at doses of 13.1 + 436 g ha-1 or less gave 90% control of all of the weed species evaluated in this study. Topramezone + atrazine (12.5 + 500 g ha-1) provides control of wild mustard and ladysthumb (Anonymous 2016), indicating either a difference in topramezone and tolpyralate activity or that control of these species in the current study is relative to the dose of atrazine. Consistent with results from previous assessment timings, common lambsquarters, velvetleaf, pigweed species and common ragweed could be controlled 90% with lower predicted tolpyralate doses than green foxtail, barnyardgrass, wild mustard and ladysthumb. In all species, the predicted tolpyralate dose for 50, 80 or 90% control was lower when applied with atrazine compared to tolpyralate applied alone. Similar results were reported with mesotrione by Hugie et al. (2008), who found that a lower dose of mesotrione was required for control of redroot pigweed when applied with atrazine compared to mesotrione applied alone.

There was variation in density of each weed species within trial sites which was reflective of natural species composition and interspecific competition within plots; however, density and biomass data generally reflected control assessments 8 WAA. Regression analyses provided a better fit to the data when conducted on species with higher natural densities (>20 m-2) within

52 trial sites or replications compared to those with lower natural densities (<5 m-2). Common ragweed, common lambsquarters, pigweed species, wild mustard and green foxtail were generally more numerous within trial areas compared to velvetleaf and ladysthumb. Tolpyralate applied alone at doses 17.6 g ha-1 or less provided a 50% reduction in density of all species except ladysthumb and wild mustard (Table 2.5). Common ragweed and green foxtail were the only species where density could be reduced by 80% with tolpyralate applied alone, while a 90% reduction in common ragweed density was achieved with tolpyralate at 20.6 g ha-1. Conversely, biomass of all species could be reduced at least 80% with tolpyralate alone (Table 2.6).

Tolpyralate at predicted doses of 2.7, 3.5, 3.8, 10.3 and 13.7 g ha-1 could reduce biomass by 90% for common lambsquarters, velvetleaf, common ragweed, green foxtail and barnyardgrass, respectively. In many cases where tolpyralate was applied alone, weeds became severely necrotic by 8 WAA but were still present in plots and therefore recorded, thus contributing to inconsistencies reflected in density and biomass data within species.

Consistent with control assessments, tolpyralate applied in combination with atrazine provided a higher level of weed control and therefore more consistent reduction in density and dry biomass than tolpyralate applied alone. Density of all species could be reduced 80% with tolpyralate + atrazine at doses of 44.7 + 1488.2 g ha-1 or less (Table 2.5). Tolpyralate + atrazine at doses 5.8 + 192.6 to 33 + 1100 g ha-1 provided a 90% reduction in density of common lambsquarters, common ragweed, green foxtail, barnyardgrass and wild mustard; however, ED90 values for reduction in velvetleaf and pigweed density could not be computed. Biomass of all species with the exception of ladysthumb could be reduced 90% with tolpyralate + atrazine at doses of 13.3 + 441.6 g ha-1 or less.

Yield and Crop Injury Crop injury with all treatments evaluated in this study was <10% and therefore considered commercially acceptable (data not presented). Where phytotoxicity did occur, injury symptoms consisted of minor leaf speckling, slight chlorosis or marginal necrosis of leaves exposed at the time of application. Symptoms were only observed in plots where tolpyralate + atrazine were applied at rates of 60 + 1000 g ha-1, or higher (data not presented).

Corn grain yields varied by year and location but were reflective of overall weed control and ranged from 0.92 t ha-1 in non-treated plots to 15.3 t ha-1 in weed-free control plots (data not presented). Tolpyralate applied alone at doses of 0.3 and 4.7 g ha-1 could maintain 50 and 80% of the yield obtained in the weed-free control plots. However, weed control was not sufficient to avoid 10% yield loss with any of the doses of tolpyralate alone based on regression analyses

(Table 2.7). This yield loss can likely be attributed to inadequate control of wild mustard with tolpyralate applied alone. Conversely, tolpyralate + atrazine at 5 + 167.8 g ha-1 was sufficient to maintain 90% of the yield obtained in the weed-free control plots. Corn is particularly susceptible to yield loss due to weed interference during emergence and early vegetative growth stages (Hall et al. 1992, Page et al. 2009). Therefore, despite complete weed control with POST applications of tolpyralate + atrazine in these studies, some level of yield loss may have been incurred due to early season weed interference prior to herbicide application. Future research could investigate the benefits of POST tolpyralate applications following application of a pre- emergence herbicide in order to mitigate this risk.

2.5 Conclusions

This research indicates that there are species-specific differences in weed sensitivity to tolpyralate. Based on predicted values calculated from regression analyses, common

54 lambsquarters, velvetleaf, pigweed species and common ragweed were controlled at least 90% with tolpyralate alone at doses below the current label rate range of 30 to 40 g ha-1. Conversely, the BED of tolpyralate for 90% control of ladysthumb and wild mustard was beyond those used in this study when tolpyralate was applied alone. Therefore, the addition of atrazine to tolpyralate applications may broaden the spectrum of weed control and improve speed of control in some species. Further insights in this regard are provided in a subsequent companion paper

Table 2.1. Soil characteristics, planting, spraying and harvest dates for trials near Ridgetown and Exeter, Ontario, Canada in 2015, 2016 and 2017. Soil Characteristics Planting Date Spray Harvest Date Location Year Type OMa pH Date (%) Ridgetown 2015 Sandy clay loam 4.2 7.3 May 14 June 17 Nov 5 2016 Sandy clay loam 3.2 7.1 May 6 June 9 Oct 18 2017 Sandy clay loam 3.9 7.2 May 15 June 21 Oct 31 Exeter 2015 Clay loam 3.6 7.7 May 6 June 9 n/ab 2016 Clay loam 3.2 7.7 May 6 June 10 Oct 7 2017 Clay loam 4.5 7.8 May 19 June 13 Oct 19 aOrganic matter. bNot harvested in 2015.

Table 2.2. Non-linear regression parameters (± se) and predicted tolpyralate or tolpyralate + atrazine dose required for 50, 80 and 90% control of velvetleaf (ABUTH), pigweed species (AMASS), common ragweed (AMBEL), common lambsquarters (CHEAL), barnyardgrass (ECHCG), ladysthumb (POLPE), green foxtail (SETVI) and wild mustard (SINAR) at 2 weeks after application in field studies conducted in Ontario, Canada in 2015, 2016 and 2017.

Equation 1 Parameters Predicted tolpyralate dosea

a b c ED50 ED80 ED90 g ai ha-1 ABUTH 86.77 (1.93) 0.57 (0.07) 86.74 (4.53) 1.5 4.5 - AMASS 86.46 (1.72) 0.56 (0.06) 86.45 (3.97) 1.5 4.5 - AMBEL 85.45 (1.31) 0.77 (0.02) 85.2 (2.78) 3.4 10.6 - CHEAL 89.76 (1.19) 0.6 (0.03) 89.67 (2.78) 1.6 4.4 - ECHCG 77.36 (2.89) 0.7 (0.06) 77.1 (6.44) 3 - - POLPE 70.69 (2.64) 0.73 (0.05) 70.37 (5.78) 3.8 - - SETVI 78.51 (1.92) 0.7 (0.04) 78.15 (4.27) 2.9 - - SINAR 64.25 (3.08) 0.92 (0.01) 65.01 (4.7) 18.6 - -

Equation 2 Parameters Predicted tolpyralate + atrazine dose

a b c ED50 ED80 ED90 g ai ha-1 ABUTH 98.36 (0.59) 98.35 (1.41) 0.02 (0.001) 1.1 + 35.7 2.5 + 84.4 3.7 + 124 AMASS 97.62 (1.19) 97.46 (2.73) 0.01 (0.001) 1.6 + 52.7 3.8 + 125.8 5.6 + 187.4 AMBEL 98.3 (0.77) 98.15 (1.73) 0.01 (0.001) 1.8 + 59.3 4.2 + 140.6 6.2 + 206.7 CHEAL 99.29 (0.44) 99.29 (1.03) 0.02 (0.001) 1.1 + 35.8 2.5 + 83.7 3.6 + 121 ECHCG 94.57 (2.04) 94.05 (4.55) 0.01 (0.001) 2.1 + 71.1 5.3 + 177.5 8.6 + 287.9 POLPE 94.01 (1.7) 92.83 (3.58) 0.01 (0.001) 2.9 + 98.1 7.5 + 248.5 12.4 + 412.9

SETVI 95.53 (1.3) 95.05 (2.8) 0.01 (0.001) 2.6 + 87.5 6.5 + 214.9 10 + 334.7

SINAR 97.59 (1.64) 96.5 (3.53) 0.01 (0.001) 2.6 + 85.2 6.2 + 205.1 9.2 + 306.3 a ED50, ED80, and ED90 denote the predicted effective dose of tolpyralate or tolpyralate + atrazine for 50%, 80% and 90% control, respectively. Where a predicted dose could not be computed by the regression equation, values are represented by a dash (-).

Table 2.3. Non-linear regression parameters (± se) and predicted tolpyralate and tolpyralate + atrazine dose required for 50, 80 and 90% control of velvetleaf (ABUTH), pigweed species (AMASS), common ragweed (AMBEL), common lambsquarters (CHEAL), barnyardgrass (ECHCG), ladysthumb (POLPE), green foxtail (SETVI) and wild mustard (SINAR) at 4 weeks after application in field studies conducted in Ontario, Canada in 2015, 2016 and 2017.

Equation 1 Parameters Predicted tolpyralate dosea

a b c ED50 ED80 ED90 g ai ha-1 ABUTH 96.15 (0.58) 0.48 (0.03) 96.15 (1.38) 1 2.4 3.8 AMASS 92.66 (1.39) 0.6 (0.04) 92.58 (3.16) 1.5 3.9 6.9 AMBEL 95.99 (1) 0.72 (0.01) 95.83 (2.2) 2.2 5.4 8.4 CHEAL 96.41 (0.63) 0.45 (0.04) 96.41 (1.52) 0.9 2.2 3.4 ECHCG 82.21 (2.53) 0.74 (0.04) 81.96 (5.53) 3 11.8 - POLPE 72.47 (2.83) 0.8 (0.03) 70.72 (5.82) 5.2 - - SETVI 90.05 (1.63) 0.78 (0.02) 89.29 (3.44) 3.2 8.7 29.6 SINAR 73.53 (4.26) 0.96 (0.009) 71.78 (5.11) 24.5 - -

Equation 2 Parameters Predicted tolpyralate + atrazine dose

a b c ED50 ED80 ED90 g ai ha-1 ABUTH 98.56 (0.44) 98.55 (1.06) 0.02 (0.001) 1 + 34.4 2.4 + 81.1 3.6 + 118.6 AMASS 97.68 (1.55) 97.52 (3.5) 0.01 (0.001) 1.8 + 60.9 4.4 + 145.4 6.5 + 216.5 AMBEL 98.55 (0.79) 98.46 (1.8) 0.01 (0.001) 1.6 + 52 3.7 + 122.8 5.4 + 179.7 CHEAL 99.2 (0.33) 99.2 (0.79) 0.02 (0.001) 0.9 + 30.3 2.1 + 70.9 3.1 + 102.6 ECHCG 94.29 (2.14) 93.82 (4.78) 0.01 (0.001) 2.1 + 69.4 5.2 + 174 8.6 + 285.2 POLPE 93.61 (1.79) 92.47 (3.78) 0.01 (0.001) 3 + 101.2 7.7 + 258 13.1 + 436.7 SETVI 96.27 (1.14) 95.8 (2.46) 0.01 (0.001) 2.6 + 85 6.2 + 207.1 9.6 + 318.6

SINAR 98.92 (1.31) 98.34 (2.73) 0.01 (0.0004) 3 + 101.2 7.2 + 238.8 10.4 + 347.8 a ED50, ED80, and ED90 denote the predicted effective dose of tolpyralate or tolpyralate + atrazine for 50%, 80% and 90% control, respectively. Where a predicted dose could not be computed by the regression equation, values are represented by a dash (-).

Table 2.4. Non-linear regression parameters (± se) and predicted tolpyralate and tolpyralate + atrazine dose required for 50, 80 and 90% control of velvetleaf (ABUTH), pigweed species (AMASS), common ragweed (AMBEL), common lambsquarters (CHEAL), barnyardgrass (ECHCG), ladysthumb (POLPE), green foxtail (SETVI) and wild mustard (SINAR) at 8 weeks after application in field studies conducted in Ontario, Canada in 2015, 2016 and 2017.

Equation 1 Parameters Predicted tolpyralate dosea

a b c ED50 ED80 ED90 g ai ha-1 ABUTH 97.61 (0.37) 0.45 (0.02) 97.6 (0.89) 0.9 2.2 3.2 AMASS 92.93 (1.74) 0.67 (0.03) 92.77 (3.96) 1.9 4.9 8.5 AMBEL 95.7 (0.86) 0.68 (0.02) 95.55 (1.94) 1.9 4.7 7.3 CHEAL 93.37 (1.25) 0.55 (0.04) 93.35 (2.96) 1.3 3.3 5.6 ECHCG 87.21 (1.78) 0.78 (0.02) 86.57 (3.73) 3.4 10.1 - POLPE 71.27 (3.2) 0.87 (0.03) 68 (5.89) 8.2 - - SETVI 92.41 (1.76) 0.79 (0.02) 91.6 (3.66) 3.3 8.5 15.5 SINAR 73.6 (4.58) 0.95 (0.01) 68.2 (5.89) 19.8 - -

Equation 2 Parameters Predicted tolpyralate + atrazine dose

a b c ED50 ED80 ED90 g ai ha-1 ABUTH 98.9 (0.24) 98.9 (0.57) 0.03 (0.002) 0.7 + 23.6 1.7 + 55.4 2.4 + 80.6 AMASS 97.13 (1.12) 96.99 (2.56) 0.01 (0.001) 1.6 + 54.7 4 + 131.5 5.9 + 198 AMBEL 98.8 (0.66) 98.7 (1.5) 0.01 (0.001) 1.7 + 56.5 4 + 133 5.8 + 193.9 CHEAL 98.3 (0.54) 98.28 (1.3) 0.02 (0.001) 1 + 34.5 2.5 + 81.6 3.6 + 120 ECHCG 93.83 (1.46) 93.28 (3.2) 0.01 (0.001) 2.5 + 81.7 6.2 + 206.4 10.4 + 345.2 POLPE 94.69 (1.78) 92.88 (3.69) 0.01 (0.001) 3.2 + 106.8 8.1 + 269.3 13.1 + 436 SETVI 94.66 (1.4) 94.27 (2.98) 0.01 (0.001) 2.8 + 93.7 7 + 233.5 11.3 + 377.3 SINAR 99.12 (1.5) 98.9 (3.13) 0.01 (0.001) 3 + 100.1 7.1 + 235 10.2 + 340.8 a ED50, ED80, and ED90 denote the predicted effective dose of tolpyralate or tolpyralate + atrazine for 50%, 80% and 90% control, respectively. Where a predicted dose could not be computed by the regression equation, values are represented by a dash (-).

Table 2.5. Non-linear regression parameters (± se) and predicted tolpyralate and tolpyralate + atrazine dose required for 50, 80 and 90% reduction in velvetleaf (ABUTH), pigweed species (AMASS), common ragweed (AMBEL), common lambsquarters (CHEAL), barnyardgrass (ECHCG), ladysthumb (POLPE), green foxtail (SETVI) and wild mustard (SINAR) density relative to non-treated control plots within blocks at 8 weeks after application in field studies conducted in Ontario, Canada in 2015, 2016 and 2017.

Equation 3 Parameters Predicted tolpyralate dosea

a b c ED50 ED80 ED90 g ai ha-1 ABUTH 1.05 (0.24) 2.44 (0.6) 0.48 (0.38) 2.6 - - AMASS 8.99 (2.06) 14.4 (5.23) 0.81 (1.89) 2.1 - - AMBEL 1.46 (3.53) 47.81 (6.47) 0.13 (0.04) 5.7 13.7 20.6 CHEAL 7.42 (1.44) 17 (3.36) 0.54 (0.39) 2.4 - - ECHCG 6.1 (3.81) 20.69 (5.7) 0.06 (0.04) 17.6 - - POLPE 4.79 (4.5) 8.84 (4.3) 0.02 (0.03) 62.2 - - SETVI 8.81 (3.87) 36.34 (5.76) 0.08 (0.03) 12.7 66.7 - SINAR 0 (0) 44.31 (7.83) 0.004 (0.005) - - -

Equation 3 Parameters Predicted tolpyralate + atrazine dose

a b c ED50 ED80 ED90 g ai ha-1 ABUTH 0.54 (0.2) 2.92 (0.49) 0.01 (0.005) 2.2 + 73.8 7.3 + 242.9 - AMASS 2.95 (1.83) 20.31 (4.39) 0.01 (0.01) 2 + 67.1 5.9 + 195.7 - AMBEL 0.23 (2.22) 49.25 (5.1) 0.01 (0.002) 2.2 + 72.5 5.1 + 169.3 7.3 + 243.8 CHEAL 1.21 (0.9) 23.18 (2.1) 0.01 (0.005) 1.5 + 49 3.6 + 120.9 5.8 + 192.6 ECHCG 0.28 (1.61) 24.08 (2.75) 0.002 (0.001) 9.6 + 321.3 22.7 + 755.5 33 + 1100 POLPE 0 (0) 11.06 (1.37) 0.001 (0.0004) 19.2 + 640.9 44.7 + 1488.2 63.9 + 2129.1 SETVI 0.75 (4.16) 46.35 (5.77) 0.01 (0.001) 2.6 + 87.6 6.2 + 207.1 9.1 + 303.9 SINAR 0 (0) 28.97 (3.43) 0.003 (0.001) 6.8 + 225.1 15.7 + 522.6 22.4 + 747.7 a ED50, ED80, and ED90 denote the predicted effective dose of tolpyralate or tolpyralate + atrazine for 50%, 80% and 90% reduction in weed density relative to the non-treated control, respectively. Where a predicted dose could not be computed by the regression equation, values are represented by a dash (-).

Table 2.6. Non-linear regression parameters (± se) and predicted tolpyralate and tolpyralate + atrazine dose required for 50, 80 and 90% reduction in velvetleaf (ABUTH), pigweed species (AMASS), common ragweed (AMBEL), common lambsquarters (CHEAL), barnyardgrass (ECHCG), ladysthumb (POLPE), green foxtail (SETVI) and wild mustard (SINAR) dry biomass relative to non-treated control plots within blocks at 8 weeks after application in field studies conducted in Ontario, Canada in 2015, 2016 and 2017.

Equation 3 Parametersa Predicted tolpyralate dose

a b c ED50 ED80 ED90 g ai ha-1 ABUTH 0.23 (0.21) 4.69 (0.53) 0.81 (0.64) 0.9 2.3 3.5 AMASS 6.67 (4.32) 44.07 (10.68) 0.53 (0.44) 1.6 4.8 -a AMBEL 2.62 (9.1) 221.3 (22.36) 0.64 (0.29) 1.1 2.6 3.8 CHEAL 3.09 (2.94) 86.6 (7.13) 0.98 (0.88) 0.7 1.8 2.7 ECHCG 1.84 (1.85) 32.72 (5) 0.22 (0.06) 3.4 8.5 13.7 POLPE 2.21 (3.34) 16.21 (4.15) 0.05 (0.03) 17.4 49.6 - SETVI 2.78 (1.58) 35.45 (3.5) 0.34 (0.09) 2.3 5.8 10.3 SINAR 28.56 (34.2) 231.9 (40.21) 0.04 (0.02) 19.1 53 -

Equation 3 Parameters Predicted tolpyralate + atrazine dose

a b c ED50 ED80 ED90 g ai ha-1 ABUTH 0.07 (0.2) 4.84 (0.52) 0.02 (0.01) 1.1 + 36.6 2.6 + 86.4 3.8 + 126.5 AMASS 1.7 (3.59) 48.93 (8.82) 0.015 (0.01) 1.4 + 47.8 3.5 + 115.4 5.3 + 175.6 AMBEL 0.5 (8.99) 223.5 (22.17) 0.02 (0.01) 1.1 + 35.5 2.5 + 82.7 3.6 + 118.7 CHEAL 0.36 (2.89) 89.32 (7) 0.03 (0.02) 0.8 + 26.7 1.9 + 62.2 2.7 + 89.4 ECHCG 1.31 (0.8) 31.91 (2.43) 0.01 (0.002) 1.8 + 59.5 4.3 + 144.7 6.7 + 223.5

POLPE 6.73 (4.17) 9.44 (9.33) 0.008 (0.02) 6.8 + 225.2 - -

SETVI 1.93 (1.69) 36.05 (3.47) 0.01 (0.002) 3.4 + 111.6 8.3 + 276 13.3 + 441.6

SINAR 0.46 (8.92) 254.2 (19.67) 0.01 (0.002) 2.1 + 71 5 + 165.2 7.11 + 237 a ED50, ED80, and ED90 denote the predicted effective dose of tolpyralate or tolpyralate + atrazine for 50%, 80% and 90% reduction in weed dry biomass relative to the non-treated control, respectively. Where a predicted dose could not be computed by the regression equation, values are represented by a dash (-).

Equation 1

Parameters Predicted tolpyralate dosea a b c ED50 ED80 ED90 g ai ha-1 Grain Yield 87.95 (1.79) 0.7 (0.06) 42 (4) 0.3 4.7 -

Equation 2

Parameters Predicted tolpyralate + atrazine dose a b c ED50 ED80 ED90 g ai ha-1

Grain Yield 95.55 (1.33) 50.8 (3.03) 0.01 (0.002) 0.3 + 8.2 2.7 + 89.7 5 + 167.8 a ED50, ED80, and ED90 denote the predicted effective dose of tolpyralate or tolpyralate + atrazine to obtain 50%, 80% and 90% of the yield obtained in the weed free plots within replications, respectively. Where a predicted dose could not be computed by the regression equation, the value is represented by a dash (-).

Chapter 3: Tolpyralate Efficacy: Part II. Comparison of three group 27 herbicides applied POST for annual grass and broadleaf weed control in corn

3.1 Abstract

Tolpyralate is a new Group 27 pyrazolone herbicide which inhibits the 4-hydroxyphenyl- pyruvate dioxygenase enzyme. In a study of the biologically-effective dose of tolpyralate from

2015-2017 in Ontario, Canada, tolpyralate exhibited efficacy on a broader range of species when co-applied with atrazine; however, there is limited published information on the efficacy of tolpyralate and tolpyralate + atrazine relative to mesotrione and topramezone, applied postemergence with atrazine, for control of annual grass and broadleaf weeds. In this study, tolpyralate applied alone at 30 g ai ha-1 provided >90% control of common lambsquarters

[CHEAL], velvetleaf [ABUTH], common ragweed [AMBEL] green/redroot pigweed [AMASS] and green foxtail [SETVI] 8 WAA. Atrazine addition was required to achieve >90% control of wild mustard [SINAR], ladysthumb [POLPE] and barnyardgrass [ECHCG] 8 WAA. Tolpyralate

+ atrazine (30 + 1000 g ai ha-1) and topramezone + atrazine (12.5 + 500 g ai ha-1) provided equivalent control 8 WAA of the eight weed species in this study; however, tolpyralate + atrazine provided >90% control of green foxtail by 1 WAA. Tolpyralate + atrazine provided 18,

68 and 67% better control of common ragweed, green foxtail and barnyardgrass, respectively, than mesotrione + atrazine (100 + 280 g ai ha-1) 8 WAA. Overall, tolpyralate + atrazine applied

POST, provided equivalent or improved control of annual grass and broadleaf weeds compared to mesotrione + atrazine and topramezone + atrazine.

3.2 Introduction

Herbicides that inhibit the 4-hydroxyphenyl-pyruvate dioxygenase (HPPD) enzyme in susceptible plants represent the most recent herbicide mode of action successfully commercialized for weed management in Ontario, Canada. Although herbicides within the triketone and isoxazole chemical families of HPPD-inhibitors have been used in North America for nearly two decades, the registration of topramezone in 2005 marked the first development of the pyrazolone family of HPPD-inhibitors (Grossmann and Ehrhardt 2007).

Four HPPD-inhibitors are available commercially for use in corn in the United States and

Canada. These include isoxazoles (isoxaflutole), triketones (mesotrione and tembotrione), and pyrazolones (topramezone) (Health Canada 2018). Tolpyralate is a new pyrazolone herbicide molecule that was registered in 2017 in the United States and Canada for use in field, pop, seed and sweet corn (Anonymous 2017a). Results from this study presented in a companion paper

(Part 1) indicate that tolpyralate exhibits strong herbicidal efficacy in field environments

(Metzger et al. 2018a); however, HPPD-inhibitors vary widely in their application timing, use rate and selectivity (Hawkes 2012; Ontario Ministry of Agriculture, Food and Rural Affairs

(OMAFRA) 2018a). Mesotrione provides PRE and POST control of several broadleaf weeds including velvetleaf, common lambsquarters, Amaranthus spp. and Polygonum spp; however, control of annual grasses including green foxtail and barnyardgrass is variable (Creech et al.

2004; OMAFRA 2018a). Tembotrione, applied POST, controls certain grass weeds, including giant foxtail (Setaria faberi Herrm.), witchgrass (Panicum capillare L.) and barnyardgrass, in addition to several annual broadleaf weeds (Bollman et al. 2008; Williams et al. 2011).

Isoxaflutole is applied PRE or early POST and provides control of several grass and broadleaf weeds (Ahrens et al. 2013; Pallet et al. 2001). Conversely, topramezone controls grass and

64 broadleaf species, but is only applied POST (Anonymous 2016; Bollman et al. 2008). Similarly, tolpyralate applied POST controls several grass and broadleaf species (Metzger et al. 2018a) but is reported to have limited residual activity in soil (Anonymous 2017a; Kikugawa et al. 2015). At this time, there is limited published information examining the interspecific activity of tolpyralate relative to other HPPD-inhibitors.

Due to variation in selectivity and residual control, HPPD-inhibitors are commonly applied with other herbicides, such as atrazine, in tank or pre-formulated mixtures (OMAFRA

2018a). Atrazine is a photosystem-II (PSII)-inhibitor and is one of the most widely-used corn herbicides; it is applied to over 55% of total corn hectares in the USA (National Agricultural

Statistics Service 2015). Atrazine provides broad spectrum annual broadleaf control, has flexible application timing, and a mode of action which is complementary to that of HPPD-inhibitors due to their shared interaction with plastoquinone within PSII (Creech et al. 2004; Hess 2000).

Interactions between mesotrione and atrazine are widely reported. For example, mixture of atrazine and mesotrione improves the control of common ragweed, Palmer amaranth

(Amaranthus palmeri S. Watson), common cocklebur (Xanthium strumarium L.), ivyleaf morningglory (Ipomoea hederacea Jacq.), yellow nutsedge (Cyperus esculentus L.), redroot pigweed and velvetleaf, compared to mesotrione applied alone (Abendroth et al. 2006; Bollman et al. 2008; Creech et al. 2004; Johnson et al. 2002). Kohrt and Sprague (2017a) found that the addition of atrazine to both mesotrione and tembotrione improved control of atrazine-resistant

Palmer amaranth biotypes. Stephenson and Bond (2012) reported that the addition of atrazine to isoxaflutole applied POST improved the control of entireleaf morningglory (Ipomoea hederacea

(L.) Jacq. var integriuscula A. Gray) and Palmer amaranth. Similarly, Bollman et al. (2008)

65 found that the addition of atrazine to topramezone provided better control of common lambsquarters compared to topramezone applied alone.

There have been few studies investigating the benefit of the addition of atrazine to tolpyralate. Tonks et al. (2015) reported that on average, the addition of atrazine to tolpyralate improved control of broadleaf species including velvetleaf, Amaranthus spp., common ragweed, common lambsquarters and kochia (Bassia scoparia (L.) A. J. Scott), 30 days after application

(DAA). However, the difference in control with tolpyralate alone or with atrazine 30 DAA varied widely by species (Tonks et al. 2015). Kohrt and Sprague (2017a) reported similar control of atrazine-resistant Palmer amaranth with tolpyralate alone and tolpyralate plus atrazine.

The results presented in the companion paper (Metzger et al. 2018a), indicate that the addition of atrazine to tolpyralate at a 1:33.3 ratio broadens the spectrum of weed control. Given that tolpyralate is reported to have limited pre-emergence activity (Anonymous 2017a), the addition of atrazine may also contribute to improved late-season control of select species; however, there is no published information in this regard. Therefore, the objectives of this study were to examine the effects of atrazine addition to tolpyralate, and to compare tolpyralate efficacy in field environments to the efficacy of two other HPPD-inhibiting herbicides currently in use.

3.3 Materials and Methods

Experimental Methods The results outlined in this paper describe the results from field studies conducted near

Ridgetown and Exeter, Ontario, Canada from 2015-2017, as described in the companion paper,

Part I (Metzger et al. 2018). A total of six experiments with four replications each were

66 conducted, arranged at each site in a randomized complete block design. Treatments investigated in this paper include tolpyralate applied at 30 g ha-1, representing the lowest current label rate

(Anonymous 2017a), and tolpyralate + atrazine applied at 30 + 1000 g ha-1, representing a 1:33.3 tank-mix ratio. This ratio was determined to be appropriate with consideration of preliminary work conducted by Tonks et al. (2015). Each of these tolpyralate treatments were compared against two current HPPD-inhibitors applied at the registered POST label rate for field corn in

Canada: mesotrione + atrazine (100 + 280 g ha-1) and topramezone + atrazine (12.5 + 500 g ha-1).

Adjuvants were included in accordance with herbicide manufacturer recommendations.

Tolpyralate applications included methylated seed oil (MSO Concentrate®, Loveland Products

Inc., Loveland, CO) at 0.50% v/v plus 28% N urea ammonium nitrate (UAN) at 2.50% v/v.

Mesotrione applications included a non-ionic surfactant (Agral® 90, Syngenta Canada Inc.,

Guelph, ON) at 0.20% v/v, while topramezone treatments included blended surfactant (Merge®,

BASF Canada Inc., Mississauga, ON) at 0.50% v/v plus UAN at 1.50% v/v. Treatments were applied when weeds reached 10 cm in height on average, using a four-nozzle handheld sprayer equipped with ULD12002 nozzles (Pentair, New Brighton, MN, USA), calibrated to deliver a

187 L ha-1 spray volume at 240 kPa.

Visible control was assessed against the non-treated control plots 1, 2, 4 and 8 WAA, with each species assigned a percent value between 0 and 100 where 0 signifies no control and

100 signifies complete plant death/absence from plots. At 8 WAA, the reduction in density and biomass of each species provided by all treatments was determined by counting and harvesting the plants contained in two, 0.5 m2 quadrats per plot. Samples were allowed to dry at 60 C, and dry biomass recorded.

For further information regarding experimental design, location characteristics and technical methods, readers are referred to the companion paper (Part 1, Metzger et al. 2018a) derived from the same field study.

Statistical Analysis For each of the eight weed species described in the first portion of this study (Metzger et al. 2018a), visible control data at 1, 2, 4 and 8 WAA, density reduction, and dry biomass reduction (8 WAA) were subjected to a mixed-model variance analysis using the GLIMMIX procedures in SAS v. 9.4 (SAS Institute, Cary, NC). A significance level of α=0.05 was declared for all tests. Variance was partitioned into random effects of environment (comprising location and year), block nested within environment, and the treatment by environment interaction, with treatment designated as the fixed effect. The appropriate model was assigned to each parameter based on the distribution and link function which best met assumptions that residuals were normally distributed, homogeneous and had mean of zero, as determined with a Shapiro-Wilk test and scatterplots of studentized residuals. Where appropriate, a normal distribution was used.

Non-Gaussian data were analyzed using the Laplace method of integral approximation. Visible control data was assigned either a normal distribution with identity link, or a beta distribution with a logit or cumulative log link, except for wild mustard control data at 4 WAA, which were arcsine square-root transformed in order to meet assumptions. Weed density and biomass were analyzed using a normal or lognormal distribution with identity link, or a Poisson or negative binomial distribution with a log link. Least-square means of each parameter were computed on the analysis scale and converted to the data scale using the ilink option for all distributions except lognormal, where data were back-transformed using the omega method within PROC GLIMMIX

(M. Edwards, Ontario Agricultural College Statistician, University of Guelph, personal communication). Least square means were compared across the four treatments using Tukey-

Kramer’s multiple range test, and letter codes assigned by specifying the lines option in the

GLIMMIX procedure.

3.4 Results and Discussion

Means comparisons included control at each assessment timing, and reduction in density and biomass provided by tolpyralate alone or with atrazine, mesotrione + atrazine and topramezone + atrazine for the eight weed species discussed in the first portion of this study

(Metzger et al. 2018a). Means comparisons are presented in Tables 3.1-3.8.

Common Lambsquarters

At 1 WAA, tolpyralate controlled common lambsquarters 60%, while the addition of atrazine to tolpyralate improved control to 93% (Table 3.1). Tolpyralate + atrazine and topramezone + atrazine provided similar control; however, tolpyralate + atrazine provided better control than mesotrione + atrazine. The numerical increase in common lambsquarters control across the four treatments 1 WAA follows the respective rate of atrazine used with each HPPD- inhibitor. However, Woodyard et al. (2009) reported similar control of common lambsquarters with atrazine applied POST at 280 and 560 g ha-1 10 DAA, suggesting the differences observed in this study 1 WAA may be secondary to the rate of atrazine. At 2 WAA, tolpyralate alone, tolpyralate + atrazine, topramezone + atrazine and mesotrione + atrazine each provided >90% control of common lambsquarters, and the four treatments were not different from one another.

These results did not vary widely from the preliminary results described previously by Tonks et al. (2015), who reported 86% control of common lambsquarters with tolpyralate. Mesotrione + atrazine applied POST at the doses used in this study have previously been reported to provide effective control (93-99%) of common lambsquarters (Armel et al. 2003; Whaley et al. 2006;

Woodyard et al. 2009). Similarly, Bollman et al. (2008) found that topramezone + atrazine applied POST, following S-metolachlor PRE, provided 100% control of common lambsquarters.

At both 4 and 8 WAA, tolpyralate alone provided similar control to tolpyralate + atrazine and both industry standard HPPD-inhibitors; however, the addition of atrazine to tolpyralate led to a greater reduction in common lambsquarters density and biomass than tolpyralate applied alone.

Topramezone + atrazine and mesotrione + atrazine provided a similar reduction in density and biomass to tolpyralate applied alone or with atrazine.

Velvetleaf

At 1 WAA, tolpyralate + atrazine provided better control of velvetleaf than tolpyralate alone; however, at 2, 4 and 8 WAA, no differences were observed between treatments (Table

3.2). These results suggest that the addition of atrazine to tolpyralate may increase speed of velvetleaf control despite the low biologically-effective dose (3.2 g ha-1) of tolpyralate for velvetleaf control determined in the first part of this study (Metzger et al. 2018a). Speed of weed control may have important physiological/yield implications, with more rapid weed control shortening the duration of weed-crop competition. At 4 and 8 WAA, all treatments provided

≥95% control of velvetleaf, and there were no differences among treatments with respect to reduction in velvetleaf density or biomass. Similar results were outlined by Tonks et al. (2015), who reported that tolpyralate, tolpyralate + atrazine and topramezone + atrazine controlled velvetleaf 87, 94 and 93%, respectively. Likewise, mesotrione has been found to exhibit excellent foliar activity on velvetleaf (Creech et al. 2004; Johnson and Young 2002).

Pigweed Species

Similar to common lambsquarters and velvetleaf, differences were observed among treatments in control of pigweed species (Amaranthus spp.) 1 WAA but not at later assessment

70 timings. The addition of atrazine to tolpyralate provided 96% control of pigweed species 1

WAA, which was superior to tolpyralate applied alone, but not different than either mesotrione + atrazine or topramezone + atrazine (Table 3.3). Tolpyralate alone provided equivalent control of pigweed species compared to tolpyralate + atrazine, topramezone + atrazine and mesotrione + atrazine at 2, 4 and 8 WAA. Tolpyralate + atrazine provided a greater reduction in pigweed density and biomass than tolpyralate alone, but results were similar to those observed with both topramezone + atrazine and mesotrione + atrazine. Similar activity with these HPPD-inhibitors has been reported in other Amaranthus spp. (Hugie et al. 2008; Kohrt and Sprague 2017a; Tonks et al. 2015; Woodyard et al. 2009). Kohrt and Sprague (2017a) observed no difference in control of atrazine-resistant Palmer amaranth with tolpyralate (40 g ha-1) compared to tolpyralate + atrazine (40 + 560 g ha-1). Similarly, Armel et al. (2003) found that mesotrione (105 g ha-1) with or without atrazine (560 g ha-1), controlled smooth pigweed (Amaranthus hybridus (L.)) 99%.

Tonks et al. (2015) reported an average of 89% control of Amaranthus spp. including Palmer amaranth, redroot pigweed and waterhemp (Amaranthus tuberculatus (Moq.) J. D. Sauer) with tolpyralate (30 g ha-1). Thus, consistent with results obtained in the BED study (Metzger et al.

2018a), tolpyralate exhibits high activity on Amaranthus species.

Common Ragweed

At 1 WAA, there was a greater range in common ragweed control than other broadleaf species. Tolpyralate + atrazine provided 34% better control than tolpyralate alone and 26% better control than mesotrione plus atrazine, but results were not significantly different from topramezone + atrazine (Table 3.4). Control with mesotrione + atrazine was similar to tolpyralate alone and topramezone + atrazine treatments. At 2 WAA, tolpyralate and tolpyralate + atrazine provided similar control of common ragweed. Tolpyralate + atrazine and topramezone + atrazine

71 controlled common ragweed equally; both were superior to mesotrione + atrazine. Similarly, at 4

WAA, tolpyralate, tolpyralate + atrazine and topramezone + atrazine each provided better common ragweed control than mesotrione + atrazine. Previous research has shown variable control of common ragweed with mesotrione, but improved control when co-applied with atrazine (Armel et al. 2003; Bollman et al. 2008). At 8 WAA, tolpyralate + atrazine controlled common ragweed 100%, which was equivalent to control with tolpyralate alone and topramezone + atrazine, and greater than control with mesotrione + atrazine. It is plausible that poorer control of common ragweed with mesotrione + atrazine at 8 WAA may be reflective of the comparatively lower rate of atrazine applied. However, tolpyralate alone provided 95% control of common ragweed in this study when no atrazine was applied, suggesting that the observed differences can be attributed to the HPPD inhibitors. Additionally, tolpyralate alone provided an equivalent reduction in common ragweed density and biomass compared with topramezone + atrazine and mesotrione + atrazine treatments. In contrast, there was a greater decrease in common ragweed density and biomass with tolpyralate + atrazine compared with other treatments. Therefore, the addition of atrazine to tolpyralate provided no significant benefit in visible control at 2, 4, and 8 WAA, but provided more complete weed necrosis at 8 WAA, contributing to a greater reduction in common ragweed density and biomass compared with all other treatments. Similarly, these results suggest that treatments with atrazine at 500 or 1000 g ha-1 provided a greater numerical decrease in common ragweed density than tolpyralate alone or mesotrione + atrazine. Therefore, it is likely that these higher rates of atrazine contributed to improved residual control of late-emerging seedlings, which were counted during harvests, but contributed little to biomass measurements. Overall, the results from this study are similar to

72 those of Tonks et al. (2015), who reported 89 and 95% control of common ragweed with tolpyralate and tolpyralate + atrazine, respectively.

Ladysthumb

There was considerable variation in control of ladysthumb observed with all treatments, potentially due to interspecific competition within plots which could have prevented thorough spray coverage of ladysthumb foliage in the lower part of the weed canopy. At 1 WAA, all treatments provided equivalent control; however, treatment separation was present at later assessment timings. Consistent with findings described in the first part of this study (Metzger et al. 2018a), the addition of atrazine to tolpyralate improved ladysthumb control at 2, 4 and 8

WAA (Table 3.5). At 2, 4 and 8 WAA, topramezone + atrazine provided control that was similar to tolpyralate alone, tolpyralate + atrazine and mesotrione + atrazine. At 4 WAA, mesotrione + atrazine provided better ladysthumb control than tolpyralate alone; however, by 8 WAA, control with both treatments was equivalent. At 8 WAA, tolpyralate + atrazine, topramezone + atrazine and mesotrione + atrazine all provided similar control of ladysthumb. Mesotrione + atrazine reduced ladysthumb density more than tolpyralate alone, but all treatments provided a similar reduction in biomass. Previous research has not investigated tolpyralate efficacy on this species; although in agreement with relatively higher BED values outlined in the first part of this study

(Metzger et al. 2018a), ladysthumb appears to exhibit greater tolerance to tolpyralate than other broadleaf species, including common lambsquarters, velvetleaf, pigweed species and common ragweed. These results suggest that the addition of atrazine is necessary to achieve adequate control (>80%) of this species. Comparable findings have been reported by Rahman et al.

(2013), who found that the addition of atrazine to topramezone was required for control of another Polygonum species (Polygonum aviculare L.).

Wild Mustard

Consistent with our previous findings (Metzger et al. 2018a), wild mustard exhibited high tolerance to tolpyralate. At 1 WAA, bleaching symptoms were visible on wild mustard plants with all treatments; however, tolpyralate + atrazine provided better control than tolpyralate alone

(Table 3.6). At 1 WAA, tolpyralate + atrazine, topramezone + atrazine and mesotrione + atrazine provided similar control of wild mustard. At 2, 4 and 8 WAA, tolpyralate + atrazine, topramezone + atrazine and mesotrione + atrazine controlled wild mustard (≥99%), which was better than tolpyralate applied alone. Similarly, density and dry weight were reduced to zero with tolpyralate + atrazine, topramezone + atrazine and mesotrione + atrazine. Consistent with these results, atrazine applied POST alone is reported to have excellent efficacy on wild mustard

(OMAFRA 2018a), suggesting there may be little benefit to inclusion of the HPPD-inhibitor.

However, mesotrione has been reported to control wild mustard when applied POST (Cornes

2005). Since atrazine was not applied alone in this study, distinctions cannot be made between the relative wild mustard control provided by atrazine versus the respective HPPD-inhibitor.

Green Foxtail

At 1 WAA, tolpyralate + atrazine was the most efficacious treatment for control of green foxtail. Control of green foxtail with tolpyralate was similar when applied alone or in combination with atrazine at most assessment timings; however, at 1 WAA, tolpyralate + atrazine provided 18% better control than tolpyralate alone (Table 3.7). Tolpyralate + atrazine also provided better green foxtail control than topramezone + atrazine or mesotrione + atrazine.

At 2, 4 and 8 WAA, tolpyralate applied alone or in combination with atrazine, and topramezone

+ atrazine provided better control of green foxtail than mesotrione + atrazine. These results are consistent with those reported in the literature, which have documented poor control of Setaria

74 spp. with mesotrione (Armel et al. 2003; Creech et al. 2004; Kaastra et al. 2008), but acceptable control with topramezone (Grossmann and Ehrhardt 2007; Bollman et al. 2008; Kaastra et al.

2008; Whaley et al. 2006). Reduction in green foxtail density and biomass was equivalent with tolpyralate applied alone or with atrazine, and was greater than topramezone + atrazine and mesotrione + atrazine. Results from the first part of this study (Metzger et al. 2018a) determined the BED of tolpyralate in green foxtail to be 29.6 g ha-1 8 WAA when applied alone. Thus, the dose of 30 g ha-1 examined in this analysis is likely to diminish any contribution to green foxtail control provided by atrazine; however, control data collected 1 WAA suggests that atrazine may improve speed of green foxtail control with tolpyralate.

Barnyardgrass

Control of barnyardgrass followed similar trends to green foxtail, but was more variable with all treatments. Tolpyralate alone or with the addition of atrazine provided statistically similar control at all assessment timings (Table 3.8). At 1 WAA, tolpyralate + atrazine and topramezone + atrazine provided similar barnyardgrass control, while mesotrione + atrazine provided poorer control. In agreement with these results, mesotrione + atrazine has not been found to provide adequate control of barnyardgrass in Ontario (OMAFRA 2018a). In contrast, both Creech et al. (2004) and De Cauwer et al. (2012) found that mesotrione provided complete barnyardgrass control; however, those studies were conducted in a greenhouse environment. At 2

WAA, tolpyralate + atrazine provided better control of barnyardgrass than topramezone + atrazine and mesotrione + atrazine. Topramezone has previously been reported to control barnyardgrass at doses similar to those used in this study (De Cauwer et al. 2012). At 4 and 8

WAA, tolpyralate alone and tolpyralate + atrazine provided similar barnyardgrass control. No statistically-significant difference was observed among treatments in barnyardgrass density or

75 biomass 8 WAA; however, the numerical differences across treatments may have biological significance.

3.5 Conclusions

The addition of atrazine to tolpyralate improved the control of ladysthumb and wild mustard, while there was no improvement in control of common lambsquarters, velvetleaf, pigweed species, common ragweed, green foxtail and barnyardgrass compared with tolpyralate applied alone. Tolpyralate + atrazine and topramezone + atrazine provided equivalent control of common lambsquarters, velvetleaf, pigweed species, common ragweed, ladysthumb, wild mustard, green foxtail and barnyardgrass. Tolpyralate + atrazine and mesotrione + atrazine provided equivalent control of common lambsquarters, velvetleaf, pigweed species, ladysthumb and wild mustard; in contrast, tolpyralate + atrazine provided better control of common ragweed, green foxtail and barnyardgrass than mesotrione + atrazine. The co-application of tolpyralate with atrazine at the

1:33.3 ratio used in this study resulted in more rapid control of all species compared to tolpyralate alone, with the exception of ladysthumb and barnyardgrass. Additionally, the rate of atrazine used with tolpyralate in these studies may have contributed to extended residual control of some late-emerging weed seedlings. However, it is unclear what ratio of tolpyralate to atrazine is required for these effects to occur. Future research on the optimal ratio of atrazine to use in combination with tolpyralate would help to maximize weed control and reduce the selection pressure for resistance to a single herbicide mechanism of action, while also minimizing environmental loading of herbicides.

Table 3.1 - CHEAL. Control 1, 2, 4 and 8 weeks after application (WAA), density and dry biomass of common lambsquarters (Chenopodium album L.) following treatment with tolpyralate, tolpyralate + atrazine, topramezone + atrazine and mesotrione + atrazine in field studies conducted in Ontario, Canada in 2015, 2016 and 2017.a Dose Density Biomass 1 WAA 2 WAA 4 WAA 8 WAA Treatment (g ai ha-1) (no. m-2) (g DM m-2)b

Tolpyralate 30 60 c 91 a 96 a 92 a 0.100 b 0.030 b

Tolpyralate + 30 + 1000 93 a 100 a 99 a 99 a 0.001 a 0.001 a atrazine

Topramezone + 12.5 + 500 87 ab 99 a 98 a 96 a 0.060 ab 0.003 ab atrazine

Mesotrione + 100 + 280 78 b 94 a 98 a 97 a 0.010 ab 0.007 ab atrazine aMeans followed by the same letter within columns are not significantly different according to Tukey-Kramer’s multiple range test α=0.05. bDM, dry matter.

Table 3.2 - ABUTH. Control 1, 2, 4 and 8 weeks after application (WAA), density and dry biomass of velvetleaf (Abutilon theophrasti Medik.) following treatment with tolpyralate, tolpyralate + atrazine, topramezone + atrazine and mesotrione + atrazine in field studies conducted in Ontario, Canada in 2015, 2016 and 2017.a Dose Density Biomass 1 WAA 2 WAA 4 WAA 8 WAA Treatment (g ai ha-1) (no. m-2) (g DM m-2)b

Tolpyralate 30 81 b 86 a 95 a 98 a 0.0 a 0.0 a

Tolpyralate + 30 + 1000 93 a 99 a 99 a 99 a 0.0 a 0.0 a atrazine

Topramezone 12.5 + 500 89 ab 95 a 97 a 99 a 0.0 a 0.0 a + atrazine

Mesotrione + 100 + 280 87 ab 98 a 99 a 99 a 0.0 a 0.0 a atrazine aMeans followed by the same letter within columns are not significantly different according to Tukey-Kramer’s multiple range test α=0.05. bDM, dry matter.

Table 3.3 - AMASS. Control 1, 2, 4 and 8 weeks after application (WAA), density and dry biomass of green/redroot pigweed (Amaranthus powelli S. Watson/Amaranthus retroflexus L.) following treatment with tolpyralate, tolpyralate + atrazine, topramezone + atrazine and mesotrione + atrazine in field studies conducted in Ontario, Canada in 2015, 2016 and 2017.a Dose Density Biomass 1 WAA 2 WAA 4 WAA 8 WAA Treatment (g ai ha-1) (no. m-2) (g DM m-2)b

Tolpyralate 30 70 b 83 a 92 a 91 a 2.5 b 0.6 b

Tolpyralate + 30 + 1000 96 a 98 a 98 a 97 a 0.0 a 0.0 a atrazine

Topramezone 12.5 + 500 83 ab 86 a 86 a 86 a 0.1 ab 0.1 ab + atrazine

Mesotrione + 100 + 280 82 ab 84 a 88 a 91 a 0.2 ab 0.1 ab atrazine aMeans followed by the same letter within columns are not significantly different according to Tukey-Kramer’s multiple range test α=0.05. bDM, dry matter.

Table 3.4 - AMBEL. Control 1, 2, 4 and 8 weeks after application (WAA), density and dry biomass of common ragweed (Ambrosia artemisiifolia L.) following treatment with tolpyralate, tolpyralate + atrazine, topramezone + atrazine and mesotrione + atrazine in field studies conducted in Ontario, Canada in 2015, 2016 and 2017.a Dose Density Biomass 1 WAA 2 WAA 4 WAA 8 WAA Treatment (g ai ha-1) (no. m-2) (g DM m-2)b

Tolpyralate 30 59 c 83 ab 97 a 95 ab 0.20 b 0.03 bc

Tolpyralate + 30 + 1000 93 a 99 a 100 a 100 a 0.00 a 0.00 a atrazine

Topramezone 12.5 + 500 80 ab 94 a 96 a 94 ab 0.01 b 0.01 b + atrazine

Mesotrione + 100 + 280 67 bc 77 b 80 b 82 b 0.40 b 0.30 c atrazine aMeans followed by the same letter within columns are not significantly different according to Tukey-Kramer’s multiple range test α=0.05. bDM, dry matter.

Table 3.5 - POLPE. Control 1, 2, 4 and 8 weeks after application (WAA), density and dry biomass of ladysthumb (Persicaria maculosa Gray) following treatment with tolpyralate, tolpyralate + atrazine, topramezone + atrazine and mesotrione + atrazine in field studies conducted in Ontario, Canada in 2015, 2016 and 2017.a Dose Density Biomass 1 WAA 2 WAA 4 WAA 8 WAA Treatment (g ai ha-1) (no. m-2) (g DM m-2)b

Tolpyralate 30 71 a 70 b 72 b 69 b 7.3 b 2.6 a

Tolpyralate + 30 + 1000 86 a 92 a 94 a 95 a 2.8 ab 0.9 a atrazine

Topramezone 12.5 + 500 73 a 78 ab 82 ab 82 ab 3.0 ab 0.3 a + atrazine

Mesotrione + 100 + 280 71 a 84 ab 92 a 92 ab 2.6 a 0.1 a atrazine aMeans followed by the same letter within columns are not significantly different according to Tukey-Kramer’s multiple range test α=0.05. bDM, dry matter.

Table 3.6 - SINAR. Control 1, 2, 4 and 8 weeks after application (WAA), density and dry biomass of wild mustard (Sinapis arvensis L.) following treatment with tolpyralate, tolpyralate + atrazine, topramezone + atrazine and mesotrione + atrazine in field studies conducted in Ontario, Canada in 2015, 2016 and 2017.a Dose Density Biomass 1 WAA 2 WAA 4 WAA 8 WAA Treatment (g ai ha-1) (no. m-2) (g DM m-2)b

Tolpyralate 30 53 b 59 b 53 b 55 b 27.5 b 62.3 b

Tolpyralate + 30 + 1000 83 a 99 a 100 a 100 a 0.00 a 0.00 a atrazine

Topramezone 12.5 + 500 76 ab 99 a 99 a 100 a 0.00 a 0.00 a + atrazine

Mesotrione + 100 + 280 73 ab 99 a 100 a 100 a 0.00 a 0.00 a atrazine aMeans followed by the same letter within columns are not significantly different according to Tukey-Kramer’s multiple range test α=0.05. bDM, dry matter.

Table 3.7 - SETVI. Control 1, 2, 4 and 8 weeks after application (WAA), density and dry biomass of green foxtail (Setaria viridis (L.) P. Beauv.) following treatment with tolpyralate, tolpyralate + atrazine, topramezone + atrazine and mesotrione + atrazine in field studies conducted in Ontario, Canada in 2015, 2016 and 2017.a Dose Density Biomass 1 WAA 2 WAA 4 WAA 8 WAA Treatment (g ai ha-1) (no. m-2) (g DM m-2)b

Tolpyralate 30 72 b 77 a 89 a 91 a 7.4 a 1.2 a

Tolpyralate + 30 + 1000 90 a 96 a 96 a 93 a 4.9 a 1.3 a atrazine

Topramezone 12.5 + 500 70 b 84 a 80 a 81 a 16 b 6.1 b + atrazine

Mesotrione + 100 + 280 32 c 35 b 33 b 25 b 36 c 51 c atrazine aMeans followed by the same letter within columns are not significantly different according to Tukey-Kramer’s multiple range test α=0.05. bDM, dry matter.

Table 3.8 - ECHCG. Control 1, 2, 4 and 8 weeks after application (WAA), density and dry biomass of barnyardgrass (Echinochloa crus-galli (L.) P. Beauv.) following treatment with tolpyralate, tolpyralate + atrazine, topramezone + atrazine and mesotrione + atrazine in field studies conducted in Ontario, Canada in 2015, 2016 and 2017.a Dose Density Biomass 1 WAA 2 WAA 4 WAA 8 WAA Treatment (g ai ha-1) (no. m-2) (g DM m-2)b

Tolpyralate 30 76 a 82 ab 86 a 84 ab 14 a 3.6 a

Tolpyralate + 30 + 1000 86 a 97 a 97 a 93 a 4.0 a 0.7 a atrazine

Topramezone 12.5 + 500 74 a 74 b 81 a 79 ab 9.9 a 3.0 a + atrazine

Mesotrione + 100 + 280 44 b 31 c 27 b 26 b 9.7 a 6.3 a atrazine aMeans followed by the same letter within columns are not significantly different according to Tukey-Kramer’s multiple range test α=0.05. bDM, dry matter.

Chapter 4: Influence of weed size and herbicide rate on the efficacy of tolpyralate plus atrazine for control of annual grass and broadleaf weeds

4.1 Abstract

Effective post-emergence (POST) herbicides and tank-mixtures are key components of integrated weed management in corn; however, herbicides vary in their efficacy based on application timing. Six field experiments were conducted over two years (2017/2018) in southwestern Ontario, Canada to determine the effects of application timing and herbicide rate on the efficacy of tolpyralate, a new 4-hydroxyphenyl-pyruvate dioxygenase inhibitor. Three rates of tolpyralate (15, 30, 40 g ai ha-1) were applied in tank-mixture with two rates of atrazine

(500, 1000 g ai ha-1) at PRE, early-POST, mid-POST and late-POST timings. Tolpyralate + atrazine at rates ≥30 + 1000 g ha-1 provided equivalent control of common lambsquarters and green pigweed applied PRE or POST, while no rate applied PRE controlled common ragweed, velvetleaf, flower-of-an-hour, ladysthumb, barnyardgrass or green foxtail. Common ragweed, common lambsquarters, velvetleaf and green pigweed were controlled equally regardless of

POST timing. In contrast, control of barnyardgrass and green foxtail declined when application was delayed to late-POST timing, irrespective of herbicide rate. Similarly, a decline in corn grain yield was observed within each tolpyralate + atrazine rate when applications were delayed to late-POST timing. Overall, these results indicate that several grass and broadleaf weed species can be controlled with tolpyralate + atrazine with an early to mid-POST application timing, before weeds reach 30 cm in height, and certain species can also be controlled PRE.

Additionally, this study provides further evidence highlighting the importance of effective, early- season weed control in corn.

4.2 Introduction

Effective weed management programs are essential in corn, as weed interference is generally the most important factor impacting grain yield (Rajcan and Swanton 2001). The critical weed-free period (CWFP) is broadly defined as the period of time in crop development during which interference from weeds will cause crop yield loss (Zimdahl 2004); however, the

CWFP can be divided into the length of time weeds can remain in the crop before yield loss occurs, and the length of time the crop must be kept weed-free to avert yield loss (Weaver and

Tan 1983). Corn is particularly vulnerable to weed interference during early vegetative growth stages, highlighting the benefit of effective soil-applied herbicide programs that prevent early- season weed emergence (Page et al. 2012; Green 2012); however, factors related to management or environment can affect onset and duration of the CWFP (Kropff and Spitters 1991; Gower et al. 2002). Selective post-emergence (POST) corn herbicides facilitate flexibility in application timing but vary in their length of residual control (Knezevic et al. 2002). Furthermore, the efficacy of several POST contact and systemic herbicides has been demonstrated to be affected by weed size or growth stage at the time of application (Steckel et al. 1997; Kegode and

Fronning 2005; Johnson and Norsworthy 2014; Soltani et al. 2016).

Contact herbicides including glufosinate, a glutamine synthetase inhibitor, and bentazon, a photosystem II (PSII)-inhibitor, are not widely-translocated within treated plants (Rojano-

Delgado et al. 2014; Stoller et al. 1975). Therefore, an inverse relationship between glufosinate and bentazon efficacy and weed size at time of application is widely reported in the literature, due to insufficient control of plant foliage not contacted by the herbicide during application

(Blackshaw 1989; Coetzer et al. 2002; Steckel et al. 1997; Stoller et al. 1975). A similar relationship has been found with some systemic herbicides. Johnson and Norsworthy (2014)

86 reported a decrease in control of johnsongrass (Sorghum halepense (L.) Pers.) with nicosulfuron, an acetolactate synthase (ALS)-inhibiting herbicide, when application was delayed from 15 to

30-45 or 60 cm timing. Similarly, a decline in the efficacy of clethodim, an acetyl coenzyme-A carboxylase (ACCase)-inhibitor, glyphosate, a 5-enolpyruvylshikimate 3-phosphate synthase inhibitor, and synthetic auxin herbicides including 2,4-D amine, dicamba and triclopyr + fluroxypyr has been reported in several weed species as weed size at the time of application increases (Johnson and Norsworthy 2014; Sellers et al. 2009; Soltani et al. 2016). In contrast,

Obrigawitch et al. (1990) reported that the efficacy of nicosulfuron on johnsongrass improved with an application at 30 or 60 cm compared to 10 cm. Similarly, several studies including

Corbett et al. (2004) and Johnson and Norsworthy (2014) have found weed size to have little effect on glyphosate efficacy, contributing to inconsistencies in this relationship across weed species, environments and herbicide active ingredients.

Tolpyralate is a Group 27 pyrazolone herbicide which inhibits the 4-hydroxyphenyl- pyruvate dioxygenase (HPPD) enzyme, impeding the biosynthesis of intermediates involved in the carotenoid biosynthetic pathway and subsequently leading to light-induced photodegradation of the photosynthetic complex and chlorophyll, manifesting as stark white bleaching of plant tissues and eventual plant death (Ahrens et al. 2013; Hawkes 2012). Tolpyralate, applied early-

POST (ePOST) to 10 cm weeds, has previously been found to exhibit efficacy on several annual grass and broadleaf weed species, particularly when co-applied with atrazine at a 1:33.3 ratio

(Metzger et al. 2018a). Atrazine disrupts passage of electrons through PSII by occupying the QB- binding niche, preventing normal binding of plastoquinone and causing proliferation of damaging reactive oxygen species and triplet chlorophyll; a mode of action that is interrelated with that of HPPD-inhibition (Hess 2000; Kim et al. 1999; Abendroth et al. 2006). The efficacy

87 of tolpyralate + atrazine relative to weed size at time of application has not previously been studied. Additionally, the efficacy of tolpyralate + atrazine applied pre-emergence (PRE) is largely unknown. Previous research with glufosinate, glyphosate, 2,4-D + dicamba, and bentazon has found that declining control of some weed species with increasing size can be overcome by increasing herbicide rate at later POST application timings (Johnson and Norsworthy 2014;

Steckel et al. 1997; King and Oliver 1992; Sellers et al. 2009; Soltani et al. 2016); however, it is unclear whether this relationship also exists with tolpyralate + atrazine. Therefore, the objective of this research was to determine the effect of application timing/weed size and herbicide rate on the efficacy of tolpyralate + atrazine tank-mixtures. An understanding of the interrelatedness of tolpyralate + atrazine rate with application timing, and the effect of these factors on overall herbicide efficacy, will aid in optimization of the application window for tolpyralate + atrazine in corn.

4.3 Materials and Methods

Experimental Methods Six experiments were conducted on field sites near Ridgetown and Exeter, Ontario,

Canada during the 2017 and 2018 growing seasons. Each experiment was organized as a two- factor randomized complete block, with application timing designated as Factor A, and herbicide rate as Factor B. Non-treated control (NTC) and weed-free control (WFC) plots were included within each level of Factor A. Weeds were controlled in WFC plots with the application of S- metolachlor/atrazine (2880 g ai ha-1) + mesotrione (140 g ai ha-1) PRE, glyphosate (900 g ae ha-1)

POST, and hand-hoeing as required. Field preparation at each experimental site consisted of fall moldboard plowing plus spring tillage with an s-tine field cultivator equipped with rolling basket harrows prior to planting. Sites were fertilized each spring according to provincially-accredited

88 soil test results and crop requirements. Plots were 3 m wide (four corn rows 76 cm apart), and 8 and 10 m long at Ridgetown and Exeter, respectively. Two glyphosate-resistant corn hybrids were selected based on length of the growing season at each location, and were DKC53-56RIB at

Ridgetown and DKC42-60RIB at Exeter (Monsanto Co., St. Louis, MO). Corn was seeded 4-5 cm deep at a population of 78,000 seeds ha-1 using a four-row conventional planter.

Treatments were applied using a CO2-pressurized backpack sprayer and 1.5 m hand boom equipped with four, ULD12002 nozzles (Pentair, New Brighton, MN, USA) spaced 50 cm apart, producing a spray width of 2 m. Treatments were applied in a spray volume of 187 L ha-1, at 255 kPa pressure. Herbicide treatments consisted of tolpyralate + atrazine applied at 15 + 500,

30 + 1000 and 40 + 1000 g ai ha-1, representing 0.5X, low and high label rates (Anonymous

2017a); hereafter, these rates are referred to as low, medium and high, respectively. Adjuvants included with tolpyralate applications were methylated seed oil (MSO Concentrate®, Loveland

Products, Loveland, CO, USA) at 0.50% v/v and urea ammonium nitrate at 2.50% v/v. Each rate of tolpyralate + atrazine was applied prior to crop/weed emergence (PRE), ePOST, mid-POST

(mPOST) and late POST (lPOST). Each POST application (ePOST, mPOST and lPOST) corresponded to average weed heights of 10, 20 and 30 cm, respectively, within NTC plots. Corn stage ranged from V3-V5 at ePOST, V5-V8 at mPOST and V6-V9 at lPOST, depending on experiment. Further details regarding the size of individual weed species at the time of each

POST application timing is presented in Table 4.1. Where a species was absent from an individual trial, or was present in insufficient density to provide meaningful data (<1-2 m-2), it was excluded and indicated by a dash (-) in Table 4.1.

Crop injury and weed control were assessed on a percent scale relative to the control plots, where 0 represents no control or injury and 100 indicates complete death of the crop or

89 weed. Crop injury was assessed 1, 2 and 4 weeks after emergence (WAE) for PRE applications, or 1, 2 and 4 weeks after application (WAA) for each POST application. Weed control of each species was assessed visually 2 and 4 WAE for PRE treatments, or 2 and 4 WAA for POST applications. At 8 weeks after lPOST applications, a final visible control assessment was conducted on all treatments. Weed density and aboveground biomass 4 weeks after lPOST applications were determined by species within a 0.5 m-2 quadrat placed at two arbitrary locations within each plot. Samples were kiln-dried at 60 C to constant mass and dry weight recorded.

Grain yield and harvest moisture was measured by harvesting the centre two rows of each plot with a small-plot combine. Grain yields were subsequently corrected to 15.5% moisture before analysis.

Statistical Analysis A mixed-model variance analysis was conducted on all response parameters for each weed species using the GLIMMIX procedure in SAS v. 9.4 (SAS Institute, Cary, NC). Variance was partitioned into the fixed effects of application timing (Factor A), herbicide rate (Factor B), and interactions therein, while environment (experiment), replication within environment, and the interaction of environment with Factor A and Factor B, were each designated as random effects; data were pooled across environments. Significance of the fixed effects was determined using an F-test, and significance of random effects determined using a restricted log-likelihood test. A significance level of α=0.05 was declared for all tests.

An appropriate distribution and link function were selected for each response parameter which best met assumptions that residuals were homogeneous, normally distributed and had a mean equal to zero, as determined by scatter plots of studentized residuals and normality plots

90 coupled with a Shapiro-Wilk test of normality. Non-Gaussian data was analyzed using the

Laplace method of integral approximation, which provides unbiased parameter estimation when the number of observations is small compared to default pseudo-likelihood models, and also facilitates direct comparison of various non-Gaussian models (Schabenberger 2007).

Least-square means of each response parameter for individual weed species were back- transformed by specifying the ilink option within the GLIMMIX procedure. Where a lognormal distribution was specified, data were back-transformed from the analysis scale using the omega procedure (M. Edwards, Ontario Agricultural College Statistician, University of Guelph, personal communication). Least-square means of Factor A, Factor B, and the interaction effect therein were separated using Tukey-Kramer’s multiple range test, with Type I error set to

α=0.05. Letter codes were assigned for presentation in tables using the pdmix800 macro (Bowley

2015) and slicediff commands. Where there was no statistically significant interaction between factors for a given weed species and assessment parameter, only the main effects are presented.

4.4 Results and Discussion

Weed Control

Common Ragweed Common ragweed (Ambrosia artemisiifolia L.) control 2 and 4 WAA was affected by herbicide rate, though the response depended on timing, resulting in a significant herbicide rate by application timing interaction (P<0.0001; Table 4.2). At 2 and 4 WAA, control of common ragweed was ≤56% with tolpyralate + atrazine applied PRE at all three rates; control improved as herbicide rate was increased from low to medium to high at 2 WAA, and from low to medium or high at 4 WAA (Table 4.3). Atrazine applied PRE controls several broadleaf weed species,

91 including common ragweed (Ontario Ministry of Agriculture, Food and Rural Affairs

[OMAFRA], 2018a); however, the increase in control observed 2 WAA with tolpyralate rate in this study indicates that tolpyralate applied PRE also has residual activity on this species. Since common ragweed control 4 WAA was equal (56%) with medium and high rates of tolpyralate applied PRE, which each included 1000 g ha-1 atrazine, the residual control provided by tolpyralate in this study appeared to be relatively short-lived. The results of this study are similar to other studies from Ontario, where control of common ragweed was variable (<80%) with atrazine applied at 1000 g ha-1 with S-metolachlor (Swanton et al. 2007). At 2 and 4 WAA, tolpyralate + atrazine applied ePOST, mPOST or lPOST provided better common ragweed control (≥85%) than PRE applications (≤56%); no differences were observed across POST application timings (Table 4.3). At 2 WAA, all three rates of tolpyralate + atrazine provided equivalent control when applied ePOST, while medium and high rates provided superior control compared to the low rate at mPOST and lPOST timings, when common ragweed was 6-22 and

14-62 cm tall, respectively. These results are consistent with other studies, where tolpyralate and tolpyralate + atrazine controlled common ragweed >90% (Sprague and Powell 2014; Tonks et al.

2015) and tolpyralate + atrazine (30 + 1000 g ha-1) applied ePOST controlled common ragweed

99% 2 WAA (Metzger et al. 2018b). Similarly, common ragweed was reported to be among the most sensitive species to tolpyralate based on a previous biologically-effective dose study examining eight weed species (Metzger et al. 2018a); however, it was unclear in that study how efficacy is impacted by size of common ragweed at the time of application.

At 8 WAA, common ragweed control was improved by 9% with the medium or high rate of tolpyralate + atrazine compared with the low rate when averaged across application timings; the rate by timing interaction was not significant (P=0.8155; Table 4.2). Common ragweed

92 control with ePOST, mPOST or lPOST applications was equivalent (≥93%), while control with the PRE application (23%) was less than with any POST timing.

There was a greater decrease in common ragweed density and biomass concomitant with tolpyralate + atrazine rate when applied at the POST timings than at the PRE timing, contributing to a statistically significant interaction between rate and timing for both parameters (P<0.0001;

Table 4.2). The application of tolpyralate + atrazine ePOST, mPOST or lPOST reduced common ragweed density and dry biomass similarly; all POST timings were superior to PRE within each respective rate, with the exception of the low rate applied ePOST for common ragweed density

(Table 4.3). Applied PRE, tolpyralate + atrazine did not reduce common ragweed density. In contrast, the high rate of tolpyralate + atrazine applied PRE reduced common ragweed dry biomass compared to the NTC. Within POST timings, all herbicide treatments reduced density and dry biomass by more than 99% compared to the NTC. With ePOST and mPOST applications, the medium and high rates of tolpyralate + atrazine provided similar reductions in density and dry biomass, and were statistically superior to the low rate. Applied lPOST, the high rate of tolpyralate + atrazine reduced density and dry biomass more than the low rate. Therefore, although there was minimal effect of common ragweed size at the time of application, when tolpyralate + atrazine was applied lPOST, there was a small numeric improvement in control with higher rates. Similarly, previous studies have reported improved control of larger, more mature weeds with higher herbicide rates (Blackshaw 1989; Lee and Oliver 1982; Steckel et al.

1997).

Common Lambsquarters Common lambsquarters (Chenopodium album L.) was controlled PRE with tolpyralate + atrazine; however, control was affected by rate, causing a significant rate by application timing

93 interaction for each assessment parameter with the exception of control 8 WAA (P=0.1600;

Table 4.4). Common lambsquarters control was >90% with all treatment combinations evaluated

2 WAA in this study, although a rate response was observed with both the PRE and lPOST application timing (Table 4.5). These results are similar to those of a previous study (Metzger et al. 2018a) which determined that the biologically-effective dose of tolpyralate + atrazine for 90% control of common lambsquarters is low (3.6 + 121 g ha-1) 2 WAA. Applied PRE, the low rate of tolpyralate + atrazine controlled common lambsquarters 91% 2 WAA but declined to 73% 4

WAA. At both 2 and 4 WAA, tolpyralate + atrazine applied PRE at the medium or high rate controlled common lambsquarters similarly; both provided better control than the low rate.

Given the similarity of the medium and high rate treatments which both included atrazine at

1000 g ha-1, compared to the low rate which included atrazine at 500 g ha-1, it is probable that lambsquarters control PRE is correlated with the rate of atrazine. Previous research has demonstrated atrazine to be highly efficacious on common lambsquarters applied PRE at rates of

1120 g ha-1, and less effective when applied at 280 or 560 g ha-1 (Bollman et al. 2006). Because tolpyralate was not applied alone in this study, it was not possible to determine the relative contribution of each herbicide for PRE control of a particular species. Applied lPOST, the high rate of tolpyralate + atrazine provided superior control to the low rate 2 WAA; no differences in rate were observed 4 WAA. Similarly, each rate of tolpyralate + atrazine controlled common lambsquarters ≥92% 2 and 4 WAA when applied ePOST or mPOST, and there was no improvement in control with the higher rates. Applied at the low rate, the ePOST application was superior to a PRE application at both 2 and 4 WAA. In contrast, tolpyralate + atrazine at the medium or high rate provided equivalent common lambsquarters control whether applied PRE, ePOST, mPOST or lPOST. Overall, these results generally corroborate previous findings

94 reported in Metzger et al. (2018a) that tolpyralate + atrazine exhibits high herbicidal activity in common lambsquarters POST.

Common lambsquarters density was reduced with both medium and high rates of tolpyralate + atrazine applied PRE; however, the low rate was not different from the NTC, presumably due to natural variation in common lambsquarters density within experiments (Table

4.5). In contrast, all rates applied PRE reduced common lambsquarters biomass compared with the NTC; the largest reduction in biomass was observed where the medium and high rates were applied. All POST rate and application timing combinations reduced common lambsquarters density and biomass ≥99% compared to the NTC. Applied ePOST, tolpyralate + atrazine at the medium rate reduced common lambsquarters density and dry biomass to near zero, which was similar to the high rate, and statistically superior to the low rate. Consistent with the current findings, tolpyralate + atrazine applied ePOST at 30 + 1000 g ha-1 in a previous experiment was found to reduce common lambsquarters biomass to near zero (Metzger et al. 2018b). At both the mPOST and lPOST timing, tolpyralate + atrazine provided a similar reduction in density and dry biomass compared with the NTC regardless of rate. At the low rate, an ePOST application resulted in a greater density and dry biomass reduction than a PRE application, while mPOST and lPOST timings were similar to the earlier timings. With the medium or high rates of tolpyralate + atrazine, density and dry biomass were slightly lower when applied ePOST compared to lPOST; however, a greater response to common lambsquarters size has been reported with other herbicides. Soltani et al. (2016), reported that a higher dose of glyphosate was required to reduce common lambsquarters biomass when applied to 30 cm-tall plants compared to 10 cm-tall plants. The small numerical differences observed between ePOST and lPOST timings in the current study likely repudiate the biological significance of the statistical

95 difference. Accordingly, visible control data collected 8 WAA showed no significant difference in application timing when averaged across rates, while the medium and high rates of tolpyralate

+ atrazine provided superior control to the low rate when averaged across application timings

(Table 4.4).

Green Pigweed Green pigweed (Amaranthus powelli S. Watson) was susceptible to tolpyralate + atrazine irrespective of application rate or timing; the interaction was not significant for any assessment parameter (P>0.0931; Table 4.6). At 2, 4 and 8 WAA, control with PRE, ePOST, mPOST and lPOST applications was similar. When averaged across application timings, tolpyralate + atrazine applied at the medium and high rates provided equivalent green pigweed control (88-

92%) and were consistently superior to control with the low rate. Interestingly, the biologically- effective dose of tolpyralate + atrazine for 90% control of a mixed population of green pigweed and redroot pigweed (Amaranthus retroflexus (L.)) was determined to be less than 15 + 500 g ha-

1 at 2, 4 and 8 WAA, although applications were only made ePOST in that study (Metzger et al

2018a); it is likely that the low rate was not sufficient for 90% control when averaged across

PRE, ePOST, mPOST and lPOST timings, as it was in the current study. Both herbicide rate and application timing were found to have significant effects on green pigweed density and dry biomass. Averaged across timings, there were no differences in density where low, medium or high rates of tolpyralate + atrazine were applied, while dry biomass was lower with the medium or high rate compared to the low rate. Averaged across rates, tolpyralate + atrazine applied PRE resulted in a greater reduction in green pigweed density and dry biomass than the lPOST application, while ePOST and mPOST applications were similar to either PRE or lPOST application timings. As with common lambsquarters, it is difficult to ascertain the relative contributions of tolpyralate and atrazine to residual green pigweed control with PRE

96 applications. Atrazine typically controls green pigweed (OMAFRA 2018a), although biotypes resistant to triazine herbicides are well-documented in Ontario (Diebold et al. 2003).

Consequently, it is possible that tolpyralate did contribute to residual control of green pigweed in this study.

Velvetleaf Velvetleaf (Abutilon theophrasti Medik.) control 2 and 4 WAA, and reduction in density and biomass varied with rate depending on the application timing; the interaction was significant

(P≤0.0138; Table 4.7). At 8 WAA however, only the application timing had a significant effect on control, and the interaction with rate was not significant (P=0.4757; Table 4.7). At 2 and 4

WAA, tolpyralate + atrazine applied at each POST timing provided better velvetleaf control than when applied at the PRE timing (Table 4.8). At 2 and 4 WAA, the medium and high rates of tolpyralate + atrazine controlled velvetleaf better than the low rate when applied PRE; however, control with all rates applied PRE was ≤45%. Consistent with these results, poor velvetleaf control with atrazine applied PRE alone has been reported previously (Bollman et al. 2006).

Applied ePOST, tolpyralate + atrazine controlled velvetleaf equally 2 and 4 WAA regardless of rate. Delaying application to mPOST resulted in a rate response from low to medium or high 2

WAA, while only the high rate of tolpyralate + atrazine improved control when application was delayed to lPOST. By 4 WAA, all three rates of tolpyralate + atrazine controlled velvetleaf equally when applied POST; delaying application from ePOST to mPOST or lPOST did not impact control. The excellent velvetleaf control observed with lPOST applications of tolpyralate

+ atrazine may have been partially due to the extended emergence pattern of velvetleaf (Mitich

1991), which meant more late-emerging seedlings were present at the late application timing.

However, velvetleaf plants were 28-42 cm tall on average at lPOST timing (Table 4.1). These

97 results provide corroborating evidence that velvetleaf is highly sensitive to tolpyralate + atrazine applied POST, as was reported in Metzger et al. (2018a).

At 8 WAA, tolpyralate + atrazine applied ePOST, mPOST or lPOST controlled velvetleaf

>90% irrespective of rate; whereas PRE applications resulted in near zero control of this species when averaged across rates (Table 4.7). Similarly, tolpyralate + atrazine applied PRE at the low, medium or high rate did not reduce velvetleaf density or dry biomass compared with the NTC

(Table 4.8). Applied ePOST, the medium and high rate of tolpyralate + atrazine reduced density compared to the NTC, while densities in plots treated with the low rate did not differ from either the NTC or the medium and high rate treatments. It is likely that these results are due to the natural variation in velvetleaf density within the experimental areas. Conversely, biomass was reduced to zero regardless of rate when applied ePOST. Similar trends were observed with mPOST applications; only medium and high rates reduced velvetleaf density, while all three rates reduced biomass, though a greater reduction in biomass was achieved with medium and high rates. Applied lPOST, all three rates of tolpyralate + atrazine provided a similar reduction in both density and biomass; each was lower than in the NTC. Overall, the results presented here are consistent with those of Metzger et al. (2018b) and Tonks et al. (2015), which observed excellent velvetleaf control with tolpyralate + atrazine tank mixtures applied ePOST; however, the current study demonstrates that velvetleaf control is maintained regardless of the POST application timings evaluated in this study.

Ladysthumb Ladysthumb (Persicaria maculosa Gray) emerged later than other broadleaf species; it was not present in the cohort of weeds assessed 2 weeks after PRE applications in the two experiments where it was assessed (Table 4.9). Staniforth and Cavers (1979) previously

98 described an extended, segmented emergence pattern with ladysthumb populations in Ontario.

Presumably due to this delayed emergence, ladysthumb was uncompetitive with other weed species, leading to variation in control with all herbicide treatments and application timings.

Askew and Wilcut (2002) also observed ladysthumb to be poorly competitive and exhibit slow early-season growth relative to other weed species in cotton. At 2 WAA, both herbicide rate and application timing had an effect on ladysthumb control; however, these factors did not interact

(P=0.0732; Table 4.9). Tolpyralate + atrazine applied at the high rate controlled ladysthumb better than the low rate when averaged across application timings, while the medium rate did not differ from either the low or high rate. When rates were averaged, ladysthumb control declined as POST applications were delayed; the ePOST timing provided greater control (94%) than the lPOST timing (62%). In another study, applications of glufosinate provided greater control of ladysthumb when applied to 2-5 cm plants, compared to when application was delayed to 8-10 cm plants (Corbett et al. 2004). Ladysthumb has also previously been found to be more tolerant to tolpyralate alone than other annual broadleaf species including velvetleaf, common lambsquarters and common ragweed (Metzger et al. 2018a). In contrast, atrazine controls ladysthumb when applied POST (OMAFRA 2018a), suggesting that the level of control observed in this study was predominately provided by atrazine rather than tolpyralate.

Ladysthumb control 4 WAA and the density and biomass data were variable and were unaffected by either rate or timing (P>0.0604; Table 4.9); no interaction was observed between main effects (P>0.0907; Table 4.9). In contrast, ladysthumb control 8 WAA generally declined with delayed POST application depending on rate, which produced a significant interaction

(P=0.0364; Table 4.9). When applied at the low rate, ladysthumb control with tolpyralate + atrazine was poor (≤53%) and no differences were observed across application timings (Table

4.10). Similarly, rate had no effect on tolpyralate + atrazine efficacy for this species when applied PRE; control was ≤28%. Applied at the medium or high rate however, tolpyralate + atrazine controlled ladysthumb more effectively when applied ePOST compared to PRE, though control with each of the POST timings was equivalent. Overall, ladysthumb control was highest with the medium or high rate of tolpyralate + atrazine applied ePOST; control was variable with lPOST applications and the three rates provided similar control. Based on the results of this study and previous determination of the biologically-effective dose of tolpyralate (Metzger et al.

2018a), it is evident that ladysthumb exhibits higher tolerance to tolpyralate compared with aforementioned broadleaf weed species (common lambsquarters, common ragweed, green pigweed). Similarly, Schӧnhammer et al. (2006) reported variable control of a related species,

Polygonum convolvulus (L.), with topramezone, a comparable pyrazolone-type herbicide to tolpyralate.

Flower-of-an-hour Flower-of-an-hour (Hibiscus trionum L.) (FOH) was present at locations near Exeter in

2017 and 2018. Flower-of-an-hour can produce over 3000 seeds plant-1, each encased in an impermeable seed coat which contributes to extended viability in soil and a fractionated emergence pattern (Chacalis et al. 2008; Johnson et al. 2003; Westra et al. 1996; Walker et al.

2010). Due to these innate biological characteristics, FOH was observed to be highly opportunistic in this study; late-emerging cohorts often thrived where all other weed species were controlled with ePOST applications. As a result of this sporadic emergence, control was affected by application timing 2 WAA (P=0.0341; Table 4.11). Averaged across rates, tolpyralate + atrazine applied ePOST and mPOST controlled FOH 93 and 88%, respectively; each was superior to the control with PRE applications. Delaying application to the lPOST

100 timing resulted in FOH control that was similar to either a PRE, ePOST or mPOST application at

2 WAA.

At 4 WAA, application timing had a significant effect on FOH control, but this effect depended on rate, causing a significant rate by timing interaction (P=0.0010; Table 4.11).

Applied PRE, medium and high rates of tolpyralate + atrazine provided similar FOH control, and were 40 and 52% higher than the low rate, respectively; no rate provided >78% control when applied PRE (Table 4.12). Interestingly, Knezevic et al. (2009) reported ≥95% control of FOH 4

WAA with several soybean herbicides applied PRE, including metribuzin (420 g ha-1). When tolpyralate + atrazine was applied ePOST, control with medium and high rates was equal (84%), and greater than with the low rate (52%). When application was delayed to mPOST, the high rate of tolpyralate + atrazine resulted in improved FOH control compared with the low rate; but no rate response was detected with lPOST applications. Control with the low rate was higher at the lPOST timing compared with the PRE timing, while control with ePOST and mPOST applications was similar to other timings. This improved control from PRE to lPOST timing with the low rate may be a result of more late-emerging FOH plants being contacted with the herbicide at the lPOST application. At the medium and high rates there was no effect of application timing, which suggests that a degree of residual control of FOH was provided by tolpyralate, atrazine or the interaction of both herbicides applied in combination.

Control of FOH 8 WAA was variable during both years of study. Within tolpyralate + atrazine rates, FOH control was equivalent whether applied PRE, ePOST, mPOST or lPOST

(Table 4.12). Knezevic et al. (2009) observed a significant decline in control of FOH when glyphosate applications were delayed from ePOST (15 cm plants) to lPOST (45 cm plants). In contrast, FOH plants in this study were comparatively smaller (12 cm) on average at lPOST

101 timings (Table 4.1), presumably diminishing the effect of application timing. Within timings, a rate response was observed with applications made PRE, ePOST and mPOST, while the three rates provided similar control 8 WAA when applied lPOST. Generally, the high rate of tolpyralate + atrazine provided greatest control of FOH; when applied mPOST, the high rate provided better FOH control than the medium rate. Previously, full-season control (98%) of FOH in corn has been observed with bromoxynil + atrazine (310 + 630 g ai ha-1) applied to 3-5 leaf

FOH plants (Westra et al. 1990), which is numerically higher than the control observed with the medium or high rate of tolpyralate + atrazine applied at a similar timing (mPOST) in this study

(Table 4.12).

Given the relatively unpredictable emergence pattern of FOH, density was difficult to accurately correlate to treatment effects. Applications of the medium or high rate PRE did not reduce FOH density compared to the NTC, while applications of the low rate resulted in FOH densities higher than that of the NTC (Table 4.12). The reason for this increase in FOH density with the low rate of tolpyralate + atrazine is probably due to the opportunistic germination and emergence of FOH. Tolpyralate + atrazine applied PRE at the low rate controlled common lambsquarters (Table 4.5) and green pigweed (Table 4.6), reducing the overall density of the weed canopy and possibly allowing FOH to germinate and emerge. In contrast, a dense weed canopy remained in NTC plots throughout the season, potentially suppressing late-germinating cohorts of FOH. In corn, FOH has been observed to exhibit weak competitive ability with plants taller than itself (Westra et al. 1990). Applied ePOST, medium and high rates of tolpyralate + atrazine reduced FOH density similarly, while only the high rate significantly reduced FOH density when applications were delayed to mPOST (Table 4.12). Applied lPOST, medium and high rates each reduced FOH density more than the low rate; though only the medium rate

102 decreased density compared with the NTC. Dry biomass of FOH plants was low; herbicide rate was the only factor to have an effect on FOH dry biomass (P=0.0283; Table 4.11). Averaged across timings, only the high rate of tolpyralate + atrazine reduced FOH dry biomass compared with the NTC; although this reduction was equivalent to that with the low and medium rates.

Tolpyralate efficacy on FOH has not previously been reported in the literature; however, the US label indicates “partial control” of this species when co-applied with atrazine (≥560 g ha-1)

(Anonymous 2017a). Atrazine (1490 g ha-1) is reported to provide some control (70-79%) of

FOH when applied POST, but <50% control when applied PRE (OMAFRA 2018a), which is generally consistent with the results of this study.

Barnyardgrass Control of barnyardgrass (Echinochloa crus-galli (L.) P. Beauv.) was poor with PRE applications at all rates; however, PRE control was improved with increasing rate at 2 WAA, leading to a significant rate by timing interaction (P<0.0001; Table 4.13). While tolpyralate efficacy PRE has not previously been reported in the literature for any weed species, atrazine generally does not control barnyardgrass (Janak and Grichar 2016; OMAFRA 2018a). At 2

WAA, PRE applications of the medium or high rate of tolpyralate + atrazine controlled barnyardgrass only 49-54%, compared to 21% with the low rate (Table 4.14). Generally, control of barnyardgrass 2 WAA was highest with ePOST and mPOST application timings. The high rate of tolpyralate + atrazine provided 98% control when applied ePOST, and 88% control when applied mPOST, which were statistically similar. At both the ePOST and mPOST timing, the high rate of tolpyralate + atrazine controlled barnyardgrass better than the low rate.

Barnyardgrass control with tolpyralate + atrazine generally declined when application was delayed from ePOST to lPOST, regardless of rate; however, no difference in control was observed when application was delayed from ePOST to mPOST. All three POST timings were

103 superior to the PRE timing, except with the medium rate. At the medium rate, ePOST applications provided 97% control, which was similar to the mPOST timing and better than the lPOST timing. A similar trend was reported by King and Oliver (1992) when the application of imazaquin, an acetolactate synthase-inhibiting herbicide, was delayed from 2-14 cm to 30 cm barnyardgrass.

At 4 and 8 WAA, tolpyralate + atrazine applied at the medium and high rate controlled barnyardgrass similarly when averaged across timings; control was consistently poorer with the low rate, leading to no significant rate by timing interaction (P=0.1405, 4 WAA; P=0.8929, 8

WAA; Table 4.13). Each POST application timing controlled barnyardgrass better than the PRE timing at both 4 and 8 WAA. The ePOST application resulted in 94% and 86% control on average 4 and 8 WAA, which was superior to the lPOST timing, and equal to the mPOST timing.

Kikugawa et al. (2015) reported better control of barnyardgrass with tolpyralate alone (30 g ha-1) when applied to plants with 5-6 leaves, compared to those at the 8 leaf stage. Similarly, Soltani et al. (2016) reported a higher biologically-effective dose of glyphosate for control of barnyardgrass when application was delayed from 10 cm to 30 cm timing.

In agreeance with control data, a greater reduction in barnyardgrass density occurred where tolpyralate + atrazine was applied ePOST or mPOST compared to PRE, while the lPOST timing was similar to all other application timings regardless of rate. No rate by timing interaction occurred for barnyardgrass density (P=0.0568; Table 4.13). Across timings, the low rate of tolpyralate + atrazine did not reduce barnyardgrass density compared with the NTC, while the high rate provided a greater density reduction than the low rate. In contrast to density, barnyardgrass dry biomass was affected by rate at POST timings but not at the PRE timing; biomass was similar to the NTC with all rates applied PRE, leading to a significant interaction

104 between rate and application timing (P=0.0166; Table 4.13; Table 4.14). The low rate of tolpyralate + atrazine did not reduce barnyardgrass dry biomass relative to the NTC, regardless of timing (Table 4.14). Conversely, the medium and high rate each reduced dry biomass more than the low rate when applied ePOST. Within the mPOST timings, only the high rate reduced barnyardgrass biomass, while within the lPOST timing, only the medium rate reduced biomass compared to the NTC. Within rates, ePOST and mPOST applications consistently reduced barnyardgrass dry biomass more than a PRE application. No differences were observed among

POST application timings with either the low or medium rates of tolpyralate + atrazine; though a greater biomass reduction occurred where the high rate was applied mPOST compared to lPOST.

Overall, the results for barnyardgrass control obtained in this study are similar to those reported previously where tolpyralate was applied with atrazine ePOST at 30 + 1000 g ha-1 (Metzger et al.

2018b; Tonks et al. 2015). Tonks et al. (2015) reported 90% control of barnyardgrass 30 d after application of tolpyralate + atrazine, while 97% control was observed with the same treatment in

Metzger et al. (2018b). The results of this study indicate that control of barnyardgrass can be maintained when application is delayed to mPOST timing, despite barnyardgrass plants being up to 29 cm in height at this timing. However, control of barnyardgrass generally declines when applications are delayed to the lPOST timing, when plants were up to 48 cm in height, highlighting the importance of timely application for control of this species.

Green Foxtail Green foxtail (Setaria viridis (L.) P. Beauv.) was among the most common weed species across sites; it was present at each trial location in both years of study (Table 4.1). Green foxtail responded differently to tolpyralate + atrazine rate depending on application timing at both 2 and

4 WAA; these differences were also present in density and biomass assessments, which contributed to a significant main effect interaction for these parameters (P<0.0273; Table 4.15).

105

Similar to barnyardgrass, control of green foxtail with PRE applications of all rates of tolpyralate

+ atrazine was poor (<50%). Regardless, a rate response secondary to the rate of atrazine was present 2 WAA with PRE applications (Table 4.16). In contrast, medium and high rates applied

PRE suppressed green foxtail equally (38-43%) by 4 WAA. At both 2 and 4 WAA, within all

POST application timings, the medium and high rate of tolpyralate + atrazine controlled green foxtail similarly; both rates were superior to the low rate. Similar to the 94% control observed in this study 2 and 4 WAA, tolpyralate + atrazine applied ePOST at 30 + 1000 g ha-1 was previously found to control green foxtail 96% at the same timings (Metzger et al. 2018b). In contrast, a significant response to application timing was observed within each tolpyralate + atrazine rate 2 and 4 WAA. At each respective rate, ePOST, mPOST and lPOST application timings were superior to the PRE timing. At 2 WAA, control with the ePOST timing was better than with the lPOST timing. Similar results have been reported in Setaria spp. with other herbicides. Soltani et al. (2016) observed better control of green foxtail with glyphosate applied to 10 or 20 cm plants, compared with applications made to 30 cm plants. Likewise, Steckel et al.

(1997) observed erratic control of 15 cm giant foxtail (Setaria faberi Herrm.) with glufosinate, compared to when applications were made to 10 cm plants. In contrast, Corbett et al. (2004) reported ≥97% control of green foxtail with glyphosate and glufosinate regardless of whether applications were made at 2-5 cm or 8-10 cm timing.

Green foxtail control 8 WAA was influenced by rate and application timing; however, the main effects acted independently of one another at this timing (P=0.1059; Table 4.15). Similar to control assessments taken 2 and 4 WAA, the medium and high rate of tolpyralate + atrazine provided equivalent control, and each was superior to the low rate when averaged across application timings. Green foxtail control with PRE applications of any rate was poor (19%).

106

Conversely, control was similar (84 to 85%) with either an ePOST or mPOST application when averaged across rates, while control declined to 67% when application was delayed to the lPOST timing, a similar trend to that observed in barnyardgrass (Table 4.13). The results of this study indicate that control of annual grasses with tolpyralate + atrazine generally declines when applications are delayed beyond a mPOST timing. At the lPOST application timings, barnyardgrass and green foxtail were from the 4 tiller to 2nd node growth stage (Zadoks GS 24-

32), while at the mPOST timings no grasses were beyond the 4 tiller stage (Zadoks GS 24) (data not presented). Consequently, growth stage in addition to size could have affected control of barnyardgrass and green foxtail. Similarly, Johnson and Norsworthy (2014) observed a decline in control of johnsongrass with nicosulfuron when application was delayed from 15 to 60 cm-tall plants, and with clethodim when application was delayed from the boot to panicle stage. Green foxtail density and dry biomass were not reduced with any rate of tolpyralate + atrazine applied

PRE; low rates resulted in an increase in dry biomass compared with the NTC (Table 4.16).

Similar to the case with FOH, this increase in biomass with the low rate can very likely be attributed to control or suppression of certain species with tolpyralate + atrazine applied PRE.

Tolpyralate + atrazine controlled green pigweed and common lambsquarters PRE (Table 4.5;

Table 4.6), allowing greater light penetration through the weed canopy and reducing interspecific competition. Furthermore, the low rate of tolpyralate + atrazine provided no appreciable suppression of green foxtail (Table 4.16). At both the ePOST and mPOST timing, green foxtail density was only reduced with the medium and high rate of tolpyralate + atrazine; all three rates reduced dry biomass. At both the ePOST and mPOST application timings, medium and high rates of tolpyralate + atrazine reduced dry biomass more than the low rate, while no differences in biomass were observed across rates within the lPOST timing. In contrast, density was not

107 reduced with any rate of tolpyralate + atrazine applied lPOST. These divergent results indicate that green foxtail was only partially controlled with lPOST applications; stunted or injured plants remained in treated plots and were therefore included in density assessments. A similar discrepancy was acknowledged in Metzger et al. (2018a) for certain weed species, including common ragweed and barnyardgrass. An effect of application timing was observed within the medium and high tolpyralate + atrazine rates for both green foxtail density and dry biomass, and within the low rate for dry biomass. Applied at the low rate, tolpyralate + atrazine reduced green foxtail biomass most effectively when applied ePOST, though this timing was not significantly different than mPOST. Each POST application of the low rate reduced green foxtail dry biomass more effectively than the PRE timing; however, no differences were observed in green foxtail density across timings at this rate. Consistent with control assessments, the medium rate of tolpyralate + atrazine reduced green foxtail density and dry biomass more effectively when applied ePOST or mPOST compared to when it was applied PRE or lPOST. In contrast, densities with PRE and mPOST applications were similar when the high rate of tolpyralate + atrazine was applied; each was lower than when application was delayed to the lPOST timing. Despite the slight reduction in green foxtail density with the high herbicide rate applied PRE, biomass was highest in PRE-treated plots across all three rates. All POST applications reduced green foxtail dry biomass compared with the control, irrespective of timing; however, the ePOST timing provided a greater biomass reduction than the lPOST timing. Overall, dry biomass evaluations demonstrate that green foxtail control with tolpyralate + atrazine is generally highest with medium or high rate applications made at either an ePOST or mPOST timing. Similarly, control of green foxtail typically declines significantly when applications are delayed to lPOST (Table

4.15), a response similar to that observed in barnyardgrass (Table 4.13).

108

Crop Injury On average, crop injury was minor (<10%) with all herbicide treatment combinations; however, injury was >30% at one trial location in 2017. Injury consisted of surfactant burn/leaf scorch on the upper surface of corn leaves exposed at the time of the lPOST application, and gradually dissipated by 4-8 WAA. While most applications were made in the morning, this application occurred near noon, under warm temperatures (27 C), factors which could have contributed to the increased leaf burn observed. No injury was observed with the PRE application timings regardless of rate, while a stepwise increase in injury concomitant with rate was observed 1 WAA with the ePOST application; no difference was observed across rates within the mPOST or lPOST applications 1 WAA (Table 4.18). Therefore, across site-years, there was a significant rate by timing interaction for corn injury 1, 2 and 4 WAA (P<0.0010;

Table 4.17). At 2 WAA, the high rate of tolpyralate + atrazine induced greater crop injury than the low rate when applied ePOST or mPOST; however, by 4 WAA the greatest injury was observed where the high rate was applied either ePOST or lPOST. At this timing, injury was

<5% on average with all rate by application timing combinations, and would therefore be considered commercially acceptable.

Grain Yield Weed interference reduced corn grain yield an average of 66% in this study. Grain yield varied by site but was reflective of overall weed control; yield ranged from 4.0-4.5 t ha-1 in NTC plots, to 12.1-12.6 t ha-1 in WFC plots (Table 4.18). Grain yield was influenced by the interaction of herbicide rate and application timing (P<0.0001; Table 4.17); but regardless of application timing, all rates of tolpyralate + atrazine resulted in corn grain yields higher than that of the

NTC. Applied PRE, the medium or high tolpyralate + atrazine rate resulted in higher yields than the low rate; however, grain yields were lower for all rates of tolpyralate + atrazine applied PRE

109 compared to the WFC. These results were expected for PRE treatments; tolpyralate + atrazine failed to adequately control six of the eight weed species evaluated when applied PRE. In contrast, grain yields were not different among the WFC, and low, medium and high rates of tolpyralate + atrazine when applied ePOST. When application occurred at the mPOST timing, there were no grain yield differences across the three rates of tolpyralate + atrazine; however, yields with the low herbicide rate was less than yield in the WFC plots. When application was further delayed to lPOST, no rate response was present; none of the applied rates of tolpyralate + atrazine reduced weed interference sufficiently to achieve yields similar to that of the WFC, likely because of the longer period of weed interference prior to application. Applied at the low rate, each of the POST application timings resulted in higher grain yield than the PRE timing; while PRE and lPOST applications resulted in similar corn grain yields when either the medium or high rate of tolpyralate + atrazine was applied. Several other studies have demonstrated the importance of early-season weed control in corn (Hall et al. 1992; Page et al. 2012; Swanton and

Weise 1991; Tursun et al. 2016; Norsworthy and Oliveira 2004). In these and other studies, authors have investigated the critical period of weed control (CPWC): the time period during crop growth where weeds must be controlled to avoid yield loss (Knezevic et al. 2002). Although it is subject to influence by a number of factors, Norsworthy and Oliveira (2004) reported the

CPWC to begin as early as the one to two leaf stage of corn, while Page et al. (2012) found the

CPWC to begin between the third and fifth leaf stages of corn, generally corresponding to when ePOST treatments were applied in this study. Additionally, Gantoli et al. (2013) found that corn yield loss increased concurrently with the amount of time that weeds were allowed to remain in the crop, a finding that supports the declining yield trend observed with lPOST applications in this study. Averaged across rates, there was a 4% decrease in corn yield incurred by delaying

110 herbicide application from ePOST to mPOST, and a 13% decrease incurred by delaying from ePOST to lPOST. Overall, these results give further supporting evidence for the importance of timely application of POST herbicides in corn to minimize yield loss, regardless of specific herbicide efficacy on larger weeds.

4.5 Conclusions

Based on this study, weed control with tolpyralate + atrazine depends on application timing, though the magnitude of difference in control depends on weed species. For five species: common ragweed, common lambsquarters, velvetleaf, barnyardgrass and green foxtail, the effect of tolpyralate + atrazine rate depended on the timing 2 WAA (Tables 4.2, 4.4, 4.7, 4.13, 4.15).

For each of these species (with the exception of barnyardgrass), and for FOH, a rate by timing interaction also occurred 4 WAA (Tables 4.2, 4.4, 4.7, 4.11, 4.15). In contrast, the effects of application rate and timing were independent at 8 WAA for all species except ladysthumb and

FOH. Despite its current registration as a POST-only herbicide, tolpyralate controlled common lambsquarters and green pigweed similarly (≥89%) 2, 4 and 8 WAA when applied PRE or ePOST at current label rates with atrazine (Table 4.4-4.6). Similarly, PRE applications at either the medium or high rate resulted in corn grain yields that were similar to those obtained with the same rates applied lPOST. Although control of common ragweed, velvetleaf, ladysthumb, FOH, barnyardgrass and green foxtail was commercially unacceptable (<80%) with tolpyralate + atrazine applied PRE in this study, a rate response secondary to the rate of atrazine was detected in common ragweed and green foxtail 2 WAA (Tables 4.3 and 4.16), suggesting that tolpyralate may provide limited, short-term residual control of select weed species. Ultimately, the results

111 presented here warrant future research examining the efficacy of tolpyralate applied PRE alone, without the confounding effect of atrazine.

Within the three POST application timings, no species showed a significant decrease in control 2 or 4 WAA with a low, medium or high rate of tolpyralate + atrazine applied mPOST compared to ePOST (Tables 4.2-4.16). Similarly, corn grain yields were equivalent with either an ePOST or mPOST application. In contrast, control of green foxtail and barnyardgrass 2 and 4

WAA, and control of ladysthumb 2 WAA, each declined with all three rates of tolpyralate + atrazine when application was delayed from ePOST to lPOST (Tables 4.9, 4.13, 4.14, 4.16). In agreeance, grain yields also declined when application of each rate of tolpyralate + atrazine was delayed from ePOST or mPOST to lPOST. In the case of FOH, which exhibited a temporal emergence pattern and opportunistic growth habit, control 4 WAA with the low rate of tolpyralate + atrazine was numerically higher with the lPOST application than with the ePOST application (Table 4.12). Because this control assessment was made 4 weeks after lPOST applications, it is likely that any residual control of FOH provided by the low rate of tolpyralate

+ atrazine applied ePOST had diminished. This likely allowed late-emerging FOH to flourish relative to plots which received an lPOST application, which intercepted late-emerging seedlings. At 8 WAA, only barnyardgrass and green foxtail exhibited a response to POST application timing; visible control declined from ePOST to lPOST, results which were generally reflected in density and dry biomass assessments, and in final corn grain yields. Additionally, this decline in control and crop yield could not be overcome by increasing the herbicide rate.

Although control of ladysthumb and FOH was not statistically different with ePOST or lPOST applications, the numerical decline in control with delayed application suggests that control of these species with tolpyralate + atrazine at label rates can be optimized with an ePOST

112 application. Interestingly, common ragweed, common lambsquarters, velvetleaf, and green pigweed could be controlled equally regardless of POST application timing, substantiating the results presented in Metzger et al. (2018a), which lists these among the most sensitive species to tolpyralate.

Applied POST, the high rate of tolpyralate + atrazine generally provided no improvement in control or corn yield relative to the medium rate. One exception, however, was control of FOH

8 WAA with mPOST applications, which was significantly improved with the high herbicide rate compared with the medium rate (Table 4.12). Yield however, was equivalent with low, medium and high rates within each POST application timing, suggesting that late-season FOH interference had little impact on corn grain yield. Weed control was generally poorer with the low rate of tolpyralate + atrazine relative to the high rate and in some cases the medium rate, and grain yield with the low rate applied mPOST and lPOST was lower than the WFC. The high tolpyralate + atrazine rate generally resulted in the highest numerical values for control and the greatest reduction in density/dry biomass, possibly indicating that consistency of weed control is improved with the high rate. Despite having no impact on grain yield in this study, more consistent weed control can reduce the likelihood of weed seed return to the soil, and potentially have important implications for weed management in subsequent seasons.

Overall, this study provides further insight into the interspecific sensitivities of weeds to tolpyralate, and facilitates the development of an appropriate application window for this herbicide when co-applied with atrazine to target the CPWC. Furthermore, results from PRE applications suggest that in the future, tolpyralate may be useful as a residual herbicide for control of certain species. Additional research in this regard will assist in maximizing the value of this herbicide to primary producers.

113

Table 4.1. POST application dates and average size of common ragweed (AMBEL), common lambsquarters (CHEAL), green pigweed (AMAPO), velvetleaf (ABUTH), ladysthumb (POLPE), flower-of-an-hour (HIBTR), barnyardgrass (ECHCG) and green foxtail (SETVI) within experiments at time of each POST application. Spray Weed height (cm) Triala Timing date AMBEL CHEAL AMAPO ABUTH POLPE HIBTR ECHCG SETVI ePOSTb Jun-21 8.5 11 12 9 7 -c 17 13 E1 mPOST Jun-26 22 19 24 18 8 - 21 18 lPOST Jun-30 36 38 38 34 21 - 42 35 ePOST Jun-17 6 9 10 8 - - 15 13 E2 mPOST Jun-21 18 18 20 17 - - 18 23 lPOST Jun-26 27 28 28 28 - - 27 29 ePOST Jun-13 4 2 - - - 2 - 8 E3 mPOST Jun-19 6 11 - - - 6 - 13 lPOST Jun-28 14 20 - - - 12 - 20 ePOST Jun-21 9 - - 9 - - 18 20 E4 mPOST Jun-26 11 - - 10 - - 29 21 lPOST Jul-3 62 - - 42 - - 48 40 ePOST Jun-6 8 7 5 9 4 - 15 13 E5 mPOST Jun-19 20 18 18 25 15 - 25 29 lPOST Jun-26 37 30 37 37 32 - 39 38 ePOST Jun-7 6 5 - - - 5 7 11 E6 mPOST Jun-15 9 10 - - - 6 11 21 lPOST Jun-21 17 21 - - - 12 15 31 aE1, E2 designates Ridgetown 2017; E4, E5 designates Ridgetown 2018; E3, E6 designates Exeter 2017, 2018, respectively. bAbbreviations: ePOST, early-postemergence; mPOST, mid-postemergence; lPOST, late-postemergence. cA dash (-) indicates that the species was absent from a particular trial location in a given year.

114

Table 4.2. Effect of rate and application timing on common ragweed (Ambrosia artemisiifolia L.) control 2, 4 and 8 WAE/WAA, and density and dry biomass reduction with tolpyralate + atrazine in field experiments conducted in Ontario, Canada in 2017/2018a. Main effects Visible control (%) Dry Density Rate (g ai ha-1) 2 WAAb 4 WAAc 8 WAAc biomass (no. m-2) (g m-2) 0 - - - 19.8 60.5 15 + 500 75 77 70 b 0.27 0.20 30 + 1000 83 88 79 a 0.00 0.01 40 + 1000 86 88 79 a 0.00 0.01 Rate p-value 0.0002 0.0003 0.0005 <0.0001 <0.0001 Timing PRE 41 48 23 b 4.55 14.5 ePOST 97 96 94 a 0.04 0.03 mPOST 96 97 95 a 0.03 0.03 lPOST 91 96 93 a 0.07 0.09 Timing p-value <0.0001 0.0003 <0.0001 0.0008 0.0001 Interaction Rate x timing p-value <0.0001 <0.0001 0.8155 <0.0001 <0.0001 Means within a column followed by the same letter are not significantly different according to Tukey’s multiple means comparison (α=0.05). aAbbreviations: WAE/WAA, weeks after emergence/weeks after application; PRE, pre-emergence; ePOST, early- postemergence; mPOST, mid-postemergence; lPOST, late-postemergence. bControl with PRE applications was assessed 2 wk after crop emergence. cAssessed 4 and 8 weeks after the lPOST application.

115

Table 4.3. Interaction of herbicide rate and application timing on control of common ragweed (Ambrosia artemisiifolia L.) 2 and 4 WAE/WAA, and density and dry biomass reduction with three rates of tolpyralate + atrazine applied PRE, ePOST, mPOST or lPOST in field experiments conducted in Ontario, Canada in 2017/2018a.

Visible control (%) 2 WAAb Application timing Rate PRE ePOST mPOST lPOST (g ai ha-1) 15 + 500 31 c Y 94 a Z 92 b Z 85 b Z 30 + 1000 43 b Y 98 a Z 97 a Z 93 a Z 40 + 1000 50 a Y 99 a Z 98 a Z 96 a Z

Visible control (%) 4 WAA Application timing Rate PRE ePOST mPOST lPOST (g ai ha-1) 15 + 500 32 b Y 92 a Z 93 a Z 92 a Z 30 + 1000 56 a Y 98 a Z 98 a Z 98 a Z 40 + 1000 56 a Y 99 a Z 98 a Z 98 a Z

Density (no. m-2) Application timing Rate PRE ePOST mPOST lPOST (g ai ha-1) 0 20.7 a 40.9 c 35.0 c 19.9 c 15 + 500 9.01 a Y 0.34 b YZ 0.11 b Z 0.06 b Z 30 + 1000 2.04 a Y 0.00 a Z 0.00 a Z 0.01 ab Z 40 + 1000 2.23 a Y 0.00 a Z 0.00 a Z 0.00 a Z

Dry biomass (g m-2) Application timing Rate PRE ePOST mPOST lPOST (g ai ha-1) 0 60.9 c 168 c 85.4 c 61.2 c 15 + 500 28.7 bc Y 0.11 b Z 0.05 b Z 0.05 b Z 30 + 1000 8.24 bc Y 0.00 a Z 0.00 a Z 0.01 ab Z 40 + 1000 5.70 ab Y 0.00 a Z 0.00 a Z 0.00 a Z Means followed by the same lower case letter within a column (a-c), or upper case letter within a row (Y-Z) for each assessment parameter are not significantly different according to Tukey’s multiple means comparison test (α=0.05). aAbbreviations: WAE/WAA, weeks after emergence/weeks after application; PRE, pre-emergence; ePOST, early- postemergence; mPOST, mid-postemergence; lPOST, late-postemergence. bControl with PRE applications was assessed 2 weeks after crop emergence.

116

Table 4.4. Effect of rate and application timing on common lambsquarters (Chenopodium album L.) control 2, 4 and 8 WAE/WAA, and density and dry biomass reduction with tolpyralate + atrazine in field experiments conducted in Ontario, Canada in 2017/2018a. Main effects Visible control (%) Dry Density Rate (g ai ha-1) 2 WAAb 4 WAAc 8 WAAc biomass (no. m-2) (g m-2) 0 - - - 47.4 69.1 15 + 500 94 88 88 b 0.26 0.12 30 + 1000 97 95 94 a 0.01 0.01 40 + 1000 98 96 94 a 0.04 0.02 Rate p-value 0.0067 0.0196 0.0237 <0.0001 <0.0001 Timing PRE 95 85 92 0.32 0.35 ePOST 99 98 96 0.04 0.03 mPOST 97 95 91 0.24 0.13 lPOST 95 94 88 1.33 0.72 Timing p-value 0.1678 0.0573 0.1118 0.0106 0.0137 Interaction Rate x timing p-value 0.0240 <0.0001 0.1600 0.0002 <0.0001 Means within a column followed by the same letter are not significantly different according to Tukey’s multiple means comparison (α=0.05). aAbbreviations: WAE/WAA, weeks after emergence/weeks after application; PRE, pre-emergence; ePOST, early- postemergence; mPOST, mid-postemergence; lPOST, late-postemergence. bControl with PRE applications was assessed 2 weeks after crop emergence. cAssessed 4 and 8 weeks after the lPOST application.

117

Table 4.5. Interaction of herbicide rate and application timing on control of common lambsquarters (Chenopodium album L.) 2 and 4 WAE/WAA, and density and dry biomass reduction with three rates of tolpyralate + atrazine applied PRE, ePOST, mPOST or lPOST in field experiments conducted in Ontario, Canada in 2017/2018a.

Visible control (%) 2 WAAb Application timing Rate PRE ePOST mPOST lPOST (g ai ha-1) 15 + 500 91 b Y 98 a Z 96 a YZ 93 b YZ 30 + 1000 96 a Z 99 a Z 98 a Z 96 ab Z 40 + 1000 97 a Z 99 a Z 98 a Z 96 a Z

Visible control (%) 4 WAA Application timing Rate PRE ePOST mPOST lPOST (g ai ha-1) 15 + 500 73 b Y 97 a Z 92 a Z 91 a Z 30 + 1000 90 a Z 99 a Z 97 a Z 95 a Z 40 + 1000 92 a Z 99 a Z 97 a Z 97 a Z

Density (no. m-2) Application timing Rate PRE ePOST mPOST lPOST (g ai ha-1) 0 56.9 b 122 c 28.1 b 86.7 b 15 + 500 2.06 b Y 0.04 b Z 0.30 a YZ 0.65 a YZ 30 + 1000 0.01 a YZ 0.00 a Z 0.06 a XY 0.19 a X 40 + 1000 0.06 a YZ 0.00 ab Z 0.03 a Z 1.24 a Y

Dry biomass (g m-2) Application timing Rate PRE ePOST mPOST lPOST (g ai ha-1) 0 69.3 c 179 c 61.4 b 92.0 b 15 + 500 1.55 b Y 0.02 b Z 0.10 a YZ 0.22 a YZ 30 + 1000 0.01 a YZ 0.00 a Z 0.01 a YZ 0.08 a Y 40 + 1000 0.04 a XY 0.00 ab Z 0.01 a YZ 0.47 a X Means followed by the same lower case letter within a column (a-c), or upper case letter within a row (X-Z) for each assessment parameter are not significantly different according to Tukey’s multiple means comparison test (α=0.05). aAbbreviations: WAE/WAA, weeks after emergence/weeks after application; PRE, pre-emergence; ePOST, early- postemergence; mPOST, mid-postemergence; lPOST, late-postemergence. bControl with PRE applications was assessed 2 weeks after crop emergence.

118

Table 4.6. Effect of rate and application timing on green pigweed (Amaranthus powelli S. Watson) control 2, 4 and 8 WAE/WAA, and density and dry biomass reduction with tolpyralate + atrazine in field experiments conducted in Ontario, Canada in 2017/2018a. Main effects Visible control (%) Dry Density Rate (g ai ha-1) 2 WAAb 4 WAAc 8 WAAc biomass (no. m-2) (g m-2) 0 - - - 28 a 50 c 15 + 500 85 b 84 b 83 b 1.5 b 0.8 b 30 + 1000 88 a 92 a 91 a 0.3 b 0.1 a 40 + 1000 91 a 92 a 92 a 0.2 b 0.1 a Rate p-value 0.0054 0.0038 0.0079 0.0008 <0.0001 Timing PRE 74 91 94 0.1 a 0.1 a ePOST 94 93 89 1.2 ab 0.7 ab mPOST 94 91 87 3.0 ab 1.4 ab lPOST 88 83 84 13 b 8.5 b Timing p-value 0.6165 0.4053 0.2927 0.0220 0.0296 Interaction Rate x timing p-value 0.5155 0.1391 0.0931 0.2822 0.4523 Means within a column followed by the same letter are not significantly different according to Tukey’s multiple means comparison (α=0.05). aAbbreviations: WAE/WAA, weeks after emergence/weeks after application; PRE, pre-emergence; ePOST, early- postemergence; mPOST, mid-postemergence; lPOST, late-postemergence. bControl with PRE applications was assessed 2 weeks after crop emergence. cAssessed 4 and 8 weeks after the lPOST application.

119

Table 4.7. Effect of rate and application timing on velvetleaf (Abutilon theophrasti Medik.) control 2, 4 and 8 WAE/WAA, and density and dry biomass reduction with tolpyralate + atrazine in field experiments conducted in Ontario, Canada in 2017/2018a. Main effects Visible control (%) Dry Density Rate (g ai ha-1) 2 WAAb 4 WAAc 8 WAAc biomass (no. m-2) (g m-2) 0 - - - 1 2.40 15 + 500 76 71 66 0.2 0.11 30 + 1000 81 79 72 0.0 0.00 40 + 1000 84 80 73 0.0 0.00 Rate p-value 0.0134 0.1825 0.1434 0.0003 <0.0001 Timing PRE 39 22 3 b 0.6 1.09 ePOST 97 94 92 a 0.1 0.04 mPOST 93 94 94 a 0.0 0.02 lPOST 92 96 93 a 0.0 0.05 Timing p-value 0.0074 <0.0001 <0.0001 0.0048 0.0030 Interaction Rate x timing p-value 0.0005 0.0004 0.4757 0.0138 0.0017 Means within a column followed by the same letter are not significantly different according to Tukey’s multiple means comparison (α=0.05). aAbbreviations: WAE/WAA, weeks after emergence/weeks after application; PRE, pre-emergence; ePOST, early- postemergence; mPOST, mid-postemergence; lPOST, late-postemergence. bControl with PRE applications was assessed 2 weeks after crop emergence. cAssessed 4 and 8 weeks after the lPOST application.

120

Table 4.8. Interaction of herbicide rate and application timing on control of velvetleaf (Abutilon theophrasti Medik.) 2 and 4 WAE/WAA, and density and dry biomass reduction with three rates of tolpyralate + atrazine applied PRE, ePOST, mPOST or lPOST in field experiments conducted in Ontario, Canada in 2017/2018a.

Visible control (%) 2 WAAb Application timing Rate PRE ePOST mPOST lPOST (g ai ha-1) 15 + 500 31 b Y 96 a Z 89 b Z 88 b Z 30 + 1000 40 a Y 98 a Z 95 a Z 92 ab Z 40 + 1000 45 a Y 98 a Z 96 a Z 94 a Z

Visible control (%) 4 WAA Application timing Rate PRE ePOST mPOST lPOST (g ai ha-1) 15 + 500 10 b Y 91 a Z 91 a Z 91 a Z 30 + 1000 26 a Y 94 a Z 96 a Z 98 a Z 40 + 1000 30 a Y 96 a Z 97 a Z 99 a Z

Density (no. m-2) Application timing Rate PRE ePOST mPOST lPOST (g ai ha-1) 0 0.74 a 1.98 b 1.53 b 2.11 b 15 + 500 3.57 a Y 0.10 ab YZ 0.20 b YZ 0.05 a Z 30 + 1000 0.52 a Y 0.03 a YZ 0.00 a Z 0.01 a Z 40 + 1000 0.43 a Y 0.01 a YZ 0.00 a Z 0.01 a Z

Dry biomass (g m-2) Application timing Rate PRE ePOST mPOST lPOST (g ai ha-1) 0 1.7 a 3.4 b 4.6 c 5.8 b 15 + 500 4.2 a Y 0.0 a Z 0.1 b Z 0.1 a Z 30 + 1000 1.2 a Y 0.0 a Z 0.0 a Z 0.0 a Z 40 + 1000 0.6 a Y 0.0 a Z 0.0 a Z 0.0 a Z Means followed by the same lower case letter within a column (a-c), or upper case letter within a row (X-Z) for each assessment parameter are not significantly different according to Tukey’s multiple means comparison test (α=0.05). aAbbreviations: WAE/WAA, weeks after emergence/weeks after application; PRE, pre-emergence; ePOST, early- postemergence; mPOST, mid-postemergence; lPOST, late-postemergence. bControl with PRE applications was assessed 2 weeks after crop emergence.

121

Table 4.9. Effect of rate and application timing on ladysthumb (Persicaria maculosa Gray) control 2, 4 and 8 WAE/WAA, and density and dry biomass reduction with tolpyralate + atrazine in field experiments conducted in Ontario, Canada in 2017/2018a. Main effects Visible control (%) Dry Density Rate (g ai ha-1) 2 WAAb 4 WAAc 8 WAAc biomass (no. m-2) (g m-2) 0 - - - 1.7 1.40 15 + 500 70 b 50 37 4.3 1.68 30 + 1000 79 ab 71 61 0.2 0.09 40 + 1000 83 a 72 59 0.1 0.04 Rate p-value 0.0310 0.6320 0.2755 0.0796 0.0604 Timing PRE -d 40 22 0.1 0.05 ePOST 94 a 90 78 1.4 0.49 mPOST 77 ab 66 66 1.0 0.50 lPOST 62 b 68 44 1.3 0.57 Timing p-value 0.0414 0.4649 0.1980 0.1437 0.1731 Interaction Rate x timing p-value 0.0732 0.9994 0.0364 0.0907 0.1557 Means within a column followed by the same letter are not significantly different according to Tukey’s multiple means comparison (α=0.05). aAbbreviations: WAE/WAA, weeks after emergence/weeks after application; PRE, pre-emergence; ePOST, early- postemergence; mPOST, mid-postemergence; lPOST, late-postemergence. bControl with PRE applications was assessed 2 weeks after crop emergence. cAssessed 4 and 8 weeks after the lPOST application. dLadysthumb was not emerged 2 weeks after PRE applications.

122

Table 4.10. Interaction of herbicide rate and application timing on control of ladysthumb (Persicaria maculosa Gray) 8 WAA with three rates of tolpyralate + atrazine applied PRE, ePOST, mPOST or lPOST in field experiments conducted in Ontario, Canada in 2017/2018a.

Visible control (%) 8 WAA Application timing Rate PRE ePOST mPOST lPOST (g ai ha-1) 15 + 500 17 a Z 53 b Z 47 b Z 32 a Z 30 + 1000 28 a Y 87 a Z 80 a YZ 49 a YZ 40 + 1000 20 a Y 93 a Z 72 ab YZ 50 a YZ Means followed by the same lower case letter within a column (a-b), or upper case letter within a row (Y-Z) for each assessment parameter are not significantly different according to Tukey’s multiple means comparison test (α=0.05). aAbbreviations: WAA, weeks after application; PRE, pre-emergence; ePOST, early-postemergence; mPOST, mid- postemergence; lPOST, late-postemergence.

123

Table 4.11. Effect of rate and application timing on flower-of-an-hour (Hibiscus trionum L.) control 2, 4 and 8 WAE/WAA and density and dry biomass reduction with tolpyralate + atrazine in field experiments conducted in Ontario, Canada in 2017/2018a. Main effects Visible Control (%) Dry Density Rate (g ai ha-1) 2 WAAb 4 WAAc 8 WAAc biomass (no. m-2) (g m-2) 0 - - - 11.3 3.31 b 15 + 500 61 54 38 14.1 0.99 ab 30 + 1000 75 75 57 3.71 0.10 ab 40 + 1000 77 81 64 3.24 0.05 a Rate p-value 0.0593 0.0444 0.0559 0.0208 0.0283 Timing PRE 38 b 57 23 9.15 0.35 ePOST 93 a 73 83 5.04 0.24 mPOST 88 a 70 58 8.47 0.80 lPOST 67 ab 80 49 4.89 0.51 Timing p-value 0.0341 0.4472 0.3563 0.0929 0.7975 Interaction Rate x timing p-value 0.3367 0.0010 0.0473 <0.0001 0.1635 Means within a column followed by the same letter are not significantly different according to Tukey’s multiple means comparison (α=0.05). aAbbreviations: WAE/WAA, weeks after emergence/weeks after application; PRE, pre-emergence; ePOST, early- postemergence; mPOST, mid-postemergence; lPOST, late-postemergence. bControl with PRE applications was assessed 2 weeks after crop emergence. cAssessed 4 and 8 weeks after the lPOST application.

124

Table 4.12. Interaction of herbicide rate and application timing on control of flower-of-an-hour (Hibiscus trionum L.) 4 and 8 WAA, and density reduction with three rates of tolpyralate + atrazine applied PRE, ePOST, mPOST or lPOST in field experiments conducted in Ontario, Canada in 2017/2018a.

Visible control (%) 4 WAA Application timing Rate PRE ePOST mPOST lPOST (g ai ha-1) 15 + 500 26 b Y 52 b YZ 59 b YZ 77 a Z 30 + 1000 66 a Z 84 a Z 70 ab Z 79 a Z 40 + 1000 78 a Z 84 a Z 81 a Z 83 a Z

Visible control (%) 8 WAA Application timing Rate PRE ePOST mPOST lPOST (g ai ha-1) 15 + 500 6 b Z 64 b Z 35 c Z 47 a Z 30 + 1000 28 ab Z 93 a Z 58 b Z 50 a Z 40 + 1000 35 a Z 92 a Z 81 a Z 50 a Z

Density (no. m-2) Application timing Rate PRE ePOST mPOST lPOST (g ai ha-1) 0 9.74 a 20.5 b 13.0 b 6.23 bc 15 + 500 19.4 b Z 12.3 b Z 14.3 b Z 11.6 c Z 30 + 1000 6.22 a Y 1.64 a Z 11.3 b Y 1.64 a Z 40 + 1000 5.96 a X 1.56 a Z 2.44 a YZ 4.82 b XY Means followed by the same lower case letter within a column (a-c), or upper case letter within a row (X-Z) for each assessment parameter are not significantly different according to Tukey’s multiple means comparison test (α=0.05). aAbbreviations: WAA, weeks after application; PRE, pre-emergence; ePOST, early-postemergence; mPOST, mid- postemergence; lPOST, late-postemergence.

125

Table 4.13. Effect of rate and application timing on barnyardgrass (Echinochloa crus-galli (L.) P. Beauv.) control 2, 4 and 8 WAE/WAA and density and dry biomass reduction with tolpyralate + atrazine in field experiments conducted in Ontario, Canada in 2017/2018a. Main effects Visible Control (%) Dry Density Rate (g ai ha-1) 2 WAAb 4 WAAc 8 WAAc biomass (no. m-2) (g m-2) 0 - - - 1.7 c 1.45 15 + 500 64 59 b 54 b 1.3 bc 0.70 30 + 1000 77 68 a 65 a 0.2 ab 0.09 40 + 1000 77 71 a 68 a 0.1 a 0.05 Rate p-value <0.0001 0.0002 0.0001 0.0030 0.0006 Timing PRE 41 17 c 9 c 4.6 b 5.01 ePOST 95 94 a 86 a 0.2 a 0.06 mPOST 85 87 ab 88 a 0.1 a 0.04 lPOST 68 67 b 66 b 0.6 ab 0.37 Timing p-value <0.0001 <0.0001 <0.0001 0.0016 <0.0001 Interaction Rate x timing p-value <0.0001 0.1405 0.8929 0.0568 0.0166 Means within a column followed by the same letter are not significantly different according to Tukey’s multiple means comparison (α=0.05). aAbbreviations: WAE/WAA, weeks after emergence/weeks after application; PRE, pre-emergence; ePOST, early- postemergence; mPOST, mid-postemergence; lPOST, late-postemergence. bControl with PRE applications was assessed 2 weeks after crop emergence. cAssessed 4 and 8 weeks after the lPOST application.

126

Table 4.14. Interaction of herbicide rate and application timing on control of barnyardgrass (Echinochloa crus-galli (L.) P. Beauv.) 2 WAE/WAA and dry biomass reduction with three rates of tolpyralate + atrazine applied PRE, ePOST, mPOST or lPOST in field experiments conducted in Ontario, Canada in 2017/2018a.

Visible control (%) 2 WAAb Application timing Rate PRE ePOST mPOST lPOST (g ai ha-1) 15 + 500 21 b Y 91 b Z 81 b Z 63 b X 30 + 1000 54 a X 97 ab Z 85 ab YZ 71 a XY 40 + 1000 49 a X 98 a Z 88 a Z 72 a Y

Dry biomass (g m-2) Application timing Rate PRE ePOST mPOST lPOST (g ai ha-1) 0 3.48 a 1.54 b 0.48 b 6.42 b 15 + 500 35.5 a Y 0.11 b Z 0.11 b Z 2.02 b YZ 30 + 1000 10.9 a Y 0.01 a Z 0.05 ab Z 0.03 a Z 40 + 1000 1.68 a X 0.03 a YZ 0.00 a Z 0.15 ab XY Means followed by the same lower case letter within a column (a-b), or upper case letter within a row (X-Z) for each assessment parameter are not significantly different according to Tukey’s multiple means comparison test (α=0.05). aAbbreviations: WAE/WAA, weeks after emergence/weeks after application; PRE, pre-emergence; ePOST, early- postemergence; mPOST, mid-postemergence; lPOST, late-postemergence. bControl with PRE applications was assessed 2 weeks after crop emergence.

127

Table 4.15. Effect of rate and application timing on green foxtail (Setaria viridis (L.) P. Beauv) control 2, 4 and 8 WAE/WAA and density and dry biomass reduction with tolpyralate + atrazine in field experiments conducted in Ontario, Canada in 2017/2018a. Main effects Visible Control (%) Dry Density Rate (g ai ha-1) 2 WAAb 4 WAAc 8 WAAc biomass (no. m-2) (g m-2) 0 - - - 65 46 15 + 500 63 59 54 b 51 19 30 + 1000 72 72 68 a 32 7 40 + 1000 77 76 71 a 31 7 Rate p-value <0.0001 <0.0001 <0.0001 0.0047 <0.0001 Timing PRE 36 32 19 c 60 61 ePOST 92 90 84 a 30 5 mPOST 84 85 85 a 33 7 lPOST 70 69 67 b 56 21 Timing p-value <0.0001 <0.0001 <0.0001 0.0171 <0.0001 Interaction Rate x timing p-value <0.0001 0.0004 0.1059 0.0273 <0.0001 Means within a column followed by the same letter are not significantly different according to Tukey’s multiple means comparison (α=0.05). aAbbreviations: WAE/WAA, weeks after emergence/weeks after application; PRE, pre-emergence; ePOST, early- postemergence; mPOST, mid-postemergence; lPOST, late-postemergence. bControl with PRE applications was assessed 2 weeks after crop emergence. cAssessed 4 and 8 weeks after the lPOST application.

128

Table 4.16. Interaction of herbicide rate and application timing on control of green foxtail (Setaria viridis (L.) P. Beauv.) 2 and 4 WAE/WAA, and density and dry biomass reduction with three rates of tolpyralate + atrazine applied PRE, ePOST, mPOST or lPOST in field experiments conducted in Ontario, Canada in 2017/2018a.

Visible control (%) 2 WAAb Application timing Rate PRE ePOST mPOST lPOST (g ai ha-1) 15 + 500 23 c X 86 b Z 78 b YZ 65 b Y 30 + 1000 36 b X 94 a Z 86 a YZ 72 a Y 40 + 1000 49 a X 96 a Z 88 a YZ 74 a Y

Visible control (%) 4 WAA Application timing Rate PRE ePOST mPOST lPOST (g ai ha-1) 15 + 500 16 b Y 82 b Z 78 b Z 61 b Z 30 + 1000 38 a Y 94 a Z 87 a Z 72 a Z 40 + 1000 43 a Y 96 a Z 90 a Z 75 a Z

Density (no. m-2) Application timing Rate PRE ePOST mPOST lPOST (g ai ha-1) 0 65 a 62 b 70 b 63 a 15 + 500 72 a Z 36 ab Z 36 ab Z 59 a Z 30 + 1000 54 a Y 10 a Z 13 a Z 51 a Y 40 + 1000 48 a Y 12 a Z 13 a YZ 52 a X

Dry biomass (g m-2) Application timing Rate PRE ePOST mPOST lPOST (g ai ha-1) 0 41 a 43 c 52 c 48 b 15 + 500 84 b X 8.1 b Z 11 b YZ 17 a Y 30 + 1000 71 ab X 1.2 a Z 2.4 a Z 16 a Y 40 + 1000 57 ab X 1.1 a Z 2.2 a YZ 15 a Y Means followed by the same lower case letter within a column (a-c), or upper case letter within a row (X-Z) for each assessment parameter are not significantly different according to Tukey’s multiple means comparison test (α=0.05). aAbbreviations: WAE/WAA, weeks after emergence/weeks after application; PRE, pre-emergence; ePOST, early- postemergence; mPOST, mid-postemergence; lPOST, late-postemergence. bControl with PRE applications was assessed 2 weeks after crop emergence.

129

Table 4.17. Corn injury 1, 2 and 4 WAE/WAA and corn grain yield with three rates of tolpyralate + atrazine applied PRE, ePOST, mPOST and lPOST in field studies conducted in Ontario, Canada in 2017/2018a. Assessment parameter Main effects Visible injury (%) Grain yield Rate (g ai ha-1) 1 WAAb 2 WAAb 4 WAA (t ha-1) 0 - - - 4.3 15 + 500 2.2 1.4 1.0 9.4 30 + 1000 2.7 1.8 1.5 10.4 40 + 1000 3.4 2.7 2.0 10.2 WFC - - - 12.3 Rate p-value 0.0020 0.0014 0.0198 <0.0001 Timing PRE 0 0 0 8.2 ePOST 4.2 2.7 1.3 10.1 mPOST 1.2 0.9 0.8 9.9 lPOST 5.6 4.1 3.9 8.9 Timing p-value 0.3407 0.3827 0.3728 0.0002 Interaction Rate x timing p-value 0.0001 <0.0001 0.0010 <0.0001 Means within a column followed by the same letter are not significantly different according to Tukey’s multiple means comparison (α=0.05). aAbbreviations: WAE, weeks after emergence; WAA, weeks after application; WFC, weed-free control; PRE, pre- emergence; ePOST, early-postemergence; mPOST, mid-postemergence; lPOST, late-postemergence. bInjury was assessed 1, 2 and 4 WAE for PRE applications.

130

Visible injury (%) 1 WAAb Application timing Rate PRE ePOST mPOST lPOST (g ai ha-1) 15 + 500 0 a Z 2.8 a Z 0.6 a Z 5.2 a Z 30 + 1000 0 a Z 4.1 b Z 1.4 a Z 5.4 a Z 40 + 1000 0 a Z 5.8 c Z 1.7 a Z 6.3 a Z

Visible injury (%) 2 WAAb Application timing Rate PRE ePOST mPOST lPOST (g ai ha-1) 15 + 500 0 a Z 1.4 a Z 0.3 a Z 3.8 a Z 30 + 1000 0 a Z 2.2 a Z 0.8 ab Z 4.2 a Z 40 + 1000 0 a Z 4.5 b Z 1.7 b Z 4.5 a Z

Visible injury (%) 4 WAAb Application timing Rate PRE ePOST mPOST lPOST (g ai ha-1) 15 + 500 0 a Z 0.4 a Z 0.4 a Z 3.2 a Z 30 + 1000 0 a Z 1.1 a Z 0.8 a Z 4.1 ab Z 40 + 1000 0 a Z 2.3 b Z 1.3 a Z 4.4 b Z

Grain yield (t ha-1) Application timing Rate PRE ePOST mPOST lPOST (g ai ha-1) 0 4.5 d 4.0 b 4.4 c 4.1 c WFC 12.6 a 12.1 a 12.3 a 12.4 a 15 + 500 6.7 c X 11.3 a Z 10.8 b Z 8.8 b Y 30 + 1000 8.8 b Y 11.7 a Z 11.4 ab Z 9.6 b Y 40 + 1000 8.5 b Y 11.6 a Z 10.9 ab Z 9.6 b Y Means followed by the same lower case letter within a column (a-d), or upper case letter within a row (X-Z) for each assessment parameter are not significantly different according to Tukey’s multiple means comparison test (α=0.05). aAbbreviations: WAA, weeks after application; PRE, pre-emergence; ePOST, early-postemergence; mPOST, mid- postemergence; lPOST, late-postemergence; WFC, weed-free control bInjury was assessed 1, 2 and 4 WAE for PRE applications.

131

Chapter 5: Multiple herbicide-resistant Canada fleabane (Conyza canadensis (L.) Cronq.) dose response to tolpyralate and tolpyralate plus atrazine, and comparison to industry standard herbicides

5.1 Abstract

Canada fleabane biotypes resistant to glyphosate and acetolactate synthase (ALS)- inhibiting herbicides are increasingly prevalent in Canada and the United States, and present a significant management challenge in field crops. Tolpyralate is a recently commercialized herbicide for use in corn which inhibits 4-hydroxyphenyl-pyruvate dioxygenase (HPPD), and there is little information on its efficacy on Canada fleabane. Six field experiments were conducted to determine the biologically-effective dose of tolpyralate and tolpyralate + atrazine and to compare label rates of tolpyralate and tolpyralate + atrazine to currently accepted herbicide standards for postemergence (POST) control of GR Canada fleabane in corn.

Experiments were conducted over a two-year period (2017, 2018) at four locations in Ontario,

Canada with documented populations of glyphosate-resistant (GR) and multiple herbicide- resistant (MR) Canada fleabane. At 8 WAA, tolpyralate at 4.8 and 22.6 g ha-1 provided 50 and

80% control respectively, while no dose ≤120 g ha-1 provided 95% control. When applied with atrazine at a 1:33.3 tank-mix ratio however, 22.3 + 741.7 g ha-1 provided 95% of MR Canada fleabane. The addition of atrazine to tolpyralate at label rates improved control of MR Canada fleabane to 98%, which is similar to the control provided by dicamba/atrazine and bromoxynil + atrazine. The results of this study conclude that tolpyralate + atrazine POST provides excellent control of GR/MR Canada fleabane in corn.

132

5.2 Introduction

Canada fleabane (Conyza canadensis (L.) Cronq.) is a fall or spring-germinated

Asteraceae species found throughout Canada and the United States, which exhibits a highly competitive and adaptable growth pattern (Weaver 2001; Buhler and Owen 1997). Canada fleabane has high fecundity and the ability to germinate on the soil surface, making it a highly successful weed in agroecosystems, and particularly adaptable to no-tillage crop production systems (Main et al. 2006; Weaver 2001; Davis et al. 2009). Canada fleabane seed production is relative to plant height; a 1.5 m-tall plant can produce in excess of 230,000 small, wind-blown seeds, which can be dispersed over 500 km from the mother plant (Weaver 2001; Shields et al.

2006).

Glyphosate-resistant (GR) biotypes of Canada fleabane were first reported in 2001 in

Delaware, and have subsequently been documented throughout much of the United States and

Ontario, Canada (Ge et al. 2011; Budd et al. 2018; VanGessel 2001). Glyphosate-resistant

Canada fleabane is present in 30 counties in Ontario, with several populations also exhibiting multiple resistance (MR) to cloransulam-methyl, an ALS-inhibiting herbicide (Budd et al. 2018).

Interference from GR Canada fleabane in corn can cause up to 69% grain yield loss based on

Ontario studies (Ford et al. 2014) underlining the importance of effective management of this species. A common management strategy implemented by growers for control of GR Canada fleabane is the use of herbicides with alternative modes of action (Scott and VanGessel 2007). In

Ontario, several herbicide tank-mixtures applied pre-plant provide effective (>90%) control of

GR Canada fleabane in corn (Brown et al. 2016; Ford et al. 2014); however, there are limited options for control of GR/MR biotypes POST in corn (Mahoney et al. 2017). In Ontario studies, only dicamba, dicamba/diflufenzopyr, dicamba/atrazine and bromoxynil + atrazine provided

133

>90% control of this species POST in corn, representing just four herbicide modes of action

(Mahoney et al. 2017).

Herbicides that inhibit 4-hydroxyphenyl-pyruvate dioxygenase (HPPD) impede biosynthesis of plastoquinone (PQ) and α-tocopherols in susceptible plants through competitive inhibition of the HPPD enzyme, inhibiting carotenoid biosynthesis and fostering light-induced generation of reactive oxygen species and triplet chlorophyll (Hawkes 2012; Ahrens et al. 2013).

Several HPPD-inhibitors including triketones, isoxazoles and pyrazolones are used in field corn; although each is unique in the spectrum of weeds controlled. Consequently, HPPD-herbicides are commonly applied to corn in tank-mixtures with photosystem-II (PSII)-inhibitors such as atrazine, which improves herbicide efficacy and broadens the weed control spectrum (Armel et al. 2005, 2007; Kim et al. 1999; Abendroth et al. 2005; Kohrt and Sprague 2017a; Metzger et al.

2018b). Photosystem-II inhibitors disrupt electron flow through PSII by competing with PQ for the binding pocket on the D1 protein (Hess 2000). Hydroxyphenyl-pyruvate dioxygenase- inhibitors are presumed to improve the efficiency of atrazine binding by limiting PQ biosynthesis, and intensify subsequent lipid peroxidation by impeding the biosynthesis of carotenoids and tocopherols (Armel et al. 2005). Therefore, HPPD-inhibitors are believed to exhibit a mode of action which is complementary to that of PSII-inhibitors (Kim et al. 1999).

Control of GR Canada fleabane with HPPD-inhibitors has been variable. Mesotrione + atrazine provided 88-97% control of GR Canada fleabane applied pre-emergence (PRE) (Armel et al. 2009; Brown et al. 2016); in contrast, mesotrione + atrazine, applied POST, provided only

76% control (Mahoney et al. 2017). Similarly, topramezone + atrazine provided only 67% control of GR Canada fleabane (Mahoney et al. 2017).

134

Tolpyralate is a new POST, pyrazolone herbicide for use in corn that was registered in

Canada and the United States in 2017 (Anonymous 2017a; Anonymous 2017b). The biologically-effective dose of tolpyralate for 90% control of common ragweed, velvetleaf, common lambsquarters, green/redroot pigweed and green foxtail has been determined to be

≤15.5 g ai ha-1; however not all species could be controlled with tolpyralate alone (Metzger et al.

2018a). The addition of atrazine to tolpyralate applications at a 1:33.3 tank-mix ratio was found to increase the speed of weed control in some species, and also provide control of additional weed species, including wild mustard and ladysthumb, which were not controlled with tolpyralate applied alone (Metzger et al. 2018b).

Little published information is available regarding the biological activity of tolpyralate or tolpyralate + atrazine mixtures in GR/MR Canada fleabane. The commercial label for tolpyralate in the USA lists “partial control” of Canada fleabane when applied alone (30-40 g ha-1) or with atrazine (≥560 g ai ha-1) (Anonymous 2017a), while Canada fleabane is not included on the

Canadian label (Anonymous 2017b). Therefore, the objectives of this study were to determine the biologically-effective dose of tolpyralate alone and tolpyralate + atrazine in GR/MR Canada fleabane, and secondly, to compare the efficacy of tolpyralate applied at the lowest current label rate with and without atrazine to current herbicide standards for POST control of GR/MR

Canada fleabane in corn.

5.3 Materials and Methods

Experimental Methods Six no-till field experiments were conducted over two years (2017, 2018) at sites near

Mull, Ridgetown, Harrow and Thamesville, Ontario, Canada which had populations of Canada fleabane previously confirmed as resistant to glyphosate and partially resistant to cloransulam-

135 methyl. At each location, plots consisted of three rows of DKC46-82RIB corn seeded 4 cm deep, on 76-cm row-spacing. Each field experiment was organized as a randomized complete block with four replications. Trial sites were fertilized according to soil test results and Ontario

Ministry of Agriculture and Food (OMAF) recommendations for field corn. Table 5.1 lists soil characteristics, planting dates, harvest dates, spray dates, and corn growth stage and Canada fleabane size and density at the times of application.

Prior to treatment application, glyphosate (450 g ae ha-1) was applied to the entire trial area using a tractor-mounted sprayer to remove all other competing weed species, and glyphosate-susceptible Canada fleabane. Weedy, non-treated control (NTC) plots only received this glyphosate application. Weed-free control (WFC) plots were treated with dimethenamid-

P/saflufenacil (735 g ai ha-1) PRE, followed by dicamba/atrazine (1500 g ai ha-1) POST, and subsequently maintained by hand-weeding as needed. All herbicide applications were made

-1 using a backpack sprayer with CO2 as the propellant, calibrated for a 187 L ha delivery volume at 240 kPa, equipped with four ULD12002 nozzles (Pentair, New Brighton, MN) spaced 50 cm apart. Herbicide treatments were applied when Canada fleabane plants reached an average of 10 cm in diameter/height. Due to the potential for Canada fleabane to germinate in the spring or autumn, corn stage at time of application varied from pre-emergence to V4, depending on field location (Table 5.1). Treatments to determine the biologically-effective dose were a titration of tolpyralate, applied at doses 3.75, 7.5, 15, 30, 60 and 120 g ai ha-1, and a titration of tolpyralate + atrazine, tank-mixed at a 1:33.3 ratio, at doses 3.75 + 125, 7.5 + 250, 15 + 500, 30 + 1000, 60 +

2000 and 120 + 4000 g ai ha-1. This tank-mix ratio was chosen with consideration of a previous dose-response study with tolpyralate + atrazine on glyphosate-susceptible weeds (Metzger et al.

2018a). Methylated seed oil (MSO Concentrate®, Loveland Products Inc., Loveland CO) at

136

0.50% v/v and 28% N urea ammonium nitrate at 2.50% v/v were included with tolpyralate applications in accordance with herbicide label recommendations (Anonymous 2017a).

Dicamba/atrazine (1500 g ai ha-1) and bromoxynil + atrazine (280 + 1500 g ai ha-1) treatments were also included, representing the current standard herbicide treatments for control of GR/MR

Canada fleabane POST in corn (Mahoney et al 2017).

Crop injury 1, 2 and 4 WAA, and visible weed control 2, 4 and 8 WAA were assessed on a percent scale (0-100) relative to the NTC plot. Canada fleabane density and biomass were determined 8 WAA by counting and cutting plants contained in a 0.25 m-2 quadrat placed randomly at two locations between the centre corn rows of each plot. Harvested Canada fleabane plants were placed into labelled paper bags, dried at 60 C, weighed and recorded as g dry matter

(DM) m-2.

Corn grain yield was determined at the end of each season by harvesting the centre two rows of each plot with a small-plot combine. Grain moisture and weight per plot was recorded, and grain yields adjusted to 15.5% moisture.

Statistical Analysis

Non-linear Regression – Log-Logistic Dose-Response

Canada fleabane visible control at 2, 4 and 8 WAA, and corn grain yield expressed as a percent of the WFC plot within replications, were each regressed against tolpyralate or tolpyralate + atrazine dose by specifying a four-parameter log-logistic model (Equation 1) within

PROC NLIN in SAS v. 9.4 (SAS Institute, Cary, NC). Canada fleabane density (plants m-2) and dry biomass (g DM m-2) were each regressed against tolpyralate dose or tolpyralate + atrazine

137 dose using an inverse exponential equation (Equation 2). Parameters generated from each regression analysis were used to compute the effective dose (R50, R80, R95) of tolpyralate or tolpyralate + atrazine combined required to give 50, 80 and 95% control, 50, 80 and 95% reduction in density and biomass and 50, 80 and 95% of corn grain yield relative to the WFC plots within each replication. When the predicted dose for any parameter was outside of the range of doses used in this study or could not be computed by the model, it was expressed as

‘Non-est.’ in Tables and Figures, as extrapolation beyond the parameters of this study would be improper.

Equation 1 – Log-logistic dosage-response model

(-b*(log(dose) – log(I ))) y = C + (D-C)/(1+e 50 )

Where:

y = response parameter

C = lower asymptote

D = upper asymptote

I50 = dose eliciting a response equidistant between C and D

b = slope about I50

Equation 2 – Inverse Exponential

y = e + f (-g*dose)

138

Where:

y = response parameter

e = lower asymptote

f = reduction in y from intercept to asymptote

g = slope

Least-Square Means Comparisons

A mixed-model variance analysis was carried out on each response parameter using

PROC GLIMMIX in SAS v. 9.4 (SAS Institute, Cary, NC). Data were combined across years and locations (collectively termed ‘environment’) for 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. Statistical significance (α=0.05 for all tests) of the fixed effect was determined using an F-test, while statistical significance of the random effects were determined using the log-likelihood ratio test. An appropriate distribution and link for each parameter which met assumptions that residuals were homogenous, had mean equal to zero and were normally-distributed was selected based on a Shapiro-Wilk test of normality, together with scatterplots of studentized residuals. Visible control 2 and 8 WAA were assigned a normal distribution with identity link; visible control 4 WAA was assigned a beta distribution with a cumulative complimentary log-link. Density and biomass data were each modelled using a lognormal distribution with an identity link. Least-square means of tolpyralate alone, tolpyralate + atrazine, dicamba/atrazine and bromoxynil + atrazine were computed for

139 each parameter on the analysis scale, and converted to the data scale using the ilink option in

PROC GLIMMIX where appropriate. Where a lognormal distribution was specified, the omega method for back-transformation was used within the GLIMMIX procedure (M. Edwards, Ontario

Agricultural College Statistician, University of Guelph, personal communication). Least-square means were compared using Tukey-Kramer’s multiple range test, and letter codes were assigned to illustrate statistically significant differences using the lines option in SAS.

5.4 Results and Discussion

Crop Injury No treatment caused injury greater than 10% on average; however, injury varied across experiments, with the highest rate of tolpyralate + atrazine (120 + 4000 g ha-1) causing up to

40% injury 1 WAA at one location in 2017 (data not presented). Injury consisted of transient white/yellow bleaching and dissipated to less than 10% by 4 WAA. Temperature during application was moderate (20 C) although crop stage may have influenced the level of injury observed at this site; corn was at the V1 (two leaf) stage at the time of application (Table 5.1).

Previous studies have reported greater crop injury when nicosulfuron + bromoxynil were applied to corn prior to the three-leaf stage (Carey and Kells 1995) possibly due to a thinner, less developed leaf cuticle in younger corn plants.

Dose-Response Tolpyralate induced white bleaching, chlorosis and necrosis of MR Canada fleabane plants within 10 days of application, with greater necrosis observed with increasing tolpyralate rate or with the addition of atrazine. These symptoms were similar to those observed in several annual weed species (Metzger et al. 2018a). Based on regression analysis, at 2 WAA, 50 and

80% control of MR Canada fleabane could be achieved with 1.5 and 16 g ha-1 tolpyralate,

140 respectively; however, no dose ≤120 g ha-1 provided 95% control at this timing (Table 5.2). In contrast, tolpyralate + atrazine at 18.5 + 616.6 g ha-1, respectively, controlled MR Canada fleabane 95%, representing 0.6X the lowest label rate. Similarly, when tolpyralate + atrazine were applied in combination, 50 and 80% control of MR Canada fleabane could be achieved 2

WAA with 1.5 + 50.8 and 4.9 + 164.5 g ha-1 tolpyralate + atrazine, respectively. Similar results were reported by Metzger et al. (2018a) in dose response studies on a range of glyphosate- susceptible (GS) weed species including common ragweed (Ambrosia artemisiifolia L.), velvetleaf (Abutilon theophrasti Medik.) common lambsquarters (Chenopodium album L.),

Amaranthus spp. and green foxtail (Setaria viridis (L.) P. Beauv), in which the addition of atrazine to tolpyralate improved control at early assessment timings. At 2 WAA, the addition of atrazine to tolpyralate allowed for a reduction in the tolpyralate rate of over 60% for 80% control of MR Canada fleabane, while similar rates of tolpyralate were required for 50% control whether atrazine was included or not.

Control of MR Canada fleabane generally improved from 2 to 4 WAA. When tolpyralate was applied alone, 50, 80 and 95% control could be achieved with 3.3, 17.5 and 116.1 g ha-1, respectively (Table 5.2). Kikugawa et al. (2015) reported 98% control of GR Canada fleabane with tolpyralate alone (30 g ha-1); however, those results were obtained from a greenhouse experiment, and applications were made to plants approximately 50% of the size of those used in this study. Similar to 2 WAA, the addition of atrazine facilitated a reduction in the required tolpyralate dose for each level of control. When tank-mixed, 2.1 + 69.9 and 5.7 + 189.9 g ha-1 tolpyralate + atrazine provided 50 and 80% control respectively, while 95% control was obtained with 17.6 + 587.9 g ha-1, representing approximately 0.6X the low label rate of tolpyralate.

141

Despite the extended emergence pattern of Canada fleabane in Ontario (Weaver 2001), and reports of tolpyralate having limited pre-emergence activity (Anonymous 2017a), few late- emerging MR Canada fleabane plants were observed in plots treated with tolpyralate (≥30 g ha-1)

8 WAA. However, regrowth of some MR Canada fleabane plants was observed where tolpyralate was applied alone. To achieve 50 and 80% control 8 WAA, 4.8 and 22.6 g ha-1 of tolpyralate was required, respectively, but no dose could provide 95% control (Table 5.2). The addition of atrazine to tolpyralate provided more complete control of MR Canada fleabane, and

95% control could be achieved with rates ≤22.3 + 741.7 g ha-1 tolpyralate + atrazine. Similar to previous assessment timings, applying atrazine in combination with tolpyralate facilitated a reduction in the rate of tolpyralate required for 50, 80 and 95% control of MR Canada fleabane.

This trend was consistent with that reported in Metzger et al. (2018a) for the biologically- effective dose of tolpyralate alone or with atrazine in GS velvetleaf, Amaranthus spp., common ragweed, common lambsquarters, green foxtail and barnyardgrass (Echinochloa crus-galli (L.) P.

Beauv.), 8 WAA.

Density and biomass data followed similar trends to visible control data 8 WAA based on regression analysis using Equation 2 (Fig. 5.1-5.4). A comparatively higher dose of tolpyralate was required for a 50, 80 or 95% reduction in MR Canada fleabane density compared to biomass, due to incompletely controlled plants remaining in plots 8 WAA. Many of these plants were severely injured, stunted or necrotic 8 WAA, and thus contributed little to total-plot biomass assessments. Tolpyralate alone reduced MR Canada fleabane dry biomass 50 and 80% compared to the NTC at 1.8 and 4.5 g ha-1, respectively (Fig. 5.1); 3.4 and 9 g ha-1 were required for an equivalent reduction in density (Fig. 5.3). Similarly, 1.2 + 38.9 and 2.7 + 91.3 g ha-1 tolpyralate + atrazine were required for a 50 and 80% reduction in MR Canada fleabane biomass,

142 respectively; 1.7 + 55.2 and 3.9 + 129.7 g ha-1 were required for equivalent reductions in density.

None of the doses of tolpyralate alone used in this study (≤120 g ha-1) provided a 95% reduction in either dry biomass or density of MR Canada fleabane 8 WAA based on the regression analysis. In contrast, with the addition of atrazine, a 95% reduction in MR Canada fleabane dry biomass and density could be achieved with 5.4 + 178.2 and 7.6 + 253.9 g ha-1, respectively (Fig.

5.2; Fig. 5.4). These rates represent approximately 0.2 and 0.3X the low label rate; they contrast with the commercial label for tolpyralate which claims only partial control of Canada fleabane

(Anonymous 2017b).

Tolpyralate and Herbicide Standards Least-square means for each control parameter were compared for four treatments applied at label rates, which included: tolpyralate alone (30 g ha-1), tolpyralate + atrazine (30 +

1000 g ha-1), dicamba/atrazine (500 + 1000 g ha-1) and bromoxynil + atrazine (280 + 1500 g ha-1)

(Table 5.3). Dicamba/atrazine and bromoxynil + atrazine treatments were included in this study based on Mahoney et al. (2017), in which these treatments provided 96 and 93% control of GR

Canada fleabane 8 WAA, respectively. At 2 WAA, tolpyralate + atrazine and bromoxynil + atrazine provided similar control of MR Canada fleabane; both provided more complete control at this assessment timing than tolpyralate alone or dicamba/atrazine. These results are consistent with Mahoney et al. (2017) in which dicamba-based herbicides provided poorer control of GR

Canada fleabane than bromoxynil + atrazine at earlier assessment timings, with control improving by 8 WAA. At 4 WAA, tolpyralate + atrazine and bromoxynil + atrazine provided 99 and 97% control of MR Canada fleabane, while tolpyralate alone and dicamba/atrazine provided

86 and 88% control, respectively. At 8 WAA, the addition of atrazine to tolpyralate increased control of MR Canada fleabane by 15 percentage points. Across all assessment timings and parameters, the addition of atrazine to tolpyralate improved MR Canada fleabane control. 143

Despite poor control of GR Canada fleabane with atrazine (1000 g ha-1) alone (Mahoney et al

2017), similar results were reported with another HPPD-inhibitor by Armel et al. (2009), who found that the addition of atrazine (280 g ha-1) to mesotrione (160 g ha-1) increased control of GS

Canada fleabane from 37 to 88%, 3 WAA. At 8 WAA, bromoxynil + atrazine and dicamba/atrazine provided control that was similar to tolpyralate alone and tolpyralate + atrazine.

In previous studies on these and other populations of GR/MR Canada fleabane in Ontario, dicamba/atrazine, and other herbicides which include dicamba provided >90% control (Byker et al. 2013b; Mahoney et al. 2017). In this study however, MR Canada fleabane plants were observed within dicamba/atrazine plots which exhibited substantial branching from axillary buds by 8 WAA. Flessner et al. (2015) reported similar findings with dicamba applied to GR Canada fleabane at low rates. Hedges et al. (2018) found that glyphosate/dicamba (2:1 ratio) at 900 and

1800 g ae ha-1 controlled GR Canada fleabane 76 and 92% respectively. Therefore, it can be postulated that the rate of dicamba used in our experiments (500 g ha-1) may be insufficient for complete control of MR Canada fleabane in some environments.

When compared to the NTC plots, all treatments with the exception of dicamba/atrazine provided a significant reduction in MR Canada fleabane dry biomass and density (Table 5.3). It is likely that natural population variability within trial sites and replications contributed to the result that dicamba/atrazine was not statistically different than the NTC; however, the large numerical difference in both density and dry biomass gives reason to consider the biological significance of control provided by dicamba/atrazine. Tolpyralate alone provided a similar reduction in MR Canada fleabane dry biomass and density compared to dicamba/atrazine, while tolpyralate + atrazine and bromoxyil + atrazine provided a greater reduction in both parameters.

144

Corn Grain Yield Corn grain yield varied by site relative to MR Canada fleabane density and biomass but was generally reflective of overall weed control. Yields were reduced by 47% relative to the

WFC where MR Canada fleabane was left uncontrolled for the entire season (Table 5.3).

Previously, Ford et al. (2014) reported a 69% grain yield reduction with GR Canada fleabane left uncontrolled in corn. Since yields were not reduced by 50% or greater as a result of MR Canada fleabane interference, R50 doses for tolpyralate alone or with atrazine could not be computed with the regression model (Table 5.2). When applied alone, 1.2 g ha-1 of tolpyralate was sufficient to maintain 80% of the yield obtained in WFC plots, while 4.0 g ha-1 provided yields equivalent to

95% of that obtained in WFC plots. Similarly, when applied in conjunction with atrazine, 80% yield potential could be achieved with 1.6 + 54.7 g ha-1, while 2.4 + 80.9 g ha-1 was sufficient to avoid yield loss greater than 5% relative to WFC plots. These R80 and R95 values are comparatively lower than those calculated for 80 and 95% visible control, indicating that although low rates of tolpyralate and tolpyralate + atrazine maintained corn grain yield, weed escapes were present, and therefore could have contributed to the soil seed bank despite not affecting yield.

Each of the four commercial treatments provided control of MR Canada fleabane that was sufficient to maintain corn grain yield equivalent to the WFC (Table 5.3). No statistically significant yield differences existed where tolpyralate, tolpyralate + atrazine, dicamba/atrazine or bromoxynil + atrazine were applied; each treatment resulted in yield that was 46-48% higher than that obtained in the NTC.

145

5.5 Conclusions

In this study, tolpyralate (22.6 g ha-1), applied POST, controlled MR Canada fleabane

80% 8 WAA, representing approximately 0.8X the low label rate. Tolpyralate applied alone at the low label rate (30 g ha-1) controlled MR Canada fleabane 83%. These results are in contrast to the current information listed on the commercial labels for tolpyralate (Anonymous 2017a;

Anonymous 2017b). Although atrazine (1000 g ha-1) has been shown to provide only 37% control of GR Canada fleabane POST (Mahoney et al. 2017), adding atrazine to tolpyralate at a

1:33.3 ratio provided >95% control of GR/MR Canada fleabane 2, 4 and 8 WAA. At 8 WAA,

22.3 + 741.7 g ha-1 provided 95% control based on regression analysis. Applied at the label rate, tolpyralate + atrazine provided 98% control of MR Canada fleabane 8 WAA, which was similar to the control provided by bromoxynil + atrazine and dicamba/atrazine. Similarly, tolpyralate, tolpyralate + atrazine and each industry standard treatment resulted in corn grain yields that were equivalent to those obtained in WFC plots. Overall, the results of this study highlight the efficacy of tolpyralate + atrazine tank mixtures on MR Canada fleabane, and applying a tank-mixture of at least two effective herbicide modes of action is prudent in order to reduce selection pressure on any single mode of action. The judicious use of this treatment in rotation with additional/alternative herbicide modes of action and in combination with biological, cultural and mechanical weed management strategies will help to preserve its long-term utility as an effective option for control of MR Canada fleabane in corn.

146

Table 5.1. Soil characteristics, planting dates, harvest dates, spray dates, corn growth stage and Canada fleabane size and density for field trials near Mull, Ridgetown, Thamesville and Harrow, ON in 2017 and 2018. Soil Characteristics Application Information Canada Canada Corn Nearest OMa Planting Harvest Spray Fleabane Fleabane Year Texture pH Growth Town (%) Date Date Date Sizeb Densityc Stage (cm) (plants m-2) Mull Loam 3.0 7.0 Jun-15 n/ad Jun-15 PREe 10 360 Sandy 2017 Thamesville 3.3 6.0 May-18 Nov-11 May-10 PRE 7 48 loam Sandy Harrow 1.9 6.8 May-15 Nov-11 May-29 V1 8 531 loam Loamy Ridgetown 2.8 6.9 May-9 Oct-30 May-21 VE 7 37 sand Sandy 2018 Thamesville 3.0 6.9 May-9 Nov-8 June-5 V4 9 325 loam Sandy Harrow 2.4 6.3 May-25 Nov-5 June-6 V2 7 41 loam aAbbreviations: OM, organic matter, PRE, pre-emergence bSize measured as height of bolting plants or rosettes. Mean of eight measurements per trial at time of application. cMean density based on two stand counts per block within each trial. dNot harvested. eTreatments applied prior to crop emergence as dictated by weed size.

147

Table 5.2. Regression parameters and predicted effective dose of tolpyralate and tolpyralate plus atrazine for 50, 80 and 95% visible control of multiple-resistant Canada fleabane (Conyza canadensis (L.) Cronq.) 2, 4 and 8 WAA, and to achieve 50, 80 and 95% of the yield obtained in weed-free control plots based on the log-logistic dose response equation (Equation 1)a. b Parameter estimates (±SE) Predicted tolpyralate dose Variable C D b I50 R50 R80 R95 (g ai ha-1)

2 WAA 0.02 (2.1) 100 (0) 0.59 (0.06) 1.49 (0.31) 1.5 16.0 Non-est.

4 WAA 0.06 (3.0) 100 (0) 0.82 (0.09) 3.25 (0.49) 3.3 17.5 116.1

8 WAA 0.23 (3.6) 100 (0) 0.90 (0.1) 4.83 (0.68) 4.8 22.6 Non-est.

Yieldc 51.2 (11.1) 105.6 (10.9) 1.08 (3.23) 1.08 (4.0) Non-est. 1.2 4.0

Predicted tolpyralate + atrazine dose

Variable C D b I50 R50 R80 R95 (g ai ha-1)

2 WAA 0.01 (1.2) 100 (0) 1.18 (0.1) 52.23 (6.12) 1.5 + 50.8 4.9 + 164.5 18.5 + 616.6

4 WAA 0.01 (1.9) 100 (1.6) 1.39 (0.25) 71.89 (9.24) 2.1 + 69.9 5.7 + 189.9 17.6 + 587.9

8 WAA 0.07 (2.4) 100 (0) 1.30 (0.17) 78.84 (10.22) 2.3 + 76.5 6.7 + 222.9 22.3 + 741.7

Yield 51.2 (11.6) 111.1 (6.1) 2.76 (19.3) 57.95 (322.7) Non-est. 1.6 + 54.7 2.4 + 80.9

aAbbreviation: WAA, weeks after application b Regression parameters: C = lower asymptote, D = upper asymptote, b = slope about I50, I50 = effective dose to elicit a 50% response, Rn, effective dose to elicit response level n. cExpressed as percent of yield in weed-free control plots within replications.

148

200 Tolpyralate 180

160

) 2

- Observed +/- SE 140 Predicted 120 Biomass = (6.763 + 97.333)(-0.427*Dose) 100 R50 = 1.8 R = 4.5 80 80 R95 = Non-est.

60 Dry m DM Biomass (g Dry 40 20 0 0 20 40 60 80 100 120 Dose (g ai ha-1) Figure 5.1. Inverse exponential function (Equation 2) and predicted effective dose of tolpyralate for a 50, 80 and 95% reduction in multiple-resistant Canada fleabane (Conyza canadensis (L.) Cronq.) dry biomass, based on six field experiments conducted in Ontario, Canada in 2017/2018.

200 Tolpyralate + Atrazine

180

) 160 2

- Observed +/- SE 140 Predicted 120 Biomass = (1.136 + 103.2)(-0.0176*Dose) 100 R50 = 1.2 + 38.9 80 R80 = 2.7 + 91.3 R = 5.4 + 178.2 60 95

Dry m DM Biomass (g Dry 40 20 0 0 500 1000 1500 2000 2500 3000 3500 4000 4500 Dose (g ai ha-1) Figure 5.2. Inverse exponential function (Equation 2) and predicted effective dose of tolpyralate + atrazine for a 50, 80 and 95% reduction in multiple-resistant Canada fleabane (Conyza canadensis (L.) Cronq.) dry biomass, based on six field experiments conducted in Ontario, Canada in 2017/2018. 149

500 Tolpyralate 450

400 Observed +/- SE

) 350 2 - Predicted 300 Density = (17.886 + 175.7)(-0.238*Dose) 250 R50 = 3.4 200 R80 = 9 R95 = Non-est.

150 Density m (plants Density 100 50 0 0 20 40 60 80 100 120 Dose (g ai ha-1) Figure 5.3. Inverse exponential function (Equation 2) and predicted effective dose of tolpyralate for a 50, 80 and 95% reduction in multiple-resistant Canada fleabane (Conyza canadensis (L.) Cronq.) density, based on six field experiments conducted in Ontario, Canada in 2017/2018.

500 Tolpyralate + Atrazine 450 400

Observed +/- SE

) 350 2 - Predicted 300 Density = (2.23 + 194)(-0.0124*Dose) 250 R50 = 1.7 + 55.2 200 R80 = 3.9 + 129.7 150 R95 = 7.6 + 253.9

Density m (plants Density 100 50 0 0 500 1000 1500 2000 2500 3000 3500 4000 4500 -1 Dose (g ai ha ) Figure 5.4. Inverse exponential function (Equation 2) and predicted effective dose of tolpyralate plus atrazine for a 50, 80 and 95% reduction in multiple-resistant Canada fleabane (Conyza canadensis (L.) Cronq.) density, based on six field experiments conducted in Ontario, Canada in 2017/2018.

150

Table 5.3. Multiple-resistant Canada fleabane (Conyza canadensis (L.) Cronq.) visible control 2, 4 and 8 WAA and reduction in plant density and dry biomass 8 WAA provided by commercial rates of tolpyralate, tolpyralate + atrazine, dicamba/atrazine and bromoxynil + atrazine applied post-emergence in field studies conducted in Ontario, Canada during 2017/2018a.

Assessment parameter Density Dry Biomass Grain yield 2 WAAb 4 WAA 8 WAA (No. m-2) (g DM m-2) (t ha-1) Rate

Treatment (g ai ha-1)

Tolpyralate 30 85 b 86 b 83 b 1 b 0.3 b 11.6 a

Tolpyralate 30 + 1000 98 a 99 a 98 a 0 a 0.0 a 11.9 a + atrazine

Dicamba/ 500/1000 79 b 88 b 87 ab 4 bc 2.0 bc 11.5 a atrazine

Bromoxynil 280 + 1500 96 a 97 a 94 ab 0 a 0.0 a 12.0 a + atrazine

NTCc 293 c 109 c 6.2 b

WFC 11.7 a 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. bAbbreviations: WAA, weeks after application; DM, dry matter; NTC, non-treated control; WFC, weed-free control. cNon-treated control plots received only glyphosate (900 g ae ha-1).

151

Chapter 6: Multiple herbicide-resistant waterhemp (Amaranthus tuberculatus (Moq.) J. D. Sauer) dose response to tolpyralate and tolpyralate plus atrazine, and comparison to industry standard herbicides

6.1 Abstract

Waterhemp is a dioecious, small-seeded, Amaranthus species with immense reproductive potential, a competitive growth pattern and vast genetic heterogeneity. Recent surveys have confirmed waterhemp biotypes in Ontario with resistance to four distinct herbicide mechanisms of action, complicating management of this species. Tolpyralate is the newest 4-hydroxyphenyl- pyruvate dioxygenase (HPPD)-inhibiting herbicide available in Ontario and this mode of action has proven activity on waterhemp. Experiments were conducted at three field locations in

Ontario in 2018, which were infested with waterhemp previously confirmed as multiple-resistant

(MR) to Group 2, 5 and 9 herbicides, to determine: 1) dose-response of MR waterhemp to tolpyralate and tolpyralate plus atrazine, and 2) the relative efficacy of tolpyralate, tolpyralate plus atrazine, dicamba/atrazine and mesotrione + atrazine for post-emergence (POST) control of this species in corn. At 12 weeks after application (WAA), tolpyralate (24.5 g ai ha-1) controlled

MR waterhemp 80%, while 46.5 + 1550.6 g ai ha-1 tolpyralate + atrazine provided 95% control.

No dose of tolpyralate (≤120 g ha-1) applied alone controlled MR waterhemp 95%. Applied at a registered rate, tolpyralate (30 g ha-1) and tolpyralate + atrazine (30 + 1000 g ha-1) provided control of MR waterhemp that was similar to dicamba/atrazine and mesotrione + atrazine; while tolpyralate + atrazine reduced MR waterhemp dry biomass more effectively than dicamba/atrazine. This study demonstrates that tolpyralate or tolpyralate + atrazine tank-

152 mixtures, applied POST, are an effective herbicide management strategy for control of three-way resistant waterhemp in corn in Ontario.

6.2 Introduction

Waterhemp (Amaranthus tuberculatus (Moq.) J.D. Sauer) is a troublesome Amaranthus species that infests agricultural land across much of the mid-western USA, and is becoming increasingly prevalent in Ontario, Canada. While tall waterhemp (Amaranthus tuberculatus

(Moq.) Sauer var. tuberculatus) is native to undisturbed habitats in southern Ontario and Western

Québec, common waterhemp (A. rudis) is non-native; it is thought to have been introduced to

Ontario on contaminated farm equipment transported from Illinois (Costea et al. 2005). Previous research in Ontario has described common waterhemp (Amaranthus tuberculatus var. rudis)

(Schryver et al. 2017a; 2017b; Vyn et al. 2006; 2007). However, due to its proclivity for outcrossing and significant genetic variability, waterhemp is now classified as a single species

(Amaranthus tuberculatus (Moq.) J.D. Sauer) (Weed Science Society of America [WSSA]

2018). All subsequent mention of ‘waterhemp’ in this study reflects this classification. Since its initial discovery, waterhemp has been documented in four southwestern Ontario counties (Benoit et al. unpublished data; Schryver et al. 2017a) and has become a management challenge in corn and soybean production.

Left uncontrolled, common waterhemp has been documented to reduce soybean yields by up to 73% and corn yields by up to 74% (Steckel and Sprague 2004; Vyn et al. 2007). Although it is currently a localized weed species in Ontario, waterhemp is capable of rapid geographical expansion, as evidenced by its pervasiveness in the mid-western United States (Horak and

Loughin 2000; Liu et al. 2012; Nordby et al. 2007). Waterhemp possesses several biological

153 characteristics that contribute to its prevalence, including high fecundity, a prolonged emergence pattern and a rapid growth habit (Costea et al. 2005; Hartzler et al. 2004; Vyn et al. 2007; Wu and Owen 2014). Hartzler et al. (2004) reported that common waterhemp can produce up to 4.8 million seeds per plant in competition with a growing crop, contributing vast reserves to the soil seed bank. Aside from direct transport of seed by equipment and human activities, waterhemp seed can also be dispersed by wildlife. Farmer et al. (2017) determined that viable common waterhemp seed could be transported distances of up to 2964 km by migratory waterfowl, highlighting the potential for rapid geographic expansion of this weed species. In Ontario, waterhemp emerges from May through September or October (Vyn et al. 2007; Schryver et al.

2017b). Once emerged, waterhemp exhibits an aggressive, indeterminate growth habit, attaining heights of over 3 m in competitive environments (Costea et al. 2005; Nordby et al. 2007). Horak and Loughin (2000) reported that common waterhemp had larger total plant volume and more branches than other Amaranthus spp. including A. retroflexus. Additionally, waterhemp is dioecious and wind-pollinated, contributing to considerable genetic variability among biotypes within populations (Costea et al. 2005; Kreiner et al. 2018; Liu et al. 2012). The resulting rapid introgression contributes to substantial phenotypic plasticity; waterhemp is able to thrive across a wide range of environments and growing conditions (Costea et al. 2005).

Further compounding the management challenges associated with waterhemp is the propensity for this species to rapidly evolve resistance to several herbicide modes of action

(Nordby et al. 2007). Globally, there are waterhemp biotypes with resistance to six herbicide modes of action, including acetolactate synthase (ALS)-inhibitors, synthetic auxins, photosystem

II (PSII)-inhibitors, 5-enolpyruvylshikimate-3-phosphate synthase (EPSPS)-inhibitors, protoporphyrinogen IX oxidase (PPO)-inhibitors, and 4-hydroxyphenyl-pyruvate dioxygenase

154

(HPPD)-inhibitors, representing WSSA Herbicide Groups 2, 4, 5, 9, 14 and 27, respectively

(Heap 2018). In Ontario, there are waterhemp biotypes with resistance to Group 2, 5, 9 and 14 herbicides (Heap 2018). Additionally, 61% of these populations have three-way multiple resistance to imazethapyr, atrazine and glyphosate, based on a 2017 survey conducted by

Schryver et al. (2017a). Its dioecious and anemophilous biology, coupled with the potential for pollen-mediated gene flow, means that waterhemp has the potential to rapidly transfer herbicides resistance genes among biotypes through outcrossing, further complicating management of this species with herbicides (Liu et al. 2012; Sarangi et al. 2017).

Due to the extended emergence pattern of waterhemp, soil-applied residual herbicides do not provide consistent season-long control of this species (Soltani et al. 2009). Additionally,

Group 2 and 9 herbicides can generally be regarded as ineffective POST treatments due to widespread resistance in the USA and Ontario (Heap 2018). Steckel and Sprague (2004) found that the critical control period for common waterhemp is up to V6 corn stage in order to avoid crop yield loss, and late emerging waterhemp plants can contribute to the soil seed bank, necessitating the use of effective POST herbicides for full-season control of waterhemp in corn

(Hartzler et al. 2004).

The Group 27 (HPPD-inhibiting) herbicides in tank-mixtures with atrazine effectively control waterhemp in corn (Soltani et al. 2009; Vyn et al. 2006). Soltani et al. (2009) reported

86% control of common waterhemp with mesotrione + atrazine applied POST; Vyn et al. (2006) observed 97-99% control with the same treatment. There has been little research conducted to evaluate these treatments for control of MR waterhemp biotypes in Ontario; however, considering the limited number of effective POST herbicides available for use in corn, HPPD- inhibitors show promise as viable alternatives for short-term management of this species.

155

Tolpyralate is a new Group 27, pyrazolone herbicide that provides control of several annual grass and broadleaf weed species when applied POST alone or with atrazine at a 1:33.3 tank-mix ratio (Metzger et al. 2018b). Previous studies conducted in Ontario have documented excellent control of other Amaranthus spp. with tolpyralate (Metzger et al. 2018a), and the commercial label for tolpyralate (30-40 g ha-1) in the USA states control of both A. rudis and A. tuberculatus, while the Canadian label states suppression of A. rudis (Anonymous 2017a;

Anonymous 2017b). However, tolpyralate efficacy on MR waterhemp biotypes has not been reported. Therefore, the first objective of this study was to determine a biologically-effective dose of tolpyralate and tolpyralate + atrazine in MR waterhemp biotypes in Ontario. The second objective was to determine the relative efficacy of tolpyralate and tolpyralate + atrazine compared to two current herbicide standards in Ontario for control of MR waterhemp POST in corn.

6.3 Materials and Methods

Experimental Methods Field experiments were conducted in 2018 at a field site near Cottam, (Essex County)

Ontario and at two sites (Corn Crib Rd and Brother Rd) on Walpole Island in Lambton County,

Ontario, Canada with infestations of waterhemp that were previously determined to be resistant to glyphosate, imazethapyr, and atrazine in greenhouse studies by Schryver et al. (2017a). More recently, gene sequencing from bio-leaf assays confirmed the hypothesis put forth by Schryver et al. (2017a), that waterhemp populations in Essex County and in Lambton County evolved resistance independent of one another (J. Kreiner, PhD candidate, University of Toronto Dept. of

Ecology and Evolutionary Biology, unpublished data). Multiple-herbicide resistance was further confirmed at each trial site during the year of study by establishing permanent quadrat

156 experiments. Quadrats were sprayed with either glyphosate (900 g ae ha-1); imazethapyr (100 g ai ha-1) + Agral® 90 non-ionic surfactant (Syngenta Canada Inc., Guelph, ON) (0.25% v/v) +

28% N urea ammonium nitrate (UAN) (2.0 L ha-1); or atrazine (1500 g ai ha-1) + Assist® oil concentrate (BASF Canada Inc. Missisauga, ON) (1.0 L ha-1), and calculating the ratio of surviving plants within each quadrat 2 weeks later. Resistance ratios generated from these supplementary quadrat experiments are listed in Table 6.1.

In the spring, trial sites were fertilized according to soil test results and crop requirements and cultivated twice with a tandem disc to prepare the seedbed. Dekalb DKC46-82RIB corn

(Monsanto Co., St. Louis, MO) was seeded 4.5 cm deep in plots which were 2.25 m wide and 8 m long, on 76 cm row spacing at a population of 83 140 seeds ha-1. Treatments were arranged in a randomized complete block with four replications. Shortly after crop emergence, glyphosate

(450 g ae ha-1) was applied to the entire trial area to control all non-target weed species.

Treatments were applied when MR waterhemp plants reached an average of 10 cm in height, using a CO2-pressurized backpack sprayer and handheld spray boom outfitted with four

ULD12002 nozzles (Pentair, New Brighton, MN). For the dose response analysis, six rates of tolpyralate (Shieldex 400SC, ISK Biosciences Corporation, Concord, OH) (3.75, 7.5, 15, 30, 60 and 120 g ha-1) were applied alone, and in a 1:33.3 tank-mixture with atrazine at rates 125, 250,

500, 1000, 2000 and 4000 g ha-1. Methylated seed oil (MSO Concentrate®, Loveland Products

Inc., Loveland CO) and 28% N UAN were included with each tolpyralate application at 0.50% v/v and 2.50% v/v, respectively. Industry standard herbicides included a pre-mixture (1:2 ratio) of dicamba/atrazine (1500 g ha-1) (Marksman® Herbicide, BASF Canada Inc. Mississauga, ON) and mesotrione + atrazine (100 + 280 g ha-1). Mesotrione + atrazine applications included a non- ionic surfactant, (Agral® 90, Syngenta Canada Inc., Guelph, ON) at (0.20% v/v) in accordance

157 with local recommendations (Ontario Ministry of Agriculture, Food and Rural Affairs

[OMAFRA], 2018a). Additionally, each replicate included a non-treated control (NTC), and a weed-free control (WFC), which was established using a pre-formulated mixture of S- metolachlor/mesotrione/bicyclopyrone/atrazine (2022 g ha-1) (Acuron™ Herbicide, Syngenta

Canada Inc. Guelph, ON) applied PRE, followed by dicamba/atrazine (1500 g ha-1) applied

POST, and subsequent hand-weeding where required.

Data collection included crop injury 1, 2 and 4 WAA, and weed control 2, 4, 8 and 12

WAA, where each parameter was visually assessed against the NTC and assigned a percent value (0-100) where 0 represents no visible crop/weed injury and 100 represents complete plant death. Waterhemp density and dry biomass were determined in each plot 8 WAA by randomly placing two, 0.25 m2 quadrats in the centre row of each plot, counting and cutting the plants contained in each quadrat, and drying the harvested biomass to constant mass at 60 C. Grain yield was taken from the centre two rows of each plot by harvesting with a small-plot combine and weighing the sample. For statistical analysis, grain yields were expressed as t ha-1 corrected to 15.5% moisture, or as a percent relative to the WFC plot within replications. Similarly, waterhemp density and dry biomass were expressed as a percent relative to the NTC plot within replications, to account for variability in waterhemp density across field sites (Table 6.1).

Statistical Analysis

Regression Analysis

Visible waterhemp control at 2, 4, 8 and 12 WAA, waterhemp density and dry biomass, and relative corn grain yield were each regressed against the applied dose of tolpyralate alone, and the combined dose of tolpyralate + atrazine, by fitting a four-parameter log-logistic model

158

(Equation 1) to the data using the NLIN procedures in SAS v. 9.4 (SAS Institute, Cary, NC).

Predicted values generated from the regression analysis were used to compute the required dose

(Rn) of tolpyralate or tolpyralate + atrazine for 50, 80 and 95% control, a 50, 80 and 95% reduction in density and dry biomass (relative to the NTC) and to achieve 50, 80 and 95% of the yield of WFC plots within replications. Where computation of the required dose was not possible with the regression model, values are substituted with “Non-est.’ in Tables.

Equation 1 – Log-logistic dosage-response model

(-b*(log(dose)-log(I ))) y = C + (D-C)/(1+e 50 )

Where:

y = response parameter

C = lower asymptote of percent control or upper asymptote of

percent density and biomass reduction

D = upper asymptote of percent control or lower asymptote of

percent density and biomass reduction

I50 = dose eliciting a response equidistant between C and D

b = slope about I50

Analysis of Variance and Least-Square Means Comparisons

159

A generalized linear mixed-model analysis of variance was performed on each variable for the purpose of comparing specific treatments including tolpyralate applied at a label rate alone or with atrazine, mesotrione + atrazine and dicamba/atrazine, using the GLIMMIX procedures in SAS (SAS Institute, Cary, NC). The NTC was excluded from analysis of visible control data, as control was assumed to be zero. Variance was partitioned into the random effects of environment, replication within environment, and the treatment by environment interaction, while treatment was designated as the fixed effect. The significance of random effects was determined with a log-likelihood ratio test, and fixed effects using an F-test, with α set to 0.05 for all tests. An appropriate model was assigned to each response parameter based on the distribution and link which best met assumptions that residuals had a mean of zero, were homogeneous and were normally distributed according to a Shapiro-Wilk test and visual inspection of scatterplots of studentized residuals. Visible control data and grain yield were modelled using a normal or lognormal distribution with identity link; percent reduction in MR waterhemp density and dry biomass were fit to a lognormal distribution. Where a lognormal distribution was specified, the data were back-transformed within the GLIMMIX procedure using the omega method (M. Edwards, Ontario Agricultural College Statistician, University of

Guelph, personal communication). Least-square means of each parameter were compared across treatments using a Tukey-Kramer test and assigned letter codes for presentation in Table 6.4.

6.4 Results and Discussion

Crop Injury On average, no treatment caused crop injury in excess of 5%; however, tolpyralate + atrazine at the highest tested dose (120 + 4000 g ha-1), caused up to 13% injury 2 WAA at one location (data not presented). Where injury did occur, symptoms consisted of blotched chlorosis

160 or temporary white bleaching of the youngest corn leaves unfurled at the time of application.

Injury symptoms diminished to <10% by 4 WAA. Greater injury was observed where the application was made under high ambient temperatures (27 C), following a period of excessive rainfall that caused temporary flooding. Temperature has been identified as one factor impacting activity of other HPPD-inhibitors in the past. Johnson and Young (2002) observed greater foliar activity with mesotrione in some weed species when applied under higher temperatures (32 C), compared with applications made under lower temperatures (18 C). Therefore, it is possible that the combination of environmental stressors at this location predisposed the corn to herbicide injury.

Dose-Response Control of MR waterhemp varied by site, presumably due to differences in waterhemp density (Table 6.1); it is probable that the high density at Cottam reduced waterhemp spray coverage. Control was generally better at the Corn Crib Rd and Brother Rd sites compared to the

Cottam site, despite notable differences in atrazine susceptibility across these sites based on quadrat experiments conducted in the year of study (Table 6.1). Quadrat results suggest that the waterhemp population at Corn Crib Rd was in the inchoate stages of segregation towards a completely atrazine-resistant population. Regardless, sites were combined for these analyses in order to gain an understanding of the average activity of tolpyralate on heterogeneous waterhemp populations throughout Ontario.

Tolpyralate applied alone caused white bleaching in the growing point of waterhemp plants within 7 days of application, while the addition of atrazine to tolpyralate induced greater leaf necrosis at this timing (data not presented). Similar observations were made in a previous study of the biologically-effective dose of tolpyralate in a mixed population of green pigweed

161

(Amaranthus powelli S. Watson) and redroot pigweed (Amaranthus retroflexus L.) (Metzger et al. 2018a). At 2 WAA, tolpyralate applied alone did not control waterhemp 80 or 95% at the tested doses in this study (≤120 g ha-1) (Table 6.2). However, waterhemp could be controlled

95% 2 WAA with tolpyralate + atrazine applied at 60.5 + 2015.5 g ha-1 (Table 6.3). At 2 WAA,

10.2 + 339.8 g ha-1 tolpyralate + atrazine was found to provide 80% control, while 2.1 + 69.7 g ha-1 was required for 50% control. Similar to a previous study (Metzger et al. 2018a), the addition of atrazine to tolpyralate increased speed of activity in MR waterhemp, and facilitated a reduction in the R50 value for tolpyralate of over 50% despite the presence of atrazine-resistant biotypes (Tables 6.2 and 6.3). In contrast, Kohrt and Sprague (2017a) reported no benefit to adding atrazine (560 g ha-1) to tolpyralate (40 g ha-1) for control of an atrazine-resistant Palmer amaranth (Amaranthus palmeri S. Watson) population. Interestingly, Rn values for tolpyralate and tolpyralate + atrazine were considerably higher for waterhemp 2 WAA than for other

Amaranthus spp. tested in previous experiments in Ontario (Metzger et al. 2018a). For example,

-1 the R50 value for tolpyralate in green/redroot pigweed was 1.5 g ha at 2 WAA (Metzger et al.

2018a), while 4.9 g ha-1 was required to control waterhemp 50% at the same timing (Table 6.3).

-1 The 3.4 g ha difference in R50 could be due to the high densities of waterhemp compared with green/redroot pigweed (average 14 plants m-2) (Metzger et al. 2018a), or could be indicative of varying sensitivity to tolpyralate among Amaranthus spp.

Subtle differences were observed in waterhemp control 4 WAA with tolpyralate + atrazine compared to the control observed at 2 WAA. Tolpyralate + atrazine at rates ≤57.1 +

1903.5 g ha-1 was sufficient to provide 95% control of waterhemp 4 WAA (Table 6.3), while the

R50 values for tolpyralate + atrazine were identical 2 and 4 WAA (Tables 6.2 and 6.3). In contrast, control with tolpyralate alone improved from 2 to 4 WAA; waterhemp could be

162 controlled 80% with tolpyralate alone (34.6 g ha-1), although no dose ≤120 g ha-1 gave 95% control (Table 6.2). These results contrast those presented by Kikugawa et al. (2015), who reported 99% control of a GR tall waterhemp biotype 29-30 days after application (DAA) with tolpyralate applied at 30 g ha-1; however, their results were from a controlled experiment, where treatments were applied to individual plants grown in a greenhouse (Kikugawa et al. 2015). In contrast, Kohrt and Sprague (2017a) reported 96% control of a Palmer amaranth biotype resistant to glyphosate, ALS-inhibitors and atrazine with tolpyralate (40 g ha-1) in a field experiment conducted using methodology similar to this study.

Despite the extended emergence pattern of waterhemp in Ontario described by Schryver et al. (2017b), control generally improved at later assessment timings with tolpyralate and tolpyralate + atrazine (Tables 6.2 and 6.3). At 8 WAA, tolpyralate alone at 26.8 g ha-1 controlled waterhemp ≥80%, while 5 g ha-1 was sufficient for 50% control at this timing (Table 6.2).

Similar to earlier assessment timings, no dose of tolpyralate alone (≤120 g ha-1) was adequate for

95% control of waterhemp. In contrast, green/redroot pigweed could be controlled 80% at 8

WAA with tolpyralate alone at 8.5 g ha-1 (Metzger et al. 2018a). When applied in conjunction

-1 with atrazine, the R50 value for tolpyralate was reduced by 60% from 5 to 2 g ha (Table 6.3).

The R80 and R95 values for tolpyralate + atrazine changed little at 8 WAA compared with 4

WAA; 9.7 + 322 g ha-1 provided 80% control, while 57 + 1901.2 g ha-1 provided 95% control.

By 12 WAA, corn at each site had matured and begun to dry down. Previous work has not examined the efficacy of tolpyralate beyond 8 WAA; this assessment timing provided an indication of full-season control. At this assessment timing, 5.1 g ha-1 of tolpyralate alone controlled waterhemp 50%, while 24.5 g ha-1 could provide 80% control; but no dose gave 95% control (Table 6.2). At 12 WAA, 46.5 + 1550.6 g ha-1 of the tolpyralate + atrazine tank-mixture

163 was adequate for 95% control of waterhemp; but this is beyond the current maximum label rate for tolpyralate of 40 g ha-1 (Anonymous 2017a). In contrast, regression analysis showed 1.9 +

64.9 g ha-1 was sufficient for 50% control 12 WAA, while 8.7 + 289.1 g ha-1 could provide 80% control, or approximately 30% of the low label rate of tolpyralate.

Density and biomass data collected 8 WAA corroborated visible control data; however, waterhemp biomass could typically be reduced 50, 80 or 95% with a comparatively lower dose of tolpyralate or tolpyralate + atrazine than was required for the equivalent reduction in density.

These results are representative of the variable weed control obtained where waterhemp density was in excess of 1000 plants m-2. When applied alone, a similar dose of tolpyralate was required for a 50% reduction in waterhemp density (3.1 g ha-1) and dry biomass (3.0 g ha-1) (Table 6.2). In contrast, 6.8 g ha-1 of tolpyralate reduced waterhemp biomass by 80% and 18.7 g ha-1 was required for the same reduction in density. Tolpyralate (30.3 g ha-1) reduced waterhemp biomass by 95%; however, no dose of tolpyralate alone (≤120 g ha-1) could reduce density by 95%. In agreeance with visible control data, the addition of atrazine to tolpyralate improved control of waterhemp and provided a greater reduction in both density and biomass of waterhemp compared to tolpyralate alone (Table 6.3). A 50, 80 and 95% reduction in waterhemp biomass could be obtained with 0.9 + 30.9, 3.3 + 110 and 13.7 + 458.2 g ha-1 tolpyralate + atrazine, respectively. In contrast, to reduce density 80 or 95%, rates of 6.7 + 223.4 and 43.2 + 1440 g ha-1 tolpyralate + atrazine were required. Similar to tolpyralate alone, tolpyralate + atrazine R50 values for waterhemp density and biomass reduction were numerically closer to one another than

were R80 or R95 values.

164

Tolpyralate, Tolpyralate plus Atrazine and Herbicide Standards Least-square means of each assessment parameter for tolpyralate alone (30 g ha-1) representing the low label rate (Anonymous, 2017a) or with atrazine (1000 g ha-1), representing a

1:33.3 tank-mix ratio, were compared to determine the effect of atrazine addition to tolpyralate for control of MR waterhemp. Tolpyralate and tolpyralate + atrazine were also compared to dicamba/atrazine (1500 g ha-1) and mesotrione + atrazine (100 + 280 g ha-1), which represent the two most efficacious POST herbicides for control of GR/MR waterhemp in Ontario (Vyn et al.

2006).

Tolpyralate + atrazine and mesotrione + atrazine provided better MR waterhemp control

(>90%) 2 WAA than tolpyralate alone and dicamba/atrazine (<70%) (Table 6.4). Consistent with these findings, Woodyard et al. (2009) reported 87-98% common waterhemp (Amaranthus rudis

Sauer) control with mesotrione + atrazine (105 + 280 g ha-1) 10 DAA, while mesotrione applied at the same rate alone provided 53-75% control; this biotype was presumably susceptible to atrazine however. Interestingly, Kohrt and Sprague (2017b) reported 88-96% control of MR

Palmer amaranth with topramezone (18 g ha-1), another pyrazolone herbicide, at 2 WAA. At 4

WAA, tolpyralate + atrazine, mesotrione + atrazine and dicamba/atrazine each controlled waterhemp similarly. Tolpyralate + atrazine provided better control at this timing than tolpyralate alone; however, control with tolpyralate alone was similar to industry standards. Vyn et al. (2006) also reported similar waterhemp control with dicamba/atrazine (1800 g ha-1) and mesotrione + atrazine (100 + 280 g ha-1) at 4 WAA. In contrast to the results of the present study,

Kohrt and Sprague (2017a) observed no difference in control of atrazine-resistant Palmer amaranth 3 WAA with tolpyralate alone compared with tolpyralate + atrazine, although a higher rate of tolpyralate (40 g ha-1), and a lower rate of atrazine (560 g ha-1) were used in their study.

Additionally, that population had a comparatively higher proportion of atrazine-resistant

165 individuals (39%) (Kohrt and Sprague 2017a) than the trial locations used in the current study

(5-26%) (Table 6.1). At 8 and 12 WAA assessment timings, tolpyralate alone, tolpyralate + atrazine, mesotrione + atrazine and dicamba/atrazine all provided ≥83% control of MR waterhemp, and were determined to be similar (Table 6.4). Although statistically similar, given the exorbitant reproductive potential of waterhemp (Hartzler et al. 2004), it is likely that the eleven percentage-point difference in control with tolpyralate compared to tolpyralate + atrazine would have important biological and agronomic implications for subsequent crops. In a previous study that included green/redroot pigweed, tolpyralate, tolpyralate + atrazine and mesotrione + atrazine applied POST at the same rates used in the present study each provided ≥91% control 8

WAA, and were determined to be similar (Metzger et al 2018b).

Tolpyralate alone, tolpyralate + atrazine, mesotrione + atrazine and dicamba/atrazine reduced waterhemp density similarly, to 0.2-5.0% of the NTC; however, density with tolpyralate alone was not significantly different than that of the NTC (Table 6.4). This finding is presumably a product of the random sampling method used for collection of waterhemp density and biomass data, which inevitably could not adequately capture the natural population variability within individual trial sites. Certainly, the 95% reduction in waterhemp density observed with tolpyralate would be biologically significant in a field environment. In contrast, each of the four herbicide treatments reduced waterhemp dry biomass compared to the NTC. Tolpyralate + atrazine reduced waterhemp dry biomass to 0.1% of that of the NTC, a greater reduction than with dicamba/atrazine, but similar to that provided by tolpyralate alone and mesotrione + atrazine. Tolpyralate alone provided a similar reduction in waterhemp dry biomass compared to both industry standards. Similar reductions in the biomass of atrazine-resistant Palmer amaranth were reported by Kohrt and Sprague (2017a) with tolpyralate, tolpyralate + atrazine and

166 mesotrione + atrazine 3 WAA. In this study however, collecting density and dry biomass 8 WAA allowed for continued emergence of waterhemp after treatments were applied. Despite the limited residual activity of tolpyralate (Anonymous 2017a), control assessments 8 and 12 WAA coupled with density/dry biomass assessments indicate that tolpyralate may provide residual control of late-emerging waterhemp.

Corn Grain Yield Yield loss where waterhemp was left uncontrolled for the duration of the study was 28% compared to the WFC; however, no statistically significant differences were observed among grain yields obtained with tolpyralate, tolpyralate + atrazine, or either industry standard herbicide

(Table 6.4). Due to yield loss being less than 50%, required doses of tolpyralate and tolpyralate + atrazine to achieve 50% of the yield in WFC plots could not be computed (Tables 6.2 and 6.3).

Applied alone, tolpyralate at 1.6 and 20 g ha-1 was sufficient to maintain 80 and 95% of the yield relative to WFC plots within replications, respectively (Table 6.2). Consistent with control assessments, a lower dose was required when atrazine was added to tolpyralate; 0.2 + 6.8 and

15.4 + 512.9 g ha-1 were required to maintain 80 and 95% relative yield, respectively (Table 6.3).

Despite no statistically significant effect of waterhemp interference on corn grain yield in this study, a 28% reduction is likely to be economically significant. Typically, waterhemp is highly competitive with corn (Steckel and Sprague 2004); however, Vyn et al (2006) also observed waterhemp to be less competitive at one experimental site in Ontario, resulting in no significant differences in corn grain yield among various herbicide treatments at that location.

6.5 Conclusions

Tolpyralate applied alone (24.5 g ha-1) controlled MR waterhemp 80% 12 WAA in this study, while no rate of tolpyralate alone provided ≥95% control. Applied at a currently-registered

167 label rate of 30 g ha-1, tolpyralate controlled MR waterhemp 83% at 8 and 12 WAA, which was similar to the control with two current standard herbicide treatments (dicamba/atrazine and mesotrione + atrazine). Despite the presence of atrazine resistance in the tested waterhemp populations, the addition of atrazine to tolpyralate at a 1:33.3 ratio improved control at 2 and 4

WAA when this tank-mixture was applied at label rates. Similarly, the co-application of tolpyralate with atrazine resulted in R50, R80, and R95 tolpyralate doses that were lower than when tolpyralate was applied alone. The results from this study are in contrast with the trends observed in a previous study of the biologically-effective dose (BED) of tolpyralate (Metzger et al. 2018a), which found relatively similar BED values for tolpyralate in green/redroot pigweed when applied alone or with atrazine. Similarly, the addition of atrazine to tolpyralate at a label rate provided no significant benefit for control of green/redroot pigweed 2 or 4 WAA, possibly indicating that A. retroflexus/A. powelli are comparatively more sensitive to tolpyralate than waterhemp is. Across all assessment parameters except 2 WAA, tolpyralate and tolpyralate + atrazine controlled MR waterhemp equivalent to mesotrione + atrazine and dicamba/atrazine. Numerically, tolpyralate + atrazine and mesotrione + atrazine provided the greatest control of MR waterhemp across assessment parameters, substantiating previous reports of excellent POST control of this species in corn with HPPD/PSII-inhibitor tank-mixtures (Soltani et al. 2009; Vyn et al. 2006). Corn grain yield was not significantly affected by waterhemp interference; however, the numerical difference in yield between treated plots and the NTC would likely have economic significance.

Considering the management challenge posed by waterhemp biotypes with resistance to up to four herbicide sites of action, HPPD-inhibiting herbicides, including tolpyralate, will likely serve as the foundation for POST herbicide management of this species in corn in Ontario. However, populations of waterhemp with resistance to HPPD-inhibitors have been confirmed in Illinois,

168

Iowa and Nebraska (Hausman et al. 2011; McMullen and Green 2011; Oliveira et al. 2017), suggesting that without careful management, resistance to this mode of action could also evolve in waterhemp populations in Ontario. In order to prolong the usefulness of HPPD-inhibiting herbicides for control of MR waterhemp in Ontario, increased emphasis should be placed on the diversification of weed management tactics, including through crop rotation and tillage, and

HPPD-inhibitors should always be co-applied with alternative, effective herbicide modes of action.

169

Table 6.1. Locations, planting, harvest and spray dates, corn growth stage and multiple herbicide-resistant waterhemp (Amaranthus tuberculatus (Moq.) J.D. Sauer) height and density at application for field experiments conducted near Cottam and on Walpole Island, Ontario, Canada in 2018. Resistance factors (%)a Application information Site GPS Corn Waterhemp Waterhemp Planting Harvest Spray coordinates 9b 5 2 growth heightc densityd date date date stage (cm) (plants m-2) 42.1494, - Cottam 88 24 84 Jun-21 V5 8 1513 Jun-2 Dec-10 82.6849

Corn Crib 42.5578, - 53 5 59 Jun-29 V6 14 578 May-28 Dec-12 Rd 82.4635

42.5612, - Brother Rd 60 26 57 Jul-9 V6 9 301 Jun-7 Nov-13 82.5004 aMean number of surviving waterhemp plants 2 weeks after application divided by number of waterhemp plants sprayed within eight quadrats per site. bHerbicide mode of action classification (Weed Science Society of America); 9, glyphosate; 5, atrazine; 2, imazethapyr. cMean height based on two or more measurements per replication within trials at time of application. dMean density based on two random stand counts in each non-treated control plot within each trial.

170

Table 6.2. Regression parameters and the predicted dose of tolpyralate required to obtain 50, 80 and 95% visible control of multiple herbicide-resistant common waterhemp (Amaranthus tuberculatus (Moq.) J.D. Sauer) 2, 4, 8 and 12 WAA, a 50, 80 and 95% reduction in density and dry biomass, and 50, 80 and 95% of the corn grain yield in WFC plots, from field experiments conducted in Ontario, Canada in 2018, based on the log-logistic dose response equation (Equation 1)a. b Regression parameters (±SE) Predicted tolpyralate dose (g ai ha-1) Variable C D b I50 R50 R80 R95

Visible control

2 WAA 0 (0) 89.78 (27.98) 0.53 (0.45) 3.15 (3.07) 4.9 Non-est. Non-est.

4 WAA 0.01 (3.73) 100 (0) 0.70 (0.08) 4.77 (0.88) 4.8 34.6 Non-est.

8 WAA 0 (0) 95.78 (8.73) 0.91 (0.32) 4.54 (1.06) 5.0 26.8 Non-est.

12 WAA 0 (0) 95.47 (8.34) 0.98 (0.35) 4.60 (1.04) 5.1 24.5 Non-est.

Densityc 100 (7.63) 1.19 (15.49) 0.80 (0.54) 3.04 (1.35) 3.1 18.7 Non-est.

Dry biomass 100 (5.39) 3.82 (3.53) 1.86 (0.74) 2.87 (0.55) 3.0 6.8 30.3

Grain yieldd 73.24 (5.81) 96.88 (5.17) 1.34 (2.10) 3.22 (2.81) Non-est. 1.6 20.0 aAbbreviation: WAA, weeks after application bRegression parameters: C = lower bound of percent control, upper bound of percent density/dry biomass reduction, D = upper bound of percent control, lower bound of percent density/biomass reduction, b = slope about I50, I50 = effective dose to elicit a 50% response, Rn, effective dose to elicit response level n. cPercent of density and dry biomass relative to non-treated control plots within replications. dYield expressed as percent relative the weed-free control plots within replications.

171

Table 6.3. Regression parameters and the predicted dose of tolpyralate plus atrazine required to obtain 50, 80 and 95% visible control of multiple herbicide-resistant common waterhemp (Amaranthus tuberculatus (Moq.) J.D. Sauer) 2, 4, 8 and 12 WAA, a 50, 80 and 95% reduction in density and dry biomass, and 50, 80 and 95% of the corn grain yield in WFC plots, from field experiments conducted in Ontario, Canada in 2018, based on the log-logistic dose response equation (Equation 1)a. b Regression parameters (±SE) Predicted tolpyralate + atrazine dose (g ai ha-1)

Variable C D b I50 R50 R80 R95

Visible control

2 WAA 0.02 (2.84) 100 (0) 0.88 (0.11) 71.84 (13.65) 2.1 + 69.7 10.2 + 339.8 60.5 + 2015.5

4 WAA 0.09 (2.79) 100 (0) 0.89 (0.12) 71.89 (13.36) 2.1 + 69.7 9.9 + 330.7 57.1 + 1903.5

8 WAA 0.07 (3.20) 100 (0) 0.88 (0.13) 68.4 (15.3) 2.0 + 66.3 9.7 + 322.0 57.0 + 1901.2

12 WAA 0.03 (3.40) 100 (0) 0.93 (0.16) 66.9 (16.0) 1.9 + 64.9 8.7 + 289.1 46.5 + 1550.6

Densityc 99.96 (5.34) 0 (0) 0.84 (0.27) 43.87 (25.02) 1.3 + 42.6 6.7 + 223.4 43.2 + 1440

Dry biomass 100 (2.05) 0 (0) 1.09 (0.23) 31.84 (12.03) 0.9 + 30.9 3.3 + 110 13.7 + 458.2

Grain yieldd 73.21 (5.59) 96.0 (7.02) 0.91 (5.72) 17.98 (209.8) Non-est. 0.2 + 6.8 15.4 + 512.9 aAbbreviation: WAA, weeks after application bRegression parameters: C = lower bound of percent control, upper bound of percent density/dry biomass reduction, D = upper bound of percent control, lower bound of percent density/biomass reduction, b = slope about I50, I50 = effective dose to elicit a 50% response, Rn, effective dose to elicit response level n. cPercent of density and dry biomass relative to non-treated control plots within replications. dYield expressed as a percent relative to weed-free control plots within replications.

172

Table 6.4. Percent visible control of multiple herbicide-resistant waterhemp (Amaranthus tuberculatus (Moq.) J.D. Sauer) 2, 4, 8, and 12 WAA, percent reduction in density and dry biomass, and corn grain yield provided by tolpyralate, tolpyralate + atrazine and two industry standard herbicides from trials conducted in Ontario, Canada in 2018a. Visible Control (%) Rate Densityb Dry biomass Grain yield Treatment 2 WAA 4 WAA 8 WAA 12 WAA (g ai ha-1) (%) (%) (t ha-1)

Tolpyralate 30 69 b 79 b 83 a 83 a 5.0 ab 1.5 ab 11.9 a

Tolpyralate + 30 + 1000 94 a 94 a 94 a 94 a 0.2 a 0.1 a 11.5 a atrazine Mesotrione + 100 + 280 91 a 90 ab 90 a 92 a 1.1 a 0.4 ab 11.7 a atrazine

Dicamba/atrazine 500/1000 68 b 83 ab 85 a 87 a 3.4 a 2.4 b 10.9 a

NTC 0 - - - - 100 b 100 c 8.6 a

WFC n/a ------11.9 a aAbbreviations: WAA, weeks after application; NTC, non-treated control; WFC, weed-free control. bDensity and biomass expressed as a percent of the non-treated control within replications. (a-c) Means within a column followed by the same letter are not statistically different at α = 0.05 using a Tukey-Kramer multiple range test.

173

Chapter 7: Effect of hybrid, application timing and herbicide rate on corn tolerance to tolpyralate plus atrazine

7.1 Abstract

A wide margin of crop safety is a desirable trait of post-emergence (POST) herbicides, and investigation of crop tolerance is a key step in evaluation of new herbicides. Six field experiments were conducted in Ontario, Canada from 2017-2018 to examine the influence of corn hybrid, application rate and application timing on the tolerance of field corn to tolpyralate, a new 4-hydroxyphenyl-pyruvate dioxygenase (HPPD)-inhibitor, co-applied with atrazine. Two corn hybrids (DKC42-60RIB and DKC43-47RIB) exhibited slightly greater visible injury from tolpyralate + atrazine application than P0094AM and P9840AM at 1-2 WAA; hybrids responded similarly with respect to height, grain moisture and yield. Applications of tolpyralate + atrazine at a 2X rate (80 + 2000 g ai ha-1) induced greater injury (≤31.6%) than the field rate (40 + 1000 g ha-1) (≤11.6%); the 2X rate applied at V1 or V3 decreased corn height and slightly increased grain moisture at harvest. On average, field rates resulted in marginally higher grain yields than

2X rates. Based on mixed-model multiple stepwise regression analysis, air temperature at application, time of day, temperature range in the 24 hours prior to application and precipitation following application were useful predictor variables in estimating crop injury with tolpyralate + atrazine; however, additional environmental variables also affected crop injury. These results demonstrate the margin of corn tolerance with tolpyralate + atrazine, which provides a basis for optimization of application timing, rate, and corn hybrid selection to mitigate the risk of crop injury with this herbicide tank-mixture.

174

7.2 Introduction

Selective herbicides with a wide margin of crop safety have a fundamental role in current corn weed management programs. Since their initial release in the 1990s, several selective herbicides that inhibit the HPPD enzyme, including triketones, isoxazoles and pyrazolones, have been commercially applied pre-emergence (PRE) and POST in corn (Edmunds and Morris 2012;

Mitchell et al. 2001). Inhibition of HPPD impedes production of homogentisic acid, an essential intermediate involved in carotenoid and tocopherol biosynthesis, leaving chlorophyll and the photosynthetic complex susceptible to photo-oxidation and causing extensive bleaching of developing plant foliage (Grossmann and Ehrhardt 2007; Hawkes 2012; Kakidani and Hirai

2003). To improve efficacy, broaden weed control spectrums, and reduce selection pressure on a single herbicide mode of action (MOA), HPPD-inhibitors are commonly applied to corn in tank- mixtures with herbicides exhibiting a different MOA, including photosystem II (PSII)-inhibitors

(Abendroth et al. 2006; Vollmer et al. 2017). Through exclusion of plastoquinone from the D1 binding niche, PSII-inhibitors such as atrazine disrupt the passage of electrons through the photosynthetic electron transport chain, causing lipid peroxidation; this MOA is considered complementary to HPPD-inhibitors (Hess 2000; Abendroth et al. 2006).

Crop selectivity is the primary factor impacting the usability of all POST herbicides for application in-crop. Previous research has confirmed varying levels of sensitivity to POST acetolactate synthase-inhibiting herbicides among various field corn hybrids (Bunting et al.

2004). Similarly, several instances of hybrid-specific sensitivities have been reported with

HPPD-inhibitors including mesotrione and topramezone applied POST to sweet corn (Bollman et al. 2008; O’Sullivan et al. 2002), due to mutations of cytochrome P450 alleles in certain hybrids (Williams II and Pataky 2010). In contrast, field corn typically exhibits excellent

175 tolerance to mesotrione (Mitchell et al. 2001). Johnson et al. (2002) reported <15% injury 1

WAA with mesotrione or mesotrione + atrazine applied POST, and injury dissipated to <8% by

4 WAA. This tolerance has been attributed to reduced uptake and more rapid metabolism of mesotrione in field corn relative to susceptible plant species (Mitchell et al 2001); although differential sensitivity of the target HPPD enzyme across species has also been reported (Hawkes

2012). Similarly, field corn has been observed to exhibit excellent tolerance to topramezone

(Grossmann and Ehrhardt 2007; Rahman et al. 2013). Topramezone selectivity has been attributed to differential sensitivity of the target enzyme in corn compared to susceptible species

(Grossmann and Ehrhardt 2007). Other HPPD-inhibitors however, including isoxaflutole, have required the addition of crop safeners (compounds which enhance crop metabolism of the herbicide), to achieve acceptable crop tolerance, particularly at POST application timings

(Ahrens et al. 2013; Sprague et al. 1999).

Tolpyralate is the latest pyrazolone HPPD-inhibiting herbicide successfully commercialized for use in all types of corn. Tolpyralate is labelled for application POST, at 30-

40 g ha-1, and is recommended as a tank-mixture with atrazine at ≥560 g ha-1(Anonymous

2017a). Tolpyralate is reported to exhibit a wide margin of crop safety based on company trials

(Tonks et al. 2015). Previous work in field corn observed <10% crop injury with tolpyralate + atrazine applied at rates up to 120 + 4000 g ha-1, representing 3X the maximum labelled rate of tolpyralate (Metzger et al. 2018a). However, no published studies have specifically examined field corn tolerance to tolpyralate. Furthermore, a limited number of hybrids were used in previous experiments, and the specific effects and interactions of tolpyralate + atrazine rate, application timing and individual corn hybrid with respect to crop safety have not been reported in the literature. Previous experiments with other herbicides have linked application timing,

176 application rate and corn hybrid to crop tolerance (Ahrens et al. 2013; Bunting et al. 2004;

Johnson et al. 2002), warranting investigation into the effects of these factors with tolpyralate.

Therefore, the goal of this study was to examine corn tolerance to tolpyralate + atrazine, by examining the effects and interactions of herbicide rate, herbicide application timing, and corn hybrid selection.

7.3 Materials and Methods

Experimental Methods Three field experiments were conducted in 2017 and in 2018 at locations near Ridgetown and Exeter, Ontario, Canada, for a total of six site-year combinations. Trial sites were moldboard plowed each fall and cultivated twice each spring, following application of fertilizer as required according to soil test results. Experimental sites were maintained weed-free for the duration of the study by applying S-metolachlor/atrazine (2880 g ai ha-1) PRE followed by glyphosate (900 g ae ha-1) POST and hand-hoeing as required. In 2017, dimethenamid-P/saflufenacil (735 g ai ha-1) was applied to the experiment near Exeter instead of S-metolachlor/atrazine. No crop injury was observed as a result of any of the PRE broadcast treatments.

Four corn hybrids were planted 4-5 cm deep, in plots that were 1.5 m-wide (two corn rows 0.76 m apart and 8 or 10 m long), organized in a split-split block design, with corn hybrid designated as the main plot and tolpyralate + atrazine rate and application timing designated as sub-plots. Each combination of sub-plots was randomized within the main plots across four replications (blocks) at each site. Due to equipment and spatial limitations, corn hybrids were seeded in a continuous pattern across the four blocks at each trial site and were not randomized.

The specific seeding arrangement of corn hybrids was selected arbitrarily during trial initiation and held constant across all six site-years. Corn hybrids were selected to represent current

177 hybrids in Ontario at the time of trial initiation: Pioneer P0094AM, Pioneer P9840AM (Pioneer

Hi-Bred International Inc., Johnston, IA), Dekalb DKC42-60RIB and Dekalb DKC43-47RIB

(Monstanto Co. St. Louis, MO). Herbicide treatments consisted of tolpyralate + atrazine applied at a label rate of 40 + 1000 g ha-1, and a 2X label rate of tolpyralate + atrazine at 80 + 2000 g ha-

1. Methylated seed oil (MSO Concentrate®, Loveland Products Inc., Loveland CO) at 0.50% v/v and urea ammonium nitrate (2.50% v/v) were included with each treatment in accordance with tolpyralate label recommendations (Anonymous 2017a). Plots that received the 2X herbicide rate were sprayed twice in immediate succession with a 1X rate, so as to simulate a spray overlap scenario. Each herbicide treatment was applied PRE, V1 corn stage, V3 corn stage and V5 corn stage as determined with the leaf collar method. Non-treated control (NTC) plots for each hybrid were included within each block. Treatments were applied using a CO2-powered backpack sprayer calibrated to deliver 187 L ha-1 application volume at 255 kPa spray pressure. At

Ridgetown, a 1 m-wide spray boom equipped with three ULD12002 nozzles (Pentair, New

Brighton, MN) spaced 50 cm apart was used. At Exeter, a 2.5 m boom fitted with identical nozzles at 50 cm spacing was used. Soil characteristics, application information and planting dates are presented in Table 7.1.

Crop injury was assessed visually 1 and 2 weeks after each application, and 4 and 8 weeks after the V5 applications. At each of these timings, total plot phytotoxicity was evaluated on a percent scale, with 0 indicating no visible injury and 100 indicating complete death of the treated plants. Plant height was assessed 2 weeks after the V5 applications, by measuring and recording the height of ten, randomly-selected corn plants per plot and calculating an average height for each plot. To minimize confounding effects introduced through uncontrolled environmental variables across sites, plant height was subsequently expressed as a percent,

178 relative to the average height of the corresponding hybrid in the NTC plot within blocks.

Potential corn stand loss was determined 2 weeks after the V5 application, by indiscriminately placing a 2 m measuring stick between the two corn rows, counting and recording the number of corn plants on either side of the stick, and calculating the average number of plants m of row-1 for each plot. At maturity, the complete plot was harvested using a small-plot combine, which recorded grain weight and moisture. For data analysis, grain yields were corrected to 15.5% moisture. As with height, grain moisture and corrected grain yields were expressed as percent relative to the NTC by dividing moisture and yield of each treatment plot by moisture and yield of the corresponding hybrid in NTC plots within blocks.

Statistical Analysis

Variance analysis

A mixed-model variance analysis was performed on each response parameter using the

GLIMMIX procedure in SAS software v. 9.4 (SAS Institute, Cary, NC). The main plot factor

(corn hybrid), subplot factors (herbicide rate, application timing) and all two and three-way interactions were designated as fixed effects in the model. Significance of these fixed effects and their interactions were determined using an F-test. Environment (comprising year and location), interaction of environment with each fixed effect, block (nested within environment), and the block by hybrid interaction were designated as random effects; their significance was determined using a restricted log-likelihood test. Significance was set to α=0.05 for all statistical analysis.

Residuals were examined to confirm assumptions that they were homogeneous, had mean equal to zero and were normally distributed, using scatter plots of studentized residual values paired with a Shapiro-Wilk test. A distribution and link function were subsequently selected for each

179 assessment parameter which best met these assumptions. Crop injury 1, 2 and 4 WAA, as well as relative plant height, relative grain moisture and relative grain yield were each determined to be normally distributed, while injury 8 WAA required lognormal transformation prior to analysis. In this case, data were back-transformed for presentation using the omega procedure (M. Edwards,

Ontario Agricultural College Statistician, University of Guelph, personal communication). Least- square means of main plot and sub-plot effects as well as their interactions were computed and separated using Tukey-Kramer’s test. Significant differences in main effect or interaction least- square means were illustrated in tables using upper and lower-case letter codes, assigned using the pdmix800 macro (Bowley 2015) and slicediff commands. When two and three-way interactions were insignificant, only the main effect least-square means are presented for a given parameter, averaged across all levels of the other two factors (Table 7.2).

Multiple Stepwise Regression

A secondary analysis was conducted on visible injury parameters 1, 2, 4 and 8 WAA to identify contributing factors that may have accentuated crop injury at these timings. A mixed model multiple stepwise regression analysis was performed with twelve distinct candidate predictor variables using the GLIMMIX procedure (SAS Institute, Cary, NC). Environment

(comprising location and year) was considered a random effect in the model; its statistical significance was tested using the log-likelihood ratio test (α=0.05). Candidate predictor variables

(Table 7.5) were selected from twenty initial variables that were unique to particular trials

(environments) or applications within trials; eight of the variables were immediately eliminated from the initial list because of probable collinearity. Candidate predictor variables from Table

7.5 were further thinned based on scatter plots (generated using PROC SGPLOT) of each variable plotted against each response parameter, which were visually inspected for potential

180 correlation. Subsequently, a mixed model was optimized (based on the AICC fit statistic) for visible injury 1, 2, 4 and 8 WAA, averaged across hybrids, rates and POST application timings through repeated, stepwise elimination of insignificant (P>0.05 using the F-test) predictor variables from the analysis. Data from PRE applications were excluded from regression analyses, as no injury was observed with PRE applications (Tables 7.2-7.4). A variance inflation (VIF) analysis was conducted on all candidate predictor variables for each assessment parameter using

PROC REG; variables were eliminated with evidence of collinearity (VIF values ≥4). Residual assumptions were confirmed using PROC UNIVARIATE, consistent with the procedure outlined for the initial variance analysis. Results of the mixed model multiple regression analyses are outlined in Tables 7.6-7.9 for each visible injury assessment parameter.

7.4 Results and Discussion

Factorial Analysis of Fixed Effects Crop injury varied with herbicide rate, application timing and corn hybrid. There was no evidence of injury with PRE applications; however, injury (>30%) was observed with certain

POST timings, particularly with 2X rates (Table 7.4) producing a rate by application timing interaction for visible injury 1, 2, 4 and 8 WAA (P<0.0001), relative plant height (P=0.0009) and for relative grain moisture (P=0.0167) (Table 7.2). Crop injury symptoms varied across application timings, but generally manifested as white bleaching or yellow chlorosis in the youngest corn leaves, which progressed to leaf necrosis in severe cases. Injury was highest 1

WAA and gradually diminished by 8 WAA as injured plants developed new, unaffected leaves and shed necrotic leaf tissue. Corn stand loss was not observed with any treatment (data not presented). Since visible injury was highest 1-2 WAA, differences in corn hybrid response were more evident at these timings. Corn injury across application timings depended on the hybrid,

181 which produced a timing by hybrid interaction at 1 WAA (P=0.0346; Table 7.2), while hybrid alone was statistically significant as a main effect at 2 WAA (P=0.0001; Table 7.2). Overall, no three-way interactions between fixed effects were statistically significant for any assessment parameter (P≥0.1972).

Visible corn injury across all four hybrids was most severe 1 WAA with V1 or V3 applications, when averaged across rates (Table 7.3). No injury was observed with PRE applications, while injury from V5 applications was ≤3.8% regardless of hybrid. O’Sullivan et al.

(2002) also reported no injury to sweet corn hybrids with mesotrione applied PRE; whereas

POST applications caused severe injury. Hybrids responded similarly to tolpyralate + atrazine applications made at V1 stage; injury ranged from 20.6 to 22.5% and consisted mainly of white bleaching, symptoms which are characteristic of HPPD-inhibitor injury (Abendroth et al. 2006;

Grossmann and Ehrdardt 2007). In accordance with label recommendations (Anonymous 2017a) tolpyralate was co-applied with atrazine in this study. Previously, Choe et al. (2014) reported an increase in sweet corn bleaching injury when atrazine was added to mesotrione, topramezone and tembotrione relative to each HPPD-inhibitor applied alone; however, this difference in injury did not translate to final ear yield. At one site in 2018 (E4; Table 7.1), substantial leaf burn was observed with V1 applications; affected foliage exhibited grey, water-soaked lesions within 4 hr of application (data not presented). These symptoms were inconsistent with typical HPPD- inhibitor bleaching injury (Abendroth et al. 2006) and were perhaps a result of the adjuvants used. When applications were made at V3, DKC43-47RIB was slightly more sensitive to tolpyralate + atrazine compared to P0094AM and P9840AM at 1 WAA (Table 7.3). By 2 WAA, both DKC43-47RIB and DKC42-60RIB were comparatively more sensitive to tolpyralate + atrazine than either P0094AM or P9840AM, irrespective of application rate or timing (Table

182

7.2). Interestingly, these hybrids are not reported to be sensitive to HPPD-inhibitors (Monsanto

Co. 2018); however, previous research has identified Ontario sweet corn hybrids that are sensitive to other HPPD-inhibitors, including mesotrione, due to the presence of mutant recessive

CYP alleles (Bollman et al. 2008; Meyer et al. 2010; O’Sullivan et al. 2002). As visible injury diminished with time, no differences in hybrid sensitivity were observed at later assessment timings or in quantitative assessment parameters (P≥0.1366; Table 7.2), suggesting that the observed differences in hybrid sensitivity in this study have negligible biological significance.

When averaged across the four hybrids 1 and 2 WAA, the 2X rate of tolpyralate + atrazine resulted in greater crop injury than the 1X rate, regardless of POST application timing

(Table 7.4). It is possible that the 2X rate of adjuvants (methylated seed oil + UAN) used in this study accentuated crop injury in addition to the effect of increasing herbicide rate. Grossmann and Ehrhardt (2007) reported an increase in weed uptake of topramezone when a non-ionic surfactant (Dash HC®, BASF SE, 2018) was added to topramezone. Similarly, Zhang et al.

(2013) reported an increase in weed control efficacy and risk of crop phytotoxicity when methylated seed oil was applied with topramezone. Therefore, the 2X adjuvant rate used in this study to simulate a spray overlap may have increased corn uptake of tolpyralate + atrazine, contributing to greater phytotoxicity. At 1 WAA, injury was 20 and 22% higher with the 2X rate than with the 1X rate, applied at V1 and V3, respectively. In contrast, the V5 application resulted in only 5.7% injury when the 2X rate was applied. Regardless of rate, corn was most susceptible to injury when tolpyralate + atrazine was applied at V1 or V3. By 2 WAA however, only the 2X rate applied at V1 or V3 caused injury greater than that observed with either PRE (0%) or V5 applications (0.8-4%). Johnson et al. (2002) observed a similar trend; crop injury with mesotrione + atrazine was higher when applied to V3 corn, compared to V4 or V5 corn.

183

However, these results were attributed to environmental conditions at the time of application rather than physiological characteristics of the corn plants.

Since 4 and 8 WAA assessments were conducted in relation to the V5 application, injury was generally more evident in plots treated at V3 and V5 timings, compared to V1 timings.

Injury with 2X rates applied at V1 decreased to 1.1% by 4 WAA, which was statistically similar to the 0% injury observed in plots treated PRE (Table 7.4). At both 4 and 8 WAA assessment timings, the rate by timing interaction was statistically significant (P<0.0001; Table 7.2). When applied at the field rate (1X), application timing had no effect on crop injury 4 or 8 WAA; although applications at V3 or V5 resulted in greater crop injury 4 WAA than applications PRE or at V1 (Table 7.4). This difference is likely due to the relatively shorter time period from the

V3 and V5 applications to the 4 WAA assessments, compared with the time from the V1 application. Similarly, an increase in injury was observed 4 and 8 WAA where the 2X rate was applied at V3 and V5, but not at earlier application timings. Regardless, crop injury was ≤4.6% 4

WAA; by 8 WAA no rate or application timing combination resulted in >0.1% injury, which would be within commercially-acceptable tolerance levels.

Consistent with early-season crop injury assessments, 2X rates of tolpyralate + atrazine applied at either V1 or V3 resulted in corn stunting relative to the NTC, while no difference in height was observed with either rate applied PRE or at V5 (Table 7.4). This effect led to a statistically significant rate by timing interaction for relative plant height (P=0.0009; Table 7.2).

Presumably due to the more severe injury observed with the 2X rate applied at V1 and V3

(Tables 7.3 and 7.4), a 4% decrease in plant height was observed where a 2X rate was applied compared to a 1X rate (Table 7.4). Similarly, a 6 and 4% decrease in height was observed where the 2X rate was applied at V1 and V3, respectively, compared to when it was applied PRE.

184

Therefore, despite plants showing <1% injury 8 WAA, plants that were most severely injured were also stunted as a result of early season herbicide injury. Furthermore, stunted plants reached anthesis later (≤7 days) than the NTC (data not presented). Despite a delay in development, negligible trends were observed in final grain yield or harvest moisture; although the rate by timing interaction was statistically significant for relative grain moisture at harvest (P=0.0167;

Table 7.2). A 1% increase in grain moisture was observed where a 2X rate of tolpyralate + atrazine was applied to V3 corn compared to a 1X rate; a 2% increase in moisture was observed where the 2X rate was applied at V3 compared with PRE (Table 7.4). The Ontario Ministry of

Agriculture, Food and Rural Affairs estimates corn drying costs to be $2.33 t-1 for each percentage point in grain moisture (OMAFRA 2018b). Therefore, despite its statistical significance, the observed difference in grain moisture is unlikely to be economically significant.

Contrary to all other assessment parameters, differences in grain yield were only detected across herbicide rates (P=0.0258; Table 7.2). A 2X rate of tolpyralate + atrazine resulted in 2% lower grain yield compared to the 1X rate. However, yield where a 2X rate was applied was still 100% of that obtained in the NTC when averaged across application timings and hybrids; no decrease in yield was observed as a result of the herbicide treatments. Similar results have been reported with sweet corn, where visible crop injury caused by mesotrione, topramezone and tembotrione did not translate to yield loss (Bollman et al. 2008; Soltani et al. 2007). Similarly, Gitsopoulos et al. (2010) reported no effect of topramezone rate or application timing on corn grain yield. In contrast, O’Sullivan et al. (2002) reported a yield decrease in a sensitive sweet corn hybrid (Del

Monte 2038) with mesotrione applied POST.

Multiple Stepwise Regression of Predictor Variables Based on multiple regression analysis, all soil parameters collected (sand, silt, clay, organic matter, pH and cation exchange capacity), were insignificant factors in modelling crop

185 injury 1, 2, 4 and 8 WAA and were therefore eliminated from the models. Conversely, several environmental variables related to temperature, precipitation and time of day were determined to be statistically significant predictors of crop injury (P≤0.0064; Tables 7.6-7.9). At 1 WAA, when injury symptoms were most pronounced, four predictor variables were significant

(P≤0.0037; Table 7.6). Injury at 1 WAA was accentuated with higher temperatures at the time of application, and with a larger temperature differential (daily high-daily low) in the 24 hours prior to application (Table 7.6); this predictor variable was also significant for injury 2 WAA (Table

7.7). In agreeance with the model, temperature differentials 24 hours prior to POST applications in this study ranged from 6.5 to 17.5 C; wider differentials were generally associated with increased injury (data not presented). Similarly, previous controlled-environment studies have determined that corn development is delayed by increasing the daily ambient temperature differential from 8.6 to 17.2 C (Coligado and Brown 1975), suggesting that wider daily temperature differentials may induce physiological stress in corn. Higher relative humidity (RH) has been suggested to increase herbicide diffusion through the leaf cuticle (Müller and Appleby

2012) and has been previously demonstrated to increase uptake and activity of several herbicides including fluroxypyr, glufosinate, acifluorfen and mesotrione (Lubbers et al. 2007; Ramsey et al.

2002; Ritter and Coble 1981; Johnson and Young 2002). However, RH was not determined to have a significant effect on crop injury in this study, despite a range from 47 to 96% RH across

POST timings (Table 7.1). At 1 WAA, crop injury was generally higher with applications made earlier in the day compared with those made in the evening (Table 7.6); though time of application in this study was arbitrary. In studies with applications made at discrete time intervals, time of day effects on herbicide efficacy have been widely reported; herbicide efficacy generally peaks during midday, possibly due to temperature, relative humidity, leaf orientation or

186 presence/absence of dew (Budd et al. 2017; Stopps et al. 2013; Stewart et al. 2009; Skuterud et al. 1998). Precipitation was also determined to be a potential factor in crop injury at each of the four assessment timings in this study (Tables 7.6-7.9). Generally, crop injury was decreased with higher cumulative precipitation in the 7 days following application; alleviation of any pre- existing moisture stress likely diminished the effects of herbicide injury. Similar to 1 WAA, visible injury at 4 and 8 WAA was associated with higher temperatures at the time of application according to the regression model (Tables 7.8 and 7.9). Johnson and Young (2002) previously reported increased foliar activity of mesotrione at higher temperatures; it is possible that a similar relationship exists with tolpyralate. Additionally, the cuticular wax on plant foliage becomes less viscous at higher temperatures, potentially allowing for greater diffusion of herbicide across the leaf membrane (Sargent 1965). While several environmental variables were identified as possible predictors of crop injury in this study, the random effect of environment was also significant in each regression model (Tables 7.6-7.9). Therefore, although the multiple regression models provide insight into factors that may have influenced crop injury with tolpyralate + atrazine, additional factors to those outlined here undoubtedly contributed to the injury observed in this study.

7.5 Conclusions

This study provides insight into the variable tolerance levels exhibited by four corn hybrids to tolpyralate + atrazine, as affected by application timing and rate. Although none of the hybrids used in this study were previously known to exhibit sensitivity to HPPD-inhibiting herbicides, DKC43-47RIB and DKC42-60RIB appear to be marginally more susceptible to short-term injury following application of tolpyralate + atrazine compared to P0094AM and

P9840AM (Tables 7.2 and 7.3). As injury diminished with time, hybrids recovered similarly; no

187 hybrid effects were observed 4 or 8 WAA, or in quantitative assessment parameters (Table 7.2).

Generally, applications of tolpyralate + atrazine at either the V1 or V3 timing induced greater crop injury than applications made at V5; injury was accentuated where a 2X rate was applied

(Table 7.4). Conversely, PRE applications caused no injury regardless of application rate or hybrid. Based on mixed model multiple stepwise regression analysis, crop injury with POST applications was generally associated with applications that occurred under warmer temperatures, and with wider daily temperature differentials recorded in the 24 hours prior to application (Tables 7.6-7.9). In contrast, injury was reduced with applications made later in the day, and when greater precipitation occurred in the 7 days following application. Soil parameters were determined to have no effect on corn injury in this study; however, the significance of the random effect of environment suggests that additional factors to those described here influenced the level of injury. The increased corn injury observed with 2X rates of tolpyralate + atrazine applied at V1 and V3 translated to a decrease in plant height, relative to the NTC, and a slight increase in grain moisture at harvest. Similarly, final relative corn grain yield was marginally higher with 1X rates compared to 2X rates, when averaged across hybrids and timings; although yield with each rate was similar to that of the NTC plots. Therefore, the application of tolpyralate

+ atrazine within the parameters of this study would be unlikely to have economic consequences, despite substantial visible injury and a degree of stunting. Future investigation into additional corn hybrids, and environmental and application variables that may affect corn tolerance to tolpyralate and tolpyralate + atrazine in field and controlled environments could aid in minimizing the risk of crop injury with this herbicide.

188

Table 7.1. Planting dates, harvest dates, soil characteristics and application information for experiments conducted near Ridgetown and Exeter, Ontario in 2017 and 2018. Soil Characteristicsb Application Information Planting Harvest Sand Silt Clay OMa CEC Temp RH Year Trial -1 pH Timing Date TOD Date Date % (meq 100 g ) (C) (%) PRE May-15 1030 17.4 53 V1 May-30 0615 16.8 80 E1c May-12 Oct-17 52 29 19 4.6 12.6 6.0 V3 Jun-7 1715 20.2 47 V5 Jun-13 2000 23.5 67 PRE May-22 0715 13.3 76 V1 Jun-5 0630 14.0 96 2017 E2 May-19 Oct-18 56 28 17 4.0 11.0 6.2 V3 Jun-12 0715 25.1 70 V5 Jun-17 0730 24.2 79 PRE May-23 0930 14.0 72 V1 Jun-5 1530 17.0 76 E3 May-18 Oct-19 41 40 19 3.6 28 7.9 V3 Jun-13 1100 28.8 47 V5 Jun-19 0845 24.9 62 PRE May-9 0700 13.5 67 V1 May-25 0730 25.4 57 E4 May-8 Oct-29 51 32 17 4.7 13.9 6.4 V3 Jun-2 0830 16.7 69 V5 Jun-8 0700 18.3 80 PRE May-25 0750 25.3 57 V1 Jun-1 0900 21.5 49 2018 E5 May-24 Nov-8 57 31 12 4.2 12.4 7.1 V3 Jun-8 0800 21.6 66 V5 Jun-18 0715 26.7 84 PRE May-16 1450 27.9 29 V1 May-26 0800 22.4 62 E6 May-11 Oct-20 39 35 26 3.4 28.3 8.0 V3 Jun-1 0900 23.5 74 V5 Jun-12 0950 23.5 65 aAbbreviations: OM, organic matter; CEC, cation exchange capacity; TOD, time of day; Temp, temperature; RH, relative humidity; PRE, pre-emergence. bRepresents average within trial sites (n=10-16 soil cores per trial; 15 cm depth). Soil characteristics were measured within replications in 2018 for multiple- regression analysis. cE1, E2, E4 and E5 denote experiments located near Ridgetown; E3, E6 denote experiments located near Exeter. 189

Table 7.2 – Least-square means and p-values for main effects and interactions of hybrid, tolpyralate + atrazine rate and application timing on visible corn injury 1, 2, 4 and 8 WAE/WAA, and on corn height, grain moisture and yield, relative to non-treated control plots within replications, from field experiments conducted near Ridgetown and Exeter, Ontario, Canada in 2017/2018a. Assessment parameter Visible injuryb Relative Relative Relative 1 WAA 2 WAA 4 WAA 8 WAA grain heightc yield Main effects moisture Corn hybrid % P0094AM 10.5 5.5 a 1.2 0.0 100 99 101 P9840AM 10.8 5.7 a 1.2 0.0 100 101 103 DKC42-60RIB 11.4 6.4 b 1.3 0.0 99 100 100 DKC43-47RIB 11.8 6.7 b 1.5 0.0 98 101 99 Hybrid p-value 0.0245 0.0001 0.2656 0.4838 0.4458 0.1230 0.4753 Rate (g ai ha-1) 40 + 1000 5.3 2.6 0.3 0.0 a 100 100 102 a 80 + 2000 17.0 9.5 2.3 0.1 b 98 100 100 b Rate p-value 0.0004 0.0034 0.0360 0.0403 0.0577 0.8800 0.0258 Timing PRE 0.0 0.0 0.0 0.1 101 100 100 V1 21.6 10.7 0.7 0.1 97 101 101 V3 19.6 11.2 2.6 0.1 99 101 100 V5 3.3 2.4 1.9 0.1 100 100 102 Timing p-value <0.0001 0.0004 0.0248 0.1038 0.0124 0.4285 0.5863 Two-way interactions Hybrid*timing 0.0346 0.1821 0.5255 0.4353 0.8561 0.8359 0.8927 Rate*hybrid 0.8699 0.3622 0.8729 0.9313 0.8319 0.8483 0.1366 Rate*timing <0.0001 <0.0001 <0.0001 <0.0001 0.0009 0.0167 0.4456 Three-way interaction Rate*hybrid*timing 0.9989 0.9930 0.7436 0.5098 0.9516 0.9975 0.1972 Means within a main effect column followed by different letters denote significant differences according to Tukey- Kramer’s multiple range test (α=0.05). aAbbreviations: WAA, weeks after application; WAE, weeks after emergence; PRE, pre-emergence; V1,3,5, vegetative growth stages 1, 3 and 5, respectively, as determined with the leaf collar method. bVisual assessments were conducted 1 and 2 weeks after emergence for PRE applications; 4 WAA/8 WAA assessments were conducted 4 and 8 weeks after V5 applications, respectively. cPlant height, grain moisture and yield were each expressed as a percent, relative to the corresponding non-treated control plots within replications.

190

Table 7.3. Interaction of tolpyralate + atrazine application timing and corn hybrid on visible injury 1 WAE/WAA in field experiments conducted in Ontario, Canada in 2017/2018a. Visible injury (%) 1 WAA Application timing Hybrid PRE V1 V3 V5 P0094AM 0.0 a Z 20.6 a Y 18.5 a Y 2.9 a Z P9840AM 0.0 a Z 21.5 a Y 18.5 a Y 3.1 a Z DKC42-60RIB 0.0 a Z 22.5 a Y 19.8 ab Y 3.4 a Z DKC43-47RIB 0.0 a Z 21.9 a Y 21.6 b Y 3.8 a Z Means followed by the same lower case letter within a row (a-b), or upper case letter within a column (Y-Z) for each assessment parameter are not significantly different according to Tukey’s multiple means comparison test (α=0.05). aAbbreviations: WAA, weeks after application; WAE, weeks after emergence; V1,3,5, vegetative growth stages 1, 3 and 5, respectively, as determined with the leaf collar method.

191

Table 7.4. Interaction of tolpyralate + atrazine rate and application timing on visible injury 1, 2, 4 and 8 WAE/WAA, corn height and grain moisture in field experiments conducted in Ontario, Canada in 2017/2018a. Visible injury (%) 1 WAA Rate Application timing (g ai ha-1) PRE V1 V3 V5 40 + 1000 0.0 a Z 11.6 a Y 8.6 a Y 0.9 a Z 80 + 2000 0.0 a Z 31.6 b Y 30.6 b Y 5.7 b Z Visible injury (%) 2 WAA Rate Application timing (g ai ha-1) PRE V1 V3 V5 40 + 1000 0.0 a Z 5.3 a Z 4.3 a Z 0.8 a Z 80 + 2000 0.0 a Z 16.1 b Y 18.1 b Y 4.0 b Z Visible injury (%) 4 WAA Rate Application timing (g ai ha-1) PRE V1 V3 V5 40 + 1000 0.0 a Z 0.3 a Z 0.6 a Z 0.5 a Z 80 + 2000 0.0 a Z 1.1 a Z 4.6 b Y 3.3 b Y Visible injury (%) 8 WAA Rate Application timing (g ai ha-1) PRE V1 V3 V5 40 + 1000 0.0 a Z 0.1 a Z 0.1 a Z 0.1 a Z 80 + 2000 0.0 a Z 0.1 a Z 0.1 b Z 0.1 b Z Relative height (%) Rate Application timing (g ai ha-1) PRE V1 V3 V5 40 + 1000 101 a Z 99 a Z 101 a Z 100 a Z 80 + 2000 101 a Z 95 b X 97 b XY 99 a YZ Relative grain moisture (%) Rate Application timing (g ai ha-1) PRE V1 V3 V5 40 + 1000 100 b Z 100 a Z 100 a Z 100 a Z 80 + 2000 99 a Z 101 a YZ 101 b Y 100 a YZ Means followed by the same lower case letter within a column (a-b), or upper case letter within a row (Y-Z) for each assessment parameter are not significantly different according to Tukey’s multiple means comparison test (α=0.05). aAbbreviations: WAA, weeks after application; WAE, weeks after emergence; V1,3,5, vegetative growth stages 1, 3 and 5, respectively, as determined with the leaf collar method. bVisible assessments were conducted 1 and 2 weeks after emergence for PRE applications; 4 WAA/8 WAA assessments were conducted 4 and 8 weeks after V5 applications, respectively. cPlant height and grain moisture were expressed as a percent, relative to the corresponding non-treated control plots within replications.

192

Table 7.5. Summary of candidate predictor variables entered in mixed-model multiple stepwise regression analysis for visible crop injury 1, 2, 4 and 8 WAA following application of tolpyralate + atrazine at two rates and three application timings to four corn hybrids in experiments conducted in Ontario, Canada in 2017 and 2018a. Predictor variable Soil parameters Abbreviation Sand (%) %Sa Silt (%) %Si Clay (%) %Cl Organic matter (%) OM pH pH Cation exchange capacity (meq 100 g1) CEC Air temperature parameters (°C)b Temperature differential 24 hours prior to application ΔTPRE Temperature differential 24 hours after application ΔTPO Air temperature at applicationc T Application parametersc Relative humidity at application (%) RH Time of application (24 hour clock) TOD Precipitation parameterb Cumulative precipitation 7 days after application (mm) PRECIP aAbbreviation: WAA, weeks after application. bTemperature differentials (daily max-daily min) and cumulative precipitation were obtained from permanent weather stations situated <1 km from trial sites. cApplication variables measured using a portable digital weather meter (Kestrel Meters, Boothwyn, PA).

193

Table 7.6. Summary of mixed-model multiple stepwise regression analysis of significant (P<0.05) predictor variables for visible crop injury 1 WAA following application of tolpyralate + atrazine at two rates and three timings to four corn hybrids, from experiments conducted near Ridgetown and Exeter, Ontario, Canada in 2017 and 2018a. Predictor variables Visible injury 1 WAA Estimate (±SE) F-value Pr>F Temperature differential 24 hr prior (ΔTPRE) 0.68 (0.218) 9.67 0.0020 Air temperature at application (T) 0.51 (0.175) 8.51 0.0037 Precipitation 7 days post (PRECIP) -0.65 (0.057) 130.49 <0.0001 Time of day (TOD) -0.01 (0.002) 38.10 <0.0001

Estimate (±SE) Chi-Sq Pr>Chi Sq Random effect Environment 16.30 (11.932) 25.31 <0.0001

Equation Injury 1 WAA = 12.37 + 0.68(ΔTPRE) + 0.51 (T) – 0.65(PRECIP) – 0.01(TOD) + 16.3 aPredictor variables deemed insignificant (P>0.05) were eliminated from the model.

194

Table 7.7. Summary of mixed-model multiple stepwise regression analysis of significant (P<0.05) predictor variables for visible crop injury 2 WAA following application of tolpyralate + atrazine at two rates and three timings to four corn hybrids, from experiments conducted near Ridgetown and Exeter, Ontario, Canada in 2017 and 2018a. Predictor variables Visible injury 2 WAA Estimate (±SE) F-value Pr>F Temperature differential 24 hr prior (ΔTPRE) 0.80 (0.141) 32.28 <0.0001 Precipitation 7 days post (PRECIP) -0.26 (0.031) 67.32 <0.0001

Estimate (±SE) Chi-Sq Pr>Chi Sq Random effect Environment 5.72 (4.030) 31.22 <0.0001

Equation Injury 2 WAA = 0.88 + 0.8(ΔTPRE) – 0.26(PRECIP) + 5.72 aPredictor variables deemed insignificant (P>0.05) were eliminated from the model.

195

Table 7.8. Summary of mixed-model multiple stepwise regression analysis of significant (P<0.05) predictor variables for visible crop injury 4 WAA following application of tolpyralate + atrazine at two rates and three timings to four corn hybrids, from experiments conducted near Ridgetown and Exeter, Ontario, Canada in 2017 and 2018a. Predictor variables Visible injury 4 WAA Estimate (±SE) F-value Pr>F Air temperature at application (T) 0.26 (0.037) 48.29 <0.0001 Precipitation 7 days post (PRECIP) -0.03 (0.012) 7.50 0.0064

Estimate (±SE) Chi-Sq Pr>Chi Sq Random effect Environment 2.43 (1.579) 135.76 <0.0001

Equation Injury 4 WAA = -3.52 + 0.26(T) – 0.03(PRECIP) + 2.43 aPredictor variables deemed insignificant (P>0.05) were eliminated from the model.

196

Table 7.9. Summary of mixed-model multiple stepwise regression analysis of significant (P<0.05) predictor variables for visible crop injury 8 WAA following application of tolpyralate + atrazine at two rates and three timings to four corn hybrids, from experiments conducted near Ridgetown and Exeter, Ontario, Canada in 2017 and 2018a. Visible injury 8 WAA Predictor variables Estimate (±SE) F-value Pr>F Air temperature at application (T) 0.07 (0.011) 37.19 <0.0001 Precipitation 7 days post (PRECIP) -0.01 (0.004) 7.83 0.0053

Estimate (±SE) Chi-Sq Pr>Chi Sq Random effect Environment 0.05 (0.038) 27.66 <0.0001

Equation Injury 8 WAA = -1.16 + 0.07(T) – 0.01(PRECIP) + 0.05 aPredictor variables deemed insignificant (P>0.05) were eliminated from the model.

197

Chapter 8: Discussion

8.1 Contributions

This research provides a broad evaluation of tolpyralate, and the results of these six field studies offer clarity in how this new herbicide may fit into weed management in corn in Ontario.

Tolpyralate efficacy was determined through dose-response studies on common annual grass and broadleaf weeds in Ontario, and separately on multiple herbicide-resistant biotypes of Canada fleabane and waterhemp, which highlighted the activity of tolpyralate on individual weed species, investigated the benefit of atrazine as a tank-mix partner, and compared tolpyralate to existing herbicide standards in the province. Through analysis of the effects of tolpyralate + atrazine application rate and timing on control of eight weed species, the optimal application window for tolpyralate + atrazine has been identified for a range of weed species, and the potential for control of certain species with tolpyralate applied PRE was determined. Perhaps most importantly to primary producers, the level of tolerance exhibited by four locally-adapted corn hybrids to tolpyralate + atrazine was established, and the effects of application rate, timing and individual hybrid investigated, facilitating the development of application strategies that mitigate the risk of crop injury with this herbicide.

With field experiments spanning three years, the biologically-effective dose of tolpyralate and tolpyralate + atrazine (1:33.3 ratio) were determined for eight common annual grass and broadleaf weed species. Based on this study, tolpyralate alone at doses of ≤15.5 g ha-1 could control velvetleaf, common lambsquarters, common ragweed, redroot/green pigweed and green foxtail 90%; while barnyardgrass, wild mustard and ladysthumb each required the addition of

198 atrazine to tolpyralate to achieve the same level of control. When applied in tank-mixture, tolpyralate + atrazine controlled all eight of the weed species examined in this study 90% at doses of ≤13.1 + 436 g ha-1, representing <44% of the lowest label rate of tolpyralate.

Through comparison of tolpyralate applied alone compared to tolpyralate + atrazine at label rates, the benefit of atrazine addition was determined for each weed species. The addition of atrazine at a 1:33.3 tank mix ratio with tolpyralate improved control of wild mustard, ladysthumb and barnyardgrass when label rates were used, corroborating results of the dose- response analysis. Additionally, two current HPPD-inhibitors (mesotrione and topramezone) each co-applied with atrazine, were compared to tolpyralate alone and tolpyralate + atrazine, to identify differences in efficacy across three HPPD-inhibitors. Tolpyralate + atrazine provided similar weed control to topramezone + atrazine; however, control of common ragweed, green foxtail and barnyardgrass was 18-67% better with tolpyralate + atrazine than with mesotrione + atrazine.

Though tolpyralate + atrazine showed excellent efficacy based on the dose-response study, it was of interest to determine the effects of application rate and timing on tolpyralate efficacy. Common ragweed, common lambsquarters, velvetleaf and green pigweed were generally controlled equally with ePOST, mPOST or lPOST application timings; while control of green foxtail and barnyardgrass, and corn grain yield each declined when application was delayed to lPOST. Despite current registration as a POST herbicide, excellent control of green pigweed and common lambsquarters was observed with tolpyralate + atrazine applied PRE, providing a potential area for future research.

199

Given its prevalence in Ontario and widespread resistance to glyphosate and ALS- inhibitors, Canada fleabane is a key species in corn weed management programs. Similarly, waterhemp resistant to four herbicide modes of action has recently been confirmed in Ontario and recent surveys have discovered this species in previously undocumented locations in the province. Tolpyralate + atrazine at 22.3 + 741.7 g ha-1 controlled multiple-resistant Canada fleabane 95%; while waterhemp required a dose higher than current label rates for equivalent control. Applied at the label rate, tolpyralate + atrazine controlled Canada fleabane 98%, which was similar to the control with current herbicide standards: dicamba/atrazine and bromoxynil + atrazine. Similarly, tolpyralate + atrazine at the label rate controlled waterhemp 94%, similar to dicamba/atrazine and mesotrione + atrazine. Although waterhemp studies will be repeated in

2019, the results of these studies indicate that tolpyralate + atrazine may have a fit as an additional POST herbicide option for management of both Canada fleabane and waterhemp in corn in Ontario.

Regardless of weed control efficacy, the utility of any POST herbicide is dependent on crop tolerance. Corn is at greatest risk of injury with tolpyralate + atrazine applied at V1 or V3; an overlap rate (2X) accentuates injury and stunting. Additionally, regression analysis using predictor variables indicated that crop injury with tolpyralate + atrazine may be influenced by temperature, precipitation and time of day. Slight differences in sensitivity were observed among corn hybrids; though these differences did not translate to a grain yield loss. Similarly, while a

2X rate resulted in slightly lower grain yield than a 1X rate on average, neither was different from the non-treated plots. The results of this study therefore suggest that despite the potential for visible symptomology, it is unlikely that any economic loss will occur as a result of tolpyralate + atrazine injury in field corn.

200

8.2 Limitations

As with all scientific research, these experiments were subject to certain constraints and limitations despite extensive planning, preparation and attention to detail on behalf of the entire research team. Enhancements to protocol, design and technique are always possible, and identifying areas for improvement is a critical part of advancing science.

The most significant limitation of the dose-response studies (Chapters 2, 5, 6) was the lack of treatments examining atrazine alone. Had tolpyralate and atrazine been applied alone, as well as in combination, investigation into the nature of this herbicide interaction would have been possible using methods described by Colby (1967) and Flint et al. (1988). Although of scientific interest, inclusion of atrazine alone treatments would have nearly doubled the size of each experiment, putting significant constraints on time and field space. Means comparisons highlight the benefit of control with tolpyralate + atrazine, but the nature of the interaction could not be determined. Similarly, because tolpyralate and atrazine were not applied alone in the rate by application timing studies, it was impossible to determine the relative contribution of each herbicide to overall weed control, particularly with PRE treatments.

An initial goal of the application rate by timing efficacy study was to examine the influence of weed size on tolpyralate efficacy; however, because this study involved eight different weed species, size of weeds at each application timing varied. Despite a general trend towards larger weeds at later application timings, applications were triggered by the average size of the entire weed population; certain weed species reached 10, 20 and 30 cm in height well before others. This approach closely mimicked what would occur in field environments; although drawing specific conclusions based on weed size was not always possible. Identifying

201 the true effect of species-specific weed size on tolpyralate efficacy would likely require experiments to be conducted in a greenhouse, where uniformity of weed size could be guaranteed.

The corn tolerance experiment was designed with three independent factors; however, the main effect of corn hybrid was not truly randomized within experiments due to equipment limitations and consideration of practicality. While this was considered and adjusted for during analysis, it is possible that a small portion of variability or uniformity within experiments was not fully accounted for.

8.3 Opportunities for Future Research

Although this research provides a reasonably comprehensive overview of tolpyralate efficacy and corn tolerance, it also identified several areas for future research. With establishment of the biologically effective doses for ten weed species, future experiments could be designed for detection of herbicide interactions which include fewer doses of each herbicide, bracketing the biologically-effective dose. Characterization of the nature of herbicide interactions could potentially allow for a reduction in the atrazine tank-mix ratio with tolpyralate while still maintaining efficacy, which would be desirable both from an environmental and economic standpoint.

Additionally, a number of herbicide tank-mix partners could be investigated for co- application with tolpyralate in place of atrazine. Alternative herbicides may have a better fit in certain weed species or geographic areas, particularly in regions with triazine-resistance, or in jurisdictions where atrazine use may be reduced due to additional regulation or restriction. Based on these experiments, tolpyralate is weak on wild mustard and ladysthumb; tank-mix partners

202 should have efficacy on these species, but also have overlapping activity on a general weed spectrum to reduce selection pressure on HPPD-inhibitors.

Despite reports of limited soil activity, results of the rate by application timing study and the Canada fleabane and waterhemp studies indicate that tolpyralate does provide residual control of some weed species. While results for common lambsquarters and green pigweed were confounded by atrazine, control of Canada fleabane and waterhemp 8 and 12 WAA with tolpyralate applied alone provides strong evidence for further investigation into the residual activity of tolpyralate.

Given the recent commercialization of tolpyralate, its use is currently limited to corn; however, future research could focus on identifying additional use patterns. The development of soybean cultivars with resistance to other HPPD-inhibitors applied PRE may provide an opportunity for determining the tolerance of these cultivars to tolpyralate + metribuzin applied

PRE and tolpyralate applied POST. Considering the efficacy of tolpyralate + atrazine on multiple-resistant Canada fleabane and waterhemp, it is plausible that a similar tank-mixture of tolpyralate + metribuzin could be a good option for management of these species in an HPPD- inhibitor-resistant soybean system. Additionally, considering the relatively high tolerance to tolpyralate exhibited by wild mustard, future research could investigate the possibility of tolerance in Brassicaceae crops such as canola and mustard.

Overall, this research provides a general overview of weed management in corn with tolpyralate which could be of substantial value to the Ontario agricultural industry, particularly if commercial release of this herbicide is expanded. Broader use of tolpyralate on a commercial scale will undoubtedly reveal additional research opportunities and also potential risks with

203 respect to resistance management. Ongoing research will serve to maximize the value of tolpyralate to primary crop producers and continue to identify integrated resistance management strategies to preserve its utility over the longer term.

204

Chapter 9: Literature Cited

Abendroth, J. A., Martin, A. R., and Roeth, F. W. 2006. Plant response to combinations of mesotrione and photosystem II inhibitors. Weed Technol. 20:267-274. Agbakoba, C. S. O. and Goodin, J. R. 1969. Effect of stage of growth of field bindweed on absorption and translocation of 14C-labelled 2,4-D and picloram. Weed Sci. 17:436-438. Ahrens, H., Lange, G., Mueller, T., Rosinger, C., Willms, L., and van Almsick, A. 2013. 4- hydroxyphenyl pyruvate dioxygenase inhibitors in combination with safeners: solutions for modern and sustainable agriculture. Angew. Chem. Int. Ed. 52:9388-9398. Anonymous 2014. ISO approves 4 new pesticide common names. AgroNews. Online. Available from: http://news.agropages.com/News/NewsDetail---12307.htm Accessed Aug. 15 2017. Anonymous. 2017a. Shieldex® 400SC Herbicide Label. SummitAgroUSA. Durham, NC. Anonymous. 2017b. Shieldex® 400SC Herbicide Label. ISK Biosciences Corporation. Concord, OH. Armel, G. R., Wilson, H. P., Richardson, R. J., and Hines, T. E. 2003. Mesotrione, acetochlor, and atrazine for weed management in corn (Zea mays). Weed Technol. 17:284-290. Armel, G. R., Hall, G. J., Wilson, H. P. and Cullen, N. 2005. Mesotrione plus atrazine mixtures for control of Canada thistle (Cirsium arvense). Weed Sci. 53:202-211. Armel, G. R., Rardon, P. L., Mccomrick, M. C., and Ferry, N. M. 2007. Differential response of several carotenoid biosynthesis inhibitors in mixtures with atrazine. Weed Technol. 21:947-953. Armel, GR, Richardson, RJ, Wilson, HP and Hines, TE (2009). Strategies for control of horseweed (Conyza canadensis) and other winter annual weeds in no-till corn. Weed Technol. 23: 379-383 Ateh, C. M. and Harvey, R. G. 1999. Annual weed control by glyphosate in glyphosate-resistant soybean (Glycine max). Weed Technol. 13:394-398. Bader, A. 1988. How to find a great herbicide. Aldrichemica Acta. 21:15. Baerson, S. R., Rodriguez, D. J., Tran, M., Feng, Y., Biest, N. A. and Dill, G. M. 2002. Glyphosate-resistant goosegrass. Identification of a mutation in the target enzyme 5- enolpyruvylshikimate-3-phosphate synthase. Plant Physiol. 129:1265-1275.

Ballaré, C. L., Scopel, A. L. and Sanchez, R. A. 1990. Far-red radiation reflected from adjacent leaves: an early signal of competition in plant canopies. Science. 247: 329-332.

205

Barry, G. F., Kishore, G. M. and Padgette, S. R. 1992 (Monsanto Co.). World Patent WO 9204449. Online. Available from: https://patentimages.storage.googleapis.com/e9/88/d1/69ade42831757a/WO1992004449 A1.pdf Nov. 11 2017.

Barrows, G., Sexton, S. and Zilberman, D. 2014. Agricultural biotechnology: the promise and prospects of genetically modified crops. J. Econ. Perspect. 28:99-120.

BASF SE. 2018. Dash® HC. BASF Agricultural Solutions UK. Online. Available from: https://www.agricentre.basf.co.uk/agroportal/uk/en/products/product_search/product_det ails_108213.html Accessed: Nov 28 2018.

Baur, P. and Aponte, J. 2014. Co-penetration of actives and adjuvants and its significance for the matched pair liaison. Pages 23-39 in K. Myung, N. M. Satchivi, C. K. Kingston eds. Retention, uptake and translocation of agrochemicals in plants. American Chemical Society. Washington, DC, USA. Beckett, T. H., Stoller, E. W. and Wax, L. M. 1988. Interference of four annual weeds in corn. Weed Sci. 36:764–769. Beckie, H. J., Harker, K. N., Hall, L. M., Warwick, S. I., Légère, A., Sikkema, P.H., Clayton, G. W., Thomas, A. G., Leeson, J. Y., Séguin-Swartz, G. and Simard, M. J. 2006. A decade of herbicide-resistant crops in Canada. Can. J. Plant Sci. 86:1243-1264.

Beckie, H. J., Harker, N., Hall, L. M., Holm, F. A. and Gulden, R. H. 2011. Risk assessment of glyphosate resistance in western Canada. Weed Technol. 25:159-164.

Beckie, H. J., Sikkema, P. H., Soltani, N., Blackshaw, R. E. and Johnson, E. N. 2014. Environmental impact of glyphosate-resistant weeds in Canada. Weed Sci. 62:385–392. Bell, M. S., Hager, A. G. and Tranel, P. J. 2013. Multiple resistance to herbicides from four site- of-action groups in waterhemp (Amaranthus tuberculatus). Weed Sci. 61:460-468.

Beltran, E., Fenet, H., Cooper, J., and Coste, C. 2001. Kinetics of chemical degradation of isoxaflutole: influence of the nature of aqueous buffers (alkanoic acid/sodium salt vs. phosphate). Pest Manag. Sci. 57:366-371. Benbrook, C. M. 2016. Trends in glyphosate herbicide use in the United States and globally. Environ. Sci. Eur. 28:3.

Bianco de Carvalho, L. and Alves, P. L. D. C. A. 2012. Pool of resistance mechanisms to glyphosate in Digitaria insularis. J. Agric. Food Chem. 60:615-622.

Bilyea, D. 2016. Ontario Weed Survey. Ridgetown, ON, Canada: University of Guelph Ridgetown Campus Blackshaw, R. E. 1989. HOE-39866 use in chemical fallow systems. Weed Technol. 3:420-428.

206

Bollman, S. L., Kells, J. J., and Penner, D. 2006. Weed response to mesotrione and atrazine applied alone and in combination preemergence. Weed Technol. 20:903-907. Bollman, J. D., Boerboom, C. M., Becker, R. L. and Fritz, V. A. 2008. Efficacy and tolerance to HPPD-inhibiting herbicides in sweet corn. Weed Technol. 22:666-674. Bowley, S. R. 2015. Variance analyses – Gaussian. Page 57 in A hitchhiker’s guide to statistics in biology – generalized linear mixed model edition. Plants et al. Kincardine, ON.

Brown, L. R., Shropshire, C. and Sikkema, P. H. 2016. Control of glyphosate-resistant Canada fleabane in corn with preplant herbicides. Can. J. Plant Sci. 96:932-934.

Bruce, J. A. and Kells, J. J. 1990. Horseweed (Conyza canadensis) control in no-tillage soybeans (Glycine max) with preplant and preemergence herbicides. Weed Technol. 4:642-647.

Bucholtz, K. P., and Doersch, R. E. 1968. Cultivation and herbicides for weed control in corn. Weed Sci. 16:232–234. Budd, C. M., Soltani, N., Robinson, D. E., Hooker, D. C., Miller, R. T. and Sikkema, P. H. 2017. Efficacy of saflufenacil for control of glyphosate-resistant horseweed (Conyza canadensis) as affected by height, density and time of day. Weed Sci. 65:275-284. Budd, C. M., Soltani, N., Robinson, D. E., Hooker, D. C. Miller, R. T. and Sikkema, P. H. 2018. 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. and Owen, M. D. K. 1997. Emergence and survival of horseweed (Conyza canadensis). Weed Sci. 45:98-101.

Bunting, J. A., Sprague, C. L. and Riechers, D. E. 2004. Corn tolerance as affected by the timing of foramsulfuron applications. Weed Technol. 18:757-762.

Byker, H. P., Soltani, N., Robinson, D. E., Tardif, F. J., Lawton, M. B. and 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. and 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 Carey, J. B. and Kells, J. J. 1995. Timing of total postemergence herbicide applications to maximize weed control and corn (Zea mays) yield. Weed Technol. 9:356-361. Chatham, L. A., Bradley, K. W., Kruger, G. R., Martin, J. R., Owen, M. D. K., Peterson, D. E., Jugulam, M. and Tranel, P. J. 2015. A multistate study of the association between glyphosate resistance and EPSPS gene amplification in waterhemp (Amarantus tuberculatus). Weed Sci. 63:569-577.

207

Choe, E., Williams, M. M. II., Boydston, R. A., Huber, J. L., Huber, S. C. and Pataky, J. K. 2014. Photosystem II-inhibitors play a limited role in sweet corn reponse to 4- hydroxyphenyl pyruvate dioxygenase-inhibiting herbicides. Agron. J. 106:1317-1323. Cobb, A. and Reade, J. P. H. 2010. The inhibition of amino acid biosynthesis. Ch. 9 in Herbicides and plant physiology. Oxford: Wiley-Blackwell. West Sussex, UK. Coetzer, E., Al-Khatib, K. and Peterson, D. E. 2002. Glufosinate efficacy on Amaranthus species in glufosinate-resistant soybean (Glycine max). Weed Technol. 16:326-331. Colby, S. R. 1967. Calculating synergistic and antagonistic responses of herbicide combinations. Weeds. 15:20-22. Cole, D., Pallett, K., and Rodgers, M. 2000. Discovering new modes of action for herbicides and the impact of genomics. Pestic. Outlook. 11:223-229. Coligado, M. C. and Brown, D. M. 1975. Response of corn (Zea mays L.) in the pre-tassel initiation period to temperature and photoperiod. Agric Meteorol. 14:357-367. Corbett, J. L., Askew, S. D., Thomas, W. E., and Wilcut, J. W. 2004. Weed efficacy evaluations for bromoxynil, glufosinate, glyphosate, pyrithiobac and sulfosate. Weed Technol. 18:443-453. Costea, M., Weaver, S. E. and 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. Cox, W. J., Hahn, R. R. and Stachowski, P. J. 2006. Time of weed removal with glyphosate affects corn growth and yield components. Agron. J. 98:349-353. Creech, J. E., Monaco, T. A., and Evans, J. O. 2004. Photosynthetic and growth responses of Zea mays L and four weed species following post-emergence treatments with mesotrione and atrazine. Pest Manag. Sci. 60:1079-1084. Culpepper, S. and York, A. 1999. Weed management in glufosinate-resistant corn (Zea mays). Weed Technol. 13:324–333. Cerrudo, D., Page, E. R., Tollenaar, M., Stewart, G. and Swanton, C. J. 2012. Mechanisms of yield loss in maize caused by weed competition. Weed Sci. 60:225–232. Dalley, C. D., Bernards, M. L. and Kells, J. J. 2006. Effect of weed removal timing and row spacing on soil moisture in corn (Zea mays). Weed Technol. 20:399–409. Davies, J. and Caseley, J. C. 1999. Herbicide safeners: a review. Pestic. Sci. 55:1043-1058. Davis, V. M., Gibson, K. D., Bauman, T. T., Weller, S. C. and Johnson, W. G. 2007. Influence of weed management practices and crop rotation on glyphosate-resistant horseweed population dynamics and crop yield. Weed Sci. 55:508-516.

208

Davis, V. M., Gibson, K. D., Bauman, T. T., Weller, S. C. and Johnson, W. G. 2009. 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.

Derpsch, R. 2008. Critical steps to no-till adoption. Pages 479-495 in T. Goddard, M. A. Zoebisch, Y. Gan, W. Ellis, A.Watson and S. Sombatpanit, eds. No-till farming systems WASWC. Devine, M. D. 2005. Why are there not more herbicide-tolerant crops? Pest Manag. Sci. 61:312– 317. Diebold, R. S., McNaughton, K. E., Lee, E. A. and Tardif, F. J. 2003. Multiple resistance to imazethapyr and atrazine in Powell amaranth (Amaranthus powelli). Weed Sci. 51:312- 318.

Dill, G. M. 2005. Glyphosate-resistant crops: history, status and future. Pest Manag. Sci. 61:219- 224.

Dill, G. M., Sammons, R. D., Feng, P. C. C., Kohn, F., Kretzmer, K., Mersheikh, A., Bleeke, M., Honegger, J. L., Farmer, D., Wright, D. and Haupfear, E. A. 2010. Glyphosate: discovery, development, applications and properties. Pages 1-34 in Nandula, V. K. ed. Glyphosate resistance in crops and weeds: history, development and management. John Wiley & Sons, Inc., Hoboken, NJ.

Duke, S. O. and Powles, S. B. 2009. Glyphosate-resistant crops and weeds: now and in the future. AgBioForum. 12:346-357.

Edmund, R. M. Jr. and York, A. C. 1987. Factors affecting postemergence control of sicklepod (Cassia obtusifolia) with imazaquin and DPX-F6025: Spray volume, growth stage, and soil-applied alachlor and vernolate. Weed Sci. 35:216-223. Edmunds, A. J. F. and Morris, J. A. 2012. Triketones. Ch. 4.3 in Modern Crop Protection Compounds. Wiley-VCH Verlag GmbH & Co. Weinheim, Germany. Evans, S. P., Knezevic, S. Z., Lindquist, J. L., Shapiro, C. A. and Blankenship, E. E. 2003. Nitrogen application influences the critical period for weed control in corn. Weed Sci. 51:408–417. Farmer, J. A., Webb, E. B., Pierce, R. A. II and Bradley, K. W. 2017. Evaluating the potential for weed seed dispersal based on waterfowl consumption and seed viability. Pest Manag. Sci. 73: 2592-2603.

Feng, P. C. C., Cajacob, C. A., Martino-Catt, S. J., Cerny, R. E., Elmore, G. A., Heck, G. R., Huang, J., Kruger, W. M., Malven, M., Miklos, J. A. and Padgette, S. R. 2010. Glyphosate-resistant crops: developing the next generation products. Pages 45-66 in Nandula, V. K. ed. Glyphosate resistance in crops and weeds: history, development and management. John Wiley & Sons Inc., Hoboken, NJ.

209

Ferrell, J. A. and Witt, W. W. 2002. Comparison of glyphosate with other herbicides for weed control in corn (Zea mays): efficacy and economics. Weed Technol. 16:701–706. Fickett, N. D., Boerboom, C. M. and Stoltenberg, D. E. 2013. Predicted corn yield loss due to weed competition prior to postemergence herbicide application on Wisconsin farms. Weed Technol. 27:54–62. Flessner, ML, McElroy, JS, McCurdy, JD, Toombs, JM, Wehtje, GR, Burmester, CH, Price, AJ, and Ducar, JT (2015). Glyphosate-resistant horseweed (Conyza canadensis) control with dicamba in Alabama. Weed Technol. 29: 633-640 Flint, J. L., Cornelius, P. L. and Barret, M. 1988. Analyzing herbicide interactions: a statistical treatment of Colby’s method. Weed Technol. 2:304-309. Ford, G. T. and Pleasant, J. M. 1994. Competitive abilities of six corn (Zea mays L.) hybrids with four weed control practices. Weed Technol. 8:124–128. Ford, L., Soltani, N., Robinson, D. E., Nurse, R. E., Mcfadden, A. and 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. 94:1231–1237. Franz, J. E., Mao, M. K. and Sikorski, J. A. 1997. Glyphosate: A unique global herbicide. American Chemical Society. Washington, DC, USA. Gaines, T. A., Zhang, W., Wang, D., Bukun, B., Chisholm, S. T., Shaner, D. L., Nissen, S. J., Patzoldt, W. L., Tranel, P. J., Culpepper, A. S., Grey, T. L., Webster, T. M., Vencill, W. K., Sammons, R. D., Jiang, J., Preston, C., Leach, J. E. and Westra, P. 2010. Gene amplification confers glyphosate resistance in Amaranthus palmeri. PNAS 107:1029- 1034.

Gantoli, G., Ayala, V. R. and Gerhards, R. 2013. Determination of the critical period for weed control in corn. Weed Technol. 27:63–71. Ge, X.,d’Avignon, A. D., Ackerman, J. J. H. and Sammons, R. D. 2010. Rapid vacuolar sequestration: the horseweed glyphosate resistance mechanism. Pest Manag. Sci. 66:345- 348.

Gianessi, L. P., Reigner, N. P. 2007. The value of herbicides in U.S. Crop Production. Weed Technol. 21:559-566. Gillespie, W. E., Czapar, G. F. and Hager, A. G. 2011. Pesticide fate in the environment: a guide for field inspectors. Illinois state water survey, Institute of Natural Resource Sustainability, University of Illinois at Urbana-Champaign. Online. Available from: http://www.isws.uiuc.edu/pubdoc/CR/ISWSCR2011-07.pdf Accessed Aug. 12 2017. Gitsopoulos, T. K., Melidis, V. and Evgenidis, G. 2010. Response of maize (Zea mays L.) to post-emergence applications of topramezone. Crop Protec. 29:1091-1093.

210

González-Torralva, F., Rojano-Delgado, A. M., Luque de Castro, M., Mülleder, N. and 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.

Gower, S., Loux, M., Cardina, J. and Harrison, S. 2002. Effect of planting date, residual herbicide, and postemergence application timing on weed control and grain yield in glyphosate-tolerant corn (Zea mays). Weed Technol. 16:488–494. Gower, S. A., Loux, M. M., Cardina, J., Harrison, S. K., Sprankle, P. L., Probst, N. J., Bauman, T. T., Bugg, W., Curran, W. S., Currie, R. S., Harvey, R. G., Johnson, W. G., Kells, J. J., Owen, M. D. K., Regehr, D. L., Slack, C. H., Spaur, M., Sprague, C. L., VanGessel, M. and Young, B. G. 2003. Effect of postemergence glyphosate application timing on weed control and grain yield in glyphosate-resistant corn: results of a 2-yr multistate study. Weed Technol. 17:821-828. Green, J.M. and Castle, L.A. 2010. Transitioning from single to multiple herbicide-resistant crops. Pg. 67-92 in: Glyphosate Resistance in Crops and Weeds: History, Development, and Management. Nandula, V.K. Ed. John Wiley & Sons, Inc., Hoboken, New Jersey, USA. Green, J. M. and Owen, M. D. K. 2011. Herbicide-resistant crops: utilities and limitations for herbicide-resistant weed management. J. Agric. Food Chem. 59:5819–5829. Green, J. M. 2012. The benefits of herbicide-resistant crops. Pest Manag. Sci. 68:1323–1331. Grichar, W. J. and Minton, B. W. 2006. Supplementary weed control using soil-applied herbicides in glyphosate-resistant maize in Texas. Crop Prot. 25:1071–1074. Grossmann, K., and Ehrhardt, T. 2007. On the mechanism of action and selectivity of the corn herbicide topramezone: a new inhibitor of 4-hydroxyphenylpyruvate dioxygenase. Pest Manag. Sci. 63:429-439. Gulden, R. H., Sikkema, P. H., Hamill, A. S., Tardif, F. and Swanton, C. J. 2009. Conventional vs. glyphosate-resistant cropping systems in Ontario: weed Control, diversity, and yield. Weed Sci. 57:665–672. Halford, C., Hamill, A. S., Zhang, J. and Doucet, C. 2001. Critical period of weed control in no- till soybean (Glycine max) and corn (Zea mays). Weed Technol. 15:737–744. Hall, M. R., Swanton, C. J. and Anderson, G. W. 1992. The critical period of weed control in grain corn (Zea mays). Weed Sci. 40:441-447. Harrison, S. K., Regnier, E. E., Schmoll, J. T. and Webb, J. E. 2001. Competition and fecundity of giant ragweed in corn. Weed Sci. 49:224–229. Hartzler, R. G., Battles, B. A. and Nordby, D. 2004. Effect of common waterhemp (Amaranthus rudis) emergence date on growth and fecundity in soybean. Weed Sci. 52: 242-245. Hartzler, B., Boerboom, C., Nice, G. and Sikkema, P. 2006. Understanding glyphosate to increase performance. GWC-2 in the glyphosate, weeds and crops series, Purdue

211

Extension. Online. Available from: https://www.extension.purdue.edu/extmedia/gwc/gwc-2.pdf Accessed Dec. 18 2017. Hatzios, K. K. and Burgos, N. 2004. Metabolism-based herbicide resistance: regulation by safeners. Weed Sci. 52:454-467. Hausman, N. E., Singh, S., Tranel, P. J., Riechers, D. E., Kaundun, S. S., Polge, N. D., Thomas, D. A. and Hager, A. G. 2011. Resistance to HPPD-inhibiting herbicides in a population of waterhemp (Amaranthus tuberculatus) from Illinois, United States. Pest Manag. Sci. 67: 258-261. Hawkes, T. R. 2012. Hydroxyphenylpyruvate dioxygenase (HPPD): the herbicide target. Ch. 4.2 in W. Krämer, U. Schirmer, P. Jeschke, M. Witschel eds. Modern Crop Protection Compounds (2nd ed., Vol. 1). Weinheim, Germany: Wiley-VCH Verlag GmbH & Co. Weinheim, Germany. Hayden, T. A. and Burnside, O. C. 1987. Effect of herbicide incorporation methods on shattercane (Sorghum bicolor) control in corn (Zea mays). Weed Sci. 35:364–372. Health Canada. 2018. Tolpyralate 400SC Herbicide Reg. No. 32901. Consumer product safety database. Online. Available from: http://pr-rp.hc-sc.gc.ca/pi-ip/result- eng.php?1=0&2=396&3=pr&4=n&5=1&6=ASC&7=T&8=E Accessed Feb. 14 2018. Heap, I. The international survey of herbicide resistant weeds. Online. Available from: www.weedscience.org Accessed Nov. 22 2018.

Hedges, B. K. 2018. Influence of glyphosate-dicamba application rate and timing on the control of glyphosate-resistant Canada fleabane in glyphosate/dicamba-resistant soybean. Ch. 6 in Control of glyphosate-resistant Canada fleabane (Conyza canadensis) and waterhemp (Amaranthus tuberculatus var. rudis) in glyphosate/dicamba-resistant soybean (Glycine max). MSc. Thesis. University of Guelph. Hellwig, K. B., Johnson, W. G. and Massey, R. E. 2003. Weed management and economic returns in no-tillage herbicide-resistant corn (Zea mays). Weed Technol. 17:239–248. Herrmann, K. M. and Weaver, L. M. 1999. The shikimate pathway. Annu. Rev. Plant Physiol. Plant Mol. Biol. 50:473-503.

Hess, F. D. and Falk, R. H. 1990. Herbicide deposition on leaf surfaces. Weed Sci. 38:280-288. Hess, F. D. 2000. Light-dependent herbicides: an overview. Weed Sci. 48:160-170. Holt, J. S. and Lebaron, H. M. 1990. Significance and distribution of herbicide resistance. Weed Technol. 4:141-149.

Horak, M. J. and Loughin, T. M. 2000. Growth analysis of four Amaranthus species. Weed Sci. 48: 347-355.

212

Huang, H., Ye, R., Qi, M., Li, X., Miller, D. R., Stewart, C. N., DuBois, D. W. and Wang, J. 2015. Wind-mediated horseweed (Conyza canadensis) gene flow: pollen emission, dispersion and deposition. Ecol. Evol. 5:2646-2658.

Huffman, J. L., Riggins, C. W., Steckel, L. E. and Tranel, P. J. 2016. The EPSPS Pro106Ser substitution solely accounts for glyphosate resistance in a goosegrass (Eleusine indica) population from Tennessee, United States. J. Integr. Agr. 15:1304-1312.

Huggins, D. R. and Reganold, J. 2008. No-till: the quiet revolution. Scientific American. 299: 70-77. Hugie, J. A., Bollero, G. A., Tranel, P. J., and Riechers, D. E. 2008. Defining the rate requirements for synergism between mesotrione and atrazine in redroot pigweed (Amaranthus Retroflexus). Weed Sci. 56:265-270. Hull, H. M., Morton, H. L. and Wharrie, J. R. 1975. Environmental influences on cuticle development and resultant foliar penetration. Bot. Rev. 41:421-452. Hussain, Z., Marwat, K. B. and Cardina, J. 2011. Common cocklebur competition in forage maize. Weed Technol. 25:151–158. Hyun, T. K., Kumar, K., Rao, K. P., Sinha, A. K., and Roitsch, T. 2010. Role of α-tocopherol in cellular signaling: α-tocopherol inhibits stress-induced mitogen-activated protein kinase activation. Plant Biotechnol. Rep. 5:19-25. ISK Biosciences Corporation. 2018. Announcements. Online. Available from: http://www.iskbc.com/ Accessed Feb. 14 2018. Janak TW and Grichar WJ (2016) Weed control in corn (Zea mays L.) as influenced by preemergence herbicides. Int. J. Agron. 2016: 1-9.

Jasieniuk M., Ahmad R., Sherwood A. M., Firestone J. L., Perez-Jones A., Lanini W. T., Mallory-Smith, C. and Stednick, Z. 2008. Glyphosate-resistant Italian ryegrass (Lolium multiflorum) in California: distribution, response to glyphosate, and molecular evidence for an altered target enzyme. Weed Sci. 56:496–502.

Jeanmart, S., Edmunds, A. J., Lamberth, C. and Pouliot, M. 2016. Synthetic approaches to the 2010–2014 new agrochemicals. Bioorganic Med. Chem. 24:317-341. Johnson, B.C. and Young, B.G. 2002 Influence of temperature and relative humidity on the foliar activity of mesotrione. Weed Sci. 50: 157-161. Johnson, B. C., Young, B. G., and Matthews, J. L. 2002. Effect of postemergence application rate and timing of mesotrione on corn (Zea mays) response and weed Control. Weed Technol. 16:414-420. Johnson, W. G., Davis, V. M., Kruger, G. R. and Weller, S. C. 2009. Influence of glyphosate- resistant cropping systems on weed species shifts and glyphosate-resistant weed populations. Europ. J. Agronomy 31:162-172.

213

Johnson, W. G., Chahal, G. S. and Regehr, D. L. 2012. Efficacy of various corn herbicides applied preplant incorporated and preemergence. Weed Technol. 26:220–229. Johnson, B. 2014 (unpublished). Tolpyralate (SL-573) for Weed Control in RR/LL Corn. Purdue University Trial Summary. Online. Available from: http://www.purdueweedsci.com/research/sites/default/files/pdf/347-14-03.pdf Accessed Aug. 14 2017. Johnson, D. B. and Norsworthy, J. K. 2014. Johnsongrass (Sorghum halepense) management as influenced by herbicide selection and application timing. Weed Technol. 28:142-150. Jordan, D. L., York, A. C., Griffin, J. L., Clay, P. A., Vidrine, P. R. and Reynolds, D. B. 1997. Influence of application variables on efficacy of glyphosate. Weed Technol. 11:354-362. Jugulam, M., Niehues, K., Godar, A. S., Koo, D. H., Danilova, T., Friebe, B., Sehgal, S., Varanasi, V. K., Wiersma, A., Westra, P., Stahlman, P. W. and Gill, B. S. 2014. Tandem amplification of a chromosomal segment harboring 5-enolpyruvylshikimate-3-phosphate synthase locus confers glyphosate resistance in Kochia scoparia. Plant Physiol. 166:1200-1207.

Kakidani, H. and Hirai, K. 2003. Three-dimensional modeling of plant 4-hydroxyphenylpyruvate dioxygenase, a molecular target of triketone-type herbicides. J. Pestic. Sci. 28:409-415. Kegode, G. O. and Fronning, B. E. 2005. Artemisa biennis (biennial wormwood) control is influenced by plant size and weed flora at time of herbicide application. Crop Prot. 24:915-920. Kikugawa, H., Satake, Y., Tonks, D. J., Grove, M., Nagayama, S. and Tsukamoto, M. 2015. Tolpyralate: New Post-Emergence Herbicide for Weed Control in Corn. Weed Science Society of America Conference Proceedings 275. Kim, J., Jung, S., Hwang, I. T. and Cho, K. Y. 1999. Characteristics of chlorophyll a fluorescence induction in cucumber cotyledons treated with diuron, norflurazon, and sulcotrione. Pest. Biochem. and Physiol. 65:73-81. King, C. A. and Oliver, L. R. 1992. Application rate and timing of aciflourfen, bentazon, chlorimuron and imazaquin. Weed Technol. 6:526-534. Knezevic, S. Z., Weise, S. F., Swanton, C. J. 1994. Interference of redroot pigweed (Amaranthus retroflexus) in corn (Zea mays). Weed Sci. 42:568–573. Knezevic, S. Z., Evans, S. P., Blankenship, E. E.,Van Acker, R. E. and Lindquist, J. L. 2002. Critical period for weed control: the concept and data analysis. Weed Sci. 50:773–786. Kohrt, J. R. and Sprague, C. L. 2017a. Response of a multiple-resistant Palmer amaranth (Amaranthus palmeri) population to four HPPD-inhibiting herbicides applied alone and with atrazine. Weed Sci. 65:534-545.

214

Kohrt, J.R. and Sprague, C. L. 2017b. Herbicide management strategies in field corn for a three- way herbicide-resistant Palmer amaranth (Amaranthus palmeri) population. Weed Technol. 31:364-372. Krausz, R. F., Kapusta, G. and Matthews, J. L. 1996. Control of annual weeds with glyphosate. Weed Technol. 10:957-962. Kreiner, J.M., Stinchcombe, J.R. and Wright, S.I. 2018. Population genomics of herbicide resistance: adaptation via evolutionary rescue. Annu. Rev. Plant Biol. 69: 611-635. Kropff, M. J., and Spitters, C. J. 1991. A simple model of crop loss by weed competition from early observations on relative leaf area of the weeds. Weed Res. 31:97–105. Lal, R., Reicosky, D. and Hanson, J. 2007. Evolution of the plow over 10,000 years and the rationale for no-till farming. Soil Tillage Res. 93:1–12. Lee, S. D. and Oliver, L. R. 1982. Efficacy of acifluorfen on broadleaf weeds. Times and methods for application. Weed Sci. 30:520-526. Lee, D. L., Knudsen, C. G., Michaely, W. J., Chin, H., Nguyen, N. H., Carter, C. G., Cromartie, T. H., Lake, B. H., Shribbs, J. M. and Fraser, T. 1998. The structure-activity relationships of the triketone class of HPPD herbicides. Pestic. Sci. 54:377-384. Lindsey, L. E., Everman,W. J., Chomas, A. J. and Kells, J. J. 2012. Evaluation of application program and timing in herbicide-resistant corn. Weed Technol. 26:617–621. Liu, W. and Tollenaar, M. 2009. Physiological mechanisms underlying heterosis for shade tolerance in maize. Crop Sci. 49:1817. Liu, J. G., Mahoney, K. J., Sikkema, P. H. and Swanton, C. J. 2009. The importance of light quality in crop-weed competition. Weed Res. 49:217–224. Liu, J., Davis, A. S., and Tranel, P. J. 2012. Pollen biology and dispersal dynamics in waterhemp (Amaranthus tuberculatus). Weed Sci. 60: 416-422. Liu, M. and Lu, S. 2016. Plastoquinone and ubiquinone in plants: biosynthesis, physiological function and metabolic engineering. Front. Plant Sci.7:1898. Lorentz, L., Gaines, T. A., Nissen, S. J., Westra, P., Strek, H. J., Dehne, H. W., Ruiz-Santaella, J. P. and Beffa, R. 2014. Characterization of glyphosate resistance in Amaranthus tuberculatus populations. J. Agric. Food Chem. 62:8134-8142.

Lubbers, M. D., Stahlman, P. W. and Al-Khatib, K. 2007. Fluroxypyr efficacy is affected by relative humidity and soil moisture. Weed Sci. 55:260-263.

Mahoney, K. J., Shropshire, C. and Sikkema, P. H. 2017. Post-emergence herbicides for control of glyphosate-resistant Canada fleabane in corn. Can. J. Plant Sci. 97:206-209.

215

Main, C. L., Mueller, T. C., Hayes, R. M. and Wilkerson, J. B. 2004. Response of selected horseweed (Conyza canadensis (L.) Cronq.) populations to glyphosate. J. Agric. Food Chem. 52:879-883.

Main, C. L., Steckel, L. E., Hayes, R. M. and Mueller, T. C. 2006. Biotic and abiotic factors affecting horseweed emergence. Weed Sci. 54:1101-1105.

Mamy, L., Gabrielle, B. and Barriuso, E. 2010. Comparative environmental impacts of glyphosate and conventional herbicides when used with glyphosate-tolerant and non- tolerant crops. Environ. Pollut. 158:3173-3178.

Massinga, R. A., Currie, R. S. and Trooien, T. P. 2003. Water use and light interception under Palmer amaranth (Amaranthus palmeri) and corn competition. Weed Sci. 51:523–531. Matsumoto, H., Mizutani, M., Yamaguchi, T. and Kadotani, J. 2002. Herbicide pyrazolate causes cessation of carotenoids synthesis in early watergrass by inhibiting 4-hydroxyphenyl pyruvate dioxygenase. Weed Biol. Manag. 2:39-45. McMullan, P. M. and Green, J. M. 2011. Identification of a tall waterhemp (Amaranthus tuberculatus) biotype resistant to HPPD-inhibiting herbicides, atrazine, and thifensulfuron in Iowa. Weed Technol. 25: 514-518. Metzger, B. A., Soltani, N., Raeder, A. J., Hooker, D. C., Robinson, D. E., and Sikkema, P. H. 2018a. Tolpyralate Efficacy: Part I. Biologically-effective dose of tolpyralate for control of annual grass and broadleaf weeds in corn. Weed Technol. 32:698-706 Metzger, B. A., Soltani, N., Raeder, A. J., Hooker, D. C., Robinson, D. E., and Sikkema, P. H. 2018b. Tolpyralate Efficacy: Part II. Comparison of three group 27 herbicides applied postemergence for annual grass and broadleaf weed control in corn. Weed Technol. 32:707-713. Meyer, M. D., Pataky, J. K. and Williams, M. M. II. 2010. Genetic factors influencing the adverse effects of mesotrione and nicosulfuron on sweet corn yield. Agron J. 102:1138- 1144. Myers, M. W., Curran, W. S., VanGessel, M. J., Majek, B. A., Scott, B. A., Mortensen, D. A., Calvin, D. D., Karsten, H. D. and Roth, G. W. 2005. The effect of weed density and application timing on weed control and corn grain yield. Weed Technol. 19:102–107. Mitchell, G., Bartlett, D. W., Fraser, T. E., Hawkes, T. R., Holt, D. C., Townson, J. K. and Wichert, R. A. 2001. Mesotrione: a new selective herbicide for use in maize. Pest Manag. Sci. 57:120-128. Mitich LW (1991) Intriguing world of weeds: Velvetleaf. Weed Technol. 5:253-255. Monsanto Company. 2018. Dekalb® Corn Hybrid Products. DKC42-60RIB and DKC43-47RIB. Online. Available from: http://www.dekalb.ca/corn/hybrids Accessed Nov. 30 2018.

216

Mueller, T. C., Massey, J. H., Hayes, R. M., Main, C. L. and Stewart, C. N. Jr. 2003. Shikimate accumulates in both glyphosate-sensitive and glyphosate-resistant horseweed (Conyza canadensis L. Cronq.). J. Agric. Food Chem. 51:680-684.

Müller, F. and Appleby, A. P. 2012. Weed control, 1. Fundamentals. Pages 173-208 in Ullmann’s Encyclopedia of Industrial Chemistry (6th Ed. Vol. 39). Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim, Germany.

Murphy, S. D., Yakubu, Y. Weise, S. F., Swanton, C. J. 1996. Effect of planting patterns and inter-row cultivation on competition between corn (Zea mays) and late emerging weeds. Weed Sci. 44:865–870. Nandula, V. K., Eubank, T. W., Poston, D. H., Koger, C. H. and Reddy, K. N. 2006. Factors affecting germination of horseweed (Conyza canadensis). Weed Sci. 54:898-902.

Nandula, V. K., Ray, J. D., Ribeiro, D. N., Pan, Z. and Reddy, K. N. 2013. Glyphosate resistance in tall waterhemp (Amaranthus tuberculatus) from Mississippi is due to both altered target-site and nontarget-site mechanisms. Weed Sci. 61(3):374-383.

National Agricultural Statistics Service (NASS) 2015. 2014 Agricultural chemical use survey. NASS Highlights. United States Department of Agriculture 2015-1.

Nelson, G. C. and Bullock, D. S. 2003. Simulating a relative environmental effect of glyphosate- resistant soybeans. Ecol. Econ. 45(2):189-202.

Nordby, D., Hartzler, B. and Bradley, K. 2007. Biology and management of waterhemp. The glyphosate, weeds and crops series no. 13. Purdue Extension. Norsworthy, J. K., Oliver, L. R. and Purcell, L. C. 1999. Diurnal leaf movement effects on spray interception and glyphosate efficacy. Weed Technol. 13:466-470. Norsworthy, J. and Oliveira, M. 2004. Comparison of the critical period for weed control in wide- and narrow-row corn. Weed Sci. 52:802-807. Nurse, R. E., Swanton, C. J., Tardif, F. and Sikkema, P. H. 2006. Weed control and yield are improved when glyphosate is preceded by a residual herbicide in glyphosate-tolerant maize (Zea mays). Crop Prot. 25:1174–1179. Obrigawitch, T. T., W. H. Kenyon, and H. Kuratle. 1990. Effect of application timing on rhizome johnsongrass (Sorghum halepense) control with DPX-V9360. Weed Sci. 38:45- 49. Oliviera, M. C., Jhala, A. J., Gaines, T., Irmak, S., Amundsen, K., Scott, J. E. and Knezevic, S. Z. 2017. Confirmation and control of HPPD-inhibiting herbicide-resistant waterhemp (Amaranthus tuberculatus) in Nebraska. 31: 67-79. Olson, K. D. and Eidman, V. R. 1992. A farmer's choice of weed control method and the impacts of policy and risk. Rev. Agric. Econ. 14:125.

217

Ontario Ministry of Agriculture, Food and Rural Affairs [OMAFRA]. 2018a. Publication 75A. Guide to weed control. Field crops. Ed. Mike Cowbrough, OMAFRA. Guelph, ON. Ontario Ministry of Agriculture, Food and Rural Affairs [OMAFRA]. 2018b. Grain Corn. Pg 6 in Publication 60. 2018 Ontario Field Crop Budgets. Online. Available from: http://www.omafra.gov.on.ca/english/busdev/facts/pub60.pdf Accessed Nov. 25 2018. O’Sullivan, J., Zandstra, J. and Sikkema, P. H. 2002. Sweet corn (Zea mays) cultivar sensitivity to mesotrione. Weed Technol. 16:421-425. Owen, M. Hartzler, R. G., Franzenburg, D. D., Lee, J. M. and Eeling, J. S. 2014 (unpublished). Weed science program 2014 weed control results. Iowa State University, Dept. of Agronomy, Ames IA, USA. Online. Available from: https://store.extension.iastate.edu/product/Weed-Science-Program-2014-Weed-Control- Results Accessed Aug. 12 2017. Page, E. R., Tollenaar, M. Lee, E. A., Lukens, L. and Swanton, C. J. 2009. Does the shade avoidance response contribute to the critical period for weed control in maize (Zea mays)? Weed Res. 49:563–571. Page, E. R., Cerrudo, D., Westra, P., Loux, M., Smith, K., Foresman, C., Wright, H. and Swanton, C. J. 2012. Why early season weed control is important in maize. Weed Sci. 60:423-430. Page, E. R., Grainger, C. M., Laforest, M., Nurse, R. E., Rajcan, I., Bae, J. and 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. Pallett, K. E., Little, J. P., Sheekey, M. and Veerasekaran, P. 1998. The mode of action of isoxaflutole: I. Physiological effects, metabolism, and selectivity. Pestic. Biochem. Physiol. 62:113-124. Pallett, K. E., Cramp, S. M., Little, J. P., Veerasekaran, P., Crudace, A. J., and Slater, A. E. 2001. Isoxaflutole: the background to its discovery and the basis of its herbicidal properties. Pest Manag. Sci. 57:133-142. Powles, S. B., Lorraine-Colwill, D. F., Dellow, J. J. and Preston, C. 1998. Evolved resistance to glyphosate in rigid ryegrass (Lolium rigidum) in Australia. Weed Sci. 46:604-607.

Rahman, A., Trolove, M. R. and James, T. K. 2013. Efficacy and crop selectivity of topramezone for post-emergence weed control in maize. Pages 470-476 in Asian-Pacific Weed Science Society Conference Proceedings. Online. Available from: https://apwss.org/documents/24th%20APWSS%20ConferenceProceeding.pdf Accessed Dec. 11 2018.

Ramsey, R. J. L., Stephenson, G. R. and Hall, J. C. 2002. Effect of relative humidity on the uptake, translocation, and efficacy of glufosinate ammonium in wild oat. Pest Biochem Physiol. 73:1-8.

218

Rajcan, I. and Swanton, C. J. 2001. Understanding maize–weed competition: resource competition, light quality and the whole plant. Field Crops Res. 71:139–150. Rajcan, I., Chandler, K. J. and Swanton, C. J. 2004. Red–far-red ratio of reflected light: a hypothesis of why early-season weed control is important in corn. Weed Sci. 52:774– 778. Reddy, K. N. and Norsworthy, J. K. 2010. Glyphosate-resistant crop production systems: impact on weed species shifts. Pages 165-184 in Nandula V. K. ed. Glyphosate resistance in crops and weeds: history, development and management. John Wiley & Sons Inc., Hoboken, NJ.

Regehr, D. L. and Bazzaz, F. A. 1979. The population dynamics of Erigeron canadensis, a successional winter annual. J. Ecol. 67:923-933.

Reid, A., Gonzalez, V., Sikkema, P. H., Lee, E. A., Lukens, L. and Swanton, C. J. 2014. Delaying weed control lengthens the anthesis-silking interval in maize. Weed Sci. 62:326–337. Riar, D. S., Norsworthy, J. K., Johnson, D. B., Scott, R. C. and Bagavathiannan, M. 2011. Glyphosate resistance in a johnsongrass (Sorghum halepense) biotype from Arkansas. Weed Sci. 59:299-304.

Riechers, D. E., Kreuz, K. and Zhang, Q. 2010. Detoxification without intoxication: herbicide safeners activate plant defense gene expression. Plant Physiol. 153:3-13. Ritter, R. L. and Coble, H. D. 1981. Influence of temperature and relative humidity on the activity of aciflourfen. Weed Sci. 29:480-485. Rojano-Delgado, A. M., Menendez, J. and De Prado, R. 2014. Absorption and penetration of herbicides viewed in metabolism studies: case of glufosinate and imazamox in wheat. Pages 159-165 in K. Myung, N. M. Satchivi, C. K. Kingston eds. Retention, uptake and translocation of agrochemicals in plants. American Chemical Society, Washington, DC, USA. Salas, R. A., Scott, R. C., Dayan, F. E. and Burgos, N. R. 2015. EPSPS gene amplification in glyphosate-resistant Italian Ryegrass (Lolium perenne spp. multiflorum) populations from Arkansas (United States). J. Agric. Food Chem. 63:5885-5893.

Sammons, R. D. and Gaines, T. A. 2014. Glyphosate resistance: state of knowledge. Pest Manag. Sci. 70:1367-1377.

Sanyal, D., Bhowmik, P. C., Anderson, R. L., Shrestha, A. 2008. Revisiting the perspective and progress of integrated weed management. Weed Sci. 56:161–167. Sarangi, D., Tyre, A. J., Patterson, E. L., Gaines, T. A., Irmak, S., Knezevic, S. Z., Lindquist, J. L. and Jhala, A. J. 2017. Pollen-mediated gene flow from glyphosate-resistant common

219

waterhemp (Amaranthus rudis Sauer): consequences for the dispersal of resistance genes. Sci. Rep. 7: 44913. Sargent, J. A. 1965. The penetration of growth regulators into leaves. Annu. Rev. Plant. Physiol. 16:1-12. Sargent, J. A. and Blackman, G. E. 1972. Studies of foliar penetration: IX. Patterns of penetration of 2,4-dichlorophenoxyacetic acid into the leaves of different species. J. Exp. Bot. 23: 830-841. Schabenburger, O. 2007. Growing up fast: SAS® 9.2 enhancements to the GLIMMIX procedure. SAS Global Forum paper 177. SAS Institute Inc., Cary, NC. Schryver M. G., Soltani, N., Hooker, D. C., Tranel, P. J., Robinson, D. E. and Sikkema, P. H. 2017a. Glyphosate-resistant waterhemp (Amranthus tuberculatus var. rudis) in Ontario, Canada. Can. J. Plant Sci. 97(6):1057-1067. Schryver M.G., Soltani, N., Hooker, D.C., Robinson, D.E., Tranel, P.J. and 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., Tranel, P. J., Robinson, D. E. and Sikkema, P. H. 2018. Glyphosate resistant waterhemp (Amaranthus tuberculatus var. rudis) control in soybean (Glycine max) using one and two-pass weed control strategies in Ontario. Weed Technol. (In Press). Schulz, A., Ort, O., Beyer, P. and Kleinig, H. 1993. SC-0051, a 2-benzoyl-cyclohexane-1,3- dione bleaching herbicide, is a potent inhibitor of the enzyme p-hydroxyphenyl pyruvate dioxygenase. FEBS Letters 318:162–166. Scott, B. A. and VanGessel, M. J. 2007. Delaware Soybean Grower Survey on Glyphosate- Resistant Horseweed (Conyza canadensis). Weed Technol. 21:270-274.

Sellers, B. A., Ferrell, J. A., MacDonald, G. E. and Kline, W. N. 2009. Dogfennel (Eupatorium capillifolium) size at application affects herbicide efficacy. Weed Technol. 23:247-250. Shah, D. M., Horsch, R. B., Klee, H. J., Kishore, G. M., Winter, J. A., Tumer, N. E., Hironaka, C. M., Sanders, P. R., Gasser, C. S., Aykent, S., Siegel, N. R., Rogers, S. G. and Fraley, R. T. 1986. Engineering herbicide tolerance in transgenic plants. Sci. 233:478-481.

Shaner, D. L. 2014. Lessons learned from the history of herbicide resistance. Weed Sci. 62:427- 431.

Shields, E. J., Dauer, J. T., VanGessel, M. J. and Neumann, G. 2006. Horseweed (Conyza canadensis) seed collected in the planetary boundary layer. Weed Sci. 54:1063-1067.

Siddall, T., Ouse, D., Benko, Z., Garvin, G., Jackson, J., McQuiston, J., Ricks, M., Thibault, T., Turner, J., VanHeertum, J. and Weimer, M. 2002. Synthesis and herbicidal activity of phenyl-substituted benzoylpyrazoles. Pest Manag. Sci. 58:1175-1186.

220

Skuterud, R., Bjugstad, N., Tyldum, A. and Tørresen, K. S. 1998. Effect of herbicides applied at different times of the day. Crop Protec. 17:41-46. Snyder, E. M., Curran, W. S., Karsten, H. D., Malcolm, G. M., Duiker, S. W., Hyde, J. A. 2016. Assessment of an integrated weed management system in no-till soybean and corn. Weed Sci. 64:712–726. Soil Conservation Council of Canada. 2017. Reduced tillage helps reduce carbon dioxide levels. Online. Available from: http://www.soilcc.ca/ggmp_feature_articles/2004/2004-02.php Accessed March 12 2017. Soltani, N., Sikkema, P. H., Zandstra, J., O’Sullivan, J. and Robinson, D. E. 2007. Response of eight sweet corn (Zea mays L.) hybrids to topramezone. Hort Sci. 42:110-112. Soltani, N., Vyn, J. D. and Sikkema, P. H. 2009. Control of common waterhemp (Amaranthus tuberculatus var. rudis) in corn and soybean with sequential herbicide applications. Can. J. Plant Sci. 89: 127-132. Soltani, N., Dille, J. A., Burke, I. C., Everman, W. J., VanGessel, M. J., Davis, V. M., Sikkema, P. H. 2016. Potential corn yield losses from weeds in North America. Weed Technol. 30:979–984. Soltani, N., Nurse, R. E., and Sikkema, P. H. 2016. Biologically effective dose of glyphosate as influenced by weed size in corn. Can. J. Plant Sci. 96:455-460. Soane, B., Ball, B., Arvidsson, J., Basch, G., Moreno, F., Roger-Estrade, J. 2012. No-till in Northern, Western and South-Western Europe: a review of problems and opportunities for crop production and the environment. Soil and Tillage Res. 118:66–87. Sprague, C. L., Penner, D. and Kells, J. J. 1999. Enhancing the margin of selectivity of RPA 201772 in Zea mays with antidotes. Weed Sci. 47:492-497. Sprague, C. and Powell, G. 2014 (unpublished). Evaluation of tolpyralate (SL-573) for weed control in corn. Michigan State University Weed Science Research Program Trial Summary. Online. Available from: https://www.canr.msu.edu/weeds/research/annual- results/2014-annual-results/C05-14%20TOLPYRALATE.pdf Accessed Dec. 11 2018. Sprankle, P., Meggitt, W. F. and Penner, D. 1975. Rapid inactivation of glyphosate in the soil. Weed Sci. 23:224-228.

Steckel, G. J., Wax, L. M., Simmons, W. and Phillips, W. H. II. 1997. Glufosinate efficacy on annual weeds is influenced by rate and growth stage. Weed Technol. 11:484-488. Steckel, L. E. and Sprague, C. L. 2004. Common waterhemp (Amaranthus rudis) interference in corn. Weed Sci. 52(3): 359-364.

Steckel, L. E., Main, C. L. and Mueller, T. C. 2010. Glyphosate-resistant horseweed in the United States. Pages 185-193 in Glyphosate resistance in crops and weeds. History, development and management. John Wiley & Sons Inc. Hoboken, NJ, USA.

221

Stephenson, D. O., and Bond, J. A. 2012. Evaluation of thiencarbazone-methyl– and isoxaflutole-based herbicide programs in corn. Weed Technol. 26:37-42. Stewart, C. L., Nurse, R. E. and Sikkema, P. H. 2009. Time of day impacts postemergence weed control in corn. Weed Technol. 23:346-355.

Stewart, C. L., Nurse, R. E., Hamill, A. S. and Sikkema, P. H. 2010. Environment and soil conditions influence pre- and postemergence herbicide efficacy in soybean. Weed Technol. 24:234-243.

Stewart, C. L., Soltani, N., Nurse, R. E., Hamill, A. S. and Sikkema, P. H. 2012. Precipitation influences pre- and post-emergence herbicide efficacy in corn. Am. J Plant Sci. 03:1193– 1204. Stoller, E. W., Wax, L. M. and Matthiesen, R. L. 1975. Response of yellow nutsedge and soybeans to bentazon, glyphosate and perfluidone. Weed Sci. 23:215-221. Stopps, G. J., Nurse, R. E. and Sikkema, P. H. 2013. The effect of time of day on the activity of postemergence soybean herbicides. Weed Technol. 27:690-695. Swanton, C. J. and Weise, S. F. 1991. Integrated weed management: the rationale and approach. Weed Technol. 5:657–663. Swanton, C. J. Harker, K. N. and Anderson, R. L. 1993. Crop losses due to weeds in Canada. Weed Technol. 7:537–542. Swanton, C. J., Weaver, S., Cowan, P.,Van Acker, R., Deen, W. and Sheshta, A. 1999. Weed thresholds: theory and applicability. J. Crop Prod. 2:9–29. Swanton, C. J., Gulden, R. H. and Chandler, K. 2007. A rationale for atrazine stewardship in corn. Weed Sci. 55:75-81. Teasdale, J. R. 1995. Influence of narrow row/high population corn (Zea mays) on weed control and light transmittance. Weed Technol. 9:113–118. Tharp, B. E. and Kells, J. J. 1999. Influence of herbicide application rate, timing and interrrow cultivation on weed control and corn (Zea mays) yield in glufosinate-resistant and glyphosate-resistant corn. Weed Technol. 13:807-813. Tharp, B. E. and Kells, J. J. 2002. Residual herbicides used in combination with glyphosate and glufosinate in corn (Zea mays) Weed Technol. 16:274–281. Timmons, F. L. 1970. A history of weed control in the United States and Canada. Weed Sci. 18:294–307. Tonks, D., Grove, M., Kikugawa, H., Parks, M., Nagayama, S. and Tsukamoto, M. 2015. Tolpyralate: An Overview of Performance for Weed Control in US Corn. Weed Science Society of America Conference Proceedings 276.

222

Travlos, I. S., Economou, G. and Kanatas, P. J. 2011. Corn and barnyardgrass competition as influenced by relative time of weed emergence and corn hybrid. Agron. J.102:1-6. Tsegaye, Y., Shintani, D. K. and Dellapenna, D. 2002. Overexpression of the enzyme p- hydroxyphenolpyruvate dioxygenase in Arabidopsis and its relation to tocopherol biosynthesis. Plant Physiol. and Biochem. 40:913-920. Tursun, N., Datta, A., Sakinmaz, M. S., Kantarci, Z., Knezevic, S. Z., Chauhan, B. S. 2016. The critical period for weed control in three corn (Zea mays L.) types. Crop Prot. 90:59–65. United States Environmental Protection Agency (USEPA). 2001. Conditional registration for mesotrione ([2-[4-(methylsulfonyl)-2-nitrobenzoyl]-1,3-cyclohexanedione). Online. Available from: https://www3.epa.gov/pesticides/chem_search/reg_actions/registration/fs_PC- 122990_04-Jun-01.pdf Accessed Aug. 16 2017. University of Illinois 2014 (unpublished). Weed Control Systems with Tolpyralate in RR/LL Corn. US 128/14/01 002 01: UCISKSYS* AS-400. Online. Available from: http://extension.cropsciences.illinois.edu/fieldcrops/weeds/field_reports/2014/Ucorn14/U C11.pdf Accessed Aug. 17 2017. Uri, N. D. 2000. Perceptions on the use of no-till farming in production agriculture in the United States: an analysis of survey results. Agric. Ecosyst. Environ. 77:263–266. Van Acker, R. C., Swanton, C. J. and Weise, S. F. 1993. The critical period of weed control in soybean [Glycine max (L.) Merr.]. Weed Sci. 41:194–200. VanGessel, M. J. 2001. Glyphosate-resistant horseweed from Delaware. Weed Sci. 49:703-705.

Vila-Aiub, M. M., Balbi, M. C., Distéfano, A. J., Fernández, L., Hopp, E., Yu, Q. and Powles, S. B. 2011. Glyphosate resistance in perennial Sorghum halapense (Johnsongrass), endowed by reduced glyphosate translocation and leaf uptake. Pest Manag. Sci. 68:430-436.

Viviani, F., Little, J. and Pallett, K. 1998. The mode of action of isoxaflutole II. Characterization of the inhibition of carrot 4-hydroxyphenylpyruvate dioxygenase by the diketonitrile derivative of isoxaflutole. Pestic. Biochem. Physiol. 62:125-134. Vollmer, K. M., VanGessel, M. J., Johnson, Q. R. and Scott, B. A. 2017. Relative safety of preemergence corn herbicides applied to coarse-textured soil. Weed Technol. 31:356- 363. Vyn, J. D., Swanton, C. J., Weaver, S. E. and Sikkema, P. H. 2006. Control of Amaranthus tuberculatus var. rudis (common waterhemp) with pre and post-emergence herbicides in Zea mays L. (maize). Crop Prot. 25: 1051-1056.

Vyn, J. D., Swanton, C. J., Weaver, S. E. and 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

223

Wakelin, A. M. and Preston, C. 2006. A target-site mutation is present in a glyphosate-resistant Lolium rigidum population. Weed Res. 46:432-440.

Walsh, M. J., Stratford, K., Stone, K., and Powles, S. B. 2012. Synergistic effects of atrazine and mesotrione on susceptible and resistant wild radish (Raphanus raphanistrum) populations and the potential for overcoming resistance to triazine herbicides. Weed Technol. 26:341- 347. Walsh, K. D., Soltani, N., Brown, L. R. and Sikkema, P. H. 2014. Weed control with postemergence glyphosate tank mixes in glyphosate-resistant soybean. Can. J. Plant Sci. 94: 1239-1244. Wang, J., Qi, M., Huang, H., Ye, R., Li, X. and Stewart, C. N. 2017. Atmospheric pollen dispersion from herbicide-resistant horseweed (Conyza canadensis L.) Aerobiologia 33:393-406.

Weaver, S. and Tan, C. 1983. Critical period of weed interference in transplanted tomatoes (Lycopersicon esculentum): growth analysis. Weed Sci. 31:476-481. Weaver, S. 1991. Size-dependent economic thresholds for three broadleaf weed species in soybeans. Weed Technol. 5:674–679. Weaver, S. E. 2001. The biology of Canadian weeds. 115. Conyza canadensis. Can. J. Plant Sci. 81:867-875.

Weed Science Society of America (WSSA). 1998. Herbicide resistance and herbicide tolerance definitions. Weed Technol. 12:789.

Weed Science Society of America. 2018. Composite list of weeds. Online. Available from: http://wssa.net/wssa/weed/composite-list-of-weeds/ Accessed: Oct. 22 2018.

Wiersma, A. T., Gaines, T. A., Preston, C., Hamilton, J. P., Giacomini, D., Buell, C. R., Leach, J. E. and Westra, P. 2015. Gene amplification of 5-enol-pyruvylshikimate-3-phosphate synthase in glyphosate –resistant Kochia scoparia. Planta 241:463-474.

Williams, M. II. 2006. Planting date influences critical period of weed control in sweet corn. Weed Sci. 54:928–933.

Williams, M. M. II., and Pataky, J. K. 2010. Factors affecting differential sensitivity of sweet corn to HPPD-inhibiting herbicides. Weed Sci. 58:289-294. Williams, M. M., Boydston, R. A., Peachey, R. E., and Robinson, D. 2011. Significance of atrazine as a tank-mix partner with tembotrione. Weed Technol. 25:299-302. Witschel, M. 2009. Design, synthesis and herbicidal activity of new iron chelating motifs for HPPD-inhibitors. Bioorganic Medic. Chem. 17: 4221-4229.

224

Woodyard, A. J., Bollero, G. A., and Riechers, D. E. 2009. Broadleaf weed management in corn utilizing synergistic postemergence herbicide combinations. Weed Technol. 23:513-518. Wu, C. and Owen, M. D. K. 2014. When is the best time to emerge: reproductive phenology and success of natural common waterhemp (Amaranthus rudis) cohorts in the Midwest United States? Weed Sci. 62: 107-117. Young, F. L., Wyse, D. L. and Jones, R. J. 1984. Quackgrass (Agropyron repens) interference on corn (Zea mays). Weed Sci. 32:226–234. Yuan, J. S., Tranel, P. J. and Stewart, C. N. Jr. 2006. Non-target-site herbicide resistance: a family business. Trends Plant Sci. 12:1360-1385

Zhang, J., Jaeck, O., Menegat, A., Zhang, Z., Gerhards, R. and Hanwen, N. 2013. The mechanism of methylated seed oil on enhancing biological efficacy of topramezone on weeds. PLOS One 8: 1-9.

Zimdahl, R. L. 2004. The effect of competition duration. Ch. 6 in Weed-Crop Competition - A Review. 2nd ed. Blackwell Publishing. Ames, IA, USA.

225

Chapter 10: Appendix

10.1 – SAS Code for Chapter 2 Non-linear Regression Analysis title 'BER of Tolpyralate'; data (*insert species*); input plot trt rep dose env cont7 cont14 cont28 cont56 dens drywt; *parameter=cont7; *parameter=cont14; *parameter=cont28; *parameter=cont56; if trt=2 then delete; if trt>8 then delete; /*if trt=3 then delete; if trt=4 then delete; if trt=5 then delete; if trt=6 then delete; if trt=7 then delete; if trt=8 then delete; if trt=15 then delete; if trt=16 then delete;*/ datalines;

*insert data* ; **/ Exponential to a Maximum; **Where a=upper asymptote, b= magnitude and c=slope (rate constant); title 'Exponential to a Maximum'; proc nlin method=marquardt; parms a=100 b=100 c=0.01; if dose=0 then model parameter=0; model parameter=a-b*(exp(-c*dose)); output out=b predicted=cp; run;

*Proc print data=b; *Run; proc corr; var parameter cp; run;

*Plot data*; proc sort data=b; by dose; run; proc sgplot data=b; scatter x=dose y=parameter; series x=dose y=cp;

226 run;

**/Exponential to a Maximum Alternate; **Where a=upper asymptote, b=slope (rate constant) and c=magnitude; title 'Exponential to a Maximum Alternate'; proc nlin method=marquardt; parms a=100 b=0.001 c=100; if dose=0 then model parameter=0; model parameter=a-c*(b**dose); output out=b predicted=dp; run;

Proc print data=b; Run; proc corr; var parameter dp; run;

*Plot data*; proc sort data=b; by dose; run; proc sgplot data=b; scatter x=dose y=parameter; series x=dose y=dp; run;

**/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 sort data=(*insert species*); by dose env rep; proc plot; plot parameter*dose; run; proc nlin method=marquardt; parms a=0 b=150 c=0.01; *bounds a<=0; bounds a>=0; model parameter=a+b*exp(-c*dose); output out=second predicted=ep; run;

Proc print data=second; Run; proc corr; var parameter ep;

227 run;

*Plot data*; proc sort data=second; by dose; run; proc sgplot data=second; scatter x=dose y=parameter; series x=dose y=ep; run;

10.2 – SAS Code for Chapter 3 GLIMMIX Analysis

title 'BER of Tolpyralate'; data (*insert species*); input plot trt rep dose env cont7 cont14 cont28 cont56 dens drywt; */ If data is percent change to decimal; /*cont7=cont7/100; cont14=cont14/100; cont28=cont28/100; cont56=cont56/100;*/ /*Replace 0 values;*/ /*if cont7=0 then cont7=0.0001; if cont14=0 then cont14=0.0001; if cont28=0 then cont28=0.0001; if cont56=0 then cont56=0.0001; if dens=0 then dens=0.0001; if drywt=0 then drywt=0.0001;

/*control data*/ if trt in (1,2,3,4,5,7,8,9,10,11,13,14) then delete; *parameter=cont7; *parameter=cont14; *parameter=cont28; *parameter=cont56; *parameter=dens; *parameter=drywt; datalines;

*insert data* ; proc sort data=(*insert species*); by env trt rep; run; proc means data=(*insert species*); class trt; var parameter; run;

*Arcsine square root trans; /*title2 'parameter (arcsine square root transformation)'; analvar1=parameter; 228 if analvar1=100 then analvar1=100-0.01; if analvar1=0 then analvar1=0+0.01; analvar2=analvar1/100; analvar=arsin(sqrt(analvar2)); */ proc glimmix data=(*insert species*); class env trt rep; model parameter=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 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;

*/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;

*/Back-transformation for lognormal distribution; /*data btdata; set mmm; omega=exp(stderr*stderr); btlsmean=exp(estimate)*sqrt(omega); btvar=exp(2*estimate)*omega*(omega-1); btse_mean=sqrt(btvar); run; proc print; run;*/

229

10.3 – SAS Code for Chapter 4 Two-Factor GLIMMIX Analysis title 'Tolpyralate Rate by Size'; data first; input plot trt rate size rep env cont14 cont28 cont56 dens drywt phyto7 phyto14 phyto28 yield; */ If data is percent change to decimal; /*cont14=cont14/100; cont28=cont28/100; cont56=cont56/100;*/ /*Replace 0 values;*/ /*if cont14=0 then cont14=0.0001; if cont28=0 then cont28=0.0001; if cont56=0 then cont56=0.0001; if dens=0 then dens=0.0001; if drywt=0 then drywt=0.0001; if pdens=0 then pdens=0.0001; if pdrywt=0 then pdrywt=0.0001;*/ /*control data*/ if trt in (2,7,12,17) then delete; *parameter=cont14; *parameter=cont28; *parameter=cont56; *parameter=dens; *parameter=drywt; *parameter=moist; *parameter=yield; datalines;

*insert data* ; proc sort data=first; by env rate size rep; run; proc means data=first; class rate size; var parameter; run;

/*Rate|Size Factorial Analysis*/ proc glimmix data=first /*method=laplace*/; class env rate size rep; model parameter = rate|size / dist= link= ; random env env*size env*rate rep(env); covtest 'envvar=0' 0 . . . ; covtest 'env*sizevar=0' . 0 . . ; covtest 'env*ratevar=0' . . 0 . ; covtest 'repvar=0' . . . 0 ; lsmeans rate|size /pdiff slicediff=rate slicediff=size adjust=tukey ilink; ods html exclude lsmeans diffs; ods output lsmeans=mmm diffs=ppp; output out=second predicted=pred residual=resid residual(noblup)=mresid student=studentresid student(noblup)=smresid;

230 run; %include 'C:\Users\bmetzger\Documents\My SAS Files\9.4\pdmix800.sas'; %pdmix800(ppp,mmm,alpha=0.05,sort=no); run;

/*data mmm; set mmm; btmean=((sin(estimate))**2)*100; se_btmean=(2*(cos(estimate))*stderr)*100; proc print; run;*/ proc sort data=second; by studentresid; proc print; run;

*/Linearity of fixed effects - both a scatter & boxplot; proc sgplot data=second; scatter y=studentresid x=rate; refline 0; proc sgplot data=second; vbox studentresid / group=rate datalabel; run;

*/Linearity of fixed effects - both a scatter & boxplot; proc sgplot data=second; scatter y=studentresid x=size; refline 0; proc sgplot data=second; vbox studentresid / group=size datalabel; run;

*/Homogeneity of effects; proc sgscatter data=second; plot studentresid*(pred env rate size rep); run;

*/Q-Q plot of normal distribution;

*/Conditional: Normal distribution, normality plot and Shapiro-Wilk test; proc univariate data=second normal plot; var studentresid; histogram studentresid/normal kernel; run;

231

10.4 – SAS Code for Chapter 5 and 6 GLIMMIX Analysis and Non-linear Regression title 'Tolpyralate for Control of Canada Fleabane/Waterhemp'; data first; input plot trt rep dose env cont14 cont28 cont56 cont84 dens drywt pdens pdrywt phyto7 phyto14 phyto28; */ If data is percent change to decimal; /*cont14=cont14/100; cont28=cont28/100; cont56=cont56/100; cont84=cont84/100; /*Replace 0 values;*/ if cont14=0 then cont14=0.0001; if cont28=0 then cont28=0.0001; if cont56=0 then cont56=0.0001; if dens=0 then dens=0.0001; if drywt=0 then drywt=0.0001; *if trt=1 then delete; if trt=2 then delete; if trt>8 then delete; /*if trt=3 then delete; if trt=4 then delete; if trt=5 then delete; if trt=6 then delete; if trt=7 then delete; if trt=8 then delete; if trt=15 then delete; if trt=16 then delete;*/ /*for glimmix*/ /*if trt in (2,3,4,5,7,8,9,10,11,13,14) then delete;*/ *parameter=cont14; *parameter=cont28; *parameter=cont56; *parameter=cont84; *parameter=dens; *parameter=drywt; *parameter=pdens; *parameter=pdrywt; *parameter=phyto7; *parameter=phyto14; *parameter=phyto28; datalines;

*insert data* ;

**/Log-logistic Regression; **Where d=upper/lower asymptote, c=lower/upper asymptote i50=ED50, b=slope; title 'Log-logistic Model'; proc sort data=first; by dose env rep; proc plot; plot parameter*dose; run;

232 proc nlin data=first; parameters d=100 c=0 i50=8 b=2 ; bounds c>=0; *bounds d<=100; IF dose=0 THEN model parameter=c; ELSE model parameter=c+(d-c)/(1+EXP(-b*(log(dose)-log(i50)))); output out=second predicted=ep; run; Proc print data=second; Run; proc corr; var parameter ep; run;

*Plot data*; proc sort data=second; by dose; run; proc sgplot data=second; scatter x=dose y=parameter; series x=dose y=ep; run;

**/Inverse Exponential for ERICA dens/drywt; **Where a=lower asymptote, b=reduction in y from intercept to asymptote and c=slope (from intercept to a); title 'Inverse Exponential'; proc sort data=first; by dose env rep; proc plot; plot parameter*dose; run; proc nlin method=marquardt; parms a=0 b=150 c=0.01; *bounds a<=0; bounds a>=0; model parameter=a+b*exp(-c*dose); output out=second predicted=ep; run;

Proc print data=second; Run; proc corr; var parameter ep; run;

233

*Plot data*; proc sort data=second; by dose; run; proc sgplot data=second; scatter x=dose y=parameter; series x=dose y=ep; run;

/*GLIMMIX Analysis of Means*/ proc sort data=first; by env trt rep; run; proc means data=first; class trt; var parameter; run; proc glimmix data=first /*nobound method=laplace*/; class env trt rep; model parameter=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 lsmeans=mmm ; output out=second predicted=pred residual=resid residual(noblup)=mresid student=studentresid student(noblup)=smresid; *ods html exclude lsmeans diffs; run; %include 'C:\Users\bmetzger\Documents\My SAS Files\9.4\pdmix800.sas'; %pdmix800(ppp,mmm,alpha=0.05,sort=no); 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;

234

histogram studentresid/normal kernel; run;

10.5 – SAS Code for Chapter 7 Split-Split Block GLIMMIX and Mixed-Model Multiple

Stepwise Regression title 'Tolpyralate Tolerance'; data first; input plot rep trt rate hybrid$ timing$ env cc sand silt clay om ph cec temp24premax temp24prerange temp24pomax temp24posrange temp rh precip7D TOD phyto7 phyto14 phyto28 phyto56 height pheight stand moist pmoist yield pyield;

*/ If data is percent change to decimal; /*phyto7=phyto7/100; phyto14=phyto14/100; phyto28=phyto28/100; phyto56=phyto56/100;*/ /*Replace 0 & 1 values;*/ /*if phyto7=0 then phyto7=0.0001; if phyto14=0 then phyto14=0.0001; if phyto28=0 then phyto28=0.0001; if phyto56=0 then phyto56=0.0001;*/ /*response parameter*/ parameter=phyto7; *parameter=phyto14; *parameter=phyto28; *parameter=phyto56; *parameter=height; *parameter=pheight; *parameter=stand; *parameter=moist; *parameter=pmoist; *parameter=yield; *parameter=pyield; *parameter=parameter+1; *if rate=0 then delete; /*for stepwise regression* *if trt in (1,2,3,4,5,6,7,8,9,10,11,12) then delete;*/ datalines;

235

*insert data*; proc sort data=first; by rate hybrid timing rep; run; proc means data=first; class rate hybrid timing; var parameter; run;

/*Split-Split Block 3-Way Factorial GLIMMIX*/ proc glimmix data=first /*nobound method=laplace*/; class env rate hybrid timing rep; model parameter= rate|hybrid|timing /dist=normal link=id; random rep(env) rep*hybrid env rate*env hybrid*env timing*env; covtest 'rep=0' 0 . . . . . ; covtest 'rep*hybrid=0' . 0 . . . . ; covtest 'env=0' . . 0 . . . ; covtest 'rate*env=0' . . . 0 . . ; covtest 'hybrid*env=0' . . . . 0 . ; covtest 'timing*env=0' . . . . . 0 ; lsmeans rate|hybrid|timing /pdiff slicediff=rate slicediff=hybrid slicediff=timing adjust=tukey ilink; *ods html exclude lsmeans diffs; ods output lsmeans=mmm diffs=ppp; output out=second predicted=pred residual=resid residual(noblup)=mresid student=studentresid student(noblup)=smresid; run; %include 'C:\Users\bmetzger\Documents\My SAS Files\9.4\pdmix800.sas'; %pdmix800(ppp,mmm,alpha=0.05,sort=no); run; proc sort data=second; by studentresid; proc print; run;*/

*/Linearity of fixed effects - both a scatter & boxplot; proc sgplot data=second; scatter y=studentresid x=rate; refline 0; proc sgplot data=second; vbox studentresid / group=rate datalabel; run;

*/Linearity of fixed effects - both a scatter & boxplot; proc sgplot data=second; scatter y=studentresid x=hybrid; refline 0; proc sgplot data=second; vbox studentresid / group=hybrid datalabel; run;

*/Linearity of fixed effects - both a scatter & boxplot; proc sgplot data=second; scatter y=studentresid x=timing; refline 0; proc sgplot data=second; vbox studentresid / group=timing datalabel; run;

*/Homogeneity of effects; proc sgscatter; plot studentresid*(pred env rate hybrid timing rep);

236 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;

*/Back-transformation for lognormal distribution; data btdata; set mmm; omega=exp(stderr*stderr); btlsmean=exp(estimate)*sqrt(omega); btvar=exp(2*estimate)*omega*(omega); btse_mean=sqrt(btvar); run; proc print; run;*/

*/Stepwise Regression of Candidate Predictor Variables; proc sort data=first; by trt; run;

*Plot data*; proc sgplot data=first; scatter x=cc y=parameter; run; proc sgplot data=first; scatter x=sand y=parameter; run; proc sgplot data=first; scatter x=silt y=parameter; run; proc sgplot data=first; scatter x=clay y=parameter; run; proc sgplot data=first; scatter x=om y=parameter; run; proc sgplot data=first; scatter x=ph y=parameter; run; proc sgplot data=first; scatter x=cec y=parameter; run;*/

237 proc sgplot data=first; scatter x=temp24prerange y=parameter; run; proc sgplot data=first; scatter x=temp24posrange y=parameter; run; proc sgplot data=first; scatter x=temp y=parameter; run; proc sgplot data=first; scatter x=rh y=parameter; run; proc sgplot data=first; scatter x=precip7D y=parameter; run; proc sgplot data=first; scatter x=TOD y=parameter; run;

/*PROC REG for Collinearity*/ proc reg data=first; model parameter= temp24prerange temp24posrange temp precip7D TOD / vif selection=stepwise slentry=0.25 slstay=0.05; run; quit;

/*Mixed-model Multiple Regression*/ proc glimmix data=first; class env; model parameter= temp24prerange temp precip7D TOD / solution; random env; covtest 'envvar=0' 0 . ; output out=second predicted=pred residual=resid residual(noblup)=mresid student=studentresid student(noblup)=smresid; run; proc sort data=second; by studentresid; proc print; run;

*/Homogeneity of effects; proc sgscatter; plot studentresid*(pred env temp24prerange temp precip7D TOD); 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;

238